Zinc induces epithelial to mesenchymal transition in human lung cancer H460 cells via superoxide anion-dependent mechanism

Human lung cancer epithelial H460 cell was obtained from the American Type Culture Collection (ATCC, Manassas, VA). H460 cell was cultured in RPMI 1640 medium in a 5 % CO₂ environment at 37 °C. The media was supplemented with 2 mM l-glutamine, 10 % fetal bovine serum and 100 units/ml of penicillin/streptomycin (Gibco, Gaithersburg, MA, USA). Zinc sulfate, dimethyl sulfoxide (DMSO), 2,7-dichlorofluorescein diacetate (DCFH₂-DA), dihydroethidium (DHE), hydroxyphenyl fluorescein (HPF), DMNQ (2,3-dimethoxy-1,4-naphthoquinone),3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and Hoechst 33342 were obtained from Sigma Chemical, Inc. (St. Louis, MO, USA). Mn (ΙΙΙ) tetrakis (4-benzoic acid) porphyrin chloride (MnTBAP) was obtained from Calbiochem (San Diego, CA, USA). Antibodies for N-cadherin, E-cadherin, vimentin, snail, slug, phosphorylated FAK (Y397), FAK, and β-actin and peroxidase-labeled secondary antibodies were obtained from Cell Signaling Technology, Inc. (Denver, MA). Mouse monoclonal antibodies for active Rho-GTP and Rac1-GTP were obtained from NewEast Biosciences (Malvern, PA, USA). Immobilon Western chemiluminescent HRP substrate was obtained from Millipore, Corp (Billerica, MA, USA) and Thermo Fisher Scientific Inc. (Rockfort, IL, USA).

Cell viability was determined by MTT colorimetric assay. Briefly, cells in 96-well plate were incubated with 500 μg/ml of MTT for 4 h at 37 °C. The supernatant was then removed and dimethylsulfoxide (DMSO) was added to dissolve the formazan product. The intensity was spectrophotometrically measured at 570 nm using an ELISA reader (Anthros, Durham, NC, USA). All analyses were performed in at least three independent replicate cultures. The optical density ratio of treated to non-treated control cells was calculated and presented in terms of relative cell viability.

Apoptotic cell death was detected by Hoechst 33342 staining. After specific treatments, cells were stained with 10 µM of the Hoechst 33342 for 30 min at 37 °C. The apoptotic cells having condensed chromatin and/or fragmented nuclei stained by Hoechst 33342 were visualized and scored under a fluorescence microscope (Olympus IX51 with DP70).

Cell morphology was investigated by seeding the cells at a density of 5 × 10⁴ cells/well onto a 12-well plate for 48 h. The cells were treated with various concentrations of zinc sulfate for 24 h. The cells were then washed with PBS, fixed with 4 % paraformaldehyde in PBS for 10 min at 37 °C, rinsed three times with PBS, and mounted with 50 % glycerol. Cell morphology was then assessed by a phase contrast microscope (Eclipse Ti-U, Nikon, Tokyo, Japan).

Cells were seeded at a density of 1 × 10⁵ cells/well onto coverslips in six-well plate and incubated overnight. After the treatment, the cells on coverslips were fixed with 4 % paraformaldehyde for 30 min and permeabilized with 0.1 % Triton-X for 20 min. Thereafter, the cells were incubated with 3 % bovine serum albumin (BSA) for 30 min to prevent nonspecific binding. The cells were washed and incubated with rabbit anti-Vimentin antibody for 1 h at room temperature. Primary antibody was removed and the cells were washed and subsequently incubated with Alexa Fluor 488 (Invitrogen) conjugated goat anti-rabbit IgG (H + L) secondary antibody for 1 h at room temperature. Samples were washed with PBS then visualized and imaged by fluorescence microscope (Olympus IX 51 with DP70, Olympus America Inc., Center valley, PA).

After specific treatments, cells were incubated in lysis buffer containing 20 mM Tris–HCl (pH 7.5), 1 % Triton X-100, 150 mM sodium chloride, 10 % glycerol, 1 mM sodium orthovanadate, 50 mM sodium fluoride, 100 mM phenylmethylsulfonyl fluoride, and a protease inhibitor cocktail (Roche Molecular Biochemicals) for 90 min on ice. The cell lysates were collected, and the protein content was determined using the BCA protein assay kit (Thermo scientific, IL, USA). Equal amounts of proteins from each sample (60 μg) were denatured by heating at 95 °C for 5 min with Laemmli loading buffer and subsequently loaded onto a 10 % SDS-PAGE. After separation, proteins were transferred onto 0.45 μM nitrocellulose membranes (Bio-Rad, Hercules, CA). The transferred membranes were blocked for 1 h in 5 % nonfat dry milk in TBST (25 mM Tris–HCl pH 7.5, 125 mM NaCl, and 0.05 % Tween 20) and incubated with the appropriate primary antibodies at 4 °C overnight. Then, the membranes were washed twice with TBST for 10 min and incubated with horseradish peroxidase-labeled isotype-specific secondary antibodies for 2 h at room temperature. The immune complexes were detected by enhancement with chemiluminescence substrate (Supersignal West Pico; Pierce, Rockfore, IL) and quantified the level of proteins using imageJ software.

Migration was determined by wound healing and transwell assays. For the wound healing assay, a monolayer of cells was cultured in a 96-well plate, and a wound space was made with a 1-mm-wide tip. After rinsing with PBS, the cell monolayers were incubated with the indicated treatments and allowed to migrate for 24 h. Micrographs were taken under a phase contrast microscope (Olympus DP70, Melville, NY), and the wound spaces were measured using Olympus DP controller software. Quantitative analysis of cell migration was performed using an average wound space from those random fields of view, and the percentage of change in the wound space was calculated using the following formula: % change = (average space at time 0 h) − (average space at time 24 h)/(average space at time 0 h) × 100. Relative cell migration was calculated by dividing the percentage change in the wound space of treated cells by that of the control cells in each experiment. For the transwell assay, the cells were seeded at a density of 5 × 10⁴ cells/well onto the upper chamber of a transwell (8 µm pore size) in a 24-well plate in serum-free medium and incubated with various concentrations of zinc. RPMI medium containing 10 % FBS was added to the lower chamber. Following the incubation, the non-migrated cells in the upper chamber were removed by cotton-swab wiping, and the cells that migrated to the underside of the membrane were stained with 10 µg/ml of Hoechst 33342 for 10 min and visualized and scored under a fluorescence microscope (Olympus IX51 with DP70).

An invasion assay was performed using a 24-well transwell unit with polycarbonate (PVDF) filters (8 µm pore size). The membrane was coated with 0.5 % matrigel on the upper surface of the chamber overnight at 37 °C in a humidified incubator. The cells were plated at a density of 2 × 10⁴ cells per well into the upper chamber of the transwell unit in serum-free medium. Medium containing 10 % FBS was added to the lower chamber of the unit. After incubation with specific test agents for 24 h at 37 °C, the medium in the upper chamber was aspirated, and the cells on the upper side of the membrane were removed with a cotton swab. The cells that invaded to the underside of the membrane were stained with 10 µg/ml of Hoechst 33342 for 10 min, visualized and scored under a fluorescence microscope (Olympus IX51 with DP70).

In vitro 3D tumorigenesis was performed in a matrigel-coated 96-well plate. A plate was coated with 0.5 % agarose and left for solidification. The cells were suspended in culture medium containing 4 % matrigel and various concentrations of zinc, and plated at a density of 3 × 10² cells/well onto a agarose-coated plate. Medium containing various concentrations of zinc were replaced every 3 days. After 10 days, the cells were visualized and scored by image analyzer under microscope (Olympus IX51 with DP70). Whole area of each well was captured in one picture and the colonies with more than 25 µm of diameter were quantified.

Intracellular ROS were determined by fluorescence microplate reader and by flow cytometry using the ROS-specific probe, superoxide anions, hydrogen peroxide and hydroxyl radicals were determined by DHE, DCFH₂-DA and HPF, respectively. For fluorescence microplate reader, cells were seeded overnight in 96-well plate. Before zinc treatment cells were incubated with 10 μM of dihydroethidium (DHE), dichlorofluorescein diacetate (DCFH₂-DA) or hydroxyphenyl fluorescein (HPF) for 30 min at 4 °C, after which they were washed and treated with various concentrations of zinc (0–50 µM) for 1 and 3 h. After incubation, the fluorescence intensity was immediately analyzed by fluorescence microplate reader (SpectraMax M5, Molecular Devices Corp., Sunnyvale, CA, USA) using a 488-nm excitation beam and a 610-nm band-pass filter for DHE, using a 480-nm excitation beam and a 530-nm band-pass filter for detecting DCF fluorescence or using a 490-nm excitation beam and a 515-nm band-pass filter for HPF. For flow cytometry, cells were seeded overnight in six-well plate. Before zinc treatment cells were incubated with 10 μM of DHE, DCFH₂-DA or HPF for 30 min at 4 °C, after which they were washed and treated with 50 µM of zinc for 1 and 3 h. After incubation, cells were washed, re-suspended in phosphate-buffered saline (PBS), and immediately analyzed for fluorescence intensity by FACScan flow cytometer (Beckton Dickinson, Rutheford, NJ) using a 488-nm excitation beam and a 610-nm band-pass filter for DHE, using a 480-nm excitation beam and a 530-nm band-pass filter for detecting DCF fluorescence or using a 490-nm excitation beam and a 515-nm band-pass filter for HPF. Mean fluorescence intensity was quantified by CellQuest software (Becton–Dickinson) analysis of the recorded histograms. Relative fluorescence was calculated as a ratio of the treated to the non-treated control fluorescence intensity.

All treatment data were normalized to non-treated controls. Data are expressed as the mean ± SD from three or more independent experiments. Multiple comparisons were examined for significant differences of multiple groups, using analysis of variance (ANOVA), followed by individual comparisons with post hoc test. Statistical significance was set at p < 0.05.

In order to investigate the effect of zinc on EMT phenotypes in human lung cancer cells, we first evaluated the non-cytotoxic concentrations of zinc. Cells were cultured in the presence or absence of zinc (0–100 µM) for 24 h, and cell viability was determined by MTT assay at 24 h. The results indicated that treatment of the cells with zinc at the concentrations ranging from 5 to 50 µM caused no significant cytotoxicity (Fig. 1a). The significant decrease in cell viability was first detected in response to 100 µM of zinc treatment with approximately 86 % of the cells remaining viable.

Proliferative effect of zinc at above concentrations was further evaluated by treating the cells with zinc for 0–72 h. Figure 1b indicates that zinc at the concentrations of 0–50 µM had no inductive effect on cell proliferation. To confirm the effect of zinc on cell toxicity, cells were similarly treated with zinc for 24 h, and apoptosis was evaluated by Hoechst 33342 staining assay. Figure 1c, d show that apoptotic cells containing condensed and/or fragmented nuclei were not detectable in response to zinc treatment at the concentrations of 5–50 µM.

The effect of zinc on EMT in H460 cells was next investigated. The alteration of cell morphology as well as hallmarks of EMT were used to monitor the effect of zinc on EMT process in lung cancer cells. Cells were treated with zinc at non-toxic concentrations for 24 h. The morphology of the cells was captured and presented in Fig. 2a The results showed that the zinc-treated cells exhibited morphology of mesenchymal-like cells with the elongated shape and loss of cell polarity. These results also suggested that the mesenchymal-like morphology is somehow dose-dependent as the more elongated cells could be found in the cells treated with high concentrations of zinc. In addition, the expression of mesenchymal marker vimentin was significantly increased in response to zinc treatment (Fig. 2b).

The switch of E-cadherin to N-cadherin and increase of EMT proteins including vimentin, slug, and snail have been shown to be important hallmarks of EMT in cancer cells [2‐5]. We next determined such cellular EMT markers in the lung cancer cells treated with zinc by western blot analysis. Obviously, treatment of the cells with zinc could reduce E-cadherin in a dose-dependent manner. Together with the fact that the significant increase of N-cadherin was found when treating the cells with 5–50 µM of zinc, these data strongly indicated that zinc could be able to mediate E-cadherin to N-cadherin switching in these cells. In addition, the upstream transcription factors of EMT namely snail and slug were determined in the zinc-treated cells. These factors were shown to bind to E-box elements in the promoter region of E-cadherin, resulting in the transcriptional repression of E-cadherin and induction of mesenchymal markers [2‐4]. Figure 2c, d indicate that zinc significantly increased the levels of slug and snail. Also, the EMT protein vimentin was found to be induced by zinc. Taken together, our results suggested that zinc could induce EMT in lung cancer cells.

One important phenotype of EMT cells is the increase in cell motility. Studies have demonstrated that EMT could enhance aggressiveness of tumor cells by increasing their ability to migrate and invade [2‐4]. To evaluate the effect of zinc on cancer cell motility, cells were left untreated or pretreated with zinc at non-toxic concentrations for 24 h and subjected to migration and invasion assays as described in “Methods” section. Wound healing migratory assay showed that zinc significantly facilitated migratory activity of the cells with the relative cell migration increased approximately 1.3- to 1.8-fold in comparison to that of non-treated control cells (Fig. 3a, b). Also, the transwell migration assay was performed to confirm the migratory effect of zinc. Figure 3c shows that zinc treatment significantly increased the number of cells passed through the membrane of well, suggesting that such element induced cell migration.

Next, we performed experiments to test the ability of the cancer cells in invading through matrigel. The transwell was pre-coated with matrigel and the zinc-treated cells were seeded on top. The cells were allowed to invade for 24 h and the invaded cells at the lower part of the membrane were determined. Figure 3d shows that zinc significantly promoted the invasion of H460 cells in a dose-dependent manner.

Enhanced tumor cell migration and invasiveness are shown to be down-stream behaviors of FAK signal [21‐24]. We next determined the effect of zinc treatment on motility regulatory proteins including FAK, activated (phosphorylated at Try397) FAK, active forms of RAC1, and RhoA by western blot analysis. The results showed that treatment of the cells with 0–50 µM of zinc for 24 h dramatically increased the activation of FAK (Fig. 3e, f). Also, its down-stream functioning proteins active Rac1 and RhoA were found to increase, accordingly. These results suggested that zinc treatment increase EMT-associated cell behaviors trough FAK-dependent pathway.

Having shown that zinc could enhance cell migration and invasion, we next tested whether zinc may augment the ability of cancer cell to initiate a new tumor. It is well known that the EMT process facilitate tumor formation at the metastatic site [25‐27] and this potential is responsible for cancer progression [25‐27]. Previous studies have shown that in vitro 3D tumorigenesis assay reflexes ability of the cancer cells in in vivo cancer condition [28]. Cells were left un-treated or pre-treated with various concentrations of zinc (0–50 µM) for 24 h and subjected to tumorigenesis assay as described in “Methods” section. After 10 days of 3D culturing, the colony number and diameter were determined. Figure 4 shows that zinc treatment enhances tumorigenic ability of the lung cancer cells as indicated by the significant increase in the number and size of colonies in zinc-treated groups. The number of colonies increased approximately 1.38- and 1.76-fold in response to zinc at 25 and 50 µM, respectively (Fig. 4b).

Studies have demonstrated the roles of ROS on cancer EMT phenotypes [15‐17]. Because zinc has been previously shown to affect the redox balance of the cells and exhibit the pro-oxidant activity in normal and cancer cells [7, 18‐20], we next evaluated whether the effect of zinc in regulation of EMT in these lung cancer cells is through ROS-dependent mechanism. To examine the specific ROS levels induced by zinc, cells were incubated with various ROS specific probes as mentioned in “Methods” section for 30 min before treatment with 0–50 µM zinc and intracellular ROS levels were determined by flow cytometry and fluorescence microplate reader. Figure 5a, d show that zinc significantly increased cellular superoxide anion, while it had minimal effect on hydrogen peroxide (Fig. 5b, e) and hydroxyl radical (Fig. 5c, f). Consistent with microplate reader determination, flow cytometric analysis revealed that intracellular superoxide anion significantly increased up to twofold in response to zinc treatment (Fig. 5g) and zinc had no detectable effect on hydrogen peroxide and hydroxyl radical (Fig. 5h, i). We further confirmed the superoxide anion-inducing effect of zinc using specific superoxide anion inhibitor MnTBAP. Results indicated that treatment with the zinc caused the superoxide anion up-regulation in the cells and such event could be abolished by the addition of MnTBAP, confirming that the major ROS induced by zinc treatment in our system was superoxide anion (Fig. 5j, k).

In order to investigate the role of superoxide anion on EMT induction, cell morphology, EMT markers, migratory behaviors, and colony formation were evaluated in the cells treated with zinc and specific inhibitor of superoxide anion. Cells were cultured in the presence or absence of MnTBAP for 1 h prior to zinc treatment and EMT phenotypes were determined. Also, to clarify role of superoxide anion, the superoxide anion inducer DMNQ was used. The results show that treatment of the cells with DMNQ or zinc alone was able to switch the morphology of lung cancer cells from epithelial to fibroblast-liked mesenchymal feature (Fig. 6a). Addition of MnTBAP to the zinc-treated cells could be able to attenuate such morphologic transformation.

Immunocytochemistry showed the increased vimentin found in DMNQ- and zinc- treated cells. And the increased vimentin signal induced by zinc was suppressed by the addition of MnTBAP (Fig. 6b). Consistently, the results of western blot analysis revealed that EMT markers including N-cadherin, vimentin, snail, and slug were found to be significantly increased in response to DMNQ and zinc treatment and such phenomenon could be reversed by the addition of MnTBAP (Fig. 6c, d).

As DMNQ was used as a superoxide anion donor, the intracellular superoxide anion generated by DMNQ was determined. Cells were incubated with DHE as mentioned in “Methods” section for 30 min and treated with 5 µM of DMNQ. DHE intensity was determined by flow cytometry. The result indicated the 1.5-fold superoxide induction in DMNQ-treated cells as shown in Fig. 6e.

The migration and invasion of the cells were further evaluated. The results were consistent with the expression levels of EMT proteins that DMNQ and zinc could be able to induce cell migration and invasion and such inductions could be abolished by superoxide anion inhibitor (Fig. 7a–d). Also, the colony size and number which were increased in response to zinc treatment were found to be significantly attenuated by the treatment of MnTBAP (Fig. 7e, f). Taken together, these results pointed out that zinc induces EMT process, migratory behaviors, and tumorigenic potential in the lung cancer cells via superoxide anion-dependent mechanism.