The adherent human colon cancer cell lines SW480 and SW620 were obtained from the Cell Bank of the Chinese Academy of Sciences. SW480 cells were cultured in RPMI-1640 medium (Kino Biological and Pharmaceutical Technology Co., Ltd.) supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific, Inc.) and maintained at 37°C in a humidified atmosphere with 5% CO₂. SW620 cells were grown in L-15 medium (Kino Biological and Pharmaceutical Technology Co., Ltd.) supplemented with 10% FBS in an incubator with the same culture conditions.

Lentivirus vectors for small interfering RNA (siRNA) were constructed to explore the function of AKT1 with assistance from Shanghai GeneChem Co., Ltd. Three AKT1-siRNAs were designed and used, with the following sequences: Sequence 1, gaTCCTCAAGAAGGAAGTCAT; sequence 2, gcATCGCTTCTTTGCCGGTAT; and sequence 3, ggACAAGGACGGGCACATTAA. After annealing, the double oligonucleotides obtained were inserted into the AgeI/EcoRI sites of the GV248 lentiviral vector with green fluorescent protein (GFP; Shanghai GeneChem Co., Ltd.). Moreover, one scrambled RNA (sequence, TTCTCCGAACGTGTCACGT) was used as the negative control (NC) and cloned into the GV248 lentiviral vector (Shanghai GeneChem Co., Ltd.). After identification and lentivirus packaging, SW480 and SW620 cells were infected with the acquired lentiviruses [multiplicity of infection (MOI)=10] for 16 h. After infection for 72 h, SW480 and SW620 cells were harvested, and the knockdown efficiency of AKT1 was detected by reverse transcription polymerase chain reaction (RT-PCR) and western blotting.

Treatment with various concentrations (10, 25, 50, and 100 ng/ml) of EGF (R&D Systems, Inc.) has been revealed to induce EMT in SW480 cells (22,23). Therefore, SW480 and SW620 cells seeded at a density of 1×10⁶ were treated with 50 ng/ml EGF for 48 h to establish an EMT model. Expression of E-cadherin, N-cadherin, and vimentin were then detected to confirm the success of the EGF-induced EMT model. Subsequent experiments were then conducted using this model.

The viability of SW480 and SW620 cells was estimated using a Cell Counting Kit-8 (CCK-8) (Nanjing KeyGen Biotech Co., Ltd.). Briefly, 1×10⁶ cells/well were seeded in 96-well plates and incubated until the cells reached 60% confluence. Resveratrol (Sigma-Aldrich; Merck KGaA) was then added at different concentrations (0, 7.5, 15, 30, 60, 120, and 240 µmol/l) to the cells and incubated for 48 h. Cells were further incubated with CCK-8 reagent for 2 h, followed by optical density (OD) measurement at a wavelength of 450 nm using a microplate reader (BioRad Laboratories, Inc.). In addition, inhibition of SW480 and SW620 cell proliferation at various treatment concentrations was calculated using the following formula: Proliferation inhibition rate (%)=(average OD in all control group duplicates-average OD in the treatment group)/average OD in the blank control group ×100%.

SW480 and SW620 cells were plated at densities of 2×10⁵ and 3.5×10⁵, respectively, into 6-well plates. After treatment, cells were cultured until the well bottoms were fully covered. Subsequently, a cell-free gap was created by scratching the well bottom using a 200-µl pipette tip. The debris produced after scratching was removed by rinsing with phosphate-buffered saline (PBS). Subsequently, 3 ml serum-free medium was added to the culture plates and the cells were incubated for 24 h at 37°C in an incubator with 5% CO₂. Finally, cell migration was observed in 3–5 randomly selected fields using an optical microscope (IX71; ×100, magnification; Olympus Corporation) and the migration distance over 24 h (in pixels) was evaluated using Image-Pro Plus software version 6 (Media Cybernetics, Inc.).

The migration and invasion of SW480 and SW620 cells were evaluated using Transwell assays. For cell invasion, 40 µl diluted Matrigel glue (Corning Incorporated) was evenly spread in the upper chamber (24-well) and allowed to solidify. Briefly, after the different treatments were completed, 2×10⁶ SW480 and SW620 cells were resuspended in 200 µl serum-free medium and plated in the upper chamber of each Transwell (6.5 mm insert; 8.0 µm polycarbonate membrane; Corning Incorporated). Subsequently, 600 µl complete medium (containing 10% FBS) was placed into the lower chamber. After incubation for 48 h at 37°C in a 5% CO₂ incubator, migrated or invaded cells that passed through the membranes were fixed in 4% paraformaldehyde (Wuhan Boster Biological Technology, Ltd.) for 20 min, washed once with PBS, and stained with 0.1% crystal violet (Sangon Biotech Co., Ltd.) for 30 min. Finally, images of the migrated and invaded cells were photographed and counted using a microscope (ix71; Olympus Corporation, ×200, magnification).

Following treatment, total RNA was extracted from the cells using TRIzol reagent (Shanghai Pufei Biotech Co., Ltd.). Reverse transcription into cDNA was then performed using the M-MLV Reverse Transcriptase kit (Promega Corporation). Real-time qPCR was used to detect the expression of each target gene and the internal reference glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Each reaction was performed in 12 µl reaction mixture, containing 0.6 µl template cDNA, 0.3 µl primer mix (5 µM), 6 µl SYBR Premix Ex Taq (Takara Bio, Inc.), and 5.1 µl RNase-Free H₂O. The reaction conditions of RT-qPCR were as follows: 95°C, 3 min, pre-modification; 95°C, 3 sec, annealing; 60°C, 30 sec, extension; intermediate cycle 40 times. The forward and reverse sequences of primers used for gene amplification were as follows: GAPDH (121 bp): 5′-TGACTTCAACAGCGACACCCA-3′ and 5′-CACCCTGTTGCTGTAGCCAAA-3′; AKT1 (287 bp): 5′-GTGCTGGAGGACAATGACTAC-3′ and 5′-TGCTGCCACACGATACCG-3′, respectively. The relative expression of each gene was calculated using the 2^−ΔΔCq method (24).

Following different treatments, total protein was extracted from cells or tissues using radioimmunoprecipitation (RIPA) lysis buffer (Beyotime Institute of Biotechnology). Equal amounts of protein extracts (~20 µg) were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and the protein bands were transferred to a polyvinylidene fluoride (PVDF) membrane (EMD Millipore). Following non-specific binding the membrane was blocked by incubation with Tris-buffered saline with Tween-20^® (TBST) containing 5% skimmed milk powder, and the membranes were then probed with appropriate primary antibodies at 4°C overnight and incubated again with a corresponding secondary horseradish peroxidase (HRP)-conjugated antibody for 1 h. AKT1 (cat. no. 4691S; 1:1,000), p-AKT (Ser473) (cat. no. 4060S; 1:1,000), GSK-3β (cat. no. 9315S; 1:1,000), p-GSK-3β (Ser9) (cat. no. 5558P; 1:1,000) and Snai1 (cat. no. 3879S; 1:500) were obtained from Cell Signaling Technology, Inc. N-cadherin (cat. no. 76011; 1:500) and E-cadherin (cat. no. 76319; 1:1,000) were purchased from Abcam, and β-actin (cat. no. 20536-1-AP; 1:1,000) was purchased from ProteinTech Group, Inc. Peroxidase Conjugated Goat anti-Mouse IgG (H+L) (cat. no. DW0990; 1:1,000), and Peroxidase Conjugated Goat anti-Rabbit IgG (H+L) (cat. no. DW-GAR007; 1:1,000) were purchased from Hangzhou Dawen Biology Co., Ltd. To visualize the bands, enhanced chemiluminescence solution (ECL; Thermo Fisher Scientific, Inc.) was added, and the integral OD of each protein band was analyzed using Image-Pro Plus version 6 software. β-actin was used as the internal control.

In accordance with the protocols for experimentation with animals (National Institutes of Health Publication; no. 85-23, revised 1996), animal experiments conducted were approved by the Institutional Animal Care and Use Committee of Zhejiang Chinese Medical University [Laboratory Animals Production License No. SCXK (Shanghai) 2013-0006, Shanghai Xipuer-Bikai Experimental Animal Co. Ltd.; Laboratory Animals Usage License No. SYXK (Zhejiang) 2013-0184, Animal Experimental Research Center, Zhejiang University of Traditional Chinese Medicine. In total, 32 male, 4-week-old BALB/c (nu/nu) mice weighing 15±1 g were obtained from Shanghai Xipuer-Bikai Experimental Animal Co., Ltd. The mice were fed a normal diet and water and were housed in a specific pathogen-free (SPF) barrier center under constant conditions (temperature, 25±2°C; humidity, 50±5%; 12-h light/dark cycles).

SW480 cells transfected with si-AKT1 or NC were collected, and the density was adjusted to 3×10⁷ cells/ml with normal saline. Each nude mouse was inoculated with 0.2 ml cell suspension into the tail vein to establish a lung metastasis model of colon cancer (25,26). Two weeks after inoculation, disease progression was assessed using live fluorescence imaging. Nude mice were further divided into control groups (NC-control and si-AKT1-control) and resveratrol groups (NC-resveratrol, si-AKT1-resveratrol) according to the total flux in in vivo imaging. During the experiment, the activity of the nude mice was observed, they were weighed daily, and the mice underwent intragastric administration according to their body weight. Nude mice in the resveratrol groups received 150 mg/kg resveratrol via gavage once daily for 2 weeks. Mice in the control groups received an equal volume of normal saline once daily for 2 weeks. Subsequently, after the last intragastric administration, and fasting and water prohibition for 4 h, lung metastasis was detected using live cell fluorescence imaging technique. The nude mice were then sacrificed by cervical dislocation and the tumor foci and lung tissue were collected for subsequent experiments.

For in vivo bioluminescence imaging, the mice were intraperitoneally injected with 150 mg/kg body weight D-luciferin (cat. no. 40901ES03; 15 mg/ml, Yeasen Corp.) in normal saline. In vivo imaging was performed half an hour after intraperitoneal injection of D-luciferin. Luminescence image acquisition was conducted using a cooled charge-coupled device (CCD) camera system (IVIS Imaging System; PerkinElmer, Inc.) with an integration time of 1–60 sec and a binning factor of 4. During imaging, the nude mice were anesthetized by continuously inhaling 2% isoflurane (cat. no. O21400; Huazhong Haiwei Gene Technology Co., Ltd.). The flux of all detected photon counts within a region of interest prescribed over the tumor area was collected to calculate the signal intensity using the Living Image software package (Xenogen Corporation).

The tumor tissues obtained were fixed in paraformaldehyde at 4°C for 24 h, dehydrated in ethanol, treated with xylene, embedded in paraffin, and sliced. Tissue slices of 4-µm thickness were subjected to H&E staining (cat. no. ZLI-9609; ZSGB-BIO; OriGene Technologies, Inc.) and histological changes in the tumor tissues were observed using a microscope (ix71; Olympus Corporation; ×200, magnification).

For immunohistochemistry staining, 4 µm-thick tissue slices were immersed in 1 mM ethylenediaminetetraacetic acid (EDTA) buffer (pH=9.0) for antigen retrieval. After removal of endogenous peroxidase activity with 3% H₂O₂, non-specific protein binding was blocked with 5% normal goat serum buffer at 37°C for 30 min. Samples were incubated at 4°C overnight with primary antibodies, including N-cadherin antibody (cat. no. 76011; 1:300), E-cadherin rabbit monoclonal antibody (mAb; cat. no. 76319; 1:300), AKT1 rabbit mAb (cat. no. 4691S; 1:300), p-AKT (Ser473; cat. no. 4060S; 1:300), GSK-3β polyclonal antibody (cat. no. 9315S; 1:300), p-GSK-3β (Ser9; cat. no. 5558P; 1:300), and Snai1 polyclonal antibody (cat. no. 3879S; 1:300). Next, the slides were incubated first with biotin-labeled goat-rabbit immunoglobulin G (IgG), followed by HRP-conjugated streptavidin for 1 h each. The reaction products were visualized after staining with diaminobenzidine (DAB; cat. no. ZLI-9065, ZSGB-BIO; OriGene Technologies, Inc.) with a concentration of 2 mg/ml at room temperature for 10 min and counterstaining with hematoxylin. The results were observed and images were captured using an inverted microscope (×200, magnification), followed by analysis of the total integral optical density (IOD) using Image-Pro Plus version 6 software.

The data are presented as the means ± standard deviation (SD) and the distribution of the data was assessed. All statistical analyses were computed using SPSS version 22.0 software (IBM Corp.). For data that were normally distributed, significant differences between the groups were evaluated via one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test. For data that did not conform to a normal distribution, the Kruskal Wallis H test was used. A value of P<0.05 indicated statistical significance. All experiments were repeated three times.

For AKT1 knockdown, lentivirus vectors containing si-ATK1 and NC were injected into SW480 and SW620 cells. As revealed in Fig. 1A, the transfection of any of the three si-ATK1 sequences significantly inhibited the expression of AKT1 (P<0.01 and P<0.001). The knockdown efficiency of sequence 1 was the highest and this was, therefore, selected for subsequent experiments. In addition, the expression of AKT1 mRNA in SW480 and SW620 cells transfected with si-AKT1 was significantly decreased compared with that in the two cell lines transfected with NC (P<0.001, Fig. 1B). Furthermore, western blotting confirmed consistent changes in AKT1 protein expression after knockdown (P<0.01, Fig. 1C and D). Thus, stable AKT1-knockdown cell lines were successfully established in SW480 and SW620 colon cancer cells.

In addition, EGF was used to induce EMT in both cell lines. Western blotting revealed that E-cadherin expression in both cell lines was markedly downregulated after treatment with 50 ng/ml EGF, whereas the expression of N-cadherin and vimentin was significantly upregulated (P<0.05 and P<0.01, Fig. 1E-G). These data indicated that the EMT model was successfully established by EGF in both cell lines.

As revealed in Fig. 2A, cells were treated with resveratrol at concentrations of 3.75, 7.5, 15, 30, 60, 120, and 240 µM. The half maximal inhibition concentration (IC₅₀) of resveratrol was 69.58 µM in SW480 cells and 77.24 µM in SW620 cells (Table I).

As revealed in Fig. 2B, resveratrol significantly decreased the cell survival rate of SW480 and SW620 cells doses >30 µM (P<0.001). In addition, the inhibitory effects increased in a dose-dependent manner. Accordingly, the maximum nontoxic concentration of resveratrol was defined as 15 µM and was used in subsequent experiments.

Whether the effects of resveratrol on cell migration and invasion were achieved via AKT1 regulation was then assessed. The scratch wound healing assay revealed that the 24-h migration distance of two cell lines in the NC-resveratrol group was significantly shorter than that in NC-control group (P<0.001, Fig. 3), indicating that resveratrol inhibited the healing ability of colon cancer cells. The 24-h migration distance of cells in the si-AKT1-control group was also significantly shorter than that in NC-control group (P<0.001, Fig. 3), indicating that AKT1 knockdown also inhibited the healing ability of colon cancer cells. However, no significant difference was observed between the 24-h migration distance in the si-AKT1-control and si-AKT1-resveratrol groups (P>0.05, Fig. 3), indicating that the ability of resveratrol to inhibit the healing ability of colon cancer cells was weakened or even nullified in AKT1-knockdown cells.

Transwell assays were also performed to detect the migration and invasion of both cell lines (Fig. 4). The number of migrated and invaded cells passing through the membrane in the NC-resveratrol group was significantly lower than that in the NC-control group (P<0.001), indicating that resveratrol markedly inhibited the migration and invasion of colon cancer cells. In addition, the number of migrated and invaded cells passing through the membrane in the si-AKT1-control group was also significantly lower than that in the NC-control group (P<0.05 and P<0.001), indicating that AKT1 knockdown also inhibited the migration and invasion of colon cancer cells. However, there was no significant difference between the number of migrated and invaded cells passing through the membrane in the si-AKT1-control and si-AKT1-resveratrol groups (P>0.05), indicating that the inhibitory effects of resveratrol on the migration and invasion of colon cancer cells were mitigated by AKT1 knockdown.

For further investigation of the effects of resveratrol on EMT in colon cancer cells as well as its downstream mechanisms, the expression of EMT markers and AKT/GSK-3β/Snail signaling pathway-related molecules were evaluated in SW480 and SW620 cells after treatment. As revealed by western blotting in Fig. 5A and D, the expression of E-cadherin in both cell lines was markedly upregulated in the NC-resveratrol and si-AKT1-control groups compared to the NC-control group, whereas N-cadherin, p-AKT1, p-GSK-3β, and Snail expression was markedly downregulated. Densitometric analysis revealed consistent changes in the protein expression of the aforementioned genes after resveratrol treatment or AKT1 knockdown (P<0.05, P<0.01 and P<0.001, Fig. 5B, C, E and F). However, protein expression in SW480 and SW620 cells was not significantly different in the si-AKT1-resveratrol and si-AKT1-control groups, indicating that the effects of resveratrol on the reversal of EMT in colon cancer cells was markedly weakened or even nullified by AKT1 knockdown (P>0.05). These data indicated that resveratrol reversed EMT in colon cancer cells by regulating the AKT/GSK-3β/Snail signaling pathway.

Lung metastasis in nude mice was evaluated using live fluorescence imaging. The total flux in mice in the NC-resveratrol or si-AKT1-control groups was significantly lower than that in the NC-control group (P<0.01 and P<0.001, Fig. 6A and B), indicating that resveratrol treatment or AKT1 knockdown could inhibit lung metastasis of colon cancer. Consistent with the in vitro findings, no significant differences in total flux in the si-AKT1-control and si-AKT1-resveratrol groups were observed (Fig. 6A and B), indicating that resveratrol did not significantly inhibit lung metastasis of colon cancer after AKT1 knockdown. In addition, our animal experiments revealed that some nude mice did not have lung metastasis, while other nude mice had small or diffuse lung metastases, which could not be counted and measured. Since we could not separate the lung tumors alone, whole lung tissue was used for subsequent experiments. H&E and immunohistochemical images of tumors were observed under a 200-fold magnifying microscope. The total protein in the animal western blotting experiment was extracted from lung tissue.

Furthermore, H&E staining revealed histopathological changes in tumor foci after treatment (Fig. 6C). The nuclei of tumor cells in the NC-control group were hypertrophic, deformed, intensely stained, and evidently heterogeneous and had several new tumor blood vessels. Tumor cell density, pathological mitotic phase, heteromorphic cells, and new tumor blood vessels were lower in the NC-resveratrol group than in the NC-control group. There was no difference between pathological mitosis and the number of heterotypic cells in the si-AKT1-control and si-AKT1-resveratrol groups. However, the si-AKT groups exhibited markedly lower pathological mitosis, number of heteromorphic cells, and new tumor blood vessels than those in the NC-control group.

To further confirm whether resveratrol inhibited lung metastasis of colon cancer in vivo via the AKT/GSK-3β/Snail signaling pathway, western blot and immunohistochemical staining assays were performed to detect the expression of EMT-markers as well as that of AKT/GSK-3β/Snail signaling pathway-related molecules in the tumor tissues of nude mice. Western blotting (Fig. 7) and immunohistochemical staining (Fig. 8) assays revealed consistent results. E-cadherin expression was upregulated in the NC-resveratrol and si-AKT1-control groups compared to the NC-control group, whereas the expression of N-cadherin, p-AKT, p-GSK-3β, and Snail was downregulated (P<0.05 and P<0.01, Figs. 7 and 8), indicating that resveratrol and AKT1 knockdown could reverse EMT in colon cancer cells in tumor tissues and promote the transformation of colon cancer cells from mesenchymal to epithelial phenotypes. However, there was no difference in the expression of these proteins in the si-AKT1-resveratrol and si-AKT1-control groups (P>0.05, Figs. 7 and 8), further confirming that the ability of resveratrol to reverse EMT in colon cancer cells in tumor tissues almost disappeared after AKT1 knockdown.