Clinical samples

The HCC tissues were collected from 78 patients who agreed to donate their clinical samples at our hospital between 2005 and 2010. This study complies with the Declaration of Helsinki and was performed according to the approval of the Institutional Review Board of the hospital (Approval No. 201712206RINC). Written consent was obtained from the patients. Information on the subjects is provided in Table 1.

Immunohistochemistry (IHC)

The collected tissue slides were incubated with B4GALT1 polyclonal antibody (1:100; Abnova, Taipei, Taiwan) at 4 °C for 16 h. Thereafter, B4GALT1 protein expression was detected using the UltraVision Quanto Detection System (Thermo Scientific, Cheshire, UK). Further, the IHC staining assessment was independently conducted by two pathologists, who were blinded to the patient outcomes.

Cell lines and cell culture

HCC cell lines, SKHEP1 (RRID: CVCL_0525), HepG2 (RRID: CVCL_0027), SNU387 ((RRID: CVCL_0250), and Huh7 (RRID: CVCL_0336) were purchased from Bioresource Collection and Research Center (Hsinchu, Taiwan), while the cell lines Hep3B (RRID: CVCL_0326), HA22T (RRID: CVCL_7046), and PLC5 (RRID: CVCL_0485) were obtained from American Type Culture Collection (Manassas, VA, USA). All the cell lines were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fisher Scientific, Grand Island, NY, USA), which was supplemented with 10% Fetal Bovine Serum (FBS) (Thermo Fisher Scientific), 100 IU/mL penicillin, and 100 μg/mL streptomycin (Thermo Fisher Scientific). The culturing of the cells was performed in a humidified tissue culture incubator with a 5% CO2 atmosphere at 37 °C. All human cell lines have been authenticated using short tandem repeat profiling within three years. All experiments were performed with mycoplasma-free cells.

Transfection and plasmid construction

To overexpress B4GALT1, pcDNA3.1/B4GALT1 plasmid was constructed and cells were transfected using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) as previously described. pcDNA3.1/myc-His (Invitrogen) was used as a control. The insert was confirmed via DNA sequencing. To transiently knock down B4GALT1, three independent siRNA oligonucleotides against B4GALT1 and negative control siRNAs with medium GC content were synthesized by Invitrogen. The siRNAs against B4GALT1 were si-B4GALT1-1: 5′-GAGGCAUGUCUAUAUCUCGCCCAAA-3′, si-B4GALT1-2: 5′-GAAGGACUAUGACUACACCUGCUUU-3′, and si-B4GALT1-3: 5′-CAACAGUUUCUAACCAUCAAUGGAU-3′. Further, for B4GALT1 knockdown, cells were transfected with 20 nmol of siRNA using Lipofectamine RNAiMAX (Invitrogen) for 2 days.

For stable B4GALT1 knockdown, the shB4GALT1/pLKO (TRCN34839) plasmid and non-targeting pLKO (TRC025) plasmid were purchased from National RNAi Core Facility (Academia Sinica, Taipei, Taiwan). Thereafter, short hairpin RNA (shRNA) plasmids were transfected for 48 h and selected using 500 ng/mL of puromycin for 10 days. The stable knockdown of B4GALT1 was then confirmed via western blot analysis.

B4GALT1 knockout in PLC5 cells using the CRISPR/Cas9 system

The CRISPR/Cas9 system was used for B4GALT1 knockout in PLC5 cells. Further, a small guide (sg) RNA for targeting B4GALT1 was designed according to the CRISPR database prediction (http://crispr.mit.edu/). The target sequence of sgB4GALT1 was 5′-CCTGTACGCATTATGGTCATTCA-3′. The success of the B4GALT1 knockout in the genome was confirmed via DNA sequencing.

Mass spectrometric analysis

LC-MS/MS analysis was performed on Orbitrap Fusion Lumos Tribrid quadrupole-ion trap-Orbitrap mass spectrometer (Thermo Fisher Scientific, San Jose, CA). Peptides were separated on Ultimate system 3000 nanoLC system (Thermo Fisher Scientific, Bremen, Germany), which was connected to the mass spectrometer. Peptide mixtures were loaded onto a 75 μm ID, 25 cm length C18 Acclaim PepMap NanoLC column (Thermo Scientific, San Jose, CA, USA) packed with 2 μm particles with a pore of 100 Å. Mobile phase A was performed using 0.1% formic acid in water, and mobile phase B was composed of 100% acetonitrile with 0.1% formic acid. A segmented gradient in 50 min from 2% to 40% solvent B was used with flow rate of 300 nL/min. Mass spectrometric analysis was performed in a data-dependent mode with Full-MS (a resolution of 120,000 at m/z = 200, AGC target 5e5, and maximum injection time of 50 msec). It was followed by high-energy collision-activated dissociation (HCD)-MS/MS of the top 15 most intense ions in 3 seconds. HCD-MS/MS was used to fragment multiply charged ions within a 1.4 Da isolation window (resolution of 15,000) at a normalized collision energy of 32. AGC target 4e4 was set for MS/MS analysis with previously selected ions dynamically excluded for 60 seconds.

For protein identification, the raw MS/MS data were searched using the Mascot and SEQUEST search algorithm via the Proteome Discoverer (PD) package (version 2.3, Thermo Scientific). The search parameters were set as follows: peptide mass tolerance, 10 p.p.m.; MS/MS ion mass tolerance, 0.02 Da. Peptides were filtered based on a 1% FDR.

Antibodies and reagents

Antibody against B4GALT1 was obtained from Abnova (Cat#PAB20512), while that against integrin β1 (CD29) was purchased from BD Transduction Laboratories (Cat#610468). Further, antibody against integrin α6 was obtained from Cell Signaling Technology Inc. (Cat#3750), while that against GAPDH was purchased from Meridian Life Science. (Cat#H86504M). Functional blocking antibodies for integrin β1 (P4C10) were obtained from Merck Millipore (Cat#MAB1987Z). Further, functional blocking antibodies for integrin α6 (CD49f) were purchased from Invitrogen (Cat#12-0495-82), and human collagen I, human collagen IV, human fibronectin, laminin, or bovine serum albumin (BSA) were purchased from Sigma Aldrich (St Louis, MO, USA). GSL-II-FITC, LEL-FITC, RCA-I-FITC, ECL-FITC, MAL-II-FITC, and PNA-FITC were purchased from Vector Laboratories (Burlingame, CA, USA).

Western blot analysis

Proteins were separated via SDS-PAGE, and then transferred onto PVDF membranes. The membrane blots then were blocked with 5% milk in TBST for 1 h at room temperature, and thereafter, incubated with primary antibodies overnight at 4 °C. The blots were then incubated with a corresponding secondary antibody conjugated with horseradish peroxidase, and signals were detected using ECL reagents and X-ray films.

Experimental metastasis model in NOD/SCID mice

Female NOD/SCID mice (4 weeks old; LASCO, Taipei, Taiwan) were injected via the tail vein with 1 × 106 HCC cells. After 60 days, the mice were sacrificed and metastatic nodules were visualized. Specifically, after sacrifice, the lungs were paraffin-embedded for hematoxylin and eosin (H&E) staining. All mice used in this study were maintained under specific pathogen-free conditions and housed in pathogen-free facilities in a 12 h light/dark cycle with ad libitum access to food and water. This animal study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of National Taiwan University College of Medicine (Approval No. 20170405).

MTT assay

HCC cells in 100 μL of complete DMEM were seeded in 96-well plates at a cell density of 3 × 103 cells per well. Thereafter, 10 μL of 5 mg/ml 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide solution (MTT, Sigma) were added to each well for the indicated times followed by incubation at 37 °C for 4 h, after which 100 μL of 10% SDS in 0.01 N HCl was added to dissolve the MTT formazan crystals. The resultant optical density was then measured spectrophotometrically at dual wavelengths (550 and 630 nm).

Transwell migration and Matrigel invasion assays

Transwell migration assays were performed using 6.5-mm polycarbonate transwell filters with an 8-μm pore size (Corning Costar Corp., Corning, NY, USA). In brief, approximately 3 × 104 cells (HA22T) or 2 × 105 cells (PLC5) in 200 μL serum-free DMEM were seeded into the upper surface of the transwell chamber, while 500 μL of 10% FBS in complete DMEM was loaded into the lower chamber of 24-well plates. Thereafter, BioCoat Matrigel invasion chambers (BD PharMingen, San Diego, CA, USA) were used for the cell invasion assay. The cells were allowed to migrate toward the transwell chamber or invade the Matrigel for 24 h. The migrated and invaded cells we then fixed with 100% methanol and stained with 0.5% (w/v) crystal violet (Sigma-Aldrich). Thereafter, the numbers of migrated and invaded cells per field were counted in at least three independent experiments (mean ± SD). For integrin functional blocking experiments, 100 μg/ml of blocking antibodies against integrin β1, integrin α6, or control mouse IgG were treated for 30 min before seeding into the upper chamber of transwells. Cells were then subjected to migration and invasion assays.

Flow cytometry

Cell surface protein expression was analyzed using the FACScan cytometer (BD PharMingen). In brief, HCC cells were detached using 5 mM EDTA and resuspended in 2% BSA/PBS. Thereafter, the cells were incubated with FITC-conjugated lectins at 1∶100 dilutions on ice for 30 min. Next, the cells were washed three times with ice-cold 2% BSA/PBS, and the fluorescence intensity of 1 × 105 cells for each sample was analyzed.

Lectin pull-down assay

For the lectin pull-down assay, cell lysates (1 mg) were incubated with Griffonia Simplicifolia Lectin II (GSL-II) biotinylated (Vector Laboratories) at 4 °C for 16 h. The next day, the sample was incubated with streptavidin beads (Vector Laboratories) for 2 h at room temperature. Finally, the pull-down proteins were analyzed via western blot analysis.

Cell adhesion assay

Cell adhesion assays were performed as previously reported [17]. In brief, 96-well plates were coated with BSA, fibronectin, laminin, collagen I, or collagen IV at concentrations of 2.5 μg/mL in PBS at 37 °C for 16 h, and then blocked with 1% BSA in PBS at 37 °C for 4 h. Next, the HCC cells were detached using 5 mM EDTA and 2 × 104 cells in 100 μL serum-free DMEM were seeded and allowed to attach for 30 min at 37 °C in a humidified 5%-CO2 incubator. Finally, the number of adherent cells in three wells were determined using an inverted microscope.

Statistical analyses

Statistical analyses were performed using GraphPad Prism software version 7 (GraphPad Software Inc., San Diego, CA, USA). The correlations between B4GALT1 expression and the clinicopathological characteristics of HCC were tested by performing chi-square tests. Further, survival curves were constructed using the Kaplan–Meier method and the Student’s t-test was performed to compare differences between two groups of quantitative variables. Data were presented as means ± SD or number (percentage), and statistical significance was set at p < 0.05.