Dibenzyl trisulfide induces caspase-independent death and lysosomal membrane permeabilization of triple-negative breast cancer cells

Abstract

The Petiveria alliacea L. (P. alliacea) plant is traditionally used in folklore medicine throughout topical regions of the world to treat arthritis, asthma, and cancer. Dibenzyl trisulfide (DTS) is one of the active ingredients within the P. alliacea plant. Triple-negative breast cancer (TNBC) is associated with a poor prognosis, particularly among women of West African ancestry, due in part to limited effective therapy. Though potent anticancer actions of DTS have been reported in a TNBC cell line, the mechanism of DTS-mediated cytotoxicity and cell death remains ill-defined. In the current study, we show that DTS exhibits cytotoxicity in a panel of triple-negative breast cancer (TNBC) cells derived from patients of European and West African ancestry. We found that DTS inhibits proliferation and migration of CRL-2335 cells derived from a patient of West African ancestry. DTS induces the expression of pro-apoptotic genes BAK1, GADD45a, and LTA in CRL2335 cells though it primarily promotes caspase-independent CRL-2335 cell death. DTS also promotes destabilization of the lysosomal membrane resulting in cathepsin B release in CRL-2335 cells. Finally, Kaplan-Meier survival curves reveal that higher expression of BAK1 and LTA in tumors from patients with TNBC is associated with longer relapse-free survival. Collectively, our data suggest that DTS confers promising antitumor efficacy in TNBC, in part, via lysosomal-mediated, caspase-independent cell death to warrant furthering its development as an anticancer agent.

1. Introduction

Breast cancer is the second leading cause of cancer mortality among women. In the US alone, breast cancer causes over 40,000 deaths with more than 280,000 new cases every year [1]. Women of West African ancestry are more than twice as likely as those from other ethnicities to die from breast cancer [2]. This breast cancer survival disparity is due, in part, to the increased likelihood that women of West African ancestry are more than twice as likely as women of other ancestries to develop aggressive triple-negative breast cancer (TNBC) [3]. TNBC is characterized by tumors that lack expression of the estrogen receptor (ER), the progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) amplification [4]. While targeted therapy options are available for patients with tumors expressing ER and/or HER2 amplification, patients with TNBC do not benefit from such therapies. Consequently, TNBC carries a poor prognosis.

Conventional anticancer therapies often entail the use of synthetic small molecules or monoclonal antibodies [5]. Studies by the World Health Organization and others have reported that the use of herbal medicines for various health challenges is rapidly expanding across the world [6-8]. This rapid global expansion is due, in part, to growing concerns regarding adverse effects associated with pharmaceuticals. Herbal remedies or medicines are often taken in tandem with prescribed pharmaceuticals, often unbeknownst to the prescribing physician [9, 10]. For example, 80% of Jamaican cancer patients have been reported to use medicinal plants along with prescription remedies, with limited physician awareness [11]. As a result, many of the adverse effects seen with pharmaceuticals could be due to alterations in metabolism as seen with drug-herbal interactions [12]. Our understanding of the mechanisms by which these plant isolates confer their actions is often limited. We therefore aimed to delineate the mechanism of anticancer action for dibenzyl trisulfide (DTS, Figure 1A).

Figure 1. Impact of Dibenzyl Trisulfide (DTS) on the viability of triple-negative breast cancer cells.

A) Structure of dibenzyl trisulfide (DTS), an active ingredient of P. alliacea. B-C) Cells were treated with DMSO or DTS (100 pM −100μM) for 48 h or 72 h respectively and the viability of the cells was examined using the Alamar Blue assay. Data are reported as the mean of at least triplicate values. Bars, SEM

DTS is isolated from a perennial shrub called Petiveria alliacea L. (P. alliacea) found growing in tropical regions such as the Caribbean [13]. This plant is traditionally prepared for internal use as a tea, liquid, or paste and for treatment of a wide range of conditions such as arthritis, asthma, cancer, colds, headaches, and snakebites [14]. DTS is one of the active phytochemicals isolated from this plant with potent activity against a variety of cancers including prostate, ovarian, and leukemia [14-16]. DTS also confers activity against the MDA-MB-231 TNBC breast cancer cell line derived from a patient of European ancestry [17, 18]. To date, such a study has not been undertaken using preclinical models derived from patients of West African ancestry. Importantly, a clinical trial (ClinicalTrials.gov Identifier: NCT04113096) is underway to investigate the effects of DTS in stage IV breast cancer patients in Jamaica where more than 90% of residents are of West African ancestry. While DTS has been shown to modulate mitogen activation protein kinase and ribosomal protein S6 kinase alpha 1 activity, the mechanism(s) by which DTS confers its anticancer actions remains ill-defined [18].

In this study, we investigated the anticancer and chemopreventive potential of one of the active components of P. alliacea, DTS. DTS and other P. alliacea extracts were screened against a panel of TNBC cell lines derived from patients of West African and European ancestry. We then evaluated the anticancer and cell death mechanisms of DTS in a TNBC cell line derived from a patient of West African ancestry that displayed the highest sensitivity to DTS. To the best of our knowledge, this is the first study investigating the potential mechanisms of DTS–mediated death in TNBC cells derived from patients of West African ancestry.

2. Materials and Methods

2.1. Cell culture and reagents

MDA-MB-468, MDA-MB-157, MDA-MB-436, MDA-MB-231 and CRL-2335 human breast cancer cell lines were obtained from the American Type Culture Collection (ATCC). The cells were cultured as previously described [19]. MDA-MB-231 and MDA-MB-436 cells are derived from patients of European ancestry while the MDA-MB-468, MDA-MB-157 and CRL-2335 cells are derived from patients of West African ancestry. Purified and refined DTS was purchased from International Laboratories (San Francisco, CA, USA). Stock solutions of inhibitors and agents were dissolved in dimethyl sulfoxide (DMSO) and stored at −20°C until use. Cells were exposed to < 0.1% DMSO to avoid impacting cell behavior.

2.2. Plant collection and preparation

The P. alliacea plant was obtained from the botanical garden at the University of the West Indies (UWI) Mona campus, identified and deposited in the UWI Herbarium. Standardized aqueous and 65% ethanolic extracts were prepared as previously described [14] from the dried and fresh plants with voucher numbers # 35931 and # 36361. All extracts were prepared following ethnomedical practices [20, 21]. Briefly, the aqueous extracts were prepared by decocting the whole plant (leaf, root and stem) for 20 minutes in deionized water (1 g/100 mL). The resulting extract was filtered, freeze-dried and stored at −20°C. The 65% ethanolic extracts were prepared by macerating the plant (leaf, root and stem) in 65% ethanol (1g/ 10mL) for 10 d and the resulting tincture was dried using a stream of nitrogen gas and stored at −20°C.

2.2. Cell Viability Assay

We evaluated DTS-mediated cytotoxicity using the Alamar Blue assay in a panel of TNBC cell lines as previously described [19]. Briefly, cells were plated in 96-well plates at their appropriate densities in a total volume of 100 μL. After 24 h of incubation, cells were treated with medium containing DMSO (0.01%), P. alliacea extracts (100 pM-100 mM) or DTS for 48 h or 72 h. Alamar Blue dye (10%) mixed with fresh medium was then added to the wells followed by an additional 4–5 h incubation. Cytotoxicity was determined using an FLx800 microplate spectrofluorometer (BioTek Instruments, USA).

2.3. Colony Formation and Wound Healing Assays

For the colony formation assay, CRL-2335 human breast cancer cells were plated at a density of 2000 cells per well in a 6-well plate. The following day, cells were treated with either DMSO or DTS (100 nM-50 μM) for 48 h. The treatment was removed, and cells were allowed to grow for 2 weeks before the cells were fixed with 10% Formalin and stained with crystal violet solution. Colonies were imaged and counted using the open-source image processing software ImageJ (National Institutes of Health, Bethesda, MD). For the wound healing assay, CRL-2335 cells were plated in 24-well plates containing Ibidi culture inserts (to create the wound) at a density 3x10⁵ cells/ml and allowed to recover overnight. The cells were then treated with DMSO, or 10 μM DTS, or 25 μM DTS for 48 h. Images were captured on an Olympus IX-71 microscope and quantified using SPOT software (Olympus Life Sciences Solutions, Waltham, MA).

2.4. Cell Morphology and Apoptosis Determination

The Annexin V/propidium iodide assay was used to detect apoptosis as previously described [22]. Briefly, CRL-2335 cells were treated with medium containing DTS (10 μM-50 μM) or DMSO for 48 h. In some studies, CRL-2335 cells were pretreated for 1 h with z-VAD-fmk (z-VAD, 100 μM) before treatment with DTS. Cell morphology was examined under relief contrast microscopy at different time points to assess the presence of apoptotic blebs before data acquisition using flow cytometry. Data were acquired using a MACSQuant Analyzer 10 (Miltenyi Biotec Inc, Auburn, CA) and analyzed using FlowJo (v9.7.5, TreeStar, Inc, Ashland, OR).

2.5. Real-time quantitative RT-PCR (qPCR) analysis

RNA was extracted from breast cancer cells after specified treatments, and cDNA synthesis was performed according to the manufacturer’s instructions and in accordance with a method previously described [19]. The primers for the GAPDH, BAK-1, GADD45A, and LTA, genes were obtained from Qiagen (Germantown, MD). Relative fold changes in gene expression were calculated using the 2^−ΔΔCt method.

2.6. Immunoblotting

CRL-2335 cells were seeded at 1-3 × 10⁶ per plate (100 mm). After 24 h, cells were treated with medium containing DMSO, 10 μM DTS, or 50 μM DTS for 48 h. After the treatment, cells were harvested, and Western blot analysis was performed as described [19]. Briefly, proteins were resolved on 4%-12% sodium dodecyl sulfate polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were blocked before an overnight incubation at 4°C in 5% milk–based buffer with mouse primary antibodies against LTA (1:1000), and rabbit primary antibodies against BAK-1 (1:1000), GADD45A (1:1000), PARP (1:1000) and caspase-3 (1:1000). Membranes were incubated with the appropriate horseradish peroxidase–conjugated secondary antibodies (Cell Signaling Technology, Danvers, MA or Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h at the appropriate dilution before imaging.

2.7. Lysosomal Membrane Permeability (LMP) Assay

The Acridine Orange (AO, ImmunoChemistry Technologies, Bloomington, MN) method was used to analyze LMP as described previously [23, 24]. AO preferentially accumulates in the lysosomes due to photon trapping. At high concentrations (lysosomes), this dye emits red/orange fluorescence and green fluorescence at low concentrations (nucleus and cytoplasm) via blue light excitation. Briefly, CRL-2335 cells were seeded at 2 x 10⁵ in a 6-well culture plate and incubated overnight. The cells were then incubated in culture medium containing 0.1% DMSO or 10 μM DTS for 48 h. To ensure that the cells did not suffer any damage in the staining process, they were washed with cold PBS following treatment. PBS was replaced with cell culture medium containing 5 μM AO and counterstained with 5 μM Hoechst 33342 before incubating for 30 min at 37°C and examined with an EVOS cell imaging system (Invitrogen, USA).

2.8. Cathepsin B Activation

The fluorogenic substrate-based assay, Magic Red (Immunochemistry Technologies, Bloomington, MN), was used to detect Cathepsin B activity using the Magic Red MR-(RR)₂ reagent as previously described [23, 24]. Briefly, CRL-2335 were treated with 0.1% DMSO or 10 μM DTS for 48 h. The cells were then exposed to the cathepsin B substrate MR-(RR)2 for 45 min, rinsed twice with cold PBS, and examined using the fluorescence microscope. Additionally, cells were counterstained with Hoechst 33342 to detect nuclear morphology.

2.9. Statistical analysis

Data are reported as mean ± SEM or mean ± SD. Statistical significance was assessed using the one-way analysis of variance (ANOVA) with Tukey’s test, the Dunnett’s test, or the Tukey-Kramer multiple comparison test when evaluating three or more groups. An unpaired Student’s t test with Welch’s correction was used to compare two groups. Statistical analysis was performed using GraphPad InStat 3.0. Differences were considered significant at p < 0.05. Kaplan-Meier Plots were constructed using the KM plotter program and hazard ratio with 95% confidence intervals and log rank P values were calculated using gene expression of tumors from cohorts of TNBC patients in accordance with St. Gallen and PAM50 databases.

3. Results

3.1. DTS but not aqueous- or ethanol-based extracts from P. alliacea reduce TNBC cell viability.

To ascertain the cytotoxic response of TNBC cells to P. alliacea, we first exposed MDA-MB-468, MDA-MB-157, MDA-MB-436, MDA-MB-231, and CRL-2335 TNBC cells to aqueous- and ethanol-based P. alliacea extracts (100 pM to 100 μM). The percent of DTS per gram of crude extract was found to be at 0.2% and 0.06% in the ethanolic and aqueous extracts, respectively, of different parts of the P. alliacea plant and in undetectable levels in the fresh plant preparations using HPLC as well as LC-MS analyses [14]. We observed no significant cytotoxicity against any of the TNBC cells exposed to either of the extracts (Table 1; Supplementary figure 1). However, when these cells were exposed to DTS (100 pM-100 μM), a significant reduction in cell viability was observed following 48 h and 72 h of exposure (Table 1; Figure 1B-C). Notably, the CRL-2335 cell line is described as an acantholytic squamous cancer that is generally more aggressive and treatment-refractory compared to the other cell lines used in this study [25, 26]. Thus, we focused our investigation on the anticancer and death promoting actions of DTS in CRL-2335 cells.

Table 1.

Relative cell viability IC₅₀ values (M) following exposure to P. alliacea extracts or DTS

	MDA-MB-157	MDA-MB-468	CRL-2335	MDA-MB-436	MDA-MB-231
GHW (AE) 48h	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004
GHW (EE) 48h	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004
DTS 48h	1.022e-006	~ 0.01383	5.022e-006	~ 8.583e-006	2.482e-006
GHW (AE) 72h	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004
GHW (EE) 72h	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004	> 1.00e-004
DTS 72h	5.354e-007	~ 0.06508	7.327e-007	3.487e-006	7.233e-006

3.2. DTS inhibits proliferation and migration in TNBC cells.

We used a colony forming assay to determine whether DTS decreases proliferation of CRL-2335 cells. DTS exposure resulted in significantly fewer colonies compared to the control at concentrations below 500 nM (Figure 2). We performed a wound healing assay involving IBIDI inserts to determine whether DTS decreases CRL-2335 cell migration (Figure 3). Our data reveal that DTS suppresses cancer cell migration at 10 μM and completely blocks it at 25 μM. These data suggest DTS not only thwarts proliferation and migration but also TNBC cell invasion and metastases.

Figure 2. Colony Forming and Wound Healing Assay Analysis.

CRL-2335 Cells were treated and allowed to grow for 2 weeks before staining with crystal violet. Colonies were counted using Image J and quantified using GraphPad Prism version 9. Data are reported as the mean of triplicates. Bars, SD. Not significant (ns), *** P <0.001, **** P < 0.0001 as compared to DMSO treatment group.

Figure 3. Wound Healing Assay of CRL-2335 Breast Cancer Cells.

Cells were plated in IBIDI culture inserts and allowed to grow to confluency before being treated with DTS (10 μM and 25 μM). A) Images were captured using an Olympus IX-71 microscope at 0h, 24h, and 48h. B) Quantification of cell migration following given treatments was performed using GraphPad Prism version 9. Bars, SD ** P < 0.01, *** P <0.001, **** P < 0.0001, relative to the DMSO-treated group.

3.3. DTS promotes caspase-independent TNBC cell death.

We next sought to determine whether DTS induces TNBC cell death using the Annexin V-FITC assay and relief contrast microscopy. We show that DTS promotes early apoptotic cell death of CRL-2335 as evidenced by Annexin V positive cells (Figure 4A-C). Morphological imaging also revealed apoptotic bodies after exposure to DTS in a dose- and time-dependent manner in MDA-MB-468 and CRL-2335 TNBC cells (Figure 4C; Supplementary Figure 2). To further elucidate the mechanism of DTS-mediated cell death, we investigated the role of caspase activation in cells treated with DTS. Pretreatment with a pan-caspase inhibitor (Z-VAD-FMK) minimally rescued cells from DTS-mediated TNBC cell death (Figure 4A-C). Taken together, these data suggest DTS promotes cell death via caspase-independent mechanisms.

Figure 4. DTS induces caspase-independent CRL-2335 cell death.

A) Flow cytometry data of cells treated with either 0.01% DMSO, 50 μM DTS, 10 μM Staurosporine (STS, positive control), or DTS in combination with caspase inhibitor Z-VAD-FMK. (Z-VAD) B) Quantification of cells in early apoptosis in cells exposed to the treatments as described in A, ** P < 0.01 as compared to 0.01% DMSO, Bars, SD. Numbers enlarged in the lower right quadrant indicate the average percentage of cells in early apoptosis. C) Morphology of CRL-2335 breast cancer cells after DTS treatment. CRL-2335 cells were exposed to DMSO, DTS (1-50 μM) and STS (10 μM) for the indicated times. Images were captured using the Olympus IX71 microscope (X40 magnification).

3.4. DTS induces pro-apoptotic gene and protein expression in CRL2335 cells.

To gain further insight into the mechanism of DTS-mediated cell death, we treated CRL-2335 cells with 0.01% DMSO, 10 μM DTS, or 50 μM DTS and analyzed the expression of three proapoptotic genes: BAK-1, GADD45a, and LTA. BAK-1 promotes apoptosis and counteracts protection from apoptosis provided by Bcl-2 [27]. GADD45a plays a role in p38-mediated p53 activation [28]. LTA induces tumor necrosis factor receptor 1-dependent apoptosis and necroptosis [29]. Our studies revealed that DTS significantly induced BAK1, GADD45a, and LTA mRNA expression at 50 μM (Figure 5). We then sought to evaluate the relationship between BAK1, GADD45a, and LTA tumor expression from patients with TNBC and survival. Figure 6 shows that higher BAK1 and LTA expression levels in tumors from patients with TNBC are associated with increased relapse-free survival according to PAM50 and St. Gallen databases [30]. We did not observe this trend with GADD45a (data not shown). Though a similar trend in overall survival for these genes was also observed, it did not achieve statistical significance. The downstream cleavage of various cytoplasmic or nuclear substrates, including PARP, by effector caspases such as caspase 3 mark many of the morphological features of apoptotic cell death [31]. Thus, we sought to determine whether DTS impacted the cleavage of PARP and caspase 3 in CRL2335 cells. We also sought to evaluate the impact of DTS on BAK1, GADD45a, and LTA protein expression in CRL2335 cells. We found that DTS induced PARP cleavage, but only minimal and non-significant caspase 3 cleavage in CRL-2335 cells (Figure 7). Importantly, DTS significantly induced the expression of proapoptotic proteins BAK-1 and LTA but not GADD45a (Figure 7A-B) in CRL-2335 cells. These data suggest that DTS not only exhibits anticancer activity in TNBC but diminishes the potential for recurrence among patients by inducing caspase-independent cell death and inducing the expression of certain pro-apoptotic genes associated with increased relapse-free survival.

Figure 5. Proapoptotic gene expression in CRL-2335 cells following treatment with DTS.

Cells were exposed to DMSO or DTS for 48h before they were harvested, RNA extracted, and quantitative PCR analysis. *P < 0.05, ** P < 0.01 Bars, SD.

Figure 6. Kaplan-Meier survival curves correlating BAK1 and LTA gene tumor expression to relapse-free survival predicted among patients with triple negative breast cancer (TNBC).

(A-B) Relapse free survival (RFS) predicted among TNBC patients with tumors that express BAK1 as significantly higher than that among those with lower expression in accordance with the PAM50 and St. Gallen databases respectively. (C-D) RFS predicted among patients with tumors expressing LTA as significantly higher than that among those with lower expression in accordance with the PAM50 and St. Gallen databases respectively.

Figure 7. Apoptosis analysis of CRL-2335 cells.

A) Immunoblot determination of pro-apoptotic protein expression in CRL-2335 cells. B) Quantification of relative protein expression. *P < 0.05, ** P < 0.01, *** P <0.001, **** P < 0.0001. Bars, SEM.

3.5. DTS promotes lysosomal membrane permeabilization and cathepsin B release in CRL2335 cells

Because our data suggest DTS mediates cell death via mechanisms that don’t rely on caspases, we sought to determine whether DTS-mediated cell death was related to lysosomal membrane permeabilization. Lysosomes are organelles that mediate the degradation of intracellular macromolecules. Various stressors can induce lysosomal membrane permeabilization (LMP) which releases lysosomal contents such as cathepsins into the cytoplasm [32]. Furthermore, LMP-induced cathepsin release activates classical caspase-dependent and -independent apoptosis [33].

To delineate the role of DTS in mediating LMP and cathepsin activity, we analyzed lysosome and cathepsin activity in cells exposed to acridine orange (AO) and Magic Red (MR) respectively. We found that DTS caused LMP in CRL-2335 cells, as evidenced by a reduction in intact, punctate red fluorescent lysosomes (Figure 8A). Similarly, DTS promoted cathepsin B release, as evidenced by diffuse red fluorescent dye staining throughout the cell (Figure 8B).

Figure 8. DTS induces lysosomal membrane permeabilization and cathepsin B release in CRL-2335 cells.

CRL-2335 cells after treatment with DMSO (CTL) or DTS (10 μM) for 48 h followed by analysis using acridine orange (A) or the Magic Red (B) assays to measure LMP and cathepsin B release, respectively (X40 magnification).

4. Discussion

TNBC carries a poor prognosis due, in part, to the scarcity of efficacious, targeted therapy. Therefore, it is crucial to develop agents with the ability to effectively treat TNBC. While aqueous- and ethanol-based preparations of P. alliacea have been used to treat a variety of medical conditions including breast cancer, our data suggest that such preparations may be less efficacious due to the low content of DTS in the ethanolic and aqueous extracts of 0.2% and 0.06% respectively. The low concentration levels of DTS may explain the low bioactivity observed in the extracts. As DTS is insoluble in water, it is no surprise that the fresh plant extract, which has a higher water content than that of the dried plant, would possess even lower concentrations of DTS. Our study does suggest that DTS itself possesses promising anticancer actions due in part to its ability to confer caspase-independent death of TNBC cells.

The effectiveness of DTS as an anticancer agent has been thoroughly documented in human pancreatic, ovarian, prostate, lung, and breast cancer cell lines [17]. In particular, the anticancer activity of DTS has been screened in MCF7 and MDA-MB-231 breast cancer cell lines with the MDA-MB-231 cell line showing more sensitivity to DTS [34, 35]. MCF-7 cells are luminal since they express hormone receptors and MDA-MB-231 breast cancer cells are classified as TNBC; both are derived from patients of European ancestry [36]. Importantly, Phase I clinical trials (ClinicalTrials.gov Identifier: NCT04113096) are underway to investigate the effects of DTS in stage IV cancer patients, including those with breast cancer. While the P. alliacea plant extracts exhibited minimal efficacy in TNBC cells, the isolated phytochemical DTS showed significant activity in these cells irrespective of the ancestry of the patients from where they were derived (Figure 1). The CRL-2335 cells demonstrated promising antiproliferative and antimigration actions (Figures 2-3) to suggest it has potential to thwart metastases in TNBC. Furthermore, while some plant isolates require concentrations of at least 100-200 μM to display activity, DTS treatment resulted in IC₅₀ values less than 10 μM in most TNBC cells irrespective of ancestry (Table 1). Previously, DTS was shown to exhibit potent cytotoxicity in MDA-MB-231 cells and our data agree with these findings [18]. Notably, studies have shown that extracts of P. alliacea as well as DTS display no cytotoxic activity against the non-tumorigenic, human hepatocellular carcinoma cell line, HB 8065 (Hep G2) [37, 38], and this suggests that DTS is selectively cytotoxic to tumorigenic cells. It is important to note that cancer cells lack the complex cytochrome P450 systems found in preclinical mouse models and in humans crucial to converting the aqueous and ethanolic extracts to active metabolites. As a result, we cannot conclude that such extracts are ineffective in humans just because we were unable to detect activity in the TNBC cell lines used in the current study.

We previously found that DTS suppressed cytochrome P450 1 (CYP1) activity and binds to the CYP1A1 active site [39]. CYP1A1 has been shown to promote the bioconversion of procarcinogens such as benzo-A-pyrene into carcinogenic metabolites which promote the progression of breast cancer after activating the aryl hydrocarbon receptor signaling pathway [40]. Thus, DTS has potential use as a chemopreventive agent. Our data show that DTS suppresses CYP1A2 expression in MDA-MB-468 and CRL2335 cells (Supplementary figure 3). We also found DTS suppressed TNBC proliferation (Figure 2) and cell migration (Figure 3). These findings suggest DTS confers chemopreventive actions in addition to its anticancer actions in TNBC.

Apoptosis can occur via extrinsic or intrinsic pathways. Extrinsic apoptosis occurs through the activation of death receptors such as CD95 or TRAIL, while intrinsic apoptosis is initiated via a diverse array of non-receptor-mediated stimuli [41]. Diverse forms of cellular stress can activate the intrinsic-apoptotic mitochondrial pathway by inducing the formation of pores in the outer mitochondrial membrane by proapoptotic members of the Bcl-2 family of proteins [42, 43]. BAK1 and the protein it encodes belong to the Bcl-2 family members that promote mitochondrial outer membrane permeabilization (MOMP). MOMP releases pro-apoptotic components such as cytochrome c into the cytosol, initiating the caspase cascade to mediate apoptosis [44]. GADD45a is a DNA damage-inducible protein that plays a role in the cell’s response to DNA damage and in the maintenance of its genomic integrity [45]. Induced Gadd45a enables the dissociation of Bcl-2 family member Bim from microtubule-associated components which are then translocated to the mitochondria to promote MOMP, release cytochrome c, and activate the caspase cascade [46]. We found that 50 μM DTS induced BAK1, GADD45a, and LTA gene and protein expression though DTS-mediated induction of GADD45a protein expression did not reach statistical significance (Figures 5 and 7). While the St. Gallen database showed that higher GADD45a tumor expression correlated with increased relapse-free survival, this association was not significantly observed when we used the PAM50 database (data not shown). LTA or tumor necrosis factor-beta (TNF-β) is a pleiotropic cytokine that mediates a myriad of antiviral, inflammatory, and immunostimulatory responses [47-50]. LTA induces apoptosis extrinsically through the death receptor p55TNFR, resulting in the subsequent activation of caspases [51]. Furthermore, increased survival was found among patients with tumors that possessed increased expression of BAK and LTA (Figure 6). This suggests that these pro-apoptotic genes are important components to improving patient outcomes and DTS-mediated induction of these genes represents an important component to its mode of anticancer action.

From our results and due to the role of BAK1 and LTA genes in mediating caspase cascade activation, we initially theorized that the anticancer actions of DTS were dependent on the activation of caspases. However, when caspase activity was inhibited via the pan caspase inhibitor Z-VAD, DTS-mediated cell death was insignificantly inhibited (Figure 4). In contrast, Z-VAD pretreatment of cells significantly suppressed staurosporine-mediated cell death (data not shown). Staurosporine has been shown to induce caspase-dependent or caspase-independent apoptosis depending on cell context [52-54].

Other cellular components are involved in the cell death pathway. Poly(ADP-ribose) polymerases (PARPs) play pivotal roles in counteracting genotoxic stress and other forms of DNA damage [55]. PARP-1, the most abundantly expressed, is one of several known cellular substrates of caspases. Cleavage of PARP-1 via caspase 3, is a hallmark of apoptosis [56-58]. Our data show that DTS induces PARP in a dose-dependent manner (Figure 7) and cleaved PARP is expressed more robustly than cleaved caspase 3. However, studies show that cathepsins and TGF-β are involved in caspase-independent PARP-1 cleavage [59]. Consequently, the lack of caspase 3 cleavage may correlate with the DTS-mediated cathepsin B release resulting in cleaved PARP-1 fragments [60]. Additionally, we observed DTS-induced expression of proapoptotic proteins (Figure 7A-B). Notably, BAK exists as an inactive monomer. Upon apoptotic stimuli, activated BAK undergoes oligomerization by forming Bcl-2 homology 3 (BH3): groove homodimers that represent the basic stable oligomeric unit [61]. For this reason, the ~50 kDa band often observed in our Western blots correlates with the dimerized form of Bak (25 kDa, anti-bak; Cell Signaling Technology, Inc.) resulting from apoptotic stimuli.

Apoptosis is a specific and specialized form of cell death characterized by distinct morphological characteristics such as cell shrinkage, pyknosis, blebbing, and formation of apoptotic bodies [62]. We observed cellular bodies closely resembling advanced apoptosis at high concentrations of DTS (Figure 4, supplementary figure 2). Furthermore, we show DTS induces LMP (Figure 8A) and cathepsin B release (Figure 8B). Interestingly, LMP is a hallmark of not only apoptosis and necrosis, but necroptotic and autophagic cell death [63-65]. Taken together, these data suggest that DTS induces TNBC cell death primarily via caspase-independent mechanisms. Caspase-independent cell death carries two advantages over caspase-dependent cell death. Firstly, caspase-independent death prevents the caspase-dependent oncogenic effects of apoptosis. Secondly, caspase-independent death can activate anti-tumor immunity, suggesting that research focused on the signals that accompany a cell during its death rather than the execution of cell death may be prove to be beneficial [66].

In conclusion, DTS represents a promising agent with potential to treat refractory forms of breast cancer including TNBC which disproportionately impacts women of West African ancestry. With the continued growth in popularity, the potential for interactions between medicinal plant products and pharmaceutical drugs is a growing health concern for many governments and regulatory bodies worldwide. Though most of our results were obtained after 48 h or 72h of treatment, it is recommended that patients use P. alliacea twice daily for an extended period [67], to allow sufficient levels to accumulate in the plasma. A thorough understanding of the pharmacokinetics of DTS, including any activated metabolites, would therefore be beneficial. It is possible that factors such as drug metabolizing enzymes (e.g., CYP1A) influence the efficacy of this phytochemical over time. It is crucial to further elucidate the mode of DTS-induced cell death and efficacy using in vivo models of TNBC as such preclinical studies will help to not only confirm and validate the results of our study but will provide a basis for further clinical evaluation of DTS as an agent to enhance clinical outcomes for patients with refractory forms of cancer.

Abbreviations

AE

Aqueous Extract

AO

Acridine Orange

BAK1

BCL2 Antagonist/Killer 1

CTL

Control

CYP1A1

Cytochrome P450 1A1 gene

CYP1A2

Cytochrome P450 1A2 gene

DMSO

Dimethyl sulfoxide

DTS

Dibenzyl Trisulfide

EE

Ethanolic Extract

ER

Estrogen Receptor

GADD45a

Growth Arrest and DNA damage Inducible Alpha

GHW

Guinea Hen Weed

HER2

Human Epidermal Growth Factor 2

LMP

Lysosomal Membrane Permeabilization

LTA

Lymphotoxin-alpha

mRNA

Messenger RNA

OS

Overall Survival

P. alliacea

Petiveria alliacea L.

PARP

Poly(ADP-ribose) polymerase

PR

Progesterone Receptor

RSF

Relapse-Free Survival

STS

Staurosporine

TNBC

Triple Negative Breast Cancer

Z-VAD

Z-VAD-FMK