Evaluation of Apoptosis and Autophagy Inducing Potential of Berberis aristata, Azadirachta indica, and Their Synergistic Combinations in Parental and Resistant Human Osteosarcoma Cells

Cancer is a multifactorial disease and hence can be effectively overcome by a multi-constituently therapeutic strategy. Medicinal plant extracts represent a perfect example of such stratagem. However, minimal studies have been done till date that portray the effect of extraction techniques on the phyto-constituent profile of plant extracts and its impact on anticancer activity. In the present study, we have evaluated the anticancer potential of methanolic extracts of Berberis aristata root and Azadirachta indica seeds prepared by various extraction techniques in human osteosarcoma (HOS) cells. Soxhlation extract of B. aristata (BAM-SX) and sonication extract of A. indica (AIM-SO) were most effective in inducing apoptosis in parental drug sensitive, as well as resistant cell type developed by repeated drug exposure. Generation of reactive oxygen species and cell cycle arrest preceded caspase-mediated apoptosis in HOS cells. Interestingly, inhibition of autophagy enhanced cell death suggesting the cytoprotective role of autophagy. Combination studies of different methanolic extracts of BAM and AIM were performed, among which, the combination of BAM-SO and AIM-SO (BAAISO) was found to show synergism (IC50 10.27 µg/ml) followed by combination of BAM-MC and AIM-MC (BAAIMC) with respect to other combinations in the ratio of 1:1. BAAISO also showed synergism when it was added to cisplatin-resistant HOS cells (HCR). Chromatographic profiling of BAM-SX and AIM-SO by high performance thin layer chromatography resulted in identification of berberine (Rf 0.55), palmitine (Rf 0.50) in BAM-SX and azadirachtin A (Rf 0.36), azadirachtin B (Rf 0.56), nimbin (Rf 0.80), and nimbolide (Rf 0.43) in AIM-SO. The cytotoxic sensitivity obtained can be attributed to the above compounds. Our results highlight the importance of extraction technique and subsequent mechanism of action of multi-constituential B. aristata and A. indica against both sensitive and drug refractory HOS cells.

inTrODucTiOn Nature has always been the mainstay of drug discovery program for human health. Around 80% of the drugs in the market for various human diseases are either natural products or nature inspired (1). Among all the human diseases, the major contribution of natural products is in the treatment of cancer (2). There are many plant-derived anticancer agents, namely vinblastine, vincristine (Catharanthus sp.), epipodophyllotoxins (Podophyllum sp.), paclitaxel (Taxus sp.), and camptothecin derivatives (Camptotheca sp.) (1,2). In-spite of multiple drugs being available in the market, cancer is still one of leading causes of fatality worldwide due to development of chemoresistance (3). Chemoresistance is one of the major challenges in treatment of all types of cancer and is thought to be inherent in certain populations of heterogeneous tumors or it may be acquired due to repeated drug exposure (4). Osteosarcoma, a common malignant bone tumor, is no exception affecting 2.7% of Indians. Surgery along with chemotherapy (methotrexate, doxorubicin, and cisplatin) is possible treatment options for osteosarcoma (5). However, these drugs develop chemoresistance on regular use, hence, strategies need to be developed to overcome the challenge of cancer and associated resistance.
In Ayurveda, various classical texts are found where plant crude extracts have been reported to be used in treatment of various tumors. The multicomponent system can often work in synergism, if explored judiciously. However, there are very few scientific studies conducted till date that facilitates understanding of the synergism of these herbal drugs (6). As a herbal medicine, Azadirachta indica (Meliaceae), known as "neem" possess phytochemicals used for anti-inflammatory properties. Isoprenoids (triterpenoids) are the major class of chemical constituents of A. indica (7) that constitutes more than 200 compounds in which azadirachtin ( Figure 1A1) is a major compound, followed by nimbolide ( Figure 1A2) and nimbin ( Figure 1A3). Another plant, Berberis aristata (Berberidaceae), is known for its antiinflammatory and immune-potentiating properties. The roots of B. aristata contain protoberberine alkaloids such as, berberine ( Figure 1B4), oxyberberine, epiberberine, palmitine ( Figure 1B5), and bis-isoquinoline alkaloids as its main constituents (8,9).
Extraction of the crude plant material is a crucial step prior to their bio-evaluation. The most common extraction techniques used in laboratory are hot percolation, ultrasonication, and maceration (10). In all the extraction techniques, an important parameter is the type of solvent used. Non-polar solvents (e.g., hexane or petroleum ether) extracts sterols, terpenes; while, polar solvents such as methanol or ethanol help in extracting out polar compounds including alkaloids and polyphenols (11). Majority of research articles do not compare and correlate the biological efficacy of plant extracts with their extraction procedures. On the other hand, they mostly report the use of any one of the above techniques for preparation and bio-evaluation, which do not necessarily represent the true scenario. Thus, contextual optimization is needed to precisely understand the best extraction procedure to obtain the most effective biological results. The present study evaluates the chemotherapeutic efficacy of various extracts of A. indica seeds, B. aristata roots, and their combinations against cisplatin sensitive and resistant osteosarcoma cells. The study also highlights the comparison and correlation of the observed biological efficacy of above plant extracts with type of extraction techniques used.

Botanical Materials
Roots of B. aristata was collected from Mandi, Himachal Pradesh and seeds of A. indica was collected from the BITS-Pilani campus, Rajasthan. The plant materials were authenticated by a botanist in NIPER, SAS Nagar, India. Samples of the same have been deposited in the institute herbarium.

chemical and reagents
Toluene, benzene, n-butanol, and ethylacetate were purchased from S. D. Fine Chemicals Ltd., Mumbai and acetic acid was purchased from Central Drug House Ltd., New Delhi. Anisaldehyde (4-methoxy benzaldehyde) was procured from Avra Synthesis Pvt. Ltd., Hyderabad.

extract Preparation
The plant materials were shed dried at room temperature and were processed properly into powder that was allowed to pass

creation of Drug-resistant Osteosarcoma cells
Human osteosarcoma cell line resistant to cisplatin (HCR) was created following similar methods described in Palvai et al. (13). Briefly, HOS cells were exposed to repeated doses of very high concentration of cisplatin (1 mg/ml); surviving cells were selected and further sub-cultured. HCR cells were periodically checked for their resistance property by MTT assay and were maintained at 37°C in 5% CO2 under inhibitory concentration (IC50) cisplatin stress.

In Vitro cytotoxicity assay
In vitro cytotoxicity was performed as described previously by Chowdhury et al. (12)

estimation of intracellular reactive Oxygen species (rOs)
Reactive oxygen species levels were measured using 2, 7dichlorofluorescein diacetate (DCFH-DA) (Sigma), which measures intracellular generation of hydrogen peroxide, a procedure widely used for estimation of ROS. The DCFH-DA passively enters the cell, where it reacts with ROS to form the highly fluorescent compound dichlorofluorescein (DCF). Approximately, 0.3 × 10 4 cells per well were seeded and treated with plant extracts for 24 h; the ROS scavenger N-acetyl cysteine (NAC, 5 mM) was added 2 h prior treatment wherever mentioned to inhibit ROS. Following exposure, cells were washed with PBS and then incubated in 100 µl of working solution of DCFH-DA (2 mM DCFH-DA stock solution was diluted to yield a 20 µM working solution) at 37°C for 30 min. The fluorescence was measured at 485 nm excitation and 530 nm emission using a microplate reader (Fluoroskan Ascent).

Annexin V/Propidium Iodide (PI) Staining
For determination of apoptosis, HOS cells were seeded in 6 cm dishes at a density of 1 × 10 6 cells/dish. The following day, the cells were treated with specific plant extracts and incubated for 48 h. Thereafter, the cells were harvested, washed with PBS, and re-suspended in 500 µl of 1X binding buffer (BD BioSciences). To detect both early and late apoptotic cells, 4 µl of AnnexinV and 10 µl of PI were added to the cells in binding buffer, followed by incubation in dark for 20 min (12). The samples were then acquired using flow cytometer (Cytoflex, Beckmann Coulter) and analysis of acquired data was performed using CytExpert software. Percentage of apoptotic cells is represented through bar diagram.

Measurement of Caspase-3 Activity
Approximately, 1 × 10 4 cells/well were seeded and exposed to plant extracts for 48 h. The activity of caspase-3 was measured using caspase-3 colorimetric protease assay kit (Invitrogen) following manufacturer's protocol (12). Briefly, the cell lysate was collected in RIPA buffer and the concentration of protein was determined using Bradford assay. Equal amount (60 μg) of protein was added to microtiter plates with caspase-3 substrate (acetyl-Asp-Glu-Val-Asp p-nitroanilide, Ac-DEVD-pNA) and absorbance was read at 405 nm using a microplate reader (Startfax 2100, Awareness Tech. Ltd.). The colorimetric assay is based on the hydrolysis of caspase-3 substrate by caspase-3 enzyme, resulting in the release of the p-nitroaniline (pNA) moiety. The concentration of the pNA released from the substrate was calculated from the absorbance values at 405 nm.

Analysis of DNA Fragmentation
A distinctive biochemical feature of apoptosis induction is the fragmentation of genomic DNA by specific nucleases like, caspase-activated DNAse (CAD). Activation of CAD by the caspase cascade in turn leads to cleavage of cellular DNA at internucleosomal linker sites, generating short fragments of ~200 base pairs known as DNA ladders. One of the classical methods used to detect DNA ladders is to examine fragmented genomic DNA on an agarose gel. To analyze DNA ladder formation, HOS cells were seeded in 60 mm dishes at a density of 2 × 10 6 cells/plate and treated with BAM-SX or AIM-SO for 48 h. Genomic DNA was extracted using Invitrogen Apoptotic DNA Ladder Detection Kit and ladder formation was analyzed in 1% agarose gel. A DNA marker was run parallel to the samples.

DAPI Staining of Nucleus
For detection of apoptotic nuclei by 4′-6-diamidino-2-phenylindole (DAPI) staining, cells were either untreated or treated with plant extracts for 48 h. Cells were grown on coverslips in 6 cm petridishes till confluency. Coverslips were then withdrawn, washed with 0.1 M PBS, and cells were fixed in methanol for 10 min at room temperature. Cells were then stained with DAPI (Sigma-Aldrich, USA) (1 mg/ml in PBS) for 20 min, washed twice with PBS, and the coverslips mounted on a glass slide. Nuclear morphology was then observed by fluorescence microscopy (Olympus, Japan). Apoptotic nuclei can be identified by the fragmented and or condensed chromatin of nuclear bodies, marked with white arrows.

Flow cytometric analysis of cell cycle Phase Distribution
For DNA content analysis, HOS cells were seeded in 6 cm dishes at a density of 2 × 10 6 cells/plate and grown overnight. After 48 h of treatment, the cells were harvested, washed with PBS, and centrifuged at 2,000 rpm for 5 min at 4°C. The pellet was then re-suspended in 100 µl of PBS and 900 µl of ice cold 70% ethanol, used as a fixative. The fixed cells were incubated at 4°C overnight. The next day, cells were centrifuged and the pellet was re-suspended in 450 µl PBS and 10 µl of PI (2 mg/ml) containing solution (12). The samples were then incubated in dark for 10 min followed by event acquisition using flow cytometer (Cytoflex, Beckmann Coulter) and analysis using CytExpert software. Percentage of cells in each phase of cell cycle was plotted and represented through bar diagram.

Monodansylcadaverine (MDC) Staining of Autophagic Vacuoles
The drug MDC, a specific autophago-lysosomal marker was used to analyze induction of autophagy after plant extract treatment (14,15). For visualization of the autophagic vacuoles by fluorescence microscopy, HOS cells were seeded on cover slips and grown overnight. Following treatment, the cells were incubated for 10 min at 37°C with 0.05 mM MDC dissolved in PBS. The cover slips containing the cells were then washed with PBS and mounted with anti-fade mountant (containing DAPI). Intracellular MDC in the form of punctate dots were visualized using fluorescence microscopy. For, fluorimetric measurement, cells after treatment were labeled with MDC for 10 min, washed with PBS, and collected in 10 mM Tris-HCl (pH 8) containing 0.1% Triton X-100. Intracellular MDC was measured by fluorescence photometry (excitation 380 nm and emission 525 nm) in a microplate reader (Fluoroskan Ascent). An increase in MDC fluorescence upon treatment was expressed as fold change with respect to control.

Stable Transfection of GFP-RFP-LC3 and Fluorescent Detection
Plasmid GFP-RFP-LC3 was kindly provided by Dr. Sovan Sarkar (Birmingham Fellow, University of Birmingham). H1299 cells were cultured in six well plates and transfected with 2 µg of GFP-RFP-LC3 purified plasmid with Lipofectamine 3000 (Invitrogen, USA). Around 24 h after transfection, the cells were selected for transfection positivity by geneticin (600 µg/ml, Gibco) selection and were maintained for several days under geneticin pressure. Successful stable transfection of the plasmid was confirmed by fluorescent detection of LC3 in surviving cells through fluorescence microscopy (Olympus, Japan). For verification of autophagy induction, cells were treated with plant extracts and intracellular green and red fluorescence representing LC3 induction and accumulation were monitored and quantitated through fluorescence microscopy. By combining an acid-sensitive GFP with an acid-insensitive RFP in the vector, the autophagic flux is better visualized by imaging decreased GFP fluorescence, with  a simultaneous increase in red fluorescence marking enhanced autophagic flux. The LysoTracker probe (Thermo Fisher Scientific, USA), which consists of a fluorophore linked to a weak base, is a fluorescent acidotropic probe, which is often used for labeling and tracking acidic organelles in live cells. When cells were 60-70% confluent, the medium from the dish was replaced with pre-warmed (37°C), probe-containing medium. Cells were then incubated for 20 min, and the loading solution was replaced with fresh medium. The cells were then observed under a fluorescence microscope and the intensity of lysoTracker fluorescence was compared with untreated control.

Bright Field images
For bright field microscopic imaging, cells were cultured at density of 2 × 10 4 cells/plate in 6 cm culture dishes, treated with phyto-extracts, and then images were captured using Olympus (CKX41) microscope at 20 × magnification.

Phytochemical analysis
The active extracts from each plant were selected for phytochemical analysis using CAMAG high performance thin layer chromatography (HPTLC) instruments equipped with winCAT software. Extracts were dried properly and re-dissolved in methanol as 1 mg/ml stock solution. TLC plates with 0.2 mm pre-coated silica gel 60F254 (Merck, Germany) of 5 × 10 cm were taken. The extract samples were spotted using Linomat 5 automated sample spotter (CAMAG) using 100 µl syringe (Hamilton, Switzerland). Sample solution (5 µl) was spotted as bands of 6 mm width with the syringe with 5 mm distance between each spot. The spotted plates were developed in glass twin trough chamber (10 cm × 10 cm, CAMAG). n-butanol: acetic acid: water (14:3:4 v/v/v) was used as mobile phase for BAM-SX (16,17), while benzene: ethyl acetate (3:7 v/v) (18) and hexane:ethyl acetate:acetic acid (6:4:0.2) were used as mobile phase for AIM-SO (18)(19)(20) to identify compounds like azadirachtin, nimbin, and nimbolide. The length of the chromatogram was 80 mm from application position. After development, the TLC plates were dried using hot plate and scanned densitometrically at 254 and 366 nm. The image of the developed TLC plate was captured using TLC Visualizer (CAMAG) under 254 and 366 nm. The developed plate of AIM-SO was derivatized using anisaldehyde reagent followed by heating the plate at 105°C for 5 min. The derivatized plate was scanned at 560 nm and image was captured under white light using visualizer. The following scan conditions were applied: slit width: 4 mm × 0.10 mm, distance from Y-position: 10 mm and distance from X-position: 10 mm. Samples were run in triplicate. Experimental conditions: temperature 25°C ± 2; relative humidity 50%.

statistical analysis
Tukey tests was used as a follow up to one-way or two-way ANOVA to compare every mean to a control mean and every   (Figure 2A). A possible reason justifying the increased cytotoxicity of BAM-SX could be due to more extractive value of BAM-SX (7.26 g) than BAM-SO (6.52 g) and BAM-MC (3.28 g). Soxhlation performed for 24 h is probably enough to extract out active constituents into the solvents. Furthermore, the underlying reason behind the comparatively less cytotoxicity of BAM-SO could be associated with ultrasound energy mediated deleterious effects on the active compounds. However, as the extractive value of BAM-SX and BAM-SO is more or less similar, hence, there is not much difference in the bioactivity of the extract. However, in contrary, AIM-SO showed enhanced cytotoxicity (IC50 value of 28.65 µg/ ml) in HOS cells than AIM-SX or MC ( Figure 2B). Probably, the amount of active constituents was more in AIM-SO compared to AIM-SX and AIM-MC. It is speculated that ultrasound waves can help the solvent to penetrate into the cell wall of the plant material extracting out more compounds for cytotoxicity. Since BAM-SX and AIM-SO showed maximum cytotoxicity, for further studies, BAM-SX and AIM-SO were selected.  metastasis (21,22). However, conversely, increasing the level of ROS has often been utilized as a strategy to tip the balance of cancer cells toward cell death (21). Under the above circumstances, the generation of enhanced ROS can play a pivotal role in the initiation of apoptosis. Existing literature hints toward the prospective potential of phytochemicals in altering cellular ROS levels and induction of anticancerous activity (1, 2). However, the potential of natural extracts of BAM and AIM for ROS generation in osteosarcoma has not been evaluated before. A more than 2.5-fold accumulation of ROS was observed in cells treated with AIM-SO compared to untreated control after 24 h of treatment, as measured fluorometrically through H2DCF-DA assay, which acts as a marker for ROS ( Figure 3A). However, BAM-SX failed to induce a significant generation of ROS at the time point studied ( Figure 3B). ROS generation however, can be dependent on and resultant of the cell type studied, and the extraction procedure involved. To further understand the role of ROS, we inhibited intracellular ROS with NAC. Pretreatment of cells with the ROS scavenger NAC significantly increased cell viability in AIM-SO treated cells, but failed to show a significant effect in BAM-SX treated cells (Figures 3C,D). From the above experiments, it can be concluded that AIM-SO induces ROS-dependent cytotoxicity, while, BAM-SX stimulates ROS-independent cell death in the HOS cells studied.

BaM-sX and aiM-sO induce sub-g 1 Population of cells, caspase activity, and apoptotic cell Death in hOs cells
Phytochemicals are known to trigger apoptosis to inhibit cancer progression and such apoptotic cell death signaling is often elicited post ROS accumulation in cancer cells (1,2). Previous studies have reported that nimbolide or berberine can induce apoptotic cascade via ROS production in various cancer cell types like colon and human choriocarcinoma cells (23)(24)(25)(26). However, the potential of BAM and AIM in induction of apoptosis in osteosarcoma cells, a highly malignant cancer type with inherent resistant properties has not been elucidated yet. Phase contrast images of cells treated with the phytoextracts showed distinct rounding up of cells posttreatment, indicative of apoptotic cells (Figures 4A,B). We were, therefore, interested to analyze induction of apoptosis through other methods. Apoptosis induction in  HOS cells, posttreatment, was measured through flow cytometry using AnnexinV-FITC-PI. A significant increase in apoptotic cells was observed in cells independently treated with BAM-SX and AIM-SO ( Figure 4C). As the extracts showed significant growth inhibitory effect on HOS cells, we determined the possible inhibitory effect of BAM-SX and AIM-SO on the cell cycle progression of HOS cells. As summarized in Figure 5A, treatment of HOS cells with BAM-SX for 48 h resulted in a significantly higher number of cells in the sub-G1 phase (12%) of the cell cycle compared to non-BAM-SX-treated control. There was a concomitant reduction in the number of cells in G2-M phase of the cell cycle. Similar, but, significantly more pronounced (24%) results were obtained in cells treated with AIM-SO ( Figure 5A). These data suggest that inhibition of cell proliferation or induction of cell death in HOS cells by BAM-SX and AIM-SO is associated with the disarray of cell cycle profile of HOS cells. Stimulation of apoptosis is known to release cytochrome c from mitochondria and lead to the activation of apaf-1 (apoptosome), which in turn cleaves the pro-enzyme caspase-9 into its active form; activated caspase-9 then activates executioner caspases, like caspase-3 ultimately leading to cell death. We evaluated caspase-3 activity through ELISA in the extract treated cells (Figure 5B). In continuation to the results obtained above, BAM-SX and AIM-SO were found to induce caspase-3-dependent apoptosis in HOS cells. Also, DNA fragmentation is often considered as a hallmark of apoptosis. As evident in Figure 5C, a smear indicative of fragmented DNA post apoptosis was observed in cells treated with the plant extracts. Furthermore, analysis of nuclear morphology by DAPI staining post extract treatment revealed the presence of condensed and fragmented nucleus in plant extract treated cells indicative of apoptosis induction (Figure 5D). Our data indicate that treatment of HOS cells with the plant extracts results in G1 arrest and associated induction of apoptosis.
autophagy acts as Protective strategy against the cytotoxic effects of BaM-sX and aiM-sO in hOs cells  (Figures 6A,B). A LysoTracker probe was also used to label and detect acidic organelles, which mainly consist of lysosomes. In our study, we observed increased green puntate dots with stronger fluorescence intensity in cells treated with the natural extracts indicating an enhanced lysosomal number and probable function in the treated cells ( Figure 6C). H1299 Cells were further stably transfected with GFP-RFP-LC3 plasmids, and we observed that posttreatment, there was a significant increase in red fluorescent punctate dots indicative of enhanced autophagy after treatment ( Figure 6D). In corroboration to above, interestingly, autophagy inhibitor chloroquine di-phosphate (CQDP) induced enhanced cell death in the phyto-extract treated cells. Cells were treated with CQDP, 2 h prior to treatment with BAM-SX and AIM-SO and cell viability was estimated using MTT assay (Figures 6E,F). Our results are indicative of the fact that autophagy is triggered post BAM-SX or AIM-SO treatment in HOS cells, which facilitate the cells to overcome the cytotoxic effect; hence autophagy here operates as a cytoprotective phenomenon. and recycling of proteins and organelles through fusion with lysosomes. The role of autophagy in cancer has been controversially discussed (27). It has been known to both support or suppress tumorigenesis more in a context-dependent manner (28). In this regard, berberine has been previously demonstrated to induce autophagy and mitochondrial apoptosis in human hepatic carcinoma cells (29); here, autophagy provided a prodeath signal, as inhibition of autophagy with 3-methyl adenine (3-MA) reduced cell death. Furthermore, Wang et al. showed that berberine can lead to high autophagic flux resulting in reduction of invasive properties and proliferative potential of glioblastoma cells (30). However, in contrary, Subramani et al. reported that nimbolide can potentiate autophagy in pancreatic cancer cells as an adaptive stress response to extend cell viability rather than to initiate apoptosis/cell death (31). Thus, autophagy can act either as a cyto-protective or as a cytotoxic phenomenon, and hence it is important to understand its role in each context independently. In our study, we observed that on treatment with BAM-SX and AIM-SO, HOS cells upregulate autophagy. Autophagy was estimated by quantitation of MDC fluorescence both microscopically and fluorimetrically (Figures 6A,B).  Another advantage of this method is that it provides complement to the algebraic analysis with an accepted graphical approach (32). The combined effect of the two cytotoxic extracts in the ratio of 1:1 was observed for 48 h in pairs of sonicated extracts, i.e., BAM-SO and AIM-SO (BAAISO) and macerated extracts, i.e., BAM-MC and AIM-MC (BAAIMC) with IC50 of 10.36 and 11.27 µg/ml, respectively (Figures 7A,C). Negligible cytotoxity was observed even after 48 h when the combination with pair of soxhlated extracts, i.e., BAM-SX and AIM-SX (BAAISX) in 1:1 ratio was performed ( Figure 7B).  (Figures 8A-F). Effect-based strategy was applied using software Compusyn version 1. The results showed that BAAISO showed dosedependent synergism as increase in concentration of BAM-SO and AIM-SO in the ratio of 1:1 (Figure 9A). On the other hand, BAAIMC showed synergism in all the doses of BAM-MC and AIM-MC in 1:1 ratio ( Figure 9B). BAAISX, combination of BAM-SX and AIM-SX showed antagonistic effect in the same ratio ( Figure 9C). Inter technique extract combination showed synergism (Figure 10B-F BSXASO, which showed antagonism as CI is more than 1 in all doses of BAM-SX and AIM-SO in 1:1 ratio (Figure 10A) ( Table 1). As BAAISO was found to be most potent against HOS cell lines compared to the other combinations, it was further tested on HCR cell lines.

BaM-sX and aiM-sO sensitizes resistant hOs cells
Human osteosarcoma is the most prevalent malignant bone tumor with a 5-year survival rate of as low as 20%. Even with an aggressive treatment regimen, disease recurrence is almost always invariable in osteosarcoma. Thus, therapy resistance remains a major concern, which needs to be promptly addressed (33,34). We, therefore, derived a cisplatin resistant in vitro model from the parental HOS cells following similar procedure described in Palvai et al. (13) and tested for the cytotoxic efficacy of BAM-SX and AIM-SO in the resistant cells. While creating the resistance model, the HOS cells were exposed to high dose of cisplatin followed by clonal selection of surviving cells post shock. This process was repeated several times to derive HOS cells that showed decreased sensitivity to cisplatin. These cells were termed as HCR, representing cells resistant to cisplatin. The concentration of cisplatin at which 50% of HOS cells died was 35 µM, while an increased viability (80%) was obtained in HCR cells at similar concentration of cisplatin, as estimated through MTT assay. To maintain clinical relevance, a resistance model showing more than twofold resistance to cisplatin was not created for this study. BAM-SX and AIM-SO were able to sensitize the resistant HCR cells with an IC50 value of 80.87 and 43.93 µg/ml, respectively, obtained only after 24 h of treatment (Figures 11A,B). As evident from the values, AIM-SO was found to be significantly more cytotoxic than BAM-SX. However, the combination of BAM-SO and AIM-SO (1:1) showed significantly increased cytotoxicity with an IC50 value of 17.34 µg/ml ( Figure 11C). The combinatorial treatment of BAM-SX and AIM-SO failed to impart enhanced cytotoxicity compared to their treatment when administered independently. Our results clearly indicate that the extracts tested in this study are potent enough to effectively sensitize cisplatin resistant cells.

Phytochemical analysis
Since BAM-SX and AIM-SO were found to induce apoptosis in HOS cells, hence, a HPTLC-based phytochemical analysis was carried out in order to understand the phytochemical profile of these extracts. There are various HPTLC methods reported for analysis of B. aristata and A. indica. However, as variation is found to occur in botanical samples due to different geographic locations and climatic conditions, identification of compounds is necessary. Reported composition of mobile phase (n-butanol:acetic acid:water; 14:3:4v/v/v) was used for BAM-SX, which showed berberine as the major components (Rf 0.55) followed by palmatine (Rf 0.50) at 254 and 366 nm (Figure 12). Berberine has been reported to affect activities of enzymes such as N-acetyltransferase, cyclooxygenase-2, and topoisomerase, by modifying gene/protein expression. These actions along with the ROS production, mitochondrial transmembrane potential, and NF-κβ activation have been found to potentiate antiproliferative and proapoptotic effects of B. aristata (24,29,35). Palmatine has also been shown to exhibit cytotoxicity in mice by selectively targeting DNA topoisomerase I (36,37).
In case of AIM-SO, two reported solvent systems were used. The first solvent system was used to identify compounds like azadirchtin A (Rf 0.36), azadirachtin B (Rf 0.56), and nimbin (Rf Frontiers in Oncology | www.frontiersin.org December 2017 | Volume 7 | Article 296    (Figures 13 and 14). Nimbolide has shown enhanced cytotoxic and apoptotic inducing potential. It does this by increasing the generation of ROS and downregulation of proliferating cell-nuclear antigen leading to apoptosis (23,31,38). Azadirachtin on the other hand has been found to decreases the over expression of cyclin-E, leading to the shortening of G1 phase of cell cycle and reducing the production of cyclin-A, a protein responsible for cell division, which leads to apoptosis (8,39).

cOnclusiOn
In spite of enormous progress in the last decade, in the field of cancer therapeutics, tumor relapse and drug refractoriness has been highly prevalent and has not yet been addressed effectively (4). Natural products and herbal therapies have demonstrated promise and prospect in prevention and treatment of various cancers, but, their true potential as remedies toward aggressive cancers has been underexplored (1,2). Hence, though the usefulness of plant extracts in attenuating cancer progression has been authenticated by the increasing number of commercially available anticancer drugs detailed studies illustrating their effectiveness against inherently resistant cancer cells has been elusive. In this study, we have evaluated the anticancerous efficacy of BAM and AIM against both drug sensitive and resistant osteosarcoma (HOS) cell types. Cells were treated with plant extracts derived through various extraction procedures to compare their potential effectiveness against the above cell types. We report that BAM extracts prepared by soxhalation technique and AIM extracts prepared by sonication were the most effective in inducing apoptosis in HOS cells. The HOS cells are reported to be inherently resistant and hence cytotoxicity of the herbal compounds has enormous significance from clinical perspective. Additionally, the extracts were able to sensitize the resistant cells as well, which were derived from the parental HOS cells. Interestingly, the exposure of independent extracts of BAM-SX and AIM-SO showed enhanced cytotoxicity in the resistant cells, while a combination of both failed to impart significant effect. We assume that the combination has some mutually inhibitory effect in terms of their cytotoxic effect. We, thereafter, explored the mode of cell death in the parental cells and observed that they induce caspase-mediated apoptosis initiated primarily by alteration of intracellular redox levels. Interestingly, a prior inhibition of autophagy, enhanced cell death, suggesting a protective role of autophagy. The role of autophagy in cancer is controversially discussed, where it can act both as a tumor promoter or a stimulator of apoptotic cell death (27). Of late, a good number of studies have highlighted the role of autophagy in survival of cancer cells under cellular stress (27,28). We analyzed the conventionally studied autophagic marker-LC3-BII and observed an accumulation of the same, indicative of enhanced autophagy upon exposure to extract; and an increase in autophagy served as a survival strategy adopted by the cells under the current scenario. Furthermore, HPTLC analysis resulted in identification of various phytochemicals such as berberine, palmitine, azadirachtin, nimbin, and nimbolide, which have been reported for anticancer properties and these compounds might be responsible for the observed activity of BAM-SX and AIM-SO. Our results suggest a high therapeutic potential of BAM-SX and AIM-SO against both sensitive and drug refractory HOS cancer cells and provides a mechanistic insight into their mode of action.