Cell culture

Cell lines were cultured in Dulbecco’s modified Eagle’s medium (Hyclone) containing 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Hyclone) at 37 °C in 5% CO₂. Passages 6–15 were used for the experiments. Cells were routinely passaged before reaching ~90% confluence. Transfections were done using the Turbofect transfection reagent (Thermo) according to DNA: reagent ratios suggested by the manufacturer.

Huh7 cells (Cat no. PTA-4583) were treated with Wnt agonist (Calbiochem) and GSK-3β Inhibitor XII, TWS 119 (Calbiochem) in concentration-dependent manner. DMSO was used as negative control.

For the generation of stable cell lines; Huh7 cells in 6-well plates were transfected separately with the following constructs: pIRES2-EGFP-MGAT1 and pIRES2-EGFP-GFP (control). Transfected cells were cultivated in the presence of G-418 geneticin sulfate solution (Hyclone). 500 μg/ml geneticin was used as working solution for selection of stable cell lines. The selection was carried on for 3–4 weeks, while the cells were passaged when confluency was reached. 250 μg/ml geneticin was used as working solution for the maintenance of stable cell lines.

Wound Healing Assay

Huh7 cells stably expressing the corresponding genes were grown to full confluency on a 6-well cell culture plate. A wound is introduced by scratching the confluent monolayer cells with a pipette tip (time = 0). Plates were washed twice with 1X PBS in order to remove detached cells and incubated with complete growth medium. Cell migration into the scraped empty space was followed at 24 h, 48 h and 72 h time intervals using bright field microscopy. At the end of 72 h, the distances across the wound were quantified by using the ImageJ software.

Xenograft experiments

Stable Huh7 cells were cultivated in DMEM without any antibiotics for 48 h before in vivo studies. For tumor formation assay, the confluent cells in 10-cm culture plates were trypsinized and washed twice with PBS. The cell pellets were resuspended in 150 μl PBS and kept on ice until injection. The cells were injected subcutaneously and bilaterally (control on one side) into the abdominal region of 6–14-week-old female SCID mice. Tumors at the injection sites were collected at different days post-injection to determine tumor volume and weight. The isolated tumors were snap-frozen in liquid nitrogen and stored at −80 °C for further use.

Total RNA extraction

Total RNA isolation was performed with GeneJet RNA Purification kit (Thermo) following the manufacturer’s instruction. Adherent cells were washed with PBS, trypsinized and centrifuged at 1600 rpm for 5 min. Cell pellets were dissolved in 600 μl lysis buffer and vortexed for 10 s. Homogenization of the lysate was accomplished by passing through a RNase-free syringe several times. Total RNA concentration was measured by nanodrop and absence of DNA contamination is verified (OD 260/280 ratio > 2). Total RNA integrity was checked by loading the samples on 1% agarose gel and confirming the presence of 18S and 28S ribosomal RNAs. Purified RNA samples were stored at −80 °C until use.

Luciferase Reporter Assay

Luciferase assay was performed according to Dual-Glo Luciferase Assay System (Promega) protocol. For the assay, cells were co-transfected with 1 μg of a pGL3-luciferase reporter plasmid including promoter of interest, 1 μg β-catenin, TCF4 or dNTCF4 expression plasmid and 100 ng of pGL3-SV40-Renilla (internal control) per well of a 12-well plate. Transfection is done by using Turbofect transfection reagent (Thermo) according to the manufacturer’s instructions. About 48 h post-transfection, cells were collected by trypsin, and then washed with 1X PBS twice. Pellet was completely resuspended in 100 μl 1X PBS and placed into 96-well plate. Firefly and Renilla luciferase activities were measured respectively, using a fluorometer. Firefly luciferase readings were normalized to Renilla luciferase readings and graphs were plotted in Microsoft Excel.

Lithium Treatment Assay

Huh7 cells were seeded into 10 cm culture dishes or 6-well plates based on the experiments. Following day, lithium chloride (LiCl) or sodium chloride (NaCl) was added to the medium of the cells to a final concentration of 25 mM. Reverse treatment was employed where all the cells treated at different time points were harvested at the same time (t = 0 h). For protein isolation from the treated cell lines, cell lysis buffer with freshly added protease inhibitors was added on to the cells (100 μl for per well of 6-well plate and 400 μl for 10 cm plate) and the cells were subjected to protein extraction procedure.

Quantitative Polymerase Chain Reaction

Real-Time PCR was performed with gene-specific primers using the Maxima SYBR Green/ROX qPCR kit (Thermo) according to the manufacturer’s protocol. The primers were chosen among the primers of MGH primer bank encompassing intron/exon junctions and blasted against the target genome. The reaction mixtures were distributed into the wells of 96-well qPCR plate and cDNAs were added last. The plate was centrifuged at 2000 rpm for 2 min in order to collect the components to the bottom of the wells. PCR reaction was started with initial denaturation at 95 °C for 10 min. Amplification cycle consists of three steps; a denaturation step at 95 °C for 15 s, an annealing step at 60 °C for 30 s and an elongation step at 72 °C for 30 s. It was repeated for 40 cycles and finished by a melting curve step. For each gene, expression levels were normalized to GAPDH as the control. Relative expression levels were calculated with the 2-ΔΔCt method, comparing the expression level of the gene of interest between the test sample and control sample.

XTT Cell Proliferation Assay

XTT Cell Proliferation Assay Kit II (Roche) was used according to manufacturer’s instruction. The stable Huh7 cells were seeded into 96-well cell culture plates at a concentration of 4 × 10³ cells per well. The cells were incubated at 37 °C incubator for 3 days. At the day of the experiment, the XTT labeling reagent was mixed with an electron-coupling reagent in a 50:1 ratio. 50 μl of XTT labeling mixture was added per well and the cells were incubated with this mixture for 2 h at 37 °C. The spectrophotometric absorbance of the samples at 475 nm were measured in a plate reader. A second measurement was performed at 650 nm as the reference wavelength.

MGAT1 is upregulated in response to Wnt/β-catenin pathway activation

The basal protein levels of MGAT1, along with a few other key proteins of cell proliferation were determined in various cell types, including carcinoma and non-carcinoma cell lines (Fig. 1a). We observed higher protein expression levels of MGAT1 in hepatocellular carcinoma cells (Hep40, Hep3B and HepG2) and colorectal carcinoma cells (HCT116) compared to other cell lines, such as Mahlavu, HEK 293FT and HeLa. It is noteworthy that hepatocellular carcinoma and colon carcinoma are the two cancer types in which the deregulation of Wnt/β-catenin signaling is mostly observed [10].

Lithium chloride (LiCl) is a widely used GSK-3β inhibitor that results in the activation of Wnt/β-catenin signaling. It interferes with the function of GSK-3β degradation complex and prevents β-catenin phosphorylation and degradation [16]. Huh7 cells were treated with LiCl and NaCl (as control) in a time-dependent manner (0 h, 24 h and 48 h). We showed that 24-h and 48-h of LiCl treatment result in increased total β-catenin and non-phosphorylated (active) β-catenin protein levels (Fig. 1b). c-Myc and Heat Shock Factor 2 (HSF2) are the known targets of β-catenin [15, 17]. In our analysis the treatment of LiCl caused an increase in the levels of c-Myc protein at the 24 h and 48 h treatments with LiCl, whereas HSF2 protein accumulated mostly at the 48 h. Also, the level of MGAT1 protein increased at the 24 h and 48 h treatments with LiCl, which is reminiscent of the impact of activated β-catenin pathway in cells (Fig. 1b).

Unlike other hepatocellular carcinoma cell lines, Huh7 cells do not show abnormal activity of Wnt/β-catenin pathway. In order to activate the pathway, Huh7 cells were transfected with the plasmids expressing either the wild-type β-catenin, S33Y-β-catenin or the combination of S33Y-β-catenin and TCF4. The protein levels of MGAT1 were determined by Western Blotting (Fig. 1c). HSF2 and c-Myc proteins are well established targets of Wnt/β-catenin pathway therefore they were used as positive controls. Non-phosphorylated β-catenin specifically recognizes β-catenin only when Serine 45 residue is not phosphorylated or not targeted for degradation; therefore, it indicates the active β-catenin levels. We also observed the higher accumulation of MGAT1 upon overexpression of S33Y-β-catenin, and the combination of S33Y-β-catenin with TCF4 in transfected cells. Likewise, overexpression of either wild-type β-catenin or S33Y-β-catenin led to an increase in the protein levels of both c-Myc and HSF2. It is worth mentioning that c-Myc displayed almost no protein expression in the cells transfected with the empty vector. Taken together, our results clearly demonstrate that MGAT1 protein is concurrently regulated with the activation of β-catenin (Fig. 1c).

MGAT1 mRNA levels increase upon Wnt/β-catenin pathway activation

In order to determine MGAT1 mRNA levels in response to Wnt/β-catenin pathway activation, Huh7 cells were treated with either Wnt agonist or GSK3β inhibitor XII (TWS 119). TWS 119 is a synthetic chemical inhibitor of GSK3β and used commonly for the activation of Wnt/β-catenin signaling [18]. Quantitative real-time PCR was performed to detect MGAT1 mRNA levels. AXIN2, being a widely known target of Wnt/β-catenin pathway, was used as positive control.

The results of q-RTPCR indicate that MGAT1 mRNA levels increased gradually with increasing concentrations of the Wnt agonist. As expected, AXIN2 mRNA was upregulated by Wnt agonist treatment (Fig. 2a). MGAT1 mRNA levels displayed a marked increase only in the cells treated with 10 μM TWS 119, but not in other concentrations. AXIN2 mRNA was also upregulated most in response to the same concentration of TWS 119 (Fig. 2b).

MGAT1 shows increased promoter activity upon β-catenin overexpression.

To get insights into the transcriptional regulation of MGAT1 gene under modulated Wnt/β-catenin pathway activation we performed luciferase reporter assay to monitor the promoter activity of this gene. For positive control of luciferase activity, TOPFLASH vector which contains three copies of wild-type TCF4 binding sites was used. On the other hand, FOPFLASH vector which contains mutant TCF4 binding sites, was used as negative control. The ratio of firefly luciferase to renilla luciferase was determined for each case, and plotted as relative luciferase activity.

Overexpression of the degradation-resistant β-catenin mutant (S33Y-β-catenin) in Huh7 cells resulted in higher luciferase activity from MGAT1 promoter compared to the negative control vector FOPFLASH, or the promoterless luciferase vector pGL3-basic (Fig. 3a).

Furthermore, S33Y-β-catenin was overexpressed in Huh7 cells in combination with either wild-type transcription factor TCF4 or its dominant negative form (dN-TCF4). Relative luciferase activity from MGAT1 promoter was determined together with the activity from the promoterless pGL3-basic vector as the negative control. Overexpression of S33Y-β-catenin together with TCF4 led to more than 8-fold increase in MGAT1 promoter activity; however, when dominant negative form of TCF4 (dN-TCF4) is overexpressed with S33Y-β-catenin, this activity was reduced nearly in half (Fig. 3b).

In order to activate the pathway, three different ligands of the canonical Wnt/β-catenin signaling pathway (Wnt1, Wnt3a and Wnt5) were overexpressed in Huh7 cell lines by transfecting the cells with each of the pLNCX-Wnt1, pLNCX-Wnt3a and pLNCX-Wnt5 plasmids. Relative luciferase activity from the MGAT1 promoter was determined in each case. Overexpression of Wnt1 led to the highest luciferase activity from MGAT1 promoter (approximately 15-fold increase compared to control), whereas overexpression of Wnt3a and Wnt5 also led to more than 10-fold increase in luciferase activity compared to the one obtained from mock transfection (Fig. 3c). As a result of these series of experiments, MGAT1 promoter activity was found to be higher in response to increased Wnt/β-catenin activity in all the different approaches used.

Assessing the migration and proliferation abilities of MGAT1-expressing stable cell lines

Stable expression of MGAT1 and GFP (as negative control) genes in Huh7 cells was established as a first step, in order to analyze the effect of MGAT1 both in vitro and in vivo.

Wound healing assay was performed in order to determine whether there existed any difference in the cell migration capacities of the stable Huh7 cells. MGAT1 and GFP overexpressing Huh7 cells were used for this assay. After introducing a scratch across the confluent and adherent layer of cells in 6-well plates, the cells were visualized under the microscope at different time points starting from 0 h up to 72 h. The widths of the scratches indicated by yellow lines were quantified by Image J software (Fig. 4a).

Differences between wound closure capacities of the stable cells were observed after 72 h. Distances across the wound were measured at each time point, and difference between the initial and final values were computed as percent wound closure for each stable cell line. Overall, MGAT1 overexpressing stable Huh7 cells demonstrated a higher percentage of wound closure (60,4%) compared to the cells overexpressing GFP as the control (42,5%) (Fig. 4b). These results substantiate the fact that MGAT1 overexpressing cells have higher proliferative capacity compared to control cells.

In order to determine whether there is any difference in the cellular proliferation properties of the stable cell lines, we perform XTT Assay. The stable Huh7 cell lines overexpressing MGAT1 and GFP (control) were grown at 96-well culture plates for three days under two different sets of serum starvation conditions (2% FBS and 5% FBS). At the end of three days, absorbance spectra for the cells treated with XTT reagent was recorded at 475 nm using an ELISA plate reader. Normalization was done by taking another set of measurements at the reference wavelength of 650 nm. The results indicate that the relative absorbance values are approximately 30% higher in MGAT1 overexpressing cells compared to the GFP overexpressing cells in both of the serum starvation conditions (Fig. 4c).

In vivo analysis of tumorigenesis

In order to test the tumor forming abilities of the stable Huh7 cell lines in vivo, we performed Xenograft assays in SCID mice. The mice were injected subcutaneously and bilaterally from their flank regions. MGAT1 overexpressing Huh7 cells were injected on the left flank of the mice, while GFP overexpressing Huh7 cells were injected on the right flank as the control (Fig.5a).

The mice were checked for tumor formation regularly. These experiments clearly indicate visible differences in tumor sizes 3–4 weeks after the injection of MGAT1 overexpressing Huh7 cells into the mice. At this point, the mice were sacrificed and tumors were isolated (Fig. 5b). Differences between the tumors were determined by weighting, and the average weights for each set are plotted (Fig. 5c). The tumors formed by MGAT1 overexpressing cells were found to be more than 3 times larger than the control tumors.