Background
Breast cancer is the most common cancer type in women, accounting for 29% of all new cancer diagnoses, and is the second most common form of cancer-related death in women worldwide [
1,
2]. The presence of metastatic disease in breast cancer patients is the single most important factor affecting patient mortality. For this reason, attempts to control the metastatic spread of tumours remains a key target for the successful treatment of this disease [
3]. However, the metastatic cascade is a complex multi-step process that is influenced by a host of factors, including the interaction between tumour cells and stromal-derived cytokines, which are key to the development and establishment of breast cancer metastases [
4‐
7]. The stromal cells of mammary tumours produce a rich array of cytokines; one of the best documented is a cytokine known as hepatocyte growth factor (HGF).
Hepatocyte growth factor plays a crucial role in promoting breast tumour progression and metastasis. Upon complexing with its specific receptor c-Met, HGF evokes an array of biological responses within cancer cells, which subsequently lead to enhanced cell migration, matrix degradation, invasiveness and induction of angiogenesis. The significance of HGF activity in cancer development and progression has also been confirmed through clinical studies; where the degree of HGF and c-Met expression was found to correlate with disease progression and poor patient prognosis [
8,
9]. The significance of the HGF/c-Met signalling in cancer has been extensively documented [
10‐
15]. Collectively, these studies demonstrate that HGF and its receptor are strong therapeutic targets for cancer treatment and investigation of factors that govern the influence of HGF/c-Met coupling may lead to novel strategies to combat the metastatic spread of tumours.
Importantly, recent studies have also recognised the potential of the c-Met receptor as a target for treating the triple negative breast cancer (TNBC) subgroup of breast cancer patients [
16,
17]. There are over one million new cases of breast cancer diagnosed globally each year, and of these approximately 15% are classed as triple-negative breast cancers as they are found to be lacking the oestrogen receptor, the progesterone receptor and the human epidermal growth factor receptor 2 [
18‐
20]. Current successful hormone-based therapies directly target these three receptors, unfortunately patients with malignancies characterised as TNBC are associated with aggressive cancers with early metastatic spread and overall poor prognosis [
21‐
23]. The poor outcome for patients with TNBC is due to the lack of a validated targeted therapy. Therefore, the recognition of a target genes, such as c-Met, is one of the most urgent current requirements for breast cancer treatment.
Traditional medicines are used by many people in the world today, generally because natural products are considered to be safe, inexpensive, and targeted towards a number of diseases. In most cases, however, neither the active components nor their mechanisms of action are well established. Frankincense is derived from the tree sap resin of the genus
Boswellia and has been valued throughout the ages to have a wealth of healing properties. Scientific literature reports that a number of the
Boswellia species, and the active component of
Boswellia (boswellic acids), display anti-cancer properties through a capacity to reduce tumour growth and metastasis in a variety of established models [
24‐
27]. Breast cancer studies have also demonstrated that
Boswellia serrata and
Boswellia sacra extracts have been developed to suppress the aggressive nature of breast cancer cells and their propensity to metastasise to secondary sites such as the brain [
28,
29]. Presently, there are clinical trials underway examining the potential benefits of
B. serrata treatment in the management of breast and colon cancer (ClinicalTrials.gov: NCT 03149081).
In this study we sought to investigate the anti-cancer properties of
Boswellia frereana, a
Boswellia species native to Somalia, on breast cancer cells.
Boswellia frereana has been reported to act as an inhibitor of matrix metalloprotease 9 (MMP-9) activity during inflammation within an articular cartilage explant model [
30]. HGF may also play a role in governing MMP-9 expression levels, as HGF antagonists have demonstrated the ability to downregulate MMP-9 activity in lung cancer cells [
31]. This is the first study to assess the potential of BFE in cancer and we examined the effects of BFE on TNBC cell proliferation, migration, matrix-adhesion, invasion, angiogenesis and the activation/phosphorylation of the c-Met receptor under the influence of HGF. Here, we report that BFE suppresses HGF-enhanced cell migration, adhesion, vessel formation and invasion of breast cancer cells in vitro through inhibition of HGF/c-Met signalling and reduction of c-Met receptor phosphorylation.
Methods
Cells and materials
This study used the BT549 and MDA-MB-231 human TNBC cell lines, which were obtained from ATCC/LGC standard (Teddington, Middlesex, UK) and a human endothelial cell line (HECV) from Interlab Cell Line Collection (ICLC, Naples, Italy). Cells were routinely cultured with Dubecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum, penicillin and streptomycin (Thermo Fisher Scientific, Paisley, UK). Breast cancer cells were passaged for less than 2 months before fresh cells were resuscitated from earlier cryogenically preserved stocks. Recombinant human hepatocyte growth factor was obtained from PeproTech (PeproTech house, London, UK) and was used at a final concentration of 10 ng/ml throughout the study unless otherwise stated.
Boswellia frereana gum resin was purchased from Hargeisa, Somaliland, and successfully extracted as described previously [
30]. Briefly, absolute ethanol was used to extract
B. frereana, followed by removal of insoluble gum resin through filtration, and subsequent evaporation and solubilisation of the BFE in ethanol. The constituents of BFE were then analysed and verified through GC–MS.
Cytotoxicity assay
BT549 and MDA-MB-231 breast cancer cells were seeded in a 96 well plate at a density of 5000 cells/well and incubated under routine conditions for 24 h. Following incubation, media was removed and replaced with media containing the appropriate concentration of BFE. The concentration of BFE added to the cells ranged from 500 μg/ml to 1 μg/ml. Breast cancer cells were then incubated for a further 72 h, followed by standard MTT assay assessment of cell viability. The sub-toxic concentration of BFE (highest concentration with 100% viability), and the LC50 values (concentration of BFE responsible for 50% cell death) were determined for each breast cancer cell line. The sub-toxic concentration of BFE in both cell lines was 10 ug/ml, and the LC50 values for BT549 and MDA-MB-231 were 27.2 μg/ml ± 2.3 and 29.8 μg/ml ± 2.1 respectively.
Cell IQ cellular migration assay
Cellular migration assays were conducted using the self-contained Cell-IQ
® system (Chip-Man Technologies Ltd, Tampere, Finland) as described in detail previously [
32]. Briefly, the Cell-IQ
® system is a fully integrated continuous live cell imaging and automated analysis platform which combines phase-contrast microscopy, environmental control (5% CO
2 and maintained at 37 °C), with an on-board Analyser software package for the quantification of migration image data.
The migratory properties of the BT549 cells were assessed to determine the impact of HGF stimulation and BFE treatment on the aggressive motile nature of these breast cancer cells. Cells were seeded into a 24-well plate (Thermo Fisher Scientific, Paisley, UK) at 50,000 per well and incubated until confluence was reached. The monolayer was then scratched using a fine plastic pipette tip to create a wound of width approximately 300 µm. The media together with the floating cells were removed and replaced with media containing the treatment groups (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]). The plate was immediately placed in the chamber of the Cell IQ unit and phase contrast images were automatically captured every 30 min for 6 h at pre‐programmed positions. Sequential images were analysed using the on-board Analyser software package.
The analysis software automatically calculated the mean width of the wound of each well at each time point. Total cell migration distance (µm) was calculated by deducting the calculated wound width from a specific time point, from the wound width at time 0, which then allowed the plotting of time course data. Whereas the migration rate (µm/h) was calculated by deducting the mean width of the wound from the mean width of the wound of the previous hour. This migration rate was calculated hourly for the duration of the experiment (6 h), and generated a mean migration rate for each treatment group for each time point. All experiments were conducted in triplicate.
Breast cancer invasion assay
We quantified the invasive nature of the breast cancer cells using the standard invasion assay procedure as described previously [
33]. Transwell chambers, equipped with a 6.5 mm diameter polycarbonate filter insert (pore size 8 μm)(Becton–Dickinson Labware, Oxford, UK), were pre-coated with 100 μg/insert of solubilised tissue basement membrane, Matrigel (Scientific Lab Supplies, Nottingham, UK). BT549 and MDA-MB-231 breast cancer cells were seeded at a density of 5000/insert and allowed to invade for 72 h in the presence of BFE (10 μg/ml) and/or HGF (10 ng/ml). Following incubation, cells that had invaded through the basement membrane were fixed (4% formaldehyde), and then stained with 0.5% crystal violet. For analysis, the cells were counted (10 fields/insert under ×40 magnification), to determine the mean number of invaded cancer cells for each treatment group.
Cell–matrix adhesion assay
The cell–matrix adhesive properties of the BFE-treated cells to an artificial basement membrane were quantified using the in vitro Matrigel adhesion assay adapted from a previously described method [
34]. Briefly, BT549 and MDA-MB-231 cells were seeded at a density of 10,000/well into a 96-well plate that had been previously coated with 5 µg of Matrigel artificial basement membrane. Cells were then routinely incubated for 30 min in the presence of the described treatments (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]). Cells were allowed to adhere to the replica cell matrix, before two washing steps in PBS to remove non-adhered cells. Adherent cells were then fixed in 4% formaldehyde and stained with 0.5% crystal violet. Cells were counted in several fields/well under ×40 magnification to determine mean number of matrix-adhered cells per treatment group.
The development of spheroids through 3D-culture is more representative of tumour characteristics grown in vivo than traditional 2D-cultured cells. BT549 breast cancer cells will grow and form dense multicellular spheroids when cultured in the appropriate conditions. We used the ‘hanging-drop’ system which takes advantage of the fact that, in the absence of a surface to attach, BT549 cells will assemble into a 3D spheroid structure. Cells were added at 10,000 per well, to a critical final volume of 50 µl per well, in the various treatments groups (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]). The aperture of each well in the plate (Perfecta3D hanging drop plate, Sigma-Aldrich, UK), is designed so that when the cells are carefully dispensed into the well, a hanging droplet is formed that is small enough to remain suspended through surface tension. The cells were then cultured under routine conditions for 4 days to allow the cells to grow and develop into spheroids within the hanging droplet. Images were collected and the mean size/area of each spheroid was calculated for each treatment group.
Proliferation assay
BT549 and MDA-MB-231 breast cancer cells were seeded in a 96 well plate at a density of 5000 cells/well and incubated under routine conditions for 24 h. Following incubation, media was removed and replaced with media containing the appropriate treatment (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]), and then incubated for a further 72 h prior to standard MTT assay assessment of viable cell number.
Human umbilical cord endothelial cells (HECV) were plated into a 24‐well plate pre‐coated with Matrigel (diluted 1:1 with serum‐free media), at a seeding density of 20,000/well. Treatments were added (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]) and tubules were allowed to form over a 24 h incubation period. Images were captured to analyse the degree of tube/vessel formation as described previously [
35].
SDS PAGE and Western blotting
BT549 breast cancer cells were seeded into a 6 well plate and allowed to grow until 70–80% confluence, whereupon cells were subjected to overnight serum starvation in a serum free culture medium. The appropriate treatments were then added to the cells (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]), for a 60 min incubation period. This was followed by standard cell lysis, SDS-PAGE and Western Blot techniques. The c-Met phosphorylation status and c-Met total protein levels were assessed using the Phospho-Met (Tyr1234/1235) and Met (25H2) Mouse mAb antibodies respectively. All antibodies were purchased from Cell Signalling Technology unless otherwise stated (Cell Signalling Technology, Leiden, Netherlands). STAT-3 and JAK-2 activity was examined with the Phospho-Stat3 (Tyr705) and the Phospho-Jak2 (Tyr1007/1008) antibodies. B-actin expression levels were used as an internal control (Santa-Cruz Biotechnologies, California, USA). Protein expression was assessed and quantified using Image J analysis software.
Gelatinase zymography
Gelatinase zymography was performed to assess the expression levels of both the inactive pro-form and biologically active form of MMP-2 and MMP-9 that are produced by the BT549 and MDA-MB-231 cell lines. Breast cancer cells were seeded into a 6 well plate and allowed to grow until 50% confluence, whereupon cells the appropriate treatments were then added to the cells (Control ± HGF [10 ng/ml]; BFE [10 μg/ml] ± HGF [10 ng/ml]), and incubated for an additional 4 days (80–90% confluence). Cell supernatant was then collected from each treatment group and samples were processed and prepared for standard gelatinase zymography as described previously [
30].
Statistical analysis
The results were assessed using one-way ANOVA test with post hoc Tukey HSD, and non-paired two-sided Student’s t-test. A p-value < 0.05 was defined as statistically significant.
Discussion
In the present study, we have demonstrated for the first time that an extract from B. frereana resin was able to suppress the influence of HGF on the BT549 and MDA-MB-231 TNBC cell lines in vitro. Upon HGF stimulation, these breast cancer cells revealed a dramatic and significant increase in motility, cell–matrix adhesion and invasiveness; this increase was governed through HGF-Met coupling and subsequent activation of the c-Met receptor. HGF stimulation did not appear to have any impact on the proliferation rate or the growth and formation of 3D spheroids within these breast cancer cells. HGF also increased the degree of tubulogenesis/vessel formation when added to a human endothelial cell line. However, the addition of the B. frereana extract quenched all these HGF-induced pro-metastatic events in vitro without affecting the growth of these cells. Further study revealed that the BFE inhibited these HGF-enhanced effects through suppression of c-Met receptor tyrosine kinase phosphorylation (pY1234/5) within these breast cancer cells.
This study also assessed a number factors that are up/down-regulated during the process of epithelial-mesenchymal transition (EMT). EMT occurs through the disassembly of epithelial cell–cell contacts and reorganisation of ECM to facilitate cell migration, invasion and metastasis. HGF is believed to act as an initiation signal for the onset of EMT, which results in the upregulation of effector molecules, such as MMP-2 and MMP-9, and activation of STAT-3 to aid EMT and the metastatic dissemination of breast cancer cells [
4,
36,
37]. Our results suggested that treatment with HGF and/or BFE did not appear to have any bearing on the activity or expression levels of these factors within these breast cancer cell lines.
Breast cancer cell invasion and establishment of metastasis are devastating events for patients with cancer. Patients classified with TNBC have a type of cancer that is characterised by the aggressive invasive clinical behaviour of the tumour which has a propensity to metastasise early and establish secondary tumours at additional sites. Furthermore, TNBC patients have limited targeted treatment options due to the fact that they do not express any of the three receptors directly targeted by the current successful therapies. For these reasons TNBC patients are generally considered to have a poor prognosis, thus emphasising the importance of recognition and validation of novel targets for the inhibition of metastasis in TNBC.
Hepatocyte growth factor and its partner c-Met play a definitive role in the development and progression of breast cancers, as well as being implicated in particularly invasive and metastatic cancers. HGF/c-Met are logical therapeutic targets in the treatment of TNBC. Strategies targeting HGF/c-Met activity have warranted further investigation for their potential in combating the metastatic spread of tumours. In recent years, the blockade of HGF-Met signalling has become one such strategy to inhibit tumour invasion and metastasis [
10,
38]. Clinical studies have demonstrated the potential of c-Met inhibitors as a novel form of targeted therapy, as an investigation into primary breast cancer patients revealed that elevated phosphorylated c-Met expression (pY1234/5) was found to be associated with TNBC patients [
39]. Several reports review the clinical trials that are presently evaluating the benefit of anti-c-Met and anti-HGF monoclonal antibodies and tyrosine kinase inhibitors in a variety of cancer types including TNBC [
9,
16,
40]. The clinical relevance of c-Met inhibitors in breast cancer is under investigation with phase II clinical trials evaluating the potential of tivantinib in patients with recurrent or metastatic TNBC [
41] and in a randomized phase II study assessing the safety and efficacy of onartuzumab and/or bevacizumab in combination with paclitaxel in patients with metastatic TNBC [
42], where the aim is to identify patients who would benefit from c-Met targeted therapy. Further progress will lead to the application of these advances in the generation of future therapies to prevent the spread of cancer in TNBC patients. To date, the most promising approach for disrupting HGF/c-Met signalling appears to be small molecular inhibitors that target the intracellular kinase domain [
43].
We demonstrate that BFE had a significant and direct effect on HGF-induced migration, adhesion, invasion and tubulogenesis through the inhibition of HGF/c-Met signalling and reduction in c-Met tyrosine kinase phosphorylation. This shift between HGF-induced c-Met activation and inhibition of c-Met signalling by BFE, could be a pivotal step controlling the metastatic influence of HGF, thus limiting breast tumour progression. The binding of HGF to c-Met and the subsequent activation of the intracellular domain occurs through a process of phosphorylation of the two tyrosine residues in the catalytical regions Y1234 and Y1235, followed by phosphorylation of two docking tyrosines (Y1349 and Y1356) [
44]. This results in the recruitment of adaptor proteins which facilitate the binding to multiple substrates [
45], which in turn leads to the activation of a variety of downstream intracellular signalling pathways (such as PI3K-AKT, MAPK, STAT and NF-κB), which are responsible for driving cell proliferation, invasion, migration, angiogenesis and cell survival [
46,
47]. The triggering of downstream signalling events upon c-Met activation appears to be cell or context specific [
16].
One study reveals that a specific c-Met tyrosine kinase activation pathway correlates with high risk patients who will go on to develop an aggressive form of the disease; this also suggests that the determination of this c-Met activation signature pathway may be used as a clinical tool to identify and predict patient response for future personalised anti-Met therapies [
48]. The use of such clinical tools will play a key role in the management of breast cancer and may be used in the process of stratifying patients accordingly. It is important to identify those subgroups of patients most likely to benefit from anti-Met therapies, such as those with TNBC or further sub-classifications of TNBC. The outcomes of current and future c-Met clinical trials are eagerly anticipated. However, Ho-Yen et al. [
10], reveal that such issues as resistance and c-Met receptor cross-talk with other receptor tyrosine kinases need to be addressed if treatment efficacy is to reach its full potential.
Authors’ contributions
CP obtained all materials, designed the study, performed all experimental work, analysed data, interpreted results and prepared the manuscript. AYA isolated BFE and performed GC–MS validation. Both authors read and approved the final manuscript.