All the cell lines used in this study (MCF7, MDA-MB-231, MDA-MB-468, DF2 and WI-38) were purchased from ATCC. Cells were grown in DMEM media supplemented with 10% fetal bovine serum, 100 μg/ml gentamycin, and 2 mM l-glutamine. To grow MCF7 cells the medium was also supplemented with 10 μg/ml insulin (NM Penfild, Denmark). Cells were grown at 37°C in 5% CO₂ atmosphere.

Ecdysterone (95% purity, Frog Tech, Russia) was dissolved in DMSO. Thus, DMSO was used as a control for all experiments with ecdysterone (0 μM Ecdy). Doxorubicin (98% purity, Sigma, United States) and 2-DG (98% purity, Sigma, United States) were dissolved in water.

For MTT experiments, 10,000 cells were planted overnight in each well of a 96-well plate. 10 wells per sample were used. A day after, ecdysterone or (and) doxorubicin were added in the required concentrations for 48 h. For cells treated with ecdysterone, DMSO was used as a control. Then 10 μL of 5 mg/ml Thiazolyl Blue (Paneko, Russia) solution was added to each well and cells were kept for 3,5 h at 37°С in CO₂ incubator. After removing the thiazol-containing medium, 150 μL isopropyl alcohol (supplemented with 40 mM HCl and 0,1% NP-40) was added to dissolve the MTT-formazan salt. The absorbance at 570 and 630 nm (reference) was measured using BioRad iMark microplate reader (BioRad, United States). Results are represented as the mean ± SD.

To perform colony-formation assay, 1,000 cells were planted per well on a 6-well plate, in triplicates. 24 h later, the cells were treated with 0, 100, 150, 250, 350 or 500 μM ecdysterone for 96 h. After treatment, fresh media was added, and cells were grown for 10 days. After the indicated time, cells were fixed with acetic acid/methanol (1:7, v/v) and stained with 0.5% crystal violet. The number of colonies was calculated. Results are represented as the mean ± SD of three biological replicates.

About 25,000 cells were seeded in 12-well plates and incubated with different amounts of 0–750 µM Ecdy for 4 days. Following the incubation, cells were trypsinized, stained with trypan blue and calculated using hemocytometer. Six replicates were used for analysis. Results are calculated as the mean ± SD; *p < 0.05.

Flow cytometry analysis of cell cycle distribution was done essentially as described previously (Lezina et al., 2014). A day after seeding, cells were treated with ecdysterone (0, 250, 500, or 750 μM) for 48 h in triplicates. After harvesting, cells were washed once with PBS, and fixed in 70% ethanol at −20°C for 1 h. The 30 min staining of DNA content was carried out by using 50 μg/ml of PI (Invitrogen, United States) and 1 μg/ml RNase A (ThermoFischer). Samples were analyzed by CytoFLEX (Beckman Coulter, United States) flow cytometer. Results were processed by CyteExpert software (Beckman Coulter, United States).

To perform the energy profiling using SeaHorse apparatus, 30,000 cells were seeded to each well (except for the background wells) of a 24-well SeaHorse plate (Agilent, United States) overnight. Four wells were used per sample. Then, 0–1,000 μM Ecdy was added for 48 h. 12 h before analysis, the Sensor Cartridge was equilibrated in XF Calibrant (Agilent, United States) at 37°C in a non-CO₂ incubator. SeaHorse XF Energy Phenotype kit (Agilent, United States) was applied for assay. SeaHorse XF base medium was supplemented with 1 mM pyruvate, 2 mM glutamine and 10 mM glucose, pH 7.4. Stressor mix consisting of olygomycin and FCCP (both Agilent, United States) was used to achieve final concentrations 1 and 2 μM in wells, respectively. Assay was run in the XFe24 Analyser device (Agilent, United States) in accordance with the manufacturer’s instructions. Data were normalized using total protein quantification by BCA assay (ThermoFischer, United States) and processed by SeaHorse XF Cell Energy Phenotype Test Report Generator (Agilent, United States). Results are represented as the mean ± SEM.

Flow cytometric determination of cell death including apoptosis was carried out by using annexin V-FITC/(PI or 7-AAD) double staining. To analyze the influence of ecdysterone on cell death, annexin V–FITC/PI kit (BD Biosciences, United States) was used, whereas in studies of combined treatments (doxorubicin and ecdysterone) annexin V–FITC/7AAD (ThermoFischer, United States) was applied in accordance with the corresponding manufacturer’s protocols. Cells were treated for 48 h with ecdysterone (0, 250, 500, and 750 μM) and doxorubicin (0, 0.15, and 0.25 μM) separately or in combination. A minimum of 5,000 cells were analyzed by CytoFLEX (Beckman Coulter, United States) flow cytometer using corresponding channels in three independent experiments. Values of the median were used for calculation. Results were represented as the mean ± SEM of three experiments.

A day after seeding, cells were treated by ecdysterone (0, 250, 500 or 750 µM) for 48 h in triplicates. Before analysis, cells were treated with 75 nM LysoTracker Red DND-99 (ThermoFischer, United States) for 2 h at 37°С in a CO₂ incubator. Then cells were washed in PBS, detached with trypsin and analyzed by flow cytometry (CytoFlex, Beckman Coulter, United States). Values of the median were used for calculation. Results were represented as the mean ± SEM of three experiments.

Autophagic flux was revealed by blocking autophagy using chloroquine followed by western-blot with anti-LC3 and p62 antibodies, as well as immunofluorescence (staining with anti-LC3 antibodies). The next day after being planted in Petri dishes or glass cover slips, cells were treated with 0–1,500 µM Ecdy for 32 h followed by a co-treatment with the same concentrations of Ecdy and 50 µM chloroquine to block autophagy for 16 h. Then cells were subjected to either western-blot or immunofluorescence.

Cell lysates were prepared using RIPA buffer (150 mM NaCl; 50 mM Tris–HCl pH 8.0; 0.5%NP-40; 1 mM PMSF, protease inhibitor cocktail). After protein quantification by BCA assay (ThermoFischer, United States), 20 ug of Laemli-diluted cell lysates were loaded on 17% SDS-PAGE, run on TRIS-Glycine running buffer, followed by transfer to a nitrocellulose membrane (Bio-RAD, United States). Following 1 h blocking in PBST-diluted 5% nonfat milk, membranes were incubated with primary antibodies: LC3B (1:1,000, #2775S, Сell Signaling, United States), p62 (1:1,000, #5114, Сell Signaling, United States) or β-actin (1:5000, A3854, Sigma-Aldrich, United States). After PBST washing, secondary anti-mouse or anti-rabbit antibodies (1:10,000; Sigma-Aldrich, United States) conjugated with horseradish peroxidase were used. ECL system (ThermoScientific, United States) and ChemiDoc Touch Imager (Bio-Rad, United States) were applied for detection.

Cells grown on glass cover slips were fixed with 4% PFA in PBS for 15 min and then washed three times in PBS, followed by 60 min incubation in permeabilization blocking solution (5% BSA, 0.3% Triton X-100 in PBS) at room temperature. Then, cells were stained with anti-LC3B antibodies (1:200, #2775S, Сell Signaling, United States) in permeabilization blocking solution for 16 h at 4°C, washed three times in PBS, incubated with the secondary antibody in permeabilization blocking solution (goat anti-rabbit, AlexaFluor488, Invitrogen, United States) for 1 h at room temperature and washed three times in PBS. Slides were mounted using ProLong Gold Antifate Mountant with DAPI (P36931, Invitrogen). Images were analyzed by confocal microscope (Olympus, FV3000, Germany).

IC50 and drug synergy were obtained using results of MTT-assay. IC50 was calculated using AAT Bioquest IC50 online calculator https://www.aatbio.com/tools/ic50-calculator. Drug interaction was assessed by Chou-Talalay algorithms (CompuSyn software, http://www.combosyn.com/; (Chou and Talalay, 1984; Сhou, 2006). Results were represented as CI (Combination Index) plots and a Table which includes values for CI and DRI (Dose Reduction Index). CI < 1 attests synergistic action of drugs; DRI estimates the extent to which the dose of one or more agents in the combination can be reduced to achieve effect levels that are comparable with those achieved with single agents.

All results are demonstrated as mean ± standard deviation (SD) or standard error of the mean (SEM) of at least three biological replicates. Statistical significance was analyzed using the Student’s t-test: *p < 0.05; **p < 0.01; n.s – not significant.

Several papers describe the ecdysterone-mediated tumor suppressive effect on some cancer cells (Konovalova et al., 2002; Martins et al., 2015). We evaluated the effect of this drug on three human breast cancer cell lines with different molecular properties: MCF7 (luminal, ER⁺Pr⁺Her2⁻), MDA-MB-231 (TNBC) and MDA-MB-468 (TNBC).

As ecdysterone is reported to exert anabolic activity in muscle tissue, which should facilitate the proliferation, we decided to carry out MTT assay to see if ecdysterone affects the proliferation of cancer cell lines studied. To do this, we have used a broad range of concentrations ranging from one to 3,000 µM. Results shown in Figures 1A–C clearly demonstrate that in our case the treatment with Ecdy has down-regulated all three cell lines starting with a concentration of 250–750 µM. No increase in cell proliferation was detected.

We have also carried out colony-formation assay. As this analysis implies the growth of colonies from single cells and they are very sensitive to any treatment, we have chosen a lower concentration of Ecdy. Results shown in Figures 1D–F confirm that the treatment with Ecdy inhibits the growth of all three cell lines. Moreover, photographs of plates with colonies clearly showed that not only the number of colonies, but also their size, was significantly reduced upon treatment with Ecdy.

The analysis of the cell cycle has shown that ecdysterone affected to varying degrees the cell cycle distributions of MCF7 and MDA-MB-231 cells, but had no effect on MDA-MB-468 cells (Figures 2A–D). It significantly increased the number of MCF7 cells in the G1-phase (Figure 2B). To a lesser extent, the same was true for MDA-MB-231 cells (Figure 2C). However, MDA-MB-468 cells were insensitive to Ecdy-induced alterations in the cell cycle (Figure 2D).

We also determined if ecdysterone affected the level of cell death. Annexin V/PI staining followed by flow cytometry (Figure 2E) demonstrated that ecdysterone increased the rate of cell death up to 1.6 fold for both MCF7 and MDA-MB-231 cells (Figures 2F,G), and 3.5 times in the case of MDA-MB-468 cells (Figure 2H). It is important to notice that although Ecdy elevated overall cell death in all cell lines, it did not increase the population of Annexin V-positive cells, suggesting that Ecdy elicited death of MCF7 cells via mechanism(s) other than apoptosis (Figure 2E). In contrast, Ecdy increased the population of Annexin V-positive (apoptotic) cells by 30% for MDA-MB-231 and by 48% for MDA-MB-468 cell line (Figure 2E).

These data suggest that the ability of ecdysterone to induce cell cycle arrest or elicit cell death of breast cancer cells presumably depends on the genetic background of a particular cell line.

As ecdysterone possesses anabolic properties in muscle tissue, we were interested to see whether it alters the energy metabolism of the breast cancer cell lines. To do this, we employed the SeaHorse energy profiling using Energy Phenotype kit.

Energy profiling of MDA-MB-231 cells treated with 0, 250, 500, 750 or 1,000 μM of ecdysterone has shown ecdysterone-mediated inhibition of respiration (basic OCR). For instance, 500 and 750 μM Ecdy decreased respiration by approximately 21 and 31%, respectively (Figure 3A). At the same time, ecdysterone did not alter glycolysis (basic ECAR, Figure 3B) but significantly decreased the metabolic potential (both stressed respiration (stressed OCR) and glycolysis (stressed ECAR), Figures 3A,B). Thus, 500 μM Ecdy mitigated stressed OCR by 23% and stressed ECAR by 18%. Stressed OCR and ECAR reflect the metabolic potential of cells–percentage increase of stressed OCR over the baseline OCR, and stressed ECAR over the baseline ECAR. Metabolic Potential (MP) is the measure of cells’ ability to meet the energy demand via respiration and glycolysis. Thereby, these results suggest that ecdysterone greatly reduced the capacity of MDA-MB-231 cells for metabolic adaptation.

In the case of MDA-MB-468 cell line, the same treatment with ecdysterone has led to an inhibition of both the baseline respiration and of glycolysis (Figures 3C,D). Even 250 μM Ecdy dampened respiration and glycolysis by 22%, whereas 750 μM Ecdy inhibited them further by 47 and 28%, respectively. Taken together, these data suggest Ecdy was able to attenuate the level of energy metabolism in TNBC breast cancer cells.

Tang and colleagues (Tang et al., 2018) has reported that ecdysterone promotes autophagy in osteoporotic rats. We have carried out staining in three ecdysterone-treated (0, 250, 500, and 750 μM, 48 h) breast cancer cell lines with LysoTracker, a fluorescent dye, reflecting the number of lysosomes which can be indicative of autophagy (Chikte et al., 2014). Flow cytometry analysis has shown the strong increase of LysoTracker fluorescence in all three cell lines in order of increase the ecdysterone concentration (Figures 4A–C).

To further study whether ecdysterone affects the autophagic flux the breast cancer cell lines were incubated with different concentrations of ecdysterone (0, 250, 500, and 750 μM) for 32 h followed by 16 h of incubation with both ecdysterone and 50 μM chloroquine to block the autophagic-mediated degradation. Treated cells were analyzed by western-blot with LC3 and p62 antibodies or by immunofluorescence microscopy to evaluate the LC3 staining. Results of both immunoblotting and immunofluorescence (Figures 4D–G; Supplementary Figure S1) demonstrate that in the case of all three lines, ecdysterone strongly induced autophagy.

Taken together, these data strongly suggest that ecdysterone induces autophagy in all breast cancer cells concomitantly with the increasing ecdysterone concentration.

Given that ecdysterone mediated the inhibition of growth of all breast cancer cell lines, we decided to examine its effect on cancer cells a part of the combined treatment with doxorubicin, a genotoxic drug which is widely applied as chemotherapeutics. To this end, we carried out MTT-assay using ecdysterone (250, 500, or 750 μM) and doxorubicin (0.15 or 0.25 μM) alone or in combination. To determine its synergistic effect we calculate the combination Index (CI) and dose reduction index (DRI) using Chou-Talalay algorithms (Chou and Talalay, 1984; Chou, 2006).

The results shown in Figures 5A–C demonstrate that ecdysterone significantly sensitizes all breast cancer cell lines to the treatment with doxorubicin. CI plots (Figures 5D–F) and Table 1 show that all three Ecdy concentrations have a pronounced synergistic (CI ranges 0.47–0.89) interaction with doxorubicin. In turn, the use of Ecdy allowed the reduction of the effective dose of doxorubicine from 1.4 to 17.9 times (Table 1), depending on the particular cell line. It is important to notice that in most cases the addition of even 250 μM ecdysterone was sufficient to down-regulate the growth of breast cancer cells 1.5–2 times more efficiently than the corresponding concentrations of doxorubicin (Figures 5A–C).

These data clearly demonstrate the ability of ecdysterone to synergize with doxorubicin to down-regulate the proliferation of breast cancer cells.

The observed synergistic effect of doxorubicin and ecdysterone likely results from cell death. To directly check this hypothesis, we treated the cell lines with ecdysterone (250 μM) and doxorubicin (0.15 or 0.25 μM) alone or in combination, followed by staining with Annexin V/7AAD and flow cytometry analysis. Since we have already shown that ecdysterone down-regulated the metabolic potential, we decided to apply 2-deoxyglucose (2-DG), a promising inhibitor of glycolysis, which currently undergoes clinical trials. We also treated cells with either 2-DG (10 mM) or ecdysterone (250 μM) alone or in combination.

Figures 6A–F and Supplementary Figures S2A–C demonstrate that the combined treatment (either doxorubicin or 2-DG with 250 μM Ecdy) both increased the level of cell death by several times relative to control, or in comparison to the treatment with an individual drug. Accordingly, co-treatment with doxorubicin and ecdysterone elevated the level of apoptosis in MCF7 cells by 23% and in MDA-MB-231 cells by 3.15 times, respectively, compared to doxorubicin alone.

Co-treatment of Ecdy with 2-DG also significantly enhanced both apoptosis and total cell death (Figures 6C–F). It increased the rate of cell death by 21% for MCF7 and 17% for MDA-MB-231 cells.

We also repeated the previously described treatment of several breast cancer cells with combinations of doxorubicin (0.25 or 0.15 μM, respectively) and ecdysterone (250, 500, and 750 μM) followed by immunofluorescence microscopy to detect the release of cytochrome C from the mitochondria upon apoptosis. Taken together, results shown in Figure 7 and Supplementary Figures S2, S3 confirm that treatment with Ecdy significantly enhanced the doxorubicin-induced release of cytochrome C in all cancer cell lines.

Taken together, these data suggest that ecdysterone mediates strong synergy with doxorubicin in attenuation of the proliferation of breast cancer cells.

To compare the effects of ecdysterone on cancerous vs. non-cancerous cells, we decided to test normal non-transformed human fibroblasts, DF2 and WI-38 cell lines after the treatment with Ecdy.

First, we assessed the effect of Ecdy on proliferation of breast cancer cell lines and non-transformed human fibroblasts. As shown in Supplementary Figure S4A,B, Ecdy down-regulated the proliferation of cancer cell lines significantly more robustly compared to normal human fibroblasts at all concentrations tested.

To study whether ecdysterone elicits effects on autophagy in non-transformed fibroblasts similar to what we have observed in breast cancer cells, we employed the previously described flow cytometry analysis with LysoTracker DND-99 Red combined with western blotting for LC3 and p62. Results shown in Supplementary Figure S6 demonstrate that Ecdy induced autophagy in fibroblasts to a much lesser extent than in breast cancer cells (Figure 4).

We have also investigated the effect of combined treatment with Ecdy and doxorubicin on DF2 and WI38 fibroblasts (Supplementary Figure S7). To this end, we have employed previously described Annexin V-FITC/7-AAD staining followed by flow cytometry. Surprisingly, the minimal effective concentration of Ecdy (250 uM) in combination with doxorubicin that induced death of breast cancer cells, had almost no effect on DF2 and WI38 fibroblasts (Figure 6; Supplementary Figure S2).

Finally, we have carried out MTT assay on DF2 and WI38 normal human fibroblasts as well as mouse embryonic fibroblasts (MEFs) treated with either Ecdy alone or in combination with doxorubicin. Supplementary Figure S8 demonstrates that in contrast to breast cancer cell lines (Figure 5), an increased ecdysterone concentration up-regulated the survival of WI-38 and MEFs cells (Supplementary Figures S7B,C). In the case of DF2 cells, the combined treatment (doxorubicin + ecdysterone) displayed even a small protection of these cells from doxorubicin (Supplementary Figure S7A). Regarding WI-38 and MEFs (Figures 7B,C), the same combined treatment has incomparable low inhibitory effect in contrast to breast cancer cells (Figure 5).

Taken together, our experiments show that ecdysterone significantly down-regulates cancer cells with no or little effect on fibroblasts.