Overexpression of Mcl-1L Splice Variant Is Associated with Poor Prognosis and Chemoresistance in Oral Cancers

Background Altered expression of Mcl-1, an anti-apoptotic member of the Bcl-2 family, has been linked to the progression and outcome of a variety of malignancies. We have previously reported the overexpression of Mcl-1 protein in human oral cancers. The present study aimed to evaluate the clinicopathological significance of the expression of three known Mcl-1 isoforms in oral tumors and the effect of targeting Mcl-1L isoform on chemosensitivity of oral cancer cells. Methods The expression of Mcl-1 isoforms- Mcl-1L, Mcl-1S & Mcl-1ES was analyzed in 130 paired oral tumors and 9 oral cell lines using quantitative real-time PCR & protein by western blotting. The Mcl-1 mRNA levels were correlated with clinicopathological parameters and outcome of oral cancer patients. The effect of Mcl-1L shRNA or Obatoclax (a small molecule Mcl-1 inhibitor), in combination with Cisplatin on chemosensitivity of oral cancer cells was also assessed. Results Anti-apoptotic Mcl-1L was predominantly expressed, over low or undetectable pro-apoptotic Mcl-1S and Mcl-1ES isoforms. The Mcl-1L transcripts were significantly overexpressed in all cancer cell lines and in 64% oral tumors versus adjacent normals (P<0.02). In oral cancer patients, high Mcl-1L expression was significantly associated with node positivity (P = 0.021), advanced tumor size (P = 0.013) and poor overall survival (P = 0.002). Multivariate analysis indicated Mcl-1L to be an independent prognostic factor for oral cancers (P = 0.037). Mcl-1L shRNA knockdown or its inhibition by Obatoclax in combination with Cisplatin synergistically reduced viability and growth of oral cancer cells than either treatment alone. Conclusion Our studies suggest that overexpression of Mcl-1L is associated with poor prognosis and chemoresistance in oral cancers. Mcl-1L is an independent prognostic factor and a potential therapeutic target in oral cancers.


Introduction
Oral cancer is the most prevalent cancer among Indian males and is predominantly associated with tobacco-chewing habit prevalent in the country [1]. Despite recent advances in treatment modalities like surgery, radiotherapy/chemotherapy, the long term survival of oral cancer patients has not changed significantly. The factors associated with poor prognosis of oral cancer include, presentation at an advanced clinical stage & uncontrolled locoregional recurrence [2]. Hence it is important to elucidate the mechanisms involved in the development and progression of oral cancer and identify molecular targets for better disease management.
Oral cancers have repeatedly been associated with apoptotic dysregulation [3]. The pro and anti-apoptotic members of the Bcl-2 family are the key regulators of cellular apoptosis and play a critical role in regulating cell survival [4]. Mcl-1 (Myeloid cell leukemia-1) is an important anti-apoptotic member of the Bcl-2 gene family, essential for development, differentiation, and proliferation [5]. Cellular expression of Mcl-1 is tightly regulated through multiple transcriptional and post-transcriptional mechanisms [6]. Increased Mcl-1 expression can produce moderate short-term viability enhancement in a broad range of cell types. Mcl-1 may promote cell survival by suppressing the release of cytochrome-c from mitochondria via heterodimerisation and the neutralization of effector pro-apoptotic BH3-only proteins such as Bak and Noxa [7]. The overexpression of Mcl-1 has been reported in a variety of malignancies including hematopoietic, lymphoid and solid tumors [8,9]. Overexpression of Mcl-1 has been associated with aggressive tumor features, resistance to treatment and poor prognosis in breast, gastric, ovarian & cervical cancers [10][11][12][13].
Although the Mcl-1 gene has been studied extensively in multiple myeloma and leukemia, there are rare reports on Mcl-1 analysis in head and neck cancer. Recent studies from our laboratory have demonstrated significant overexpression of Mcl-1 protein in oral cancer cell lines, premalignant lesions (OSF) and oral tumors by immunohistochemistry [14]. We have also demonstrated high PCNA and Mcl-1 protein expression to be associated with poor prognosis in oral cancer patients treated with definitive radiotherapy [15]. However, the situation is complex due to the existence of three distinct Mcl-1 isoforms having contrasting functions namely anti-apoptotic Mcl-1L and proapoptotic Mcl-1S & Mcl-1ES [16]. Interestingly, our lab has recently reported the association of anti-apoptotic Mcl-1L isoform with survival and radioresistance of oral squamous carcinoma cells [17].
Mcl-1 has also been shown to play a role in chemoresistance of a variety of cancers, but the role of its isoforms in chemoresistance has not been studied in cancers including oral cancers. Radiation followed by chemotherapy is the common treatment modality for oral cancer and Mcl-1 overexpression has been shown to provide resistance to conventional chemotherapeutic drugs like Cisplatin [18]. Several reports have shown that Mcl-1 promotes cell survival and targeting Mcl-1 via BH3-mimetic molecules can induce cell death in Cisplatin resistant cancer cells [19,20]. Recently, Obatoclax, a BH3 mimetic small molecule inhibitor has been shown to antagonize Mcl-1 protein and overcome Mcl-1 mediated resistance to apoptosis [21,22] [28] cell lines were also used in study. The cells were maintained in MEM or IMDM supplemented with 10% FBS & 1% standard antibiotic mixture in 5% CO 2 incubator at 37uC as described earlier [17,23,29].

Oral tissues
This study was approved by the Institutional Review Board of Tata Memorial Centre and Sharad Pawar Dental College, India. Written informed consents were obtained from all the study participants. Treatment naïve primary oral tumor samples and their adjacent normal tissues were obtained from 130 patients undergoing surgery at the Head and Neck Unit and from the ICMR National Tumor Tissue Repository, Tata Memorial Centre, Mumbai, India. Inflamed oral tissues (n = 20) were collected as healthy normal tissues from patients undergoing minor dental surgical procedure at the Department of Oral Pathology, Sharad Pawar Dental College, Wardha, India. All the tissues were frozen immediately in liquid nitrogen and stored at 280uC until analysis. The clinicopathological characteristics of the cohort are illustrated in Table 1.

RNA isolation and Real-time PCR
RNA from cell pellets and tissues were extracted using TRI reagent (Sigma, USA) according to the manufacturer's protocol. The RNA was dissolved in DEPC-treated water and contaminating DNA was removed by DNaseI treatment (Sigma, USA). RNA integrity was analyzed by electrophoresis and samples were preserved at 280uC until analysis, as described earlier [17]. cDNA was synthesized with 500 ng total RNA, using a First Strand cDNA synthesis kit (MBI Fermentas, Canada) according to the manufacturer's instructions. The cDNA was then subjected for quantitative real-time PCR using Taqman universal PCR master mix (ABI, 4304437) and Mcl-1 isoform specific gene expression probes (ABI, Mcl-1L:Hs00172036_m1; Mcl-1S:Hs00766187_m1 & Mcl-1ES:4331348). The amplification was done on ABI-7900HT real time PCR system (Applied Biosystems, USA). GAPDH (ABI, Hs99999905_m1) was used as internal control to normalize inter sample variation in RNA input & amplification efficiency. All amplification reactions were done in triplicates, using DEPC treated water as negative controls. The analysis of gene expression data was done by using the Comparative C T method of relative quantification [30].

Western blotting
Proteins were extracted from cell pellets and tumor tissues as described earlier [14]. Briefly, 30 mg of frozen tissue was pulverized in mortal & pestle using liquid Nitrogen, dissolved in chilled ProteoJET Mammalian Cell Lysis Reagent containing ProteoBlock Protease Inhibitor Cocktail (Fermentas, USA) and sonicated on ice. The tissue or cell lysates were cleared by centrifugation at 14000 rpm at 4uC. The protein estimation was performed by the Bradford method. The tissue/cell lysates (30-50 mg) were resolved on 12% SDS-PAGE gels and transferred onto PVDF membranes (Millipore, USA). Membranes were blocked with 5% skimmed milk in TBS for 2 hrs. and incubated overnight at 4uC with mouse monoclonal antibody against Mcl-1L (1:1000) (Santa Cruz Biotechnology, USA). The membranes were stripped and reprobed with rabbit polyclonal antibody against ßactin (1:2000) (Santa Cruz Biotechnology, USA) as loading control. Secondary antibodies used were Horseradish peroxidase conjugated IgG (1:5000) (Santa Cruz Biotechnology, USA). Proteins were visualized with enhanced chemiluminescence kit (GE Healthcare, US). Densitometry analysis of developed X-ray film was performed using ImageJ software (NIH, Bethesda, MD).

MTT assay
Cells were seeded at 2000 cells per well in 96-well plates containing DMEM with 10% FBS. After 24 h, cells were either left untreated or treated with different concentrations of Cisplatin (0.025-10.0 mg/ml) (MP Biomedicals, India) or Obatoclax (0, 1.0-10.0 nM) (Selleck chem USA). The 50% inhibitory concentrations (IC50) for Cisplatin and Obatoclax were calculated from the survival curves for all the cell lines. DMSO containing medium served as the respective control. Following incubation of cells with Obatoclax and/or Cisplatin for 72 h, the cell growth was assessed by the MTT assay, as per manufacturer's instructions (Sigma, USA). Briefly, the spent medium was removed and 50 ml of MTT solution (1 mg/ml) was added to each well. The plates was incubated for 4 hrs at 37uC in CO 2 incubator and the formazan crystals in viable cells were dissolved by using dimethyl sulfoxide (DMSO, 100ul per well). The plate was agitated on an orbital shaker for 10 min and the absorbance was measured at 540 nm using reference wavelength of 690 nm on micro plate reader (Spectramax, USA). Five wells were used for each drug concentration, and the experiment was repeated three times. In another experiment, to evaluate whether combination of treatments would achieve higher growth inhibition than individual treatment, oral cancer cells were treated with doxycycline to knockdown Mcl-1 expression or Obatoclax (5 nM) to inhibit Mcl-1 activity, followed by Cisplatin at a concentration lower than the IC50 dose. All the experiments were performed in triplicates.

Trypan blue assay & Confocal Microscopy
Cells were seeded into 24-well plates at a density of 5610 4 per well. The Mcl-1 expression was targeted either by Doxycycline or Obatoclax treatments or in combination with Cisplatin as described above. Cells were trypsinzed after 48 hrs. of treatment

Statistical analysis
Statistical analysis was done using a licensed version of SPSS software package 15.0. To statistically correlate the expression of Mcl-1 isoforms with clinicopathological parameters, the data was dichotomized into two groups namely: the Mcl-1 isoform high expressers and low expressers. For comparison the mean expression of Mcl-1 isoforms in healthy normals and histologically normal adjacent tissues were used. The correlation between clinicopathological parameters (gender, age, site, habits, size, node, metastasis, recurrence) was done using the Chi Square test (x 2 ). Kaplan Meier curves were used for evaluating the overall survival and the difference within the groups was calculated with Log Rank test of significance. The predictive parameters in the univariate analysis were incorporated into multivariate analysis using Cox's proportional hazard test to identify the factors that were independent predictors of survival. The Graphpad Prism5 software was used to plot graphs and the data of two similar & dissimilar groups were statistically analyzed by Student-t test & Mann Whitney test. The variation among group means was analyzed by one-way analysis of variance (ANOVA) test. The data are reported as mean 6SD. The p value ,0.05 was considered statistically significant.

Correlation of Mcl-1 expression with clinicopathological parameters & outcome
Univariate analysis revealed significant correlation of high Mcl-1L expression with node positivity (p = 0.021) and advanced tumors (p = 0.013). However, no significant correlation was observed between expression of Mcl-1 isoforms and gender, age, tobacco/alcohol habits, primary site & differentiation of oral cancer patients (Table 2). Notably, low expression of pro-apoptotic Mcl-1S isoform showed a trend towards positive significance with poorly differentiated tumors (p = 0.053).

Correlation of Mcl-1 expression with overall survival
The Kaplan-Meier survival curves of low and high expressers of Mcl-1L showed a statistically significant difference (p = 0.002). Patients expressing high anti-apoptotic Mcl-1L exhibited poor overall survival versus those expressing low Mcl-1L (Figure 3a). Inversely, patients expressing high pro-apoptotic Mcl-1S exhibited significantly better overall survival (p = 0.051) as compared to those having low Mcl-1S (Figure 3b). Notably, the relative ratio of Mcl-1L/Mcl-1S also showed a positive correlation (p = 0.006) with the poor overall survival of oral cancer patients (Figure 3c). Additionally, the univariate analysis also revealed poor overall survival in node positive versus node negative oral cancer patients (p = 0.003). The other parameters like age, tobacco/alcohol habits and differentiation did not significantly influence overall survival of these patients.   (Figure 6b). Moreover, the MTT proliferation assay also revealed a significant (p,0.05) reduction in cell growth (2-4 fold) after combined treatment of Cisplatin and Obatoclax as compared to either treatments alone (Figure 6c) thereby, suggesting a synergistic effect of the combined treatment on cell viability and growth of oral cancer cells.

Discussion
In the present study, we assessed the expression of The dysregulation of apoptosis regulating genes has been reported to play a key role in the development and progression of several human malignancies. The anti-apoptotic Mcl-1 is an important member of the apoptosis regulating Bcl-2 family and has been shown to be overexpressed in variety of cancers including, cervical, ovarian, pancreatic, hepatocellular, non-small cell lung, testicular germ cell cancers and melanomas [8].  As observed in the present study for Mcl-1 isoforms, high Mcl-1 protein expression and its association with poor prognosis has been demonstrated in cervical, gastric, lung and ovarian cancers [11][12][13]32]. However, a previous study from our lab could not find the above association at the Mcl-1 protein level [14]. The present study was undertaken to elucidate the expression profile of Mcl-1 isoforms in oral cell lines and a large cohort of tumors, which demonstrated the association of anti-apoptotic Mcl-1L expression with prognosis in oral cancer. This is the first study, demonstrating the correlation of Mcl-1L splice variant with outcome of oral cancer patients and Mcl-1L as an independent prognostic marker for oral cancers.
Mcl-1 is tightly regulated at the transcriptional, post transcriptional and translational levels [6] and the exact mechanism of Mcl-1 overexpression in oral cancers is not known. The pro-apoptotic Mcl-1S & Mcl-1ES isoforms were expressed at low or undetectable levels as compared to the predominantly expressed Mcl-1L and did not correlate with clinicopathological parameters of oral cancer patients. Notably, the short pro-apoptotic Mcl-1S only binds to Mcl-1L possibly neutralizing its anti-apoptotic function [33]. Moreover, the recently identified pro-apoptotic Mcl-1ES isoform may also bind and neutralize the anti-apoptotic function of full length Mcl-1L isoform, though it is expressed at very low or undetectable levels. We therefore analyzed the relative ratios of  [11,34]. Moreover, the prognostic significance of high Mcl-1 protein expression has also been demonstrated in breast, ovarian, non-small cell lung cancer and several hematological malignancies [10,13,35].
Mcl-1 is overexpressed in a variety of human malignancies and proposed to be a potential therapeutic target [8]. Mcl-1 overexpression also appears to be a key factor in the resistance of various cancer types to conventional treatments, including radiation and chemotherapy. The down regulation of Mcl-1 has been shown to sensitize neuroblastoma cells to cytotoxic chemotherapy and resistant melanoma cells to Fas mediated apoptosis [36,37]. Moreover, in our present study, shRNA mediated down regulation of Mcl-1L has shown to chemosensitize oral cancer cells (AW8507, UPCI:SCC040 & UPCI:SCC029B) to Cisplatin indicating a crucial role for Mcl-1 in treatment resistance in oral cancers. Studies targeting Mcl-1 via antisense therapy have also shown to chemosensitize hepatocellular carcinoma cells in vitro [38,39]. Although our study indicates depletion of Mcl-1L expression from 90% to 20% by shRNA ( fig.4a)      Recent studies from our lab have shown a role for Mcl-1L isoform in radioresistance of oral squamous carcinoma cells and as a prognostic factor in prediction of disease-free survival of oral cancer patients treated with definitive radiotherapy. [15,17]. Hence, depletion of Mcl-1 appears to be a prerequisite to radio/ chemo sensitize cancer cells that overexpress Mcl-1 protein, to promote apoptosis. Several attempts have been made to target the anti-apoptotic Bcl-2 family proteins including Mcl-1 in tumors using a variety of inhibitors. Among these, the most promising inhibitor, a BH3 mimetic ABT-737 has failed to overcome resistance due to Mcl-1 overexpression. Further studies indicate that, Mcl-1 down-regulation alone can potentiate ABT-737 lethality in leukemia cells [44]. In the present study a pan Bcl-2 small molecule inhibitor, Obatoclax (GX15-070) which is known to antagonize Mcl-1 and overcome Mcl-1 mediated resistance to apoptosis was used [21]. Obatoclax has been shown to sequester anti-apoptotic Bcl-2 family proteins and up-regulate expression of Puma, Noxa & Bim which further induced apoptosis in neoplastic mast cells [45]. In a recent study, Obatoclax has been shown to reduce levels of anti-apoptotic proteins including Mcl-1 and induced pro-survival autophagy & apoptosis in Head & Neck Squamous Carcinoma Cells [46]. This is the only study available supporting the importance of BH3 mimetic small molecule inhibitors in treatment of oral cancer patients. The present study has demonstrated a synergistic effect of Obatoclax and Cisplatin on oral cancer cell viability and growth, indicating that elimination of anti-apoptotic Mcl-1L is crucial for chemosensitization the oral cancer cells.
This is the first study that demonstrates expression of the three Mcl-1 splice variants in oral cancer cell lines & tumor tissues. Moreover, high Mcl-1L mRNA was significantly associated with advanced tumor size, nodal involvement & poor outcome of oral cancer patients. Also, depleting or inhibiting Mcl-1L levels could successfully chemosensitize oral cancer cells to Cisplatin treatment. These studies suggest that Mcl-1L splice variant is overexpressed in oral cancers and its targeting may be important in the treatment outcome of oral cancer patients.

Conclusion
Our study for the first time assessed Mcl-1 splice variants in oral cancer and demonstrated, anti-apoptotic Mcl-1L isoform to be a chemoresistance and prognostic factor and a potential therapeutic target in oral cancers.