The combination of sorafenib and everolimus shows antitumor activity in preclinical models of malignant pleural mesothelioma

Abstract

Background: Malignant Pleural Mesothelioma (MPM) is an aggressive tumor arising from mesothelial cells lining the pleural cavities characterized by resistance to standard therapies. Most of the molecular steps responsible for pleural transformation remain unclear; however, several growth factor signaling cascades are known to be altered during MPM onset and progression. Transducers of these pathways, such as PIK3CA-mTOR-AKT, MAPK, and ezrin/radixin/moesin (ERM) could therefore be exploited as possible targets for pharmacological intervention. This study aimed to identify 'druggable' pathways in MPM and to formulate a targeted approach based on the use of commercially available molecules, such as the multikinase inhibitor sorafenib and the mTOR inhibitor everolimus.

Methods: We planned a triple approach based on: i) analysis of immunophenotypes and mutational profiles in a cohort of thoracoscopic MPM samples, ii) in vitro pharmacological assays, ii) in vivo therapeutic approaches on MPM xenografts. No mutations were found in 'hot spot' regions of the mTOR upstream genes (e.g. EGFR, KRAS and PIK3CA).

Results: Phosphorylated mTOR and ERM were specifically overexpressed in the analyzed MPM samples. Sorafenib and everolimus combination was effective in mTOR and ERM blockade; exerted synergistic effects on the inhibition of MPM cell proliferation; triggered ROS production and consequent AMPK-p38 mediated-apoptosis. The antitumor activity was displayed when orally administered to MPM-bearing NOD/SCID mice.

Conclusions: ERM and mTOR pathways are activated in MPM and 'druggable' by a combination of sorafenib and everolimus. Combination therapy is a promising therapeutic strategy against MPM.

Figure 1

Immunohistochemical evaluation of the presence and amounts of P-ERM and P-mTOR in MPM samples. Representative fields of epitheloid (A, C) and sarcomatous (B, D) tumor subtypes expressing P-mTOR (A, B) and P-ERM (C, D) are shown. P-mTOR is present in the cytoplasm of malignant mesothelia clusters. P-ERM are diffusely and uniformly expressed on the membrane of neoplastic mesothelial cells.

Figure 2

Inhibition of mTOR and ERK1/2 MAPK pathways by sorafenib (S), everolimus (E) and their combination. A. representative western blot images obtained with in MES-MM98 cells B. Histograms showing quantification of phosphorylated proteins after normalization with the respective total proteins and the housekeeping (vinculin).

Figure 3

Sorafenib and everolimus effect on (A) ERM activation and (B) MES-MM98 cell migration; black line untreated cells, blue line everolimus 10nM, green line sorefenib 2.5 μM, and red line combination-treated cells.

Figure 4

Silencing of ezrin in MES-MM98 cells. A western blot analysis of ezrin protein expression. B. In vivo time course XCelligence analysis of migrating H226 cells after 24 h from mock (control) and ezrin-specific siRNA treatment C. Wound healing assay with MPM cells after 72 hours from treatment with control siRNA (siCTRL), and ezrin-specific siRNA (siEZRIN). The histogram depicts cell migration, calculated as percentage of wound healing after 24 h from the scratch In the text: “We demonstrated that the down-regulation of ezrin slightly reduced cell viability after 24, 48, 72 hours (p < 0.001) and impinged on the migration ability of MPM cells (Figure 4 B,C,D).

Figure 5

Synergistic anti-proliferative effects of sorafenib and everolimus on MES-MM98 cell lines, calculated according to the number of viable cells in culture based on quantitation of the ATP present. (A) Representative dose-effect curve of MES-MM98 cells treated with scalar doses of sorafenib, everolimus or their combination. The dose effect curve was calculated by CalcuSyn software after 72 h of treatment. (B) Photograph of a representative colony growth of MES-MM98 cells after 10 day-incubation in complete medium alone (NT) or with sorafenib (SOR 2.5 μM), everolimus (EV 10 nM), or their combination (SOR 2.5 μM + EV 10 nM). (C) Analysis of the total surface area occupied by colonies/well after treatment with escalating doses of sorafenib (0, corresponding to control, 1.25 μM and 2.5 μM), alone or in combination with everolimus (10 nM). *p < 0.05 vs NT; †, p < 0.05 vs both single agents and control. Results are expressed as mean ± standard deviation of triplicate experiments.

Figure 6

Mechanisms of apoptosis induction by sorafenib and everolimus combination. (A) Bar graph of cell cycle distribution in MES-MM98 cells after 48 h treatment with sorafenib (2.5 μM), everolimus (10 nM) and their combination, compared to untreated control. (B) Induction of apoptosis in MES-MM98 cells, in the absence (blue bars) or in the presence (red bars) of the ROS scavenger BHA. (C) Activation of AMPK by sorafenib treatment. Everolimus potentiated the effect of sorafenib on AMPK activation, reflecting the enhancement of mTOR inhibition and the pro-apoptotic effect of sorafenib and everolimus mediated by c-Jun and p38 activation, triggered by ROS production. PARP cleavage confirmed the apoptotic status. (D) Propidium iodide and Annexin V staining of MES-MM98 cells treated with sorafenib (2.5 μM), everolimus 10 nM, and the p38 inhibitor SB202190. These experiments are representative of 4 different tested cell lines

Figure 7

ROS production in MES-MM98 cells left untreated (control) or treated with everolimus 10 nM, sorafenib 5 μM and their combination for 24 h and then stained with the ROS-specific MitoSOX™ probe. (A) Confocal microscopy photomicrographs. Nuclei are counterstained with Hoechst 33342 bar = 10 μm. (B) Histograms of flow cytometry analysis.

Figure 8

In vivo antitumor effect in human MPM xenograft models. (A) After tumor establishment, 6 mice per group were randomized to daily receive sorafenib 5 mg/kg/day, everolimus 1 mg/kg/day, their combination or vehicle alone. Tumor volume was monitored as described in the “material and methods” section and is reported in graph as mean ± St. Dev (Y error bar). *, p < 0.05 vs vehicle alone; †, p < 0.05 vs either single agent and vehicle alone. (B) Vascular specific CD31 staining and (C) TUNEL assay on vehicle (NT), everolimus 1 mg/kg/day (EV), sorafenib 5 mg/kg/day (SOR), and combination (CB) treated xenografts.