Moesin expression by tumor cells is an unfavorable prognostic biomarker for oral cancer

Moesin is a member of the ERM (ezrin, radixin and moesin) family of proteins that plays a role in cellular morphology, cell adhesion, controlling adherent junctions and cell motility, the key events of the carcinogenesis processes [1‐3]. Specifically, the metastatic process involves moesin and other ERM proteins, leading to changes in cell morphology, cell to cell adhesion and in actin filament reorganization [4].

Recently, the overexpression of moesin in tumors has been correlated with metastasis and poor prognosis for the patient [5‐8], including those with oral squamous cell carcinomas [9, 10]. Moesin probably participates in necessary conformational changes for an appropriate cell configuration, flexible enough to allow extravasation [11]. These processes occur through the phosphorylation of the C-terminal domain of moesin, which binds to actin filaments [12]. In the other domain, N-terminal, transmembrane molecules, such as podoplanin, are able to activate it, consequently inducing Rho phosphorylation. These downstream signals result in loss of adhesion, motility and higher rates of cell proliferation [13].

Podoplanin is a transmembrane glycoprotein and its overexpression has been associated with enhanced cancer cell motility, tumor invasion and poor patient prognosis in head and neck tumors [14‐21]. The role of podoplanin in the tumor invasion process was first hypothesized by Martin-Villar et al. (2005), who reported the activation of ERM proteins by podoplanin through Rho-A phosphorylation. This connection is responsible for the maintanance of the ERM proteins in an open and active state conformation, strengthening the anchorage of podoplanin to cytoskeleton filaments.

To gain further understanding of the participation of moesin in tumor invasion process and its association with podoplanin in this pathway, for the first time, we analyzed the immunohistochemical association of moesin and podoplanin in oral squamous cell carcinomas with clinicopathological features and patients’ prognosis.

This study was based on the analysis of eighty-four surgical specimens of patients who underwent surgical treatment for primary oral squamous cell carcinoma (OSCC) at the Head and Neck Surgery and Otorhinolaryngology Department of the A.C. Camargo Cancer Hospital, São Paulo, Brazil, from 1963 to 2012. Tumors were selected according to the following inclusion criteria: (1) oral squamous cell carcinoma located in the tongue, floor of the mouth, inferior gingiva and retromolar area, confirmed by biopsy; (2) patients not submitted to other previous treatment; (3) complete clinical data and follow up; (4) tumor tissue available for microscopic analysis. Clinical data were obtained from the medical records of the A.C. Camargo Cancer Hospital and included age, ethnic group, gender, tobacco and alcohol consumption, TNM stage (UICC: union for international cancer control, 2004), treatment (surgery, post-operative adjuvant radiotherapy), localization of the tumor and clinical follow-up (local recurrence, regional recurrence and death). Histopathological analysis included the following variables: vascular embolization, perineural infiltration, muscular infiltration, bone infiltration and lymph node involvement (pN+). Additionally, the histopathological grade of malignancy of oral squamous cell carcinomas was determined in hematoxylin & eosin stained sections [22].

For the present study, the sample consisted of 84 patients with oral squamous cell carcinomas, predominantly male (85.7%), white (95.2%), with ages varying from 33 to 95 years (mean of 58 years). According to risk factors for OSCC, alcohol (67.9%) and tobacco (84.5%) consumption were reported by the majority of the patients. The most common location for OSCC was the tongue (54.8%), followed by floor of the mouth (28.6%). Moreover, most patients were clinically classified as T2 (59.5%) or T3 (40.5%), and N+ (48.8%), at the time of physical examination.

Only three patients (3.6%) were not submitted to elective neck dissection, 78.6% were submitted to ipsilateral neck dissection and 17.8% were dissected bilaterally. Thirty-eight patients (45.2%) who were submitted to neck dissection were positive for lymph node metastasis (pN+), shown by histopathological analysis.

Local recurrence occurred in 25% of the patients, while regional recurrence was present in 16.7% and local and regional recurrences, simultaneously, occurred in 6% of the patients.

Most tumors, 86.9% were classified as well/moderately differentiated; and 13.1% were classified as poorly differentiated, according to the histopathological grade of tumor malignancy [22].

To date, the intracellular location of moesin during oral carcinogenesis has been poorly understood. Moesin translocation from plasma membrane to the cytoplasm of neoplastic cells has been demonstrated in a previous study [24] and it may reduce the ability to form cell-cell contacts, as well as, influence the cytoskeleton remodeling and tumor invasion process, when overexpressed in the cytoplasm [24]. Furthermore, podoplanin linkage with the N-terminal moesin domain activates the ERM proteins, consequently inducing Rho-A phosphorylation and in the maintenance of the ERM proteins in an active form (open conformation), in the cytoplasm [25].

The moesin immunoprofile in the invasive front of oral squamous cell carcinomas, clinical stages II and III, was predominantly cytoplasmic and strong. Keratin pearls and some areas with more differentiated neoplastic cells showed negative moesin expression. Confirming these results, other authors described the membranous and cytoplasmic moesin expression in oral squamous cell carcinomas, [5, 9, 10]. Belbin et al. (2005) showed that membranous and cytoplasmic moesin expression increased when normal epithelium was compared with dysplastic epithelium and/or with tumor samples [10]. The authors affirmed that moesin expression was significantly associated with head and neck squamous cell carcinomas progression. In addition, as found in the present study, the loss of moesin expression in more differentiated cells was found in keratin pearls [5, 26].

The expression of moesin was not associated with the clinical, demographic and microscopic features analyzed (p > 0.05), reinforcing previous findings [5, 7, 9, 26]. Kobayashi et al. (2004) found a significant association of moesin with the size of the tumor, but the sample analyzed varied from T1 to T4 tumors. Differently, the present study was composed of T2 and T3 tumors and no association could be found with moesin expression, as it was restricted to a limited size of tumors. According to our previous experiences with analyses of protein expression by oral neoplastic cells, the OSCCs clinically classified as T2 and T3 are better for verifying their influence on patients’ prognosis, than T1 (initial tumors) or T4 (advanced tumors).

Regarding podoplanin expression by tumor cells, as previously observed by de Vicente et al. (2015) and Tsuneki et al. (2013), the present results showed higher podoplanin expression in the earlier stages of tumors and in highly differentiated malignant cells. These studies [27‐29] verified that podoplanin expression was inversely correlated with the degree of neoplastic epithelial cell differentiation. In agreement with the cited studies, the results found in the present study confirmed a weak/absent podoplanin expression in the majority of tumors with muscular infiltration and lymph node involvement (pN+).

Initial studies about podoplanin related strong podoplanin expression to the worst prognosis and more advanced stages of oral squamous cell carcinomas [14‐16, 30, 31]. Instead, the present results showed the opposite side of those findings, and the authors suggest further studies should be conducted about podoplanin expression in less differentiated tumors to validate these findings.

Considering demographic and clinical features, the strong expression of podoplanin was observed mainly in those patients submitted to post-operative radiotherapy (p = 0.004) [14, 18, 19, 28, 30]; however, further studies are necessary for better evaluation of radiotherapy and podoplanin expression.

Concerning clinical stages T and N, our results failed to find an association between podoplanin expression and the clinical stages of the tumor, probably because we included only T2 and T3 tumors, while in other studies, the tumor stages varied from I to IV [14, 16, 19, 20, 30].

The overall survival and the prognostic value of moesin and podoplanin expression in oral squamous cell carcinomas were analyzed in this study. Moesin expression was considered a significant prognostic factor for oral squamous cell carcinomas. The overall survival rate in 5 years was 22.7% for those patients with the strong moesin expression, while for patients with weaker moesin expression, the overall survival rates in 5 years was 38.5%. In ten years, overall survival rates for strong moesin expression were 6.8% and 23.8% for weak expression. Moreover, the patient with strong moesin expression presented a 1737 times higher risk of dying, compared with those with the weak expression. These results are consistent with those found by Kobayashi et al. (2004) and Schlecht et al. (2012) in oral squamous cell carcinomas and in breast cancer [6, 32].

Interestingly and corroborating the above results, a recent study conducted by Li et al. (2015) observed the knockdown of moesin in oral squamous cell carcinomas cell lines and a significantly reduction in migration and invasion [33]. Furthermore, moesin silencing showed an increase in cell-cell adhesion. By cell spreading assay, moesin inhibition reduced filopodia formation, indicating the role of moesin in cytoskeletal modifications. This study was in agreement with the present results, which indicated that the weak expression of moesin could be related to higher survival rates in oral squamous cell carcinoma patients.

Podoplanin expression, in turn, was not considered as a significant prognostic factor for oral squamous cell carcinomas, as related by de Vicente et al. (2015) and Dos Santos Almeida et al. (2013). Although previous results considered [14‐16, 30, 31] that strong podoplanin expression was related to poor prognosis, we could not predict the prognostic significance of podoplanin in those tumors, probably because strong podoplanin expression was related to initial tumors in earlier stages of the neoplastic cell differentiation process [27‐29]. Therefore, considering the latest controversial studies concerning podoplanin expression, we suggest that further studies with a larger sample should be conducted to confirm the potential prognostic value of this protein in oral squamous cell carcinomas.

The weak immunohistochemical expression of moesin by neoplastic epithelial cells could be used as a favorable prognostic marker for oral squamous cell carcinomas. Additionally, the weaker expression of podoplanin was associated with lymph node involvement and muscular infiltration in oral squamous cell carcinomas, indicating that podoplanin expression occurs in more differentiated cells and at earlier stages of the tumors. Moreover, malignant cells of oral squamous cell carcinomas expressed both podoplanin and moesin, but these proteins probably act individually in tumor invasion process.