Escin induces caspase-dependent apoptosis and autophagy through the ROS/p38 MAPK signalling pathway in human osteosarcoma cells in vitro and in vivo

Osteosarcoma is one of the most malignant neoplasms in adolescents, and it generally develops multidrug resistance. Escin, a natural mixture of triterpene saponins isolated from Aesculus hippocastanum (horse chestnut), has demonstrated potent anti-tumour potential in vitro and in vivo. In the present study, we found that escin inhibited osteosarcoma proliferation in a dose- and time-dependent manner. Additionally, escin-induced apoptosis was evidenced by the increased expression of caspase-related proteins and the formation of apoptotic bodies. Escin also induced autophagy, with elevated LC3, ATG5, ATG12 and Beclin expression as well as autophagosome formation. Inhibition of escin-induced autophagy promoted apoptosis. Moreover, p38 mitogen-activated protein kinases (MAPKs) and reactive oxygen species (ROS) were activated by escin. A p38 MAPK inhibitor partially attenuated the autophagy and apoptosis triggered by escin, but a ROS scavenger showed a greater inhibitory effect. Finally, the therapeutic efficacy of escin against osteosarcoma was demonstrated in an orthotopic model. Overall, escin counteracted osteosarcoma by inducing autophagy and apoptosis via the activation of the ROS/p38 MAPK signalling pathway; these findings provide evidence for escin as a novel and potent therapeutic for the treatment of osteosarcoma.

Osteosarcoma is one of the most malignant primary bone tumours, primarily affecting adolescents and young adults. 1,2 Whereas conventional chemotherapy is an important treatment option for osteosarcoma, it has many disadvantages. 3 For example, chemotherapeutics tend to be highly toxic to normal tissues, leading to anaemia, neutropenia, thrombocytopenia, heart damage and a series of adverse reactions, thereby decreasing the patient survival rate. 4,5 Therefore, new agents with fewer side effects and better therapeutic benefits are needed.
Escin, the main active compound of horse chestnuts, 6 has been widely used to treat swelling, exudation, and inflammation. 7 Recently, accumulated experimental evidence has shown that escin is a potential cancer chemotherapeutic and has elucidated some of the underlying mechanisms. [8][9][10] However, the principal mechanism underlying its anti-tumour effect is not fully understood. Furthermore, the effects of escin effect on human osteosarcoma and its underlying mechanisms have not been reported.
Apoptosis is the classic mechanism by which chemotherapy agents induce cell death; 11 however, numerous studies have shown that these drugs also induce autophagy. 12 Autophagy is usually considered to be a critical process by which cells recycle ineffectual cellular components to maintain homeostasis. This process plays a critical role in programmed cell death type II, which can be exploited to suppress tumour growth. 13 Paradoxically, autophagy also serves as a cell survival pathway. 14,15 Previous studies 16,17 have demonstrated the dual roles of autophagy in neoplastic development; it can function as a suppressor of early-stage carcinoma and a promoter of later stage disease. 18 Hence, the relationship between apoptosis and autophagy has not been fully elucidated, especially in the context of escin-induced osteosarcoma cell death.
Increasing evidence supports that mitogen-activated protein kinases (MAPKs) mediate apoptosis and autophagy in response to chemotherapy. 19,20 MAPKs are key players in the regulation of cell proliferation and cancer development. Activation of p38 MAPK could augment the processes of apoptosis and autophagy. 21 Meanwhile, MAPK signal transduction cascades could be regulated by reactive oxygen species (ROS). [22][23][24] Intracellular redox status has been reported to be associated with carcinogenesis and antineoplastic processes. 25,26 ROS are small and highly reactive molecules that play key roles in regulating physiological processes and maintaining redox balance. 27 Although an appropriate amount of intracellular ROS conditions can improve cell proliferation. Recent studies 28,29 have reported that excessive amounts of ROS cause oxidative damage to cells, especially malignant cells, by inducing autophagy and apoptosis. Accordingly, targeting relevant signalling pathways, especially the ROS and p38 MAPK pathways, may be a good choice for osteosarcoma treatment.
In the current study, we demonstrated that escin has antitumour potential against osteosarcoma. We further investigated the underlying mechanisms, and the results suggest that escin may induce apoptosis and autophagy in osteosarcoma through the ROS/p38MAPK pathway.

Results
Escin has an anti-proliferation effect on osteosarcoma.
To investigate the effect of escin on tumour growth, four osteosarcoma cell lines MNNG/HOS, Saos-2, MG-63 and U2-OS were exposed to different concentrations of escin for 24 or 48 h. Then, cell viability was evaluated using the MTS assays ( Figure 1a). The IC50 of escin at 24 h was 30.44 μM Figure 1 Escin inhibits human osteosarcoma cell proliferation through apoptosis. (a) Four osteosarcoma cell lines and HUVEC cells were treated with different concentrations of escin for 24 or 48 h. MTS assays were performed. The control group was treated with 0.1% DMSO. The data are presented as the mean of six replicates, and each experiment was performed three times. (b,c) The HOS and Saos-2 osteosarcoma cell lines were incubated with escin or control for 24 h, stained with an Annexin V-PE/7-AAD kit and analysed by flow cytometry. The proportion of apoptotic cells is presented as the mean ± S.D. of three replicate experiments. (d) Apoptotic morphology was observed by transmission electron microscopy. The red arrows indicate apoptotic bodies and nuclear condensation. Scale bar, 2 μm. *Po0.05 compared with control for HOS cells, 29.93 μM for Saos-2 cells, 25.51 μM for MG-63 cells, and 32.40 μM for U2-OS cells. Moreover, in the previous studies, 30 escin shows cytotoxic effects on control cell especially after treatment with escin at 48 h. Furthermore, our study showed that escin has less cytotoxic effects on control cell at 24 h than 48 h, so 24 h treatment is beneficial for cancer studies. The IC50 value of escin for HUVEC cells was 95.81 μM for 24 h and 37.92μM for 48 h. These results indicated that escin could inhibit the proliferation of osteosarcoma cells in a time-and dose-dependent manner. MNNG/HOS and Saos-2 cells harbour mutant p53 and are more chemoresistant to current chemotherapeutics; 31,32 these cells were used in subsequent assays. We chose concentrations of 20, 30 and 40 μM escin and a timepoint of 24 h for the following experiments based on data from previous studies. [33][34][35] Escin induces apoptosis in osteosarcoma cells. To determine whether apoptosis is responsible for escininduced cell death, we detected apoptotic cells, apoptosisrelated proteins and morphological changes. Annexin-PE/PI staining results indicated that escin induced the apoptosis of osteosarcoma cell lines ( Figure 1b). The apoptosis percentage of HOS cells in 40 μM group is 70.8%, which is significantly higher than 5.13% in control group. The Saos-2 cells apoptosis percentage in 40 μM group is 56.93%, which is 11 times greater than the apoptosis percentage in control group. Consistent with the MTS results, escin induced the apoptosis of osteosarcoma cells in a dose-dependent manner ( Figure 1c). As shown in Figure 1d, nuclear condensation and apoptotic bodies were clearly observed. Figure 2a shows the appearance of apoptotic chromatin condensation and DNA fragmentation. Overall, these results demonstrate that escin induces apoptosis in osteosarcoma cells.
Escin induces caspase-dependent apoptosis through mitochondrial-mediated pathways. To explore the molecular mechanism of escin-induced apoptosis, we performed western blots to assess apoptosis-related proteins. As shown in Figure 2b and Supplementary figure 1, the protein levels of cleaved caspase-3, -7, -8, -9 and poly(ADPribose) polymerase (PARP) was 1.85, 1.28, 1.2, 1.53 and 2.03 times higher than control group of HOS and 2.7, 1.37, 1.41, 1.66 and 3.21 times higher than control group of Saos-2 after treatment with 40 μM escin, respectively. Bcl-2 family members are known to be involved in the mitochondrial-mediated apoptosis pathway, which activates the caspase cascade. 36 Thus, we investigated Bax, Bcl-XL and Bcl-2 key members of the Bcl-2 family. Escin of 40 μM group upregulated Bax protein levels (1.46 times higher than control group in HOS and 2.61 times higher than control group in Saos-2) and downregulated the expression level of Bcl-XL (decreasing 63% than control group in HOS and 51% than control group in Saos-2) and Bcl-2 (decreasing 39% than control group in HOS and 53% than control group in Saos-2) (Figure 2b and Supplementary figure1). These results indicate that escin induces caspasedependent apoptosis through mitochondrial-mediated pathways.
Escin induces autophagy in osteosarcoma cells. As autophagy is involved in an alternative cell death mechanism, we explored its role in escin-induced tumour inhibition. First, transmission electron microscopy (TEM) was used to directly confirm autophagosome formation ( Figure 3a). LC3 is an important regulator of autophagy that promotes autophagosome formation and ATG5 and ATG12 form a complex that is essential for autophagy. 37 The ATG5-ATG12 complex promotes the conjugation of LC3BI to phosphatidylethanolamine to form LC3BII, which is then recruited to autophagosomal membranes. 38 Figure 3b shows that escin upregulated the levels of LC3BII, ATG-5, ATG-12 and Beclin-1 and downregulated LC3IB levels, indicating that escin regulates autophagy. Next, we investigated whether escin-induced autophagy contributes to cell death. Cell viability was assessed in the presence of 3-MA (an autophagy inhibitor) or z-VAD-fmk (an apoptosis inhibitor). 3-MA moderately reduced escin-induced cell death, while z-VAD-fmk strongly reduced cell death ( Figure 3c). These data suggest that escin-induced autophagy promotes cell death and that escin causes both apoptotic and autophagic cell death.
Escin inhibits osteosarcoma through the p38 MAPK pathway. The activated MAPK signalling pathway can induce cell death in cancer. 39 Thus, we explored whether this pathway is activated by escin in osteosarcoma. The expression levels of p38, JNK and ERK-2 were determined by western blot analysis. Escin upregulated p38 expression in a dose-and time-dependent manner (Figure 4a) but had a minimal impact on JNK and ERK-2 (data not shown). Furthermore, when cells were precultured with the p38 MAPK inhibitor, the proportion of apoptotic cells and the expression of apoptosis-and autophagy-related proteins were decreased (Figures 5b-e). These results indicated that the p38 MAPK pathway is essential for escin-induced apoptosis and autophagy.
Escin-induced ROS initiates apoptosis and autophagy in osteosarcoma cells through the ROS/p38 MAPK pathway. ROS usually plays an important role in regulating apoptosis and autophagy. 40,41 Thus, Escin-treated osteosarcoma cells were stained with DCFH-DA to assess ROS generation. As shown in Figures 4b and c, escin induced ROS generation in a dose-dependent manner. The ROS scavenger N-acetyl L-cysteine (NAC) was used to confirm the role of ROS in escin-induced cell death. NAC decreased ROS levels (Figure 4d). Less autophagic vacuoles and chromatin condensation were observed in cells pretreated with NAC than the p38/MAPK inhibitor (Figure 5a). Moreover, NAC reversed escin-induced death and apoptosis proportion to a greater extent than the p38/MAPK inhibitor (Figures  5b-d). In addition, NAC inhibited the escin-induced activation of apoptosis and autophagy-related proteins and strongly blocked p38 phosphorylation (Figure 5e). These results revealed that ROS may be the important factor upstream of p38 MAPK that initiates escin-induced apoptosis and autophagy.
Escin inhibits osteosarcoma in vivo. To simulate the circumstance of osteosarcoma in vivo, we established an in vivo orthotopic model of osteosarcoma by inoculating osteosarcoma cells (Saos-2 cells transferred with luciferase) into the tibia of nude mice. Tumour size was calculated based on luminescence intensity. Escin at doses of 1.4 mg/kg and 2.8 mg/kg caused a decrease in tumour luminescence intensity after 7 days of drug administration, and there was a significant difference between the two groups after 21 days (Figures 6a and b). X-ray analysis showed that escin minimized osteoclasia during osteosarcoma development (Figure 6c). The tumour-located in the right leg was excised (Figure 6d). The H&E, Ki-67 staining and terminal deoxynucleotidyl transferase-mediated (d)-UTP nick-end labelling (TUNEL) assays revealed more tumour cell death after escin treatment. The mean optical density was calculated using Image-pro software, and this immunohistochemical analysis confirmed the increased expression of LC3, caspase-3 and p38/MAPK induced by escin (Figure 6e). These results indicated that escin inhibits the growth of osteosarcoma in vivo.

Discussion
Osteosarcoma is one of the most malignant neoplasms in adolescents. Current therapies incorporate surgery and combination chemotherapy, which cures nearly 70% of patients. However, the survival of metastatic osteosarcoma patients has remained virtually unchanged over recent decades, with an overall 5-year survival rate of nearly 20%. 42 Thus, there is an urgent need to discover innovative drugs for the treatment of osteosarcoma. An increasing number of studies have shown that escin has anti-tumour potential against various types of cancer. 8,[43][44][45] However, the corresponding mechanisms remain to be elucidated. In this study, we confirmed that escin inhibited osteosarcoma growth by inducing apoptosis and autophagy through the ROS/p38 MAPK pathway.
Apoptosis is the main pathway of drug-induced cell death. In this study, we found that escin activated caspase-dependent apoptosis through mitochondrial-mediated pathways, upregulated the expression of the pro-apoptotic protein Bax and Autophagy is involved in type II programmed cell death, particularly in apoptosis-deficient cells, and can be exploited to suppress tumour growth. 46 In accordance with previous studies 8,47-49 that used TEM observation to show apoptotic and autophagic death, our study used TEM observation and western blot analysis to confirm the formation of autophagosomes and the elevated expression of autophagic markers after escin treatment, respectively. Beclin-1 is a pivotal protein in autophagy, and the upregulation of Beclin-1 can lead to autophagy. 50,51 We found that escin upregulated Beclin-1 expression. While many studies 52,53 have demonstrated that autophagy functions as a survival mechanism in cancer development, our results showed that escin-induced cell death was moderately diminished by the autophagy inhibitor 3-MA, demonstrating the contribution of autophagy to cell death instead of survival.
To further explore the upstream pathways, we tested the potential ones that might be involved in the process of apoptosis and autophagy. The p38/MAPK pathway was reported to be involved in regulating apoptosis and autophagy. 54 Recent studies have found that higher p38/ MAPK expression correlates with the increased expression of autophagic markers, such as LC3B and ATG5/ATG12, and apoptotic markers, such as caspase-3 and PARP. 21,55,56 In the It is well documented that ROS regulate the expression of p38/MAPK. [57][58][59] In the present study, increased ROS generation was observed after escin treatment. In addition, the ROS scavenger NAC strongly reversed escin-induced apoptosis and autophagy; in this context, NAC was more potent than the inhibitor of p38/MAPK phosphorylation. Taken together, we concluded that escin induced the autophagy and apoptosis of osteosarcoma cells through the ROS/p38 pathway.
It was reported previously that escin treatment resulted in significant tumour inhibition in vivo. [60][61][62] Our in vivo study indicated that escin significantly inhibited osteosarcoma growth during the 3-week treatment period. Moreover, immunohistochemical analysis confirmed increased expression of LC3, caspase-3 and p38/MAPK after escin treatment. Furthermore, X-ray analysis showed that osteoclasia induced by osteosarcoma was minimized by escin.
In conclusion, escin can inhibit osteosarcoma cell proliferation by inducing autophagy and apoptosis mediated by the   Figure 6 Escin inhibits the growth of human osteosarcoma xenografts in vivo. Saos2-luc cells were inoculated orthotopically in the right tibia of BALB/c-nu mice. After 3 days, the tumour in the tibia was assessed using an in vivo imaging system, and the luminescence intensity was used as an indicator of tumour size. Then, the mice were separated randomly into three groups. The next day, the mice began receiving daily intraperitoneal injections of PBS or escin (1. Morphological apoptosis. Morphological changes of apoptosis were measured by fluorescence microscopy using Hoechst 33258 staining. Briefly, cells were treated with different concentrations of escin for 24 h and then exposed to Hoechst 33258 for approximately 10 min. Then, the cells were washed with PBS twice. Fluorescence microscopy (Olympus, Tokyo, Japan) was used to visualize morphological changes, such as chromatin condensation.
Transmission electron microscopy detection. TEM was used to observe the cell ultra-structure. Changes regarding apoptotic body and nuclear condensation indicated apoptosis, and the formation of autophagosome indicated autophagy. Cells treated with escin were first fixed with 2.5% glutaraldehyde and subsequently fixed with 1% osmium tetroxide. Cells were embedded in Epon after dehydrated with different concentrations of alcohol. Ultrathin sections (0.5 μm) were generated for observation under a transmission electron microscope. Western blot analysis. Cells were treated with different concentration of escin (0-40 μM) for 24 h. Then, the cells were collected and lysed with RIPA lysis buffer, which included a protease inhibitor cocktail (Sigma-Aldrich). The lysing process lasted for 30 min on ice, and the solution was vortexed every 10 min. Cell lysates were centrifuged at 13 000 × g for 15 min at 4°C. Then, the supernatant was collected, and the total amount of protein was quantified using a BCA protein assay kit (Pierce, Waltham, MA, USA). Equal protein was separated by 8-12% SDS-PAGE at 80 V for 1.5 h. Then, the proteins were transferred onto 2.2 μm polyvinylidene fluoride membranes (Millipore) at 250 mA for 2 h in a humid atmosphere. The membranes were blocked with 5% bovine serum albumin (Sigma-Aldrich) solution for 1 h at room temperature. Then, the membranes were incubated with the primary antibody (Cell Signaling Technology) at the recommended dilution at 4°C overnight. After being washed with TBST for 30 min, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Cell signaling technology) at room temperature for 1 h. The target bands were developed by enhanced chemiluminescence kit (Millipore).
Xenograft orthotopic model. Male BALB/c-nu mice (4 weeks old) were obtained from Shanghai Laboratory Animal Center of Chinese Academy of Sciences. The mice were maintained under specific pathogen-free conditions. Saos-2 cells were transfected with luciferase (Saos-2-luc) for in vivo imaging. Saos-2-luc cells (5 × 10 6 ) suspended in 50 μl of PBS were injected into the right tibia of each mouse. On the third day, the luminescence intensity was measured using an in vivo bioluminescence imaging system. When the luminescence intensity reached 1 × 10 6 p/s, mice were randomly divided into three groups (five mice per group). Mice with a very low or high luminescence signal were killed. Then, the mice received daily intraperitoneal injections of 100 μl of PBS, 1.4 mg/kg of escin or 2.8 mg/kg of escin. The luminescence intensity was measured weekly. After 21 days of dosing, all mice were killed. The tumours were excised and fixed in 10% formalin for further analysis. All treatments were approved by the Research Ethics Committee of the Second Affiliated Hospital of Zhejiang University School of Medicine, China.
Tumour histology. The excised tumour were fixed with 4% formalin, dehydrated in a graded alcohol series and embedded in paraffin. Tumour tissue was cut into serial sections (3 μm) that were stained with H&E after deparaffinization for histological analysis.
In situ TUNEL staining. A TUNEL kit (Roche Diagnostics, Mannheim, Germany) was used to evaluate apoptosis. Briefly, after deparaffinization and hydration, the sections were incubated with proteinase K (20 μg/ml) for 15 min at room temperature. After permeabilization, slides were incubated with the TUNEL reaction mixture. HRP-conjugated Fab fragments were used to detect labelled DNA strand breaks.
Immunohistochemistry analysis. For immunohistochemical analysis, sections were deparaffinized in xylene and hydrated with gradient increased alcohol. To block endogenous peroxidase activity, the sections were treated with 3% H 2 O 2 for 15 min. Next, sections were immersed in boiling sodium citrate buffer (pH 6.0) for 10 min. Then, the slides were treated with 10% goat serum for 15 min at room temperature and incubated with antibodies against cleaved caspase 3, Ki-67, LC3-B, and phospho-p38 MAPK at 4°C overnight. After washing five times with PBS, the slides were incubated with biotin-labelled secondary antibody at 37°C for 30 min. Immunoreactivity was detected by SABC method. A DP70 CCD camera (Olympus) couple with an AX-70 microscope (Olympus) was used to record images. Image-pro-plus (version. 6.0, Media Cybernetics) was used for digital image analysis. The measure parameters contained mean density, total area and IOD. The optical density was calibrated and the area of interest assigned value for hue, 0-30; saturation, 0-255; intensity, 0-230. Then the image was transformed into a greyscale image, and values were measured.
In vivo bioluminescence assay. For in vivo imaging, 200 μl of luciferin (15 mg/ml) was injected intraperitoneally into the mice. After approximately 10 min, the mice were anaesthetized with isoflurane. In vivo imaging was performed using an IVIS 200 imaging system, and the results were analysed with Living Image Software (Version 3.0.4, Xenogen, Hopkinton, MA, USA). Total flux of the region of interest was measured in photons (p)/s for each mouse.
X-ray. X-rays were utilized to identify the tumour-associated osteoclasia in the tibia. The mice with an orthotopic tumour underwent X-ray analysis. X-rays were