Cell culture and genetic manipulations
Hepatocellular carcinoma cells C3A (HepG2/C3A, a clonal derivative of HepG2, ATCC, Manassas, VA), and Hep3B (Hep 3B2.1–7, ATCC) were grown with 4.5 g/L glucose DMEM (Gibco, Thermo Fisher Scientific, Waltham, MA), plus 10% FBS and 100 U/ml penicillin and streptomycin (Invitrogen, Waltham, MA). Cells were cultured in a humidified incubator at 37ºC with 5% CO2, and tested periodically for mycoplasma by PCR.
When indicated, alpha lipoic acid (0.5 or 1 mM, Santa Cruz Biotechnology Inc., Santa Cruz, CA) was added.
For reducing AMPKα1 protein expression in C3A cells, 21 nucleotide duplexes targeting two specific sequences and a scrambled control were designed and synthesized by Ambion SilencerTMsiRNA (Ambion, Thermo Fisher Scientific). The target sequences were AMPKα1 1842–1864: AACATTTCTGCATATTAGGCTCCTGTCTC and 2659–2681: AAGAGCTGAGTTGCATATACTCCTGTCTC, as we previously reported19. Experiments with silenced cells were carried out 24–48 h after transfection using siPORT NeoFX (Ambion, Thermo Fisher Scientific) with the Negative Control siRNA (Ambion) or the synthesized siRNAs against AMPKa1, and decrease (60 to 80%) in AMPKα expression for both duplexes was confirmed by immunoblotting.
For stable knockdown of p53 in C3A cells, transduction with retrovirus-based plasmids harboring the following specific and control shRNAs were carried out: shp53 (GACUCCAGUGGUAAUCUAC), shp53-3′UTR (GGUGAACCUUAGUACCUAA), and control shLacZ (GUGACCAGCGAAUACCUGU), as we previously described29. Plasmid DNA was prepared using the Wizard DNA purification kit (Promega, Madison, WI). p53 diminution (80–90%) was confirmed by qRT-PCR.
Colony formation assay
Low density cell inoculum (1000 cells) was seeded in 35 mm dishes of 6-well plates and grew up to 15 days to allow the formation of visible colonies. After washing the colonies were fixed with cold methanol and toluidine blue staining was performed for 10 min. After removing the excess of colorant, the dishes were photographed and the number of colonies were counted and compared.
MTT assay
HCC cells were plated at the same density by triplicate in 96-well microplates and incubated 2 h with methylthiazolyldiphenyl-tetrazolium bromide (MTT, Sigma Chemical Co., St Louis, MO) at the indicated time. The metabolite produced from viable cells was dissolved in DMSO and detected at 540 nm in a microplate reader (Beckman Coulter LD400), as we described19. Results were expressed as percentage of absorbance in cells at 0 time.
Annexin V/propidium iodide assay
Annexin V assay (BD Biosciences, San José, CA) was used following the manufacturer recommendations and the samples were prepared as we previously described18. Briefly, after 48 h of treatments, homogenization in the medium and harvest, 100,000 cells were gently re-suspended and phosphatidylserine externalization and cell death was detected by staining with Annexin V-FITC and PI, respectively. The stained cells were subjected to flow cytometric analysis (Cell Sorter BD FACSAria II, BD Biosciences).
Cell cycle analysis
The proportion of cells in each cell cycle population at 48 h of culture in different conditions was analyzed by determining the cellular DNA content by flow cytometry (Cell Sorter BD FACSAria II, BD Biosciences), as we previously described18. In brief, 1 × 106 cells were fixed in 70% ethanol, washed and stained with 50 µg/ml propidium iodide (Sigma Chemical Co.) in a solution containing 0.1% sodium citrate, 0.02 mg/ml RNAse, and 0.3% NP-40. WinMDi and Cylchred softwares were used for data analysis.
Wound healing assay
Collective cell migration was determined by wound healing assay as we previously described30, with slight modifications. Briefly, cells were plated (6-well plates) at 1.5 × 106 cells/ml and growth for 24 h. Monolayers were wounded by dragging a 200-µl pipette tip, washed and treated. Captures of microscopy images (Zeiss Axiovert 25) of wounds were obtained at 0 h and 24 h after wounding. Lineal advance of the wound front (µm) was assessed by using ImageJ.
Invasion studies
For invasion assays, transwell chambers (Biofil, Beijing, China) with 8 µm pore size polyester membrane filters were coated with 250 µg/ml Matrigel (Corning, Corning, NY). Suspension of cells (1 × 105) in DMEM containing 1% FBS were plated into the upper chamber, and the lower chamber was filled with complete DMEM medium containing 10% FBS. Cells were treated for 48 h and after this period those cells that reached the lower chamber were fixed with methanol, stained with 1% toluidine blue-1% borax for 5 min, and counted by microscopy captures, as we previously described19.
Immunofluorescence microscopy
Cells were grown on coverslips, treated for 24 h and, at the end of experiments, fixed, permeabilized and blocked with 0.3% Triton X-100-1% albumin. Then, cell samples were incubated with anti p53 (sc-126, Santa Cruz Biotechnology), washed, incubated with anti-mouse secondary antibody (Invitrogen) and then incubated for nuclei staining with 4′,6-diamidino-2-phenylindole (DAPI, Molecular Probes, Eugene, OR), and mounted with ProLong (Molecular Probes). Fluorescence was detected by a confocal microscopy (Nikon C1SiR with inverted microscope Nikon TE200). Nuclear intensity of p53 staining was assessed in the Z-stack projections by using ImageJ tools.
Western Blotting
Cell lysates were obtained after scrapping and washing the cultured cells subjected to the indicated treatments during the indicated time. The harvest cells were incubated in 100 µl of ice cold RIPA buffer with protease and phosphatase inhibitors for 30 min and sonicated. Protein concentration was determined by the Lowry's method. Equal amounts of total protein (30–50 µg) were loaded, resolved by electrophoresis on SDS-polyacrylamide gels31, and transferred to polyvinylidene difluoride membranes (Perkin Elmer Life Sciences, Boston, MA, USA). Molecular weights were estimated using “rainbow” molecular weight markers. The membranes were blocked with 5% non-fat milk/0.3% Tween/PBS, washed and incubated O.N. with each primary antibody [α tubulin (T-5168 Sigma Chemical Co.); p53(DO-1), GAPDH (FL-335) (sc-126 and sc-25778 Santa Cruz Biotechnology); AMPKα, Phospho-AMPKα (Thr172), SNAIL/C15D3 (2532, 2535, 3879 Cell Signaling Technology); or E-cadherin (610182 BD Biosciences)]. After incubation of the membranes with the appropriate secondary antibody, the bands were detected by chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ).
Real-time quantitative PCR
cDNA was obtained from Trizol (Invitrogen) extracts by retrotranscription with retrotranscriptase M-MLV (Promega) following the manufacturer conditions, and real-time PCR was performed using SYBR Green PCR master mix (Promega) according to the following conditions: 2 min at 95 ºC for one cycle; and 30 s at 95 ºC, 20 s at 60 ºC, 30 s at
72 ºC for 40 cycles. Results were analyzed using the comparative Ct method. Values were
normalized to GAPDH expression. The following real-time PCR forward and reverse primers were used, respectively: TP53: CTCCTCTCCCCAGCCAAAGA, GGAACATCTCGAAGCGCTCA; CDH1: AAGAAGCTGGCTGACATGTACGGA, CCACCAGCAACGTGATTTCTGCAT; VIM: CCCTGCAATCTTTCAGACAG, ATCTGAGCCTGCAGCTCCT; SNAI1: CCAGAGTTTACCTTCCAGCAGCC, CAGGACAGAGTCCCAGATGAGCAT; GAPDH: TCTCTGCTCCTCCTGTTC, GCCCAATACGACCAAATCC.
Bioinformatic analyses
RNAseq data from HCC patients (371 tumor and 50 non tumor samples) were obtained from TCGA datasets accessible from http://gdac.broadinstitute.org/. Data were organized and processed for differential gene expression (DGE) analysis for a selected group of EMT related genes with EDAseq and edgeR, respectively. Other packages from tydiverse were used for data management. Mean log2 fold change (log2FC), p, and false discovery rate (FDR) values were obtained. Values of log2FC > 1 or < 1, and -log2FDR > 1 (FDR < 0.01) were set as cutoffs. A heatmap plot from the 50 patients with paired tumor and non-tumor samples was performed for the indicated genes using complexHeatmap, and the results were splitted in two groups by the status of TP53 gene. In the same dataset a comparison of the distribution of frequencies of log2FC values for each gene was performed between those patients with wild type (n = 37) or mutant TP53 (n = 13) using the Kolmogorov–Smirnov test. This analysis was performed with SAS Studio, and a p-value < 0.05 was considered statistically significant.
Further expression profiles of the genes studied from the same TCGA patients were analyzed through the tools of the UALCAN data portal (http://ualcan.path.uab.edu). These tools match expression and clinical data allowing comparison of expression levels in different groups of patients, as cancer stages and others. Data processing and statistical analysis were described by the authors32. In addition, expression plus survival assays of these genes from the same dataset of TCGA were also analyzed with the visualization tools of The Human Protein Atlas (www.proteinatlas.org).
For those groups of patients with low or high expression (below or above the median) of the selected genes with wild type or mutant TP53 survival analysis were performed using survival and survminer R tools. p < 0.05 was set as a significant difference among the Kaplan-Meier curves. When it was necessary, hazard ratios (HR) were calculated and represented as a forest plot and the p value for HR between the reference group versus each of the others was obtained.
Statistical analysis
In all the studies except for bioinformatics, data were expressed as mean ± S.E.M. and statistical comparisons were made by using Student t test. p < 0.05 was considered statistically significant.