Background
While recent decades have witnessed therapeutic advances, the clinical outcome of gastric cancer (GC) is still disappointing in view of the facts that a majority of GC patients has advanced to late stage at diagnosis and that current chemotherapy only offers limited survival advantage.
GC is a very aggressive malignancy representing the third leading cancer mortality worldwide [
1]. Advanced stage at initial diagnosis of GC is commonly seen in a large percentage of GC patients presenting unresectable disease or distant metastases. Moreover, it is one of the most challenging clinical tasks to effectively manage and treat advanced GC patients. Conventional systemic chemotherapy has limited efficacy for advanced GC cancer with only a minority of the patients achieving a satisfactory response [
2]. Thus, there is certainly a need to identify novel biomolecules for possible GC early diagnosis, prognosis prediction and potential targets for development of novel therapeutic agents that target such pivotal molecular signaling pathways participating in gastric carcinogenesis and progression.
The endoplasmic reticulum (ER) protein 29 (ERp29) is expressed ubiquitously and abundantly in eukaryotic cells and normally serves as a molecular chaperone in protein secretion from the ER [
3‐
5]. Protein structural analysis demonstrates that N-terminal domain of ERp29 involves dimerization essential to its function in unfolding and escort of secretory proteins [
6] while the C-terminal domain is necessary for substrate binding and secretion [
7,
8].
ERp29 biological and pathological functions in carcinogenesis of epithelial cancers were in controversy. Several studies reported that ERp29 functioned as a tumor suppressor since its expression suppressed tumor formation in mice [
9,
10] and was inversely correlated with tumor development in breast, lung and gallbladder cancer [
11,
12]. In contrast, ERp29 expression could also sustain tumor cell survival against genotoxic insults by chemotherapy and radiation therapy [
13‐
15]. Intriguingly, ERp29 is found to be involved in inducing mesenchymal–epithelial transition (MET) of cancer cells and epithelial morphogenesis implicating its another important role in predisposing cancer cells to survival and metastasis as well [
9,
11]. Regardless, a functional link between ERp29 expression and GC progression has yet to be described. In this study, we evaluated the expression of ERp29 in primary GC tumors and analyzed its prognostic significance in the GC patients. In addition, we explored the function of ERp29 in GC growth and invasion in vitro and metastasis in vivo. We report here that loss of ERp29 expression was commonly observed in GC and strongly correlated with poor clinical outcome. Knocking down ERp29 promoted GC cell invasion and metastasis. Mechanistically, ERp29 suppression promoted EMT in GC cells as evidenced by a profound reduction of E-cadherin and ZO-1 expression, an increase of Snail and Twist expression and an activation of the PI3K/AKT pathway.
Methods
Clinical samples and immunohistochemical analysis
Human GC samples and their adjacent nontumorous gastric tissues were obtained from surgical resection performed at the First Affiliated Hospital of Fujian Medical University (Fuzhou, China) during the period of 2008 to 2010. The resected specimens were either frozen in liquid nitrogen and stored at −80 °C freezer immediately or fixed in 10% formalin for paraffin embedding. Written informed consent was obtained from all patients according to the Declaration of Helsinki and this study had been approved by our institutional review board and regulatory authorities. Clinicopathological classification and staging were determined by the standards of American Joint Committee on Cancer (AJCC) Seventh Edition of GC TNM Staging [
16]. Tissue cores were extracted from 148 gastric tumors for construction of a tissue microarray (TMA) with at least two tissue cores per sample of 1 mm diameter. A rabbit anti-ERp29 monoclonal antibody (1:100, Abcam, ab42002) was used for immunohistochemical staining of formalin-fixed, paraffin-embedded tissue sections cut from TMAs. To assess the degree of nuclear or cytoplasmic staining, a 5-tiered scale was employed according to the average percentage of positively stained cells. Value 0, 1, 2, 3, 4 represented ≤5%, 5% -25%, 26%–50%, 51%–75% and ≥75% positive cells, respectively. Assigned values were then multiplied with the staining intensity of 0 (no staining), 1 (weak staining, light yellow), 2 (moderate staining, yellow brown), or 3 (strong staining, brown) to obtain a score ranging from 0 to 12. A score equal to or less than 3 was considered low expression of ERp29, and a score greater than 3 was considered high ERp29 expression.
Western blot analysis
Western & IP cell lysis buffer (Beyotime, Shanghai, China) with PMSF (Amresco, Solon, Ohio, USA) was used to lyse tissues or cells for 30 min on ice following centrifugation at 12000 g for 10 min at 4 °C. BCA Protein Assay Kit (Thermo Scientific, Waltham, MA, USA) was employed to measure total proteins in the supernatants collected from the centrifugation. The equal amount of proteins were separated on 10% SDS-PAGE and transferred to a 0.45 μm PVDF membrane (Amersham Hybond, GE Healthcare, München, Germany) followed by blocking in 0.5% albumin from bovine serum (Amresco, Solon, Ohio, USA) overnight at 4 °C with specific antibodies. The primary antibodies used in the study were as follows: rabbit anti-ERp29, rabbit anti-pan-AKT and mouse anti-β-actin diluted at 1:2000 (Cell Signaling Technology, Danvers, MA, USA); rabbit anti-E-cadherin, rabbit anti-ZO-1, rabbit anti-snail, rabbit anti-Ser473-AKT, rabbit anti-Thr308-AKT, rabbit anti-GSK-3β, rabbit anti-phospho-GSK-3β(Ser9), rabbit anti-mTOR and rabbit anti-phospho-mTOR all diluted at 1:1000 (Cell Signaling Technology); rabbit anti-Vimentin diluted at 1:1500 (Abcam, ab92547). After 3 times washing in TBST for 10 min each time, the membrane was then incubated with the respective secondary antibodies at room temperature for 1 h and the immunoblot was developed through enhanced chemiluminescence (Lulong biotech, Xiamen China).
RNA extraction and quantitative real-time PCR
RNA in cultured cells or frozen tissues was extracted using Trizol reagent (Ambion, Carisbad CA,USA) and reverse transcribed to cDNA by RT Reagent Kit (TaKaRa, Dalian, China). Quantitative real-time PCR was carried out in Mx3000P QPCR system (Agilent Technologies, Santa Clara, CA, USA) by using SYBR Premix EX Taq kit (Takara, Shiga, Japan). Primers for human Slug, Snail, E-cadherin and Vimentin were designed (Additional file
1: Table S1) for measuring the relative mRNA expression of the respective genes by the 2
-△△ct method after normalization to endogenous β-actin.
Cell lines
Human gastric cancer cell lines SGC7901 (NTCC411001) and MGC803 (NTCC411124) were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) and cultured in RPMI-1640 (Gibco) medium supplemented with 10% FBS (Gibco) at a humidified atmosphere of 5% CO2 at 37 °C.
Plasmids and generation of stable GC cell lines
Opening reading frame of human ERp29 gene was PCR-amplified and inserted into lentiviral expression vector pCDH-CMV-MCS-EF1-RFP-Puro (System Biosciences, Mountain View, CA, USA). The resulting plasmid or empty vector without insert was co-transfected with lentiviral packaging plasmids pMDL, pVSVG and pRev into 293 T cells. 48 h post co-transfection the supernatants were collected for infecting MGC803 or SGC7901 cells cultured in 6-cm dishes. The clones surviving from puromycin selection were expanded into cell clones as being ERp29 over-expressing cells (MGC803-pERp29 or SGC7901-pERp29), or empty control cells (MGC803-pCDH or SGC7901-pCDH). For generation of ERp29 knocking down clones, shERp29 fragment was cloned into pSuper-retro-puro plasmid (Oligoengine, Seattle, Washington, USA) and the resulting recombinant plasmid or empty vector with no inserts was co-transfected into 293 T cells with lentiviral packaging plasmids pIK (Invitrogen Carlsbad, CA). The supernatants collected from the co-transfection culture were used to infect MGC803 or SGC7901 cells. The clones resistant to puromycin were expanded into the cell clones as being ERp29 knockdown cells (MGC803-pshERp29 or SGC7901-pshERp29), or empty control cells (MGC803-pSuper or SGC7901-pSuper). ERp29 expression levels in these cell lines were evaluated by western blot analysis. The sequences of the primers and oligonucleotides used are listed in Additional file
1: Table S1.
Cell proliferation assay
Cells were seeded into 96-well plate at a density of 2 × 103 cells per well and incubated for 24, 48, 72, 96 or 120 h. The proliferation of cells was evaluated by the Cell Counting Kit-8 (CCK-8, Dojindo, Kuma-moto, Japan). 10 μl CCK-8 reagent was added into each well and incubated for 4 h. The absorbance from each well was determined using a microplate reader at the wavelength of 450 nm (Bio-Tek, Winooski, VT, USA).
5 × 102 cells were grown in 60-mm plate containing complete growth medium for 14 days and the colonies formed that contained 50 or more cells were counted after staining with crystal violet for 5 min. For the soft agar colony formation assay, the cell suspension containing 5 × 103 in 2 × DMEM with 20% FBS were mixed with equal volume of 0.7% agarose and immediately laid on top of an underlayer of 0.5% agarose made in 1× DMEM supplemented with 10% FBS. The plates were cultured for 5 to 21 days when the surviving colonies (>50 cells per colony) were counted and photographed with a Qimaging micropublisher 5.0 RTV microscope camera (Olympus, Tokyo, Japan).
Cell migration and invasion assay
For the migration assay, 4 × 104 cells suspended in serum-free media were plated onto the upper chamber of Transwell insert (8-mm pore size; BD Bioscience). As for the invasion assay, equal cells were plated onto the Transwell insert coated with Matrigel (BD Bioscience). The medium supplemented with 20% FBS in the lower chamber functioned as chemoattractant. 24 h after incubation at 37 °C the cells in the upper surface of chambers were removed with cotton swab and then the cells that successfully migrated or invaded through the pores and located on the lower surface of filter were stained with 0.1% crystal violet in 20% methanol, photographed, and counted using a Qimaging Micropublisher 5.0 RTV microscope camera (Olympus).
Wound-healing assay
Cells were grown to nearly 100% confluence in 6-well plate and scratch was made through the cell monolayer by a 200 μl disposable pipette tip. After washing three times with HBSS, the cells were cultured in fresh growth medium and incubated for 0, 24 or 48 h at which point wound closure was photographed, respectively.
Cell suspension containing 2 × 106 MGC-ERp29 or MGC-vector cells in 0.2 ml serum-free RPMI-1640 was prepared and injected intravenously via the lateral tail vein in female BALB/c nude mice. 12 weeks after injection, all mice were euthanized and the lungs and liver were resected. Metastasis on the lungs and liver was thoroughly examined under dissecting microscope and using histopathologic analysis. The in vivo studies were approved by the Fujian Medical University Institutional Animal Care and Use Committee.
Immunofluorescent staining
For immunofluorescent staining, cells were seeded onto 8-μm-thick section slides and fixed in 4% ice-cold paraformaldehyde for 10 min after overnight culturing. Afterwards, the cells were blocked with 10% normal goat serum (ZSGB Biotech, China) for 10 min and incubated with antibodies against Vimentin and E-cadherin overnight at 4 °C. On the next day, cells were washed three times and incubated with Alexa Fluor 488 conjugated goat anti-rabbit secondary antibody (1:200, 2 mg/ml, Invitrogen). DAPI (2 mg/ml, Invitrogen) was used to counterstain the nuclei and cells were visualized with a laser scanning confocal microscope (Leica, Germany).
Statistical analysis
SPSS 17.0 for Windows was used to perform statistical analysis and all data were expressed as mean ± SD from 3 separate assays. Pearson’s chi-square test and Spearman’s rank-order correlation were employed to analyze an association between ERp29 expression and the clinicopathological parameters. Kaplan-Meier analysis was performed to plot the survival curves. Differences were considered significant when p values were smaller than 0.05.
Discussion
The results obtained from this study provide several lines of evidence supporting that the expression of ERp29 influences the behavior of GC. We showed that ERp29 was downregulated in primary GC tumors and the downregulation of ERp29 was associated with tumor stage and grade as well as poor survival of the patients. In the experimental models, specific knockdown of ERp29 in the GC cells enhanced migration, invasion and metastatic colonization of the lungs and liver. Knockdown of ERp29 in the epithelial GC cells promotes acquisition of EMT traits via activation of PI3K/Akt signaling pathway. These findings imply that ERp29 is likely to functionally serve as a tumor suppressor and that its loss promotes EMT and tumor progression.
EMT is one of the most significant biologic processes in the initial invasion step during cancer metastasis, which allows polarized epithelial cells to become irregular mesenchymal cells. After a series of biochemical changes that induce a morphological transformation, epithelial cells reduced intercellular adhesion and enhanced migratory and invasive capabilities [
17‐
20]. A hallmark of EMT is characteristic of reduced expression of the epithelial markers E-cadherin and ZO-1 but elevated expression of the mesenchymal marker Vimentin. Meanwhile, transcriptional modulators such as Snail, Slug, Twist and β-catenin were up-regulated in EMT [
21‐
23]. ERp29 induced EMT has been found in basal-like MDA-MB-231 breast cancer cells [
11]. Consistently, we found the upregulation of E-cadherin but downregulation of Vimentin when ERp29 was stably overexpressed in the GC cell lines. In contrast, knockdown of ERp29 endowed the GC cells with EMT characteristics exemplified by loss of E-cadherin, upregulation of mesenchymal marker Vimentin, and enhanced expression of E-cadherin transcription repressors Snail and Twist.
A central feature of EMT is manifested by activation of the PI3K/Akt pathway in tumor cell lines and clinical samples [
24]. AKT participates in a wide array of oncogenic processes such as cell growth and survival, cell cycle progression, metabolism and EMT [
11]. In the case of the Akt-induced EMT, the cell is characterized by loss of apico-basolateral cell polarization and cell–cell adhesion, increased cell motility but decreased cell–matrix adhesion, and alterations in the distribution or production of specific markers [
25]. Akt usually becomes highly activated by its phosphorylation at both the Thy308 and Ser473 sites, which would decrease the expression of E-cadherin and the tight junction protein ZO-1 in cancer cells, increase metastasis in vivo, induce EMT [
25,
26]. An important contribution of PI3K/Akt activation to the aggressive phenotypes of tumor cells is further supported by the observation that the PI3K inhibitor LY294002 can reverse these effects [
27]. In line with this, we found that PI3K/Akt activation is a characteristic feature of the GC cells that have undergone an EMT. Although it is not clear whether a defective ERp29 is the sole determinant of the invasive phenotype, the capability of forced expression of ERp29 in reversal of the invasive phenotype and the enhancement of Akt and mTOR phosphorylation triggered by loss of ERp29 both indicate that ERp29 may play a critical role in this process. GSK3β is the kinase primarily responsible for phosphorylating β-catenin for degradation. GSK-3β phosphorylation by Akt on Ser9 inhibits its activity and prevents β-catenin from degradation [
28,
29]. A significant increase in the level of Ser9-phosphorylated and therefore inactive GSK-3β was found in the ERp29 knockdown cells implicating coordinate activation of Wnt/β-catenin signaling pathway while the exact mechanism of how ERp29 suppresses β-catenin activity remains to be elucidated. It may be noteworthy that GSK3β could regulates Snail and Slug phosphorylation for degradation respectively through β-Trcp-mediated ubiquitination and the CHIP pathway [
30,
31].
Intriguingly, while we have demonstrated that ERp29 depletion induces a significant enhancement of the GC cell migration and invasion by acquisition of EMT traits through activation of PI3K/Akt signaling pathway, it does not affect the proliferation rate of the cells. PI3K/Akt signaling is well known for its capacity to promote tumor progression by activating cell proliferation and growth as well as metastasis. However, it has also been reported that inhibition of PI3K/Akt in breast, non-small cell lung and gastric cancer cells restrains tumor invasion and metastasis independent of its growth inhibitory effects [
32‐
34]. Although the detailed mechanism is yet to be expounded, our observation that knockdown of ERp29 has a strong influence on GC cell migration and invasion rather than direct effects on proliferation further supports the notion that processes governing tumor metastasis downstream of PI3K/Akt may be separable from those driving proliferation.
Conclusions
In summary, this is the first study to date demonstrating that ERp29 may functionally serve as a tumor suppressor in GC. Downregulation of ERp29 in GC tumors and its strong correlation with tumor progression and patients’ prognosis merit further efforts to develop it as a diagnostic and prognostic biomarker. Furthermore, ERp29 may be exploited as a potential target in cancer therapy since it regulates PI3K/Akt and β-catenin pathways, which are well-established and recognized signaling events associated with malignant transformation.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.