HepG2 and SNU449 cells were generously provided by Dr Mehmet Ozturk, Department of Molecular Biology and Genetics, Bilkent University, Ankara, Turkey. HepG2 cells represent an early stage of liver cancer and are derived from a well-differentiated tumor from a 15-year-old Caucasian male. HepG2 is classically described as an HCC cell line, but has also been suggested to have originated from hepatoblastoma (26). SNU449 represents an intermediate stage of HCC and is derived from a grade II–III/IV HCC from a 52-year-old Asian male, positive for HBV DNA. The cells were cultured in RPMI-1640 basal medium (Lonza Group, Ltd.) supplemented with 10% FBS (Invitrogen; Thermo Fisher Scientific, Inc.), 100 U/ml penicillin and 100 mg/ml streptomycin (Invitrogen; Thermo Fisher Scientific, Inc.). The cells were incubated in a humidified incubator at 37°C with 5% CO₂. Images were captured using an inverted fluorescence phase contrast microscope at ×100, ×200 or ×400 magnifications (Olympus IX70; Olympus Corporation); some images were converted to grey-scale and were adjusted for brightness.

The NELF-B overexpression plasmid, pCMV5-HCOBRA1, was a generous gift from Dr Rong Li, University of Texas Health Science Center, San Antonio TX, USA. An empty pCMV5 plasmid was used as a negative control. Briefly, pCMV5-HCOBRA1 was digested with EcoRI and SalI (New England Biolabs, Inc.). The empty vector (4.7 Kb) was purified using the QIAquick Gel Extraction kit (Qiagen Sciences, Inc.), following the manufacturer's protocol. The blunting of the 5-overhangs resulting from the digestion was performed by filling the ends with the Klenow fragment of DNA polymerase I (New England BioLabs, Inc.). For each µg of DNA, 1 unit of Klenow was added, followed by incubation at 25°C for 15 min and heating at 67°C for 20 min. T4 DNA ligase (Promega Corporation) was used to re-ligate the blunted vector. The pEGFP-N1 expression plasmid was used to assess the transfection efficiency. The plasmid was provided by Dr Ahmed Osman, Ain Shams University, Cairo, Egypt, and was initially purchased from Clontech Laboratories, Inc.

HepG2 cells were transfected using Lipofectamine^® 3000 (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Briefly, HepG2 cells were cultured in 6-well plates and incubated for 24 h until they were 70–90% confluent. Before transfection, 3.75 µl Lipofectamine 3000 was added to 125 µl Opti-MEM reduced serum medium (Gibco; Thermo Fisher Scientific, Inc.) and the mixture was incubated for 5 min at room temperature. Meanwhile, 2.5 µg of either the NELF-B overexpression plasmid pCMV5-HCOBRA1 or pCMV5 empty plasmid was diluted in a mixture of 125 µl Opti-MEM and 5 µl Lipofectamine 3000. The two dilutions were mixed and incubated for 20 min to allow for complex formation. The DNA-Lipofectamine complex (250 µl) was then added dropwise to each well containing cells in 2 ml of antibiotic-free medium (RPMI + 10% FBS). Overexpression efficiency was assessed based on the percentage of green fluorescent cells, as well as the protein and mRNA expression levels of NELF-B following transfection. Transfection efficiency was measured 24 and 48 h post-transfection, and was found to be 45 and 60%, respectively using an inverted fluorescence phase contrast microscope at ×100 magnification (Fig. S1). At 48 h, NELF-B expression in HepG2 cells transfected with the NELF-B overexpression plasmid (pCMV5-HCOBRA1) was increased 4-fold compared with that in cells transfected with pCMV5 empty plasmid (P≤0.0001; Fig. 1A). The cells were collected 48 h post-transfection for downstream analysis.

siGENOME SMARTPool siRNA (cat. no. M-015839-02; GE Healthcare Dharmacon, Inc.), a pool of four siRNAs targeting different exons of the NELF-B mRNA, was used for the knockdown of NELF-B. Allstars' negative control siRNA (cat. no. SI03650318; Qiagen, Inc.) was used as a control. Approximately 1.5×10⁵ cells suspended in antibiotic-free medium (RPMI + 10% FBS) were mixed with the transfection complex prior to seeding (reverse transfection). The complex was prepared by adding 40 nM siRNA and 3.75 µl Lipofectamine 3000 to 500 µl Opti-MEM medium, followed by incubation for 15–20 min at room temperature. The medium was changed 24 h after transfection, and fresh antibiotic-free medium (RPMI + 10% FBS) was added. The optimum knockdown conditions were determined by performing the transfection using different siRNA concentrations (25 and 40 nM) and post-transfection incubation time points (48, 72 and 96 h) (data not shown). The use of 40 nM siRNA, and an incubation of 96 h after transfection, resulted in a decrease in the protein levels of NELF-B by an average of 96% (P≤0.001; Fig. 1C), which was considered to be the optimal knockdown. The cells were collected 96 h following transfection for downstream analysis.

Cell viability was determined using the trypan blue exclusion assay (27). HepG2 cells were harvested and counted 48 h post-transfection to monitor the effect of NELF-B overexpression on cell proliferation. For monitoring the effect of NELF-B silencing on cell proliferation, the number of SNU449 cells was counted 96 h post-transfection. Cells were mixed with 0.4% trypan blue in PBS at a ratio of 1:1, and viable cells were counted using a hemocytometer (Hausser Scientific) and an inverted fluorescence phase contrast microscope at ×200 magnification.

TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was used for extraction of total RNA, according to the manufacturer's recommendations. RevertAid First Strand cDNA synthesis kit (Thermo Fisher Scientific, Inc.) was used for cDNA synthesis, following the manufacturer's protocol. Random primers were used for reverse transcription of 0.5 µg of the total RNA in a final volume of 20 µl. MyTaq DNA Polymerase kit (Bioline; Meridian Bioscience) was used for semi-quantitative RT-PCR using 1 µl cDNA, according to the manufacturer's recommendations. All the genes were analyzed under similar PCR conditions, except for the cycle number and annealing temperature (Table I). Initial denaturation was performed at 94°C for 3 min, followed by cycles of 94°C for 30 sec, annealing temperature for 30 sec and 72°C for 45 sec, and a final extension at 72°C for 7 min. Amplicons were electrophoresed on a 2–2.5% agarose gel and visualized using the Gel Doc EZ System (Bio-Rad Laboratories, Inc.). Intensities of bands generated in RT-PCR were quantified using ImageJ software version 1.51j8 (National Institutes of Health) (28). The band intensities were normalized to those of respective loading controls (GAPDH or β-tubulin). Images were converted to grey scale.

qPCR amplification was performed using SYBR-Green as a DNA-specific binding dye, and continuous monitoring of fluorescence was done. Each PCR reaction (10 µl) consisted of 2X SYBR-Green I Master Mix (Applied Biosystems; Thermo Fisher Scientific, Inc.), 1 µl cDNA and 0.5 µl each primer. All the primers used for the selected genes are listed in Table II. The thermal cycler was set up for 40 cycles of the following amplification program: 50°C for 2 min, 95°C for 2 min, 95°C for 15 sec and 60°C for 1 min. The instrument was set to perform the default dissociation step under the following conditions: 95°C for 15 sec, 60°C for 1 min and 95°C for 15 sec. All experiments were performed three times using the 7500 Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). Data analysis was performed using the 2^−ΔΔCq method (29) with GAPDH used as the reference gene.

Cells were washed with ice-cold PBS before being lysed in 1X ice-cold Laemmli Lysis Buffer (50 mM Tris pH 6.8, 2% SDS and 10% glycerol) supplemented with 100X Halt Protease Inhibitor Cocktail (Thermo Fisher Scientific, Inc.; 10 µl protease inhibitor per ml of Laemmli Lysis buffer was used). Protein concentrations were determined using the BCA Protein Assay kit (Pierce; Thermo Fisher Scientific, Inc.), following the manufacturer's protocol. Whole-cell lysates (20–30 µg/lane) were separated via 10% SDS-PAGE and the proteins were blotted onto a nitrocellulose membrane. The membranes were blocked in 5% non-fat dry milk, and then incubated overnight with primary antibodies at 4°C. Subsequently, the membranes were incubated for 1 h with the secondary antibody at room temperature. Colorimetric detection of the tested proteins was performed using BCIP/NBT Phosphatase colorimetric substrate (KPL). Primary antibodies used were as follows: Mouse monoclonal anti-GAPDH (Abcam; cat. no. ab8245; 1:10,000 in 5% non-fat dry milk), mouse monoclonal anti-β-tubulin (Sigma-Aldrich; Merck KGaA; cat. no. T7816; 1:20,000 in 5% non-fat dry milk), rabbit monoclonal anti-NELF-B (anti-COBRA1; Abcam; cat. no. ab167401; 1:1,000 in 3% non-fat dry milk) and rabbit monoclonal anti-NELF-E (Abcam; cat. no. ab170104; 1:1,000 in 5% non-fat dry milk). Secondary antibodies used were polyclonal goat anti-mouse (KPL; cat. no. 4751-1806; 1:10,000 in 5% non-fat dry milk) and polyclonal goat anti-rabbit (KPL; cat. no. 4751-1516; 1:10,000 in 5% non-fat dry milk), both have alkaline phosphatase conjugates. For all purposes, non-fat dry milk used in western blot analysis was dissolved in 1X TBS-0.1% Tween-20. Intensities of bands generated were quantified using ImageJ software version 1.51j8 (National Institutes of Health) (28). The band intensities were normalized to those of respective loading controls (GAPDH or β-tubulin).

Classic wound healing assay (30) was performed to assess the effect of the overexpression/knockdown of NELF-B on the cell migration rate. After 48 h from NELF-B overexpression, and 96 h post-NELF-B silencing, the spent medium was discarded, a scratch was made in the confluent (~90%) cell monolayer using a sterile 20 µl yellow tip. The cells were gently washed twice with PBS and incubated for 24 h in fresh 10% FBS RPMI medium. The migration of cells was monitored using an inverted fluorescence phase contrast microscope at ×100 magnification (Olympus IX70; Olympus Corporation); images were converted to greyscale. The wound area was measured using the TScratch software version 1.0 (31), and the percentage of wound closure was calculated using the following equation (32):

where WC % is the percentage of wound closure, WA 0 h is the wound area at 0 h, and WA 24 h is the wound area at 24 h.

Transwell Boyden 24-well chambers (ThinCert™; Greiner Bio-One) were used to determine the effect on the invasive capacity of cells. Cells were harvested 96 h post-transfection, and 2×10⁵ cells were suspended in 100 µl RPMI-1640 medium (supplemented with 1% FBS) and seeded in the upper chamber of the Transwell cell culture insert (8-µm pore size). The cell culture insert was coated with 40 µg collagen I (SERVA Electrophoresis GmbH; cat. no. 47254), allowed to dry overnight in an incubator at 37°C, and rehydrated with 1% FBS-containing medium, 30 min before the seeding of cells. The lower chamber was filled with 600 µl RPMI-1640 medium supplemented with 10% FBS. Cells were placed inside the upper chamber and incubated for 22 h at 37°C. Subsequently, cotton swabs were used to remove cells on the upper side of the membrane, and the cells that invaded to the lower side of the membrane were fixed using 4% formaldehyde for 10 min at room temperature, and subsequently stained with DAPI for 10 min at room temperature in the dark (KPL, Inc.; cat. no. 71-03-01; 1:1,000 in PBS). The images were captured with a fluorescence microscope at ×200 magnification in five random fields for each condition, and the average of cell counts was compared.

Relative gene expression was calculated in reference to the negative control siRNA or pCMV5-empty and represented as fold-change. Data are presented as the mean ± SD of two to four independent experiments. Statistical significance for comparison between two groups was performed using unpaired Student's t-test (two-tailed). GraphPad Prism 8.0 (GraphPad Software, Inc., La Jolla California USA) was used to generate graphs. P<0.05 was considered to indicate a statistically significant difference.

Downregulation of NELF-B in the early-stage liver cancer HepG2 cell line has been previously shown to inhibit the migration and proliferation of the cells (25). To investigate whether ectopic expression of NELF-B in the same cell line would lead to an opposite effect to that induced by its downregulation, NELF-B was overexpressed in HepG2 cells. To further investigate the role of NELF-B in the progression of liver cancer, loss of function analysis was performed on the intermediate-stage HCC cell line, SNU449. The efficiency of overexpression and knockdown was assessed by measuring the protein expression levels by western blot analysis. HepG2 cells transfected with the NELF-B overexpression plasmid (pCMV5-HCOBRA1) exhibited a 4-fold increase in NELF-B expression compared with the pCMV5-empty-transfected control cells (P≤0.0001; Fig. 1A); however, NELF-E expression was not significantly altered upon overexpression of NELF-B in HepG2 cells (P>0.05; Fig. 1B). SNU449 cells transfected with NELF-B siRNA showed a significant decrease in protein expression by an average of 96% compared with that in the negative siRNA-transfected control cells (P≤0.001; Fig. 1C). Consistent with the protein expression levels, the mRNA expression levels of NELF-B assessed by RT-PCR exhibited a 2.2-fold increase in HepG2 cells following NELF-B overexpression and an average of 93.5% decrease in SNU449 cells following NELF-B-knockdown compared with their respective controls (P≤0.0001 and P≤0.001; Fig. 1D and E).

The effect of overexpression and knockdown of NELF-B on the expression levels of other NELF subunits was assessed using RT-PCR. The expression levels of NELF-A, NELF-C/D and NELF-E were not significantly affected following the overexpression of NELF-B (P>0.05, Fig. 1D). In SNU449 cells, the expression levels of NELF-C/D and NELF-E were not significantly altered, while NELF-A expression was significantly decreased by an average of 30% upon knockdown of NELF-B (P≤0.001; Fig. 1E).

No morphological changes were observed after the overexpression or knockdown of NELF-B compared with the control cells (Fig. 2A and B). However, cells in which NELF-B was knocked down appeared less dense than their negative controls, suggesting a decrease in the proliferation rate and/or survival of cells (Fig. 2B). To further examine the effect of deregulation of NELF-B on cell proliferation, cells were counted at time points when optimum overexpression/knockdown was observed post-transfection (48 and 96 h for HepG2 and SNU449 cells, respectively). The count of NELF-B-overexpressing HepG2 cells was similar to that of cells transfected with the empty vector, and no significant difference in Ki-67 mRNA expression was observed (P>0.05; Fig. 2C). By contrast, the cell count was significantly decreased upon silencing of NELF-B in SNU449 by an average of 58% compared with the negative control (P≤0.001; Fig. 2D). Consistent with these results, a significant decrease in the expression levels of the proliferation marker, Ki67 (by an average of 51%), compared with the negative control was also observed (P≤0.05; Fig. 2D).

Cancer progression involves epithelial-mesenchymal transition (EMT), a process during which the migratory and invasive abilities of cells increase (33). The present study investigated the association between NELF-B and EMT in liver cancer. Wound healing assay was used to examine the effect of overexpression and knockdown of NELF-B on cell migration. Scratches were made in cell monolayers, 48 and 96 h post-transfection in HepG2 and SNU449 cells, respectively, and wound closure was monitored 24 h later. The overexpression of NELF-B in HepG2 cells resulted in a 1.4-fold higher migration rate compared with that in the negative control group; however, the difference was not statistically significant (P>0.05; Fig. 3A). In SNU449 cells, knockdown of NELF-B significantly decreased the cell migration rate by an average of 49% compared with that in the negative control group (P≤0.001; Fig. 3B).

Transwell invasion assay was performed to assess the effect of NELF-B-knockdown on the invasive capacity of cells through the extracellular matrix equivalent collagen. The knockdown of NELF-B significantly decreased the number of cells that invaded through collagen by an average of 30% compared with the control (P≤0.01; Fig. 3C).

RT-qPCR was used to assess some of the primary EMT markers, including N-cadherin, vimentin and β-catenin (34), in SNU449 cells following NELF-B-knockdown. In line with the results of invasion and migration assays, the expression levels of N-cadherin and vimentin were significantly decreased by an average of 75 and 70%, respectively, compared with the negative control (P≤0.01; Fig. 3D). Another critical gene involved in EMT is the signal transducer molecule, β-catenin. β-catenin expression is usually elevated in cancer and the protein is localized in the nucleus, where it drives the expression of downstream genes involved in the EMT process (35–37). A significant decrease was observed in the expression levels of β-catenin post-NELF-B-knockdown by an average of 85% (P≤0.05; Fig. 3D). The expression levels of the transcription factor FOXF2 were also analyzed, which is reported to enhance the invasion and migration of HCC cells (38). FOXF2 expression was significantly decreased by an average of 60% following NELF-B-knockdown (P≤0.05; Fig. 3D).

To examine whether NELF-B affected apoptosis in liver cancer, the mRNA expression levels of survivin were measured, which is known to be deregulated in cancer and serves critical roles in the survival and proliferation of cancer cells (39). The expression levels of wild-type survivin have been previously shown to be decreased upon knockdown of NELF-B in HepG2 cells (25). In the present study, the mRNA expression levels of survivin were examined following the overexpression and knockdown of NELF-B. No significant difference in survivin expression was detected upon overexpression of NELF-B in HepG2 cells (P>0.05; Fig. 4A). By contrast, survivin expression was decreased by an average of 43% following silencing of NELF-B in SNU449 cells (P≤0.05; Fig. 4B).

DAPI staining revealed fragmented nuclei in cells in which NELF-B was silenced (Fig. 4D). Fragmented nuclei are considered a marker for apoptosis (40), which is in agreement with the decreased survivin expression and decrease in cell counts.

Several types of cancer, such as breast, gastric, prostate and colorectal cancers exhibit aberrant expression levels of trefoil factor family members, such as TFF1 and TFF3 (41–45). In HCC, the downregulation of TFF1 and the upregulation of TFF3 are frequently observed (46,47). In the present study, the mRNA expression levels of these genes were assessed after knockdown of NELF-B in SNU449 cells. A significant increase of 1.67 fold was observed in the expression levels of TFF1 (P≤0.01; Fig. 4C), while TFF3 expression was decreased by an average of 21% (P>0.05, Fig. 4C) compared with the negative control group. Notably, the overexpression of NELF-B in HepG2 cells did not significantly alter the expression levels of either of these genes (data not shown).