Abstract
Background/Aim: Up-regulation of CD109 occurs in malignant tumours, although its role is unknown. Here we aimed to evaluate the significance of CD109 expression in oropharyngeal squamous cell carcinoma (OPSCC). Patients and Methods: Immunohistochemical analysis was performed to assess CD109 expression in 169 patients with OPSCC. We assessed the effects of small interfering RNA (siRNA)-mediated inhibition of CD109 expression on the proliferation and invasiveness of the human papillomavirus 16-positive (HPV16+) head and neck SCC cell line UM-SCC-47. Results: Expression of CD109 was associated with higher tumour differentiation in p16+ OPSCC (p=0.0036), and the CD109+ subgroup experienced significantly shorter progression-free survival (p=0.03). UM-SCC-47 cells with siRNA-mediated inhibition of CD109 expression showed reduced invasiveness (p=0.07). Conclusion: CD109 expression is associated with poor prognosis of HPV16+ OPSCCs.
CD109 is a glycosylphosphatidylinositol-anchored glycoprotein present on the surface of the cell membrane (1). It is usually not expressed in normal tissues, except in tissues with exceptionally high expression levels, such as vascular endothelium. Nevertheless, high CD109 expression occurs in diverse malignant tumours, including squamous cell carcinoma (SCC) (2–11). However, the role of CD109 in malignant tumours is insufficiently understood.
The therapeutic effects in and survival rates of patients with lung carcinoma and epithelial ovarian cancer decrease in association with high levels of CD109 expression (2, 12). However, few studies focused on the expression of CD109 in head and neck SCC (HNSCC). CD109 is highly expressed in oral, nasopharyngeal, and laryngeal cancer (13–15). Oral dysplastic lesions with high CD109 expression are at high risk of malignant transformation into carcinomas (13), and elevated serum levels of CD109 in patients with HNSCC are associated with lymph node metastasis and poor prognosis (16). However, the relationship between the expression of CD109 and prognosis in patients with oropharyngeal (OP) SCCs is unknown.
The status of human papillomavirus (HPV) infection in OPSCC serves as a biomarker for predicting the effects of treatment and prognosis of HNSCC (17, 18). However, no other useful biomarkers have been established (19). Therefore, new additional biomarkers are required to more effectively predict prognosis to enable the determination of an appropriate treatment strategy.
This study aimed to investigate the significance of the association between the expression of CD109 and the prognosis of OPSCC, and to decipher the underlying mechanism.
Patients and Methods
Patients. Among 192 patients with OPSCC treated at the Department of Otorhinolaryngology, Head and Neck Surgery, Kitasato University Hospital, Sagamihara, Kanagawa, Japan, between 2005 and 2016, 169 were enrolled, excluding two patients who did not agree to participate in the study and 21 patients without available SCC specimens. Seventy-five patients were mainly treated with surgery and 94 were mainly treated with chemoradiotherapy. All cases were histopathologically diagnosed at Kitasato University Hospital. Formalin-fixed paraffin-embedded tissue specimens collected through biopsy or surgery were analysed. Stages were classified according to the seventh edition of the Union for International Cancer Control TNM classification (2009) (20). Clinical data were acquired from patients’ medical records. Progression-free survival (PFS) was defined as the time from the end of treatment to disease progression or death.
Antibodies. Mouse monoclonal anti-human CD109 (11H3) antibody was purchased from Immuno-Biological Laboratory Co, Ltd. (Fujioka, Gunma, Japan). Mouse monoclonal anti-CD109 (C-9) antibody was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Anti-p16INK4a antibody (CINtec p16 Histology) was purchased from Roche Diagnostics K.K. (Tokyo, Japan). The horseradish peroxidase (HRP)-labelled polymer-binding secondary antibody (EnVision+ Dual Link rabbit/mouse) was purchased from Agilent (Santa Clara, CA, USA). Anti-β-actin (G043) antibody was purchased from Applied Biological Materials (Richmond, BC, Canada).
Immunohistochemistry (IHC). IHC was performed using formalin-fixed, paraffin-embedded 4-μm thick tissue sections. CD109 immunostaining was performed as follows: Sections were deparaffinised with the xylene substitute Neo-Clear (Merck KGaA, Darmstadt, Germany) and rehydrated with alcohol. The antigen was activated by immersing the sections in water at 98°C for 40 min in Tris-EDTA Buffer, pH 9.0. Endogenous peroxidase activity was blocked using 0.3% H2O2 in methanol for 15 min. The specimens were reacted with Protein Block Serum-Free (Agilent) at room temperature for 10 min and then with the anti-CD109 (11H3) antibody overnight at 4°C. The next day, samples were incubated with the secondary antibody for 30 min at room temperature. The reactions were visualised using diaminobenzidine (Agilent). Meyer’s haematoxylin was used to visualise nuclei.
IHC analysis of p16 expression was performed as follows: Sections were deparaffinised using Neo-Clear and rehydrated with alcohol. Endogenous peroxidase was blocked using 3% H2O2 in methanol for 5 min. The antigen was activated by boiling in a microwave oven for 15 min in Target Retrieval Solution, pH 6.0 (Agilent). The specimens were reacted with Protein Block Serum-Free (Agilent) at room temperature for 10 min and then incubated with an anti-p16 antibody overnight at 4°C. The next day, specimens were incubated with the secondary antibody for 30 min at room temperature. The reactions were visualised using diaminobenzidine (Agilent). Meyer’s hematoxylin was used to stain nuclei.
The expression levels of CD109 and p16 were evaluated by YM and ST, and MI and ST, respectively. Samples with ≥10% of tumour cells exhibiting membrane-associated and cytoplasmic CD109 staining were considered CD109+. In accordance with the eighth edition of the American Joint Committee on Cancer, >75% tumour cells exhibiting diffuse, strong p16 nuclear staining were judged p16+.
Cell lines. The human HPV16+ HNSCC cell lines UM-SCC-47 and UM-SCC-104, purchased from Merck KGaA, were maintained in Dulbecco’s modified Eagle’s medium (Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Life Technologies) and non-essential amino acids (Life Technologies). Cell lines were cultured at 37°C in an atmosphere containing 5% CO2.
Transfection using small interfering RNA (siRNA). For siRNA transfection, UM-SCC-47 cells (1.5×104/well) were added to a 24-well plate, cultured for 24 h and then transfected with a CD109 siRNA (siGENOME Human SMARTpool; Horizon Discovery, Cambridge, UK) or a negative-control siRNA (siGENOME non-Targeting siRNA #1; Horizon Discovery) (final concentration 10 nM) in the presence of Lipofectamine RNAiMAX Reagent (Thermo Fisher Scientific, Waltham, MA, USA). The efficiency of siRNA transfection was assessed using western blotting.
Western blotting. Cells were collected in 1.5-ml microcentrifuge tubes and homogenised in RIPA Buffer (50 mM Tris-HCl pH 7.6, 150 mM NaCl, 1% Nonidet P40 Substitute, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate) containing Protease Inhibitor Cocktail (Nacalai Tesque, Kyoto, Japan). Cells were homogenised and then centrifuged at 15,000 rpm for 10 min at 4°C. The cell lysates, in the supernatant, were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12.5% e-PAGEL-HR; ATTO, Tokyo, Japan), and the separated proteins were electrophoretically transferred onto an Immobilon-P membrane (Clear Blot Membrane-P plus; ATTO). Membranes were incubated with anti-CD109 (C-9, 1:100) or anti-β-actin (G043, 1:1,000) antibodies overnight at 4°C and then with secondary antibodies (EnVision+/HRP, Dual Link Rabbit/Mouse, HRP-conjugated; Agilent). Protein bands were visualised using a chemiluminescence detection system (Amersham ECL Prime Western Blotting Detection Reagent; Cytiva, Tokyo, Japan).
Cell proliferation assay. Cell proliferation was quantified using a Cell Counting Kit-8 (CCK-8; Dojindo Laboratories, Kamimashiki, Kumamoto, Japan). UM-SCC-47 cells (1.5×104/well) were cultured in a 24-well plate for 24 h and then transfected with CD109 siRNA or negative-control siRNA. The control (nontransfected) UM-SCC-47 cell line was added to a 24-well plate (1.5×104 cells/well) and then incubated in the same solution (without siRNA) as the transfected cell line. CCK-8 solution (50 μl) was added to each well at 0, 24, 48 and 72 h after transfection and plates were incubated at 37°C for an additional 2 h. The absorbance of samples at 450 nm was measured using a Varioskan LUX microplate reader (Thermo Fisher Scientific). Triplicate experiments were performed.
Invasion assay. Invasion assays were performed using a Corning Biocoat Matrigel invasion chamber (Corning, Tewksbury, MA, USA). UM-SCC-47 cells transfected with CD109 siRNA or negative-control siRNA were collected 48 h after transfection. UM-SCC-47 cells were resuspended in an FBS-free medium. The cell suspension (500 μl) adjusted to 1.0×104 cells/ml was added to the upper chamber, and the lower chamber was filled with medium (750 μl) containing 10% FBS. Cells were cultured for 44 h. A cotton swab was used to remove cells that did not transit the membrane from the top of the Matrigel. Cells that transited the membrane were fixed, visualised using Diff-Quik staining and counted using an inverted microscope. The invasion rate was expressed as the ratio of the number of cells that invaded through the Matrigel to the number of cells that migrated through the control membrane. Triplicate experiments were performed.
Statistical analysis. The associations between the levels of CD109 and p16 expression or clinicopathological features (data from medical records) were analysed using Fisher’s exact test or the Mann–Whitney test. The Kaplan–Meier method using the log-rank test was performed to analyse survival. The results of cell proliferation and invasion assays were analysed using Student’s (two-tailed) t-test; p<0.05 was considered significant. EZR software version 1.54 (Division of Haematology, Saitama Medical Centre, Jichi Medical University, Saitama, Japan) was used for all statistical analyses (21).
Declarations. This was a retrospective study conducted after the review and approval of the Institutional Review Board of the Kitasato University School of Medicine and Hospital (application number: B16-272). Information related to this study, including the option to opt-out, was posted on bulletin boards and on the Kitasato University Hospital’s website. Only clinical data and pathological specimens of patients who did not opt-out were used. The study was registered with the University Hospital Medical Information Network (UMIN, https://upload.umin.ac.jp) (registration number: UMIN 0041971).
Results
Patient characteristics. Patient characteristics are summarised in Table I. The study included 136 males and 33 females, with a median age of 64 years (range=30-90 years). The median observation period was 4.6 years (range=0.3-13.2 years). The subsite of the lesion was a lateral wall in 121 (71.6%) patients. Disease stages were as follows: 50 (29.6%) patients with early stages (0 to II) and 119 (70.4%) patients with advanced stages III and IV; and among patients in the latter stages, 109 (64.5%) had lymph node metastasis.
Expression of p16 and CD109 in OPSCC. IHC analysis of 169 OPSCC specimens revealed that 143 (84.6%) expressed CD109 and 90 (53.3%) expressed p16. Typical CD109 staining pattern of tumour tissue is shown in Figure 1. The immunoreactivity of CD109 was mainly localised to the cell membrane. There was no significant difference in p16 expression between the CD109+ and CD109− groups (p=0.20).
Relationship between CD109 expression and prognosis in OPSCC. Firstly, significant differences were not observed when we analysed PFS rates of patients treated with surgery or chemoradiotherapy (Figure 2A). Next, when we evaluated the prognostic significance of CD109 expression in all patients with OPSCC, there was no significant difference in PFS between CD109+ and CD109− groups (Figure 2B). We then evaluated the prognostic significance of CD109 expression in patients with OPSCC classified according to p16 status. The CD109+ group experienced significantly shorter PFS compared with the CD109− group in patients with p16+ OPSCC (p=0.03) (Figure 2C). There was not a significant relationship between CD109 expression and PFS in patients with p16− OPSCC (Figure 2D). These results suggest that the expression of CD109 was associated with poor prognosis of p16+ OPSCC.
Relationship between CD109 expression and clinicopathological features of patients with OPSCC. The relationships between CD109 expression and clinicopathological features were analysed in all patients with OPSCC (Table I). The CD109− group had a significantly larger number of poorly differentiated SCCs than the CD109+ group (p=0.0005). Furthermore, there were significantly more patients with lymph node metastases in the CD109− group than in the CD109+ group (p=0.02). The relationships between CD109 expression and clinicopathological features were then analysed according to p16 (Table II and Table III). The CD109−p16+ group included a significantly larger number of poorly differentiated SCCs than did the CD109+p16+ group (p=0.004). There was no significant difference in the clinical stage, lymph node metastasis, age of diagnosis, sex, subsite, treatment, drinking history, smoking history or performance status between the CD109+p16+ and CD109−p16+ groups.
These results suggest that the expression of CD109 was associated with the state of differentiation of p16+ OPSCCs, not with progression and metastasis (Table II). There was no significant relationship between CD109 expression and clinicopathological features of patients with p16− OPSCC, although the CD109−p16− group tended to have more poorly differentiated SCCs than the CD109+p16− group (p=0.08) (Table III).
Expression of CD109 in head and neck SCC cell lines and siRNA-mediated inhibition of CD109 expression. We employed western blotting to analyse the biological significance in vitro of CD109 expression in p16+ OPSCC using the human HPV16+ HNSCC cell lines UM-SCC-47 and UM-SCC-104. Expression of CD109 was detected in these cell lines as two bands corresponding to 180 kDa and 190 kDa (Figure 3A). When we analyzed the UM-SCC-47 cell line transfected with negative-control siRNA or CD109 siRNA, western blotting revealed reduced expression of CD109 in CD109 siRNA-transfected UM-SCC-47 cells compared with cells transfected with the negative-control or with the nontransfected cells (Figure 3B).
Effects of inhibiting CD109 expression on cell proliferation and invasion. When we analysed the effect of siRNA-mediated inhibition of CD109 expression on cell proliferation, there was no significant difference in proliferative potential between UM-SCC-47 cells transfected with the CD109 siRNA and those transfected with the negative-control siRNA or with nontransfected UM-SCC-47 cells (Figure 4). Next, we analysed the effects of inhibiting CD109 expression on cell invasion. UM-SCC-47 cells transfected with CD109 siRNA tended to be less invasive than those transfected with the negative-control siRNA (p=0.07) and were significantly less invasive than the control nontransfected UM-SCC-47 cells (p=0.008) (Figure 5). These results suggest CD109 promotes the invasiveness of UM-SCC-47 cells.
Discussion
High-level expression of CD109 occurs in SCCs of organs such as the lung, oesophagus, skin, gallbladder, and uterine cervix, as well as in malignant melanoma, urothelial carcinoma, certain breast carcinomas, epithelial ovarian cancer, certain haematopoietic tumours, and gliomas (2–11). In the head and neck region, high levels of CD109 are expressed in oral, nasopharyngeal, and laryngeal carcinomas (13–15). Moreover, oral dysplastic lesions that express high levels of CD109 frequently undergo malignant transformation to carcinomas (13).
Here, when we evaluated clinical specimens of OPSCC, we detected CD109 expression in 85% of cases. There was no significant correlation between CD109 expression and p16 expression. When we evaluated the expression of CD109 in HNSCC cell lines, we detected high levels of CD109 in the HPV16+ SCC cell lines UM-SCC-47 and UM-SCC-104. CD109 was detected as two bands of approximately 180 kDa and 190 kDa, consistent with other reports (22, 23). To our knowledge, this study is the first to demonstrate CD109 expression in OPSCC as well as the first report to investigate the correlation between CD109 and p16 expression in these tumours.
In oral, urothelial, skin and oesophageal carcinomas, the expression of CD109 is higher in well-differentiated than in poorly differentiated SCCs, suggesting that CD109 is required for differentiation in SCC (3, 4, 13, 24). Moreover, CD109 suppresses the acquisition of mesenchymal phenotypes in SCC cell lines by suppressing the epithelial–mesenchymal transition through inhibition of the transforming growth factor-β signalling pathway (25).
Our present clinical data show that the CD109− group had significantly more poorly differentiated SCCs than the CD109+ group in patients with p16+ OPSCC. Furthermore, we found that the CD109− group tended to have more poorly differentiated SCCs than the CD109+ group in patients with p16− OPSCC. These results are consistent with a study showing that CD109 suppresses the epithelial–mesenchymal transition (25). Further research is required to characterise the relationship between CD109 and the degree of differentiation of tumour cells.
Evidence indicates that high levels of CD109 expression in epithelial ovarian cancer, lung adenocarcinoma, glioblastoma, diffuse large B-cell lymphoma, and myxofibrosarcoma are associated with poor prognosis (2, 5, 6, 12, 26). However, the clinical relationship between CD109 expression in HNSCC and survival is undefined. Our clinical data show that the CD109+p16+ group experienced significantly shorter PFS than the CD109−p16+ group. Therefore, CD109 shows potential as a biomarker of poor prognosis of patients with p16+ OPSCC and may help identify those at high risk among such patients. There was no significant difference in PFS between the CD109+p16− and CD109−p16− groups, which is likely explained by an insufficient number of patients in the CD109−p16− group.
Here we investigated the association between CD109 expression and the clinical features of patients with p16+ OPSCC to determine why CD109+p16+ OPSCC has a poor prognosis. There was no difference in clinical stage and lymph node metastasis between the CD109+ and the CD109− groups. Thus, the expression of CD109 was associated with poor prognosis, although it did not appear to promote the progression or metastasis of p16+ OPSCC.
Furthermore, there was no significant difference between the CD109+ and CD109− groups associated with chemoradiotherapy or surgery as the main treatment. Among 17 patients who experienced disease progression, 10 and seven were treated with chemoradiation or surgery, respectively. These results suggest that the poor prognosis of CD109+p16+ OPSCC was not caused by differences in treatment methods.
To further investigate the poor prognosis of CD109+p16+ OPSCC, the effect of CD109 knockdown on the proliferative and invasive potential of HPV+ HNSCC cell lines was investigated in vitro. CD109 knockdown mediated by a CD109-specific siRNA did not change the proliferative capacity, although invasiveness was reduced. These observations suggest that CD109 promotes the invasion of HPV+ HNSCC cells, resulting in poor prognosis.
In glioblastoma-derived cell lines, CD109 was shown to promote epidermal growth factor receptor (EGFR) signalling and cell migration and invasion (22). Knockdown of CD109 expression also suppressed EGFR signalling in lung carcinoma cell lines (12). EGFR is expressed in 90% of HNSCCs (27), suggesting that EGFR and phosphorylated (activated) EGFR serve as biomarkers for assessing the prognosis of OPSCC (28). Moreover, EGFR signalling promoted cell invasion in vitro (29, 30). Together, these findings indicate that CD109 may promote the invasion of HPV+ HNSCC by promoting EGFR signalling.
CD109 is highly expressed in patients with ovarian epithelial cancer and chronic myelogenous leukaemia who are less responsive to chemotherapy (2, 31). However, we are unaware of any reports that demonstrate an association of therapeutic efficacy for p16+ OPSCC that expresses CD109. The possibility that CD109 expression in p16+ OPSCC is associated with poor therapeutic efficacy is consistent with our demonstration here that CD109+p16+ OPSCC was associated with poor prognosis. Determination of the underlying mechanism will require further investigations of the effects of CD109 on EGFR signalling and the effects of therapy.
Footnotes
Authors’ Contributions
ST designed this study. TY supervised throughout the study. ST, KM, and SM performed experiments and collected data. TK provided intellectual contribution in vitro studies. ST, MI, and YM evaluated the immunohistochemistry slides. All Authors were involved in data interpretation. ST wrote the article. YM and TY reviewed and revised the article.
Conflicts of Interest
There are no conflicts of interest to declare regarding this study.
- Received February 4, 2022.
- Revision received February 26, 2022.
- Accepted March 2, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.