Introduction

Oncology Reports

1021-335X 1791-2431

D.A. Spandidos

10.3892/or.2020.7795

or-44-06-2757

Articles

Analysis of the correlation between alterations in N-glycans and invasiveness in liver cancer cell lines

Takahashi

Hidenori

1 Kamiyama

Toshiya

1 Hirane

Nozomi

2 Kobayashi

Nozomi

1 Aiyama

Takeshi

1 Nagatsu

Akihisa

1 Shimada

Shingo

1 Orimo

Tatsuya

1 Kakisaka

Tatsuhiko

1 Fukai

Moto

1 Yokoo

Hideki

1 Kamachi

Hirofumi

1 Nishimura

Shin-Ichiro

2 Taketomi

Akinobu

1Department of Gastroenterological Surgery I, Hokkaido University Graduate School of Medicine, Sapporo, Hokkaido 060-8638, Japan 2Graduate School of Life Science and Frontier Research Center for Post-Genome Science and Technology, Hokkaido University, Sapporo, Hokkaido 001-0021, Japan

Correspondence to: Dr Toshiya Kamiyama, Department of Gastroenterological Surgery I, Hokkaido University Graduate School of Medicine, North 5, West 7, Kita-ku, Sapporo, Hokkaido 060-8638, Japan, E-mail: t-kamiya@med.hokudai.ac.jp

122020

08102020

44 6 2757 2769 23032020 31072020

2020

The N-glycoforms of glycoproteins modify protein function and control a number of biological pathways. The aim of the present study was to investigate the correlation between alterations in N-glycans and cancer aggressiveness in terms of cancer cell invasion ability. The expression of urokinase-type plasminogen activator (uPA) and N-acetylglucosaminyltransferase V (GnT-V) in liver cancer cell lines was analyzed by western blotting. Cell invasiveness was analyzed by Matrigel invasion assays. uPA and GnT-V expression in liver cancer cell lines was knocked down by RNA interference. Furthermore, uPA was overexpressed in liver cancer cells using lentiviral vectors, and a mutant strain of HepG2 cells overexpressing uPA deficient in N-glycans was established. A glycoblotting-assisted matrix-assisted laser desorption/ionization-time-of-flight/mass spectrometry-based quantitative analysis of liver cancer cell lines was performed, in which invasiveness was altered by modifying the expression of uPA and GnT-V. N-glycan profiles were found to differ between the highly invasive liver cancer cell line HLE and the less invasive cell line HepG2. The expression of several N-glycans, including a form with m/z=1892, was changed according to invasiveness controlled by knockdown and overexpression of uPA. The invasiveness of HepG2 cells with mutant uPA did not increase regardless of the level of expression of uPA. Following GnT-V knockdown and N-glycan alteration, uPA expression did not change, whereas cell invasiveness decreased. One N-glycan (m/z=1892) was common among N-glycans in the comparative analysis between HLE and HepG2, HLE and uPA knockdown HLE, HepG2 and uPA-overexpressing HepG2, and HLE and GnT-V knockdown HLE cells and among N-glycan profiles in human uPA. Therefore, N-glycosylation is an important factor controlling invasiveness of liver cancer cells, and a specific N-glycan (m/z=1892) associated with the invasion of liver cancer cells via uPA was identified in the present study.

glycomics liver cancer invasion urokinase-type plasminogen activator N-acetylglucosaminyltransferase V

Introduction

Glycosylation is one of the most common post-translational protein modifications. During the process of glycosylation, sugars are attached to proteins or lipids and cause immense structural and functional variations in the majority of eukaryotic cell proteins (1,2). N-glycans modify proteins and control several biological pathways (3). Alterations in N-glycan profiles have been suggested to play important roles in the proliferation, differentiation, invasion and metastasis of malignant cells (4). Glycan species may be analyzed and characterized using mass spectrometry (MS), and profiling of these molecules when they are secreted or shed from cancer cells is also performed (5,6). A novel glycomics method, SweetBlot, which facilitates high-throughput and large-scale glycome analysis using an automated glycan purification system, has been developed (7). This approach enables quantitative profiling of serum N-glycans. Using this quantitative N-glycomics procedure based on glycoblotting technology, which is both highly accurate and can be conducted on a large scale, specific N-glycan alterations were reported in the sera of 369 patients with liver cancer. These specific alterations were independent risk factors for shorter overall and disease-free survival (8). These alterations were also correlated with invasiveness based on analyses of clinicopathological factors. Specific glycosylation patterns related to the malignancy of human brain tumors at different stages were also reported (9). Altered N-glycan profiles may affect the malignancy of several types of cancer (10–14). Compared with normal cells, cancer cells frequently express N-glycans at different levels and with fundamentally different structures; therefore, elucidation of these glycosylation patterns may provide a novel approach to cancer therapy.

The invasiveness of cancer cells is a major factor determining malignancy. Urokinase-type plasminogen activator (uPA) and N-acetylglucosaminyltransferase V (GnT-V) are important factors associated with cancer malignancy. uPA is a serine protease that converts plasminogen to plasmin. Through this enzymatic effect, uPA plays an important role in fibrinolysis, which mediates cancer cell invasiveness (15). uPA was found to be significantly upregulated in liver cancer tissues in comparison with normal liver tissues in patients with liver cancer (16). Another study suggested that uPA activity was the most important factor affecting liver cancer invasion in the tissue fibrinolysis system controlled by uPA and a strong predictor of liver cancer recurrence (17). It was also reported that the concentration of uPA is correlated with invasiveness in liver cancer cell lines, and that controlling uPA expression makes it possible to interfere with invasiveness (18). GnT-V is a glycosyltransferase that forms β1,6 branches. β1,6-branching also affects cellular pathways involved in adhesion, motility, angiogenesis and apoptosis. Thus, GnT-V expressed on cancer cells acts as a master regulator of metastasis (19,20) and affects malignancy in several types of cancer (21). GnT-V also affects the expression and activation of matriptase and uPA (22–24).

However, the association between alterations in N-glycans and cancer malignancy in terms of invasiveness is incompletely understood. The aim of the present study was to investigate the correlation between invasiveness and alterations in N-glycans using liver cancer cell lines with modified uPA and GnT-V expression.

Materials and methods <sec> <title>Cell culture

The human liver cancer cell lines HLE and HepG2 were obtained from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan) and they were authenticated by short tandem repeat profiling. The cells were cultured in DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 100 U/ml penicillin and streptomycin (Gibco; Thermo Fisher Scientific, Inc.) and 10% FBS at 37°C under 5% CO₂. The cells were harvested using trypsin and pelleted by centrifugation (1,500 × g, 3 min, 4°C).

Establishment of uPA and GnT-V knockdown HLE cells

Stealth siRNA transduction particles targeting uPA (PLAUHSS108076, forward sequence, 5′-GCCCUCCUCUCCUCCAGAAGAAUUA-3′ and reverse, 5′-UAAUUCUUCUGGAGGAGAGGAGGGC-3′; and PLAUHSS108078, forward sequence, 5′-CCAUCCCGGACUAUACAGACCAUCU-3′ and reverse, 5′-AGAUGGUCUGUAUAGUCCGGGAUGG-3′) and GnT-V (MGAT5HSS106510, forward sequence, 5′-GAAAGCGGAAGAAAGUCCUCGUUCA-3′ and reverse, 5′-UGAACGAGGACUUUCUUCCGCUUUC-3′; and MGAT5HSS181096, forward sequence, 5′-CGCUGGAGUCAUGACAGCUUAUGAU-3′ and reverse, 5′-AUCAUAAGCUGUCAUGACUCCAGCG-3′) and negative control (Stealth™ RNAi Negative Control Duplex) were purchased from Thermo Fisher Scientific, Inc. siRNAs were transfected into HLE cells with Lipofectamine RNAiMAX Reagent (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Approximately 30×10⁴ cells in 10% FBS-supplemented medium were seeded in a 10-cm dish at 37°C under 5% CO₂. Following incubation for 24 h, stealth siRNA (10 nM) was added and cells were transfected for 48 h.

Establishment of uPA-overexpressing cells

Whole uPA primer, a primer for amplifying human uPA (forward, 5′-TCGACTCGAGATGAGAGCCCTGCTGGC-3′, reverse, 5′-TCGACTCGAGTCAGAGGGCCAGGCCATTC-3′, Thermo Fisher Scientific, Inc.), was designed based on the uPA gene sequence (1,296 bp). The cDNA of uPA was amplified by reverse transcription-PCR (RT-PCR) with the cDNA of HLE cells as template, whole uPA primer and QIAquick Spin (Qiagen GmbH). The cDNA of uPA was cloned into the pLVSIN-IRES-ZsGreen1 vector (Takara Bio, Inc.) according to the manufacturer's instructions. Lentiviral supernatants were produced using the Lenti-X 293T Packaging System (Takara Bio, Inc.) and used for the transduction of HepG2 cells. HepG2 cells were transduced with supernatants containing the empty vector for the negative control. The insert was checked by sequencing with a Big Dye Termination Cycle Sequencing kit (Applied Biosystems; Thermo Fisher Scientific, Inc.). Fluorescence-activated cell sorting was performed twice to select stable clones.

Establishment of uPA-mutant HepG2 cells

uPA has a glycosylation site including asparagine at the 302nd amino acid residue. This asparagine was replaced with glutamine by a substitution point mutation with uPA mutation primer (5′-CCGGATAGAGATAGTCGGTAGATTGCTCTTTTCC-3′, Thermo Fisher Scientific, Inc.) and QIAprep Spin Mini Kit (Qiagen GmbH). The mutant plasmid was transfected into HepG2 cells, which resulted in the production of uPA without N-glycan modification, as the N-glycan is bound to the amide nitrogen atom on the side chain of asparagine.

Western blot analysis

Cellular proteins were extracted using the Subcellular Protein Fractionation Kit for Cultured Cells (Pierce; Thermo Fisher Scientific, Inc.). Samples were mixed with SDS sample buffer (Bio-Rad Laboratories, Inc.), boiled at 95°C for 5 min, separated by SDS-PAGE, and transferred to PVDF membranes (GE Healthcare). The membranes were immunoblotted using primary antibodies against uPA (goat monoclonal anti-human, cat. no. sc-19088, diluted 1:750; Santa Cruz Biotechnology, Inc.), GnT-V [rabbit polyclonal anti-human, cat. no. EP6274 (2), diluted 1:200; Abcam] and GAPDH (rabbit monoclonal anti-human, cat. no. D16H11, diluted 1:2,000; Cell Signaling Technology, Inc.). The antibodies were diluted with 1X Tris-buffered saline with 0.1% Tween-20. The secondary antibodies were diluted 1:5,000 for uPA and GAPDH and 1:2,000 for GnT-V.

Cell invasion assay

Invasion assays were performed in 24-well Matrigel invasion chambers (Corning, Inc.) according to the manufacturer's instructions. Medium containing 10% FBS was added to the lower chambers. Approximately 3×10⁴ HLE cells or 15×10⁴ HepG2 cells in FBS-free medium were seeded in the upper chambers. After 48 h, cells on the upper chamber were removed with a cotton swab. The membranes were fixed with 99% methanol for 2 min and stained with staining solutions I and II for 2 min each at room temperature using the Diff-Quik staining kit (Sysmex Corporation), and the cells were counted at a magnification of ×20 using a fluorescence microscope (Keyence Corporation).

Extraction of glycoproteins

For N-glycan analysis, glycoproteins were extracted as described previously (25). Briefly, cell pellets consisting of 1×10⁶ cells were homogenized using an Ultrasonic Homogenizer (Taitec Corporation) in 100 mM Tris-acetate buffer (100 µl; pH 7.4) supplemented with 2% SDS as a surfactant for the complete dissolution of cell pellets. Reductive alkylation of the cellular proteins was performed by the addition of 500 mM tris(2-carboxyethyl)phosphine (Sigma-Aldrich; Merck KGaA) at room temperature for 60 min, followed by the addition of 200 mM iodoacetamide (Sigma-Aldrich; Merck KGaA) at room temperature for 30 min. After reductive alkylation, ethanol precipitation was carried out by adding a 4-fold volume of cold ethanol and incubating for 3 h at −30°C. Supernatants and precipitated proteins were separated by centrifugation at 20,000 × g for 10 min at 4°C, and the precipitates were again washed with cold ethanol. Collected precipitates containing glycoproteins/N-glycans were dried at 37°C for 10 min.

For urokinase from human urine (P.N.: U0633-25UG, Lot: SLBN9831V, Sigma-Aldrich; Merck KGaA), the sample was dissolved in 100 mM solution of ammonium bicarbonate/0.1% Triton X-100/1% SDS containing 120 mM 1,4-dithiothreitol (10 µl). After incubation at 60°C for 30 min, 123 mM iodoacetamide (10 µl) was added to the mixture, followed by incubation at room temperature to enable reductive alkylation. After 60 min, the mixture was treated with 10 µl trypsin (Sigma-Aldrich; Merck KGaA) at 37°C overnight, followed by heat inactivation of the enzyme at 90°C for 10 min. After cooling to room temperature, the mixture was treated with 10 µl trypsin (Sigma-Aldrich; Merck KGaA) at 37°C overnight.

Pre-treatment of uPA

For urokinase from human urine (P.N.: U0633-25UG, Lot: SLBN9831V, Sigma-Aldrich; Merck KGaA), the sample was dissolved in 100 mM solution of ammonium bicarbonate/0.1% Triton X-100/1% SDS containing 120 mM 1,4-dithiothreitol (10 µl). After incubation at 60°C for 30 min, 123 mM iodoacetamide (10 µl) was added to the mixture, followed by incubation at room temperature to enable reductive alkylation. After 60 min, the mixture was treated with 5 µG of trypsin (FUJIFILM Wako Pure Chemical Corporation) at 37°C overnight, followed by heat inactivation of the enzyme at 90°C for 10 min. After cooling to room temperature, the mixture was treated with 100 U of PNGase F from Flavobacterium meningosepticum (New England BioLabs Japan, Inc.) at 37°C overnight. Then, the sample mixture was dried up by SpeedVac (Thermo Fisher Scientific, Inc.).

Glycoblotting analysis Experimental procedures: Cell culture N-glycomics based on glycoblotting

N-glycans from samples of cell culture and urokinase from human urine were purified by glycoblotting using BlotGlyco H, which are commercially available synthetic polymer beads with high-density hydrazide groups (Sumitomo Bakelite Co., Ltd.). All procedures used the SweetBlot automated glycan purification system containing a 96-well plate platform (System Instruments Co., Ltd.).

N-glycan purification and modification by glycoblotting

Glycoblotting of sample mixtures containing urokinase from human urine was performed in accordance with previously described procedures. Commercially available BlotGlyco H beads (250 µl; 10 mg/ml suspension; Sumitomo Bakelite Co., Ltd.) were aliquoted into the wells of a MultiScreen Solvinert hydrophilic polytetrafluoroethylene 96-well filter plate (EMD Millipore). After removal of water using a vacuum pump, PNGase F-digested samples were applied to the wells, followed by the addition of 20 µl of internal standard (Sialylated glycan 2.5 µM) and 180 µl of 2% acetic acid in acetonitrile. The filter plate was then incubated at 80°C for 45 min to capture the N-glycans onto the beads by chemically stable and reversible hydrazine bonds. The beads were then washed using 200 µl of 2M guanidine-HCl in 10 mM ammonium bicarbonate, followed by washing with the same volume of water and 1% triethylamine in methanol. Each washing step was performed twice. The N-glycan-linked beads were then incubated with 10% acetic anhydride in methanol for 30 min at room temperature so that unreacted hydrazide groups would become capped by acetylation. After capping, the reaction solution was removed under vacuum, and the beads were serially washed with 200 µl of 10 mM HCl, methanol and dioxane (this is a pretreatment for sialic acid modification). On-bead methyl esterification of carboxyl groups in sialic acid was carried out with 100 µl of 20 mM 3-methyl-1-p-tolyltriazene (Tokyo Chemical Industry Co., Ltd.) in dioxane at 60°C for 90 min to dryness. After methyl esterification of the more stable glycans, the beads were serially washed in 200 µl of dioxane, water, methanol and water. The captured glycans were then subjected to a transiminization reaction with 20 µl of aminooxy-functionalized peptide reagent (aoWR; 20 mM in Milli Q; Sumitomo Bakelite Co., Ltd.) in 180 µl 2% AcOH/CAN solution for 45 min at 80°C. After this reaction, WR-tagged glycan was eluted in 100 µl of water and purified by Mass PREPTM HILIC µEluent Plate (Waters Corporation); the recovered sample was dried up under vacuum.

Matrix-assisted laser desorption/ionization (MALDI)-time-of-flight (TOF) MS analysis

The N-glycans purified by glycoblotting were diluted with 5 µl of 1% AcOH in 5% ACN/water. A total of 1 µl of sample was mixed with 1 µl of 2,5-dihydroxy benzoic acid (FUJIFILM Wako Pure Chemical Corporation) as ionic liquid matrices, and spotted onto a MALDI target plate. The analytes were then subjected to MALDI-TOF MS analysis using an Ultraflex time-of-flight mass spectrometer III (Bruker Daltonics, Inc.) in a reflector, using the positive-ion mode and typically accumulating 2,000 shots. The N-glycan peaks in the MALDI-TOF MS spectra were selected using FlexAnalysis v. 3 (Bruker Daltonics, Inc.). The glycan structures were estimated using the GlycoMod Tool (http://br.expasy.org/tools/glycomod/).

Statistical analysis

One-way ANOVA with Tukey's post hoc test was performed for experiments that involved morethan two groups, and Student's t-test was performed for comparisons between two groups. Statistical analyses were performed using JMP Pro 12.0.1 for Windows (SAS Institute, Inc.). P<0.05 was considered to indicate statistically significant differences.

Results <sec> <title>uPA expression, invasiveness and glycoblotting analysis of HLE and HepG2 cells

Cell invasion assays revealed that HLE cells exhibited higher invasiveness compared with HepG2 cells (Figs. 1B and 2). Western blot analysis was performed to examine the expression of uPA in liver cancer cell lines. The expression of uPA in the highly invasive HLE cell line was higher compared with that in the less invasive HepG2 cell line (Fig. 1A). Glycoblotting analysis was performed in HLE and HepG2 cells to examine the correlation between invasiveness and N-glycan expression. In this assay, 75 N-glycans were identified, and the concentrations of 10 N-glycans (m/z=1851, 1892, 2029, 2054, 2070, 2216, 2521, 2887, 3192 and 3252) were higher in HLE compared with HepG2 cells (Fig. 3).

uPA expression, invasiveness and glycoblotting analysis of uPA knockdown HLE cells

The association between uPA expression and invasiveness was examined. uPA knockdown HLE cell lines were established using two specific siRNAs against uPA (HLE uPA siRNA1 and HLE uPA siRNA2) and a control HLE cell line (HLE control). Western blot analysis demonstrated that the expression of uPA was decreased in HLE uPA siRNA1 and HLE uPA siRNA2 cells compared with that in HLE control cells (Fig. 4A). Matrigel invasion assays demonstrated that the invasiveness of HLE uPA siRNA1 and HLE uPA siRNA2 cells was significantly lower compared with that of HLE control cells (P<0.01; Figs. 4B and 5). By glycoblotting analysis, 86 N-glycans were identified in the uPA knockdown HLE and HLE control cells. The concentrations of 14 N-glycans (m/z=1502, 1664, 1826, 1988, 2150, 2312, 2474, 1810, 1542, 1689, 1705, 1851, 1892 and 2095) decreased by >20% in HLE uPA siRNA1 and HLE uPA siRNA2 cells compared with those in HLE control cells (Fig. 6).

uPA expression, invasiveness and glycoblotting analysis of uPA-overexpressing HepG2 cells

A HepG2-uPA cell line (HepG2uPAWT) that overexpressed uPA was constructed using a lentivirus vector and a HepG2 control cell line (HepG2control) using an empty vector. A HepG2-uPA cell line (HepG2uPAMUT) that overexpressed a mutant uPA protein and had glutamine as the 302nd amino acid residue instead of asparagine to prevent N-glycosylation was also constructed. Western blot analysis demonstrated that the expression of uPA was increased in HepG2uPAWT and HepG2uPAMUT cells compared to that in HepG2control cells (Fig. 7A). HepG2uPAWT cells expressed higher levels of uPA compared with HepG2uPAMUT cells. The invasiveness of HepG2uPAWT cells was significantly increased compared with that of HepG2control cells (P<0.01), whereas the invasiveness of HepG2uPAMUT cells deficient in N-glycans did not increase (Figs. 7B and 8). By glycoblotting analysis, 72 N-glycans were identified in the uPA-overexpressing HepG2 and HepG2control cells. The concentrations of 22 N-glycans (m/z=1502, 2312, 2474, 1543, 1689, 1705, 1851, 1982, 2010, 2095, 2156, 2172, 2334, 2375, 2512, 2668, 2681, 2741, 2827, 2887, 3192 and 3497) increased by >20% in HepG2uPAWT cells compared with those in HepG2control cells (Fig. 9). These 22 N-glycans were also increased in HepG2uPAMUT cells compared with those in HepG2control cells.

uPA expression, invasiveness and glycoblotting analysis of GnT-V knockdown HLE cells

The association between N-glycan alterations and invasiveness was next examined. GnT-V knockdown HLE cell lines were established using specific siRNAs against GnT-V (HLE GnT-V KD) and a control HLE cell line (HLE control). GnT-V knockdown was verified by western blot analysis and it was observed that uPA expression did not change in HLE GnT-V KD and HLE control cells (Fig. 10A). Matrigel invasion assays demonstrated that the knockdown of GnT-V significantly reduced invasiveness in HLE cells (P<0.05), despite the lack of change in uPA expression (Figs. 10B and 11).

By glycoblotting analysis, 97 N-glycans were identified in GnT-V knockdown HLE and HLE control cells. The concentrations of 18 N-glycans (m/z=1892, 2403, 2563, 2681, 2827, 2887, 2928, 3030, 3192, 3351, 3395, 3411, 3497, 3557, 3656, 3716, 3862 and 3922) decreased by >20% in HLE GnT-V KD cells compared with those in HLE control cells (Fig. 12).

N-glycan profiles in human uPA

A total of 16 N-glycans, 1746, 1892, 1908, 2038, 2054, 2070, 2095, 2152, 2172, 2299, 2445, 2459, 2501, 2563, 2604 and 2705, were identified by glycoblotting analysis (Figs. 13 and 14).

Comparing N-glycans in the analysis of HLE/HepG2, uPA knockdown HLE, uPA-overexpressing HepG2, GnT-V knockdown HLE cells and N-glycan profiles in human uPA

By comparing N-glycans in the analysis of HLE/HepG2 cells, 75 N-glycans were identified, and the concentrations of 10 N-glycans were higher in HLE cells compared with those in HepG2 cells. The concentrations of 13 N-glycans decreased by >20% in HLE uPA siRNA1 and HLE uPA siRNA2 cells compared with those in control HLE cells. The concentrations of 22 N-glycans increased by >20% in HepG2uPAWT cells compared with those in HepG2control cells. These 22 N-glycans were also increased in HepG2uPAMUT cells. The concentrations of 18 N-glycans decreased by >20% in HLE GnT-V KD cells compared with those in HLE control cells. A total of 16 N-glycans were identified by glycoblotting analysis of human uPA. One common N-glycan (m/z=1892) was correlated with invasiveness based on glycoblotting analysis of HLE/HepG2, uPA knockdown HLE, uPA-overexpressing HepG2 and GnT-V knockdown HLE cells.

Discussion

To investigate the variations in the N-glycome profile according to the invasiveness of liver cancer cell lines, a glycoblotting-assisted MALDI-TOF/MS-based quantitative analysis was performed. N-glycan profiles were different between the highly invasive liver cancer cell line HLE and the less invasive HepG2 cell line. The expression of several N-glycans was changed according to cell invasiveness controlled by the knockdown and overexpression of uPA. The invasiveness of HepG2uPAMUT cells deficient in N-glycans did not increase regardless of the level of uPA expression. After GnT-V knockdown and N-glycan alterations, uPA expression did not change, whereas cell invasiveness decreased. Of note, one N-glycan (m/z=1892), was common in the comparative analysis among liver cancer cell lines with different invasiveness, and among N-glycan profiles in human uPA. In liver cancer cells, N-glycosylation was one of the important factors controlling invasiveness, and a specific N-glycan (m/z=1892) that was associated with the invasiveness of liver cancer cells via uPA was identified in the present study.

uPA is a mosaic glycoprotein composed of a COOH-terminal chymotrypsin-like serine protease domain and a modular non-catalytic NH₂-terminal region that mediates the specific cofactor activities of these proteins, exemplified by the cellular binding of uPA (26). The modular composition of the amino-terminal fragment of uPA includes an epidermal growth factor-like module and a kringle module, with the former being responsible for the specific cell binding of uPA (27). It has been reported that uPA was also correlated with invasion and growth in liver cancer cell lines (18). In the present study, the expression level of uPA in the highly invasive liver cancer cell line HLE was higher compared with that in the less invasive cell line HepG2, and the N-glycan profiles were different between these two cell lines. The invasiveness of uPA knockdown HLE cells decreased, and N-glycan profiles were different between HLE and uPA knockdown HLE cells. The invasiveness of uPA-overexpressing HepG2 cells increased, and the N-glycan profiles were also different between HepG2 and uPA-overexpressing HepG2 cells. These data demonstrated that the differences in invasiveness among different liver cancer cell lines may be attributed to the glycosylation changes. Moreover, a specific N-glycan (m/z=1892) was common among N-glycans in the comparative analysis between HLE and HepG2, HLE and uPA knockdown HLE, HepG2 and uPA-overexpressing HepG2, and HLE and GnT-V knockdown HLE cells. In addition, the N-glycan profiles of human uPA were evaluated, and they all shared one common N-glycan (m/z=1892). Therefore, N-glycan (m/z=1892) appears to play an important role in cell invasiveness controlled by uPA. These glycomic alterations may prove to be useful for the evaluation of tumor malignancy.

The glycosylation patterns of uPA have been previously analyzed (28,29). Pro-uPA is a glycoprotein containing 411 amino acids with 24 cysteine, 22 arginine and 27 lysine residues (28). N-glycosylation of human uPA occurs at Asn302 in the protease domain (30). The N-glycans of uPA contain Man, Gal, Fuc, GlcNAc and NeuSAc, as well as GalNAc residues (31). In the present study, the overexpressed mutant uPA protein, which had glutamine as the 302nd amino acid residue instead of asparagine and was deficient in N-glycans, altered the invasiveness of liver cancer cells. Therefore, the glycosylation site at asparagine in the 302nd amino acid residue of uPA was found to be an important factor in controlling cancer cell invasiveness, although the glycosylation pattern of uPA does not affect its interaction with plasma proteins directly involved in its fibrinolytic function (32); additionally, the fibrinolytic activity of recombinant non-glycosylated single-chain uPA has been demonstrated to be similar to that of the glycosylated urinary and recombinant proteins (33).

GnT-V, encoded by the MGAT5 gene (34), is involved in the formation of a β1,6-branched antenna in complex-type glycans. GnT-V is the strongest enzyme associated with cancer progression and metastasis among all glycosyltransferases involved in the formation of multiantennary N-glycans (35–37). The expression of GnT-V was found to be correlated with metastasis in resected colorectal cancer specimens by immunohistochemical analysis (38). Formation of multiantennary N-glycans by GnT-V has been observed in excised liver cancer tissues (39). In addition, increased levels of GnT-V have been observed on immunohistochemical analysis of liver cancer tissues (40). The GnT-V levels were also found to be correlated with the metastatic potential of gastric (41), endometrial (42) and mucinous ovarian cancer (43). Matriptase is a serine protease that plays important roles in cell migration, extracellular matrix degradation, and the activation of uPA and hepatocyte growth factor (22,23). It was also reported that GnT-V mediated metastasis via the upregulation and increased stabilization of matriptase by adding β1-6GluNAc branching (44,45). However, in the present study, knockdown of GnT-V altered N-glycans, including specific N-glycans (m/z=1892) bound to uPA, and reduced the invasiveness of HLE cells without altering uPA expression. These data demonstrated that GnT-V modified N-glycan-binding to uPA, and N-glycosylation is necessary for uPA-mediated cancer cell invasiveness. Therefore, knocking down GnT-V may reduce the invasiveness of HLE cells through N-glycan alterations, in addition to the activation of uPA by matriptase, which was also upregulated by GnT-V. Moreover, N-glycan (m/z=1892) was revealed to be specific among N-glycan alterations when comparing HLE and GnT-V knockdown HLE cells, and was common among N-glycan alterations modulating uPA expression. It was recently reported that the spatial accessibility of N-glycans of substrate glycoproteins to the GnT-V active site likely defines the potential for site-specific modification (46).

There was a limitation to the present study: The expression of the GnT-V protein in HepG2 cells and overexpression of GnT-V in HepG2 cells was not evaluated. If the expression of GnT-V in HepG2 cells is low, overexpression of GnT-V in HepG2 cells and detection of the corresponding indicators may further support the conclusions of the present study. This point will be the focus of a future study.

In conclusion, N-glycan alterations occurred in response to changes in the cancer cell invasiveness triggered by uPA and GnT-V, and a specific N-glycan (m/z=1892) was found to be associated with this change in invasiveness. The function of uPA was controlled by N-glycan alterations mediated by GnT-V, even if not directly through the effect of uPA activation by matriptase. Therefore, glycomic alterations may be useful for the diagnosis of malignancy, and a specific N-glycan (m/z=1892) may be a candidate as a novel biomarker and target for molecular therapy.

Acknowledgements

The authors would like to thank the staff of Department of Gastroenterological Surgery I, Hokkaido University Graduate School of Medicine and Graduate School of Life Science and Frontier Research Center for Post-Genome Science and Technology, Hokkaido University, for their kind cooperation.

Funding

The present study was supported by JSPS KAKENHI (grant no. JP 16K10561).

Availability of data and materials

The datasets generated and/or analyzed during the present study are available from the corresponding author on reasonable request.

Authors' contributions

ToK made substantial contributions to conception and design. HT, NK and TA analyzed and interpreted the data regarding the alterations in N-glycans and invasiveness in liver cancer cell lines. NH performed the histological examination of the kidney. AN, SS, TO, TaK, MF, HY, HK, SN and AT helped supervise the findings of this study. HT and ToK were major contributors to writing the manuscript. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References 1

Pilobello

Mahal

Deciphering the glycocode: The complexity and analytical challenge of glycomics

Curr Opin Chem Biol113003052007

10.1016/j.cbpa.2007.05.002

17500024

Ohtsubo

Marth

Glycosylation in cellular mechanisms of health and disease

Cell1268558672006

10.1016/j.cell.2006.08.019

16959566

Helenius

Aebi

Intracellular functions of N-linked glycans

Science291236423692001

10.1126/science.291.5512.2364

11269317

Lin

Zhao

Wang

Han

Pan

Ren

Dynamic analysis of N-glycomic and transcriptomic changes in the development of ovarian cancer cell line A2780 to its three cisplatin-resistant variants

Ann Transl Med82892020

10.21037/atm.2020.03.12

32355733

Stephanie

Jennifer

Kamila

Yoann

Manfred

N-glycomic and transcriptomic changes associated with CDX1 mRNA expression in colorectal cancer cell lines

Cells82732019

10.3390/cells8030273

Yue

Feifei

Yun

Zhixin

A quantitative N-glycoproteomics study of cell-surface N-glycoprotein markers of MCF-7/ADR cancer stem cells

Anal Bioanal Chem412242324322020

10.1007/s00216-020-02453-7

32030495

Miura

Hato

Shinohara

Kuramoto

Furukawa

Kurogochi

Shimaoka

Tada

Nakanishi

Ozaki

BlotGlycoABCTM, an integrated glycoblotting technique for rapid and large scale clinical glycomics

Mol Cell Proteomics73703772008

10.1074/mcp.M700377-MCP200

17986439

Kamiyama

Yokoo

Furukawa

Kurogochi

Togashi

Miura

Nakanishi

Kamachi

Kakisaka

Tsuruga

Identification of novel serum biomarkers of hepatocellular carcinoma using glycomic analysis

Hepatology57231423252013

10.1002/hep.26262

23322672

Furukawa

Tsuda

Okada

Kimura

Piao

Tanaka

Shinohara

Comprehensive glycomics of a multistep human brain tumor model reveals specific glycosylation patterns related to malignancy

PLoS One10e01283002015

10.1371/journal.pone.0128300

26132161

Kang

Madera

Alley

JrGoldman

Mechref

Novotny

Glycomic alterations in the highly-abundant and lesser-abundant blood serum protein fractions for patients diagnosed with hepatocellular carcinoma

Int J Mass Spectrom3051851982011

10.1016/j.ijms.2010.11.007

23788846

Vanhooren

Liu

Franceschi

Gao

Libert

Contreras

Chen

N-glycan profiles as tools in diagnosis of hepatocellular carcinoma and prediction of healthy human ageing

Mech Ageing Dev13092972009

10.1016/j.mad.2008.11.008

19070631

Kirmiz

Clowers

Chew

Lam

Ferrige

Alecio

Borowsky

Sulaimon

A serum glycomics approach to breast cancer biomarkers

Mol Cell Proteomics643552007

10.1074/mcp.M600171-MCP200

16847285

Alley

JrVasseur

Goetz

Svoboda

Mann

Matei

Menning

Hussein

Mechref

Novotny

N-linked glycan structures and their expressions change in the blood sera of ovarian cancer patients

J Proteome Res11228223002012

10.1021/pr201070k

22304416

Lee

Bose

Ahn

Son

Kim

Chung

Kim

Lee

Breast cancer diagnosis by analysis of serum N-glycans using MALDI-TOF mass spectroscopy

PLoS One15e02310042020

10.1371/journal.pone.0231004

32271809

Mekkawy

Pourgholami

Morris

Involvement of urokinase-type plasminogen activator system in cancer: An overview

Med Res Rev349189562014

10.1002/med.21308

24549574

Zhou

Liang

Peng

Zhen

Expression and clinical significance of coagulate and fibrolysis factors in tissue and plasma from hepatocellular carcinoma patients

Ai Zheng25143314382006(In Chinese)

17094916

Itoh

Hayashi

Kanamaru

Morita

Suzuki

Wang

Zhou

Rui

Yamamoto

Kuroda

Itoh

Clinical significance of urokinase-type plasminogen activator activity in hepatocellular carcinoma

J Gastroenterol Hepatol154224302000

10.1046/j.1440-1746.2000.02150.x

10824888

Kamiyama

Une

Uchino

Hamada

Hepatocyte growth factor enhances the invasion activity of human hepatocellular carcinoma cell lines

Int J Oncol126556591998

9472107

Chakraborty

Pawelek

GnT-V, macrophage and cancer metastasis: A common link

Clin Exp Metastasis203653732003

10.1023/A:1024007915938

12856724

Handerson

Camp

Harigopal

Rimm

Pawelek

Beta1,6-branched oligosaccharides are increased in lymph node metastases and predict poor outcome in breast carcinoma

Clin Cancer Res11296929732005

10.1158/1078-0432.CCR-04-2211

15837749

Handerson

Pawelek

Beta1,6-branched oligosaccharides and coarse vesicles: A common, pervasive phenotype in melanoma and other human cancers

Cancer Res63536353692003

14500369

Lee

Dickson

Lin

Activation of hepatocyte growth factor and urokinase/plasminogen activator by matriptase, an epithelial membrane serine protease

J Biol Chem27536720367252000

10.1074/jbc.M007802200

10962009

Takeuchi

Harris

Huang

Yan

Coughlin

Craik

Cellular localization of membrane-type serine protease 1 and identification of protease-activated receptor-2 and single-chain urokinase-type plasminogen activator as substrates

J Biol Chem27526333263422000

10.1074/jbc.M002941200

10831593

Ito

Akinaga

Yamanaka

Nakagawa

Kondo

Dickson

Lin

Miyauchi

Taniguchi

Miyoshi

Co-expression of matriptase and N-acetylglucosaminyltransferase V in thyroid cancer tissues-its possible role in prolonged stability in vivo by aberrant glycosylation

Glycobiology163683742006

10.1093/glycob/cwj084

16464868

Fujitani

Furukawa

Araki

Fujioka

Takegawa

Piao

Nishioka

Tamura

Nikaido

Ito

Total cellular glycomics allows characterizing cells and streamlining the discovery process for cellular biomarkers

Proc Natl Acad Sci USA110210521102013

10.1073/pnas.1214233110

23345451

Henkin

Marcotte

Yang

The plasminogen-plasmin system

Prog Cardiovasc Dis341351641991

10.1016/0033-0620(91)90010-J

1832499

Appella

Robinson

Ullrich

Stoppelli

Corti

Cassani

Blasi

The receptor-binding sequence of urokinase. A biological function for the growth-factor module of proteases

J Biol Chem262443744401987

3031025

Buko

Kentzer

Petros

Menon

Zuiderweg

Sarin

Characterization of a posttranslational fucosylation in the growth factor domain of urinary plasminogen activator

Proc Natl Acad Sci USA88399239961991

10.1073/pnas.88.9.3992

2023947

Bergwerff

Van Oostrum

Kamerling

Vliegenthart

The major N-linked carbohydrate chains from human urokinase. The occurrence of 4-O-sulfated, (alpha 2–6)-sialylated or (alpha 1–3)-fucosylated N-acetylgalactosamine(beta 1–4)-N-acetylglucosamine elements

Eur J Biochem228100910191995

10.1111/j.1432-1033.1995.tb20351.x

7737145

Kasai

Arimura

Nishida

Suyama

Primary structure of single-chain pro-urokinase

J Biol Chem26012382123891985

2931434

McLellan

Vetterlein

Roblin

The glycoprotein nature of human plasminogen activators

FEBS Lett1151811841980

10.1016/0014-5793(80)81163-X

7190507

Sarubbi

Nolli

Robbiati

Soffientini

Parenti

Cassani

The differential glycosylation of human pro-urokinase from various recombinant mammalian cell lines does not affect activity and binding to PAI-1

Thromb Haemost629279331989

10.1055/s-0038-1651030

2512681

Lijnen

Nelles

Collen

Biochemical properties of recombinant mutants of nonglycosylated single chain urokinase-type plasminogen activator

Biochim Biophys Acta115937431992

10.1016/0167-4838(92)90072-L

1390910

Saito

Nishikawa

Ihara

Soejima

Wada

Sekiya

Niikawa

Taniguchi

cDNA cloning and chromosomal mapping of human N-acetylglucosaminyltransferase V+

Biochem Biophys Res Commun1983183271994

10.1006/bbrc.1994.1045

8292036

Miyoshi

Terao

Kamada

Physiological roles of N-acetylglucosaminyltransferase V(GnT-V) in mice

BMB Rep455545592012

10.5483/BMBRep.2012.45.10.190

23101508

Taniguchi

Miyoshi

Ikeda

Ihara

Implication of N-acetylglucosaminyltransferases III and V in cancer: Gene regulation and signaling mechanism

Biochim Biophys Acta14552873001999

10.1016/S0925-4439(99)00066-6

10571019

Cui

Huang

Jin

Jing

Shi

Liu

Yuan

Luo

N-glycosylation by N-acetylglucosaminyltransferase V enhances the interaction of CD147/basigin with integrin β1 and promotes HCC metastasis

J Pathol24541522018

10.1002/path.5054

29431199

Murata

Miyoshi

Kameyama

Ishikawa

Kabuto

Sasaki

Hiratsuka

Ohigashi

Ishiguro

Ito

Expression of N-acetylglucosaminyltransferase V in colorectal cancer correlates with metastasis and poor prognosis

Clin Cancer Res6177217772000

10815896

Mehta

Norton

Liang

Comunale

Wang

Rodemich-Betesh

Koszycki

Noda

Miyoshi

Block

Increased levels of tetra-antennary N-linked glycan but not core fucosylation are associated with hepatocellular carcinoma tissue

Cancer Epidemiol Biomarkers Prev219259332012

10.1158/1055-9965.EPI-11-1183

22490318

Ito

Miyoshi

Sakon

Takeda

Noda

Tsujimoto

Ito

Honda

Takemura

Wakasa

Elevated expression of UDP-N-acetylglucosamine: Alphamannoside beta1,6 N-acetylglucosaminyltransferase is an early event in hepatocarcinogenesis

Int J Cancer916316372001

10.1002/1097-0215(200002)9999:9999<::AID-IJC1118>3.0.CO;2-F

11267972

Tian

Miyoshi

Kawaguchi

Shaker

Ito

Taniguchi

Tsujimoto

Matsuura

The implication of N-acetylglucosaminyltransferase V expression in gastric cancer

Pathobiology752882942008

10.1159/000151709

18931531

Yamamoto

Ino

Miyoshi

Shibata

Takahashi

Kajiyama

Nawa

Nomura

Nagasaka

Kikkawa

Expression of N-acetylglucosaminyltransferase V in endometrial cancer correlates with poor prognosis

Br J Cancer97153815442007

10.1038/sj.bjc.6604044

17971775

Takahashi

Yamamoto

Ino

Miyoshi

Nagasaka

Kajiyama

Shibata

Nawa

Kikkawa

High expression of N-acetylglucosaminyltransferase V in mucinous tumors of the ovary

Oncol Rep22102710322009

19787216

Ihara

Miyoshi

Murata

Nakahara

Honke

Dickson

Lin

Taniguchi

Prometastatic effect of N-acetylglucosaminyltransferase V is due to modification and stabilization of active matriptase by adding beta 1–6 GlcNAc branching

J Biol Chem27716960169672002

10.1074/jbc.M200673200

11864986

Ihara

Miyoshi

Nakahara

Sakiyama

Ihara

Akinaga

Honke

Dickson

Lin

Taniguchi

Addition of beta1-6 GlcNAc branching to the oligosaccharide attached to Asn 772 in the serine protease domain of matriptase plays a pivotal role in its stability and resistance against trypsin

Glycobiology141391462004

10.1093/glycob/cwh013

14551220

Nagae

Kizuka

Mihara

Kitago

Hanashima

Ito

Takagi

Taniguchi

Yamaguchi

Structure and mechanism of cancer-associated N-acetylglucosaminyltransferase-V

Nat Commun933802018

10.1038/s41467-018-05931-w

30140003

Figure 1.

Expression of uPA and invasiveness. (A) uPA expression was detected in HLE cells, but not in HepG2 cells, by western blot analysis. uPA and GAPDH were exposed for 40 and 10 sec, respectively. (B) Cell invasion assay showed that HLE cells exhibited higher invasiveness than HepG2 cells. uPA, urokinase-type plasminogen activator. Experiment was repeated three times and the statistical significance is denoted by *P<0.05.

Figure 2.

Invasion assay of (A) HLE and (B) HepG2 cells.

Figure 3.

The concentration of 10 N-glycans was higher in HLE compared with HepG2 cells. Hex, hexose; HexNAc, N-acetylhexosamine; Man, mannose; GlcNAc, N-acetylglucosamine; NeuAc, N-acetylneuraminic acid.

Figure 4.

uPA knockdown and HLE cell invasiveness. (A) Two siRNAs against uPA effectively downregulated uPA protein levels in HLE cells as detected by western blot analysis. uPA and GAPDH were exposed for 60 and 6 sec, respectively. (B) Knockdown of uPA in HLE cells inhibited invasion. uPA, urokinase-type plasminogen activator; HPF, high-power field. Each experiment was repeated three times and the statistical significance is denoted by *P<0.05.

Figure 5.

Invasion assay of HLE cells with uPA knockdown. (A) HLE control. (B) HLE uPA siRNA1. (C) HLE uPA siRNA2. uPA, urokinase-type plasminogen activator.

Figure 6.

Decreased N-glycans in HLE cells were correlated with changes in invasiveness. Hex, hexose; HexNAc, N-acetylhexosamine; Man, mannose; GlcNAc, N-acetylglucosamine; NeuAc, N-acetylneuraminic acid.

Figure 7.

uPA overexpression and HepG2 cell invasiveness. (A) uPA expression increased in HepG2uPAWT cells, which overexpressed uPA using lentiviral vectors, and HepG2uPAMUT cells. uPA and GAPDH were exposed for 30 and 20 sec, respectively. (B) The invasion of HepG2uPAWT cells, which overexpressed mutant uPA proteins and had glutamine as the 322nd amino-acid residue instead of asparagine, was increased, while that of HepG2uPAMUT cells was not increased. uPA, urokinase-type plasminogen activator; WT, wild-type; MUT, mutant; HPF, high-power field. Each experiment was repeated three times and the statistical significance is denoted by *P<0.05.

Figure 8.

Invasion assay of HepG2 cells with uPA overexpression. (A) HepG2 control. (B) HepG2 uPA/MUT. (C) HepG2 uPA/WT. uPA, urokinase-type plasminogen activator; WT, wild-type; MUT, mutant.

Figure 9.

Change of concentration of N-glycans in HepG2 cells with uPA overexpression. Hex, hexose; HexNAc, N-acetylhexosamine; Man, mannose; GlcNAc, N-acetylglucosamine; NeuAc, N-acetylneuraminic acid. uPA, urokinase-type plasminogen activator. (A) Concentrations of N-glycans (m/z=1502-2156). (B) Concentrations of N-glycans (m/z=2172-3497).

Figure 10.

GnT-V knockdown and HLE cell invasiveness. (A) siRNAs against GnT-V effectively downregulated GnT-V protein levels in HLE cells as detected by western blot analysis. uPA expression did not decrease with GnT-V knockdown. GnT-V, uPA and GAPDH were exposed for 60, 60 and 20 sec, respectively. (B) Knockdown of GnT-V in HLE cells inhibited invasion without decreasing uPA expression. uPA, urokinase-type plasminogen activator; GnT-V, N-acetylglucosaminyltransferase V; HPF, high-power field. Experiment was repeated three times and the statistical significance is denoted by *P<0.05.

Figure 11.

Invasion assay of HLE GnT-V knockdown. (A) HLE control. (B) HLE GnT-V knockdown. GnT-V, N-acetylglucosaminyltransferase V.

Figure 12.

Change of concentration of N-glycans in HLE cells with GnT-V knockdown. Hex, hexose; HexNAc, N-acetylhexosamine; Man, mannose; GlcNAc, N-acetylglucosamine; NeuAc, N-acetylneuraminic acid; GnT-V, N-acetylglucosaminyltransferase V. (A) Concentrations of N-glycans (m/z=1892-3192). (B) Concentrations of N-glycans (m/z=3351-3922).

Figure 13.

N-glycans in uPA from human urine. Hex, hexose; HexNAc, N-acetylhexosamine; Man, mannose; GlcNAc, N-acetylglucosamine; NeuAc, N-acetylneuraminic acid. uPA, urokinase-type plasminogen activator.

Figure 14.

N-glycoforms of urokinase from human urine. Tag: aoWR, Matrix: 20 mg/mi DHB in 70% EtOH, Internal Standard: 50 pmol. There was one common N-glycan (m/z=1892) correlated with invasiveness based on glycoblotting analysis of HLE/HepG2, uPA knockdown HLE, uPA-overexpressing HepG2 and GnT-V knockdown HLE cells. uPA, urokinase-type plasminogen activator.