Overexpression of BST2 is associated with nodal metastasis and poorer prognosis in oral cavity cancer
Objective: Bone marrow stromal cell antigen 2 (BST2) was one of the proteins which were found related to tumor metastasis in our previous proteomic study. Now we want to examine its clinical role on the oral cavity squamous cell carcinomas (OSCC).
Study Design: Individual retrospective cohort study and basic research.
Methods: Immunohistochemical analysis, Western blotting, and quantitative real-time polymerase chain reaction were used to demonstrate the expression levels of BST2 on 159 OSCC tumors. RNA interference was utilized for cell migration and proliferation study in vitro.
Results: BST2 expression was significantly higher in OSCC cells of metastatic lymph nodes and primary tumor cells, compared to adjacent normal epithelia. Higher BST2 expression was associated with positive N stage, advanced overall stage, perineural invasion, and tumor depth (P = 0.049, 0.015, 0.021, and 0.010, respectively). OSCC patients with higher BST2 expression had poorer prognosis for disease-specific and disease-free survival (P = 0.009 and 0.001, respectively). Multivariate analyses also demonstrated that higher BST2 expression is an independent prognostic factor of disease-specific and disease-free survival (P = 0.047 and 0.013, respectively). In-vitrosuppression of BST2 expression in OEC-M1 cells showed that BST2 contributes to tumor migration of OSCC cells.
Conclusions: The findings in this study indicate that BST2 expression in OSCC tumors is an independent prognostic factor of patient survival and associated with tumor metastasis.
Keywords: oral cancer; head and neck; BST2; metastasis; OSCC; tumor marker; squamous cell carcinoma.
Level of evidence: NA
Oral cavity squamous cell carcinoma (OSCC) is the most common head-and-neck cancer worldwide and remains a difficult malignancy to treat because of approximate 50% mortality rates in spite of recent advances in the contemporary management over the past three decades.1-4 Cervical lymph node dissemination has been long regarded as one of the major poor prognosticators in patient with OSCC and is a major obstacle to OSCC management. 5-8 Previously, some transcriptomic studies analyzing OSCC tumors have established the expression signatures of OSCC metastasis.9-13 On the other hand, our group also used a proteomic approach by isobaric tags for relative and absolute quantitation to identify proteins that are differentially expressed between laser capture microdissected primary and metastatic OSCC tumors and some potential markers of OSCC metastasis were therefore discovered.14 From the candidate proteins derived from by this approach, bone marrow stromal cell antigen 2 (BST2) was identified as one of the potential molecules associated with OSCC metastasis and need further validation in the clinical scenario.14
The BST2 gene is located on chromosome 19p13.2 and codes for a 35kDa transmembrane glycoprotein consisting 180 amino acids. It is also known as the HM1.24 antigen, CD317, and tetherin.15,16 BST2 has been found to be expressed on mature B cells but not on other normal tissues of health individuals.15 The BST2 overexpression has been identified in several cancer types such as multiple myeloma, endometrial cancer, gastric cancer, and glioblastoma multiforme.17-23 Moreover, BST2 was discovered associated with the tumor progression and metastasis by previous findings from the pancreatic endocrine tumors with liver metastases and the breast cancers with bone metastasis.24,25 In vitro, overexpression of BST2 increased invasion, migration, cell proliferation and apoptosis evasion in breast cancer cells.26,27 Thus, based on the findings of these previous studies and our proteomic discovery, we also hypothesized that BST2 is overexpressed in OSCC and plays a specific role in tumor cell modulation. Accordingly, our study is designed to investigate the expression and the role of BST2 in the OSCC tumors. Immunohistochemical analysis and quantitative real-time PCR were used to demonstrate the expression levels of BST2 on OSCC tumors. We also employed RNAi techniques to suppress the BST2 expression to investigate the effects of BST2 modulation on the OSCC cell line in vitro.
Materials and Methods
Patient Characteristics and Clinical Specimens
One houndred and thirty-eight male and 21 female were diagnosed as OSCC patients at the Chang Gung Memorial Hospital (Tao-Yuan, Taiwan) between 2002 and 2007 and enrolled in this study. Patient age at diagnosis ranged from 22.0 to 84.0 years (mean, 51.9±12.3). The associated subsites of the oral cavity were buccal mucosa (62 patients), gum (21), hard palate (5), lip (5), floor of the mouth (5), and tongue (61). Patients with at least one of the following conditions were considered ineligible: unresectable or inoperable cancer, other primary cancer (synchronous or metachronous), recurrent cancer, distant metastasis, prior history of malignancy, treatment with neoadjuvant therapy, medical contraindication for surgery, or individuals lost to follow-up. Lesions diagnosed as carcinoma in situ, verrucous carcinoma or a histologically basaloid subtype were also not included in the study. All patients provided informed consent prior to study participation, and the study was approved by the Institutional Review Board of Chang Gung Memorial Hospital. Patients underwent standard preoperative work-ups according to institutional guidelines, including detailed medical history, complete physical examination, computed tomography or magnetic resonance imaging scans of the head and neck, chest radiographs, bone scan, and abdominal ultrasound. Primary tumors were excised with adequate margins under intraoperative frozen section control. Surgical defects were immediately reconstructed via free flap or local flap by plastic surgeons, if necessary. Following surgical treatment, pathological TNM classification of all tumors was established according to the American Joint Committee on Cancer Staging Manual (2007). After discharge, all patients had regular follow-up visits every 2 months for the first year, every 3 months for the second year, and every 6 months thereafter.28-30
RNA Extraction and Quantitative Real-time RT-PCR Detection of BST2
According to the manufacturer’s protocol, total RNA of sixty paired OSCC tumor and adjacent normal tissues was extracted and purification by using RNAzol B reagent (Tel-Test, Friendwood, TX) and an RNeasy cleanup kit (Qiagen, Valencia, CA). 5ug of total RNA was used for first-strand cDNA synthesis and then adding to a reaction mixture comprising of commercially purchasable primers (BST2 Hs00171632_m1 and normalization control B2M, Hs00984230_m1 and ACTB, Hs99999903_m1 ; Assay-on-Demand, Applied Biosystems, Foster City, CA), TaqMan Universal PCR Master Mix, and RNase-free water. 7900 HT Sequence Detection System was used for performing quantitative real-time RT-PCR and the result was analyzed with SDS version 2 (Applied Biosystems, Foster City, CA). All experiments were repeated in duplicate, and the mean fold-change of each sample was calculated.
Western blot analysis
Proteins were extracted from culture cells with RIPA buffer (50 mM Tris pH 8, 0.0150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.2 % Na-deoxylate, 1x protease cocktail (Sigma-Aldrich, St. Louis, MO), and the concentrations determined using the BCA protein assay kit (Perice Biotechnology) Samples were separated on 12% SDS gels, transferred to PVDF membranes (GE Healthcare Life Sciences, Buckinghamshire, UK), and probed using rabbit monoclonal anti-BST2 antibody (Epitomics, Burlingame, CA), mouse monoclonal beta-actin antibody (MAB1501, Chemicon, Billerica, MA), and mouse monoclonal GAPDH antibody (Novus, Littleton, CO). The beta-actin and GAPDH signals were used as the loading controls. Densitometer quantitation was also measured to demonstrate the relative fold differences of prtein levels.
Enzyme-linked immunosorbent assay (ELISA) measurement
ELISA kit for human BST2 (CUSABIO, Hubei, China) was used for determining the BST2 levels in the tested samples. Human recombinant BST2 was used as the standard. Briefly, 100 μl of samples or standard were added to a murine monoclonal antibody against BST2 coated microtiter plates and incubated for 2 h at room temperature. After the plates were washing with wash butter for three times, a biotin-conjugated polyclonal antibody was added and further incubated for 1 h at room temperature. The plates were washed three times again and 100 μl of horseradish peroxidase-avidin was added at room temperature for 1 h. After the plates were washed five times and 90 μl of tetramethylbenzidine was added to each well. The reaction was terminated by adding of 2 N sulfuric acid, and the optical density of each well was measured by a microplate reader set to 450 nm. Each experiment was performed in triplicate.
For immunohistochemistry, formalin-fixed and paraffin-embedded tissues were cut into 4 ïm sections, deparaffinized, rehydrated, and prepared for antigen retrieval. Slides of consecutive sections were incubated with the appropriate antibodies: rabbit monoclonal anti-BST2 antibody (diluted 1:30, Epitomics, Burlingame, CA) at room temperature for 1 hour. After incubation, slides were washed three times with phosphate buffered saline, incubated with horseradish peroxidase polymer antibody (Invitrogen, Carlsbad, CA) at room temperature for 10 min, and developed by the addition of 3,3’-Diaminobenzidine tetrahydrochloride (DAB) reagent (Dako, Glostrup, Denmark) as the chromogen and hematoxylin as the counterstain. A ScanScope CT automated slide-scanning system (Aperio Technologies, Vista, CA) was used for obtaining images of stained slides. Expression of BST2 was scored using a combined scoring method accounting for both percentage of stained cell s and staining intensity and.31-33 Strong to negative staining intensities were scored from 3 to 0. For each intensity score, cells staining at that specific level were visually estimated and calculated as a percentage. The resultant combined score was calculated as the sum of the intensity scores multiplied by the percentage of stained cells. All specimens were evaluated independently by our pathologists (Liang Y and Hseuh C) who had blinded to clinical origin of the specimen.
Knockdown of BST2 Using RNA Interference (RNAi)
SMARTpool small interfering RNAs (siRNA) were purchased from Thermo Scientific (Dharmacon, Lafayette, CO) .RNAi specifically targeting human BST2 (No. L-011817-00-0005, Dharmacon) and a scrambled control RNAi (No. D-001810-10-05, Dharmacon) were purchased from Thermo Fisher Scientific (Rockford, IL). RNAi (at a final concentration 400 nM) was mixed with Lipofectamine RNAiMAXTM (Invitrogen, Carlsbad, CA) and Opti-MEM medium (Invitrogen, Carlsbad, CA) without serum, incubated for 20 min at room temperature, and then added to OEC-M1 cells that were seeded at a density of 1 × 105 cells per well in six-well plates. After incubation for 6 h at 37°C, transfer fresh culture medium (RPMI medium containing 10% FBS) was added to each well. After transfection for 48 h, cells were harvested for analysis of cell migration, and invasive capacity.
Cell Proliferation Assay
Cell proliferation ability was determined using the methylthiazoltetrazolium (MTT) assay (Bionovas Biotechnology, Toronto, Canada). After transfection for 24 h, cells were seeded at a density of 5 x103cells/well in RPMI medium containing 10% FBS. At the end of another 48-hour incubation period, the medium was exchanged with MTT solution at a final concentration of 1 mg/mL, and cells placed in a culture incubator at 37oC for 1 h. After washing twice with PBS, cells were solubilized with 0.1 ml of DMSO at 37oC for 1 h. The converted dye was measured at 540 nm. Three independent experiments were performed in quadruplicate. The average value of the control experiment was taken as 100% proliferation, and used to calculate the percentage of cell proliferation for each treatment.
Cell Migration Assay
Cell migration was evaluated using a chemotaxis chamber (Corning, Lowell, MA) with a polycarbonate membrane (8-μm pore size) placed between the two chambers. Transfected OEC-M1 cells (1 × 105) in 200 μl of serum free culture medium were applied to the upper chamber and 600 μl of RPMI medium containing 10% FBS medium was added to the lower chamber. Chambers were incubated at 37°C for 16 h, and then the membrane was fixed with methanol for 10 min and then stained with GIEMSA (Sigma-Aldrich, St. Louis, MO). Cotton swab was used for removing the cells on the upper surface of the filter, and the cells which migrated through the membrane were counted in eight different visual fields under a light microscope (magnification: 200x). Each migration assay was performed in triplicate during three independent experiments.
All statistical data display as means ± SD. The significance was examined by Wilcoxon test. The Wilcoxon signed ranks test was employed for comparison of the relative signal intensity of quantitative real-time RT-PCR and immunohistochemical staining scores between paired tumor and pericancerous normal mucosa samples. All patients had regular follow-up evaluations at our department until Apr 2012 or death. Survival analysis was plotted using the Kaplan-Meier method, and differences evaluated using the log-rank test. The specific risk factors for disease-specific survival were analyzed with multivariate regression. Statistical analyses were performed using SAS software (version 9.1; SAS institute, Cary, NC). All P values were two-sided, and statistical significance accepted at P<0.05.
Overexpression of BST2 in tumor cells of OSCC tissues.
Expression of BST2 was examined by quantitative real-time RT-PCR in 60 paired OSCC tumor and adjacent normal tissues. Transcripts for BST2 were significantly elevated in OSCC tumor specimens as compared with adjacent normal tissue (155 ± 205 vs. 38 ± 53, P < 0.0001; Supplementary Fig. 1A). The presence of BST2 was further confirmed by Western blot. As depicted in Supplementary Figure 1B, the BST2 protein was detected in 4 pairs of OSCC tissue lysate, and the BST2 expression significantly increased in the tumor tissues compared with the adjacent normal counterparts. To distinguish which cell types in the tumor mass expressed BST2, we performed immunohistochemical staining of tissue sections. BST2 was highly expressed in the cytoplasm of tumor cells, but was relatively absent from the infiltrating lymphocytes (Fig. 1A). Moreover, the paired adjacent normal epithelium samples showed lower or no expression of BST2 (Fig. 1B). Statistical analysis of the 133 paired samples available from these 159 patients demonstrated that BST2ï€ expressionï€ was significantly higher in tumor cells versus normal epithelial cells (133.5 ± 52.2 vs. 3.0 ± 21.1, P< 0.0001; Fig. 1B). Furthermore, BST2 levels appeared to be higher in 27 metastatic tumors of lymph nodes, compared to those in the corresponding primary tumors (157.8 ± 40.2 vs. 133.5 ± 52.2, P= 0.533; Fig. 1B), indicating that BST2 is more highly expressed in tumor cells of metastatic lymph nodes and primary tumor cells and almost absent in normal oral epithelia.
Association of BST2 expression with various clinicopathological manifestations
Next, we evaluated the relationships between increased BST2 expression and various clinicopathological characteristics of OSCC patients (Table 1). Higher BST2 expression was significantly associated with higher pN status, advanced overall stage, positive perineural invasion, and greater tumor depth (P=0.049, 0.015, 0.021, and 0.010, respectively; Table 1). However, we observed no association between BST2 overexpression in OSCC tumors and patient age, sex, T stage, differentiation or bone invasion. Consistent with our hypothesis, BST2 overexpression was significantly (P=0.049) associated with nodal metastasis (pN status).
Association of BST2 expression with overall survival (OS), disease-specific survival (DSS), and disease-free survival (DFS)
Based on expression data obtained from IHC, patients were stratified into two groups (high vs. low expression using 160 out of 300 as the cut-off value), and the possible association of BST2 expression with patient OS evaluated. Survival analysis revealed that the five-year OS rates for patients stratified into high and low BST2 expression subgroups were 66.8% and 51.2%, respectively. These differences in OS were not significant, compared in the log-rank test (P=0.059; Figure 2A). However, the Kaplan-Meier plots evaluated 5-year DSS rates for patients stratified by high and low BST2 expression as 74.0% and 53.7%, respectively. These differences in DSS were statistically significant, as observed with the log-rank test (P= 0.009; Figure 2B). Moreover, five-year DFS rates for patients stratified based on high or low BST2 expression were also significantly different in the log-rank test (75.9% and 51.2%, respectively; P= 0.001) (Figure 2C). BST2 expression was additionally a significant predictor of DFS and DSS in univariate analysis with the Cox proportional regression model. To further ascertain whether BST2 expression can be applied as an independent predictor of patient survival, multivariate analysis was performed using age, gender, pT status, pN status, overall stage, perineural invasion, tumor differentiation and BST2 expression as parameters in the Cox proportional regression model. Our results indicated that pT status, pN status and BST2 expression are independent predictors of DSS (P= 0.017, 0.002, and 0.047, respectively; Table 2). Similarly, we also found that pT status, pN status and BST2 expression are independent predictors of DFS (P= 0.014, <0.001, and 0.013, respectively; Table 3).
BST2 Promotes OSCC Cell Migration In Vitro
To evaluate the biological significance of BST2 overexpression in OSCC using an in vitro system, endogenous expression of BST2 in OSCC cells was knocked down using gene specific RNAi. The effects of RNAi were determined by the ELISA method in the supernatants of OEC-M1 cells transfected with either BST2-specific RNAi (si-BST2) or a scrambled sequence control RNAi. As shown in Supplementary Figure 2A, the levels of endogenous BST2 was significantly reduced (P=0.003; Supplementary Figure 2A) in si-BST2–transfected cells as compared with cells transfected with the control RNAi. Control and si-BST2–transfected cells were further analyzed for cell proliferation and migration. As shown in Supplementary Figure 2B, the cell proliferation ability in OEC-M1 cells was not significantly decreased in the si-BST2–transfected cells compared to the control RNAi transfected cells (P =0.435; Supplementary Fig 2B). However, the capability of OEC-M1 cell migration attenuated by addition of si-BST2 (70.2% reduction, P = 0.001; Supplementary Figure 2C). Collectively, these findings indicated that overexpression of BST2 in vitro can mediate cell migration in OSCC cells.