This article is mentioned in:

Introduction

Hypoxia frequently occurs in a wide range of solid tumors. Hypoxia is responsible for the promotion of tumor motility, invasion and metastasis, genomic instability and cell survival responses, through the regulation of gene expression downstream of hypoxia-inducible factor 1 (HIF1) (1–3). HIF1 is a heterodimeric transcription factor composed of an ‘α’ and a ‘β’ subunit. HIF1α is a cytoplasmic protein responsive to O2 levels, whereas HIF1β is a nuclear protein, which is constitutively expressed independent of O2 levels. Under hypoxic conditions, HIF1α is translocated to the nucleus where it forms a heterodimer with HIF1β subunits, resulting in the activation of HIF1 and its complex downstream genetic program, which may be responsible for numerous hypoxia-associated cellular changes (4). Hypoxia has been identified to promote metastasis of tumor cells by induction of epithelial-mesenchymal transition (EMT) (5,6). EMT is a process by which epithelial cells lose their cell polarity and cell-cell adhesion, and gain migratory and invasive properties, thus becoming mesenchymal stem cells. Therefore, EMT is an essential process that results in tumor metastasis. At the molecular level, EMT is characterized by decreases in the expression levels of epithelial markers, such as E-cadherin, and an increase in mesenchymal markers, such as vimentin and snail family transcriptional repressor 1 (Snail1), as well as an increase in migratory and invasive behavior (7–9). Under hypoxic conditions, expression of the EMT marker proteins has previously been observed to be mutually dependent on the expression and activation of HIF1α (8,9).

Previous studies have demonstrated that hypoxia stimulates the production of reactive oxygen species (ROS) in tumor cells and hypoxia-generated intracellular ROS may activate and stabilize HIF1α, thus contributing to tumor growth and metastasis, in addition to tumor recurrence and therapeutic resistance (10–13). In this regard, ROS-induced HIF1α activation may be a major target for future cancer therapeutic approaches.

The present study examined the effects of hypoxia-induced intracellular ROS on the EMT of HT29 human colorectal cancer cells in the presence or absence of the antioxidant dieckol. The hypoxia-inducing agent, CoCl2 was used to induce an increase in ROS levels and HIF1α expression in HT29 cells, and led to the transition of cellular phenotypes, corresponding to the fundamental events in EMT. These effects were markedly attenuated following dieckol treatment. Therefore, the present study demonstrated that ROS-induced activation of HIF1α was a key factor contributing to the induction of EMT in HT29 cells.

Materials and methods

Cell culture and reagents

HT29 human colorectal cancer cells, obtained from the American Type Culture Collection (Manassas, VA, USA), were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA), supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.), 100 U/ml penicillin and 100 mg/ml streptomycin, in a 5% CO2 atmosphere at 37°C. For the cell-culture experiments, cells were passaged three times and detached using Trypsin-EDTA. Transwell migration and Matrigel invasion chambers were purchased from BD Biosciences (Franklin Lakes, NJ, USA). Primary antibodies against HIF1α (cat. no. 610958), E-cadherin (cat. no. 7870), vimentin (cat. no. 5741), Snail1 (cat. no. AV33314) and β-actin (cat. no. MAB1501) were obtained from BD Biosciences, Santa Cruz Biotechnology, Inc. (Dallas, TX, USA), Cell Signaling Technology, Inc. (Danvers, MA, USA), and Sigma-Aldrich; Merck Millipore (Darmstadt, Germany), respectively. Dihydroethidium (DHE) was obtained from Molecular Probes (Thermo Fisher Scientific, Inc.). Dieckol was prepared as previously described by Lee et al (14). All remaining chemicals and reagents were purchased form Sigma-Aldrich; Merck Millipore, unless stated otherwise.

Cell viability assay

The cytotoxicity of CoCl2 and/or dieckol was determined using the 3-(4, 5-dimetylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) formazan assay. HT29 cells were seeded in 12-well plates at a density of 2.0×105 cells/well in DMEM containing 10% FBS. The medium was changed 36 h after seeding to serum-free low glucose (1,000 mg/l) DMEM and the cells were incubated with 0–200 µM CoCl2 for 36 h. The cells were then incubated with or without 0–100 mg/ml (0–216 µM) dieckol for 12 h. Subsequently, the medium was carefully removed and 160 µl MTT (0.5 mg/ml final concentration) solution was added to each well prior to incubation for an additional 4 h at 37°C in 5% CO2. The medium was aspirated without the formazan crystals and 1 ml dimethyl sulfoxide was added to each well. The absorbance was then measured using a microplate reader (iMark; Bio-Rad Laboratories, Inc., Hercules, CA, USA) at 540 nm.

Cell migration and invasion assay

Cell migration was determined using a Transwell chamber. HT29 cells were treated with or without 50 µM CoCl2 in serum-free low glucose (1,000 mg/l) DMEM for 24 h. The cells were trypsinized and collected. Subsequently, 300 µl cell suspension (1.5×105 cells) was incubated at 37°C in serum-starved DMEM containing 50 µM CoCl2 and/or 25 mg/ml dieckol and added to the upper compartment of the Transwell chamber in a humidified atmosphere with 5% CO2 for 24 h. DMEM containing 10% FBS was used as a chemoattractant in the lower compartment. Following a 24 h incubation at 37°C, cells on the upper surface of the membrane were removed and the cells that had migrated below the surface of the membrane were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS), stained with crystal violet (1 g dissolved in 10% ethanol) for 30 min, washed twice with PBS and counted under a phase contrast microscope using a 10x objective lens. The number of migrated cells in five randomly selected fields were counted. The migration assays were performed in triplicate.

For the invasion assay, cells were seeded in the Transwell invasion chambers, which were coated with Matrigel (1 mg/ml). DMEM supplemented with 10% FBS was added to the lower compartment of the chamber. Cells were serum-starved and treated with 50 µM CoCl2 and/or 25 mg/ml dieckol as aforementioned, then trypsinized and collected. Subsequently, 300 µl of each cell suspension was added to the upper compartment of the chamber and incubated at 37°C in a humidified atmosphere with 5% CO2 for 24 h. Cells that had invaded below the surface of the membrane were fixed, stained, washed, and counted. The number of cells in five randomly selected fields was counted. The invasion assays were performed in triplicate.

ROS quantification

HT29 cells (2.0×105 cells/well) were serum-starved for 24 h and were then incubated in the presence or absence of 50 µM CoCl2 for 36 h, with the medium being replaced twice, followed by treatment with or without 25 mg/ml dieckol for 3 h. Cellular ROS levels were determined using DHE staining and the Muse oxidative stress kit (Merck Millipore) according to the manufacturer's protocol.

Western blotting

Following the aforementioned treatments, cells were washed twice with PBS, harvested and solubilized in 2X sodium dodecyl sulfate (SDS) protein sample buffer containing 100 mM Tris-HCl (pH 6.8), 200 mM DTT, 4 SDS, 0.4% bromophenol blue, and 20% glycerol. Protein was quantified using the Bradford Protein assay kit II (Bio-Rad Laboratories, Inc.). Equal quantities of protein (25 µg) were separated by 10% SDS-polyacrylamide gel electrophoresis. The resolved proteins were then transferred to polyvinyldifluoride membranes (Merck Millipore). The membranes were blocked by incubation with 1% bovine serum albumin in TBS-Tween-20 (TBST; 10 mM Tris-HCl, 150 mM NaCl pH 7.5 containing 0.1% Tween-20) at room temperature for 1 h and were then incubated with primary antibodies against HIF1α (1:500), E-cadherin (1:200), vimentin (1:500), Snail1 (1:200) and β-actin (1:1,000) for 1 h at room temperature. The membranes were washed three times with TBST and were then incubated for 30 min with goat anti-rabbit or goat anti-mouse secondary antibody conjugated to alkaline phosphatase (1:1,000; Santa Cruz Biotechnology, Inc.; cat. nos. sc-2007 and sc-2008). The respective proteins were detected by colorimetric reactions using 5-bromo-4-chloro-3′-indolyphosphate/nitro-blue tetrazolium substrate.

Cell morphology and immunofluorescence assay

Cells were collected at different time points (2–48 h) for analysis of cell morphological changes and at 48 h for immunofluorescence and were fixed in 3.7% formaldehyde for 5 min. Fixed cells were permeabilized in 0.1% Triton X-100 and were incubated with antibodies specific to E-cadherin (1:200), vimentin (1:100), and Snail1 (1:100) for 1 h at room temperature. The cells were then washed and incubated with secondary antibodies for 30 min at room temperature. Immunostaining was detected using the goat anti-mouse and goat anti-rabbit secondary antibodies labeled with Alexa Fluor 488 (1:500; Molecular Probes; Thermo Fisher Scientific, Inc.; cat. nos. A11029 and A11034). Actin was stained using rhodamine phalloidin (Thermo Fisher Scientific, Inc.). Cells were examined using the Zeiss LSM 510 confocal imaging system (Carl Zeiss, Oberkochen, Germany).

Statistical analysis

Data were analyzed using InStat statistics software version 6.0 (GraphPad Software, Inc., La Jolla, CA, USA). Statistical comparisons were performed using one-way analysis of variance followed by Bonferroni post-hoc test for multiple comparisons. P<0.05 was considered to indicate a statistically significant difference.

Results

Dieckol attenuates CoCl2-induced intracellular ROS levels in HT29 cells

Previous studies have suggested that hypoxia may stimulate the generation of ROS in tumor cells, and increased ROS promotes tumor progression; however, details of the underlying molecular mechanism remain to be elucidated (10–13). The investigation of ROS levels in tumor cells and antioxidants as potential dietary supplements for the prevention of tumor progression and potential cancer treatment has grown rapidly. The present study used HT29 human colorectal cancer cells to assess the association between hypoxia-induced intracellular ROS generation and tumor progression. CoCl2 was used as a reagent to mimic the hypoxic microenvironment of tumor cells (8,15). CoCl2 and dieckol did not affect the viability of HT29 cells at concentrations below 100 µM and 25 mg/ml (54 µM), respectively (Fig. 1). As presented in Fig. 2, treatment of HT29 cells with 50 µM CoCl2 resulted in significantly increased intracellular ROS levels to ~140% compared with the control (P<0.05; Fig. 2B), suggesting that CoCl2 successfully mimicked the hypoxic environment of tumor cells. In addition, in order to determine whether intracellular ROS levels were associated with hypoxia-induced tumor progression, HT29 cells were treated with 25 mg/ml dieckol in the absence and presence of 50 µM CoCl2 treatment. Dieckol treatment significantly attenuated the CoCl2-induced increased cellular ROS levels (P<0.05; Fig. 2B). These findings suggest that HT29 cells may be responsive to CoCl2-induced hypoxia and the antioxidant dieckol effectively reduced the CoCl2-induced intracellular ROS levels.

Figure 1. Effects of CoCl2 and DK on viability of HT29 cells. (A) HT29 cells were treated with various CoCl2 or DK concentrations for 48 h and were then assessed using an MTT assay. (B) HT29 cells were treated with 50 µM CoCl2 and/or 25 mg/ml (54 µM) DK for 48 h. All data are presented as the mean ± standard deviation from three independent experiments. DK, dieckol.

Figure 2. DK attenuated CoCl2-induced ROS levels in HT29 cells. (A) HT29 cells (2×105 cells) were serum starved and then incubated in the presence or absence of 50 µM CoCl2 for 36 h with medium replaced twice, followed by treatment with or without 25 mg/ml DK for 3 h. Cellular ROS levels were determined by dihydroethidium staining and flow cytometry. (B) Quantification of ROS levels in the different treatment groups. All data are presented as the mean ± standard deviation from three independent experiments. *P<0.05 vs. control group; †P<0.05 vs. CoCl2 group. DK, dieckol; ROS, reactive oxygen species.

Dieckol inhibits CoCl2-induced EMT-associated phenotypic changes of HT29 cells

The effects of hypoxia-induced intracellular ROS on the EMT of HT29 cells were also examined. EMT is characterized by epithelial cells acquiring a mesenchymal appearance, such as the loss of cell-cell contact, scattering from cell clusters, acquisition of mesenchymal properties and the downregulation of epithelium-specific proteins, such as E-cadherin, while simultaneously expressing mesenchymal proteins such as vimentin and Snail1. As presented in Fig. 3A, HT29 cells displayed typical epithelial characteristics with well-developed cell-cell contact in the control group, without CoCl2 treatment, whereas HT29 cells subjected to 50 µM CoCl2 had reduced cell-cell contact and formed scattered clusters. To examine whether the morphological changes in CoCl2-treated HT29 cells were consistent with EMT, immunofluorescence microscopy was performed, examining E-cadherin, an epithelial cell marker, and vimentin and Snail1, mesenchymal cell markers. As presented in Fig. 3B-D, E-cadherin signaling was primarily localized at the cell-cell contact area in the control group (without CoCl2 treatment). Following treatment with 50 µM CoCl2 E-cadherin levels were decreased and diffused throughout the cytoplasm, whereas nuclear vimentin signaling was dispersed to the cytoplasm and membrane fraction. Nuclear translocation of the transcription factor Snail1 was clearly observed (Fig. 3D), suggesting that the morphological changes observed in CoCl2-treated HT29 cells may be identified as the onset of EMT. Cells were treated with 50 µM CoCl2 in order to determine whether CoCl2-induced intracellular ROS levels were associated with EMT and subsequently incubated in the presence or absence of 25 mg/ml dieckol. Dieckol treatment inhibited the CoCl2-induced mesenchymal morphologic changes and the redistribution of E-cadherin, vimentin and Snail1, indicating that the CoCl2-induced increase in ROS levels may be associated with the occurrence of EMT in HT29 cells.

Figure 3. DK inhibited CoCl2-induced EMT-associated phenotypic changes of HT29 cells. (A) HT29 cells (2.0×105 cells/well) were serum starved and incubated in the presence or absence of 50 µM CoCl2 for 36 h, followed by treatment with or without 25 mg/ml DK for 12 h. Morphological changes of cells were detected by phase contrast microscopy (scale bar, 10 µm). Representative images of all cells were collected after 3 h treatment with 25 mg/ml DK. CoCl2 treatment induced changes in intracellular localization of typical EMT marker proteins. Intracellular localization of (B) E-cadherin, (C) vimentin, and (D) Snail1 was determined using antibodies specific to each protein and Alexa Fluor 488 (green) conjugated secondary antibodies. Actin was immunostained using rhodamine phalloidin (red). Cells were visualized by laser confocal scanning microscopy. Representative images from three independent experiments have been presented. Scale bar, 10 µm. DK, dieckol; EMT, epithelial-mesenchymal transition.

Dieckol decreases CoCl2-induced migration and invasion of HT29 cells

EMT promotes the migration and invasion of cancer cell (5–7). In order to investigate the effects of CoCl2-induced hypoxia on migration and invasion of HT29 cells, the migration of HT29 cells was examined using Transwell chambers. DMEM with 10% FBS was used as a chemoattractant in the lower chamber. As presented in Fig. 4A, treatment with 50 µM CoCl2 significantly increased migration of HT29 cells by ~1.7-fold compared with the control (P<0.05), suggesting that CoCl2-induced hypoxia may be associated with increased HT29 cell migration. However, when HT29 cells were treated with 25 mg/ml dieckol following incubation with CoCl2, the CoCl2-induced increased migration of cells was significantly inhibited (P<0.05; Fig. 4A). Invasion of HT29 cells on reconstituted 3D Matrigel matrix significantly increased by ~1.4-fold in the presence of CoCl2 compared with the control group (P<0.05; Fig. 4B) and subsequent treatment with dieckol significantly inhibited the CoCl2-induced invasive motility of cells (P<0.05; Fig. 4B). These findings suggest that the reduction of intracellular ROS levels following dieckol treatment may decrease the hypoxia-induced migration and invasion of HT29 cells.

Figure 4. DK decreased CoCl2-induced migration and invasion of HT29 cells. (A) Inhibitory effects of DK on the CoCl2-induced migration of HT29 cells. Cell migration assays were performed with HT29 cells incubated in the presence or absence of 50 µM CoCl2 and/or 25 mg/ml DK. Cells that migrated through the pores to the lower surface of the membrane were fixed, stained and counted. Five random microscopic fields were counted in each treatment group. *P<0.05 vs. control; †P<0.05 vs. CoCl2. (B) Invasion assay was performed with HT29 cells incubated with 50 µM CoCl2 and/or 25 mg/ml DK. All data are presented as the mean ± standard deviation from three independent experiments. *P<0.05 compared with control; †P<0.05 compared with CoCl2. DK, dieckol.

Dieckol suppresses the HIF1α signaling pathway, required for CoCl2-induced EMT of HT29 cells

Subsequently, the CoCl2-induced morphological changes and increased cell motility of HT29 cells were verified by quantification of the changes in the intracellular expression levels of EMT marker proteins. As aforementioned, EMT is initially characterized by the downregulation of the epithelial marker, E-cadherin, and by an upregulation of mesenchymal markers, such as vimentin. In particular, under hypoxia-induced EMT, selective transcription factors HIF1α and Snail1 are major factors in regulating the expression of EMT marker proteins (8). Previous studies silenced HIF1α by RNA interference (RNAi), which resulted in an increase in E-cadherin, a decrease in vimentin and reduced cell invasion (16,17). The HIF1α-dependent regulation of E-cadherin expression is induced by upregulation of Snail1 (18,19). Silencing of HIF1α reduced nuclear localization of Snail1, which resulted in increased E-cadherin expression. In accordance with these reports, when HT29 cells were cultured in 50 µM CoCl2, HIF1α expression levels increased (Fig. 5). Expression levels of vimentin and Snail1 were also increased, whereas the expression of E-cadherin was decreased. Conversely, dieckol treatment reversed the CoCl2-induced changes in expression levels of all EMT marker proteins. These findings suggest that ROS levels and HIF1α signaling are important for CoCl2-induced EMT in HT29 cells.

Figure 5. DK suppressed the HIF1α signaling pathway required for CoCl2-induced EMT in HT29 cells. HT29 cells were incubated in the presence or absence of 50 µM CoCl2 and/or 25 mg/ml DK. Protein expression levels of HIF1α, E-cadherin, vimentin and snail1 in the different treatment groups were analyzed using western blotting. A representative image from three independent experiments has been presented. DK, dieckol; HIF1α, hypoxia-induced factor 1α; EMT, epithelial-mesenchymal transition

Discussion

Metastasis is required for tumor progression to occur. Recent developments in early diagnosis and surgical techniques have revealed that >90% of mortality in cancer patients is due to the metastatic spread of primary tumors (20). It is of note that in patients with colorectal cancer, ~50% of patients with an early diagnosis have already developed metastasis. Colorectal cancer may metastasize to the liver, lung, peritoneum and various other distant organs. The metastasis rates and mortality of colorectal cancer have markedly increased, resulting in it becoming the third most common type of cancer worldwide (21,22). Conventional chemotherapy increases the survival rate of patients with colorectal cancer; however, there is a wide range of acute and long-term side effects that may substantially affect the quality of life of patients. Non-toxic chemotherapeutic agents derived from natural products may be used for future treatment of colorectal cancer.

Dieckol is a natural and non-toxic polyphenol compound and its antioxidant properties have previously been examined (23,24). A previous study revealed the anti-inflammatory and anticancer activities of dieckol are based on the induction of apoptosis and inhibition of aberrant proliferation by oxidative stress (14). The present study demonstrated that dieckol inhibited the hypoxia-induced metastatic properties of tumor cells by the downregulation of intracellular ROS levels and reduction of HIF1α signaling. CoCl2-induced hypoxia increased ROS generation and promoted EMT in HT29 cells, subsequent treatment with dieckol inhibited hypoxia-induced EMT.

Previous studies have reported that hypoxia-induced EMT occurs downstream of HIF1α (25,26) and hypoxia-induced intracellular ROS generation activates various signaling components upstream of HIF1α such as phosphatidylinositol 3-kinase/AKT serine/threonine kinase 1 (27,28) and mitogen-activated protein kinase (29–31). Expression of the dominant negative HIF1α mutant, von Hippel-Lindau tumor suppressor, or its silencing has been identified to increase E-cadherin expression (16), whereas E-cadherin RNAi increased the migration and invasion of tumor cells in hypoxic conditions (18), indicating that a HIF1α-dependent downregulation of E-cadherin may be a critical factor for the occurrence of EMT during hypoxia. It has also been previously reported that Snail1, another strong transcriptional repressor of E-cadherin, may be regulated by interaction with HIF1α (18). Snail1 has been determined to be overexpressed in various types of human cancer (31). The upregulation of Snail1 may subsequently result in decreased E-cadherin levels and ultimately EMT (18,32). HIF1α silencing has also been identified to affect nuclear localization and activation of Snail1 and tumor cell invasion (19). Therefore, it is possible that hypoxia-induced EMT-associated events may be due to the activation of HIF1α and Snail in conjunction with the reduction of E-cadherin. The findings of the present study clearly supported this hypothesis as the association between the expression of HIF1α and Snail1, E-cadherin and vimentin in hypoxia-induced invasive tumor cell migration was evident.

In conclusion, the present study provided evidence that hypoxia-induced EMT of HT29 human colorectal cancer cells may be dependent on cellular ROS levels and the HIF1α signaling pathway. In addition, the results suggested that the natural, non-toxic antioxidant, dieckol, may be considered a novel therapeutic agent for the treatment of metastatic colorectal cancer.

Acknowledgements

The present study was supported by the Basic Science Research Program through the National Research Foundation of Korea, funded by the Ministry of Education, Science and Technology (grant nos. 2012R1A1A2041500 and 2015R1D1A3A01018506).

References