Cell culture and reagents
OSCC cell lines YD-9 and YD-38 were obtained from the Korean Cell Line Bank (Seoul, Korea). YD-10B cells were a gift from Dr. J.I. Yook (Yonsei University College of Dentistry, Seoul, Korea). Oral cancer-associated fibroblasts (CAFs) were generously provided by Dr. X.L. Zhang (Yonsei University College of Dentistry, Seoul, Korea). Oral CAFs were cultured in DMEM and F-12 Ham mixed in a 3:1 ratio and supplemented with 10% FBS and 1% penicillin/streptomycin. All OSCC cells were cultured in RPMI supplemented with 10% FBS and 1% penicillin/streptomycin at 37 °C under 5% CO2 in a humidified incubator. Gefitinib was purchased from Selleckchem (Houston, TX). REV was purchased from Sigma-Aldrich (St Louis, MO). All other reagents were of the purest grade available.
siRNA and plasmid transfection
Cells were transiently transfected using Lipofectamine 3000 or RNAiMAX (Invitrogen, Carlsbad, CA) as described previously11. RCP and Zeb1 constructs were in the pEGFP-C3 and pCMV6A vectors, respectively. β1 integrin cDNA was added to the pcDNA3 vector. Each empty vector was used as the negative control. RCP and β1 integrin constructs were provided by Jim Norman (Beaston Cancer Institute, Glasgow, UK) and Dr. Y.S. Lee (Ewha Woman’s University, Seoul, Korea), respectively. The vector for Zeb1 was purchased from Switchgear Genomics (Carlsbad, CA). siRNAs against RCP, Zeb1, N-cadherin, and β1 integrin were purchased from Sigma-Aldrich, and β-catenin siRNA was purchased from Santa Cruz Biotechnology (Dallas, Texas). Control scrambled siRNA was purchased from Invitrogen. Lentivirus packaging and transduction were performed with a Second Generation packaging mix kit (Abm, Richmond, BC, Canada) according to the manufacturer’s protocol. Briefly, the recombinant lentivirus was produced by transfecting HEK293T cells with expression and packaging vectors. The supernatant medium was harvested twice at 24 h and 48 h. The prepared YD-10B cells were transduced with the virus-containing supernatant with polybrene (5 mg/ml, Merck Millipore, Darmstadt, Germany). The permanent RCP transfectants were selected using puromycin (1 μg/ml, Thermo Fisher Scientific Inc., Rockford, IL) for 7 days, and the expression of RCP was evaluated by immunoblotting.
Isolation of total cellular RNA and subsequent analysis were performed as previously described11,17. Complementary DNA was amplified using an iQ5 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA) with the following primer sets: RCP, 5′-GGATGTCTCCGAATCTTCCA-3′ (forward) and 5′-CCGTCATCAGAGACAGCAAA-3′ (reverse); MT1-MMP, 5′-TTGGACTGTCAGGAATGAGG-3′ (forward) and 5′-GCAGCACAAAATTCTCCGTG-3′ (reverse); MMP2, 5′-ATGACAGCTGCACCACTGAG-3′ (forward) and 5′-AGTTCCCACCAACAGTGGAC-3′ (reverse); MMP9, 5′-GTGCCATGTAAATCCCCACT-3′ (forward) and 5′-CTCCACTCCTCCCTTTCCTC-3′ (reverse); uPA, 5′-GTGGCCAAAAGACTCTGAGG-3′ (forward) and 5′-GCCGTACATGAAGCAGTGTG-3′ (reverse); Zeb1, 5′-AAGAATTCACAGTGGAGAGAAGCCA-3′ (forward) and 5′-CGTTTCTTGCAGTTTGGGCATT-3′ (reverse); and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5′-A CAGTCAGCCGCATCTTCTT-3′ (forward) and 5′-ACGACCAAATCCGTTGACTC-3′ (reverse). The GAPDH gene was used as a control for calculating dCt. The real-time PCR data were analyzed using the 2−(ddCt) method.
Immunoblotting was analyzed as previously described18. Briefly, the cell lysates were resolved by SDS-PAGE. PVDF membranes with proteins were blocked and incubated for 30 min. Antibodies against RCP (1:1000, 12849S), Snail (1:1000, 3879S), N-cadherin (1:1000, 4061S), Zeb1 (1:1000, 3396S), p-EGFR (1:1000, 4407S), EGFR (1:1000, 2232 S), MT1-MMP (1:1000, 13130S) and nonphospho (NP) β-catenin (1:1000, 19807S) were purchased from Cell Signaling Inc. (Danvers, MA). Antibodies against GAPDH (1:3000, sc-47724), Twist (1:1000, sc-15393), β-catenin (1:1000, sc-7963), and β1 integrin (1:1000, sc-53711) were purchased from Santa Cruz Biotechnology. An antibody against E-cadherin (1:1000, 610182) was obtained from BD Biosciences (Franklin Lakes, NJ). The immunoreactive bands were exposed via ECL (Thermo Fisher Scientific Inc.) using an Amersham Imager 600 (GE Healthcare, Buckinghamshire, UK).
Immunofluorescence was performed as previously described19. Briefly, immunofluorescence detection was conducted with antibodies against RCP (1:500, Cell Signaling Technology), E-cadherin (1:500, Santa Cruz Biotechnology), β1 integrin (1:500, Santa Cruz Biotechnology), MT1-MMP (1:500, Santa Cruz Biotechnology), and LAMP1 (1:500, Abcam, Cambridge, UK) overnight. The cells were washed with ice-cold PBS and incubated with Cy3-conjugated goat anti-rabbit IgG (1:500, Jackson ImmunoResearch, PA), Cy2-conjugated goat anti-mouse IgG (1:500, Jackson ImmunoResearch) and Cy5-conjugated goat anti-rat IgG (1:500, Jackson ImmunoResearch). The nuclei of cells were marked with 4′,6′-diamidino-2-phenylindole (DAPI, Invitrogen). The cells were observed by confocal microscopy (x600, LSM710, Carl Zeiss).
In vitro invasion and wound healing assay
Invasion and wound healing analyses were performed as previously described11,20. Briefly, 0.4 × 106–1 × 106 cells were loaded in the upper well of the invasion chamber. After incubation for 13–15 h at 37 °C, the cells that invaded the membrane were fixed and stained with Diff-Quik reagents (Sysmex Co., Kobe, Japan). The invaded cells were calculated by counting the number of cells in three arbitrary high-power fields for each replicate (×200) with a light microscope. The results were derived from three independent experiments. For wound healing analysis, the cells were seeded in 6-well plates for 24 h and then transfected with the indicated vector and siRNA. After serum starvation, the cells were scraped with a 200 μl pipette tip. The gap distance of cells was observed after 24 h incubation in the same locations.
Three-dimensional (3D) Matrigel culture
Three-dimensional (3D) Matrigel culture was executed as previously described11. Briefly, cells were suspended in 2% Matrigel and put over a layer of polymerized 100% Matrigel at 1 × 104 cells/ml in an eight-well chamber slide (Nunc, Littleton, CO). RPMI culture medium was replaced once every two days. Cultures were analyzed after 7 days of cultivation.
Three-dimensional (3D) gel invasion assay
3D gel invasion was analyzed as described previously with some modifications21,22,23. Cells were embedded in a mixture of 20% type I collagen (Nitta Gelatin Inc, Cellmatrix Type I-P, Japan) and Matrigel (BD Biosciences) in Transwell (0.3 μm pore size, Corning, Acton, MA). Oral CAFs and YD-10B cells were labeled with DiI (Oral CAFs; Thermo Fisher Scientific, Waltham, MA) and DiO (YD-10B cells; Thermo Fisher Scientific), respectively. Oral CAFs (2 × 104) and YD-10B cells (2 × 104) were mixed in 200 μl of medium (1:1 mixture of RPMI and medium of oral CAFs) supplemented with 0.2% FBS and plated on gels. The low chamber of the Transwell was filled with 800 μl of the above medium with 10% FBS. After 5 days, the embedded gel was sectioned without a fixture, and the cells were analyzed by fluorescence confocal microscopy. In these images, the distance of invaded cells was measured from eight different positions and calculated by the ZEN blue edition program of Carl Zeiss Microscopy GmbH. The distance in μm was calculated as described previously24.
The cell viability assay was determined as previously described20,25. Briefly, cells were seeded in 96-well plates, serum-starved and treated with or without REV for 24 h. The absorbance was calculated at 540 nm using a SYNERGY/HTX ELISA plate reader (BioTek, Winooski, VT).
Proteome extraction assay
Proteome extraction assays were performed according to the manufacturer’s protocol (Merck-539790). Briefly, YD-10B cells were transfected with the indicated vectors for 72 h. After washing with cold washing buffer, the cells were added to fraction buffers to isolate supernatants with proteins in the cytosol (buffer I), membrane (buffer II) and nucleus (buffer III). Supernatants from buffers I and III were resolved by SDS-PAGE.
Data are shown as the means ± standard deviation (SD). Differences between two groups were assessed with SigmaPlot software (Systat Software, San Jose, CA) using Student’s t-test, and statistical significance was set at p-values less than 0.05. Differences among three or more groups were estimated by analysis of variance followed by Bonferroni multiple comparison tests.