Indole-alkaloid–rich fraction of Ervatamia coronaria leaf extract regresses breast cancer by inducing apoptotic cell death

Abstract

Current anti-breast cancer therapies often cause severe toxicity to normal cells and promote chemoresistance, while their exorbitant costs limit their accessibility to many patients. This underscores the need for safer, effective, and affordable alternative treatment strategies. In this study, we evaluated the anti-breast cancer potential of an alkaloid-rich dichloromethane fraction of Ervatamia coronaria (DFE) leaf extract, identified by LC-MS to contain eight indole-alkaloids as its major constituents. Using 4T1 cell-induced tumour allografts in BALB/c mice, alongside cell-based assays including cell viability, scratch assay, immunoblotting, immunohistochemistry, and scanning electron microscopy, we investigated its mechanisms of action. DFE treatment induced cell cycle arrest in the sub-G1 phase, triggering apoptosis, with little effect on normal cells. Mechanistically, elevated mitochondrial ROS were identified as the primary driver of toxicity, as pre-treatment with NAC reversed DFE’s effects. Additionally, DFE downregulated AKT signalling in breast cancer cells. Importantly, DFE significantly reduced 4T1 tumour growth in vivo, both alone and in combination with doxorubicin, without exerting significant toxicity on healthy mice. These findings support further evaluation of DFE in clinical models.

Graphical abstract

1. Introduction

Breast cancer is a major global pub lic health concern, particularly among women. In 2022, there were an estimated 2.3 million new cases worldwide, and 670,000 deaths. Incidence rates have increased in many countries. Alarmingly, in developing countries, new cases and death rates are projected to rise by 38 % and 68 %, respectively, by 2050 [1]. GLOBOCAN 2022 identified breast cancer as the most common malignancy in India, with 192,020 reported cases and 98,337 deaths, ranking third globally [[2], [3]]. Data from the National Cancer Registry Programme, conducted by the Indian Council of Medical Research (ICMR), show an incidence rate as high as 30 cases per 100,000 people in urban areas, compared to 5 cases per 100,000 in rural areas [3].

Increased health awareness, along with advances in early screening, is reducing the incidence of disease and improving early detection leading to better survival. Usually, besides surgery and radiation therapy, systemic therapies or chemotherapy are adopted to treat breast cancer [4]. But cancer chemotherapy regimens currently in use often face limitations due to associated general cytotoxicity and chemoresistance. Additionally, current treatments impose an economic burden on patients and their families. These facts altogether underscore the need to search for relatively safer, more efficient, and affordable alternative treatment strategies [5]. Throughout history, natural products have played an essential role in the treatment of cancer. Natural products, such as Taxol [6], and Vinblastine [7], have saved the lives of many cancer patients. However, these drugs are not without limitations, partly due to their effects on healthy cells. Taxol, for example, is known to cause several toxic side effects, including neutropenia, thrombocytopenia, hypersensitivity reactions, peripheral neuropathy, and cardiac complications [8]. The toxic effects of vinblastine include leukopenia, diarrhoea, nausea, vomiting [9], and syndrome of inappropriate secretion of antidiuretic hormone [10]. Research has been ongoing to improve the efficacy of these drugs further [11]. Here, we have tested the anti-breast cancer efficacy of an alkaloid-rich dichloromethane fraction obtained from the methanolic extract of Ervatamia coronaria (Ec) leaves (this fraction is hereafter referred to as DFE). Ec, a member of the Apocynaceae family, is found in many tropical Asian countries, including India, China, Japan, and Myanmar, as well as Australia [12]. This plant is widely used in traditional and folk medicine due to its effectiveness in treating inflammation and wounds [13]. Ethanolic and methanolic extracts of different parts of this plant exhibited anti-cancer, gastroprotective, and anti-inflammatory activities [[14], [15], [16], [17]]. Many alkaloids found in Ec, such as voacamine, voaphylline, voacristine, voacangine, ibogamine, and tabernamine, have been implicated in analgesic, antitumor, and anti-inflammatory activities, highlighting their therapeutic relevance in folk medicine [18]. However, the molecular bases underlying these activities remain poorly understood. A recent study demonstrated that an alkaloid-rich fraction (AFE) from Ec suppresses colorectal cancer by modulating cellular AMPK–mTOR signaling pathway [19]. The current study analyzes the anti-breast cancer properties of DFE both in vivo and in vitro, revealing a unique mode of action. This study demonstrated that DFE-induced elevation of cellular ROS levels is the primary cause of its toxicity to breast cancer cells. Results suggest that the DFEtreatment induces elevated levels of ROS in breast cancer cells, which triggers apoptotic death through mitochondrial outer membrane permeabilization (MOMP) and the release of cytochrome c. In contrast, immortalised normal cells resisted DFE activity, showing no elevation of intracellular ROS. In line with these results, DFE efficiently regresses 4T1-induced tumour allografts by triggering apoptotic death in mice with little effect on the animals.

2. Results

2.1. The dichloromethane fraction of Ervatamia coronaria (DFE) is toxic to breast cancer cells

2.1.1. LC-MS analysis of DFE

Earlier studies demonstrated anti-colorectal cancer activities in an alkaloid-rich fraction of Ervatamia coronaria [19]. LC-MS analysis of DFE, which we tested here for its anti-breast cancer activity, identified 17 different molecules consisting of eight different indole-alkaloids as major compounds such as stemmadenine, voafinidine, voaphylline, isovoacristine, vallesamine, pericyclivine, o-acetyl-vallesamine, and 19,20-dihydrotabernamine (Fig. 1A-B).

Fig. 1.

[A.] LC-MS profile of DFE [B.] Identified eight major compounds of DFE obtained from LC-MS analysis are indicated by arrowheads. [C.] Table representing a list of all (17) compounds identified in DFE by LC-MS analysis using Agilent PCDL mass hunter molecule library through the molecular feature extractor as reference.

2.1.2. Cytotoxicity analysis of DFE

The cell viability assay was used to test the sensitivity of breast cancer cells MCF7, MDA-MB-231, and NKE cells (normal cells) exposed to increasing concentrations of DFE for 24 h. The breast cancer cells exhibited increased sensitivity with increasing concentrations of DFE, whereas NKE cells showed little sensitivity up to 100 µg/ml. The IC₅₀ of DFE was 51.07 μg/ml for MCF7, 61.00 μg/ml for MDA-MB-231, and 230.8 μg/ml for NKE cells. Selectivity index in comparison with NKE cells for MDA-MB-231 and MCF7 is 3.78 and 4.52, respectively (Fig. 2A).

Fig. 2.

DFE kills both MCF7 and triple-negative MDA-MB-231 cells [A.] Bar graph representing cytotoxicity of DFE in MDA-MB-231, MCF7, and human kidney epithelial (NKE) cells, measured by MTT assay after treatment with increasing concentrations of DFE (0–200 µg/ml) for 24 h. Error bar, mean ± SD (n = 3); ns, not significant; ***, P = 0.0002, ****, P < 0.0001. IC₅₀ and selectivity index (SI) values of treated cells for 24 h as indicated (boxed). [B.] Representative phase-contrast image showing the effect of DFE on the migration potential of MDA-MB-231 and MCF7 cells obtained by wound healing assay. Both MCF7 and MDA-MB-231 cells were treated with DMSO/vehicle or DFE at the indicated concentrations for different durations after the scratches were made. [C.] Morphology of cells in the sensitivity assay, obtained by variable pressure scanning electron microscopy following the indicated treatments for 24 h. The lower panel shows a magnified view, as indicated. Scale bar and magnification for the upper panel are 25 µm and 1000X, respectively. Bar graphs (on the right-hand side, 1 K µm² =1000 µm²) showed changes in mean cellular area of MCF7, MDA-MB-231, and NKE cells, respectively, after treatment with DMSO/vehicle and DFE at different concentrations for 24 h, as indicated. 1, 2, and 3 represent the control, DFE 50 µg/ml, and DFE 100 µg/ml, respectively. Error bar, mean ± SD (n = 6 cells from 3 randomly chosen fields (2 cells from each field) (N = 3). Significance consideration according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001. Bar graphs (right-hand side) showed the changes in the number of mean rounded cells (n = 3) of MCF7, MDA-MB-231, and NKE cells, respectively, after treatment with DMSO/vehicle and DFE at different concentrations for 24 h, as indicated. 1, 2, and 3 represent the control, DFE 50 µg/ml, and DFE 100 µg/ml, respectively. MRC is the abbreviation for mean rounded cells/field. Significance consideration according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001. [D.] Super-resolution laser scanning confocal images show the formation of agglomerated stabilized actin fibres (Alexa Fluor 488 ^R tagged- actin poison phalloidin in green and nucleus was stained with Hoechst 33342, in blue) in MCF7, MDA-MB-231, and NKE cells after treatment with DMSO/vehicle or DFE at the indicated concentrations for 16 h. Latrunculin B was used as a positive control. Bar graph representing the estimation of thickness (relative actin filament thickness or RAFT) of individual agglomerated fibres. Error bar, mean ± SD (n = 6 cells from 3 randomly chosen fields). Scale bar-25 µm. Magnification at 100x oil lens.

2.1.3. DFE inhibits breast cancer cell migration

Cancer cell migration or metastasis of cancer cells plays a crucial role in cancer-induced patient death [[20], [21], [22]]. We therefore analysed the effect of DFE on the migration of MCF7 and MDA-MB-231 by scratch assay. Results indicated that DFE induced dose-dependent inhibition of cell migration in breast cancer cells (Fig. 2B).

2.1.4. Ultrastructural analysis of DFE-induced destruction of breast cancer cells

The morphological changes of cell surfaces (MCF7, MDA-MB-231, and NKE) were analysed by scanning electron microscopy (SEM). Smooth and normal cell surfaces were found in the control or DMSO group, with high focal adhesion. Intercellular communications through extended lamellipodia were found in control cell samples (red arrow), essential for maintaining cellular homeostasis. Distinct morphological changes in MCF7 and MDA-MB-231 cells were observed in treatment groups with less cellular density, shortening of cellular area (cell shrinkage), fewer focal adhesions, cell rounding, membrane blebbing (yellow arrow), and formation of apoptotic bodies (green arrow) (Fig. 2C). In contrast, few morphological changes, except cellular area at an ultrastructural level, were found in NKE upon DFE treatment at the same concentrations, indicating that DFE induced programmed cell death in the breast cancer cells with little effect on their normal counterparts. The average cellular area was decreased by 3.48-fold and 10-fold in MDA-MB-231 cells and by 2.53- and 5.12-fold in MCF7 cells after treatment with 50 and 100 µg/ml DFE for 24 h, respectively, compared to 1.24- and 1.34-fold in NKE in mean cellular area, which was found statistically insignificant (Fig. 2C).

2.1.5. DFE induces actin agglomeration in breast cancer cells

Cytoskeletal structures of cancer cells are targeted by anticancer drugs such as latrunculin, cytochalasin, and Taxol (paclitaxel) [23]. Significant agglomeration of cellular F-actin filaments and stabilization of fibres were found with increasing dose of DFE treatment. At 50 and 100 µg/ml, DFE increased F-actin agglomeration in MCF7 cells by 1.23- and 2.91-fold, respectively, and in MDA-MB-231 cells by 2.82- and 4.46-fold, respectively, compared to the vehicle control group. In contrast, DFE treatment did not cause significant changes in actin structure in NKE cells (Fig. 2D). Latrunculin B, was used as a positive control, disrupted microfilament structures by preventing their polymerization, leading to loss of actin filament integrity, which could be visible at the cell periphery (white arrowheads). In contrast, DFE stabilized actin filaments, which appeared as thickened actin bundles over the nucleus (white arrows). These results suggested that the DFE’s action mechanism is opposite to that of Latrunculin B. DFE did not exert any effects on microtubules, while Taxol, used as a positive control in the experiment, affected it (Supplementary figure 2A-C). DFE-induced agglomeration of actin filaments was also found in 4T1 cells (Supplementary figure 3A). Previously, actin fibre agglomeration resulting in cell death induced by high oxidative stress was reported [24].

2.2. DFE reduced 4T1-induced tumour allograft alone and in synergy with doxorubicin in mice

The 4T1 cell-induced mouse model is widely used for studying triple-negative breast cancer (TNBC) due to its close resemblance to the human disease [25]. Like human TNBC, 4T1 cells lack estrogen, progesterone, and HER2 receptors and form highly tumorigenic, invasive tumours that metastasize to distant organs [26]. The tumor progression in this model closely parallels that observed in advanced-stage TNBC patients, considering it ideal for evaluating the therapeutic efficacy of DFE [26]. Prior to testing in the in vivo tumour model, we tested the sensitivity of the murine triple-negative breast cancer cell line 4T1 to DFE. The results indicated that 4T1 cells exhibited increasing death with increasing concentrations of DFE. IC₅₀ of DFE for 4T1 cells was 58.80 μg/ml and the selectivity index (SI) in comparison to NKE is 3.92 (Fig. 3A). Distinct morphological changes were also observed in 4T1 cells upon DFE treatment (Fig. 3B).

Fig. 3.

DFE regresses 4T1 cell-induced tumour in BALB/c mice. [A.] Sensitivity of 4T1 cells to DFE at the indicated doses (µg/ml) was tested through MTT assay in a 24 h period. Doxorubicin was used as a positive control. The effects of the combination of DFE and doxorubicin were also shown. Error bar, mean ± SD (n = 3); Significance considered according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001. IC_50, R-square value and selectivity index (SI) of 4T1 cells compared to NKE after being treated with DFE for 24 h (below). [B.] The morphology of cells in the sensitivity assay was determined by scanning electron microscopy. Doxorubicin was used as a positive control. A combination of DFE and doxorubicin was also used in the study. Magnification – 1000x (1KX) for the initial figure (upper panel) and variable magnification (described below) for single cell morphology (from box with white boundary in upper panel) [C.] Represented image of tumours excised from BALB/c female mice of each experimental group (n = 3) after treatment with the vehicle or DFE (100 mg/kg BW) or doxorubicin (2 mg/kg BW) or combination of DFE and doxorubicin (50 mg/kg BW+ 1 mg/kg BW) for 10 days. [D.] Line graph representing the time-dependent effect of DFE treatment on tumour volume (in mm³). (Red arrows signify applied doses and blue arrow signifies day of sacrifice (DOS).) Error bar, mean ± SD (n = 3), Significance considered according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001. [E.] The bar graph represents tumour weights after treatment. Error bar, mean ± SD (n = 3), Significance considered according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001. [F.] Representative H&E, periodic-acid Schiff (PAS), and IHC images of tumour tissues (40x) of different treatment groups as indicated. Scale bar-50 µm. Bar graph representing the CASPASE 3 percentage in tissues with different treatment groups (right-hand side). [G.] Representative images of analysis of apoptosis in tumour tissues by TUNEL staining.

Next, we tested the efficacy of DFE alone and in combination with doxorubicin (Dox) on the 4T1 induced tumour allograft in BALB/c female mice. Animals with palpable tumours, which developed after one week of injection of 1.5 × 10⁶ 4T1 cells under the breast skin of each animal, were randomly divided into 4 groups (n = 3). Mice of each group were injected with DMSO (vehicle), DFE [100 mg/kg body weight (BW)], Dox (2 mg/kg BW) alone (as positive control) or in combination of DFE (50 mg/kg BW) and Dox (1mg/kg BW) once in every two days (total three doses). Tumour volume was measured on the day of treatment by a digital slide calliper. On completion of treatment, on the 10th day, animals were euthanised, and tumour weights were recorded after extraction. The isolated tumours exhibited a significant reduction in size in the treatment groups, compared to the vehicle control group (Fig. 3C). Tumour volume was significantly reduced in all treatment groups compared to vehicle control (Fig. 3D). Tumor weight was decreased approximately 8-fold and 13-fold in the DFE and Dox treated groups, respectively, compared to the control, while the combination treatment resulted in a reduction of about 3.3-fold (Fig. 3E). The reason for the apparent inefficacy of the combination treatment was not understood, and further studies are required to address this issue (Fig. 3C-E). However, an earlier study did not report better efficacy of treatment of the combination of drugs against solid tumours in comparison to the drugs alone. Gautam et al. [27] applied tyrosine kinase inhibitor neratinib (NER), along with c-MET inhibitor cabozantinib (CBZ) in combination (NER+CBZ) as well as alone (NER/CBZ) against migration-positive solid breast tumour, induced by SKBrM3(BrM3+) cells in female BALB/c mice. They demonstrated that the combination was significantly effective in reducing tumour volume compared to the vehicle control group. Still, no significant difference (NS) was found in the regression of tumour volume or weight between the combination group (NER+CBZ) and the neratinib (NER) alone group [27]. Nevertheless, additional study in the future will clarify if drug-drug antagonism could underlie this result [28]. The H&E staining revealed significant unstained regions in the tissue, less densely nuclear areas, and pleomorphic nuclei in both DFE- and Dox-treated groups (Fig. 3F). Metabolic reprogramming is a key survival strategy of the tumour microenvironment, supporting progression and chemoresistance. Elevated glycogen accumulation has been observed in several cancers, including kidney, breast, uterus, and bladder, particularly during neoplastic transformation [29]. High glycogen enables tumour cells to adapt to oxygen- and nutrient-deprived conditions [30]. Thus, inhibition of glycogen synthesis, indicated by lower cellular glycogen, is indicative of suppression of tumorigenesis. Periodic acid-Schiff (PAS) staining analysis showed a lower presence of stored glycogen content in all treated groups than the vehicle control (Fig. 3F), indicative of less availability of glycogen as an energy source, essential for tumorigenesis. Similarly, a decrease in pro-caspase 3 level was also found in the treated groups compared to the vehicle control group (Fig. 3F). Terminal deoxynucleotidyl transferase dUTP nick end labelling assay (TUNEL) detects DNA fragmentation by labelling the 3′‑hydroxyl termini in the double-stranded DNA breaks in apoptotic cells. Higher levels of TUNEL-positive fields detected in the DFE, doxorubicin (positive control), and combination group compared to the vehicle control group, confirmed cellular apoptosis upon DFE treatment in tumour tissues (Fig. 3G).

2.3. DFE was well tolerated in healthy BALB/c mice

Sixteen BALB/c female mice were randomly distributed into four groups (n = 4). These four groups were then treated intraperitoneally with PBS (vehicle), DFE (100 mg/kg body weight), or dox alone (2 mg/kg body weight) or in combination of DFE (50 mg/kg body weight) and Dox (1 mg/kg body weight) in two days interval for 30 days (total 10 doses). Body weight changes and serum parameters were noted to check the status of the hepatic, cardiac, and renal health in these animals. (Supplementary figure 3C-F). Previous studies with this plant extract showed good tolerance in healthy mice [19]. A comparative table indicated differences in toxicity between AFE and DFE (Supplementary Table 1). Another independent group did not find any significant toxicity in Swiss albino mice with crude methanol extract of Ervatamia coronaria (MEEC) leaves [31]. No significant changes in serum creatinine level were found in both the DFE alone and combination groups, except for an average increase in the Dox-treated samples group (Supplementary figure 3D). Serum glutamic oxaloacetic transferase (SGOT), also known as aspartate aminotransferase (AST), is expressed in various tissues. Its presence in blood is often used as a measure of liver, heart, and other organ function. The highest SGOT activity was observed in the heart, muscle, kidney, brain, and liver [32]. In contrast to DFE, a significantly high AST level was observed in the Dox group, although the combination group showed a relatively much safer range of AST (Supplementary figure 3E).

In contrast to the DFE-treated group, a significant glutamic pyruvate transferase (SGPT) hike was observed in the Dox alone group. The SGPT level in the combination group was found closest to the vehicle control group. SGPT, also known as alanine aminotransferase (ALT), is normally found in liver and heart cells. It starts to be released into the blood when there is damage in the polyhedral hepatic cells and/or in cardiomyocytes [33,34](Supplementary figure 3F).

H & E staining analysis of the isolated liver indicated a slightly higher hepatic portal pressure in the DFE alone-treated group, compared to the control and dox-treated group. The dox-treated group, on the other hand, showed a much greater hepatic damage compared to all groups. The dox-treated group exhibited the highest hepatic portal pressure, as indicated by the significantly increased hepatic portal vein (HPV) area compared to the control or DFE-treated group (Supplementary figure 3G). A much larger sinusoidal space (blue arrow head), severe hepatic cell damage (white arrow head), was also found compared to all other groups, indicating severe hepatotoxicity of doxorubicin (Supplementary figure 3G). Histological analysis of heart tissues indicated no such significant changes in the DFE group and considered in mononuclear infiltration grade 1(MNI G1) with minimal effects compared to the vehicle-treated group. The Dox-treated group, on the contrary, showed a signature of cardiotoxicity. Dox induced severe mononuclear (>20 infiltrations at 400x/field) infiltration at the interstitial space, which is considered in mononuclear infiltration grade 3 (MNI G3), an attribute of idiopathic myocarditis or severe damage category [35] (Supplementary figure 3G) and Zenker’s degeneration (necrosis) [36] was found. The combination group, on the other hand, was the closest to vehicle-treated control (Supplementary figure 3G).

Histological analysis of renal tissues showed that the DFE group showed no significant damage in the renal tissue. In contrast, in the Dox group, some of the Bowman’s capsule in the glomerular region was disrupted (red arrows). The combination group showed no significant changes compared to the vehicle control (Supplementary figure 3G).

2.4. DFE induced cytotoxicity through inducing mitochondrial and whole cellular ROS generation and altering mitochondrial membrane potential in breast cancer cells

We undertook tests to investigate whether a higher level of ROS is generated in DFE-treated cells. Cells were stained with MitoSox Red to check the ROS production in mitochondria. Significant MitoSox Red⁻positive cells and increased MitoSox Red intensity were observed in both MCF7 and MDA-MB-231 cells treated with increasing concentrations of DFE. Interestingly, DFE at the same treatment condition did not cause a significant increase in MitoSox Red intensity in NKE cells, suggesting their relatively lower ROS level and insensitivity to DFE treatment (Fig. 4A-B). DFE induced elevated levels of total cellular ROS in both the breast cancer cells, as indicated by dichlorofluorescein (DCF)-positive cells analysed by FACS (Fig. 5A). Pre-treatment with N-acetyl cysteine (NAC), a ROS quencher, significantly reduced DCF-positive cells, confirming its ROS-quenching effect and protection against DFE-induced damage, as further supported by SEM analysis (Supplementary figure 4A-B). In contrast, no DCF-positive cells were observed under similar treatment conditions, indicating little ROS production in NKE cells (Fig. 5A). Significant protection from DFE-induced toxicity by pre-treatment with NAC, indicated elevated levels of intracellular ROS as the mediator of death of breast cancer cells. Overall, 50 µg/ml DFE was identified as the optimum dose for ROS induction in breast cancer cells. The impact of ROS on cell survival was evidenced by the protective effect of NAC pre-treatment against DFE-induced damage (Fig. 5B).

Fig. 4.

DFE treatment induces mitochondrial ROS. [A.] Mitosox Red™ staining after treatment with DFE for 6 h, indicating ROS generation, as visualized in MCF7, MDA-MB-231, and NKE cells by confocal microscopy (at 63x oil lens). Nuclear co-staining was done with Hoechst 33342 (blue). Doxorubicin was used as a positive control—scale bar-25 µm. [B.] Representative bar graphs estimating the corresponding ROS levels (Mitosox Red ™ intensity). Error bar, mean ± SD (n = 3); ns, not significant; Significance considered according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001.

Fig. 5.

Pre-treatment of cells with N-acetyl cysteine (NAC) protects cells from DFE-induced toxicity [A.] Total cellular ROS was determined by DFCDA staining through flow cytometry. Breast cancer cell lines MCF7, MDA-MB-231, and NKE were treated with 50µg/ml DFE for 6 h with or without pre-treatment with NAC. Doxorubicin (2.5 µM) was used as a positive control. DFE treatment in breast cancer cell lines produced highly fluorescent DCF-positive cells, which were determined by the entry of DCF-positive cells in the DCF-positive threshold box or by the right-shift of the population curve, compared to the control. NAC pre-treatment reduced the DCF-positive cell population. [B.] Representative bar graph showing protection of breast cancer cells from DFE-induced death, upon pre-treatment with N-acetyl cysteine (NAC). Cells were treated with 10 mM NAC for 3 h before DFE treatment for 24 h. Doxorubicin was used for the positive control. Error bar, mean ± SD (n = 3); Significance considered according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001.

DFE treatment at indicated doses for 12 h increases mean TMRM intensity by nearly twofold in MDA-MB-231 cells, indicating mitochondrial hyperpolarization (↑ΔΨm). Conversely, DFE reduces TMRM intensity in MCF7 cells by 1.48 and 9.39 times at 50 and 75 µg/ml, respectively, indicating depolarization (↓ΔΨm) (Supplementary figure 5 A-B). Both hyperpolarization [37] and depolarization [38] of the mitochondrial membrane can trigger the release of apoptotic factors, leading to apoptosis. The variation in mitochondrial membrane potential changes between the two breast cancer cell lines (MCF7 and MDA-MB-231) following DFE treatment may stem from differences in their cellular bioenergetics. MCF7 cells (luminal-like) primarily generate ATP through oxidative phosphorylation under normoxia, while MDA-MB-231 cells (basal-like) rely on glycolysis, following the Warburg effect [39]. Lunetti et al. [40] demonstrated that the respiratory control rate (RCR), a direct measure of mitochondrial respiratory efficiency, was approximately twice as high in luminal-like MCF7 cells compared to basal-like MDA-MB-231 cells across all three mitochondrial complexes (Complex I, II, and IV). They also observed that lactate release was roughly double in MDA-MB-231 cells at all time points, indicating higher glycolytic activity [40]. In essence, MCF7, because of its balanced metabolic profile and reliance on OXPHOS, is better equipped to handle disruption of mitochondrial functions, whereas MDA-MB-231 relies more on glycolysis and has greater mitochondrial activity and mass [41], making it more susceptible to ROS generation in response to different stressors, including anti-cancer compounds [42], which could further contribute to mitochondrial dysfunction and hyperpolarization. DFE treatment may modify these mitochondrial bioenergetics, inducing reactive oxygen species (ROS) production or alteration of MMP, resulting in mitochondrial transition pore (mito-TP) formation or mitochondrial outer membrane permeabilization (MOMP), which could lead to the release of apoptotic factors, like cytochrome C, into the cytosol and subsequent cell death. A significant reduction (potentially due to fragmentation) (Supplementary figure 5 A-B) in mitochondrial length was observed in both breast MCF7 and MDA-MB-231 after DFE treatment, indicative of cell death [43]. In contrast, DFE had minimal effects on NKE cells (Supplementary figure 5 A-B). FACS analysis revealed a right-shift (increased TMRM high population) with increasing DFE concentrations in MDA-MB-231 cells, suggesting mitochondrial membrane hyperpolarization, whereas MCF7 cells showed a left-shift (increased TMRM low population), indicating depolarization (Supplementary figure 5C). In contrast, similar DFE treatment did not move the NKE cell population from the TMRM-low gate (black box) into the TMRM-high gate (blue box), indicating that DFE does not alter the MMP of NKE cells (Supplementary figure 5C). DFE-induced alterations in MMP were also analyzed by JC1 staining (Supplementary figure 6). In MCF7 cells, DFE treatment (for 12 h) caused a concentration-dependent decrease in JC1 aggregates (red) and an increase in JC1 monomers (green), lowering the red-to-green ratio consistent with mitochondrial depolarization. In contrast, MDA-MB-231 cells showed an increase in JC1 aggregates and a decrease in JC1 monomers, resulting in a higher red-to-green ratio consistent with mitochondrial hyperpolarization. Minimal changes were detected in NKE cells (Supplementary figure 6).

2.5. DFE induces apoptotic death in breast cancer cells

The TUNEL assay and reduction of caspase-3 suggested a DFE-triggered apoptotic death in 4T1 induced-tumour cells. To follow that up, we investigated whether DFE indeed induced apoptosis in breastcancer cells. Therefore, we exposed breast cancer cells to different concentrations of DFE and investigated the status of apoptotic markers in those cells. We observed a dose-dependent increase in the cleavage of PARP1(Fig. 6A-B), cleaved CASPASE-3 (Supplementary figure 7D and E) formation, as well as a decrease in CASPASE-7 levels, clearly indicating DFE-induced apoptosis in these cells (Fig. 6A-B). In contrast, little change in the cleavage status of PARP1 indicated insensitivity of NKE cells to DFE treatment (Fig. 6C). Cell cycle arrest at the sub-G1 phase is in line with the results that DFE-induced apoptotic death in breast cancer cells (Fig. 6D) [44].

Fig. 6.

DFE-induced cell death in MCF7 and MDA-MB-231 cells involves apoptotic and autophagy mechanisms [A.-B.]. Immunoblots of whole cell extracts of MCF7 and MDA-MB-231 cells were pre-treated with indicated concentrations of DFE for 24 h. Cellular markers tested as indicated. Beta-actin was used as an internal loading control. Relative fold changes of protein expression levels given below, were normalized with the loading control, Beta-actin. [C.] Immunoblots of indicated cellular marker, PARP1 of extracts of NKE cells pre-treated with indicated doses of DFE and doxorubicin (positive control) for 24 h. Beta-actin was used as an internal loading control. Relative fold changes of protein expression are given below, which were normalized with the loading control, Beta-actin. [D.] DFE induces cell cycle arrest at the sub-G1 phase in both MCF7 and MDA-MB-231 cells. Cytofluorimetric analysis of MCF7 and MDA-MB-231 cells treated with DMSO (vehicle control) and DFE at different concentrations as indicated for 24 h. [E.] Cytofluorimetric analysis of cytochrome C release after treatment of DFE and/or vehicle for 20 h. Error bar, mean ± SD (n = 3), P = 0.0047 for MCF7 and P = 0.0011 for MDA-MB-231; Significance considered according to *P<0.05, **P <0.01, *** <0.001, ****P<0.0001. [F.] Bar graphs representing cell viability of respective cell lines after treatment with DFE (50 µg/ml). Cells were pre-treated with Z-VAD-FMK (pan-caspase inhibitor) or chloroquine (autophagy inhibitor) as indicated.

Further investigations revealed the induction of autophagic death in these cells, accompanied by the formation of LC3II and a reduction in Beclin-1 and ATG7 levels. This indicated DFE-induced autophagy to be ATG-independent (Fig. 6A-B, Supplementary figure 7D and E) [[45], [46], [47]].

The release of Cytochrome C from the mitochondria to the cytosol is a hallmark of apoptotic cell death. Therefore, we tested Cytochrome C release from the mitochondria (caused by mitochondrial outer membrane permeabilization) through FACS [48]. We observed that, compared to the vehicle control, DFE treatment (at 50 µg/ml) resulted in a 4-fold and 2-fold increase in the release of cytochrome C, in MDA-MB-231 and MCF7, respectively. This indicated the DFE-induced killing of breast cancer cells is mitochondria-dependent (Fig. 6E).

To analyse the contributions of apoptosis and autophagy in cell death, the breast cancer cells were pre-treated with Z-VAD-FMK and chloroquine, respectively, before DFE exposure, followed by an MTT assay to assess cell viability. Z-VAD-FMK was found to offer significant protection from DFE-induced toxicity (about 28 % protection in MDA-MB-231 cells and about 32 % in MCF7 cells, respectively) in contrast to chloroquine. This indicated that DFE-induced killing of breast cancer cells was mainly mediated by apoptosis. An insignificant protection offered by chloroquine from DFE-induced death indicated that autophagy had little or no role in the killing mechanism (Fig. 6F).

In addition, annexin V-FITC and propidium iodide (PI) dual staining indicated DFE-induced apoptosis in breast cancer cells. Our results, however, showed early and late apoptosis in MCF7 and MDA-MB-231 cells, respectively (Fig. 7). This difference in cell death between these two cell lines (MCF7 and MDA-MB-231) was probably due to differences in the involvement of executioner caspases, which are located downstream of the intrinsic apoptotic pathway. MCF7 lacks the expression of CASPASE 3, because a 47-base pair deletion of the caspase-3 gene [49] causes exon skipping at pre-mRNA splicing, resulting in a premature stop codon. Several groups have found almost no expression [50] or very low expression of CASPASE 3 in MCF7 cells [51]. Our study also found very little expression of cleaved CASPASE-3 in MCF7 cells (Supplementary figure 7D). However, intrinsic or mitochondria-induced apoptosis does occur in MCF7 cells, primarily mediated by executioner caspases, including CASPASE 7 and CASPASE 6 [52,53]. Boucher et al. [52] found CASPASE 7 to be a potent player in PARP-1 cleavage in CASPASE 3-deficient MCF7 cells. However, compared to CASPASE 7, CASPASE 3 is considered more promiscuous and a major executioner caspase, especially during the demolition phase of apoptosis [54], which could explain the DFE-induced differences in cell death in these cells. Several groups found an early apoptotic population in MCF7 cells treated with various anti-cancer compounds [55,56]; however, the present study also observed a small percentage of late apoptotic population in DFE-treated MCF7 cells (Fig. 7, Fig. 8).

Fig. 7.

DFE induces apoptotic death in MCF7 and MDA-MB-231 cells, as observed by annexin V-propidium iodide (PI) dual staining. [A.-D.] Confocal images of MCF7 (A) and MDA-MB-231 (C) cells pre-treated with indicated doses of DFE (in µg/ml), DMSO, or doxorubicin (1.25 µM, as positive control) for 24 h were stained with annexin V and propidium iodide (PI) (laser scanning confocal microscope images through 63x oil lens). Hoechst 33342 stained the nuclei—scale bar- 25 µm. Bar graphs B and D represent the estimation of panels A and C, respectively. BF, bright fields. The dose-dependent increase of cell death was represented by a scatter plot based on LAS X analysis of the obtained images. Bar graphs representing the relative intensity of annexin V and propidium iodide (PI) in treatment groups. Error bar, mean ± SD (n = 3); *P<0.05, **P <0.01, *** <0.001, ****P<0.0001.

Fig. 8.

Apoptosis assay using flow cytometry after Annexin-FITC/propidium iodide (PI) dual staining. The X-axis represents annexin V-FITC labelled population, and the Y-axis represents PI-labelled population. The lower left quadrant represents viable cells (An-, PI-), the lower right column represents early apoptotic population (An+, PI-), the upper right column represents late apoptosis (An+, PI +), and the upper left column represents dead cells/ necrotic population (An-, PI+). MCF, MDA-MB-231, and NKE cells were treated with DFE (50 µg/ml) in the presence or absence of N-acetyl cysteine (10 mM) (NAC) for 24 h. Doxorubicin (5 µM) was used as a positive control.

To elucidate the role of ROS in DFE-induced apoptosis, we treated MCF7, MDA-MB-231, and NKE cells with DFE (50 µg/ml) in the presence or absence of NAC, followed by Annexin V-FITC/PI dual staining and subsequent FACS analysis. NAC pre-treatment caused about 18 % and 58 % recovery in cell survival in DFE-treated MCF7 and MDA-MB-231 cells, respectively (Fig. 8). MDA-MB-231 cells are more susceptible to ROS-induced death than MCF7 cells, which is probably due to the lack of glutathione peroxidase 4 (GPX4) in MDA-MB-231 cells compared to MCF7 cells [57,58]. Due to such redox imbalance, MDA-MB-231 cells become more susceptible to ROS generation and ROS-induced cell death compared to MCF7 cells. No such NAC-mediated improvement in cell survival was observed in NKE cells, as the same concentration of DFE (50 µg/ml) failed to produce significant ROS in them (Fig. 8). The fluorescence intensities of annexin V-FITC and PI were found to increase with increasing dose of DFE in MDA-MB-231 cells (Fig. 7). Interestingly, however, the PI fluorescence was much more intense and was detected from the membranes of MDA-MB-231 cells, instead of their nuclei (Supplementary figure 8). Notably, the PI signal is usually detected from the nuclei of necrotic cells [59]; although PI is also known to associate with damaged cell membranes that have less integrity [https://www.ptglab.com/products/CoraLite-488-Annexin-V-andPIApoptosisKitPF00005.htm?srsltid=AfmBOor5ZwO3pbOViqRkonRJr5e8zA1Gia0bUK94TWMqycDlrS4kH]. As FACS identifies overall fluorescence only, this predominance of the PI signal over the annexin V-FITC signal resulted in a shift of the late apoptotic cell population (annexin V-FITC +ve and PI +ve quadrant) to the necrotic cell population (annexinV-FITC -ve and PI +ve quadrant) in FACS analysis. (Fig. 8). Thus, our results ruled out the possibility of DFE-induced necrotic death in MDA-MB-231 cells and established DFE-induced late apoptosis as the sole mode of death. A distinct shrinkage of DFE-treated MDA-MB-231 cells (indicated by red arrow) (Supplementary figure 8) compared to the vehicle control group further corroborated DFE-induced apoptosis in this cell line, as cell swelling is known to be a hallmark of necrosis [60]

3. Discussion

Our study highlighted an aspect of the molecular basis of DFE-induced death in both triple-positive (MCF7) and triple-negative breast cancer (MDA-MB-231) cells, with little effect on NKE cells (Fig. 2A). DFE treatment significantly blocked the migration potential of breast cancer cells (Fig. 2B) and distinctively induced apoptotic death in these cells. Ultrastructure analysis through scanning electron microscopy (SEM) indicated the absence of inter-cellular communications due to destruction of lamellipodial structures, loss of cell volume, and shrinkage, protuberances of smooth cellular membrane, loss of focal adhesion, cell rounding and formation of blebbed bodies in DFE treated cells as the physio-morphological attributes of apoptosis (Fig. 2C) [61,62]. These physio-morphological attributes of apoptosis are supported by cytochrome C release (Fig. 6E), PARP1 cleavage, CASPASE activation (Fig. 6A and B; Supplementary figure 7D and E), and annexin/PI positivity (Fig. 7, Fig. 8) in both cell lines. In NKE cells, no such changes in cell surface morphology were found at the same treatment condition, except for some cellular shrinkage after 24 h treatment (Fig. 2C). Fluorophore-tagged phalloidin staining of F-actin indicated that DFE treatment produced distinct actin agglomeration in both MDA-MB-231 and MCF-7 cells, but interestingly, no such agglomeration was found in NKE (Fig. 2D). No particular evidence of alteration of microtubule dynamics was found in any cell type upon DFE treatment (Supplementary figure 2A-C). DFE induced apoptosis in 4T1, the murine TNBC cells in vitro in a dose-dependent manner (Fig. 3A and B), and also induced actin agglomeration (Supplementary figure 3A). Many anti-cancer compounds, especially those from natural sources, disrupt cytoskeletal structures, including microtubules, actin (microfilaments), intermediate filaments, and septins [23], which are essential for maintaining cell shape, spatial organization of cellular components, physical interactions with the external environment, and cell movements [63]. In our study, we found significant DFE-induced apoptosis in 4T1-induced tumour allografts of BALB/c female mice, with reasonable tolerance of DFE in healthy female mice (Figs. 3C-G and Supplementary figure 3C-G). Analysis of tumour tissue samples clearly showed tumour regression mechanisms when DFE was applied alone or in combinatorial doses (Fig. 3F-G).

DFE-induced apoptosis of breast cancer cells was caused by elevating the level of intracellular ROS (Fig. 4, Fig. 5), as pre-treatment with NAC abrogated the anticancer effect of DFE (Fig. 5A, Supplementary figure 4A-B). DFE treatment alters the mitochondrial membrane potential (MMP) (Supplementary figure 5 and 6) in both breast cancer cells (but not in NKE cells), leading to mitochondrial outer membrane permeabilization (MOMP) and cytochrome C release, resulting in cell death. However, minimal changes in MMP were observed in NKE cells following DFE treatment at the same concentrations (Supplementary figure 5 and 6), suggesting that the DFE-induced alteration of MMP is specific to breast cancer cells. Cancer cells, including PC3 and MCF7 cells, become vulnerable to sudden ROS bursts due to their defective antioxidant response [[64], [65], [66]]. DFE-induced downregulation of SOD1 helps accumulate higher cellular ROS levels and mitochondria-dependent apoptosis (Supplementary figure 5–7, Fig. 6E) [[67], [68], [69]]. It appears that DFE-treated cells also downregulate AKT-1, triggering apoptosis (Supplementary figure 7) [70]. Many anticancer drugs (chemotherapeutics), including natural products, have been shown to function by altering the cellular antioxidant pathway [[65], [66]].

Additional studies will be necessary to identify the source responsible for the upregulation of ROS following DFE treatment. A previous study demonstrated that treatment of prostate cancer cells with a polyphenol-rich fraction of Bergenia ligulata led to increased catalytic activity of monoamine oxidase-A (MAO-A), leading to toxic levels of ROS production [[64], [65]].

The DFE reported here consists of 8 major indole alkaloids per LC-MS/MS analysis, which include stemmadenine, voafinidine, voaphylline, isovoacristine, vallesamine, pericyclivine, o-acetyl-vallesamine, and 19,20-dihydrotabernamine (Fig. 1A-B and Supplementary Table 2) among a total of 17 identified compounds (Fig. 1C).

Earlier, an LC-MS analysis of a methanolic (alkaloid-rich) fraction of Ec, termed AFE, identified six major compounds: 3-oxo-voacangine, voacamine, 19,20-dihydrotabernamine, criophylline, dregamine, and N-methylalanine [19] (Supplementary Table 2). AFE was shown to regress colorectal cancer by mechanisms independent of intracellular ROS production [19]. In addition, unlike DFE, primary AFE action involved inducing death in colorectal cancer through autophagy without targeting actin structure. An apparent difference in chemical composition between these two preparations could be due to the plant materials being harvested from widely different locations, as well as the use of different organic solvents, which have varying polarities and eluotropic strengths (Ɛº) such as dichloromethane vs. methanol. In addition, Dutta et al. [19] collected Ec leaves from a southern state of India, whereas this study used Ec leaves collected from West Bengal, an eastern state of India. Climate, weather, soil nature, and geographical factors are known to influence the constitution and the levels of secondary metabolites in plants [71].

While many of the compounds reported here have already been shown to possess anticancer properties, future investigation will determine whether the observed effect is due to the individual or combinatorial functions of the compounds, as well as provide deeper insight into the observed anticancer activity. It is also possible that the responsible molecule remains undetected by the LC-MS/MS analysis.

Many of these phytochemicals are also being tested in various stages of clinical trials, offering promising prospects [72]. Nevertheless, the therapeutic application of many phytochemicals is constrained by their poor solubility, low bioavailability, and limited stability [[73], [74]] —the fundamental challenges that future research must overcome (Fig. 9).

Fig. 9.

Schematic diagram of DFE activity.

4. Materials and methods

Flow chart, summarising materials and methods

4.1. Preparation of extract and fractions

Dark green Ervatamia coronaria (Ec) leaves were collected from Hooghly district (22.90ºN 88.39ºE), West Bengal, India, in August 2020. A proper herbarium specimen was prepared, and identification was carried out by botanist Professor Swadesh Sarkar of Chandernagore Government College, West Bengal, India. The prepared herbarium specimen was deposited in the Central National Herbarium (CNH) of the Botanical Survey of India (BSI), Kolkata, India, and certified (No: CNH/Tech.II/2024/127) for future reference. After drying in the shade, leaves were powdered by a mechanical grinder. The powdered leaves were soaked in methanol (100 g of powder in 500 ml of methanol in a 1-litre capacity flask) for 4 days with occasional mixing before collecting the extract by filtration. The extraction of the powder with methanol was repeated several times until the filtrate appeared almost colourless. The methanolic extract (filtrates pooled together) was thus collected and concentrated by a vacuum evaporator. Then the dried methanolic fraction was further fractionated by using 250 ml of hexane, followed by 250 ml of Dichloromethane (DCM), and finally 250 ml of methanol (Supplementary figure 1A). All the fractions were then dried and dissolved individually in DMSO and stored in −20 degrees in small aliquots. The fractions were then applied to breast cancer cell lines MCF7, MDA-MB-231, 4T1, and standard cell line (NKE) at varying concentrations for 24 h by MTT assay. The DCM fraction of Ec leaves (hereafter referred to as DFE) exhibited the highest sensitivity against MCF7, MDA-MB-231, and 4T1 cells. LC-MS analyzed compounds present in DFE. The fractionation procedure was done using the method described by Dutta et al. [19] with some modifications [19].

4.2. UPLC-Q-TOF-MS/MS (LCMS) analysis

UPLC-Q-TOF-MS/MS (LCMS) was performed as described by Dutta et al. [19] to analyze the composition of the DFE with minor changes.

4.3. Cell culture

Human breast cancer cell lines MDA-MB-231, MCF7 and regular cell line NKE (normal kidney epithelial) /or murine breast cancer cell line 4T1(all procured from ATCC), were cultured with DMEM (Gibco) and RPMI1640 (Gibco), respectively supplemented with 10 % fetal bovine serum (cat no- 16,000–044, Gibco), 1 mM l-glutamine, 50 μg/ml penicillin, streptomycin (50 μg/ml, Himedia) and amphotericin B (2.5 μg/ml, Himedia), gentamycin (50 mg/ml, Himedia) and non-essential amino acids (Himedia) at 37 °C with 5 % CO₂ in a humidified incubator (Heracell vios 160i from Thermo Scientific) [[19], [65]].

4.4. Cytotoxicity assay

Cytotoxicity of DFE was biochemically tested by 3-(4,5-diethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT). Cells were treated with desired concentrations of DFE at 70 percent confluency in a 48-well plate in a humidified incubator with 5 % CO₂ for 24 h. After the desired experiment, the respective media was removed and MTT solution (0.5 mg/ml) (Himedia RM-1131) was added for 4 h at 37 °C. MTT media was then removed, and formazan crystals were dissolved in DMSO, and reading was taken at 570 nm in a UV–VIS spectrophotometer (Multiscan Go from ThermoScientific) [19].

4.5. Scratch assay

Scratch assay was done using the method described by Liang et al. [75] and Ghosh et al. [65] with some modifications [[65], [75]]. 2 × 10⁵ cells (MCF-7 and MDA-MB-231) were plated in 6-well plates until about 80 % confluency. Cells were treated with different doses of DFE (20 and 40 μg/ml) or DMSO for 24 h in 10 % FBS-containing media. Next, scratches were made in each well by a 20 μl pipette tip, followed by replacing the growth media with 5 % FBS. Next, the images were taken with a Leica inverted microscope.

4.6. Scanning electron microscopy

Sample preparation for scanning electron microscopy was done according to methods described by Ali et al. [76] and Heckman et al. [77] with some modifications [76,77]. Briefly, the cells were seeded (2 × 10^4 cells/cover slip) onto sterile cover slips kept in a 6-well plate. The next day, these cells were treated with different concentrations of DFE for 24 h at 37 °C. Cells were then fixed rapidly after removal of DFE-containing media with prewarmed (at 37 °C) 3.7 % (V/V) EM grade glutaraldehyde (Sigma-Aldrich) in 0.1 M PBS (pH 7.3) for 15 min at room temperature. Following PBS wash, the cells were dehydrated by treating sequentially with increasing concentrations (30 %, 50 %, 70 %, 90 %, and 100 %) of ethanol (Merck) for three times each for 30 min at room temperature, followed by hexamethyldisilazane (HMDS) (SRL) for 10 min. Then the processed cover slips were transferred to a new 6-well plate to air-dry for 24 h at room temperature. Finally, the samples were sputter-platinum coated (Quorum 150T ES) and examined by a Zeiss EVO 18 special edition variable pressure scanning electron microscope.

4.7. Actin-phalloidin staining

Cells (at 70 % confluency) after being treated with DFE, and/or DMSO/ latrunculin B (100 nM) (cat no L5288, Sigma-Aldrich) as positive control for 16 h, were washed with 1x PBS to fix with 3.7 % paraformaldehyde (in PBS pH-7.4) for 10 min at room temperature. After 1x PBS wash, permeabilization of cells was made by applying 0.1 % Triton X-100 (in PBS pH-7.4) at room temperature for 15 min, followed by 1x PBS wash to remove Triton X-100. Cells after incubation with 2 % BSA in PBST (1x PBS+ 0.1 % Tween-20) for 1 h were washed in PBS, followed by treatment with 5 units/ml phallotoxin, phalloidin (Phalloidin iFluor 488 from abcam # cat no ab176753) for 30 min at room temperature. After 1x PBS wash, the nuclei of the cells were stained with Hoechst 33342 and the cells were then visualized and analysed by Leica Stellaris 5 confocal microscope and Leica Application Suite-X (LAS-X) software, respectively. Sample preparation was done according to the methods described by Sen et al. [78] with some modifications [78].

4.8. Immunocytochemistry (ICC)

Cells (MCF7, MDA-MB-231, and NKE) were grown on a confocal dish (SPL LIFESCIENCES Cat No 200350) until 70 % confluency was achieved. On the very next day, cells were treated with desired concentrations of DFE and DMSO as a vehicle control for 16 h/18 h(18 h incubation was done for AKT-1 and SOD-1). After treatment completion, DFE and /or DMSO-containing media were removed, and cells were washed with 1x PBS. Then, the cells were fixed with pre-warmed 3.7 % paraformaldehyde for 10 min at room temperature. Cell samples were then washed with 1x PBS (3 times) for the removal of paraformaldehyde. Permeabilization of cells was then made by applying 0.1 % Triton X-100 (in PBS pH 7.4) at room temperature for 20 min, followed by 1x PBS wash for 3 times for 5 min each to remove Triton X-100. Blocking of cell samples to eliminate non-specific signals was done by using 1 % BSA in PBST (1x PBS+ 0.1 % Tween-20) for 1 h at room temperature. Then the cell samples were incubated with desired primary antibodies (β-TUBULIN, cat no AC008, SOD-1, cat no A0274, and AKT-1, cat no A17909 from ABclonal, USA) diluted in 1 % BSA in PBST at 4 °C overnight, followed by decanting the primary antibody-containing solution and 1x PBS wash for 3 times. Finally, the cell samples were treated with secondary antibody (Rhodamine-TRITC Goat anti-rabbit, catalogue no AS040 from ABclonal, USA) diluted in 1 % BSA in PBST at room temperature for 1 h in the dark, followed by 1x PBS wash for 3 times in the dark. Nuclear staining was done with Hoechst 33342 (5µg/ml). A Leica Stellaris 5 confocal microscope was used to examine cell samples, and the obtained images were analysed by Leica Application Suite-X (LAS-X) software. Sample preparation was done according to the methods described by Lundström et al. [79] with some modifications [79].

4.9. 4T1 orthotopic allograft model in BALB/c female mice

Female BALB/c mice of 4–6 weeks (average weight 25 g approx.) were maintained in a regulated environment with 23ºC ± 2ºC, 50 % humidity, and 12 h light/dark cycle with proper diet and water. Mice were divided randomly into 4 groups of three animals each (n = 3). Tumours were developed for 7 days by injecting 1.5 million freshly harvested healthy 4T1 cells subcutaneously per animal. Groups were treated with DMSO, DFE (100 mg/kg body weight), doxorubicin (2 mg/kg body weight), or a combination of DFE and doxorubicin (DFE-50 mg/kg body weight + dox-1mg/kg body weight) intraperitoneally for 1 week in 2-day intervals (in total 3 doses). The animals were euthanised on the 10th day. The weight and volume of excised tumours harvested from the animals were recorded [19]. Permission for animal experiments was taken from the Bose Institute animal ethical committee (reference number- IAEC/BI/84/2018) [[19], [65]].

4.10. H and E, periodic acid Schiff (PAS) staining, immunohistochemistry (IHC), and TUNEL assay

Tumour samples collected from all four groups were fixed in neutral buffer formalin (NBF), pH 7.36, and then rinsed in 1XPBS. After proper processing, paraffin embedding of tumour samples was done with the standard protocol. Sections were made by an advanced rotary microtome. Tumour sections were then deparaffinized in 100 % xylene, followed by rehydration and finally stained with haematoxylin and eosin (H and E) as well as Periodic-acid Schiff (PAS). Stained slides of tumour samples were then analysed by a Leica inverted phase contrast microscope (Leica DMIL) for morphological and total glycogen content, respectively.

Immunohistochemistry was performed according to the protocol described by Dutta et al. [19]. Paraffin sections were deparaffinized by 100 % xylene, followed by xylene: ethanol=1:1, 100 % ethanol, 90 % ethanol, 70 % ethanol, and 50 % ethanol. IHC for pro-caspase 3 (catalogue no A16794 from ABclonal, USA) was then performed by the abcam IHC kit according to the manufacturer’s protocol [[19], [65]].

Tunnel assay was done according to the protocol described by Liu et al. [80] with some modifications [80]. Paraffin sections were deparaffinized by 100 % xylene, followed by xylene: ethanol=1:1, 100 % ethanol, 90 % ethanol, 70 % ethanol, and 50 % ethanol. Then the tissue samples were treated with proteinase K at 37 ºC for 2 h to remove extracellular matrix components. Tissue samples were then washed with 1X PBS for 3 times. Finally, the tissue samples were treated with BrdUTP/TdT cocktail from Roche, Mannheim, Germany, In Situ cell death detection kit (TUNEL) (ref no 11684795910). Tissue samples were then washed with 1X PBS and mounted with Fluoroshield solution, and tissue samples were visualized with an Olympus CKX53 fluorescence microscope.

4.11. Toxicity study in healthy BALB/c female mice

BALB/c female mice of 4–6 weeks old (average weight- 20–25 g) were taken and separated randomly into 4 groups having 4 animals in each group. Animals were treated with DMSO (as vehicle control), DFE (100 mg/kg body weight), doxorubicin (2 mg/kg body weight) or combination of DFE and doxorubicin (DFE-50 mg/kg bodyweight + dox-1mg/kg body weight) intraperitoneally (IP) for 1 month in 2 days interval (total 10 doses) until animals were euthanized. Bodyweight was recorded on the day of each treatment, throughout the treatment period. After completion of treatment, animals were sacrificed, and the indicated organs and serum samples were collected. Serum parameters, including SGOT, SGPT, and creatinine, were estimated using respective kits (Erba Mannheim kits) according to the manufacturer's protocol [65]. Collected indicated organs were processed, and H&E staining was done for histological analysis. Permission for animal experiments was taken from the Bose Institute animal ethical committee (reference number- IAEC/BI/84/2018).

4.12. Whole cell lysate preparation and immunoblot

Cells (MCF7, MDA-MB-231, and NKE) were harvested after desired treatment with a cell scraper on ice and in ice-cold PBS. After harvesting, cells were washed with PBS 3 times to remove serum-containing media. Then the cells were lysed by lysis buffer (20 mM Tris–HCl, pH 7.5, 1 % Triton X-100, 150 mM NaCl, 5 % glycerol, one mM Phenylmethylsulfonylfluoride (PMSF), 10 μg/mL leupeptin, 10 μg/mL aprotinin, 20 mM each of sodium fluoride and sodium orthovanadate, pH 7.4) to have the whole cell lysate. Concentration of proteins in the whole cell lysate was measured by using Bradford reagent (Bio-Rad) by UV–VIS spectrophotometer (Multiscan Go from Thermo Scientific) at 595 nm [65]. (Antibodies used with catalogue number: ATG7 D12B11, CASPASE 7 #9492 from Cell Signalling Technology, USA, PARP-1 cat no A0942, BECLIN-1 cat no A7353, Abclonal, USA, CLEAVED CASPASE 3 cat no 9664S, from CST USA, LC3B (KO validated) cat no A7198, β-ACTIN Catalogue no# BB-AB0024 from BioBharti LifeScience, India, HRP-Goat anti-rabbit secondary antibody was from BioBharti LifeScience, India, Catalogue no# BB-SAB01B.)

4.13. Cell cycle analysis

Cells grown to 70 % confluency were treated with DFE at indicated concentrations, or DMSO as vehicle control, for 24 h. Cells were collected by trypsinization and kept at −20ºC overnight in 80 % ethanol for fixing. The next day, cells were resuspended in 1X PBS at 4ºC for 2 h to rehydrate. The cells were then treated with 20 µM RNase A at 37ºC for 2 h. Finally, after incubation with Propidium iodide (PI) (supplied with BD Pharmingen apoptosis kit) for 20 min at room temperature in the dark, the cell-cycle distribution of cells was done by BD FACS Verse analyser [65].

4.14. Cytochrome C release assay and sorting through FACS

Cells grown to 70 % confluency in a 6-well plate were treated with indicated concentrations of DFE and /or DMSO as a vehicle control for 20 h. After treatment completion, DFE and /or DMSO-containing media were removed, and cells were washed with 1x PBS and trypsinized to collect in a 1.5 ml microfuge tube, followed by a 1x PBS wash. Then, the cells were fixed with pre-warmed 3.7 % paraformaldehyde for 10 min at room temperature. Cell samples were then washed three times with 1x PBS to remove paraformaldehyde. Permeabilization of cells was then made by applying 0.1 % Triton X-100 (in PBS pH 7.4) at room temperature for 20 min, followed by 1x PBS wash for 3 times for 5 min each to remove Triton X-100. Blocking of cell samples to eliminate non-specific signals was done with 2 % BSA in PBST (1x PBS+ 0.1 % Tween-20) for one hour at room temperature. Then, the cell samples were incubated with anti-cytochrome C (Catalog No 4272T from CST, USA) antibody diluted in 1 % BSA in PBS-T (1:100) at 4 °C overnight, followed by decanting the primary antibody-containing solution and washing with 1x PBS three times. Finally, the cell samples were treated with secondary antibody (1:1000) (Alexafluor-488 Goat anti-rabbit, reference no A11034, Invitrogen, USA) diluted in 1 % BSA in PBST at room temperature for 1 h in the dark, followed by 1x PBS wash for 3 times in the dark. Cells were then analysed by the BD FACS Verse instrument through the FITC channel [81].

4.14.1. Live cell imaging

4.14.1.1. MitoSox Red ™ staining

Cells grown in a confocal dish (SPL LIFESCIENCES Cat No 200350) at 70 % confluency were treated with different concentrations of DFE /DMSO (vehicle control) /2.5 µM doxorubicin (positive control) for 6 h. Following treatment, the cells were washed thoroughly with 1X PBS. Then, the cell samples were incubated with 5 µM MitoSox Red ™ reagent (Molecular Probes, Invitrogen) for 10 min in the dark at 37º C in a humidified chamber (in dark) attached with a confocal microscope unit (from Oko lab ™), followed by a gentle 1X PBS wash for 3 times. Nuclei were stained with Hoechst 33342. Finally, live cell imaging was performed using a Leica Stellaris 5 super-resolution confocal microscope. The obtained results were analyzed using LAS-X software, and graphs were plotted with GraphPad Prism 8.0 software. Sample preparation was done according to the methods described by Zhang et al. [82] with some modifications [82].

4.14.1.2. Tetramethyl-rhodamine methyl ester (TMRM) staining

Cells were grown in confocal dishes (SPL LIFESCIENCES Cat No 200350) up to 70 % confluency and treated with the desired concentrations of DFE /DMSO (vehicle control)/2.5 µM doxorubicin (positive control) for 12 h. Following treatment, cells were washed three times with 1x PBS (pH 7.4) to remove as much drug/vehicle-containing media as possible. Then, the cells were incubated with the TMRM reagent (Invitrogen, Thermofisher Scientific) at a final concentration of 40 nM for 30 min at 37 °C in a humidified chamber (in the dark), from Oko lab ™ attached to a confocal system. Three 1x PBS washes followed this. Nuclei were stained with Hoechst 33342 (Invitrogen), and then live-cell imaging was performed using a Leica Stellaris 5 super-resolution confocal microscope. The obtained results were analyzed using LAS-X software [83], and graphs were plotted with GraphPad Prism 8.0 software.

4.14.1.3. Annexin V-FITC and propidium iodide staining

Annexin V-FITC/PI dual staining (Elabscience REF E-CK-A211, USA) was done (Hoechst 33342 staining was also done for nuclear localization) for the detection of the type and magnitude of DFE-induced apoptosis in both MCF7 and MDA-MB-231 cell lines. Cells were grown in a confocal dish (SPL LIFESCIENCES Cat No 200350) up to 70 % confluency and treated with the desired concentrations of DFE / DMSO (vehicle control)/1.25 µM doxorubicin (positive control) for 24 h. After completion of the treatment, cells were washed with 1xPBS (pH- 7.4) followed by application of annexin V-FITC/PI, diluted in 1x binding buffer for 20 min in the dark at room temperature, and then rinsed with 1x binding buffer (last wash buffer contained Hoechst 33342), according to the manufacturer's protocol. Finally, live cell-imaging was done by Leica Stellaris 5 super-resolution confocal microscope, and the obtained results were analysed by LAS-X software as described [84] with some modifications.

4.14.1.4. JC 1 staining

Cells were grown in confocal dishes (SPL LIFESCIENCES Cat No 200350) up to 70 % confluency and treated with the desired concentrations of DFE /DMSO (vehicle control) for 12 h. Following treatment, cells were washed three times with 1x PBS (pH 7.4) to remove as much drug/vehicle-containing media as possible. Then, the cells were incubated with JC-1 reagent (Invitrogen, Thermo Scientific, USA) at a final concentration of 10 μg/ml for 30 min at 37 °C in a humidified chamber (in the dark), from Oko lab ™ attached to a confocal system. This was followed by 1x PBS wash for 3 times. Nuclei were stained with Hoechst 33342 (Invitrogen), and then live-cell imaging was performed using a Leica Stellaris 5 super-resolution confocal microscope [85]. The analyzed results were plotted using GraphPad Prism 8.0 software.

4.15. TMRM sorting by FACS analysis for understanding changes in MMP

Cells grown to 70 % confluency in a 6-well plate were treated with indicated concentrations of DFE and /or DMSO as a vehicle control for 12 h. Cells were then washed with 1X PBS, followed by resuspension in 500 µL 1X PBS, and finally stained with TMRM reagent (Invitrogen, Thermo Scientific) at a final concentration of 40 nM for 30 min at 37 °C in the dark. Cells were analysed by BD FACS verse instrument through PE bandpass filter [86].

4.16. DCFDA sorting by FACS for quantification of total cellular ROS

Cells grown to 70 % confluency in a 6-well plate were treated with indicated concentrations of DFE (in the presence or absence of NAC) (NAC pre-treatment was done for 3 h) and /or DMSO as a vehicle control for 6 h. Cells were then washed with 1X PBS, followed by resuspension of the cells in 500 µl 1X PBS by scraping. They were finally stained with the DCFDA reagent (Invitrogen, Thermofisher Scientific) at a final concentration of 10 µM for 30 min at 25 °C in the dark. Doxorubicin 2.5 µM was used as a positive control. The BD FACS Verse instrument analyzed cells through the FITC channel [87].

4.17. Annexin- V FITC / PI dual staining assay for FACS analysis

Cells grown to 70 % confluency in a 6-well plate were treated with indicated concentrations of DFE (in the presence or absence of NAC) and /or DMSO as a vehicle control for 24 h. Cells were then washed with 1XPBS, followed by resuspension in 1X binding buffer, followed by FITC-Annexin V/PI (BD Pharmingen) for 25 min in the dark at room temperature. Doxorubicin 5.0 µM was used as a positive control. Finally, cells were analyzed by the FACS Verse instrument [65].

4.18. Statistical procedures

Plotted graphs were statistically analysed by either one-way ANOVA or Student’s t-test with GraphPad Prism 8.0 software. Data were presented statistically as mean ± SD. P < 0.05 is considered significant.

Funding

This research received no external funding.

CRediT authorship contribution statement

Chirantan Majumder: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Anirban Manna: Data curation. Satyajit Halder: Methodology, Data curation. Somesh Roy: Methodology, Data curation. Subhash C Mandal: Writing – review & editing. Kuladip Jana: Writing – review & editing. Mahadeb Pal: Writing – review & editing, Writing – original draft, Visualization, Project administration, Conceptualization.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Data availability

Data will be made available on request.