Aberrant activation of the mTOR pathway and anti-tumour effect of everolimus on oesophageal squamous cell carcinoma

Background: The mammalian target of rapamycin (mTOR) protein is important for cellular growth and homeostasis. The presence and prognostic significance of inappropriate mTOR activation have been reported for several cancers. Mammalian target of rapamycin inhibitors, such as everolimus (RAD001), are in development and show promise as anti-cancer drugs; however, the therapeutic effect of everolimus on oesophageal squamous cell carcinoma (OSCC) remains unknown. Methods: Phosphorylation of mTOR (p-mTOR) was evaluated in 167 resected OSCC tumours and 5 OSCC cell lines. The effects of everolimus on the OSCC cell lines TE4 and TE11 in vitro and alone or in combination with cisplatin on tumour growth in vivo were evaluated. Results: Mammalian target of rapamycin phosphorylation was detected in 116 tumours (69.5%) and all the 5 OSCC cell lines. Everolimus suppressed p-mTOR downstream pathways, inhibited proliferation and invasion, and induced apoptosis in both TE4 and TE11 cells. In a mouse xenograft model established with TE4 and TE11 cells, everolimus alone or in combination with cisplatin inhibited tumour growth. Conclusion: The mTOR pathway was aberrantly activated in most OSCC tumours. Everolimus had a therapeutic effect both as a single agent and in combination with cisplatin. Everolimus could be a useful anti-cancer drug for patients with OSCC.

Oesophageal squamous cell carcinoma (OSCC), the major histological type of oesophageal cancer in East Asian countries, is one of the most aggressive malignant tumours (Enzinger and Mayer, 2003). Despite the development of multimodal therapies, including surgery, chemotherapy, and radiotherapy, the prognosis remains poor even for patients who undergo complete carcinoma resection. The limited improvement in outcomes achieved by conventional therapies urges us to seek innovative strategies, especially those involving molecular targeting, for treating OSCC.
The mammalian target of rapamycin (mTOR) is a 289-kDa serine/threonine kinase involved in cellular growth and homeostasis (Bjornsti and Houghton, 2004;Abraham and Gibbons, 2007;Menon and Manning, 2008;Wouters and Koritzinsky, 2008). Mammalian target of rapamycin is activated by phosphorylation as a part of the phosphatidylinositol-3 kinase/AKT signalling pathway (Mita et al, 2003;Chan, 2004;Dancey, 2006) and in turn phosphorylates and activates eukaryotic translation factor 4E (elF4E) and p70 ribosomal S6 kinase (p70S6 kinase), leading to the translation of proteins required for cell cycle progression (Hidalgo and Rowinsky, 2000;Panwalkar et al, 2004). The presence and prognostic significance of aberrant mTOR activation have been reported for several types of human carcinomas (Easton and Houghton, 2006;Herberger et al, 2007;Hou et al, 2007;Hudes, 2009;Hirashima et al, 2010). Our group previously showed an association between high phosphorylated mTOR (p-mTOR) expression and poor prognosis in 143 resected OSCC samples (Hirashima et al, 2010). Mammalian target of rapamycin has recently been recognised as an important and attractive target for anti-cancer therapy (Boulay et al, 2004;Bianco et al, 2006;Johnston, 2006;Antonarakis et al, 2010;Sparks and Guertin, 2010). Everolimus, an oral mTOR inhibitor, has shown particularly promising results in experimental studies, inhibiting tumour growth and displaying anti-angiogenic effects (Carmeliet and Jain, 2000;Bianco et al, 2008;Manegold et al, 2008;Lane et al, 2009). Combination therapy using everolimus and cisplatin has also been reported to be effective (Beuvink et al, 2005;Mabuchi et al, 2007;Hou et al, 2010;Ma et al, 2010). Many clinical trials using everolimus for several types of cancers are currently underway (Yee et al, 2006;Fouladi et al, 2007;Gridelli et al, 2007;Johnson et al, 2007;Awada et al, 2008;O'Donnell et al, 2008;Tabernero et al, 2008;Tanaka et al, 2008;Yao et al, 2008;Wolpin et al, 2009). Everolimus has been already approved for the treatment of advanced renal cell carcinoma in patients whose disease has progressed during or after VEGF-targeted therapy (Coppin, 2010). Nonetheless, to our knowledge, no study has examined the therapeutic effect of everolimus on OSCC using in vitro and in vivo assays.
We therefore conducted this study with three main aims. First, we examined the importance of mTOR activation in OSCC by determining the overall prevalence of p-mTOR expression in OSCC specimens and cell lines. Second, we evaluated the therapeutic effect of everolimus on OSCC cell lines by both in vitro and in vivo assays. Third, we specifically assessed the effect of everolimus in combination with cisplatin, which is one of the most frequently used chemotherapeutic drugs, on OSCC cells.

Reagents and antibodies
Everolimus was provided by Novartis Pharma AG (Basel, Switzerland) and formulated at 2% (w/v) in a microemulsion vehicle. For in vivo analysis, everolimus was diluted to the appropriate concentration in double-distilled water just before administration by gavage. For in vitro analyses, everolimus was prepared in DMSO just before addition to cell cultures. Antibodies recognising mTOR, phospho-mTOR (Ser 2448 ), p70s6k, phospho-p70s6k (Thr 389 ), 4E-BP1, phospho-4E-BP1 (Thr 70 ), and b-actin were purchased from Cell Signaling Technology (Boston, MA, USA).

Patients
The present study involved 167 consecutive patients who underwent surgical resection of OSCC at the Kumamoto University Hospital from January 1996 to December 2007. None of these patients underwent endoscopic mucosal resection, palliative resection, preoperative chemotherapy, preoperative radiotherapy, or preoperative chemoradiotherapy. This study was approved by the Institute Review Board of the Graduate School of Medical Science, Kumamoto University (Approval number: 236; 2 August 2008).

Immunohistochemistry for p-mTOR
The method of immunohistochemical staining for p-mTOR was described previously (Hirashima et al, 2010). Of the 167 tumours, 51 showed no p-mTOR expression, 84 showed weak expression, and 32 showed strong expression. As the aim of the immunohistochemistry in this study was to evaluate the prevalence of p-mTOR expression in OSCC tissues, both weak and strong p-mTOR expression were defined as positive.

Cell culture
Oesophageal squamous cell carcinoma cell lines (TE series) were obtained from the Cell Resource Center for Biomedical Research, Tohoku University. Cell cultures were grown in the recommended medium with 10% foetal bovine serum and incubated in 5% CO 2 at 37 1C.

Cell cycle analysis
The cell cycle distribution of TE4 and TE11 cells treated with everolimus (20 nM) or vehicle (control) for 48 h was analysed by flow cytometry using a BD FACSCalibur (BD Bioscience, San Jose, CA, USA) according to previously published methods Milella et al, 2004).

Apoptosis analysis
TE4 and TE11 cells were treated with everolimus (20 nM) or vehicle (control) for 48 h and then apoptosis was assessed by flow cytometry using Annexin V-FITC (BD Bioscience) and propidium iodide (PI) staining according to previously published methods (Del Milella et al, 2004).

Invasion analysis
To evaluate the effect of everolimus on cell invasiveness, a Matrigel Invasion Chamber (BD Bioscience) was used according to the manufacturer's protocol. Matrigel-coated chambers containing 8 mm pore-size filters were fitted into 24-well tissue culture plates. Briefly, cells of each type (TE4, 1.0 Â 10 5 cells ml À1 ; TE11, 5.0 Â 10 5 cells ml À1 ) were seeded into the Matrigel-coated chambers in RPMI-1640 medium with everolimus (20 nM) or vehicle (control) and incubated at 37 1C in 5% CO 2 for 24 h. The invasive cells on the bottom sides of the filters were stained using Toruijin blue dye, and the numbers of cells in five randomly selected fields at Â 200 magnification were counted.

Subcutaneous xenograft model
All the procedures involving animals and their care were approved by the Animal Care and Use Committee of Kumamoto University. These procedures meet the standards required by the United Kingdom Coordinating Committee for Cancer Research (UKCCCR) guidelines (Workman et al, 2010). A subcutaneous xenograft model was used to assess the therapeutic effect on OSCC cells of everolimus either as a single agent or in combination with cisplatin, one of the chemotherapeutic drugs most frequently used for OSCC in the clinical setting. Six-week-old nude mice (BALB/c) (n ¼ 24) were inoculated subcutaneously in the right or left flank with 5 Â 10 6 TE4 cells and TE11 cells in 200 ml of PBS. Some mice showed insufficient tumour growth and were therefore excluded from the study, leaving a total of 22 mice used for the single in vivo experiment. When the tumours reached approximately 50 -70 mm 3 , the mice were randomised into four treatment groups (n ¼ 5 -6 mice per group). The first group was treated twice a week with placebo. The second group was treated twice a week with everolimus (5 mg kg À1 ). The third group was treated every 2 weeks with cisplatin (3 mg kg À1 ). The fourth group was treated twice a week with everolimus (5 ml kg À1 ) and every 2 weeks with cisplatin (3 mg kg À1 ) ( Figure 4A). The validity of these everolimus and/or cisplatin protocols has been demonstrated in an ovarian cancer model. Everolimus was administered by oral gavage using an animal-feeding needle. Cisplatin was injected intraperitoneally. Body weight was measured every 3 days. Calliper measurements of the longest perpendicular tumour diameters were made weekly using a digital calliper, and tumour volumes were estimated using the following formula: V ¼ L Â W Â D Â p/6, where V is the tumour volume, L the length, W the width, and D the depth (Mabuchi et al, 2007).

Statistical analysis
For the in vitro assays, including the cell proliferation assay, cell cycle ratio assay, apoptosis assay, and invasion assay, statistical analyses were performed using Mann -Whitney's U-test for unpaired samples. For the in vivo experiment, body weight and tumour volume were compared among placebo-, everolimus-, cisplatin-, and everolimus plus cisplatin-treated mice using the Wilcoxon exact test. Statistical analysis was performed with Stat View-J 5.0 software (Abacus Concepts, Inc., Berkeley, CA, USA). A two-sided significance level of Po0.05 was used for all the statistical analyses.

Phosphorylated mTOR expression in OSCC specimens and cell lines
We assessed p-mTOR expression (i.e., mTOR activation) by immunohistochemistry. Of the 167 OSCC specimens, 116 (70%) were positive for p-mTOR expression ( Figures 1A and B). The high percentage of p-mTOR-positive tumours supports the crucial role of mTOR activation in the pathogenesis of OSCC.
All five human OSCC cell lines (TE1,4,9,11,and 13) examined in the current study showed p-mTOR expression in vitro; the expression level was highest in TE4 cells and lowest in TE11 cells ( Figure 1C). Therefore, both TE4 and TE11 cells were used in the following experiments.

Therapeutic effect of everolimus on OSCC cell lines in vitro
Everolimus (20 nM) treatment for 48 h significantly inhibited the proliferation of both TE4 and TE11 cells ( Figure 3A). In order to clarify the effect of everolimus on the cell cycle, OSCC cells were treated with everolimus (20 nM) and then subjected to cell cycle analysis by flow cytometry. An accumulation of cells in the G 0 /G 1 phase and a reduction in the S-phase fraction were observed in both TE4 and TE11 cells treated with everolimus (20 nM) for 48 h ( Figure 3B). Everolimus (20 nM) also significantly increased the proportion of early apoptotic cells (Annexin V-FITC positive, PI negative) compared with that of vehicle-treated cells in both TE4 and TE11 cells treated for 48 h ( Figure 3C), indicating that everolimus could induce early apoptosis in these cell lines. Western blot analysis utilising antibodies for Bad and PARP also showed the induction of apoptosis by everolimus (Supplementary Figure 1); everolimus (20 nM) increased the expression of Bad and cleaved PARP protein. Finally, we performed an in vitro invasion assay using Matrigel Invasion Chambers and found that everolimus (20 nM) significantly decreased the numbers of invading TE4 and TE11 cells compared with those of vehicle-treated cells ( Figure 3D).

Everolimus inhibits tumour growth in a mouse subcutaneous xenograft model
The mean tumour volumes on day 36 in a mouse xenograft model established with TE4 cells were 1314±134, 311±87, 542±161, and 159±21 mm 3 in mice treated with placebo, everolimus, cisplatin, and everolimus plus cisplatin, respectively (Table 1, Figure 4B). Treatment with everolimus or cisplatin alone decreased the tumour burdens by 83% and 68%, respectively, compared with that of placebo-treated mice ( Figure 4C), indicating that everolimus used as a single agent has marked anti-tumour activity. Moreover, treatment with cisplatin plus everolimus decreased the tumour burden by 92% ( Figure 4C), suggesting that the use of everolimus and cisplatin as a combination therapy might be promising. Similar results were obtained for TE11 cells (Supplementary Figure 2). The weight changes of the mice over the course of the treatments did not differ significantly among the four groups, as shown in Supplementary Figure 3. In addition, we confirmed histologically that there were no differences in the levels of injury to the organs, including liver, kidney, pancreas, lung, intestine, and skin, among these four groups (Supplementary Figure 4). We continued to follow these mice for 2 months. Although all of the mice in the placebo group died within 2 months, no mouse in any of the other three groups died during this period.

DISCUSSION
Mammalian target of rapamycin is a key regulator of cell growth and proliferation and as such is regarded as a promising target for anti-cancer therapy (Kapoor, 2009;Scott et al, 2009). In this study, we made three intriguing findings. First, most OSCC tumours were positive for p-mTOR expression, supporting a role for mTOR activation in the pathogenesis of OSCC. Second, everolimus, an oral mTOR inhibitor, had a therapeutic effect on OSCC cell lines in vitro. Third, combination therapy with everolimus and cisplatin showed an additive effect on OSCC cells in vivo. Our findings certainly suggest that everolimus could be useful as an anti-cancer drug for patients with OSCC. Previous studies have shown the importance of mTOR activation in OSCC specimens: Boone et al, (2008) detected activated mTOR in 25% of patients with OSCC, a subset of patients that might potentially benefit from mTOR-inhibiting therapy. Yoshioka et al, (2008) demonstrated that 48% of OSCC tumours showed high levels of p-mTOR phosphorylation. In the current study utilising 167 OSCC samples, about 70% of the OSCC tumours showed p-mTOR (i.e., mTOR activation). This discrepancy might be due to a difference in the method used to evaluate mTOR phosphorylation or in the cutoff for p-mTOR positivity. Nonetheless, these two previous studies and the current study certainly support the hypotheses that mTOR activation is important in the pathogenesis of OSCC and that mTOR inhibitors might be useful for OSCC treatment.
Everolimus, an orally bioavailable derivative of rapamycin, is a promising drug for cancer therapy. However, to our knowledge, no previous study has utilised in vitro and in vivo models to evaluate the therapeutic efficacy of everolimus. First, we demonstrated that everolimus suppressed down-stream signalling (i.e., phosphorylation of p70s6 kinase and 4E-BP1) and significantly inhibited cell proliferation and invasion of mTOR-activated OSCC cell lines in vitro. Second, we showed that inhibition of mTOR signalling by everolimus induced G 0 /G 1 arrest and apoptosis, suggesting that everolimus might inhibit anti-apoptotic or survival signalling in OSCC cell lines. Third, we found that treatment with everolimus significantly inhibited tumour growth in vivo. Taken together, these results indicate that everolimus as a single agent could have significant anti-tumour efficacy against OSCC cells.
The effects of everolimus were more prominent in TE11 cells (p-mTOR-low) than in TE4 cells (p-mTOR-high). The malignant characteristics of OSCC cells are likely acquired not only through the mTOR signalling pathway but also through a wide variety of other signalling pathways. Although the activation level of the mTOR pathway was lower in TE11 cells than in TE4 cells, TE11 cells might depend more heavily on the mTOR pathway for their malignant behaviour. On the other hand, although TE4 cells showed a high level of mTOR activation, they might rely more on the other signalling pathways than on the mTOR pathway. In this study, the mTOR pathway was activated in all five cell lines assessed. If we could obtain an OSCC cell line without mTOR activation, those cells might well be resistant to everolimus. Future studies are necessary to confirm our findings as well as to elucidate the biological mechanisms by which the mTOR activation level affects the therapeutic efficacy of everolimus.
Interestingly, we also found an additive effect of everolimus and cisplatin on OSCC cells in an in vivo model. A similar effect has been reported for other types of carcinomas. Beuvink et al, (2005) reported that everolimus could sensitise cells to cisplatin by inhibiting induction of p21 expression by p53. Ma et al, (2010) showed that everolimus exerts an additive-to-synergetic effect on cisplatin-induced growth inhibition in nasopharyngeal carcinoma. Unfortunately, our current experiment could show only an additive effect rather than a synergetic effect. We expect that additional experiments in the future might be able to show a synergetic effect. However, we were at least able to recognise that combining everolimus and cisplatin might be a useful therapeutic strategy. As cisplatin is one of the most important chemotherapeutic drugs for OSCC treatment, our finding may have significant clinical implications.
In conclusion, most OSCC tumours showed mTOR activation, suggesting that mTOR could be a promising target for anti-cancer therapy against OSCC. Everolimus had a therapeutic effect on OSCC cells both in vitro and in vivo, and combination therapy with everolimus and cisplatin showed an additive effect. Although further experimental studies are necessary to confirm our findings, the current study certainly provides further rationale for future clinical trials of everolimus (in combination with cisplatin) in OSCC patients.