Introduction
Breast cancer is a leading cause of cancer-related mortalities globally. Over 266,000 new breast cancer cases are projected to occur solely in the United States in 2018 accounting for over 40,000 deaths [
1]. In the United Kingdom, breast cancer is the most common type of cancer in women, with nearly 53,700 new cases in 2013 and a one in eight estimated lifetime risk of diagnosis [
2]. Radiotherapy is a primary treatment modality for breast cancer patients. This approach uses the fractionated delivery of high-energy X-ray beams to generate highly reactive free radicals in target tumour tissue. This causes DNA damage via lipid peroxidation or oxidative cellular respiration. Radiation-induced damage activates several signal transduction pathways whose primary role is to detect genomic injury leading to cell cycle arrest and DNA repair. In the event of substantial damage, the endogenous apoptotic machinery of the cells is triggered to inhibit further replication of the damaged DNA [
3]. Radiotherapy also causes damage in healthy cells and can potentially trigger new cancer-initiating DNA mutations in local tissue. Therapeutic selectivity is, therefore, a vital issue in cancer therapy, and an ideal anticancer agent should be toxic to cancerous cells but exert minimal toxicity in healthy cells.
Watercress (
Nasturtium officinale) belongs to the family of Brassicaceae together with broccoli, brussels sprouts and kale. Epidemiological studies suggest a link between the consumption of Brassica vegetables and a reduced risk for many types of cancers [
4] including breast cancer [
5,
6]. Watercress has a complex phytochemical profile characterised by high amounts of carotenoids, flavonols and glucosinolates [
7] and is the main dietary source of phenethyl isothiocyanate (PEITC). Crude extracts of watercress have been shown to demonstrate strong antioxidant capacity in vitro [
8,
9] and have been associated with the inhibition of the three stages of carcinogenesis: initiation, proliferation and metastasis in in vitro cancer cell models [
10‐
12].
PEITC has been extensively shown to have direct anticancer effects in in vitro cancer models. PEITC exists in watercress as gluconasturtiin before mastication of the leaves, which exposes the parent compound to myrosinase resulting in the production of PEITC. It causes cell cycle arrest and mitochondrial damage in a wide variety of cell lines and it is a potent inducer of apoptosis [
13‐
16]. Combined treatment of cancer cells with PEITC and established chemotherapeutic agents such as cisplatin and doxorubicin potentiates their cancer-killing properties [
17,
18]. These findings support the potential of PEITC to be used as an adjuvant treatment during radiotherapy in cancer patients. Due to its highly electrophilic nature PEITC reacts with cellular thiols such as glutathione, the major intracellular antioxidant, depleting the cells of their antioxidant content and impacting cell survival [
13,
19,
20]. As radiotherapy works primarily by inducing DNA damage through the formation of free radicals, the ability of PEITC to deplete the radical scavenger glutathione is likely to contribute to its radio-sensitising properties.
Metabolic regulation is a determining factor of the cell growth machinery and cancer cells have adapted to several oncogenic signals to modify their metabolic phenotype to support their needs for growth, survival and malignant transformation [
21]. To our knowledge, limited work has been performed on the effects of isothiocyanates or of crude watercress extract on cancer cell energetics and metabolism. In this work, a metabolic profiling approach has been used to study the biochemical response of MCF-7 breast cancer cells and immortalised but non-tumorigenic MCF-10A cells to increasing doses of watercress extract (lacking PEITC) and PEITC alone. The impact of PEITC or watercress extract on the biomolecular events exerted by X-ray irradiation exposure was then investigated in these breast cells. Combining high-resolution metabolic phenotyping with measures of cell viability and DNA damage enables the radio-sensitising or radio-protective potential of watercress and its components to be evaluated.
Materials and methods
Cell culture
The MCF-7 human breast adenocarcinoma cell line was purchased from the American Type Culture Collection (ATCC) (LGC standards, Middlesex, UK). Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% (v/v) foetal bovine serum, 2 mM glutamine, 50 U/ml penicillin and 50 U/ml streptomycin and 1% non-essential amino acids. The MCF-10A, non-tumorigenic breast epithelial cell line was purchased from ATCC (LGC standards, Middlesex, UK). Cells were maintained in Ham’s F12:DMEM (1:1), 20 ng/ml epidermal growth factor (EGF) (PeproTech, London, UK), 0.1 µg/ml cholera toxin, 10 µg/ml insulin, 500 ng/ml hydrocortisone, 5% horse serum and 50 U/ml penicillin and 50 U/ml streptomycin.
Treatments and irradiation
For the watercress extract, fresh watercress samples were obtained directly from Vitacress Salads Ltd. (Andover, UK). Samples were snap frozen in liquid nitrogen and stored at − 80 °C. 2 g of leaf and 2 g of stem were weighed and placed in a 20 ml syringe that had the plunger removed and a circular 25 mm glass microfiber filter placed at the bottom. The syringe was then placed inside a 50 ml centrifuge tube without the lid and centrifuged at 1500
g for 30 min to collect the extract. This crude watercress extract was then filtered through a 0.22 µm filter and used in the cultures. Phytochemical characterisation of the watercress extract has previously been published [
7]. As PEITC is produced after consumption following exposure to myrosinase, it is absent from the watercress extract. To examine the metabolic effects of PEITC, 30 mM of PEITC was made up in DMSO fresh on the day of use. MCF-7 and MCF-10A cells were seeded at 1 × 10
5 cells per well into six well plates and treated at 80% confluence. Cells were exposed to the watercress extract at 6.25, 12.5, 25 and 50 µl/ml and PEITC at 5, 10, 20, 30 µM for 24 h. Following the watercress extract/PEITC treatment period, the cells were exposed to 5 Gy X-ray radiation using an orthovoltage X-ray unit (Gulmay Medical D3225, Xstrahl, UK). The irradiator was at a stable distance from the cell culture plates and the irradiator field was approximately 20 × 20 cm. The cell culture plates were placed in the centre of the irradiation field. Following radiation treatment cells were returned in the incubator and were allowed to rest for 1 h. The cells were then collected and used in the experiments.
Cell proliferation and viability
DAPI staining
For the determination of cell proliferation MCF-7 and MCF-10A cells were seeded in 96-well microplates at 5 × 103 cells per well and incubated at 37 °C with 5% CO2 and 95% humidity for 24 h. Cells were exposed to the respective treatments and then permeabilized with 100 µl of ice-cold methanol for 5 min at room temperature. Methanol was removed and the plates were allowed to air-dry for 15 min in a hood, followed by addition of 100 µl of DAPI in PBS (70 µl of 30 mM DAPI stock solution in 10.43 ml of PBS). Cells were incubated in the dark for 30 min at 37 °C and absorption was measured using GENios microplate reader (TECAN Group Ltd., Mannedorf, Switzerland) with absorbance at 340 nm and emission at 465 nm. The experiment was performed in triplicate with three technical replicates per experiment.
MTT assay
Cell viability was assessed using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide]-based in vitro toxicology assay kit (Sigma–Aldrich, Dorset, UK) according to the manufacturer’s instructions, in the case of the irradiation experiments. The experiment was performed in triplicate with three technical replicates per experiment.
The metabolic profiles of MCF-7 and MCF-10A cells were measured using 1H NMR spectroscopy. Cells were seeded and treated as described above. Media was transferred into Eppendorf tubes and cells on the surface of the plate were washed twice using 1 ml of cold (4 °C) PBS and were quenched using 1 ml of ice-cold methanol (maintained on dry ice). Cells were allowed to lyze for 2 min and were detached from the plate using a cell scraper and transferred into an Eppendorf tube. Methanol quenching was repeated to maximize metabolite recovery. A vacuum concentrator (SpeedVac) was used to dry down the cell suspensions before reconstitution in 80 µl of phosphate buffer (pH 7.4) in 100% deuterium oxide containing 1 mM of the internal standard, 3-(trimethylsilyl)-[2,2,3,3,−2H4]-propionic acid (TSP).
For every sample, a standard one-dimensional NMR spectrum was acquired using a 600 MHz Bruker NMR spectrometer, with water peak suppression using a standard pulse sequence [recycle delay (RD)-90°-t1–90°-tm-90°-acquire free induction decay (FID)]. For each spectrum, 256 scans and 8 dummy scans were obtained, collected in 64K data points with a spectral width of 12.001 ppm. 1H NMR spectra were manually corrected for phase and baseline distortions and referenced to the TSP singlet at δ 0.0. Spectra were digitised using in-house Matlab (version R2016a, The Mathworks, Inc.; Natwick, MA) scripts. Metabolites were identified using an in-house database of standards. Multivariate modelling, including principal component analysis (PCA) and orthogonal projections to latent structure discriminant analysis (OPLS-DA), were performed using in-house scripts in Matlab.
Cell cycle
Watercress extract and PEITC-treated MCF-7 and MCF-10A cells were collected by centrifugation and were then fixed in 70% (v/v) fresh ice-cold ethanol. The samples were then stored at − 20 °C until analysis. For the analysis, samples were centrifuged and the cell pellets were resuspended in 200 µl of PBS and 25 µl of 1 mg/ml RNAse added to the suspensions. The samples were incubated at 37 °C for 30 min and 2.5 µl of 400 µg/ml of propidium iodide dye were added to the cells which were then incubated for a further 30 min at room temperature under dark conditions. The final volume of the cell suspensions was adjusted to 600 µl with PBS. Cellular DNA content of 15,000 cells was quantified via flow cytometry. The flow cytometry analysis was performed using the FL2 channel on a BD Accuri™ C6 flow cytometer (Germany). Data analysis was facilitated using the Flow Jo software (version 7.6, Tree star Inc, Oregon, USA). Cell cycle progression was evaluated accounting for the percentage of all cells in each of the phases (G1, S, G2/M).
Comet assay
The Comet assay was used for the measurement of DNA strand breaks in single cells. Treated cell suspensions were adjusted to a concentration of 1 × 106 cells/ml and 50 µl were resuspended in 85 µl of warm low-melting point agarose (0.85% w/v) and applied on Trevigen Comet slides. The slides were allowed to solidify at 4 °C for 10 min. The slides were then transferred into a staining jar, lysis buffer was added (2.5 M NaCl, 0.1 M EDTA, 0.01 M Tris and 1% (v/v) Triton X—added just prior to use—pH 10), and the cells were lysed for 1 h at 4 °C. Following lysis of the cells, the slides were placed in a horizontal electrophoresis tank and incubated for 20 min in alkaline buffer (0.3 M NaOH, 1 mM EDTA—pH 13) at 4 °C in dark conditions. Subsequent to DNA unwinding, electrophoresis was carried out at 26 V, 300 mA for 20 min at 4 °C. Immediately after electrophoresis, the slides were washed in neutralising buffer (0.4 M Tris—pH 7.5) three times for 5 min. Slides were stained with ethidium bromide (20 µl/ml) and DNA migration from the nucleus was visualised with a fluorescence microscope (Olympus BX51). The computer-based image analysis software, Komet 4.0 (Andor Technology, South Windsor, CT) was used to calculate % tail DNA, the proportion of DNA migrated from the head to the tail of the comet. The mean value from 75 randomly scored cells was taken as an index of damage for each replicate well.
Discussion
This study combined in vitro experiments with high-resolution metabolic phenotyping to show that watercress and a derivative of watercress, PEITC, can enhance the outcomes of radiotherapy via different molecular mechanisms. The protective effect of watercress in healthy cells observed here is unlikely to be a result of PEITC or any other isothiocyanate since they are not present in the extract, as a result of the high volatility of these compounds as well as the snap freezing of the plant material, which inactivates the myrosinase enzyme that converts the glucosinolate precursor to the reactive isothiocyanate. As such, PEITC was absent from the watercress extract and its effects were studied independently. Both the extract and PEITC were found to modulate the metabolome of healthy and cancerous breast cells, influencing phospholipid and amino acid metabolism as well as modulating the energy and antioxidant status of these cells. These perturbations occurred alongside cell cycle arrest, DNA damage and compromised cell viability in cancerous cells. Subsequently, it was shown that through its ability to deplete glutathione, PEITC sensitizes breast cancer cells to radiation-induced damage, whereas the watercress extract protected healthy cells from IR toxicity by increasing intracellular glutathione.
Baseline metabolic variation was clear between the cancerous MCF-7 cell line and the non-transformed MCF-10A cell line. Greater lactate content in the cancerous cells reflects a higher rate of glycolysis in these cells for energy generation, consistent with the Warburg effect. MCF-7 cells also contained greater amounts of glutamine, an amino acid required for protein and nucleotide synthesis. Similarly, phosphocholine was present in higher amounts in the MCF-7 cells, relative to the non-transformed cells, and likely relates to the higher proliferation rate in these cells and the increased synthesis of new cellular membranes, reflected in the higher proportion of cells in the S phase of the cell cycle compared with the MCF-10A cells. The cancerous cells also contained greater amounts of glutathione compared to the healthy cells. Glutathione serves as an intracellular antioxidant and its increased abundance facilitates the maintenance of appropriate cellular redox status by keeping the amount of ROS at a level that enables cell proliferation and successful progression through survival pathways as a result of post-translational modifications [
22,
23]. Conversely, if the levels of ROS become extremely high this causes detrimental oxidative stress via macromolecular damage, senescence [
24] and loss of mitochondrial membrane potential leading to apoptosis [
25]; a collection of events that can have lethal effects on cells. To counter the outcomes of oxidative stress, cancer cells increase their antioxidant content (mainly glutathione), limiting the accumulation of ROS at excessively high levels preventing irreparable damage [
26]. Naturally, proliferating cells acquire oncogenic mutations, which favour anomalous energy metabolism and protein translation leading to aberrantly increased ROS [
27]. Through further mutations and adjustments, cancer cells tightly orchestrate the cycling of ROS and antioxidant production in a manner that permits cell survival and maintenance of ROS at moderate quantities.
Radiotherapy is an important treatment modality in breast cancer. This approach generates reactive free radicals, which damage DNA, ultimately resulting in cell death. Cancer cells posses several aberrant signalling pathways that can result in drug resistance or failure of therapeutic outcomes. Current research suggests that combination therapy can kill cancer cells more efficiently via diverse mechanisms simultaneously [
28]. Isothiocyanates, such as PEITC, have a range of cellular targets for cancer-related outcomes, including cell cycle arrest, apoptosis, and anti-angiogenic effects [
28]. As such, PEITC and its dietary source watercress are an attractive option for combinatorial therapeutic methods.
In the cancerous MCF-7 cells, IR caused DNA damage that resulted in G2 cell cycle arrest but there was no significant impact on cell survival suggesting a potential resistance of these cells to IR killing. Our results are consistent with those of Jänicke et al. [
29] who observed the same cell cycle arrest and failure of IR to activate the mitochondrial intrinsic apoptosis pathway. However, pre-treatment with PEITC resulted in significant G1 arrest parallel to increased DNA damage and a significant loss of cell viability. This is likely mediated by the ability of PEITC to induce p53 activity in MCF-7 cells [
30], which is a potent regulator of G1 cell cycle arrest. PEITC can also induce apoptosis from the mitochondria in breast cancer cells by caspase activation, as well as changes in the Bac/Bcl-2 ratio following the release of cytochrome
c, all significant elements of the intrinsic apoptotic pathway [
31]. In contrast, lower DNA damage was seen in the MCF-10A cells following PEITC and IR exposure.
Differences in glutathione content between MCF-7 and MCF-10A cells may contribute to the variation seen in response to PEITC and IR. Consistent with other studies, treatment of MCF-7 cells with IR-depleted intracellular glutathione [
32‐
34]. In contrast, MCF-10A cells responded to IR induced stress by increasing their glutathione content. IR generates ROS, which are quenched in part through the glutathione response reducing the potential of ROS to exert oxidative DNA damage. Elevations in intracellular glutathione in MCF-10A cells can be considered part of a protective response by upregulating the metabolic antioxidant capacity of these cells. This may explain their ability to better recover from IR-induced damage compared to MCF-7 cells and may explain the lower DNA damage observed in the healthy cells in this study.
PEITC appears to induce a biphasic response in the glutathione abundance of MCF-7 cells, with increased concentrations at low doses and depletion at the two higher doses. The ability of isothiocyanates to act as both pro-oxidants and indirect antioxidants may explain these dose-dependent fluctuations. Prolonged exposure to low isothiocyanate concentrations can induce phase II enzymes that regulate antioxidant gene expression [
35], increasing glutathione synthesis and abundance. However, at higher doses PEITC depletes cells of glutathione through sustained intracellular conjugation and efflux [
19,
20]. Glutathione depletion accompanied by compromised mitochondrial function ultimately results in excessive oxidative stress, as demonstrated by increased DNA damage with higher PEITC exposure and may explain the observed cell cycle arrest and cell cytotoxicity in MCF-7 cells treated with PEITC. Thus, the ability of PEITC to deplete glutathione availability sensitizes the MCF-7 cells to IR-induced damage resulting in G1 cell cycle arrest, greater DNA damage and reduced cell viability.
Interestingly, PEITC did not deplete MCF-10A cells of glutathione and these cells were also less sensitive to PEITC-induced DNA damage. PEITC has previously been shown to selectively kill cancer cells over non-tumorigenic cell lines due to their lower antioxidant status [
36‐
38]. When MCF-10A cells were exposed to IR and PEITC (10 µM), glutathione was depleted as it scavenges IR-derived ROS and to compensate for this loss glutathione synthesis was up-regulated. This metabolic adaption may explain the lower DNA damage seen in MCF-10A cells compared to MCF-7 cells following IR.
Cellular membranes are a primary target of IR due to the impact ROS can have on lipid bilayers, of which phosphocholine is a main constituent. Following IR, phosphocholine was increased in the MCF-10A cells. This increase may reflect the efforts of the cell to maintain membrane integrity, which is violated by ROS produced during IR. This is consistent the greater abundance of glutathione in MCF-10A cells and their greater resistance against oxidative damage. Conversely, phosphocholine was reduced in the MCF-7 cells in response to IR. Decreases in phosphocholine have been observed in tissues after chemotherapy and radiation treatment and have been correlated with positive therapy outcomes [
32,
39‐
41].
In both cell types, the watercress extract increased the cellular glutathione content. This is likely to be a result of the complex mixture of compounds in the watercress extract such as phenolics and flavonoids with proven antioxidant properties. Flavonoids increase the expression of γ-glutamylcysteine synthetase, which is directly proportional to glutathione abundance [
42]. Watercress is also a rich source of folate [
43] which can be used in one-carbon metabolism pathway, adding to the cellular glutathione pool. Several in vitro studies have shown the anti-genotoxic properties of watercress extracts [
10,
44]. In the MCF-10A cells, pre-treatment with the watercress extract had a protective effect against IR exposure, evidenced by lower DNA damage compared to MCF-7 cells additional to increased glutathione content. This suggests enhanced antioxidant activity and hence a protective effect. The presence of additional antioxidant compounds in watercress help to preserve the glutathione content of the cells.
PEITC and watercress strongly interact with the metabolism of amino acids in both cell lines. PEITC at the higher doses, but not watercress, induced a strikingly selective increase in the pool of amino acids in MCF-7 cells, but not in the MCF-10A cells. These effects were generally maintained after IR. Rapidly dividing cells rely heavily on the maintenance of their biosynthetic potential as well as redox status for survival. Continuous shuttling of carbon molecules through amino acids such as glycine, methionine, threonine and serine, in the one-carbon metabolism pathway, which has a central role in cell proliferation and cancer progression, ensures the availability of the building blocks necessary for the construction of new cellular components. This also sustains the formation of reducing power compounds for redox balance. Accumulation of amino acids in the PEITC-treated MCF-7 cells is suggestive of a blockage in one-carbon metabolism pathway resulting in the inability of these cells to maintain their needs in macromolecules necessary for proliferation, increasing their susceptibility to IR damage.
Amino acids are key components for the protein translational requirements of cancer cells. Elevated mRNA translation is a key driver of carcinogenesis and PEITC has been recently shown to increase eIF2a phosphorylation and inhibit mTORC1 activity resulting in inhibition of translation in MCF-7 cells and in B cells from chronic lymphocytic leukaemia [
45]. Further investigation is needed to understand if the accumulation of amino acids seen in the PEITC-treated cells is a cause or consequence of translational inhibition.
mTORC1 is master regulator of protein translation, which is a known target of PEITC [
46,
47]. PEITC causes mitochondrial damage that essentially increases the AMP/ATP ratio due to energy depletion, which in turn activates AMPK. AMPK acts upstream of mTORC1, ultimately inactivating it and suppressing translation. MCF-10A cells have a lower basal mTORC1 activity as compared to the MCF-7 cells [
48] suggesting that PEITC has a stronger affinity for cells with increased rates of translation.
The chemopreventive potential of watercress and its effects against oxidative stress have been investigated in a number of in vivo studies. Although pharmacokinetic data for PEITC following the ingestion of watercress is limited, Ji et al. reported a mean maximal PEITC plasma concentration (
Cmax) of 929 nM following the consumption of 100 g of watercress [
49] while Alwi et al. [
50] reported a
Cmax of 297 nM with 80 g watercress. In the study by Alwi et al., this single 80 g of portion of watercress was sufficient to reduce the phosphorylation of 4E-binding protein 1, a key factor in angiogenesis. In another study, a single 50 g portion of watercress effectively attenuated the immunoreactivity of a proinflammatory cytokine macrophage migration inhibitory factor [
51]. Daily intake of 85 g of watercress for 8 weeks has been shown to decrease DNA damage in peripheral blood lymphocytes and lipid peroxidation [
8,
9] These in vivo studies demonstrate that dietary intake of watercress is sufficient to modulate anticarcinogenic pathways.