Background
Ezrin, encoded by the
Vil2 gene, is a member of the ERM family; it provides a functional link between the plasma membrane and the cortical actin cytoskeleton of the cell. Ezrin plays important roles in cell motility, morphogenesis, adhesion, survival and apoptosis [
1‐
6]. It also participates in crucial signal transduction pathways [
7]. Ezrin binds to cell surface glycoproteins, such as CD43, CD44, ICAM-1 and ICAM-2, through interacting with their amino (N)-terminal domains. Ezrin also binds to filamentous actin through its carboxyl (C)-terminal domains [
8]. Ezrin has been linked to molecules that control the phosphatidylinositol-3-kinase, AKT, Erk1/2 MAPK and Rho pathways, which are functionally involved in signaling events regulating cell survival, proliferation and migration. Phosphorylation of ezrin induces its translocation from the cytoplasm to the plasma membranes of microvillus and confers the ability of binding to plasma membrane and actin filaments [
9‐
12].
Ezrin is expressed in a variety of normal and neoplastic cells, including many types of epithelial, lymphoid and glial cells [
5,
13,
14]. In melanoma cells, ezrin has been shown to be localized in phagocytic vacuoles, suggesting that its association with the actin cytoskeleton is crucial for the phagocytic activity [
15]. Phagocytic behavior is usually considered to be an indicator of high-grade malignancy in melanomas. In addition, immunohistochemical analysis has demonstrated a significant correlation between increased ezrin immunoreactivity and a high histological grade in astrocytoma [
16]. In a complementary DNA (cDNA) microarray analysis of highly and poorly metastatic rhabdomyosarcomas, ezrin was indicated to be a key regulator of metastasis [
17]. Ezrin overexpression has also been considered as an independent predictor of adverse outcome of gastrointestinal stromal tumors [
18]. These results indicated that ezrin expression level is closely associated with malignant progression of cancer.
Consistent with these reports, suppression of ezrin protein expression and disruption of its function significantly reduced lung metastasis in a mouse osteosarcoma model [
19]. Furthermore, high-level ezrin expression in canine osteosarcomas has been associated with early development of metastasis [
20]. Ezrin silencing by small hairpin RNA could reverse the metastatic behavior of human breast cancer cells [
21]. Taken together, the observed effects of ezrin overexpression and silencing on the cell malignant transformation indicate a role for ezrin in regulating tumor metastasis and progression [
22].
In pancreatic carcinomas, a high-level ezrin expression is associated with high metastatic potential; membrane translocation of ezrin might play a role in the progression from borderline tumor to malignant transformation. Patients with pancreatic ductal adenocarcinoma (PDAC) with membranous ezrin expression exhibited poorer prognosis compared to those without membranous ezrin expression, and ERM protein was more likely to be present in poorly differentiated cancers [
23‐
26]. A recent study showed that overexpression of pEzrin (Tyr353) in pancreatic cancers was associated with positive lymph node metastasis, less differentiation, pAkt overexpression and shorter survival times [
27]. Ezrin can interact with cortactin to form podosomal rosettes in pancreatic cancer cells, thereby playing a role in pancreatic cancer invasion [
28]. However, the mechanisms of ezrin-mediated tumor development still require further elucidation. In this study, we investigated the effect of ezrin on the motility and invasion ability of the pancreatic cancer cell line MiaPaCa-2, as well as the expression of ezrin in pancreatic duct adenocarcinoma, chronic pancreatitis and normal pancreatic tissues.
Materials and methods
Antibodies and plasmids
Rabbit polyclonal anti-ezrin antibody was purchased from Upstate technology (Lake Placid, NY). Rabbit polyclonal anti-phosphorylated Ezrin (Tyr353), mouse monoclonal anti-AKT, anti-phospho-AKT (Ser473), anti-p44/42 MAPK (Erk1/2) and anti-phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) antibodies were purchased from Cell Signaling Technology (Beverly, MA, USA). The mouse monoclonal antibody VSV-G (P5D4) was purchased from Roche Applied Science (Indianapolis, USA). The mouse monoclonal antibody GAPDH was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The secondary antibodies, including the rhodamine-conjugated goat anti-mouse, FITC-conjugated goat anti-mouse, horseradish peroxidase-conjugated anti-mouse and anti-rabbit antibodies, were purchased from ZhongShan Biotechnology (Beijing, China). The pcb6 vector that contains the cDNA encoding VSV-G-tagged ezrin was kindly provided by Dr. Monique Arpin [
14].
Plasmid-based silencing of ezrin expression
The mammalian expression vector, pSilencer 2.1-U6 (Ambion, Austin, Texas, USA) was used for expressing of siRNA in MiaPaCa-2 cells. Briefly, two primer pairs were synthesized, with the first pair encoding the nucleotides, GGGCCAAGTTCTACCCTGAAG (376-396, No. 1) followed by a 9 base "loop", TTCAAGAGA and an inverted repeat and the second pair encoding the nucleotides, GGCTTTCCTTGGAGTGAAA (849-867, No. 2) followed by the loop and the inverted repeat. A nonspecific 21-nucleotide siRNA scrambled to the first pair, GACCGAGTCCGAAGTCAGCT (No. 3) was used as a control. The primer pairs were annealed and inserted into the BamH I and Hind III sites of pSilencer 2.1-U6 and transformed into JM109 competent cells (Promega, Madison, WI, USA). Positive clones were identified and verified by restriction enzyme analysis and sequence analysis.
Cell culture and cell transfection
The pancreatic adenocarcinoma cell line MiaPaCa-2 (American Type Culture Collection, Manassas, Virginia, USA) was grown in DMEM (GIBCO, Grand Island, New Yolk, USA) supplemented with 10% fetal calf serum (FCS) and 1% L-glutamine (Invitrogen, Karlsruhe, Germany) and maintained at 37°C in 5% CO2. All transfections reactions were performed using Lipofectamine 2000 (Invitrogen; Carlsbad, CA) in accordance with the manufacturer's instructions. Stable transfectants were selected with 800 μg/mL G418 (Sigma-Aldrich, St. Louis, MO, USA), and individual clones were isolated.
Scanning electron microscopy
Cells were cultured on coverslips and harvested after 24 hours. Cells were then washed with phosphate buffered saline (PBS) and fixed with 2.5% glutaraldehyde at 4°C for 12 hours. After thoroughly washing with PBS, the fixed cells were dehydrated through an ethanol series and dried at room temperature. The samples were coated with a thin film of silver and examined under a scanning electron microscope (JEOL/JSM-6000F, JEOL Ltd., Tokyo, Japan).
Western blotting
Cell lysates (30 μg protein) resolved on 10% SDS-PAGE were transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA). For immunoblotting, we used antibodies against ezrin, VSV-G, phospho-ezrin(Tyr353), phospho-p44/42 MAPK(Erk1/2) (Thr202/Tyr204), p44/42 MAPK (Erk1/2), phospho-AKT (Ser473), AKT and GAPDH. The immunoreactive proteins were visualized using the ECL western blotting system (Amersham International, little Chalfont, UK), and densitometric analysis was performed using the Image Pro-Plus Software.
Indirect immunofluorescence
Cells were plated on glass coverslips for 24 hours, fixed with 3.7% paraformaldehyde for 20 minutes and then permeabilized with PBS containing 0.05% Triton X-100 for 10 minutes. The cells were then blocked with 1% BSA in PBS for 1 hour, followed by adding of primary antibodies diluted in blocking buffer at 4°C overnight at the following concentrations: anti-ezrin (serum was diluted 1:150) and anti-VSV-G (serum was diluted 1:75). Subsequently, the cells were washed with PBS and then incubated for 1 hour in either the goat-anti-mouse IgG TRITC-conjugated antibodies or the goat-anti-rabbit IgG FITC-conjugated antibody, both of which were diluted in the blocking buffer (1:60). Afterwards, 4',6-diamidino-2-phenylindole (DAPI) was used for nuclear counter-staining. Finally, the cells were mounted in the fluorescent mounting medium (Applygen Technologies Inc., Beijing, China) and viewed with under a fluorescence microscope (BH2-RFCA; Olympus Optical Co., Ltd, Tokyo, Japan).
Cell growth assay and flow cytometry analysis
In vitro cell growth was assessed using the Dojindo Cell Counting Kit-8 (Dojindo Laboratory, Kumamoto, Japan) according to the supplier's recommendations. Clones were plated in tissue culture plates at a density of 1 × 103 cells in 0.1 mL of culture medium per well and grown in DMEM with 10% FCS in 5% CO2 at 37°C. The number of cells per well was quantified by daily measurement of the absorbance at 450 nm for 7 days after plating. All experiments were performed in triplicate on three separate occasions. Replicate growth curves were plotted for each of the clones and compared to control cells grown under identical culture conditions. To determine the cell cycle distribution, 5 × 105 cells were plated in 60-mm dishes and cultured for periods of up to 2 days. The cells were then collected by trypsinization, fixed with 70% ethanol, washed with PBS, resuspended in 1 mL of 0.01 M PBS with RNase and 50 μg/mL propidium iodide, incubated for 20 minutes in the dark at room temperature and analyzed by flow cytometry using a FACS Calibur (Becton Dickinson, Bedford, MA).
An equal amount of 1% Noble agar solution pre-warmed to 40°C was added to DMEM containing 20% FCS pre-warmed to 37°C to make a 0.5% agar solution. After rapid mixing by inversion, the resultant solution was added to 24-well plates (0.5 mL/well). After reaching 70 to 80% confluence, the cells were trypsinized, washed with D-Hanks three times and diluted in Noble agar solution (0.35% Noble agar in DMEM with 10% FCS) at 37°C. The cell suspensions were then added into 24-well plate with a 0.5% agar layer (200 cells in 0.5 mL) (three wells per condition). The plates were incubated at 37°C with 5% CO2 for three weeks. The colony formation ability under each condition was assessed using untreated cells as control.
Transfilter migration and invasion assays
Transfilter assays were performed with 8.0-μm pore inserts in 24-well BioCoat Chambers (Becton Dickinson) using 5 × 104 cells in serum-free DMEM. The DMEM medium with 10% FCS was placed in the lower chambers as a chemoattractant. For invasion assays, Matrigel-coated transwell chambers were used. For migration and invasion assays, the cells were removed from the upper surface of the filter by scraping with a cotton swab after 12 and 24 hours in culture respectively. Migrated cells and invasive cells were fixed and stained with the crystal violet reagent. Mean values of the data obtained from three separate chambers were presented.
Tumor transplantation and spontaneous/experimental metastasis
Female BALB/c nude mice (body weight, 15 to 17 g) were bred under specified pathogen-free conditions (26°C, 70% relative humidity and a 12-h light/12-h dark cycle) in a germ-free environment with free access to food and water. To examine the effects of ezrin on tumor cell proliferation and metastasis in vivo, Mia ez22-B, Mia pcb6, Mia ezsi-scram and Mia ezsi-E (5 × 106 cells/100 μL normal sodium/mouse) were used. For spontaneous metastasis, the cells were injected into the inferior of pancreas capsule of the nude mice, whereas for experimental metastasis, the cells were injected into the tail vein of the nude mice. The mice were monitored every 2 to 3 days and sacrificed 10 weeks after injection. Tumors were excised, and metastasis in the lung, viscera, liver, draining lymph nodes and other organs were assessed. These tumors were embedded into paraffin. Histological analysis of the tissue sections stained with hematoxylin and eosin were performed to confirm the presence of metastasis in the various organs. Based on the gross and histological analyses, animals were assessed as positive or negative with respect to metastasis. Animal handling and experimental procedures were approved by the Peking Union Medical College Hospital animal experiments committee. It was also in accordance with the recommendations by the regional and country animal ethics committee.
Patients, specimens and immunohistochemistry
This study was approved by the Institutional Review Board of Peking Union Medical College Hospital, Chinese Academy of Medical Sciences (CAMS) and Peking Union Medical College (PUMC). Surgically resected specimens from 70 patients (age range, 29 to 78 years) with PDAC were examined. This patient population represented a randomly selected subgroup from a clinical series including all patients who underwent surgical resection between June 1998 and December 2005 in the Department of Surgery at Peking Union Medical College Hospital. The diagnosis of PDAC, histological grading and pathologic staging were re-evaluated and/or confirmed by two independent pathologists. PanIN lesions (n = 34) and CP (n = 28) were assessed and graded in the pancreatic tissues adjacent to the tumor in hematoxylin and eosin-stained slides.
Immunostaining for ezrin was performed using the primary rabbit polyclonal antibody against human ezrin (diluted 1:150) at 4°C overnight after antigen retrieval in 10 mM sodium citrate buffer (pH 6.0) for 15 minutes at 95°C, followed by incubation with an HRP-labeled anti-rabbit antibody for 1 hour. Immunostaining and clinicopathologic features were evaluated microscopically by two pathologists. Ezrin-specific immunoreactivity was scored by estimating the percentage of labeled tumor cells as follows: score 0, < 25% positive cancer cells; score +, 25-50% positive cancer cells; score ++, 50-75% positive cancer cells; and score +++, > 75% positive cancer cells. Specimens were considered positive for ezrin expression when the scores were + to +++ and were considered negative for ezrin expression when the score was 0. Pictures were collected using the MicroView MVC2000 image apparatus and software.
Statistical analysis
Each experiment was performed three to four times. All of the data were expressed as mean ± SD. Statistical analysis was performed using the Microsoft Excel software package. Comparisons between groups were conducted using Welch's t test. Correlation of ezrin immunoreactivity with clinicopathologic parameters were analyzed by Fisher's exact test. Differences were considered statistically significant at P < 0.05.
Discussion
Ezrin is the best characterized member in the ERM family; it shares the common membrane-binding N-terminal FERM domain with band-4.1 family members [
32]. Ezrin linking the cell membrane to actin cytoskeleton allows a cell to interact with its microenvironment and provides an "intracellular scaffolding" that facilitates signal transduction through a number of growth factor receptors and adhesion molecules [
2,
11,
33]. Positioned at the cell membrane-cytoskeleton interface, ezrin may be a nexus in the metastatic phenotype, playing a central, necessary and early role in the process of metastasis [
22]. Upon threonine and tyrosine phosphorylation, ezrin assumes an active, "open" conformation and, in turn, moves to the cell membrane and directly or indirectly tethers F-actin to the cell membrane. Ezrin resides at the nexus of multiple pathways regulating cellular behavior that can influence metastatic potential, including cell survival, motility, invasion and adherence. Ezrin participates in several crucial signal transduction pathways, including the MAPK, AKT, Rho kinase and CD44 pathways, promoting cytoskeletal reorganization and subsequent morphogenetic alterations [
3,
5,
8,
11]. High-level ezrin expression was observed in many tumor cell lines, such as breast carcinoma and rhabdomyosarcoma cell lines [
19‐
21]. Ezrin overexpression was also been observed in borderline lesions and pancreatic cancer tissues and associated with tumor malignant transformation and metastatic potential [
23‐
26]; however, its role and mechanisms remain elusive.
The invasion of cells into the surrounding tissue is a multi-step action that requires cell-cell contact, cell motility and degradation of the extracellular matrix by matrix metalloproteinases [
34,
35]. Here we demonstrated that ezrin was involved in the cytoskeleton modulation by SEM, showing the ezrin-induced changes in cell protrusions, cell microvilli and pseudopodia compared to the control cells. Consistent with these results, the chamber migration and invasion assays confirmed that ezrin expression could alter the cell migration and invasion abilities of pancreatic cancer cells. Ezrin is a cytoskeletal protein that might affect the assembly of cytoskeletal elements at the cytoplasmic face of the membrane and the nuclear skeleton, which would then facilitate cell migration and invasion. These results were in agreement with the previous reports demonstrating that changes in cytoskeleton might be a key factor in regulating neoplastic progression and tumor growth [
13,
22,
32,
36].
Our results showed that increased level of the ezrin protein was correlated with an increase in anchorage-independent growth of tumor cells, consistent with the previous finding in glioma cells [
13]. We also established experimental and spontaneous mice models and showed that ezrin overexpression could enhance tumor metastasis
in vivo, consistent with our observations in the cell motility/invasion and soft agar colony formation assays
in vitro.
Our results also showed that ezrin overexpression could induce metastasis in vivo in the spontaneous metastasis mouse model; however, ezrin silencing exerted no obvious effect on the metastatic potential of MiaPaCa-2 cells. These observations might be explained by the fact that MiaPaCa-2 cells is a cell line with low metastatic potential; therefore, the effect of ezrin silencing on metastasis may not be obvious. In addition, ezrin silencing might affect other signal pathway.
A striking feature of ezrin overexpression was the increased formation of surface protrusions that play essential roles in cell motility [
37]. Aside from the increased formation of protrusions, we proposed another possible mechanism for the effects of ezrin on cell invasion. Our results showed that phosphorylated Erk1/2 was markedly increased in the Mia ez22-B cells, although no significant changes were observed in the Mia ezsi-E cells compared to the controls. Activation of Erk1/2 by ezrin in MiaPaCa-2 cells indicates that the Erk1/2 MAPK pathway is one of the mediators of ezrin signaling. Therefore, the Erk1/2 pathway may also be involved in ezrin-induced cell motility, invasion and morphological changes in pancreatic ductal adenocarcinoma. This result is consistent with previous studies showing that the Erk1/2 pathway is involved in the reappearance of the actin cytoskeleton, allowing for the extension of ruffles into the active protrusions that are required for cell motility and, therefore contributing to the alteration of cell motility and invasion [
29,
30]. Although the phosphatidylinositol 3-kinase/Akt pathway is involved in ezrin-mediated cell survival [
11], we did not find corresponding evidence in either the ezrin-overexpressing or the ezrin-silencing MiaPaCa-2 cells. Although a recent study has shown that overexpression of pEzrin(Tyr353) in pancreatic cancers is associated with positive lymph node metastasis, less differentiation, pAkt overexpression, and shorter survival times [
27], we did not observe any change of ezrin phosphorylation in either the ezrin overexpressing or the ezrin silencing MiaPaCa-2 cells. Therefore, phosphorylation of ezrin may not affect the motility and invasion ability of MiaPaCa-2 cells
in vitro. Ezrin overexpression may be sufficient to confer metastatic potential [
38], and ezrin silencing may reverse metastatic behavior, through other ways [
21]. These underlying mechanisms require further elucidation.
Immunohistochemical analysis demonstrated that ezrin expression was elevated in PDAC samples compared to normal pancreatic tissues, which provided additional evidence supporting a functional role of ezrin in pancreatic cancer development. We also observed that ezrin was highly expressed in precancerous lesions, such as PanINs (97.1%, 33/34) and tubular complexes in CP (85.7%, 24/28). These observations have further significance. The detection and treatment of early-stage, non-invasive PanINs has a major impact on pancreatic cancer survival. PanINs are morphologically classified into three grades, according to nuclear polarity, nuclear size (pleomorphism) and hyper-chromatic staining [
39,
40]. Despite complex associations between tumor cells and PanINs, histological and molecular evidence suggests that PanINs can gradually progress to PDAC, bearing genetic traits such as Ras mutations, Cyclin D1 overexpression and loss of p16 expression [
41,
42]. Because CP is also an independent risk factor for PDAC development [
41], high ezrin expression proportion in CP and PanINs suggests that ezrin might be involved in the earliest stages of PDAC pathogenesis and could potentially serve as an indicator for those lesions progressing to the more advanced stage--PDAC. Ezrin was also expressed in the intercalated ducts in the pancreatic tissue that was adjacent to the adenocarcinoma, which was considered to be the origin in the pancreatic ducts and acini, as well as the starting point of PDAC development [
43,
44]. The results indicate that ezrin may play an important role in the early development of pancreatic ductal carcinoma.
In conclusion, we successfully overexpressed and silenced the ezrin protein expression in MiaPaCa-2 cells, and found that changes in the ezrin protein level were correlated with changes in the formation of dynamic cell protrusions, motility, invasion and the ability of anchorage-independent growth, which are all tumor cells features. Based on these results, we propose that ezrin, by participating in the formation of cell protrusions and the enhancement of anchorage-independent growth ability, might promote invasion and metastasis in carcinogenesis. These processes may be attributed to the activation of the Erk1/2 pathway [
29,
30]. The high ezrin expression proportion in CP, PanINs and ezrin expression in intercalated duct cells suggest that ezrin might be involved in the earliest stages of PDAC pathogenesis and could potentially serve as an indicator for those lesions progressing to the more advanced stage, PDAC. These results indicate that blocking ezrin function may represent a novel and effective strategy for preventing pancreatic cancer progression, invasion and metastasis. In pancreatic precancerous lesions, such as PanINs and chronic pancreatitis, blocking ezrin function may have therapeutic effects that prevent these two diseases from progressing to pancreatic cancer.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YM participated in the design of the study, performed experiments, analyzed the data and drafted the manuscript. ZL performed the immunohistochemical evaluations and participated in writing the manuscript. SY contributed to study design and conducted the animal studies. QZ performed the experiments, analyzed the data and drafted the manuscript. YM contributed to data analysis. JC planned the study, supervised the statistical calculations, performed the immunohistochemical evaluations, coordinated the study and drafted the manuscript. All authors read and approved the final manuscript.