Fucosyltransferase 8 deficiency suppresses breast cancer cell migration by interference of the FAK/integrin pathway

Abstract

BACKGROUND AND OBJECTIVE:

Fucosyltranferase 8 (FUT8), which catalyzes core fucosylation of glycopeptides, plays important roles in cancer development. In this study, we aimed to explore the influence of FUT8 expression on migration ability of human breast cancer cells and its potential mechanisms.

METHODS:

The core fucosylation levels in normal and FUT8 deficient MCF-7 cells were analyzed by lectin LCA blots. Then, the cell adhesion assay, transwell and wound healing experiments were conducted. The phosphorylation of FAK and core fucosylation of E-cadherin and its downstream integrins in the FAK/integrin pathway were measured. Moreover, the expression levels of nuclear $\beta$ -catenin, MMP-2, and MMP-9 were also measured.

RESULTS:

The core fucosylation levels were significantly reduced by inhibited FUT8. FUT8 deficiency suppressed the adhesion, migration and invasion of MCF-7 cells; the potential mechanisms might involve three aspects. FUT8 deficiency inhibited FAK/integrin pathway by suppressing core fucosylation of E-cadherin. In addition, FUT8 deficiency reduced nuclear $\beta$ -catenin accumulation. The suppression of MMP-2 and MMP-9 expression also accounted for FUT8 deficiency inhibiting breast cancer cells migration.

CONCLUSIONS:

FUT8 deficiency suppressed migration of MCF-7 cells by impacting core fucosylation of E-cadherin and the downstream FAK/integrin pathway. Therefore, FUT8 is a potential biomarker for breast cancer detection and treatment.

1. Introduction

Breast cancer has become a vital cause of cancer-associated morbidity and mortality among woman worldwide [1] and the incidence of breast cancer has been increasing at a fast pace for the past few decades [2]. Although some improvements have been achieved in oncological immunotherapies, gene therapies and target-oriented therapies, large amounts of patients still suffer from tumor proliferation, metastasis and malignancy [3]. Therefore, it is necessary to identify effective biomarkers that can be used for prevention, early detection and treatment in breast cancer.

Core fucosylation is a post-translational modification process that plays an important role in the pathological progression of breast, hepatocellular carcinoma and colon carcinoma [4, 5]. Fucosyltranferase 8 (FUT8) is the only pivotal enzyme that can transfer the fucose residue to the innermost N-acetylglucosamine (GlcNAc) residue of the N-sugar chain in glycoproteins [6] to form $\alpha$ 1, 6-fucosylation, and modulation of its level can impact cancer pathological progress through regulation of $\alpha$ 1, 6-fucosylation. Actually, the modulation of FUT8 expression is closely associated with cancer cell proliferation in colorectal cancer, metastasis in lymphoma and malignancy in lung cancer [7, 8, 9, 10]. Previous research in breast cancer patients reported that the expressions of FUT8 in cancer tissues are significantly higher than that in adjacent normal tissues, and suggests that FUT8 can be regarded as a prognostic biomarker for breast cancer patients [11]. Therefore, downregulating FUT8 expression may be an effective option to interfere the development of breast cancer. However, up to now, the influence of downregulation of FUT8 expression on the migration capability of breast cancer cells and its potential mechanism is not yet fully understood.

In the present study, to explore the effect of FUT8 expression on breast cancer cell migration, FUT8 deficient MCF-7 cells were constructed according to our previous work [12]. The core fucosylation levels, migration and invasion capabilities of breast cancer cells with normal or deficient FUT8 expression were investigated. To probe the potential mechanism, the phosphorylation of FAK and core fucosylation of E-cadherin and its downstream integrins in FAK/integrin pathway were compared in MCF-7-NC, shFUT8-1 and shFUT8-2 cells. In addition, the expression of MMP-2 and MMP-9 was also monitored.

2. Materials and methods

2.1 Materials

Antibodies against FUT8, EGFR, integrin- $\alpha$ 3 and integrin- $\alpha$ 5 were obtained from Santa (Cruz, USA). Antibodies against E-cadherin, MMP-2, MMP-9 and $\beta$ -catenin were bought from Abcam (Cambridge, UK). Antibodies against integrin- $\beta$ 1, p-FAK, FAK, lamin B, $\beta$ -actin and GAPDH were purchased from Cell Signaling Technology (MA, USA). Biotinylated Lens culinaris agglutinin (LCA) lectin and HRP-conjugated streptavidin were acquired from Vector laboratories, Inc. (CA, USA). Protein A-agarose were purchased from Pierce (WI, USA). The HRP-conjugated secondary antibody was bought from Bioss, Inc. (Beijing, China). All cell culture reagents were bought from Gibco BRL Technologies (MD, USA). All chemical reagents were analytical grade.

2.2 Cell culture

Human breast cancer MDA-MB-231, MCF-7, ZR-75-1 and Bcap-37 cell lines were acquired from the American Type Culture Collection (Manassas, VA, USA). MDA-MB-231 and Bcap-37 cells were cultured in DMEM medium. ZR-75-1 cells were maintained in RPMI 1640 medium and MCF-7 cells were propagated in MEM medium. All cell culture mediums were supplemented with 10% (V/V) fetal bovine serum (FBS), 0.1 mg/mL streptomycin and 100 U/mL penicillin. The cells were cultured in a 37 ${}^{\circ }$ C humidified incubator with 5% CO ${}_2$ , fed every 2 days and sub-cultured once they reached 90% confluency.

2.3 Transfection FUT8 gene lentiviral RNA interference vectors

The FUT8 gene lentiviral RNA interference vectors were transfected into MCF-7cells to construct FUT8 deficient MCF-7 cells (shFUT8-1 and shFUT8-2 with different short hairpin RNA sequence (5’-3’) as TG GAGTGATCCTGGATATA and CTATAATGACGGA TCTATA) according to our previous work [12]. The linear PGC-LV-GFP carriers were also transfected into MCF-7 cells (MCF-7-NC) for negative control.

2.4 Real-time PCR

Total RNA in cells were extracted using TRIzol ${}^{\text{TM}}$ Plus RNA purification kit (Thermo Fisher Scientific, Inc.), according to the manufacturer’s instructions. 1 $\mu$ g total RNA was reversed transcribed to generate cDNA by a ReverTra Ace- $\alpha$ kit (Toyobo, Shanghai, China) and Real-time PCR (RT-PCR) was conducted with Powere SYBR ${}^{\text{MT}}$ Green PCR Master Mix (Applied Biosystems, CA, USA). Specific primers used for genes were as follows: FUT8 forward, 5’-CCATTTC AGGTTTGTTTGGTAG-3’; and reverse, 5’-ATTGGT CCCGCTTCTCACTT-3’; $\beta$ -actin forward, 5’-CTGG GACGACATGGAGAAAA-3’; and reverse, 5’-AAGG AAGGCTGGAAGAGTGC-3’. All data were analyzed by the ABI7300 system SDS software (Applied Biosystems) and the relative expression intensity of mRNAs were calculated by the 2 ${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{Ct}}$ method.

2.5 Western blot analysis and immunoprecipitation analysis

Total proteins in cells were extracted by ice cold RIPA buffer with PMSF. Cells were lysed for 30 min and then centrifuged at 15000 rpm at 4 ${}^{\circ }$ C for 10 min. For immunoprecipitation, the supernatants were incubated with 20 $\mu$ L of 50% (v/v) Protein A-agarose for 4 h and then precipitated complexes were boiled in laemmli sample buffer for solubilizing. The protein samples were separated by SDS-PAGE and then transferred to PVDF membranes. After blocking with 5% skim milk in TBST for 1 h at room temperature, the PVDF membranes were incubated overnight in 4 ${}^{\circ }$ C, with primary antibodies diluted in TBST: FUT8 (1:1000), MMP-2 (1:1000), MMP-9 (1:1000), integrin- $\beta$ 1 (1:1000), integrin- $\alpha$ 3 (1:1000), integrin- $\alpha$ 5 (1:1000), E-cadherin (1:1000), EGFR (1:500), $\beta$ -catenin (1:1000), lamin B (1:1000), p-FAK (1:500), FAK (1:500), GAPDH (1:1000) and $\beta$ -actin (1:1000). After overnight incubation, the PVDF membranes were incubated in HRP-conjugated secondary antibody (1:1000) at room temperature for 1 h. The protein brands were stained with Immobilon ${}^{\text{TM}}$ Western Chemiluminescent HRP Substrate detection reagent (Millipore, MA, USA) and captured using Image Lab ${}^{\text{TM}}$ software (Bio-Rad, VA, USA).

2.6 LCA lectin blot analysis

Protein extraction steps were the same as 2.5. The protein samples were fractioned by SDS-PAGE and then transferred to PVDF membranes. After blocking with 0.1% gelatin and 0.05% tween 20 in PBS (pH 7.5) for 1 h at room temperature, the PVDF membranes were incubated overnight in 4 ${}^{\circ }$ C, with biotinylated LCA lectin, and next incubated with HRP-conjugated streptavidin. The images were acquired by the same way as in 2.5.

2.7 Cell adhesion assay

Cells were plated in Matrigel coated 24-well plates (Corning Inc., USA) at the density of 1 $\times$ 10 ${}^4$ cells/well with complete medium and allowed to adhere for 2 h. Subsequently, the medium with non-adhered cells were discarded and PBS was applied to wash plates three times to remove loosely attached cells. The adhered cells were then stained by 0.1% crystal violet and the cell adhesion capability was expressed by the mean absorbance of optical density (OD) 570 nm per assay by microplate reader (Bio Tek instruments, Inc., VT, USA).

2.8 Transwell assay

For invasion assay, cells were incubated on the top chambers of 24-well transwell plates (8.0 $\mu$ m-pore, Corning Inc., USA) with serum-free medium while the bottom chambers were added 10% FBS medium. After 20 h incubation, cells that migrated to the bottom surface of the top chamber membranes were stained with 0.1% crystal violet and photographed by a microscope. The values of invasion were expressed with the mean absorbance of OD 570 nm per assay.

2.9 Wound healing assay

Cells were plated in 6-well plates with a density of 2 $\times$ 10 ${}^5$ cells/well and incubated in complete medium for 24 h. A pipette tip was used to scratch line wounds on the completely confluent monolayer cells. Then, cells were incubated in serum-free medium and photographed at 0, 24 and 48 h by an inverted microscope. The images were analyzed by the Scion image software (version 4.0.3.2, Scion Corporation, MD, USA). The migration ability was expressed as migration % $=$ (the distance at 0 h (D0) – the distance at 24 or 48 h (D24 or D48))/D0.

2.10 Statistical analysis

Data were expressed as mean $\pm$ standard deviation (SD) and a two-tailed, unpaired student’s $t$ -test was used to statistical comparison between groups. $P<$ 0.05 was considered statistically significant.

3. Results

3.1 MCF-7 cells highly expressed FUT8 among four breast cancer cell lines

The FUT8 relative expressions in MDA-MB-231, MCF-7, ZR-75-1 and Bcap-37 cell lines were examined. The results of RT-PCR and Western blot revealed that FUT8 mRNA and protein expression was highest in MCF-7 cells, followed by MDA-MB-231, ZR-75-1, and Bcap-37 as the lowest (Fig. 1A and B). MCF-7 cells were chosen to construct FUT8 deficient cells due to the high levels of FUT8 expression relative to the other breast cancer cell lines.

Figure 1.

MCF-7 cells express higher FUT8 than other human breast cancer cells. (A) RT-PCR and Western blot. (B) analysis of gene and protein expressions of FUT8 in MDA-MB-231, MCF-7, ZR-75-1 and Bcap-37 cell lines. (C) Lectin blot analysis of $\alpha$ 1, 6-fucosylation levels in cells with normal (MCF-7, MCF-7-NC) and deficient (shFUT8-1, 2) expression FUT8. Each bar represents the mean $\pm$ S.D. of three independent experiments; ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01, ${}^{***}P<$ 0.001 compared to the MCF-7 cells.

3.2 Lentiviral RNA interference of FUT8 downregulated core-fucosylation levels

Our previous work revealed that FUT8 expression in shFUT8-1 and shFUT8-2 cells is efficiently downregulated when compared with MCF-7-NC and MCF-7 cells, and shFUT8-2 cells express lower FUT8 levels than shFUT8-1 cells [12].

To validate the effects of FUT8 interference, core fucosylation levels were detected by lectin blots. The results showed that non-carriers transfection had no influence on the core fucosylation level in MCF-7-NC cells, because the core fucosylation level in MCF-7-NC was similar to the core fucosylation level in MCF-7. The shFUT8-1 and shFUT8-2 cells displayed much lower core fucosylation levels, which indicated lentiviral RNA interference vectors decreased core fucosylation levels through inhibiting FUT8 expression (Fig. 1C).

3.3 FUT8 deficiency suppressed the adhesion ability of MCF-7 cells

A non-specific cell adhesion assay was applied to investigate the influence of FUT8 expression on MCF-7 cell adhesion. As is shown in Fig. 2A, the absorbance of OD 570 nm in MCF-7-NC cells were 0.83 $\pm$ 0.08. However, the absorbance of OD 570 nm decreased significantly (0.53 $\pm$ 0.03 for shFUT8-1 and 0.57 $\pm$ 0.05 for shFUT8-2) in shFUT8-1 and shFUT8-2 cells compared to MCF-7-NC, which indicates that cell adhesion is severely reduced by the silencing FUT8.

Figure 2.

FUT8 deficiency inhibits the adhesion, invasion and migration of MCF-7 cells. (A) Cell adhesion evaluation of MCF-7 cells after FUT8 interference via cell adhesion assay. (B) The images and quantitative analysis of invaded MCF-7-NC, shFUT8-1 and shFUT8-2 cells by transwell assay. (Scale bars $=$ 20 $\mu$ m; Magnification, 200 $\times$ ). (C) The images and quantitative analysis of migrated MCF-7-NC, shFUT8-1 and shFUT8-2 cells by wound healing assay. (Scale bars $=$ 200 $\mu$ m; Magnification, 10 $\times$ ). Each bar represents the mean $\pm$ S.D. of three independent experiments; ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01, ${}^{***}P<$ 0.001 compared to the MCF-7-NC cells.

3.4 FUT8 deficiency suppressed the invasion ability of MCF-7 cells

To investigate the effect of FUT8 expression on invasion ability of human breast cancer cells, the transwell assay was carried out. As presented in Fig. 2B, MCF-7-NC cells exerted high invasion ability, while FUT8 deficiency effectively decreased the numbers of MCF-7 cells migrating through the transwell membrane to the bottom chamber. Thus, FUT8 deficient MCF-7 cells exerted weaker invasion potential than MCF-7-NC cells.

3.5 FUT8 deficiency suppressed the migration of MCF-7 cells

The migratory capabilities of MCF-7-NC, shFUT8-1 and shFUT8-2 cells from one side of the wound to the other side were investigated by wound healing assay. Compared with MCF-7-NC cells, the migration rate (%) of shFUT8-1 and shFUT8-2 cells changed little in 24 h (Fig. 2C). The migration rate (%) of shFUT8-1 and shFUT8-2 cells decreased significantly by 48 hours, showing that MCF-7 cell migration is inhibited in the absence of FUT8.

3.6 FUT8 deficiency suppressed the phosphorylation of FAK

FAK activation by E-cadherin stimulates the potential of migrating cells, so the phosphorylation of FAK directly impacts the migration of cancer cells [33]. The Western blot results in Fig. 3A show that the phosphorylation of FAK was enormously inhibited by downregulating the expression of FUT8 in shFUT8-1 and shFUT8-2 cells, compared with MCF-7-NC cells.

Figure 3.

(A) The phosphorylation levels of FAK in MCF-7-NC, shFUT8-1 and shFUT8-2 cells, analyzed by Western blot. (B) The protein expression and core fucosylation of integrin- $\beta$ 1, integrin- $\alpha$ 3, integrin- $\alpha$ 5, E-cadherin and EGFR in FUT8 normal and deficient cells by Western blot and immunoprecipitation assays. Each bar represents the mean $\pm$ S.D. of three independent experiments; ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01, ${}^{***}P<$ 0.001 compared to the MCF-7-NC cells.

3.7 FUT8 deficiency affected expression and core fucosylation levels of substrates

The Western blot results in Fig. 3B showed that the absence of FUT8 had no obvious influence on the expressions of integrin- $\beta$ 1, integrin- $\alpha$ 3, E-cadherin nor EGFR but did increase the expression of integrin- $\alpha$ 5. The LCA lectin immunoprecipitation revealed that the core fucosylation levels of integrin- $\beta$ 1, integrin- $\alpha$ 3, E-cadherin and EGFR significantly decreased, however it increased in integrin- $\alpha$ 5. Therefore, FUT8 interference decreased the core fucosylation levels of integrin- $\beta$ 1, integrin- $\alpha$ 3, E-cadherin and EGFR, but increased the core fucosylation of integrin- $\alpha$ 5.

3.8 FUT8 deficiency suppressed nuclear $\beta$ -catenin accumulation

$\beta$ -catenin, as well as E-cadherin, is a principle marker of the epithelial-to-mesenchymal transition which accelerates tumor metastasis. In this study, we separately detected $\beta$ -catenin expression in the cytoplasm and nucleus (Fig. 4A). The results revealed that FUT8 interference has no effect on the expression of total $\beta$ -catenin. However, the amount of nuclear $\beta$ -catenin significantly decreased in shFUT8-1 and shFUT8-2 cells compared with MCF-7-NC, and shFUT8-2 cells showed stronger reduction than shFUT-1 cells. This result suggested that FUT8 interference inhibits nuclear $\beta$ -catenin accumulation.

Figure 4.

(A) The total $\beta$ -catenin and nuclear $\beta$ -catenin quantifications in MCF-7-NC, shFUT8-1 and shFUT8-2 cells, detected by Western blot. (B) Western blot analysis of protein expression of MMP-2 and MMP-9 in MCF-7-NC, shFUT8-1 and shFUT8-2 cells. Each bar represents the mean $\pm$ S.D. of three independent experiments; ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01, ${}^{***}P<$ 0.001 compared to the MCF-7-NC cells.

3.9 FUT8 deficiency downregulated expressions of MMP-2 and MMP-9

MMPs have close associations with cancer cell invasion and migration, and the deduction of MMP-2 and MMP-9 has been reported to inhibit cancer cell invasion. In our results, shFUT8-1 and shFUT8-2 cells had decreased expression of MMP-2 and MMP-9 compared with MCF-7-NC cells (Fig. 4B). The decreased expression levels of MMP-9 in shFUT8-1 and shFUT8-2 cells were comparable, while shFUT8-2 cells showed a higher reduction in MMP-2 expression than shFUT8-1 cells.

4. Discussion

Glycoproteins have been considered for the past decade as specific biomarkers for diagnosis and prognostic evaluation of cancers [5]. FUT8 is the only enzyme able to catalyze $\alpha$ 1, 6-fucosylation of glycopeptides, and its regulation effect on cancer development causes much concern. The glycoprotein substrates of FUT8 can gather to adhesion factor L1CAM and this assembly impacted L1CAM cleavage in melanoma cells. Therefore, FUT8 promoted melanoma invasion by regulating the activity of L1CAM and thus it was considered as a driver of melanoma metastasis [13]. Compared with control subjects, the mRNA expression of FUT8 was higher not only in lung tissue specimens, but also in plasma of lung cancer patients, so FUT8 could be regarded as a circulating biomarker for the early detection of lung cancer [14]. FUT8 could also play a prognostic biomarker role in stage II and III p53-negative colorectal cancer patients, where high expression of FUT8 was closely associated with increased disease free survival [15]. Studies in breast cancer tissues showed that the over-expression of FUT8 was significantly associated with less time to disease-related mortality and reduced time to recurrence, and so is also considered as a prognostic biomarker in breast cancer [11, 16]. In this study, the FUT8 deficient cells displayed suppression of adhesion, migration and invasion in the human breast cancer cell line MCF-7; hence, we propose that the expression of FUT8 is not only useful as a prognostic biomarker, but also as a detection and treatment biomarker in breast cancer.

E-cadherin is the marker of epithelial cells and its decreased expression reduces the amount of epithelial cell junctions, increases epithelial-mesenchymal transition and promotes cells acquiring the capabilities of migration and invasion [17]. In our results, the expression of E-cadherin changed little in shFUT8-1 and shFUT8-2 cells, compared with MCF-7-NC cells, which is in agreement with the former observation that FUT8 expression has no significant influence on E-cadherin expression in miR-10 overexpressed MCF10A cells [18]. Therefore, the suppressed migration and invasion by interfering FUT8 was achieved without inhibiting EMT in MCF-7 cells. Core fucosylation of E-cadherin can regulate nuclear $\beta$ -catenin accumulation, which was validated in FUT8 overexpressed and FUT8-RNAi lung cancer cells, but does not occur in MDA-MB-231 cells [19]. The activation of $\beta$ -catenin in cancer cells promotes the potential to develop resistance to treatment, self-renewal and migration [20]. Here, the results demonstrated that decreased FUT8 expression reduces core fucosylation of E-cadherin and that the inhibition of nuclear $\beta$ -catenin accumulation also occurred in MCF-7 cells.

Core fucosylated integrin- $\alpha$ 3 $\beta$ 1 is the major receptor of laminin 5 and works as important signal transducer in laminin 5 induced cell migration and adhesion. In FUT8 ${}^{-/-}$ embryonic fibroblasts, the core fucosylation of integrin- $\alpha$ 3 $\beta$ 1 is almost abolished, which is associated with greatly decreased migration [21]. The activation of integrin- $\alpha$ 5 $\beta$ 1, a major fibronectin receptor, promotes membrane complex formation and cell adhesion, which is preferable to cell migration [22]. This activation of integrin- $\alpha$ 5 $\beta$ 1 needs the $N$ -glycosylation of integrin $\alpha$ 5 and core fucosylation of integrin- $\beta$ 1 [23]. Once reduced FUT8 expression decreases the core fucosylation of integrin- $\beta$ 1, the integrin- $\alpha$ 5 $\beta$ 1 will be inactivated and the continuous inactivation of integrin- $\alpha$ 5 $\beta$ 1 may in turn stimulate the expression and core fucolysation of integrin- $\alpha$ 5 [24]. In agreement with this data, decreased core fucosylation of integrin- $\alpha$ 3, $\beta$ 1 and increased expression and core fucosylation of integrin- $\alpha$ 5 was observed in shFUT8 cells.

The studies of FAK and endocytosed integrins revealed that FAK regulates integrin activation, sustains integrin active conformation and promotes its recycling into new adhesions, with the help of focal adhesions [25]. The recycling of activated integrins happens in the initiating stage of migrating cells. Therefore, the activation of FAK promotes migration potential. The phosphorylation of FAK can be achieved by core fucosylation of E-cadherin through regulating the SRC/FAK pathway, so core fucosylation of E-cadherin decides the fate of cell migration by FAK/integrin pathway [26, 27]. In our results, insufficient FUT8 expression inhibited core fucosylation of E-cadherin and the downstream FAK/integrin pathway. Next, the reduced phosphorylation of FAK repressed core fucosylation of integrins and reduced the adhesion ability of human breast cancer cells.

The biological activities, such as potentiating cancer malignancy, of epidermal growth factor receptors (EGFR) are also regulated by FUT8-catalyzed core fucosylation [28]. The core fucosylation of EGFR and its downstream signals induced by EGF are inhibited in FUT8 ${}^{-/-}$ embryonic fibroblasts, and this inhibition can be reversed by reinduction of FUT8 gene [28]. Our present study revealed that the expression levels of EGFR barely vary, but the core fucosylation of this protein was significantly inhibited in FUT8 deficient human breast cancer cells.

Metalloproteinases (MMPs) play pivotal roles in cancer metastasis, invasion and angiogenesis through changing the microenvironment of tumor cells [29, 30, 31]. Among the members of MMP family, MMP-2 exerts tumor promoting effects and its genetic polymorphism increases the rate of breast, lung, colorectal and gastric cancer incidence [32, 33, 34, 35]. MMP-9 is also vital in tumor incidence and metastasis, and its genetic polymorphism increases the risk of bladder cancer and tumor malignancy [36, 37]. In our result, FUT8 interference suppressed the expressions of MMP-2 and MMP-9, which indicated that reducing FUT8 expression can inhibit the deterioration of tumor microenvironment catalyzed by MMP-2 and MMP-9.

In conclusion, FUT8 deficient cells were utilized to investigate the influence of FUT8 expression on human breast cancer cells migration capability. Our results revealed that FUT8 deficiency represses human breast cancer cell migration through three mechanisms. Firstly, FUT8 deficiency inhibited FAK/ integrin pathway by suppressing core fucosylation of E-cadherin; Secondly, FUT8 deficiency reduced nuclear $\beta$ -catenin accumulation also by the regulation of E-cadherin; and thirdly, FUT8 deficiency suppressed MMP-2 and MMP-9 expression and their catalyzation for tumor microenvironment deterioration.