Overexpression of Golgi Phosphoprotein 2 Is Associated With Poor Prognosis in Oral Squamous Cell Carcinoma

Overexpression of Golgi Phosphoprotein 2 Is Associated With Poor Prognosis in Oral Squamous Cell... Abstract Objectives The aims of this study were to investigate the relationship between Golgi phosphoprotein 2 (GOLPH2) and oral squamous cell carcinoma (OSCC) and explore the clinical significance of GOLPH2 in OSCC. Methods Tissue microarrays from human OSCC samples were stained for GOLPH2 expression and clinicopathologic features. Kaplan-Meier analysis was used to compare the survival of patients with high GOLPH2 expression and patients with low GOLPH2 expression. Results We found GOLPH2 is highly expressed in OSCC tissue, and the GOLPH2 expression in metastatic lymph nodes is higher than in tumor tissue. Our data indicate that patients with higher GOLPH2 expression have poor overall survival compared with those with lower GOLPH2 expression. This study demonstrated that GOLPH2 was associated with CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Conclusions These findings suggest GOLPH2 is a potential marker for estimating the patient’s prognosis and may be a target for molecular-targeted therapy against OSCC. Cancer stem cell, Epithelial-mesenchymal transition, GOLPH2, Immune checkpoint, Oral squamous cell carcinoma Oral squamous cell carcinoma (OSCC) accounts for more than 90% of oral cancers (occurring in the mouth, lip, and tongue) and is the eighth most common cancer in humans. OSCC accounts for approximately 2.5% of all new cancer cases and 1.9% of all cancer deaths annually.1,2 Many important risk factors have been identified so far for OSCC, including immune status, smoking, alcohol use, and the synergistic effects of tobacco and alcohol.1,3 Despite recent advances in cancer treatment methods, the 5-year survival rate of OSCC is approximately 50% in the advanced stage of the disease.4,5 Although there are conventional tumor parameters, such as tumor size, nodal status, and the existence of systemic metastasis, new molecular markers are warranted to provide more information on the tumor biology, to allow a better prognosis and possibly predict the stratification of patients. Golgi phosphoprotein 2 (GOLPH2), also known as Golgi membrane protein 1 (GOLM1) or GP73, is a Golgi phosphoprotein of yet unknown function. The 73-kDa Golgi apparatus–associated protein is encoded by the GOLM1 gene and was first described in the liver tissue of a patient with giant cell hepatitis.6 The structure includes several areas of possible glycosylation.7 Due to its localization at the Golgi apparatus, the proposed functions include protein modification, cell signaling, intracellular transport function, or merely local structural tasks.8 Initially, GOLPH2 upregulation was reported in neoplastic and nonneoplastic liver pathologies, and many studies postulated that its serum level is a novel marker of hepatocellular carcinoma.9 Subsequently, enhanced GOLPH2 expression was reported in other cancer settings such as gastric cancer, esophageal cancer, cutaneous melanoma, renal cell cancer, and prostate cancer.10-14 GOLPH2 has been implicated in several important pathologic processes, including cancer cell proliferation and metastasis.15,16 The epithelial-mesenchymal transition (EMT) is a process by which epithelial cells lose their cell polarity and cell-cell adhesion and is a key developmental program that is often activated during cancer invasion and metastasis.17 Cancer stem cells (CSCs) represent a minor subpopulation of tumor cells that share some features with the normal stem cells of the tissue from which tumor derives and have the properties of self-renewal, multiple differentiations, and tumor initiation.18 CSCs cause relapse and metastasis by giving rise to new tumors and are therefore tumorigenic in contrast to other nontumorigenic cancer cells.19 This property conferred by the EMT is thus a doubly dangerous duo for the patient, as it not only enables the carcinoma cells to enter the bloodstream but also increases the tumorigenic and proliferative potential of these cells.20 CD44, SOX2 and Slug, the marker of CSCs and EMT, were correlated with OSCC in previous studies.21-23 The immune system plays a key role in the development, establishment, and progression of cancer.24 In a recent study, GOLPH2 was proved to have a correlation with the immune system and be a novel regulator of interleukin 12 production and macrophage polarization.16 Immune checkpoint pathways can be coopted by cancer to evade immune destruction, particularly targeting T cells.25 In our previous studies, we demonstrated the expression status of B7-H3, B7-H4, TIM3, and VISTA, four of the inhibitory immune checkpoints in OSCC,26-29 but the correlation between GOLPH2 and immune checkpoint was unclear. Although the expression of GOLPH2 has been studied in several types of cancer, the association between GOLPH2 and OSCC remains unknown. The central aim of this study was to evaluate the potential diagnostic and prognostic value of GOLPH2. Materials and Methods Detailed procedures are provided in the Supplementary Materials and Methods (all supplemental materials can be found at American Journal of Clinical Pathology online). Tissue Microarray Construction Full ethical approval was granted by the School and Stomatology of Wuhan University Medical Ethics Committee (principal investigator: Zhi-Jun Sun) to use patient samples in this study. The specimens of the tissue microarrays were obtained from January 2008 to August 2015 at the Department of Oral and Maxillofacial Surgery, School and Hospital of Stomatology, Wuhan University. In the present study, all patients were diagnosed with OSCC by two independent pathologists. Clinical follow-up data, as annually assessed for survival time, were available for all patients. All information, including hospital records, histology slides, and pathology reports, was retrieved and reviewed. The clinical stage of the OSCC was classified according to the guidelines of the Union for International Cancer Control.30 The grading scheme of the World Health Organization was used to determine the histologic grading. The OSCC cancer cohort consisted of 43 normal oral mucosae, 48 oral epithelial dysplasias, 165 patients with primary OSCC (excluding recurrent, presurgical inductive chemotherapy and presurgical radiotherapy), 12 patients with recurrent OSCC, 10 patients with OSCC who received presurgical radiotherapy (without presurgical chemotherapy), and 17 patients with OSCC who received presurgical TPF (docetaxel, cisplatin, and fluorouracil) inductive chemotherapy without presurgical radiotherapy by previous protocol.31 Forty-one pathologically confirmed metastatic lymph nodes from 41 neck dissection specimens were constructed in the tissue microarray (TMA). Clinical features, including TNM stage, histologic grade, and overall survival, were available for all cases. Human papillomavirus (HPV) infection status was diagnosed by p16 immunostaining and the HPV DNA in situ hybridization technique. Immunohistochemistry and Evaluation The TMAs were performed on 4-μm-thick paraffin sections, which were then subjected to deparaffinization and rehydration. For epitope unmasking, the sections were boiled in 0.01 mol/L citric acid buffer solution (pH 6.0) for 1.5 minutes at high pressure. Then, the sections were blocked for endogenous peroxidase activity by incubating sections in 3% hydrogen peroxide solution for 20 minutes at room temperature, and 10% goat serum was added to the sections to block nonspecific binding. The TMAs were incubated overnight at 4°C in a humidified chamber with primary antibody for GOLPH2 (1:500; GeneTex, San Antonio, TX), CD44 (1:100; Cell Signaling Technology, Danvers, MA), SOX2 (1:300; Cell Signaling Technology), Slug (1:200; Cell Signaling Technology), B7-H3 (1:800; Cell Signaling Technology), B7-H4 (1:800; Cell Signaling Technology), TIM3 (1:400; Cell Signaling Technology), VISTA (1:400; Cell Signaling Technology), or isotype-matched immunoglobulin G controls. A positive slide was created for each experiment. Subsequently, a secondary biotinylated IgG antibody solution and an avidin-biotin-peroxidase reagent were added to the slides. Peroxidase activity was detected with a 3,3′-diaminobenzidine solution. Then, the slides were counterstained by immersion in hematoxylin for 1 to 2 minutes. Slides were scanned by the ScanScope CS scanner (Aperio, Vista, CA) and then processed with background subtraction and white balance. The membrane, nuclear, or pixel immunohistochemical staining was quantified using Aperio Quantification software. For each sample, fields of high expression were scanned, and the membrane and nuclear immunostaining were calculated by the followed formula: (1 × percentage of weakly positive staining) + (2 × percentage of moderately positive staining) + (3 × percentage of strongly positive staining). Then, the scores were normalized between 0 and 300. The histoscores were converted to the range of –3 to 3 using Microsoft Excel (Microsoft, Redmond WA). Cluster 3.0 (Stanford University, Stanford, CA) with average linkage was used for hierarchical analysis, and the results were visualized using Java TreeView 1.1.3 (Stanford University). Statistical Analysis The data analyses were performed by GraphPad Prism version 5.0 for Windows (GraphPad Software, La Jolla, CA). One-way analysis of variance (ANOVA) followed by post-ANOVA Tukey multiple-comparison tests and unpaired t tests were used to analyze the differences among immunohistochemical staining within each group. The best cutoff of survival curves was calculated from the website Cutoff Finder (http://molpath.charite.de/cutoff/index.jsp) by a previously reported protocol.32 The survival of the patients in the two groups was compared using a Kaplan-Meier analysis and log tests. Data were graphically represented as the mean values ± SEM. The Cox proportional hazards regression model was used for multivariate analysis to assess the significance of overall differences. The statistical significance level for all comparisons was set at P < .05. Results Expression of GOLPH2 in OSCC Is More Activated Than in Normal Mucosa and Dysplasia and Is Associated With Patient Survival To determine whether overexpression of GOLPH2 was correlated with OSCC, the Oncomine database, a publicly available cancer data set, was used. In studies by Toruner et al33 and Ginos et al,34 messenger RNA (mRNA) expression indicated that the mRNA expression of GOLPH2 was significantly increased in OSCC compared with normal squamous cells (P < .005, Supplementary Figure 1). Then, the tissue microarrays were used to evaluate the expression in OSCC tissues. Immunostaining revealed that cytoplasmic patterns of GOLPH2 staining were mainly found in the tumor cells, and partial expression was found in the stromal components Figure 1A, which is consistent with a previous study about GOLPH2 expression in other types of cancer.9-13 In addition, we detected that GOLPH2 had a stronger staining intensity in OSCC than in dysplasia (n = 43, P < .01) and in normal mucosae (n = 48, P < .005; Figure 1B). Subsequently, we used Cutoff Finder to find the optimized cutoff by a previously reported protocol,32 and a histoscore of 120.6 was calculated for the analysis as the best cutoff for survival. The Kaplan-Meier survival analysis in Figure 1C indicated that GOLPH2 expression was significantly correlated with poor overall survival in patients with OSCC at the best cutoff (P = .0005). When the median expression was used as the cutoff, a significant prognostic value for patient survival was not detectable, but a trend for longer survival times of patients with lower GOLPH2 expression was apparent in the Kaplan-Meier curve (Supplementary Figure 2E, P > .05). A multivariate Cox proportional hazard model using variables that included sex, age, pathologic grade, tumor size, lymph node stage, and GOLPH2 expression identified variables as prognostically significant by univariate analysis and demonstrated that high expression of GOLPH2 was a predictor of reduced survival (P = .034; Table 1). Table 1 Multivariate Analysis for Overall Survival in Patients With Primary Oral Squamous Cell Carcinomaa Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b CI, confidence interval; GOLPH2, Golgi phosphoprotein 2; HR, hazard ratio. aCox proportional hazards regression model. bP < .05. View Large Table 1 Multivariate Analysis for Overall Survival in Patients With Primary Oral Squamous Cell Carcinomaa Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b CI, confidence interval; GOLPH2, Golgi phosphoprotein 2; HR, hazard ratio. aCox proportional hazards regression model. bP < .05. View Large Figure 1 View largeDownload slide Increased expression of Golgi phosphoprotein 2 (GOLPH2) in human oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of GOLPH2 in OSCC tissue compared with normal mucosa (×4 and ×20). B, Quantification of the immunohistochemical histoscore of GOLPH2 in oral squamous cell carcinoma (OSCC, n = 165) compared with dysplasia (DYS, n = 48) and oral mucosa (MUC, n = 43). C, Kaplan-Meier curve according to low (n = 136) and high (n = 29) expression of GOLPH2. The best cutoff (120.6) was used. *P < .01. **P < .001. Figure 1 View largeDownload slide Increased expression of Golgi phosphoprotein 2 (GOLPH2) in human oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of GOLPH2 in OSCC tissue compared with normal mucosa (×4 and ×20). B, Quantification of the immunohistochemical histoscore of GOLPH2 in oral squamous cell carcinoma (OSCC, n = 165) compared with dysplasia (DYS, n = 48) and oral mucosa (MUC, n = 43). C, Kaplan-Meier curve according to low (n = 136) and high (n = 29) expression of GOLPH2. The best cutoff (120.6) was used. *P < .01. **P < .001. Expression of GOLPH2 in Metastatic Lymph Nodes Is Higher Than in Primary Tumor Tissue We observed that GOLPH2 had a stronger immunoreactivity in metastatic lymph nodes than in primary tumor tissue Figure 2A. The quantification also confirmed this result (P < .01; Figure 2B). On further assessment, GOLPH2 was not significantly associated with pathologic grade (I, II, and III; P > .05; Figure 2C). There was no significant difference in tumor size (T1 + T2 vs T3 + T4; P > .05; Figure 2D) and lymph node metastasis status (N1 vs N2 + N3; P > .05; Figure 2E), but we found a slight association from both parameters in the graph. Subsequently, no significant differences were observed for GOLPH2 expression from samples with different HPV infection status (P > .05; Figure 2F). There was no significant correlation between GOLPH2 expression in primary OSCC and preoperation inductive TPF (P > .05; Figure 2G), preoperation radiotherapy (P > .05; Supplementary Figure 2A), and recurrent OSCC (P > .05; Supplementary Figure 2B). Furthermore, no significant differences in GOLPH2 expression were observed for tobacco smoking (P > .05; Supplementary Figure 2C) and alcohol consumption (P > .05; Supplementary Figure 2D) in primary OSCC. Figure 2 View largeDownload slide Higher expression of Golgi phosphoprotein 2 (GOLPH2) in oral squamous cell carcinoma (OSCC) tissue than metastatic lymph nodes. A, A representative patient (patient 29) for whom the immunochemical histoscore was greater in OSCC tissue than in the metastatic lymph node (LN) (×4 and ×20). B, Histoscore of GOLPH2 in OSCC tissue (n = 165) compared with tissue of metastatic LNs (n = 41). *P < .01. C, Quantification of the immunohistochemical histoscore of GOLPH2 by pathologic grade (I, n = 43; II, n = 84; and III, n = 38). D, Quantification of immunochemical histoscore of GOLPH2 between T1 + T2 (n = 115) and T3 + T4 (n = 50). E, Quantification of the immunohistochemical histoscore of GOLPH2 between N1 (n = 105) and N2 + N3 (n = 60). F, Quantification of immunohistochemical histoscore of GOLPH2 between human papillomavirus (HPV)–positive (n = 16) and HPV-negative (n = 149) samples. G, Quantification of immunohistochemical histoscore of GOLPH2 in primary OSCC (n = 165) compared with preoperative inductive TPF (docetaxel, cisplatin, and fluorouracil) (n = 17). All data are presented as the means ± SEM. NS, not significant. Figure 2 View largeDownload slide Higher expression of Golgi phosphoprotein 2 (GOLPH2) in oral squamous cell carcinoma (OSCC) tissue than metastatic lymph nodes. A, A representative patient (patient 29) for whom the immunochemical histoscore was greater in OSCC tissue than in the metastatic lymph node (LN) (×4 and ×20). B, Histoscore of GOLPH2 in OSCC tissue (n = 165) compared with tissue of metastatic LNs (n = 41). *P < .01. C, Quantification of the immunohistochemical histoscore of GOLPH2 by pathologic grade (I, n = 43; II, n = 84; and III, n = 38). D, Quantification of immunochemical histoscore of GOLPH2 between T1 + T2 (n = 115) and T3 + T4 (n = 50). E, Quantification of the immunohistochemical histoscore of GOLPH2 between N1 (n = 105) and N2 + N3 (n = 60). F, Quantification of immunohistochemical histoscore of GOLPH2 between human papillomavirus (HPV)–positive (n = 16) and HPV-negative (n = 149) samples. G, Quantification of immunohistochemical histoscore of GOLPH2 in primary OSCC (n = 165) compared with preoperative inductive TPF (docetaxel, cisplatin, and fluorouracil) (n = 17). All data are presented as the means ± SEM. NS, not significant. Protein Expression of GOLPH2 Was Conspicuously Associated With CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA in OSCC Tissue We focused on specific antibody expression in human OSCC tissue microarrays. In previous studies, CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA had high expressions in OSCC compared with normal tissue.26-29,35-37 As shown in Figure 3A, higher expressions of these molecules were confirmed in our TMAs. Immunohistochemical staining showed that CD44 and SOX2, markers of CSCs, were overexpressed in tumor tissue (Figure 3A). Slug, an EMT marker, was overexpressed in the TMAs (Figure 3A). TIM3 and VISTA were highly expressed in the tumor-infiltrating immune cells in the tumor microenvironment (Figure 3A). B7-H3 and B7-H4 were highly expressed in the immune cells and tumor cells (Figure 3A). Hierarchical clustering further confirmed that GOLPH2 expression was closely associated with B7-H4 expression Figure 3B. Interestingly, the Pearson correlation coefficient test revealed that GOLPH2 expression was statistically associated with CD44 (P < .0001, r = 0.3855), SOX2 (P = .0005, r = 0.2692), Slug (P = .0005, r = 0.2662), B7-H3 (P = .0006, r = 0.2635), B7-H4 (P < .0001, r = 0.5939), TIM3 (P < .0001, r = 0.4520), and VISTA (P = .0002, r = 0.2909) Figure 4. Figure 3 View largeDownload slide High expression of CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA in oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of Golgi phosphoprotein 2 (GOLPH2), CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA (×4 and ×20). B, Hierarchical clustering of GOLPH2, CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA immunohistochemical results in human OSCC with statistics (n = 196). Figure 3 View largeDownload slide High expression of CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA in oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of Golgi phosphoprotein 2 (GOLPH2), CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA (×4 and ×20). B, Hierarchical clustering of GOLPH2, CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA immunohistochemical results in human OSCC with statistics (n = 196). Figure 4 View largeDownload slide Golgi phosphoprotein 2 (GOLPH2) has a positive correlation with CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Correlations among GOLPH2 expression and CD44 (A), SOX2 (B), Slug (C), B7-H3 (D), B7-H4 (E), TIM3 (F), and VISTA (G) expression in human oral squamous cell carcinoma tissue microarrays. In dot plots, each dot represents a specimen according to the data (n = 165) from the issue microarrays. A, P < .0001, r = 0.3855. B, P = .0005, r = 0.2692. C, P = .0005, r = 0.2662. D, P = .0006, r = 0.2635. E, P < .0001, r = 0.5939. F, P < .0001, r = 0.4520. G, P = .0002, r = 0.2909. Figure 4 View largeDownload slide Golgi phosphoprotein 2 (GOLPH2) has a positive correlation with CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Correlations among GOLPH2 expression and CD44 (A), SOX2 (B), Slug (C), B7-H3 (D), B7-H4 (E), TIM3 (F), and VISTA (G) expression in human oral squamous cell carcinoma tissue microarrays. In dot plots, each dot represents a specimen according to the data (n = 165) from the issue microarrays. A, P < .0001, r = 0.3855. B, P = .0005, r = 0.2692. C, P = .0005, r = 0.2662. D, P = .0006, r = 0.2635. E, P < .0001, r = 0.5939. F, P < .0001, r = 0.4520. G, P = .0002, r = 0.2909. Discussion In the present study, our results demonstrated that GOLPH2 was highly and aberrantly expressed in human primary OSCC tissue compared with dysplasia and normal mucosa. Furthermore, we observed GOLPH2 was more highly expressed in metastatic lymph nodes than in primary tumor tissue. Survival analysis indicated that patients with high GOLPH2 expression had poor overall survival compared with those with low GOLPH2 expression. In addition, the data in the present study demonstrated that the expression of GOLPH2 correlated with that of CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Our findings support that GOLPH2 may be a prognostic indicator for OSCC. To our knowledge, this is the first systematic study of GOLPH2 expression in OSCC in the clinical setting. Cell invasion is a major feature of malignancy, and the Golgi apparatus plays an important role in cell migration and acts as a hub for different signaling pathways.38 Therefore, elucidation of the regulatory networks of the Golgi apparatus will provide important information for the development of therapeutic strategies against cancer migration, invasion, and metastasis. GOLPH2, a cis-Golgi-localized protein with unknown cell function, is overexpressed and usually an indicator for poor prognosis in several types of cancer.9-14 Previous studies have indicated that GOLPH2 plays a role in regulating tumor cell migration and invasion, important functions for the metastatic process in many kinds of cancer.14,39,40 A recent study41 has shown that GOLPH2 promotes hepatocellular carcinoma invasion by the activation of matrix metalloproteinases (MMPs), which play a crucial role in degradation of basement membranes and the extracellular matrix. Furthermore, studies have proved that MMP molecules are correlated with OSCC metastasis.42-44 Previous results have indicated that adhesion molecules such as CD44 are involved in positioning activated MMPs on the cell surface of invasive tumor cells.45 A recent study confirmed that GOLPH2 expression leads to increased expression of EMT-related proteins and that GOLPH2 silencing reduces hepatocellular carcinoma cell migration in vitro.40 Molecular profiling has shown that cancer cells and tumors that exhibit the CSC phenotype also express genes associated with the EMT.46 In the present study, GOLPH2 expression was associated with CSC markers (CD44 and SOX2) and an EMT marker (SLUG). CSCs play an important role in tumor metastasis. CD44, a transmembrane glycoprotein and a key cell surface marker of CSCs in head and neck carcinoma, has possessed unique properties in functional assays for CSC self-renewal and differentiation.47,48 SOX2, which also has been recognized for causing “stemness” characteristics in cancer cells, can lead to cell immortality and account for the self-renewal and invasive properties of cancer cells.49,50 Slug, a C2H2-type zinc finger transcription factor that is also an EMT marker, has been identified in aggressive cancers.51 Our data support this point, as GOLPH2 expression in the lymph nodes was higher than in primary OSCC tissue. Moreover, patients with higher GOLPH2 expression had shorter survival times compared with patients with lower GOLPH2 expression. This finding may reveal novel ways that GOLPH2 regulates tumor cell migration and invasion, important functions for the metastatic process in OSCC. The immune system plays a crucial role in the carcinogenesis of OSCC.24 The immune system can recognize and eliminate cancer, but it is held in check by inhibitory receptors and ligands. These immune checkpoint pathways, which normally maintain self-tolerance and limit collateral tissue damage during antimicrobial immune responses, can be coopted by cancer to evade immune destruction, particularly targeting T cells that are specific for tumor antigens.52,53 GOLPH2 overexpression attenuated antitumor Th1 lymphocyte response in a study on gastric cancer.54 B7-H3, B7-H4, TIM3, and VISTA, which can act as inhibitory immune checkpoints, play an important role in carcinogenesis.55 As shown in previous and present studies, these molecules have a high expression in OSCC.26-29 In the present study, we also showed that GOLPH2 expression was correlated with the expression of B7-H3, B7-H4, TIM3, and VISTA. Furthermore, we can detect close correlations between GOLPH2 and B7H4 from the cluster graph. In our previous study, higher expression of B7-H4 confers poor prognosis of patients with OSCC.28 In our data, a similar result was found, with high expression of GOLPH2 signifying a poor prognosis. To our knowledge, this is the first report of GOLPH2 correlations with immune checkpoints. Conclusion Our study suggests that GOLPH2 is a novel predictive marker for clinical outcome in OSCC. Meanwhile, the positive correlation among GOLPH2 and CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA preliminarily identifies the GOLPH2-associated molecules that may play a potential role in tumor invasion and migration. Because this is a small OSCC cohort, further study is still needed. Our data also provide evidence that GOLPH2 is a potential marker for estimating the patient’s prognosis and may be a novel target for molecular-targeted therapy against OSCC. Functional studies are needed to evaluate the mechanisms involved. This work was supported by the National Natural Science Foundation of China (81472528, 81472529). Z.-J.S. was supported by the Fundamental Research Funds for the Central Universities of China (2042017kf0171) (Outstanding Young Scholars). References 1. Ali J , Sabiha B , Jan HU , et al. Genetic etiology of oral cancer . Oral Oncol . 2017 ; 70 : 23 - 28 . Google Scholar CrossRef Search ADS PubMed 2. Torre LA , Bray F , Siegel RL , et al. Global cancer statistics, 2012 . CA Cancer J Clin . 2015 ; 65 : 87 - 108 . Google Scholar CrossRef Search ADS PubMed 3. Gupta S , Kong W , Peng Y , et al. Temporal trends in the incidence and survival of cancers of the upper aerodigestive tract in Ontario and the United States . Int J Cancer . 2009 ; 125 : 2159 - 2165 . Google Scholar CrossRef Search ADS PubMed 4. Inagi K , Takahashi H , Okamoto M , et al. Treatment effects in patients with squamous cell carcinoma of the oral cavity . Acta Otolaryngol Suppl . 2002 ;547: 25 - 29 . 5. Shingaki S , Takada M , Sasai K , et al. Impact of lymph node metastasis on the pattern of failure and survival in oral carcinomas . Am J Surg . 2003 ; 185 : 278 - 284 . Google Scholar CrossRef Search ADS PubMed 6. Kladney RD , Bulla GA , Guo L , et al. GP73, a novel Golgi-localized protein upregulated by viral infection . Gene . 2000 ; 249 : 53 - 65 . Google Scholar CrossRef Search ADS PubMed 7. Maccioni HJ , Quiroga R , Spessott W . Organization of the synthesis of glycolipid oligosaccharides in the Golgi complex . FEBS Lett . 2011 ; 585 : 1691 - 1698 . Google Scholar CrossRef Search ADS PubMed 8. Isaji T , Sato Y , Fukuda T , et al. N-glycosylation of the I-like domain of beta1 integrin is essential for beta1 integrin expression and biological function: identification of the minimal N-glycosylation requirement for alpha5beta1 . J Biol Chem . 2009 ; 284 : 12207 - 12216 . Google Scholar CrossRef Search ADS PubMed 9. Riener MO , Stenner F , Liewen H , et al. Golgi phosphoprotein 2 (GOLPH2) expression in liver tumors and its value as a serum marker in hepatocellular carcinomas . Hepatology . 2009 ; 49 : 1602 - 1609 . Google Scholar CrossRef Search ADS PubMed 10. Kristiansen G , Fritzsche FR , Wassermann K , et al. GOLPH2 protein expression as a novel tissue biomarker for prostate cancer: implications for tissue-based diagnostics . Br J Cancer . 2008 ; 99 : 939 - 948 . Google Scholar CrossRef Search ADS PubMed 11. Liu G , Zhang Y , He F , et al. Expression of GOLPH2 is associated with the progression of and poor prognosis in gastric cancer . Oncol Rep . 2014 ; 32 : 2077 - 2085 . Google Scholar CrossRef Search ADS PubMed 12. Donizy P , Kaczorowski M , Biecek P , et al. Golgi-related proteins GOLPH2 (GP73/GOLM1) and GOLPH3 (GOPP1/MIDAS) in cutaneous melanoma: patterns of expression and prognostic significance . Int J Mol Sci . 2016 ; 17 : 1619 . Google Scholar CrossRef Search ADS 13. Fritzsche FR , Riener MO , Dietel M , et al. GOLPH2 expression in renal cell cancer . BMC Urol . 2008 ; 8 : 15 . Google Scholar CrossRef Search ADS PubMed 14. Byrne AM , Bekiaris S , Duggan G , et al. Golgi phosphoprotein 2 (GOLPH2) is a novel bile acid-responsive modulator of oesophageal cell migration and invasion . Br J Cancer . 2015 ; 113 : 1332 - 1342 . Google Scholar CrossRef Search ADS PubMed 15. Zhang Y , Hu W , Wang L , et al. Association of GOLPH2 expression with survival in non-small-cell lung cancer: clinical implications and biological validation . Biomark Med . 2017 ; 11 : 967 - 977 . Google Scholar CrossRef Search ADS PubMed 16. Zhang W , Kim H , Lv J , et al. Golgi phosphoprotein 2 is a novel regulator of IL-12 production and macrophage polarization . J Immunol . 2018 ; 200 : 1480 - 1488 . Google Scholar CrossRef Search ADS PubMed 17. Mani SA , Guo W , Liao MJ , et al. The epithelial-mesenchymal transition generates cells with properties of stem cells . Cell . 2008 ; 133 : 704 - 715 . Google Scholar CrossRef Search ADS PubMed 18. Shackleton M , Quintana E , Fearon ER , et al. Heterogeneity in cancer: cancer stem cells versus clonal evolution . Cell . 2009 ; 138 : 822 - 829 . Google Scholar CrossRef Search ADS PubMed 19. Visvader JE , Lindeman GJ . Cancer stem cells in solid tumours: accumulating evidence and unresolved questions . Nat Rev Cancer . 2008 ; 8 : 755 - 768 . Google Scholar CrossRef Search ADS PubMed 20. Singh A , Settleman J . EMT, cancer stem cells and drug resistance: an emerging axis of evil in the war on cancer . Oncogene . 2010 ; 29 : 4741 - 4751 . Google Scholar CrossRef Search ADS PubMed 21. Faber A , Barth C , Hörmann K , et al. CD44 as a stem cell marker in head and neck squamous cell carcinoma . Oncol Rep . 2011 ; 26 : 321 - 326 . Google Scholar PubMed 22. Huang CF , Xu XR , Wu TF , et al. Correlation of ALDH1, CD44, OCT4 and SOX2 in tongue squamous cell carcinoma and their association with disease progression and prognosis . J Oral Pathol Med . 2014 ; 43 : 492 - 498 . Google Scholar CrossRef Search ADS PubMed 23. Zhang J , Cheng Q , Zhou Y , et al. Slug is a key mediator of hypoxia induced cadherin switch in HNSCC: correlations with poor prognosis . Oral Oncol . 2013 ; 49 : 1043 - 1050 . Google Scholar CrossRef Search ADS PubMed 24. Ferris RL . Immunology and immunotherapy of head and neck cancer . J Clin Oncol . 2015 ; 33 : 3293 - 3304 . Google Scholar CrossRef Search ADS PubMed 25. Wherry EJ , Kurachi M . Molecular and cellular insights into T cell exhaustion . Nat Rev Immunol . 2015 ; 15 : 486 - 499 . Google Scholar CrossRef Search ADS PubMed 26. Mao L , Fan TF , Wu L , et al. Selective blockade of B7-H3 enhances antitumour immune activity by reducing immature myeloid cells in head and neck squamous cell carcinoma . J Cell Mol Med . 2017 ; 21 : 2199 - 2210 . Google Scholar CrossRef Search ADS PubMed 27. Wu L , Deng WW , Huang CF , et al. Expression of VISTA correlated with immunosuppression and synergized with CD8 to predict survival in human oral squamous cell carcinoma . Cancer Immunol Immunother . 2017 ; 66 : 627 - 636 . Google Scholar CrossRef Search ADS PubMed 28. Wu L , Deng WW , Yu GT , et al. B7-H4 expression indicates poor prognosis of oral squamous cell carcinoma . Cancer Immunol Immunother . 2016 ; 65 : 1035 - 1045 . Google Scholar CrossRef Search ADS PubMed 29. Liu JF , Ma SR , Mao L , et al. T-cell immunoglobulin mucin 3 blockade drives an antitumor immune response in head and neck cancer . Mol Oncol . 2017 ; 11 : 235 - 247 . Google Scholar CrossRef Search ADS PubMed 30. Sobin L, Wittekind C, Greene F, et al. TNM Classification of Malignant Tumours. 5th ed. New York, NY: John Wiley & Sons; 2002. 31. Zhong LP , Zhang CP , Ren GX , et al. Randomized phase III trial of induction chemotherapy with docetaxel, cisplatin, and fluorouracil followed by surgery versus up-front surgery in locally advanced resectable oral squamous cell carcinoma . J Clin Oncol . 2013 ; 31 : 744 - 751 . Google Scholar CrossRef Search ADS PubMed 32. Budczies J , Klauschen F , Sinn BV , et al. Cutoff Finder: a comprehensive and straightforward web application enabling rapid biomarker cutoff optimization . PLoS One . 2012 ; 7 : e51862 . Google Scholar CrossRef Search ADS PubMed 33. Toruner GA , Ulger C , Alkan M , et al. Association between gene expression profile and tumor invasion in oral squamous cell carcinoma . Cancer Genet Cytogenet . 2004 ; 154 : 27 - 35 . Google Scholar CrossRef Search ADS PubMed 34. Ginos MA , Page GP , Michalowicz BS , et al. Identification of a gene expression signature associated with recurrent disease in squamous cell carcinoma of the head and neck . Cancer Res . 2004 ; 64 : 55 - 63 . Google Scholar CrossRef Search ADS PubMed 35. Wu TF , Chen L , Bu LL , et al. CD44+ cancer cell-induced metastasis: a feasible neck metastasis model . Eur J Pharm Sci . 2017 ; 101 : 243 - 250 . Google Scholar CrossRef Search ADS PubMed 36. Qiao B , He B , Cai J , et al. The expression profile of Oct4 and Sox2 in the carcinogenesis of oral mucosa . Int J Clin Exp Pathol . 2014 ; 7 : 28 - 37 . Google Scholar PubMed 37. Ma SR , Wang WM , Huang CF , et al. Anterior gradient protein 2 expression in high grade head and neck squamous cell carcinoma correlated with cancer stem cell and epithelial mesenchymal transition . Oncotarget . 2015 ; 6 : 8807 - 8821 . Google Scholar PubMed 38. Millarte V , Farhan H . The Golgi in cell migration: regulation by signal transduction and its implications for cancer cell metastasis . Sci World J . 2012 ; 2012 : 498278 . Google Scholar CrossRef Search ADS 39. Kojima S , Enokida H , Yoshino H , et al. The tumor-suppressive microrna-143/145 cluster inhibits cell migration and invasion by targeting GOLM1 in prostate cancer . J Hum Genet . 2014 ; 59 : 78 - 87 . Google Scholar CrossRef Search ADS PubMed 40. Liu Y , Zhang X , Sun T , et al. Knockdown of Golgi phosphoprotein 2 inhibits hepatocellular carcinoma cell proliferation and motility . Oncotarget . 2016 ; 7 : 21404 - 21415 . Google Scholar PubMed 41. Jin D , Tao J , Li D , et al. Golgi protein 73 activation of MMP-13 promotes hepatocellular carcinoma cell invasion . Oncotarget . 2015 ; 6 : 33523 - 33533 . Google Scholar PubMed 42. Ha NH , Park DG , Woo BH , et al. Porphyromonas gingivalis increases the invasiveness of oral cancer cells by upregulating IL-8 and MMPs . Cytokine . 2016 ; 86 : 64 - 72 . Google Scholar CrossRef Search ADS PubMed 43. Ahmed Haji Omar A , Haglund C , Virolainen S , et al. MMP-7, MMP-8, and MMP-9 in oral and cutaneous squamous cell carcinomas . Oral Surg Oral Med Oral Pathol Oral Radiol . 2015 ; 119 : 459 - 467 . Google Scholar CrossRef Search ADS PubMed 44. Pu Y , Wang L , Wu H , et al. High MMP-21 expression in metastatic lymph nodes predicts unfavorable overall survival for oral squamous cell carcinoma patients with lymphatic metastasis . Oncol Rep . 2014 ; 31 : 2644 - 2650 . Google Scholar CrossRef Search ADS PubMed 45. Yu Q , Stamenkovic I . Localization of matrix metalloproteinase 9 to the cell surface provides a mechanism for CD44-mediated tumor invasion . Genes Dev . 1999 ; 13 : 35 - 48 . Google Scholar CrossRef Search ADS PubMed 46. Mladinich M , Ruan D , Chan CH . Tackling cancer stem cells via inhibition of EMT transcription factors . Stem Cells Int . 2016 ; 2016 : 5285892 . Google Scholar CrossRef Search ADS PubMed 47. Chen YW , Chen KH , Huang PI , et al. Cucurbitacin I suppressed stem-like property and enhanced radiation-induced apoptosis in head and neck squamous carcinoma–derived CD44(+)ALDH1(+) cells . Mol Cancer Ther . 2010 ; 9 : 2879 - 2892 . Google Scholar CrossRef Search ADS PubMed 48. Rodrigo JP , Domínguez F , Alvarez C , et al. Clinicopathologic significance of expression of CD44S and CD44V6 isoforms in squamous cell carcinoma of the supraglottic larynx . Am J Clin Pathol . 2002 ; 118 : 67 - 72 . Google Scholar CrossRef Search ADS PubMed 49. Li X , Wang J , Xu Z , et al. Expression of Sox2 and Oct4 and their clinical significance in human non-small-cell lung cancer . Int J Mol Sci . 2012 ; 13 : 7663 - 7675 . Google Scholar CrossRef Search ADS PubMed 50. Brcic L , Sherer CK , Shuai Y , et al. Morphologic and clinicopathologic features of lung squamous cell carcinomas expressing Sox2 . Am J Clin Pathol . 2012 ; 138 : 712 - 718 . Google Scholar CrossRef Search ADS PubMed 51. Camp ER , Findlay VJ , Vaena SG , et al. Slug expression enhances tumor formation in a noninvasive rectal cancer model . J Surg Res . 2011 ; 170 : 56 - 63 . Google Scholar CrossRef Search ADS PubMed 52. Topalian SL , Drake CG , Pardoll DM . Immune checkpoint blockade: a common denominator approach to cancer therapy . Cancer Cell . 2015 ; 27 : 450 - 461 . Google Scholar CrossRef Search ADS PubMed 53. Pardoll DM . The blockade of immune checkpoints in cancer immunotherapy . Nat Rev Cancer . 2012 ; 12 : 252 - 264 . Google Scholar CrossRef Search ADS PubMed 54. Tang QF , Ji Q , Tang Y , et al. Golgi phosphoprotein 2 down-regulates the th1 response in human gastric cancer cells by suppressing IL-12A . Asian Pac J Cancer Prev . 2013 ; 14 : 5747 - 5751 . Google Scholar CrossRef Search ADS PubMed 55. Schildberg FA , Klein SR , Freeman GJ , et al. Coinhibitory pathways in the B7-CD28 ligand-receptor family . Immunity . 2016 ; 44 : 955 - 972 . Google Scholar CrossRef Search ADS PubMed © American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Clinical Pathology Oxford University Press

Overexpression of Golgi Phosphoprotein 2 Is Associated With Poor Prognosis in Oral Squamous Cell Carcinoma

Loading next page...
 
/lp/ou_press/overexpression-of-golgi-phosphoprotein-2-is-associated-with-poor-RBq0xO4BY2
Publisher
Oxford University Press
Copyright
© American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0002-9173
eISSN
1943-7722
D.O.I.
10.1093/ajcp/aqy029
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives The aims of this study were to investigate the relationship between Golgi phosphoprotein 2 (GOLPH2) and oral squamous cell carcinoma (OSCC) and explore the clinical significance of GOLPH2 in OSCC. Methods Tissue microarrays from human OSCC samples were stained for GOLPH2 expression and clinicopathologic features. Kaplan-Meier analysis was used to compare the survival of patients with high GOLPH2 expression and patients with low GOLPH2 expression. Results We found GOLPH2 is highly expressed in OSCC tissue, and the GOLPH2 expression in metastatic lymph nodes is higher than in tumor tissue. Our data indicate that patients with higher GOLPH2 expression have poor overall survival compared with those with lower GOLPH2 expression. This study demonstrated that GOLPH2 was associated with CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Conclusions These findings suggest GOLPH2 is a potential marker for estimating the patient’s prognosis and may be a target for molecular-targeted therapy against OSCC. Cancer stem cell, Epithelial-mesenchymal transition, GOLPH2, Immune checkpoint, Oral squamous cell carcinoma Oral squamous cell carcinoma (OSCC) accounts for more than 90% of oral cancers (occurring in the mouth, lip, and tongue) and is the eighth most common cancer in humans. OSCC accounts for approximately 2.5% of all new cancer cases and 1.9% of all cancer deaths annually.1,2 Many important risk factors have been identified so far for OSCC, including immune status, smoking, alcohol use, and the synergistic effects of tobacco and alcohol.1,3 Despite recent advances in cancer treatment methods, the 5-year survival rate of OSCC is approximately 50% in the advanced stage of the disease.4,5 Although there are conventional tumor parameters, such as tumor size, nodal status, and the existence of systemic metastasis, new molecular markers are warranted to provide more information on the tumor biology, to allow a better prognosis and possibly predict the stratification of patients. Golgi phosphoprotein 2 (GOLPH2), also known as Golgi membrane protein 1 (GOLM1) or GP73, is a Golgi phosphoprotein of yet unknown function. The 73-kDa Golgi apparatus–associated protein is encoded by the GOLM1 gene and was first described in the liver tissue of a patient with giant cell hepatitis.6 The structure includes several areas of possible glycosylation.7 Due to its localization at the Golgi apparatus, the proposed functions include protein modification, cell signaling, intracellular transport function, or merely local structural tasks.8 Initially, GOLPH2 upregulation was reported in neoplastic and nonneoplastic liver pathologies, and many studies postulated that its serum level is a novel marker of hepatocellular carcinoma.9 Subsequently, enhanced GOLPH2 expression was reported in other cancer settings such as gastric cancer, esophageal cancer, cutaneous melanoma, renal cell cancer, and prostate cancer.10-14 GOLPH2 has been implicated in several important pathologic processes, including cancer cell proliferation and metastasis.15,16 The epithelial-mesenchymal transition (EMT) is a process by which epithelial cells lose their cell polarity and cell-cell adhesion and is a key developmental program that is often activated during cancer invasion and metastasis.17 Cancer stem cells (CSCs) represent a minor subpopulation of tumor cells that share some features with the normal stem cells of the tissue from which tumor derives and have the properties of self-renewal, multiple differentiations, and tumor initiation.18 CSCs cause relapse and metastasis by giving rise to new tumors and are therefore tumorigenic in contrast to other nontumorigenic cancer cells.19 This property conferred by the EMT is thus a doubly dangerous duo for the patient, as it not only enables the carcinoma cells to enter the bloodstream but also increases the tumorigenic and proliferative potential of these cells.20 CD44, SOX2 and Slug, the marker of CSCs and EMT, were correlated with OSCC in previous studies.21-23 The immune system plays a key role in the development, establishment, and progression of cancer.24 In a recent study, GOLPH2 was proved to have a correlation with the immune system and be a novel regulator of interleukin 12 production and macrophage polarization.16 Immune checkpoint pathways can be coopted by cancer to evade immune destruction, particularly targeting T cells.25 In our previous studies, we demonstrated the expression status of B7-H3, B7-H4, TIM3, and VISTA, four of the inhibitory immune checkpoints in OSCC,26-29 but the correlation between GOLPH2 and immune checkpoint was unclear. Although the expression of GOLPH2 has been studied in several types of cancer, the association between GOLPH2 and OSCC remains unknown. The central aim of this study was to evaluate the potential diagnostic and prognostic value of GOLPH2. Materials and Methods Detailed procedures are provided in the Supplementary Materials and Methods (all supplemental materials can be found at American Journal of Clinical Pathology online). Tissue Microarray Construction Full ethical approval was granted by the School and Stomatology of Wuhan University Medical Ethics Committee (principal investigator: Zhi-Jun Sun) to use patient samples in this study. The specimens of the tissue microarrays were obtained from January 2008 to August 2015 at the Department of Oral and Maxillofacial Surgery, School and Hospital of Stomatology, Wuhan University. In the present study, all patients were diagnosed with OSCC by two independent pathologists. Clinical follow-up data, as annually assessed for survival time, were available for all patients. All information, including hospital records, histology slides, and pathology reports, was retrieved and reviewed. The clinical stage of the OSCC was classified according to the guidelines of the Union for International Cancer Control.30 The grading scheme of the World Health Organization was used to determine the histologic grading. The OSCC cancer cohort consisted of 43 normal oral mucosae, 48 oral epithelial dysplasias, 165 patients with primary OSCC (excluding recurrent, presurgical inductive chemotherapy and presurgical radiotherapy), 12 patients with recurrent OSCC, 10 patients with OSCC who received presurgical radiotherapy (without presurgical chemotherapy), and 17 patients with OSCC who received presurgical TPF (docetaxel, cisplatin, and fluorouracil) inductive chemotherapy without presurgical radiotherapy by previous protocol.31 Forty-one pathologically confirmed metastatic lymph nodes from 41 neck dissection specimens were constructed in the tissue microarray (TMA). Clinical features, including TNM stage, histologic grade, and overall survival, were available for all cases. Human papillomavirus (HPV) infection status was diagnosed by p16 immunostaining and the HPV DNA in situ hybridization technique. Immunohistochemistry and Evaluation The TMAs were performed on 4-μm-thick paraffin sections, which were then subjected to deparaffinization and rehydration. For epitope unmasking, the sections were boiled in 0.01 mol/L citric acid buffer solution (pH 6.0) for 1.5 minutes at high pressure. Then, the sections were blocked for endogenous peroxidase activity by incubating sections in 3% hydrogen peroxide solution for 20 minutes at room temperature, and 10% goat serum was added to the sections to block nonspecific binding. The TMAs were incubated overnight at 4°C in a humidified chamber with primary antibody for GOLPH2 (1:500; GeneTex, San Antonio, TX), CD44 (1:100; Cell Signaling Technology, Danvers, MA), SOX2 (1:300; Cell Signaling Technology), Slug (1:200; Cell Signaling Technology), B7-H3 (1:800; Cell Signaling Technology), B7-H4 (1:800; Cell Signaling Technology), TIM3 (1:400; Cell Signaling Technology), VISTA (1:400; Cell Signaling Technology), or isotype-matched immunoglobulin G controls. A positive slide was created for each experiment. Subsequently, a secondary biotinylated IgG antibody solution and an avidin-biotin-peroxidase reagent were added to the slides. Peroxidase activity was detected with a 3,3′-diaminobenzidine solution. Then, the slides were counterstained by immersion in hematoxylin for 1 to 2 minutes. Slides were scanned by the ScanScope CS scanner (Aperio, Vista, CA) and then processed with background subtraction and white balance. The membrane, nuclear, or pixel immunohistochemical staining was quantified using Aperio Quantification software. For each sample, fields of high expression were scanned, and the membrane and nuclear immunostaining were calculated by the followed formula: (1 × percentage of weakly positive staining) + (2 × percentage of moderately positive staining) + (3 × percentage of strongly positive staining). Then, the scores were normalized between 0 and 300. The histoscores were converted to the range of –3 to 3 using Microsoft Excel (Microsoft, Redmond WA). Cluster 3.0 (Stanford University, Stanford, CA) with average linkage was used for hierarchical analysis, and the results were visualized using Java TreeView 1.1.3 (Stanford University). Statistical Analysis The data analyses were performed by GraphPad Prism version 5.0 for Windows (GraphPad Software, La Jolla, CA). One-way analysis of variance (ANOVA) followed by post-ANOVA Tukey multiple-comparison tests and unpaired t tests were used to analyze the differences among immunohistochemical staining within each group. The best cutoff of survival curves was calculated from the website Cutoff Finder (http://molpath.charite.de/cutoff/index.jsp) by a previously reported protocol.32 The survival of the patients in the two groups was compared using a Kaplan-Meier analysis and log tests. Data were graphically represented as the mean values ± SEM. The Cox proportional hazards regression model was used for multivariate analysis to assess the significance of overall differences. The statistical significance level for all comparisons was set at P < .05. Results Expression of GOLPH2 in OSCC Is More Activated Than in Normal Mucosa and Dysplasia and Is Associated With Patient Survival To determine whether overexpression of GOLPH2 was correlated with OSCC, the Oncomine database, a publicly available cancer data set, was used. In studies by Toruner et al33 and Ginos et al,34 messenger RNA (mRNA) expression indicated that the mRNA expression of GOLPH2 was significantly increased in OSCC compared with normal squamous cells (P < .005, Supplementary Figure 1). Then, the tissue microarrays were used to evaluate the expression in OSCC tissues. Immunostaining revealed that cytoplasmic patterns of GOLPH2 staining were mainly found in the tumor cells, and partial expression was found in the stromal components Figure 1A, which is consistent with a previous study about GOLPH2 expression in other types of cancer.9-13 In addition, we detected that GOLPH2 had a stronger staining intensity in OSCC than in dysplasia (n = 43, P < .01) and in normal mucosae (n = 48, P < .005; Figure 1B). Subsequently, we used Cutoff Finder to find the optimized cutoff by a previously reported protocol,32 and a histoscore of 120.6 was calculated for the analysis as the best cutoff for survival. The Kaplan-Meier survival analysis in Figure 1C indicated that GOLPH2 expression was significantly correlated with poor overall survival in patients with OSCC at the best cutoff (P = .0005). When the median expression was used as the cutoff, a significant prognostic value for patient survival was not detectable, but a trend for longer survival times of patients with lower GOLPH2 expression was apparent in the Kaplan-Meier curve (Supplementary Figure 2E, P > .05). A multivariate Cox proportional hazard model using variables that included sex, age, pathologic grade, tumor size, lymph node stage, and GOLPH2 expression identified variables as prognostically significant by univariate analysis and demonstrated that high expression of GOLPH2 was a predictor of reduced survival (P = .034; Table 1). Table 1 Multivariate Analysis for Overall Survival in Patients With Primary Oral Squamous Cell Carcinomaa Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b CI, confidence interval; GOLPH2, Golgi phosphoprotein 2; HR, hazard ratio. aCox proportional hazards regression model. bP < .05. View Large Table 1 Multivariate Analysis for Overall Survival in Patients With Primary Oral Squamous Cell Carcinomaa Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b Parameter HR (95% CI) P Value Sex 1.266 (0.594-2.698) .542 Age 1.194 (0.664-2.147) .554 Pathologic grade  II vs I 7.424 (1.768-31.181) .006b  III vs I 8.292 (1.867-36.830) .005b Tumor size  T2 vs T1 1.267 (0.478-3.361) .635  T3 vs T1 2.248 (0.803-6.292) .123  T4 vs T1 1.624 (0.441-5.976) .466 Node stage  N1 + N2 vs N0 1.711 (0.97-3.019) .064 GOLPH2 expression 1.954 (1.054-3.622) .034b CI, confidence interval; GOLPH2, Golgi phosphoprotein 2; HR, hazard ratio. aCox proportional hazards regression model. bP < .05. View Large Figure 1 View largeDownload slide Increased expression of Golgi phosphoprotein 2 (GOLPH2) in human oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of GOLPH2 in OSCC tissue compared with normal mucosa (×4 and ×20). B, Quantification of the immunohistochemical histoscore of GOLPH2 in oral squamous cell carcinoma (OSCC, n = 165) compared with dysplasia (DYS, n = 48) and oral mucosa (MUC, n = 43). C, Kaplan-Meier curve according to low (n = 136) and high (n = 29) expression of GOLPH2. The best cutoff (120.6) was used. *P < .01. **P < .001. Figure 1 View largeDownload slide Increased expression of Golgi phosphoprotein 2 (GOLPH2) in human oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of GOLPH2 in OSCC tissue compared with normal mucosa (×4 and ×20). B, Quantification of the immunohistochemical histoscore of GOLPH2 in oral squamous cell carcinoma (OSCC, n = 165) compared with dysplasia (DYS, n = 48) and oral mucosa (MUC, n = 43). C, Kaplan-Meier curve according to low (n = 136) and high (n = 29) expression of GOLPH2. The best cutoff (120.6) was used. *P < .01. **P < .001. Expression of GOLPH2 in Metastatic Lymph Nodes Is Higher Than in Primary Tumor Tissue We observed that GOLPH2 had a stronger immunoreactivity in metastatic lymph nodes than in primary tumor tissue Figure 2A. The quantification also confirmed this result (P < .01; Figure 2B). On further assessment, GOLPH2 was not significantly associated with pathologic grade (I, II, and III; P > .05; Figure 2C). There was no significant difference in tumor size (T1 + T2 vs T3 + T4; P > .05; Figure 2D) and lymph node metastasis status (N1 vs N2 + N3; P > .05; Figure 2E), but we found a slight association from both parameters in the graph. Subsequently, no significant differences were observed for GOLPH2 expression from samples with different HPV infection status (P > .05; Figure 2F). There was no significant correlation between GOLPH2 expression in primary OSCC and preoperation inductive TPF (P > .05; Figure 2G), preoperation radiotherapy (P > .05; Supplementary Figure 2A), and recurrent OSCC (P > .05; Supplementary Figure 2B). Furthermore, no significant differences in GOLPH2 expression were observed for tobacco smoking (P > .05; Supplementary Figure 2C) and alcohol consumption (P > .05; Supplementary Figure 2D) in primary OSCC. Figure 2 View largeDownload slide Higher expression of Golgi phosphoprotein 2 (GOLPH2) in oral squamous cell carcinoma (OSCC) tissue than metastatic lymph nodes. A, A representative patient (patient 29) for whom the immunochemical histoscore was greater in OSCC tissue than in the metastatic lymph node (LN) (×4 and ×20). B, Histoscore of GOLPH2 in OSCC tissue (n = 165) compared with tissue of metastatic LNs (n = 41). *P < .01. C, Quantification of the immunohistochemical histoscore of GOLPH2 by pathologic grade (I, n = 43; II, n = 84; and III, n = 38). D, Quantification of immunochemical histoscore of GOLPH2 between T1 + T2 (n = 115) and T3 + T4 (n = 50). E, Quantification of the immunohistochemical histoscore of GOLPH2 between N1 (n = 105) and N2 + N3 (n = 60). F, Quantification of immunohistochemical histoscore of GOLPH2 between human papillomavirus (HPV)–positive (n = 16) and HPV-negative (n = 149) samples. G, Quantification of immunohistochemical histoscore of GOLPH2 in primary OSCC (n = 165) compared with preoperative inductive TPF (docetaxel, cisplatin, and fluorouracil) (n = 17). All data are presented as the means ± SEM. NS, not significant. Figure 2 View largeDownload slide Higher expression of Golgi phosphoprotein 2 (GOLPH2) in oral squamous cell carcinoma (OSCC) tissue than metastatic lymph nodes. A, A representative patient (patient 29) for whom the immunochemical histoscore was greater in OSCC tissue than in the metastatic lymph node (LN) (×4 and ×20). B, Histoscore of GOLPH2 in OSCC tissue (n = 165) compared with tissue of metastatic LNs (n = 41). *P < .01. C, Quantification of the immunohistochemical histoscore of GOLPH2 by pathologic grade (I, n = 43; II, n = 84; and III, n = 38). D, Quantification of immunochemical histoscore of GOLPH2 between T1 + T2 (n = 115) and T3 + T4 (n = 50). E, Quantification of the immunohistochemical histoscore of GOLPH2 between N1 (n = 105) and N2 + N3 (n = 60). F, Quantification of immunohistochemical histoscore of GOLPH2 between human papillomavirus (HPV)–positive (n = 16) and HPV-negative (n = 149) samples. G, Quantification of immunohistochemical histoscore of GOLPH2 in primary OSCC (n = 165) compared with preoperative inductive TPF (docetaxel, cisplatin, and fluorouracil) (n = 17). All data are presented as the means ± SEM. NS, not significant. Protein Expression of GOLPH2 Was Conspicuously Associated With CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA in OSCC Tissue We focused on specific antibody expression in human OSCC tissue microarrays. In previous studies, CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA had high expressions in OSCC compared with normal tissue.26-29,35-37 As shown in Figure 3A, higher expressions of these molecules were confirmed in our TMAs. Immunohistochemical staining showed that CD44 and SOX2, markers of CSCs, were overexpressed in tumor tissue (Figure 3A). Slug, an EMT marker, was overexpressed in the TMAs (Figure 3A). TIM3 and VISTA were highly expressed in the tumor-infiltrating immune cells in the tumor microenvironment (Figure 3A). B7-H3 and B7-H4 were highly expressed in the immune cells and tumor cells (Figure 3A). Hierarchical clustering further confirmed that GOLPH2 expression was closely associated with B7-H4 expression Figure 3B. Interestingly, the Pearson correlation coefficient test revealed that GOLPH2 expression was statistically associated with CD44 (P < .0001, r = 0.3855), SOX2 (P = .0005, r = 0.2692), Slug (P = .0005, r = 0.2662), B7-H3 (P = .0006, r = 0.2635), B7-H4 (P < .0001, r = 0.5939), TIM3 (P < .0001, r = 0.4520), and VISTA (P = .0002, r = 0.2909) Figure 4. Figure 3 View largeDownload slide High expression of CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA in oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of Golgi phosphoprotein 2 (GOLPH2), CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA (×4 and ×20). B, Hierarchical clustering of GOLPH2, CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA immunohistochemical results in human OSCC with statistics (n = 196). Figure 3 View largeDownload slide High expression of CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA in oral squamous cell carcinoma (OSCC). A, Representative H&E and immunohistochemical staining of Golgi phosphoprotein 2 (GOLPH2), CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA (×4 and ×20). B, Hierarchical clustering of GOLPH2, CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA immunohistochemical results in human OSCC with statistics (n = 196). Figure 4 View largeDownload slide Golgi phosphoprotein 2 (GOLPH2) has a positive correlation with CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Correlations among GOLPH2 expression and CD44 (A), SOX2 (B), Slug (C), B7-H3 (D), B7-H4 (E), TIM3 (F), and VISTA (G) expression in human oral squamous cell carcinoma tissue microarrays. In dot plots, each dot represents a specimen according to the data (n = 165) from the issue microarrays. A, P < .0001, r = 0.3855. B, P = .0005, r = 0.2692. C, P = .0005, r = 0.2662. D, P = .0006, r = 0.2635. E, P < .0001, r = 0.5939. F, P < .0001, r = 0.4520. G, P = .0002, r = 0.2909. Figure 4 View largeDownload slide Golgi phosphoprotein 2 (GOLPH2) has a positive correlation with CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Correlations among GOLPH2 expression and CD44 (A), SOX2 (B), Slug (C), B7-H3 (D), B7-H4 (E), TIM3 (F), and VISTA (G) expression in human oral squamous cell carcinoma tissue microarrays. In dot plots, each dot represents a specimen according to the data (n = 165) from the issue microarrays. A, P < .0001, r = 0.3855. B, P = .0005, r = 0.2692. C, P = .0005, r = 0.2662. D, P = .0006, r = 0.2635. E, P < .0001, r = 0.5939. F, P < .0001, r = 0.4520. G, P = .0002, r = 0.2909. Discussion In the present study, our results demonstrated that GOLPH2 was highly and aberrantly expressed in human primary OSCC tissue compared with dysplasia and normal mucosa. Furthermore, we observed GOLPH2 was more highly expressed in metastatic lymph nodes than in primary tumor tissue. Survival analysis indicated that patients with high GOLPH2 expression had poor overall survival compared with those with low GOLPH2 expression. In addition, the data in the present study demonstrated that the expression of GOLPH2 correlated with that of CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA. Our findings support that GOLPH2 may be a prognostic indicator for OSCC. To our knowledge, this is the first systematic study of GOLPH2 expression in OSCC in the clinical setting. Cell invasion is a major feature of malignancy, and the Golgi apparatus plays an important role in cell migration and acts as a hub for different signaling pathways.38 Therefore, elucidation of the regulatory networks of the Golgi apparatus will provide important information for the development of therapeutic strategies against cancer migration, invasion, and metastasis. GOLPH2, a cis-Golgi-localized protein with unknown cell function, is overexpressed and usually an indicator for poor prognosis in several types of cancer.9-14 Previous studies have indicated that GOLPH2 plays a role in regulating tumor cell migration and invasion, important functions for the metastatic process in many kinds of cancer.14,39,40 A recent study41 has shown that GOLPH2 promotes hepatocellular carcinoma invasion by the activation of matrix metalloproteinases (MMPs), which play a crucial role in degradation of basement membranes and the extracellular matrix. Furthermore, studies have proved that MMP molecules are correlated with OSCC metastasis.42-44 Previous results have indicated that adhesion molecules such as CD44 are involved in positioning activated MMPs on the cell surface of invasive tumor cells.45 A recent study confirmed that GOLPH2 expression leads to increased expression of EMT-related proteins and that GOLPH2 silencing reduces hepatocellular carcinoma cell migration in vitro.40 Molecular profiling has shown that cancer cells and tumors that exhibit the CSC phenotype also express genes associated with the EMT.46 In the present study, GOLPH2 expression was associated with CSC markers (CD44 and SOX2) and an EMT marker (SLUG). CSCs play an important role in tumor metastasis. CD44, a transmembrane glycoprotein and a key cell surface marker of CSCs in head and neck carcinoma, has possessed unique properties in functional assays for CSC self-renewal and differentiation.47,48 SOX2, which also has been recognized for causing “stemness” characteristics in cancer cells, can lead to cell immortality and account for the self-renewal and invasive properties of cancer cells.49,50 Slug, a C2H2-type zinc finger transcription factor that is also an EMT marker, has been identified in aggressive cancers.51 Our data support this point, as GOLPH2 expression in the lymph nodes was higher than in primary OSCC tissue. Moreover, patients with higher GOLPH2 expression had shorter survival times compared with patients with lower GOLPH2 expression. This finding may reveal novel ways that GOLPH2 regulates tumor cell migration and invasion, important functions for the metastatic process in OSCC. The immune system plays a crucial role in the carcinogenesis of OSCC.24 The immune system can recognize and eliminate cancer, but it is held in check by inhibitory receptors and ligands. These immune checkpoint pathways, which normally maintain self-tolerance and limit collateral tissue damage during antimicrobial immune responses, can be coopted by cancer to evade immune destruction, particularly targeting T cells that are specific for tumor antigens.52,53 GOLPH2 overexpression attenuated antitumor Th1 lymphocyte response in a study on gastric cancer.54 B7-H3, B7-H4, TIM3, and VISTA, which can act as inhibitory immune checkpoints, play an important role in carcinogenesis.55 As shown in previous and present studies, these molecules have a high expression in OSCC.26-29 In the present study, we also showed that GOLPH2 expression was correlated with the expression of B7-H3, B7-H4, TIM3, and VISTA. Furthermore, we can detect close correlations between GOLPH2 and B7H4 from the cluster graph. In our previous study, higher expression of B7-H4 confers poor prognosis of patients with OSCC.28 In our data, a similar result was found, with high expression of GOLPH2 signifying a poor prognosis. To our knowledge, this is the first report of GOLPH2 correlations with immune checkpoints. Conclusion Our study suggests that GOLPH2 is a novel predictive marker for clinical outcome in OSCC. Meanwhile, the positive correlation among GOLPH2 and CD44, SOX2, Slug, B7-H3, B7-H4, TIM3, and VISTA preliminarily identifies the GOLPH2-associated molecules that may play a potential role in tumor invasion and migration. Because this is a small OSCC cohort, further study is still needed. Our data also provide evidence that GOLPH2 is a potential marker for estimating the patient’s prognosis and may be a novel target for molecular-targeted therapy against OSCC. Functional studies are needed to evaluate the mechanisms involved. This work was supported by the National Natural Science Foundation of China (81472528, 81472529). Z.-J.S. was supported by the Fundamental Research Funds for the Central Universities of China (2042017kf0171) (Outstanding Young Scholars). References 1. Ali J , Sabiha B , Jan HU , et al. Genetic etiology of oral cancer . Oral Oncol . 2017 ; 70 : 23 - 28 . Google Scholar CrossRef Search ADS PubMed 2. Torre LA , Bray F , Siegel RL , et al. Global cancer statistics, 2012 . CA Cancer J Clin . 2015 ; 65 : 87 - 108 . Google Scholar CrossRef Search ADS PubMed 3. Gupta S , Kong W , Peng Y , et al. Temporal trends in the incidence and survival of cancers of the upper aerodigestive tract in Ontario and the United States . Int J Cancer . 2009 ; 125 : 2159 - 2165 . Google Scholar CrossRef Search ADS PubMed 4. Inagi K , Takahashi H , Okamoto M , et al. Treatment effects in patients with squamous cell carcinoma of the oral cavity . Acta Otolaryngol Suppl . 2002 ;547: 25 - 29 . 5. Shingaki S , Takada M , Sasai K , et al. Impact of lymph node metastasis on the pattern of failure and survival in oral carcinomas . Am J Surg . 2003 ; 185 : 278 - 284 . Google Scholar CrossRef Search ADS PubMed 6. Kladney RD , Bulla GA , Guo L , et al. GP73, a novel Golgi-localized protein upregulated by viral infection . Gene . 2000 ; 249 : 53 - 65 . Google Scholar CrossRef Search ADS PubMed 7. Maccioni HJ , Quiroga R , Spessott W . Organization of the synthesis of glycolipid oligosaccharides in the Golgi complex . FEBS Lett . 2011 ; 585 : 1691 - 1698 . Google Scholar CrossRef Search ADS PubMed 8. Isaji T , Sato Y , Fukuda T , et al. N-glycosylation of the I-like domain of beta1 integrin is essential for beta1 integrin expression and biological function: identification of the minimal N-glycosylation requirement for alpha5beta1 . J Biol Chem . 2009 ; 284 : 12207 - 12216 . Google Scholar CrossRef Search ADS PubMed 9. Riener MO , Stenner F , Liewen H , et al. Golgi phosphoprotein 2 (GOLPH2) expression in liver tumors and its value as a serum marker in hepatocellular carcinomas . Hepatology . 2009 ; 49 : 1602 - 1609 . Google Scholar CrossRef Search ADS PubMed 10. Kristiansen G , Fritzsche FR , Wassermann K , et al. GOLPH2 protein expression as a novel tissue biomarker for prostate cancer: implications for tissue-based diagnostics . Br J Cancer . 2008 ; 99 : 939 - 948 . Google Scholar CrossRef Search ADS PubMed 11. Liu G , Zhang Y , He F , et al. Expression of GOLPH2 is associated with the progression of and poor prognosis in gastric cancer . Oncol Rep . 2014 ; 32 : 2077 - 2085 . Google Scholar CrossRef Search ADS PubMed 12. Donizy P , Kaczorowski M , Biecek P , et al. Golgi-related proteins GOLPH2 (GP73/GOLM1) and GOLPH3 (GOPP1/MIDAS) in cutaneous melanoma: patterns of expression and prognostic significance . Int J Mol Sci . 2016 ; 17 : 1619 . Google Scholar CrossRef Search ADS 13. Fritzsche FR , Riener MO , Dietel M , et al. GOLPH2 expression in renal cell cancer . BMC Urol . 2008 ; 8 : 15 . Google Scholar CrossRef Search ADS PubMed 14. Byrne AM , Bekiaris S , Duggan G , et al. Golgi phosphoprotein 2 (GOLPH2) is a novel bile acid-responsive modulator of oesophageal cell migration and invasion . Br J Cancer . 2015 ; 113 : 1332 - 1342 . Google Scholar CrossRef Search ADS PubMed 15. Zhang Y , Hu W , Wang L , et al. Association of GOLPH2 expression with survival in non-small-cell lung cancer: clinical implications and biological validation . Biomark Med . 2017 ; 11 : 967 - 977 . Google Scholar CrossRef Search ADS PubMed 16. Zhang W , Kim H , Lv J , et al. Golgi phosphoprotein 2 is a novel regulator of IL-12 production and macrophage polarization . J Immunol . 2018 ; 200 : 1480 - 1488 . Google Scholar CrossRef Search ADS PubMed 17. Mani SA , Guo W , Liao MJ , et al. The epithelial-mesenchymal transition generates cells with properties of stem cells . Cell . 2008 ; 133 : 704 - 715 . Google Scholar CrossRef Search ADS PubMed 18. Shackleton M , Quintana E , Fearon ER , et al. Heterogeneity in cancer: cancer stem cells versus clonal evolution . Cell . 2009 ; 138 : 822 - 829 . Google Scholar CrossRef Search ADS PubMed 19. Visvader JE , Lindeman GJ . Cancer stem cells in solid tumours: accumulating evidence and unresolved questions . Nat Rev Cancer . 2008 ; 8 : 755 - 768 . Google Scholar CrossRef Search ADS PubMed 20. Singh A , Settleman J . EMT, cancer stem cells and drug resistance: an emerging axis of evil in the war on cancer . Oncogene . 2010 ; 29 : 4741 - 4751 . Google Scholar CrossRef Search ADS PubMed 21. Faber A , Barth C , Hörmann K , et al. CD44 as a stem cell marker in head and neck squamous cell carcinoma . Oncol Rep . 2011 ; 26 : 321 - 326 . Google Scholar PubMed 22. Huang CF , Xu XR , Wu TF , et al. Correlation of ALDH1, CD44, OCT4 and SOX2 in tongue squamous cell carcinoma and their association with disease progression and prognosis . J Oral Pathol Med . 2014 ; 43 : 492 - 498 . Google Scholar CrossRef Search ADS PubMed 23. Zhang J , Cheng Q , Zhou Y , et al. Slug is a key mediator of hypoxia induced cadherin switch in HNSCC: correlations with poor prognosis . Oral Oncol . 2013 ; 49 : 1043 - 1050 . Google Scholar CrossRef Search ADS PubMed 24. Ferris RL . Immunology and immunotherapy of head and neck cancer . J Clin Oncol . 2015 ; 33 : 3293 - 3304 . Google Scholar CrossRef Search ADS PubMed 25. Wherry EJ , Kurachi M . Molecular and cellular insights into T cell exhaustion . Nat Rev Immunol . 2015 ; 15 : 486 - 499 . Google Scholar CrossRef Search ADS PubMed 26. Mao L , Fan TF , Wu L , et al. Selective blockade of B7-H3 enhances antitumour immune activity by reducing immature myeloid cells in head and neck squamous cell carcinoma . J Cell Mol Med . 2017 ; 21 : 2199 - 2210 . Google Scholar CrossRef Search ADS PubMed 27. Wu L , Deng WW , Huang CF , et al. Expression of VISTA correlated with immunosuppression and synergized with CD8 to predict survival in human oral squamous cell carcinoma . Cancer Immunol Immunother . 2017 ; 66 : 627 - 636 . Google Scholar CrossRef Search ADS PubMed 28. Wu L , Deng WW , Yu GT , et al. B7-H4 expression indicates poor prognosis of oral squamous cell carcinoma . Cancer Immunol Immunother . 2016 ; 65 : 1035 - 1045 . Google Scholar CrossRef Search ADS PubMed 29. Liu JF , Ma SR , Mao L , et al. T-cell immunoglobulin mucin 3 blockade drives an antitumor immune response in head and neck cancer . Mol Oncol . 2017 ; 11 : 235 - 247 . Google Scholar CrossRef Search ADS PubMed 30. Sobin L, Wittekind C, Greene F, et al. TNM Classification of Malignant Tumours. 5th ed. New York, NY: John Wiley & Sons; 2002. 31. Zhong LP , Zhang CP , Ren GX , et al. Randomized phase III trial of induction chemotherapy with docetaxel, cisplatin, and fluorouracil followed by surgery versus up-front surgery in locally advanced resectable oral squamous cell carcinoma . J Clin Oncol . 2013 ; 31 : 744 - 751 . Google Scholar CrossRef Search ADS PubMed 32. Budczies J , Klauschen F , Sinn BV , et al. Cutoff Finder: a comprehensive and straightforward web application enabling rapid biomarker cutoff optimization . PLoS One . 2012 ; 7 : e51862 . Google Scholar CrossRef Search ADS PubMed 33. Toruner GA , Ulger C , Alkan M , et al. Association between gene expression profile and tumor invasion in oral squamous cell carcinoma . Cancer Genet Cytogenet . 2004 ; 154 : 27 - 35 . Google Scholar CrossRef Search ADS PubMed 34. Ginos MA , Page GP , Michalowicz BS , et al. Identification of a gene expression signature associated with recurrent disease in squamous cell carcinoma of the head and neck . Cancer Res . 2004 ; 64 : 55 - 63 . Google Scholar CrossRef Search ADS PubMed 35. Wu TF , Chen L , Bu LL , et al. CD44+ cancer cell-induced metastasis: a feasible neck metastasis model . Eur J Pharm Sci . 2017 ; 101 : 243 - 250 . Google Scholar CrossRef Search ADS PubMed 36. Qiao B , He B , Cai J , et al. The expression profile of Oct4 and Sox2 in the carcinogenesis of oral mucosa . Int J Clin Exp Pathol . 2014 ; 7 : 28 - 37 . Google Scholar PubMed 37. Ma SR , Wang WM , Huang CF , et al. Anterior gradient protein 2 expression in high grade head and neck squamous cell carcinoma correlated with cancer stem cell and epithelial mesenchymal transition . Oncotarget . 2015 ; 6 : 8807 - 8821 . Google Scholar PubMed 38. Millarte V , Farhan H . The Golgi in cell migration: regulation by signal transduction and its implications for cancer cell metastasis . Sci World J . 2012 ; 2012 : 498278 . Google Scholar CrossRef Search ADS 39. Kojima S , Enokida H , Yoshino H , et al. The tumor-suppressive microrna-143/145 cluster inhibits cell migration and invasion by targeting GOLM1 in prostate cancer . J Hum Genet . 2014 ; 59 : 78 - 87 . Google Scholar CrossRef Search ADS PubMed 40. Liu Y , Zhang X , Sun T , et al. Knockdown of Golgi phosphoprotein 2 inhibits hepatocellular carcinoma cell proliferation and motility . Oncotarget . 2016 ; 7 : 21404 - 21415 . Google Scholar PubMed 41. Jin D , Tao J , Li D , et al. Golgi protein 73 activation of MMP-13 promotes hepatocellular carcinoma cell invasion . Oncotarget . 2015 ; 6 : 33523 - 33533 . Google Scholar PubMed 42. Ha NH , Park DG , Woo BH , et al. Porphyromonas gingivalis increases the invasiveness of oral cancer cells by upregulating IL-8 and MMPs . Cytokine . 2016 ; 86 : 64 - 72 . Google Scholar CrossRef Search ADS PubMed 43. Ahmed Haji Omar A , Haglund C , Virolainen S , et al. MMP-7, MMP-8, and MMP-9 in oral and cutaneous squamous cell carcinomas . Oral Surg Oral Med Oral Pathol Oral Radiol . 2015 ; 119 : 459 - 467 . Google Scholar CrossRef Search ADS PubMed 44. Pu Y , Wang L , Wu H , et al. High MMP-21 expression in metastatic lymph nodes predicts unfavorable overall survival for oral squamous cell carcinoma patients with lymphatic metastasis . Oncol Rep . 2014 ; 31 : 2644 - 2650 . Google Scholar CrossRef Search ADS PubMed 45. Yu Q , Stamenkovic I . Localization of matrix metalloproteinase 9 to the cell surface provides a mechanism for CD44-mediated tumor invasion . Genes Dev . 1999 ; 13 : 35 - 48 . Google Scholar CrossRef Search ADS PubMed 46. Mladinich M , Ruan D , Chan CH . Tackling cancer stem cells via inhibition of EMT transcription factors . Stem Cells Int . 2016 ; 2016 : 5285892 . Google Scholar CrossRef Search ADS PubMed 47. Chen YW , Chen KH , Huang PI , et al. Cucurbitacin I suppressed stem-like property and enhanced radiation-induced apoptosis in head and neck squamous carcinoma–derived CD44(+)ALDH1(+) cells . Mol Cancer Ther . 2010 ; 9 : 2879 - 2892 . Google Scholar CrossRef Search ADS PubMed 48. Rodrigo JP , Domínguez F , Alvarez C , et al. Clinicopathologic significance of expression of CD44S and CD44V6 isoforms in squamous cell carcinoma of the supraglottic larynx . Am J Clin Pathol . 2002 ; 118 : 67 - 72 . Google Scholar CrossRef Search ADS PubMed 49. Li X , Wang J , Xu Z , et al. Expression of Sox2 and Oct4 and their clinical significance in human non-small-cell lung cancer . Int J Mol Sci . 2012 ; 13 : 7663 - 7675 . Google Scholar CrossRef Search ADS PubMed 50. Brcic L , Sherer CK , Shuai Y , et al. Morphologic and clinicopathologic features of lung squamous cell carcinomas expressing Sox2 . Am J Clin Pathol . 2012 ; 138 : 712 - 718 . Google Scholar CrossRef Search ADS PubMed 51. Camp ER , Findlay VJ , Vaena SG , et al. Slug expression enhances tumor formation in a noninvasive rectal cancer model . J Surg Res . 2011 ; 170 : 56 - 63 . Google Scholar CrossRef Search ADS PubMed 52. Topalian SL , Drake CG , Pardoll DM . Immune checkpoint blockade: a common denominator approach to cancer therapy . Cancer Cell . 2015 ; 27 : 450 - 461 . Google Scholar CrossRef Search ADS PubMed 53. Pardoll DM . The blockade of immune checkpoints in cancer immunotherapy . Nat Rev Cancer . 2012 ; 12 : 252 - 264 . Google Scholar CrossRef Search ADS PubMed 54. Tang QF , Ji Q , Tang Y , et al. Golgi phosphoprotein 2 down-regulates the th1 response in human gastric cancer cells by suppressing IL-12A . Asian Pac J Cancer Prev . 2013 ; 14 : 5747 - 5751 . Google Scholar CrossRef Search ADS PubMed 55. Schildberg FA , Klein SR , Freeman GJ , et al. Coinhibitory pathways in the B7-CD28 ligand-receptor family . Immunity . 2016 ; 44 : 955 - 972 . Google Scholar CrossRef Search ADS PubMed © American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

American Journal of Clinical PathologyOxford University Press

Published: May 16, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off