Tissue factor/FVIIa activates Bcl-2 and prevents doxorubicin-induced apoptosis in neuroblastoma cells

Two neuroblastoma cell lines SH-EP1 and SK-N-SH were used. The SH-EP1 is a substrate-adherent (S) type clone of SK-N-SH [30]. The cell lines were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 50 units/ml penicillin, and 50 μg/ml streptomycin, at 37°C and in a humidified atmosphere containing 5% carbon dioxide.

The full-length TF ORF containing a hemagglutinin (HA) tag at the 5' end was inserted into the pcDNA3.1 vector (Invitrogen, Carlsbad, CA) between the BamH1 and Xba1 sites to create pcDNA3.1/TF. Other constructs that express truncated TF tagged at the N-terminal end with HA, i.e. pcDNA3.1/TF (extro) containing extracellular domain and transmembrane domain and pcDNA3.1/TF (intro) containing cytoplasmic domain and transmembrane domain were derived from pcDNA3.1/TF by replacing the full-length TF ORF with corresponding PCR products that lack the nucleotides for either the last 40 or the first 256 amino acids of TF, respectively.

For the transient gene transfection assay, SK-N-SH cells were plated at a density of 1 × 10⁶/well in a 6-well plate one day before transfection. Transfection was performed with Lipofectamine 2000™ (Invitrogen, CA), according to instructions by the manufacturer. A total of 4 μg of plasmid DNA was used for transfection in each well, and the total DNA supplied was kept constant in all cases by adjusting with empty vector. The nearly confluent cells were harvested 48 h after transfection, and used in experiments analyzing gene expression and sensitivity to treatment with doxorubicin.

Both TF siRNA and STAT5 siRNA were used to inhibit endogenous TF and STAT5 expression, respectively. As previously reported, the sequence of TF siRNA was 5'-GCGCTTCAGGCACTACAAA-3' [31]; this was synthesized for us by Dharmacon (Chicago, IL). The STAT5 siRNA was purchased from Invitrogen. A mismatched siRNA was used as a control. Transfection of siRNA into cells was performed using siPORT™NeoFX™ Reagent (Ambion, Austin TX). Briefly, cells were trypsinized and diluted in RPMI 1640 medium supplemented with 10% FBS to a density of 1 × 10⁵ per ml. The siRNA solution and siPORT™NeoFX™ Reagent were diluted in OPTI-MEMI media and mixed following the manufacturer's protocols, then the prepared cells were added to the plates containing these siRNA/siPORT™NeoFX™ complexes. Cells were incubated at 37°C until ready to assay. All treatments were performed in triplicate.

First, whole cell protein samples were prepared by lysing cells in a buffer composed of 150 mM NaCl, 50 mMTris (pH 8.0), 5 mM EDTA, 1% (vol/vol) Nonidet P40, 1 mM phenylmethylsulfonyl fluoride (PMSF), 20 ug/mL aprotinin, and 25 ug/mL leupeptin for 30 min at 4°C. After this denaturation, equal amounts of the protein extracts were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a nitrocellulose membrane. After blocking the membrane with buffer containing 20 mM Tris-HCl (pH 7.5) and 500 mM NaCl-5% nonfat milk for 1 hour at room temperature, it was subsequently incubated with primary antibodies for 3 hours at room temperature, followed by washes and treatment with horseradish peroxidase-labeled secondary antibody. These blots were developed using a chemiluminescent detection system (ECL; Amersham Life Science, Buckinghamshire, England). After stripping, the membrane was re-probed with an anti-β-actin antibody to control for equal protein loading and protein integrity.

The soft agarose method was used to measure colony formation. Briefly, a bottom layer of low-melting point agarose solution containing 0.5% agarose in a final concentration of 1 × RPMI 1640 medium supplemented with 10% FBS and the various concentrations of test reagents such as FVIIa (American Diagnostica Inc. Stamford, CT) and siRNA, was poured into gridded 35 mm dishes and allowed to gel. The top layer contained prepared (trypsinized, siRNA transfected, counted) cells, 0.35% agarose, and the 1 × medium as diluent. The cells in soft agarose were then cultured at 37°C in a humidified atmosphere containing 5% carbon dioxide. After 1 to 2 weeks, the cultures were fixed with formalin and the colonies were scored under phase microscopy (using a cutoff value of 50 viable cells).

The effects of TF transfection and TF inhibition by siRNA on cell growth and apoptosis of tumor cells after treatment with doxorubicin in the presence or absence of activated FVIIa were determined using a water-soluble tetrazolium salt (WST-1) assay and the annexin-V assay, respectively. Doxorubicin concentrations used in the experiments were those achieved with therapeutic doses, and the concentration of FVIIa used (10 uM) corresponds to the normal plasma concentration of zymogen FVII. For the WST-1 assay, cells were cultured in 96-well microtiter plates along with different concentrations of doxorubicin, for 44 hr. Then, WST (25 μg/well) was added and the cells were incubated for an additional 4 hr. The final optical density (OD) in the wells was read with a microplate reader at a test wavelength of 450 nm and a reference wavelength of 620 nm.

For the annexin-V assay, tumor cells incubated with or without doxorubicin were harvested by trypsinization, washed with PBS, and then stained with FITC-annexin-V (Oncogene, San Diego, CA) and PI for 30 min according to the manufacturer's instructions. Stained cells were detected using the FACScan (Becton Dickinson) and analyzed using WinList software (Verity Software House, Inc).

We have examined TF protein in two neuroblastoma cell lines and found that the level of TF expression was remarkably variable among the different cell lines. As shown in Fig. 1A, SH-EP1 cells expressed very high levels of TF, whereas SK-N-SH cells lacked TF expression as detected by western blot assay. SH-EP1 cells were much more resistant to doxorubicin than SK-N-SH cells. The difference in mean cell survival between the two cell lines after 48-hr of drug exposure was significantly greater for the SH-EP1 line in either presence or absence of FVIIa (Fig 1B). Consistent with these observations, the apoptosis assay showed a lower percentage of annexin-V positive SH-EP1 cells after 48 hr of drug treatment compared to SK-N-SH cells (13 vs. 53 %, p < 0.01) (Fig. 1C). Interestingly, for SH-EP1 cells but not for SK-N-SH cells, cell survival and apoptosis observed after doxorubicin were significantly different depending upon the presence or absence of FVIIa in the assays (80 vs. 64% for cell survival and 13 vs. 31% for apoptosis in Fig. 1B and 1C, respectively, p < 0.05).

To examine whether silencing of endogenous TF by TF siRNA had any effect on Bcl-2 and Bcl-XL expression and resistance of SH-EP1 cells to doxorubicin, we used a specific TF siRNA shown to effectively inhibit TF expression in a previous report [31]. SH-EP-1 cells were cultured in medium with 10 nM FVIIa for all experiments. As seen in Fig. 2A, transfection of TF siRNA into high-expressor SH-EP1 cells effectively downregulated expression of TF as well as Bcl-2 in a dose-dependent manner. Furthermore, inhibition of the endogenous TF by TF siRNA suppressed SH-EP1 cell growth, as compared with control siRNA (Fig. 2B). Also, transfection of TF siRNA sensitized SH-EP1 cells to doxorubicin-induced apoptosis, as measured both activation of caspase-3 and cleavage of the death substrate PARP. As demonstrated in Fig. 2C, significant cleavage of caspase-3 and PARP was detected in TF siRNA-transfected cells 8 hours after doxorubicin treatment, whereas cleavage of these two proteins was not observed in the control siRNA-treated cells even after 24 hours of doxorubicin treatment. Likewise, in the quantitative apoptosis assay, 24 hours post-treatment with doxorubicin and TF siRNA, there was an increased percentage of annexin-V positive SH-EP1 cells compared to that observed after doxorubicin plus control siRNA treatment (Fig. 2D).

To further explore the role of TF expression in apoptosis, expression of Bcl-2 and sensitivity to doxorubicin was evaluated in TF-transfected SK-N-SH cells in the presence of FVIIa. TF was efficiently transfected and expressed in SK-N-SH cells (Fig. 3A). FVIIa treatment of either TF-transfected SK-N-SH cells or endogenously- TF expressing SH-EP1 cells significantly induced Bcl-2 expression, whereas expression of Bax (a member of the Bcl-2 family) remained unchanged (Fig. 3B). To investigate whether induction of Bcl-2 by TF/FVIIa can affect sensitivity to doxorubicin-induced cell death, we performed a WST cytotoxicity assay in TF-transfected-SK-N-SH cells treated with FVIIa. As shown in Fig. 3C, transfection of TF alone into SK-N-SH cells did not increase resistance to doxorubicin. However, in presence of FVIIa resistance of these transfected cells to doxorubicin was significantly enhanced in an FVIIa dose-dependent manner (Fig. 3D).

Since the JAK/STAT5 and PI3k/Akt signaling pathways have been reported to be involved in regulation of cell survival mediated by the TF/FVIIa interaction, we tested the phosphorylation of STAT5 (p-STAT5) and Akt (p-Akt) in TF-transfected SK-N-SH cells treated with FVIIa. FVIIa treatment induced phosphorylation of STAT5, which was detectable 5 min post-addition of FVIIa and reached a maximum effect 20 min thereafter (Fig. 4A and 4B). Similarly, induction of Akt phosphorylation was detected only in TF-transfected but not control-transfected SK-N-SH cells. Expression of pan-STAT5 and pan-Akt was not affected by FVIIa, suggesting that the observed increases in phosphorylation of STAT5 and Akt were most likely due to active phosphorylation of existent molecules, but not a product of upregulation.

In order to investigate whether the cytoplasmic domain of TF is involved in activation of STAT5, two expression plasmids containing a cytoplasmic domain deletion [pcDNA3.1/TF (extro)] and only the cytoplasmic domain of TF [pcDNA3.1/TF (intro)] were tested. In presence of FVIIa, transfection of the pcDNA3.1/TF (extro) lacking the cytoplasmic domain into SK-N-SH cells was able to induce phosphorylation of STAT5 comparable to that in cells transfected with the plasmid containing the full-length TF. In contrast, transfection of pcDNA3.1/TF (intro) under similar stimulation of FVIIa did not induce STAT5 phosphorylation (Fig. 4C). These results indicate that the cytoplasmic tail of TF is not the part of the molecule involved in phosphorylation of STAT5.

We also performed similar experiments of TF-transfected cells to study the effect of stimulation with FVIIa in the presence of the specific JAK inhibitor 1 and PI3k inhibitor LY294002, in order to examine whether activation of STAT5 and Akt were altered, respectively. As shown in Fig. 4D, JAK inhibitor 1 did block phosphorylation of STAT5, whereas LY294002 inhibited Akt phosphorylation in TF-transfected SK-N-SH cells treated with FVIIa. These results suggest that both pathways are also involved downstream of the TF/FVIIa interaction in human cells and can be blocked by specific pathway inhibitors.

Since we found that TF/FVIIa interaction induces the JAK/STAT5 and PI3K/Akt survival signaling pathways as well as the anti-apoptotic factor Bcl-2, we evaluated whether there are associations between Bcl-2 induction and either induction of STAT5 or Akt activation. We tested for the expression of Bcl-2 in FVIIa-stimulated, TF-transfected SK-N-SH cells in the presence of either JAK inhibitor 1 or LY294002. As shown in Fig. 5A, we found that JAK inhibitor 1, but not LY294002, blocked the TF/FVIIa-induced upregulation of Bcl-2. This JAK inhibitor 1-mediated suppression of Bcl-2 was dose-dependent. Our results imply that TF/FVIIa-upregulation of Bcl-2 is mediated by the JAK/STAT5 pathway.

To further prove that the JAK/STAT5 pathway contributes to TF/FVIIa-induced upregulation of Bcl-2, we also tested for Bcl-2 expression in FVIIa-stimulated, TF-transfected SK-N-SH cells, in which the STAT5 was silenced by a prior transfection of STAT5 siRNA. As can be seen in Fig 5B, transfection of the STAT5 siRNA inhibited the expression of endogenous STAT5 expression, and consequently blocked TF/FVIIa-mediated upregulation of Bcl-2 (Fig. 5C). This appears to suggest that not only JAK, but also STAT5 is involved in the upregulation of Bcl-2 following the interaction of FVIIa with TF.