The human breast cancer cell line, MDA-MB-231 was obtained from the Korean Cell Line Bank (Seoul, Korea). RPMI-1640 medium, fetal bovine serum (FBS), and antibiotics were purchased from Gibco (Grand Island, NY, USA). Human TNF-α, 3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), dimethyl sulfoxide (DMSO) and anti β-actin antibody were purchased from Sigma (St. Louis, MO, USA). Antibodies against MMP-2, MMP-9, TIMP-1, TIMP-2, PAI-1, phospho-JNK1/2, phospho-ERK1/2, and phospho-p38 were obtained from Cell Signaling Technology (Beverly, MA, USA). Antibodies against uPA-1, JNK1/2, ERK, and p38 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Cultrex basement membrane extract (BME) cell invasion assay kit and gelatin precast gels were from Trevigen Inc. (Gathersburg, MD, USA) and Bio-Rad (Hercules, CA, USA), respectively.

The ethanolic extract of N. sublateritium was prepared as described (28). The extract was sequentially fractionated using the organic solvents, hexane, dichloromethane, n-butanol and ethyl acetate. The hexane fraction was concentrated, passed through a 0.2-μm filter and dried using a freeze-drier. Dried HFNS was reconstituted in vehicle for cell culture studies. The HPLC chromatogram of HFNS confirmed the non-polar characteristics of the extract at late retention times (27)

Monolayer cultures of MDA-MB-231 cells were maintained in RPMI-1640 medium supplemented with 10% (v/v) FBS and antibiotics. Cells were maintained in an atmosphere of 95% air and 5% CO₂ at 37°C. The effect of HFNS on cell proliferation and viability was determined by the MTT colorimetric method and propidium iodide (PI) staining, respectively (27). Cells were treated with DMSO (control) or HFNS for 24 or 48 h. Results are expressed as the ratio of the number of live cells with HFNS treatment relative to that observed after DMSO-treatment.

Cells were plated in a 6-well plate and allowed to form a confluent monolayer for 24 h. Cells were then serum starved for 24 h and the monolayer in each cell was scratched with a pipette tip, and washed with serum-free medium to remove floating cells. Cells were then treated with TNF-α (20 ng/ml) in the presence of various concentrations of HFNS for 18 h. The cells were allowed to migrate across the scratch, and photographed at three randomly selected sites per well through an inverted microscope (x40 magnification).

The invasion assay was performed using Cultrex^® 96-well BME Cell invasion assay kit. Briefly, duplicate transwell chambers with 8-μm pore polycarbonate filters were coated with 50 μl of ice-cold 0.8X BME in coat buffer and incubated overnight at 37°C. To monitor cell migration, 1×10⁵ cells were seeded on to BME-coated filter, which was then inserted into the upper chamber, containing serum-free media. The lower chamber was filled with 500 μl of medium containing TNF-α, as well as various concentrations of HFNS. The control well contained media only. After incubation for 24 h, the cells on the underside of the filter were quantified using Calcein-AM according to the assay kit manual. Cells that had migrated to the bottom of the membrane were visualized and counted using an inverted fluorescence microscope.

The activities of MMP-2 and MMP-9 medium released from MDA-MB-231 cells into the medium were measured using a gelatin zymography protease assay. Cells were serum starved for 18 h and then treated with TNF-α or TNF-α combined with HFNS for the indicated time or concentration for 12 h. The cell culture medium was then collected and concentrated using Microcon YM-10 filters (Millipore, Billerica, MA, USA). The concentrated sample was subjected to electrophoresis on 7% polyacrylamide gels containing gelatin, in the absence of a reducing agent.

Cells were treated with different HFNS concentrations in the presence of TNF-α for the indicated time. The cell culture medium was collected and concentrated. Cell lysates were prepared by centrifugation at 12,000 × g for 20 min, as previously described (18). Proteins were resolved on SDS-PAGE and transferred onto a PVDF membrane. Immunoblotting was performed with antibodies specific for uPA-1 PAI-1, TIMP-1, TIMP-2, MMP-2 and MMP-9. Changes in total protein level and phospho-JNK1/2, phospho-ERK1/2, and phospho-p38 levels were also determined using specific antibodies. The immunoreactive bands were visualized using enhanced chemiluminescence. The band intensity was quantified using a densitometer followed by normalization to the density of β-actin.

MDA-MB-231 cells were treated with TNF-α or TNF-α plus HFNS of the indicated concentration and for the indicated time, and then total RNA was extracted using the BCP Phase Separation Reagent (MRC Inc., Cincinnati, OH, USA). The RNA was converted to cDNA using a reverse transcription-PCR kit containing oligo(dT) primers (Intron Biotech., Seoul, Korea) according to the manufacturer’s protocol (28). The primers and PCR conditions are listed in Table I. PCR products were electrophoresed on 1% agarose gels and visualized under ultraviolet light after ethidium bromide staining.

Cells were treated with the different HFNS concentrations in the presence of TNF-α. Nuclear extracts of the cells were prepared and the proteins were subjected to electrophoretic mobility shift assays (EMSA) as previously described (28). Briefly, 2 μg of nuclear extract was combined with 0.25 mg/ml poly(dI)-poly(dC)-non-specific competitor in Gel Shift binding buffer (20% glycerol, 5 mM MgCl₂, 2.5 mM EDTA, 2.5 mM DTT, 250 mM NaCl, and 50 mM Tris-HCl); to this was added IRDye 700-labeled AP-1 or NFκB oligonucleotide (LI-COR Inc., Lincoln, NE, USA). After incubation at room temperature for 30 min, the protein-DNA complexes were separated from the free probe on a pre-run 8% polyacrylamide gel. The signal was then detected and quantified using the Odyssey Infrared Imaging System (LI-COR Inc.).

Differences in the measured variables between the control and HFNS-treated groups were determined using a one-way analysis of variance (ANOVA) followed by Dunnett’s or Bonferroni’s test for multiple comparisons. P-values of <0.05 were considered significant.

In a previous study, we observed that HFNS significantly inhibited growth of various human cancer cell lines (27). The HFNS concentration required for 50% inhibition of viability and proliferation of MDA-MB-231 cells was 200 μg/ml for a 24 h treatment. To verify the effects of HFNS, MDA-MB-231 cells were exposed to different concentrations of HFNS and were then stimulated with TNF-α for 24 or 48 h. As shown in Fig. 1, the viability of MDA-MB-231 cells was not affected by up to 100 μg/ml HFNS treatment, indicating that these were non-cytotoxic concentrations of HFNS in the presence of TNF-α. A significant decrease in cell viability was observed after a 24 h exposure of 200 μg/ml HFNS: cell viability was reduced by ~39.7% relative to that of cells treated with TNF-α only (Fig. 1A). The MTT assay indicated that proliferation of MDA-MB-231 cells was markedly inhibited by HFNS treatment, even at 50 μg/ml for 24 h, confirming the anti-proliferating activity of HFNS on MDA-MB-231 cells (Fig. 1B). These results suggested that the change in cell viability observed at 100 μg/ml HFNS can be attributed mainly to the inhibition of proliferation. Moreover, HFNS could regulate the metastatic properties of MDA-MB-231 cells without causing significant cell death.

Next, we investigated the effects of HFNS on TNF-α-stimulated MDA-MB-231 cells, considering the highly invasive nature of TNBC. We investigated whether HFNS could inhibit the migration and invasive potential of TNF-α-stimulated MDA-MB-231 cells, using a wound-healing repair assay and an in vitro transwell assay, respectively. After cell monolayers were wounded in the repair assay, TNF-α-treated cells migrated to the cleared area, but treatment with HFNS dose-dependently inhibited TNF-α-stimulated migration of MDA-MB-231 cells (Fig. 2A). Moreover, the invasive activity of MDA-MB-231 cells was significantly regulated by HFNS treatment in a dose-dependent manner. As shown in Fig. 2B, in the transwell chamber assay, HFNS induced a decrease in the fluorescence of MDA-MB-231 invasive cells in the lower chamber, reached through a Matrigel-coated membrane. More specifically, the fluorescent intensity was reduced by 51.4% relative to that seen with TNF-α-treated control cells following a 24 h exposure to 50 μg/ml of HFNS. These results indicated that HFNS has an inhibitory effect on the migration and invasiveness of TNF-α-stimulated MDA-MB-231 cells.

Because MMPs have emerged as critical regulators of metastasis, by their role in degrading the basement membrane, we investigated the effect of HFNS treatment on the activity of secreted MMP-9 and MMP-2 in TNF-α-stimulated MDA-MB-231 cells. The effect of HFNS on the gelatinase activity of MMP-2 and MMP-9 was analyzed using zymogram gels containing gelatin, the preferred substrate. The activity of MMPs present in the cell culture supernatant was identified by digestion of the substrate at a molecular weight corresponding to that attribute to specific MMPs (MMP-9: 92 kDa), which is seen as a clear band.

As can be seen in Fig. 3, in conditioned medium from cells treated with only TNF-α, the intensity of the MMP-9 band was decreased in a dose- and time-dependent manner. Treatment of cells with 100 μg/ml HFNS resulted in a rapid decrease in the intensity of this band (~34.2% decrease, compared to the levels of the TNF-α-treated control cells) within 2 h, indicating inhibition of MMP-9 activity. In contrast, no detectable amount of MMP-2 activity was seen with either TNF-α or HFNS treatment (Fig. 3B). This suggests that the inhibitory activity of HFNS on the migration and invasion of MDA-MB-231 cells may be due mainly to inhibition of MMP-9 activity.

Because HFNS treatment resulted in inhibition of MMP-9 activity, we also evaluated the secretion of MMP-9, MMP-2, uPA-1, and the MMP inhibitors, viz., TIMP-1, TIMP-2, and PAI-1 in MDA-MB-231 cells. In immunoblot analysis, the levels of MMP-9 and uPA-1 were significantly decreased by HFNS treatment, in accordance with the decreased MMP-9 activity. There were no noticeable changes in MMP-2 levels in TNF-α- or HFNS-treated MDA-MB-231 cells (data not shown).

The level of MMPs and uPA-1 are inversely related to the levels of TIMPs and PAIs, respectively. Interestingly, the treatment with HFNS induced a modest increase in levels of TIMP-2 and PAI-1 in a dose-dependent manner, although the change in the level of TIMP-2 was not statistically significant. However, upregulated TIMP2 and PAI-1 at transcript level was also observed by semi-quantitative RT-PCR analysis after treatment of cells with HFNS (Fig. 4B).

Previous reports have demonstrated that natural product-derived agents could regulate metastasis through suppression of MAPKs (13,29,30). To gain insights into the mechanism underlying the anti-metastatic effects of HFNS in MDA-MB-231 cells, we investigated the MAPK signaling pathways, and determined the effects of HFNS on JNK, p38-MAPK, and ERK1/2 activation. In response to TNF-α stimulation, JNK and p38 proteins were phosphorylated within 30 min and the levels reached ~2.7- and 5.1-fold those in the control cells, respectively (Fig. 5A). The activation of p38 MAPK was prolonged for >2 h after exposure to TNF-α-simulated MDA-MB-231 cells (Fig. 5A), although JNK activation declined gradually. However, cells exposed to 100 μg/ml HFNS exhibited a rapid and complete inactivation of JNK and p38, which was not induced by changes in their total protein level (Fig. 5A). In contrast, phosphorylation of ERK was not affected by HFNS treatment or by the ERK inhibitor PD98059 (Fig. 5). These results indicated that HFNS-mediated supression of MAPKs was selective for the regulation of metastatic MDA-MB-231 cells.

Furthermore, we confirmed the functional significance of MAPK downregulation using pharmacologic inhibitors of JNK1/2 (SP600125), p38 MAPK (SB202190), and ERK1/2 (PD98059). The activation of MAPKs in response to TNF-α was attenuated by pretreatment of cells with SP600125 or SB202190, but not PD98059, indicating inactivation of JNK1/2 or p38, but not ERK1/2, respectively. The pharmacological inactivation of MAPK also caused downregulated expression and enzymatic inactivation of MMP-9 as shown in Fig. 5B. The results were comparable to those of HFNS-treated MDA-MB-231 cells. Collectively, these results pointed toward an important regulatory role for JNK1/2 and p38 MAPK anti-metastatic effect of HFNS on MDA-MB-231 human breast cancer cells.

The expression of MMP-encoding genes is regulated by the transcription factors AP-1 or NFκB (3,11,21,29), which are upregulated by TNF-α in a variety of cancer cell lines, including breast cancer cells (28,31,32). To confirm whether the repressive effect of HFNS on TNF-α-induced MMP-9 or uPA-1 expression is mediated via AP-1 or NFκB motifs, TNF-α-stimulated MDA-MB-231 cells were dose-dependently treated with HFNS for 18 h. The DNA-binding activity of these transcription factors was then determined using nuclear extracts and EMSA, TNF-α treatment caused an increase in AP-1 and NFκB binding activity (Fig. 6A). The activity of NFκB was markedly inhibited at treatment of cells with 100 μg/ml HFNS; more specifically, the initial TNF-α-induced activation was followed by 25.4% inhibition (Fig. 6B). The binding activity of AP-1 was sustained at first and then declined to 27.1% of that in the TNF-α-treated control after treatment with 200 μg/ml HFNS (Fig. 6B). Based on these results, it is plausible that HFNS treatment inhibits the binding of transcription factors to their DNA response element in response to TNF-α signaling, leading to downregulation of MMP-9 expression.