Human GBM cell line known as U87 were purchased from Sigma-Aldrich (St. Louis, MO, USA). A human primary GBM cell line previously known as U87 was used to derive cell lines. Cells were cultivated in a high glucose (4.5 g/L) Dulbecco's Modified Eagle Medium that had 10% heat inactivated fetal calf serum, 100 mg/mL streptomycin, and 100 U/mL penicillin. Cells were grown at 37℃ in an atmosphere that was humidified and composed of 95% air and 5% CO₂ [16].

The 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) assay kit (CellTiter 96 Aqueous One Solution, Promega, Modison, WI, USA) was used to measure cell viability. Immediately before the experiments, the curcumin was stored at −20℃ and then dissolved in dimethyl sulfoxide and diluted to the target concentration. The U87 cells were seeded in a 96-well plate at a density of 5,000 cells/well and were then allowed to adhere overnight. Subsequently, the U87 cells were incubated in fetal bovine serum (FBS)-free media that contained different concentrations of curcumin (0, 10, 20, 30, 40, and 50 µM/L). Then, an MTS reagent was added to each well; and the 96-well plates were incubated at 37℃ over various durations (6, 24, 48, and 72 hours) in a humidified incubator maintained with 5% CO₂. The optical density was measured to be 490 nm with a microplate spectrophotometer.

Western blot analysis was used to quantify fascin expression. U87 cells were incubated for 6 hours in FBS-free media that contained 0 and 10 µM/L curcumin, and then were washed with phosphate-buffered saline (PBS). Cells were collected after 6, 12, and 24 hours, and then were lysed with a lysis buffer solution [25 mM Tris-Cl pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), and protease inhibitor cocktail, Roche, Mannheim, Germany]. Cell lysate was centrifuged at 13,000 revolutions per minute for 30 minutes at 4℃, and then the supernatant were separated and the proteins were quantified with a bicinchoninic acid protein assay (Sigma, St. Louis, MO, USA). Protein samples were separated with a 10% SDS-polyacrylamide electrophoresis gel and then were moved to membranes (Millipore, Bedford, MA, USA). The membranes were blocked for 1 hour at room temperature with a 5% non-fat dry milk in a Tris-buffered saline that contained 0.1% Tween-20. After incubation at 4℃ overnight with a 1:1,000 dilution of mouse monoclonal anti-fascin (Santa Cruz Biotechnology, Santa Cruz, CA, USA), followed by a peroxidase-conjugated secondary antibody at room temperature for 1 hour, signals were then detected using an enhanced chemiluminescence system (Amersham Biosciences, Buckinghamshire, UK). An ImageJ computer-assisted image analyzer was used to calculated band densities (version 1.146r, National Institutes of Health, Bethesda, MD, USA).

To determine the effect of curcumin on the upstream pathway of fascin expression, sandwich-ELISA was used to analyze the STAT3, a transcription factor connected to a master regulator of human gliomas central to maintaining a tumor's initiating capacity and ability to invade brain parenchyma [18]. After U87 cells were incubated in FBS-free media containing 10 µM/L curcumin for 6 hours and cleaned with PBS, the cells were seeded on a 100 mm plate at a density of 8×10⁵ cells/well and were permitted to adhere overnight. The cells were collected after 6, 12, and 24 hours and lysed with an ice-cold lysis buffer solution (150 mM NaCl, 25 mM Tris-Cl pH 7.6, 1% NP-40, 0.1% SDS, 1% sodium deoxycholate, and protease inhibitor cocktail, Roche, Mannheim, Germany). The concentrations of total or phosphorylated STAT3 in the samples were determined using sandwich-ELISA kits [Human STAT3 ELISA Kit; LifeSpan Biosciences, Inc., Seattle, WA, USA and PathScan Phospho-Stat3 (Tyr705) ELISA kit; Cell Signaling Technology Inc., Danvers, MA, USA]. After coating of the microplate wells, 100 µL of the lysates were added to a microplate well, and then they were incubated for 2 hours at 37℃ before the well was cleaned, and then a detection antibody was added for 1 hour. Subsequently, 100 µL of a secondary polyclonal antibody conjugated to horseradish peroxidase was added into each well. The plate was then incubated at room temperature for 30 minutes. Later, a 200 µL of substrate solution (tetramethylbenzidine) was incubated in each well for 10 minutes at 37℃. The reaction was then terminated through the addition of 100 µL of stop solution (2N sulfuric acid). The color in the wells changed from blue to yellow. A microplate reader was used to measure the optical density of each well (VersaMax, Molecular Device Inc., Sunnyvale, CA, USA) set to 450 nm.

Ninety-six well plates were incubated with laminin at 10 µg/mL overnight at 4℃. Unbound sites were restricted with 0.1% bovine serum albumin (BSA) for 1 hour. U87 cells, which were temporarily treated with 10 µM/L and 20 µM/L curcumin for 6 hours, were seeded at a density of 5×10⁴ cells/well on BSA-coated or laminin plates. The cells were allowed to adhere in a humidified incubator supplied with 5% CO₂ at 37℃ for 6, 12, and 24 hours. After cleaning with PBS, the cells that remained were fixed for 10 minutes with 4% paraformaldehyde, and then were stained with 5% crystal violet for 20 minutes at room temperature. Cells were solubilized with 1% SDS; then a microplate reader (VersaMax) set to 595 nm was used to measure the absorbance of each well.

U87 cells were first plated in six well dishes at a density of 3×10⁵ cells/well, and then were allowed to attach and reach a subconfluency of 80%. Thereafter, the cells were incubated using a starvation medium that contained 10 µM/L and 20 µM/L curcumin for 6 hours. A yellow pipet tip was used to perform a scratch through a cell monolayer in FBS-free media. Cells were photographed using an Axioplan-2 epifluorescence microscope (Carl Zeiss Vision GmbH, Munchen, Germany) after 6, 12, and 24 hours. A computer monitor was used to view the images using a Zeiss Plan-Apochromat 40× objective (Carl Zeiss Vision GmbH, Munchen, Germany). An ImageJ computer-assisted image analyzer was used to count the area of the scratch.

A Matrigel coated modified Boyden chamber (Biocoat™ Matrigel™ Invasion Chamber; Becton Dickinson GmbH, Heidelberg, Germany) was used to evaluate cell invasion. Approximately 2.5×10⁴ control and U87 cells treated with 10 µM/L and 20 µM/L curcumin for 6 hours were seeded into an upper well of the chamber that contained a 500 µL serum free culture medium. A 500 µL culture medium with 10% FBS was then introduced to the bottom of the well. After 6, 12, and 24 hours, non-invading cells in the top chamber were cautiously removed with a cotton swab and cells on the bottom of the chamber were fixed with 4% paraformaldehyde for 15 minutes at room temperature, and then were stained at room temperature for 10 minutes with 0.5% crystal violet. Subsequent to the crystal violet being rinsed with distilled water, the cells were photographed with an Axioplan-2 epifluorescence microscope (Carl Zeiss Meditec Incorporation, Jena, Germany). The images were displayed on a computer monitor that used a Zeiss Plan-Apochromat 100× objective (Carl Zeiss).

To determine the morphological changes of GBM cells by curcumin, immunofluorescence was performed. After temporary exposure of 0 and 10 µM/L curcumin for 6 hours, U87 cells were cleaned with PBS three times. After 6 and 24 hours, the cells were fixed for 10 minutes with 4% paraformaldehyde and then were permeabilized for 5 minutes at room temperature with 0.5% Triton X-100 in PBS. Nonspecific binding was blocked for 1 hour at room temperature by 1% BSA in PBS. Subsequently, the cells were incubated for 1 hour with an anti-Fascin monoclonal antibody. After cleaning, the cells were then incubated at room temperature for 30 minutes with Alexa Fluor 488 (Molecular Probes, Eugene, OR, USA)-conjugated goat anti-mouse IgG secondary antibodies. A computer monitor using a Zeiss Plan-Apochromat 40× objective was used to view the images.

All statistical comparisons were computed using SPSS 21.0 (IBM Corp., Armonk, NY, USA). Data were expressed as mean±standard error of the mean. Repeated measure ANOVA was used to compare groups. Null hypotheses of no difference were rejected if p-values were less than 0.05.

In terms of concentrations, U87 cells were treated with different concentrations of curcumin (0, 10, 20, 30, 40, and 50 µM/L) during same time (24 hours). MTS assays showed that cell viabilities decreased at all concentrations. However, cell viabilities at 10 µM/L and 20 µM/L curcumin showed minimal decreases compared to control group (10 µM/L: 80.8% and 20 µM/L: 78.2%). At concentrations higher than 20 µM/L, cell viabilities steeply declined (Fig. 1A).

In terms of exposure times, U87 cells were treated with 10 µM/L and 20 µM/L curcumin over various times (6, 24, 48 and 72 hours). Although cell viabilities decreased over all times, the cell viabilities at 10 µM/L and 20 µM/L curcumin during 6 hours revealed minimal decreases compared to control group (10 µM/L: 93.9% and 20 µM/L: 90.5%). After 6 hours, cell viabilities revealed consistent decreases below 80% (Fig. 1B).

Consequently, according to the cell viability assays, 10 µM/L and 20 µM/L curcumin and 6 hours were chosen as the minimally lethal concentrations and exposure time to minimize cell death to an acceptable level.

When GBM cells were temporarily treated with 10 µM/L curcumin during 6 hours as the lowest minimally lethal concentration and exposure time, fascin expression temporarily decreased after 12 hours (18.4%, p=0.024). Subsequently, it gradually recovered (Fig. 2A). Under the same circumstances, the change of phosphorylated STAT3 level also reflected temporarily decreased patterns of fascin expression at 12 hours (19.7%, p=0.010). Whereas concentrations of total STAT3 according to a time sequence were similar (Fig. 2B).

Therefore, curcumin is likely to suppress fascin expression, even with a minimally lethal concentration and short exposure time. And, curcumin may suppress fascin expression in GBM cells via phosphorylated STAT3 downregulation.

U87 cells were temporarily treated with 10 µM/L and 20 µM/L curcumin for 6 hours, and then the curcumin was washed out. Then attachment, migration and invasion assays were analyzed according to a time sequence, such as 6, 12, and 24 hours.

The attachment ability of GBM cells showed a statistically significant decrease after 6 hours (20 µM/L curcumin at 12 hours: 14.7%, p=0.001; 10 µM/L and 20 µM/L curcumin at 24 hours: 12.6% and 24.4%, p=0.001 and p<0.001) (Fig. 3). The migration ability of GBM cells showed a statistically significant decline after 6 hours, and the closure percentage of GBM cells showed statistical differences when compared to control (10 µM/L and 20 µM/L at 12 hours: 24.2% and 58.1%, p=0.031 and p<0.001; at 10 µM/L and 20 µM/L at 24 hours: 25.0% and 54.3%, p=0.002 and p<0.001) (Fig. 4). Additionally, the invasion ability of GBM cells showed a statistically significant decline after 6 hours (10 µM/L and 20 µM/L at 12 hours: 28.8% and 43.2%, p=0.046 and p<0.001; 10 µM/L and 20 µM/L at 24 hours: 54.1% and 73.8%, p=0.001 and p<0.001) (Fig. 5).

U87 cells were temporarily treated with 10 µM/L curcumin during 6 hours as the lowest minimally lethal concentration and exposure time, and then curcumin was washed out. Morphological changes of GBM cells were determined by immunofluorescence at 6 and 24 hours (Fig. 6).

Immunofluorescence showed change of cell shape from typical spindle form of uninfluenced cells to a more polygonal appearance, and a significant reduction in filopodia formation of GBM cells. And, even though filopodia remained, the length was shorter when compared to controls. Furthermore, there was no recovery of the morphological change in GBM cells according to a time sequence.