Silencing the YB-1 Gene Inhibits Cell Migration in Gastric Cancer In Vitro


Correspondence to: Boon-Huat Bay, Department of Anatomy, Yong Loo Lin School of Medicine, National University of Singapore, 4 Medical Drive, MD10, Singapore 117 597. Fax: +65-67787643. E-mail:


The Y-Box-Binding Protein-1 (YB-1) is known to regulate the processes of transcription, translation, cellular response to drug treatment and viral infection as well as DNA repair among others. As gastric cancer is a common cancer with a high incidence in countries in Asia, we evaluated the association of YB-1 with the malignant potential of gastric cancer cells in vitro. YB-1 mRNA expression levels were first determined by real-time RT-PCR in two adherent gastric cancer cell lines (viz., MKN7 and NUGC3 gastric cancer cells) and a normal GES-1 gastric epithelial cell line. Poorly differentiated NUGC3 gastric cancer cells were found to have the highest YB-1 gene expression among the adherent cells. YB-1 gene expression was also observed to be higher in non-adherent SNU5 gastric cancer cells compared to more aggressive SNU16 cells. Silencing of the YB-1 gene by siRNA in NUGC3 cells was associated with a significant reduction of the YB-1 protein by more than 55% as verified by Western blot analysis. Down-regulation of YB-1 protein expression was further demonstrated qualitatively by immunocytochemistry and immunofluorescence staining. Silencing of the YB-1 gene induced significant inhibition of cell migration in NUGC3 cells by 60% but did not influence cell invasion. Although epithelial-mesenchymal-transition (EMT) is known to be associated with the migratory phenotype in cancer cells, there was no change in the expression of EMT genes when YB-1 expression was modulated. YB-1 appears to have an integral role in cancer cell migration, a process which is important for gastric cancer metastasis. Anat Rec, 296:891–898, 2013. © 2013 Wiley Periodicals, Inc.

According to global cancer statistics, gastric cancer ranks among the top five most common cancers worldwide, and is one of the leading causes of cancer-related mortality (Garcia et al., 2007). Despite the observation of a significant declining trend in its incidence in developed countries (Jemal et al., 2011), gastric cancer still pose a major challenge to the health of people in some parts of the world (Gigek et al., 2012). Generally, incidence rates remain high in developing countries, and in the Asian region, in particular, Japan, Korea, and China (Parkin, 2004). However, low incidence rates are present in North American and Western European regions (Bertuccio et al., 2009). The occurrence of the disease varies throughout the world and is twice that in males compared to females. Gastric cancers can be further classified into two categories based on the Lauren classification, viz., diffuse and intestinal types (Chiaravalli et al., 2012). Helicobacter pylori infection has been defined as a major gastric carcinogen for humans (Rathbone and Rathbone, 2011). The development of gastric cancer has also been attributed to smoking, dietary factors and multienvironmental factors (Jemal et al., 2011).

The Y-Box-Binding Protein 1 (YB-1) was first identified in 1988 as a transcription factor that binds to the Y-box (inverted CCAAT box) of the MHC class II promoters (Didier et al., 1988). YB-1 (also known as YBX1), a 43 kDa protein containing 324 amino acids, belongs to the Cold-Shock-Domain (CSD) protein family (Kuwano et al., 2004). The YB-1 gene which comprises 8 exons spanning 19 kb of genomic DNA is known to be located on chromosome 1p34 (Kohno et al., 2003; Eliseeva et al., 2011). The protein structure of YB-1 consists of three functional domains, which are the well conserved CSD, N-terminal domain, and Carboxyl-terminal tail domain (C-terminal domain). The N-terminal domain, which is rich in proline and alanine is believed to participate in transactivation of genes, whereas the C terminal domain is involved in protein-protein interactions (Okamoto et al., 2000) and binds single-stranded DNA/RNA in vitro (Matsumoto and Bay, 2005). Specific and nonspecific RNA binding via RNP1 and RNP2 RNA binding motifs and nonspecific binding to DNA are mediated by the CSD (Kohno et al., 2003; Evdokimova et al., 2006). The pleiotropic functions of YB-1 include regulating transcription, translation, DNA repair and cellular response to drug treatment, cell proliferation, and viral infection (Matsumoto and Wolffe, 1998; Okamoto et al., 2000; Kohno et al., 2003; Inoue et al., 2012). YB-1 proteins are known to be mainly situated in the cytoplasm but relocate to the nuclei in the presence of environmental stress such as heat, anticancer drugs, hyperthermia or UV radiation (Kohno et al., 1992; Uchiumi et al., 1993; Bargou et al., 1997; Koike et al., 1997; Basaki et al., 2007; Cohen et al., 2010).

YB-1 has been reported to be overexpressed in a variety of cancer cells as the protein is known to affect cell proliferation (Jurchott et al., 2003; Yu et al., 2010) and promote the spread of cancer (Kohno et al., 2003). Since the seminal paper showing expression of YB-1 in breast cancer by Royer et al. (Bargou et al., 1997), increased expression of the YB-1 protein has been reported in other cancers including melanoma (Schittek et al., 2007), prostate cancer (Gimenez-Bonafe et al., 2004), glioblastoma (Faury et al., 2007), colorectal carcinoma (Shibao et al., 1999) and neuroblastoma (Wachowiak et al., 2010).

In this study, we determined YB-1 mRNA expression in gastric cell lines and analyzed the association of YB-1 with cell migratory and invasive potential in gastric cancer cells in vitro.


Cell Culture

NUGC3 and MKN74 gastric cancer cell lines obtained from the American Type Culture Collection (ATCC, Rockville, MD) and normal GES-1 gastric epithelial cell line from the Beijing Institute for Cancer Research (BICR, Beijing, China) were cultured in RPMI 1640 medium containing10% fetal bovine serum (FBS) in an incubator at 37°C with 5% CO2/95% air. Non-adherent SNU5 and SNU 16 gastric cancer cell lines are gifts from Dr Yoshi Ito (Cancer Science Institute, Singapore).

Quantitative Real-Time RT-PCR

Total RNA was extracted using the Qiagen RNeasy Mini Kit (Hilden, Germany) and reverse transcribed into cDNA. The primer sequences for YB-1 and genes known to be associated with epithelial-mesenchymal-transition (EMT), viz. KRT18 which encodes cytokeratin 18, CDH1 (E-cadherin), KRT14 (cytokeratin 14) ZO-1 (zonula occludens 1), FN1 (fibronectin 1), and CDH2 (N-cadherin) are listed in Table 1. The housekeeping gene was glyceraldehyde −3-phosphate dehydrogenase (GAPDH). The transcript levels were quantified using QuantiTect SYBR Green Master Mix (Qiagen). PCR cycling conditions were programmed as initial denaturation at 95°C for 15 min with another 45 cycles of denaturation at 94°C for 15 sec. The annealing process was then carried out at 60°C for 25 sec followed by extension at 72°C for 15 sec. The ΔΔCT and 2-ΔΔCT method were used for relative quantification, where ΔCT refers to the difference between the CT values of the target gene and the housekeeping GAPDH gene.

Table 1. The primer sequences for RT-PCR
GeneOrientationSequences (5′→3′)

Transfection with YB-1 Small Interfering RNA (siRNA)

Human YBX1 ON-TARGETplus SMARTpool siRNA and nontargeting pool siRNA were obtained from Dharmacon (Chicago, IL). A negative siRNA sequence that does not target any human gene was used as the control siRNA. Cells were initially seeded in six-well culture dish at a density of 2.5 × 105 cells for 24 h in complete culture medium without antibiotics. NUGC3 gastric cancer cells were then transfected with siYB-1 and non-targeting pooled siRNA by DharmaFECT 1 transfection reagent at a concentration of 20 nmol L−1. Subsequently, growth medium containing 10% FBS was changed after 24 h incubation before harvesting at 48-h post transfection for further analysis.

Western Blot Analysis

The cell lysate was obtained by using an ice-cold extraction reagent (from Pierce, Rockford, IL) which contains EDTA and a protease inhibitor. The lysate was then centrifuged at 13,000 rpm for 10 min at 4°C and the protein concentration was determined by the Bradford method. Subsequently, 20 μg of protein samples were separated by 10% SDS-PAGE and transferred onto a PVDF membrane. After which, primary YB-1 antibody (Tay et al., 2009) was added at a dilution of 1:1,000 for 1 h at room temperature (RT) followed by the secondary antibody which was horseradish peroxidase-conjugated anti-rabbit IgG (Sigma–Aldrich, St. Louis, MO) at a dilution of 1:6,000 for 1 h. For the loading control, monoclonal mouse beta-actin primary antibody at a dilution of 1:5000 (Sigma-Aldrich) was used. The western blots were developed on X-ray films and the optical density of the protein bands were determined after scanning the X-ray film and analysis by the Bio-Rad (Hercules, CA) GS-710 Calibrated Imaging Densitometer using the Quantity One-4.11 software.


NUGC3 cells were fixed in 4% paraformaldehyde (PFA) for 8 min before permeabilization with phosphate buffered saline 0.2% Triton X (PBS-TX) and blocking in 5% normal goat serum for 1 h at RT. This was followed by the addition of primary YB-1 antibody from the RIKEN Institute (Tay et al., 2009) at 1:250 dilution and incubating overnight at 4°C. Subsequently, cells were incubated with biotinylated secondary antibody (1:200 dilution) for 1 h at RT. The immunostaining was visualized using the avidin–biotin-complex technique with diaminobenzidine as the substrate. Hematoxlin was used to counterstain the cells which were then mounted on microscope slides with permount. Negative controls with omission of the primary antibody were also included.


NUGC3 cells were seeded on four-chambered coverglass (Lab-tek Chambered Coverglass System; Nalge Nunc, Tokyo, Japan) at a density of cells 6 × 104 cells per chamber. At about 70% confluency, cells were fixed with 4% PFA and permeabilized with 0.1% Triton X-100 in PBS (PBS-TX). Nearly 1% bovine serum albumin (BSA) prepared in PBS was added to block non-specific binding of antibody before incubation with primary YB-1 antibody in a 1:250 dilution and Cy3 conjugated anti-rabbit IgG antibody (1:200 dilution) (Sigma–Aldrich) for 1 h at RT. Cells were visualized under confocal microscopy and fluorescence images were captured (Olympus Fluoview™ confocal microscope, Tokyo, Japan).

Cell Migration Assay

The assays were carried out using 6.5-mm Transwell polycarbonate membrane inserts with 8.0-μm pore size (Corning, Lowell, MA). Rehydration of Transwell inserts were carried out with RPMI medium at 37°C with 5% CO2 in 24-well plate overnight. The medium was removed from the inserts and wells in the next day. Treated NUGC3 cells were reseeded to the upper chamber of each well at a density of 8 × 104 with serum free RPMI at a volume of 200 μL. About 600 μL of RPMI medium containing 20% FBS were filled in the lower chamber of each well as a chemoattractant. After 24-hr incubation at 37°C, in a 5% CO2 incubator, the Transwell inserts were washed with PBS, fixed with 100% methanol for 15 min and air dried after washing with PBS. The staining was performed with 0.5% (w/v) of crystal violet in 20% methanol in PBS for 30 min. A cotton swab was used to remove the non-migrating cells lying on the upper surface of the insert membrane before counting the migrated cells under a 10× objective of a stereo microscope.

Cell Invasion Assay

The assay was performed using matrigel invasion chambers with 8.0-μm pore size membranes (BD Biosciences, Bedfortd, MA). Overnight rehydration of inserts was carried out with free RPMI medium at 37°C with 5% CO2 in 24-well plate after thawing to room temperature. RPMI medium with 20% FBS in 600 μL was added to the lower chamber after rehydration. Treated NUGC3 cells were seeded at a density of 8 × 104 cells with serum free RPMI at a volume of 200 μL in the upper chamber. The invasion chambers were incubated for 24 hr at 37°C with 5% CO2 humidified conditions. The rest of the procedure is as described in the migration assay.

Statistical Analysis

The SPSS software Version 17.0 for Windows was used for statistical analysis. For comparing means of three or more groups, the One-way ANOVA with post hoc Tukey's test was used. Comparison between two variables was analyzed with Student's t test. P <0.05 was considered as statistically significant.


YB-1 mRNA is Expressed in Gastric Cell Lines

As shown in Fig. 1, YB-1 mRNA was found to be expressed in adherent MKN74 and NUGC3 gastric cancer cells and normal GES-1 gastric epithelial cells. Poorly differentiated invasive NUGC3 cells were found to have the highest YB-1 mRNA expression using GES-1 normal epithelial cells as the calibrator cell line (P = 0.0006). In non-adherent gastric cancer cell lines, YB-1 mRNA expression was significantly higher in SNU16 gastric cancer cells than SNU5 cells (P = 0.0012; Fig. 2).

Figure 1.

Differential expression of YB-1 mRNA in adherent gastric cancer cell lines. Relative expression of YB-1 mRNA levels as determined by quantitative real-time RT-PCR in MKN74, and NUGC3 gastric cancer cell lines. Normal GES-1 gastric epithelial cell line was used as calibrator cell line and GAPDH for calculating the ΔCT values of target genes. Values are means of triplicates. Error bar = SEM. ***P < 0.001.

Figure 2.

Expression of YB-1 mRNA in non-adherent gastric cancer cell lines. Relative expression of YB-1 mRNA levels as determined by quantitative real-time RT-PCR in SNU5 and SNU16 gastric cancer cell lines. SNU5 was used as the calibrator cell line and GAPDH for calculating the ΔCT values of target genes. Error bar = SEM. ** P < 0.01.

Silencing of YB-1 Gene in NUGC3 Gastric Cancer Cells Reduces YB-1 Protein Expression

The YB-1 gene was silenced in NUGC3 gastric cancer cells which had the most abundant expression in the adherent gastric cancer cells examined. The silencing efficiency of siYB-1 in NUGC3 cells was obtained by Western blot analysis at 48 hr post transfection. There was a significant reduction in YB-1 protein expression level by 63% in siYB-1 treated cells compared with siNegative treated cells (P = 0.0013; Fig. 3). In addition, immunocytochemistry also revealed the inhibition of YB-1 expression in the cytoplasm of siYB-1 treated NUGC3 cells (Fig. 4A) which was further verified by confocal immunofluorescence microscopy showing an obvious reduction of YB-1 immunostaining in the cytoplasm of siYB-1-treated NUGC3 cells (Fig. 4B).

Figure 3.

YB-1 protein expression after silencing of the YB-1 gene in NUGC3 gastric cancer cells. Western blots of YB-1 and β-actin (housekeeping protein) and bar chart showing relative YB-1 protein expression. Error bar = SEM. **P < 0.01.

Figure 4.

Localization of YB-1 protein in NUGC3 gastric cancer cells by immuno-cytochemischemistry and immunofluorescence staining. (A) Immunohistochemical expression of YB-1 protein. YB-1 protein was visualized by the brown diaminobenzidine staining. Nucleus was counterstained with hematoxylin. Negative control experiment was performed without the primary antibody. Scale bar = 100 μm. (B) Immunofluorescence staining of YB-1 protein by confocal microscopy. Red Cy3 immunofluorescence represents specific YB-1 staining and DAPI dye stains the nuclei blue. Scale bar = 50 μm.

Silencing of YB-1 Gene Inhibits Cell Migration but not Invasion

Cell migration was significantly inhibited by ∼60% in siYB-1 treated cells in comparison to control siNegative treated cells at 48-hr post transfection (P = 0.048; Fig. 5). Although a lower invasive potential was observed in siYB-1 treated cells, the result was not statistically significant in comparison with the siNegative treated cells (Fig. 6).

Figure 5.

Effect of downregulating of YB-1 protein on migration of NUGC3 gastric cancer cells. (A) NUGC3 cells were treated with siNegative and siYB-1 siRNAs. Micrographs were taken at ×15 and ×100 magnifications, respectively. (B) Bar chart showing the number of cells which migrated in the siYB-1 treated groups was significantly lower than that in siNegative control group. Data are presented as means ± SEM of triplicates. *P < 0.05.

Figure 6.

Effect of downregulating of YB-1 protein on invasion of NUGC3 gastric cancer cells. (A) NUGC3 cells were treated with siNegative and siYB-1 siRNAs. Micrographs were taken at ×15 and ×100 magnifications respectively. (B) The number of cells which invaded in the siYB-1 treated groups was lower than that in siNegative control group but without statistical significance. Data are presented as means ± SEM of triplicates.

Migratory Phenotype Associated with YB-1 is not Related to Epithelial–Mesenchymal–Transition (EMT)

As EMT is known to promote cell migration, the expression of six EMT related genes which encode for cytokeratin 14, cytokeratin 18, E-cadherin and zonula occludens (which are epithelial markers), fibronectin and N-cadherin (which are mesenchymal markers) were analyzed. However, following downregulation of the YB-1 gene in NUGC3 gastric cancer cells, there was no significant difference observed in the expression of the genes that are known to mediate EMT (Fig. 7).

Figure 7.

Expression of EMT-related genes after silencing the YB-1 gene in NUGC3 gastric cancer cells. The housekeeping gene GAPDH was used for calculating the ΔCT values of target genes. YB-1 was used as a positive control. Error bar = SEM. *P < 0.05.


In this study, we observed a higher YB-1 mRNA expression in poorly differentiated NUGC3 gastric cancer cells compared with the moderately differentiated and less aggressive MKN74 gastric cancer cells. This was also the trend observed in non-adherent gastric cancer cells where SNU16 gastric cancer cells had higher YB-1 gene expression than SNU5 cells which are known to be less aggressive (Park et al., 1997; Lin et al., 2005). In addition, the YB-1 protein was observed to be mainly localized to the cytoplasm. It has recently been suggested that different YB-1 antibodies may show distinct staining patterns according to accessibility of the epitopes which is dependent on the nature of the YB-1 complexes (Woolley et al., 2011).

We also demonstrated the inhibition of cell migratory ability after knockdown of YB-1 in malignant NUGC3 gastric cancer cells. Metastasis in cancer patients is normally associated with unsuccessful treatment and poor prognosis (Chambers et al., 2002). Cell migration in epithelial cancers consists of different mechanisms, such as mesenchymal (integrin/MMP dependent), ameboid (integrin/MMP independent), and collective migration (Friedl, 2004). The various migration processes are facilitated by a group of molecules considered as mesenchymal markers, where alteration of these molecules may lead to cancer metastasis (Voulgari and Pintzas, 2009). For instance, N-cadherin is known to aid epithelial cancer progression by accelerating the acquisition of cell motility and enhancing cell invasion (Nieman et al., 1999; Tran et al., 1999; Hulit et al., 2007; Kotb et al., 2011).

EMT was initially thought to be an important biological activity of embryogenesis by changing cell–cell adhesion, disrupting the extracellular matrix, remodeling cell matrix adhesion, modifying the cytoskeleton and facilitating cell movements during the formation of new tissue types in organ development (Moustakas and Heldin, 2007; Guarino et al., 2009). However, EMT has been directly associated with tumor progression as polarized epithelial cancer cells acquire dynamic migratory ability and mesenchymal-like characteristics, becoming flat, spindle-shaped and motile (Huber et al., 2005; Thiery and Sleeman, 2006). In 2009, Evdokimova et al. demonstrated that overexpression YB-1 can induce upregulation of Snail1 and subsequently mediate the process of EMT in breast epithelial cells (Evdokimova et al., 2009). However, our current results showed that down-regulation of YB-1 expression in gastric cancer cells did not modulate expression of EMT genes. This is in agreement with Wu et al., (2012) who reported that upregulation of YB-1 in gastric cancer tissues enhanced liver metastasis but was not associated with EMT. The same authors observed no correlation of YB-1 expression with E-cadherin and vimentin which are established EMT markers.

Some recent reports with regard to enhancers of gastric cancer cell migration include cancer associated orthothopic fibroblasts which have been shown to stimulate migratory ability (possibly via TGFβ) of gastric cancer cells in vitro (Fuyuhiro et al., 2012). Overexpression of P21-activated protein kinase (Pak1) has been observed to increase gastric cancer cell motility (Li et al., 2012). Radixin which is known to mediate cytoskeletal remodeling has been observed to increase cell migration in gastric cancer (Zheng et al., 2012). Vasodilator stimulated phosphoprotein, a key regulator of actin cytoskeleton has been implicated in promoting gastric cancer metastasis via enhanced cell migration and invasion (Wang et al., 2011). Targeting of ROCK1 which acts downstream of Rho GTPases (and known to regulate cytoskeletal reorganization) has been demonstrated to suppress cell migration and invasion (Zheng et al., 2011).

In conclusion, we have shown positive expression of YB-1 in various gastric cancer cells in vitro with higher YB-1 expression in more aggressive cell lines. Additionally, silencing the expression of YB-1 gene in gastric cancer inhibits cell migration but not invasion. This work provides new insights into the functional significance of YB-1 in gastric cancer cells in vitro. It would appear that the migratory potential mediated by YB-1 in gastric cancer cells is not associated with EMT. Future investigations are required to elucidate the mechanistic role of YB-1 in enhancing migration in gastric cancer.


The authors thank Ms Olivia Jane Scully and Ms Song Lin Bay from the Department of Anatomy, National University of Singapore for technical assistance.