Alpha-mangostin inhibits the migration and invasion of A549 lung cancer cells

Materials

Human lung adenocarcinoma cell line A549 cells and normal human bronchus diploid cell line CCD-14Br cells were purchased from Japanese Collection of Research Bioresources (JCRB) cell bank (Osaka, Japan). 3,3′-Dioctadecyloxacarbocyanine perchlorate (DiO), 1,1′-Dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI), and antibiotics were purchased from Sigma-Aldrich (St. Louis, MO). Cell harvesting solution TrypLE express and fetal bovine serum (FBS) were purchased from Life Technologies Japan Ltd. (Tokyo, Japan). α-Mangostin was purchased from Wako Pure Chemical Industries Ltd. (Osaka, Japan). Cell counting kit-8 was purchased from Dojindo Molecular Technologies, Inc. (Kumamoto, Japan). The cone probe (BL-AC-40TS-C2; spring constant: around 0.05 N/m) was purchased from Olympus (Tokyo, Japan). Other reagents were purchased from Sigma-Aldrich, Wako Pure Chemical Industries Ltd., or Life Technologies Japan Ltd.

Cell culture

The cells were cultured in DMEM containing 10% FBS and antibiotics (100 units/mL penicillin G and 100 µg/mL streptomycin sulfate) in humidified atmosphere of 95% air and 5% CO₂ at 37°C

Determination of cell viability

The viability of cells after treatment with various concentrations of α-mangostin was evaluated by the cell counting kit-8 as recommended by the manufacturer. Briefly, cells were seeded on a 96-well-plate at 10⁴ cells/well (24 h culture experiments) or 1.5 ×10³ cells/well (time course experiments) with 100 µL medium and cultured for 24 h, so as to allow the cells to adhere to the plate. The culture medium was replaced by 100 µL of fresh culture medium diluted with various concentrations of α-mangostin for 24–96 h treatment. The medium was replaced by adding 100 µL fresh medium diluted with 10 µL of cell counting kit-8 solution to each well. The cells were cultured for suitable time periods for each cell type. The plate absorbance was then measured at 450 nm using a microplate reader. Ratio of cell viability V_c was calculated as: ${\displaystyle V_c=\frac{Ab{s}_{target}-Ab{s}_{background}}{Ab{s}_{ctrl}-Ab{s}_{background}}}$ where V_c is the cell viability ratio, Abs_target is the absorbance of α-mangostin-treated cells, Abs_ctrl is the absorbance of control cells, Abs_background is the absorbance of the background.

Measurement of mechanical properties of cells

The cultured cells treated with α-mangostin for 24 h were manipulated by atomic force microscopy (AFM) (Nanowizard III; JPK Instruments AG, Berlin, Germany) at room temperature. Combining optical microscopy (IX-71; Olympus) and AFM allows the probe to be placed on a particular region of the cell surface. In this study, the AFM probe was indented at the top of the cell surface with a loading force of up to 0.5 nN and velocity of 5 µm/s. The Young’s modulus of the cell was calculated using the Hertz model (Hertz, 1881). The force-distance curve for a region up to about 1 µm of cell surface indentation was fitted using JPK data processing software (JPK Instruments AG) as: ${\displaystyle F=\frac{E}{1-{\nu }^2}\frac{2\tan \mathrm{\alpha }}{\pi }{\delta }^2,}$ Where F = force, δ = depth of the probe indentation, ν = Poisson’s ratio (0.5), α = half-angle of the cone probe (9°), and E = Young’s modulus. More than 25 cells were used per experiment, and 25 points were examined on the surface of each cell. The logarithmic Young’s modulus values for each group were compared by nonparametric analyses of variance followed by Kruskal-Wallis H test and Steel pairwise comparison test. Young’s modulus of the polystyrene tissue culture surface was more than 1 ×10⁷ Pa (Haghparast, Kihara & Miyake, 2015). The range of Young’s moduli of cell surface was in the order of about 10² to 10⁴ Pa. Thus, we were convinced that the surface stiffness of the cells could be measured by this method without affecting the rigidity of culture surface.

Wound healing assay

A549 cells were plated on 35-mm culture dishes at a density of 2 ×10⁵ cells using a regular cell culture medium and cultured for 24 h. After the cells achieved confluence, a wound was created by scratching through the middle of the dish with a 200-µL tip (Yuan, Wu & Lu, 2013). Cells were gently rinsed twice with the culture medium to remove any floating cell debris. The medium alone was added to the control dish; the medium diluted with α-mangostin at final concentrations of 5 and 10 µM was added to the treatment dish. The first image acquisition (t = 0 h) was then done by using phase contrast microscopy. Cells were then cultured for the next image acquisition (t = 12 h and t = 24 h). The data were analyzed by ImageJ software (NIH, Bethesda, MD); the ratio of recovered area, A_r, which was covered by cells was calculated as ${\displaystyle A_r=\frac{A_{\mathrm{covered}}}{A_0},}$ where, A ₀ is the scratched area at t = 0 h, A_covered is the area covered cells at different incubation times (t = 12, 24 h). The covered areas were compared by nonparametric analyses of variance followed by Kruskal-Wallis H test and Steel pairwise comparison test.

Invasion assay

In order to evaluate the invasion ability of cancer cells, a co-culture system was established by culturing both cancer and normal cell lines together. Briefly, cells were trypsinized, spun, and resuspended with fresh medium. Then, the suspended cells were fluorescently labeled with DiI for A549 cells or DiO for CCD-14Br cells for their membrane at 37°C for 1 h in the dark. Cells were then spun at 200× g for 5 min; the medium was then removed, and cells were resuspended with fresh medium and spun one more time. The labeled cells were cultured in 12-well-plates at a density of 2 ×10⁴ and 4 ×10⁴ cells/well with CCD-14Br and A549 cells, respectively. For monoculture, A549 or CCD-14Br cells were prepared separately at the same density. After incubation for 24 h, the cells were exposed to α-mangostin at 0, 5, and 10 µM concentrations. The cells were observed and image acquisition was done by using fluorescence microscopy at the first time point (t = day 0). Then, the cells were cultured for 1, 2, 3, 4, 5, and 6 days and image acquisition was done for each day.

The fluorescence images were analyzed by ImageJ software. After binarization of each fluorescent color image, the cell area estimated by DiI or DiO fluorescence was calculated. Finally, we evaluated the cell area as cell coverage ratio, A_c: ${\displaystyle A_c=\frac{\text{Fluorescence detectable pixels}}{\text{Total image pixels}}.}$ The inverse values of cell doubling time, $t_d^{-1}$ (d ⁻¹), were calculated from proliferation slopes in early stage.

Cytotoxicity of α-mangostin on A549 and CCD-14Br cells

The cytotoxicity of α-mangostin on A549 cells was evident by a dose- and time-dependent inhibition of cell viability and growth (Shih et al., 2010). We first evaluated the cytotoxicity of α-mangostin by treatment of both the non-small cell lung cancer A549 and pulmonary normal diploid CCD-14Br cell lines with various concentrations of α-mangostin for 24 h. The cell viability was evaluated by the reduction of formazan dye produced from WST-8 in the presence of an electron mediator by the activities of dehydrogenases in cells. α-Mangostin exhibited cytotoxic effects on A549 and CCD-14Br cells at higher concentrations (Fig. 1A). The half-maximal effective concentration (EC₅₀) values of α-mangostin for the cytotoxicity of A549 and CCD-14Br cells were 19 µM and 22 µM, respectively (Fig. 1). Furthermore, we examined the cytotoxic effects of α-mangostin on these cells in time course experiments (24–96 h culture). The EC₅₀ values of α-mangostin for the cytotoxic effects for A549 and CCD-14Br cells decreased in a time-dependent manner (Figs. 1B and 1C). The EC₅₀ values of α-mangostin for the cytotoxic effects for A549 cells almost plateaued at 48 h of culturing and was about 10 µM (Fig. 1B). On the other hand, the EC₅₀ values of α-mangostin for the cytotoxic effects for CCD-14Br cells gradually decreased (Fig. 1C). Thus, CCD-14Br cells were less sensitive to the cytotoxic effects of α-mangostin than the A549 cells. Compared with the untreated control, A549 and CCD-14Br cells treated with α-mangostin at concentrations below 5 µM were not significantly different, proving that these dosages are non-toxic to these cells. However, those treated with more than 10 µM of α-mangostin displayed some cytotoxicity (Fig. 1). Therefore, we used less than 10 µM of α-mangostin for subsequent experiments.

α-Mangostin decreases cell surface stiffness

We then examined the effect of α-mangostin on the mechanical properties of cells using AFM. Alterations in cell activities or states often entail a change in the mechanical properties of cells (Haghparast et al., 2013; Shimizu et al., 2012), and the mechanical alterations are largely attributable to the actin cytoskeleton (Dai & Sheetz, 1995; Sugitate et al., 2009). Thus, analyzing the alteration in mechanical properties of cells can reveal the changes in characteristics of their underlying actin networks as well as their states (Haghparast, Kihara & Miyake, 2015; Haghparast et al., 2013; Kihara et al., 2011). AFM indentation is a sensitive method for analyzing the surface mechanical properties of cells (Haghparast, Kihara & Miyake, 2015; Haghparast et al., 2013). Fig. 2 shows the distribution of the Young’s modulus of the cells treated with α-mangostin. The distribution of Young’s modulus of normal CCD-14Br cells (fibroblast-like morphology) has a logarithmic average value of 8.9 kPa, which was clearly higher than that of cancerous A549 cells with logarithmic average value of 3.6 kPa (Fig. 2). The difference in mechanical properties of normal and cancer cells is based on the difference in F-actin structures at the apical surface of these cell types (Cross et al., 2007; Haghparast et al., 2013; Lekka et al., 2012).

Addition of 5 µM or 10 µM of α-mangostin to the cells reduced the distribution of the Young’s moduli in A549 as well as CCD-14Br cells in a dose-dependent manner (Fig. 2). Particularly, the Young’s modulus of A549 cells significantly decreased (Fig. 2). Thus, even though its concentration does not affect cell viability in a short-time period, α-mangostin clearly reduced the surface rigidity of cells. Furthermore, A549 cells, whose surface stiffness was originally soft, were more sensitive to the effect of α-mangostin than the CCD-14Br cells.

α-Mangostin inhibit cell migration

The effect of α-mangostin on lung cancer A549 cell motility was measured by the wound healing assay (Liu et al., 2013), an established method for studying directional cell migration in vitro. The migration of several types of cancer cells was reportedly inhibited after treatment with α-mangostin (Shih et al., 2010; Wang, Sanderson & Zhang, 2012; Yuan, Wu & Lu, 2013). In the untreated control group, cells exhibited marked cell migration in the wounded area, whereas the wounds treated with α-mangostin showed a delayed healing (Fig. 3). The ratio of recovered area of wound closure in the untreated cells was about 0.47 at 12 h and almost 0.98 after 24 h (Fig. 3B). On the other hand, α-mangostin, at 5 and 10 µM, reduced the ratio of recovered area of wound closure to approximately 0.36 after 12 h, and 0.87 after 24 h (Fig. 3B). This result indicated that α-mangostin inhibited the migration of A549 cells in vitro.

α-Mangostin suppresses the invasion of cancer cells

The co-culturing system of A549 and CCD-14Br cells was used as a model for evaluating cancer cell invasiveness, which imitate the initial invasive progress of cancer cells. Then we examined the potential effects of α-mangostin on cancer cell invasion (Fig. 4A). Besides, the monoculture of each cell type was also conducted to compare their results with those of the co-culture (Fig. S1).

When only cancerous A549 cells were cultured, the cells grew and covered almost all of the plate surface after six days under α-mangostin concentration of less than 5 µM; however, the cells treated with 10 µM of α-mangostin showed a gradual decrease in the area covered by cells (Fig. S1, A549 and Fig. 4B, A549 mono-culture). Conversely, when CCD-14Br cells were cultured under α-mangostin condition, the cells gradually proliferated (Fig. S1, CCD-14Br and Fig. 4B, CCD-14Br mono-culture). These results suggest that A549 cells were relatively sensitive to high concentration of α-mangostin and showed gradual cell death at 10 µM of α-mangostin after three days culture. These results are roughly corresponding to our cytotoxic experiments (Figs. 1B and 1C).

After A549 and CCD-14Br cells were co-cultured, the A549 cells gradually grew and almost covered the whole area at six days (Figs. 4A, ctrl and 4B, Co-culture of ctrl condition). Conversely, the covered area of CCD-14Br cells increased within several days, and after coming in contact with A549 cells, the area of CCD-14Br cells gradually decreased (Figs. 4A, ctrl and 4B, Co-culture of ctrl condition). A549 cancer cells killed and eroded the normal CCD-14Br cells and resulted in a decline in the area of the CCD-14Br cells. In contrast, in the presence of α-mangostin, A549 cells could not cover the whole surface area of the culture plate, and the coverage area of the CCD-14Br cells did not decrease (Figs. 4A, 5 µM of α-mangostin and 10 µM of α-mangostin and 4B, Co-culture of 5 and 10 µM of α-mangostin conditions). Thus, invasive activities of A549 cells on CCD-14Br cells were inhibited by treatment with α-mangostin. Furthermore, in the presence of 10 µM of α-mangostin, A549 cells proliferated gradually but did not show any cell death as seen with the monoculture (Fig. 4B, 10 µM of α-mangostin). Thus, normal CCD-14Br cells probably rescue A549 cells from the cytotoxic effects of highly concentrated α- mangostin. These results demonstrate that by using a co-culturing system, we can conspicuously observe the invasive activities of cancer cells acting on normal cells and α-mangostin exhibited its potential effect in repressing cancer cell invasion.