Pharmacological inhibition of store-operated calcium entry in MDA-MB-468 basal A breast cancer cells: consequences on calcium signalling, cell migration and proliferation

Abstract

Store-operated Ca²⁺ entry is a pathway that is remodelled in a variety of cancers, and altered expression of the components of store-operated Ca²⁺ entry is a feature of breast cancer cells of the basal molecular subtype. Studies of store-operated Ca²⁺ entry in breast cancer cells have used non-specific pharmacological inhibitors, complete depletion of intracellular Ca²⁺ stores and have mostly focused on MDA-MB-231 cells (a basal B breast cancer cell line). These studies compared the effects of the selective store-operated Ca²⁺ entry inhibitors Synta66 and YM58483 (also known as BTP2) on global cytosolic free Ca²⁺ ([Ca²⁺]_CYT) changes induced by physiological stimuli in a different breast cancer basal cell line model, MDA-MB-468. The effects of these agents on proliferation as well as serum and epidermal growth factor (EGF) induced migration were also assessed. Activation with the purinergic receptor activator adenosine triphosphate, produced a sustained increase in [Ca²⁺]_CYT that was entirely dependent on store-operated Ca²⁺ entry. The protease activated receptor 2 activator, trypsin, and EGF also produced Ca²⁺ influx that was sensitive to both Synta66 and YM58483. Serum-activated migration of MDA-MB-468 breast cancer cells was sensitive to both store-operated Ca²⁺ inhibitors. However, proliferation and EGF-activated migration was differentially affected by Synta66 and YM58483. These studies highlight the need to define the exact mechanisms of action of different store-operated calcium entry inhibitors and the impact of such differences in the control of tumour progression pathways.

Electronic supplementary material

The online version of this article (10.1007/s00018-018-2904-y) contains supplementary material, which is available to authorized users.

Introduction

The existence of a mechanism by which the depletion of intracellular Ca²⁺ stores triggers a Ca²⁺ influx pathway that is then used to refill Ca²⁺ stores was proposed by Putney in 1986 [1], with store-operated Ca²⁺ entry, one of the terms used to describe this phenomenon. The molecular components that drive Ca²⁺ influx to refill the endoplasmic reticulum (ER) intracellular Ca²⁺ stores after depletion, were identified a little over a decade ago [2]. Since the original finding, the process of store-operated Ca²⁺ entry (SOCE) has been defined further, with a variety of regulators identified [3]. Store-operated Ca²⁺ entry is typically initiated by the sensing of reduced levels of Ca²⁺ in the ER by stromal interaction molecule 1 (STIM1). Subsequently, STIM1 is oligomerised and locates to sites of the ER that are in close proximity to the plasma membrane. STIM1 then interacts with the hexamers of the plasma membrane protein and Ca²⁺ channel component Orai1 to promote Ca²⁺ influx [4–6].

Although the refilling of Ca²⁺ stores could be seen as a simple “house-keeping” function, it is clear that Orai1-mediated Ca²⁺ influx is more than this, given its important role in key cellular processes, including specific immune system events, lactation and male fertility [7–9]. Indeed, decoding of Ca²⁺ changes in cellular compartments allows Orai1-mediated Ca²⁺ influx to differentially regulate transcription by NFAT1 and NFAT4 [10]. These characteristics of the store-operated Ca²⁺ influx pathway have defined it as a legitimate pharmacological target for drug discovery programs to improve human health. The conditions where store-operated Ca²⁺ entry inhibitors have been proposed to potentially have therapeutic benefit are varied, and range from vascular thrombosis to inflammatory bowel disease [11] and pancreatitis [12] to renal fibrosis [13].

Another disease state where Orai1-mediated Ca²⁺ influx has been recently investigated is cancer [14]. Remodelling of the expression of components of store-operated Ca²⁺ entry is a feature of many cancers including those of the colon [15], liver [16] and oesophagus [17]. However, the potential remodelling of Orai1-mediated Ca²⁺ influx is not ubiquitous, as reflected in the apparent switch from store-operated Ca²⁺ entry to more Orai3-mediated Ca²⁺ influx in some prostate cancers [18]. Differences may even be evident between breast cancer subtypes, which differ significantly in their prognosis and therapy. For example, Orai1 levels are higher in basal compared to non-basal breast cancers [19], and cell line data suggest that Orai3 is more important than Orai1 in store-operated Ca²⁺ entry in estrogen receptor-positive breast cancer cells compared to estrogen receptor-negative breast cancer cells [20]. In the specific context of Orai1-mediated Ca²⁺ influx in basal breast cancer, which has a significant overlap with triple negative breast cancers, where the need for new therapies has been consistently highlighted [21], Orai1 silencing reduces the proliferation and invasiveness of MDA-MB-231 triple negative breast cancer cells [19, 22]. This combined with the association between high STIM1 and low STIM2 with poorer prognosis in basal breast cancers [19], has highlighted the potential importance of Orai1 in this breast cancer subtype.

Future therapies for basal/triple negative breast cancer that seek to exploit the remodelling of Orai1-mediated Ca²⁺ influx will most likely involve pharmacological inhibitors of this Ca²⁺ influx pathway. Although encouraging, studies assessing the effects of store-operated Ca²⁺ entry inhibitors in triple negative breast cancer cell lines in vitro or in vivo have had some limitations [19, 22]. First, studies have often focused on the most common triple negative breast cancer cell line model (MDA-MB-231 cells) and have not explored other triple negative breast cancer cell lines. Effects on Ca²⁺ influx have involved pronounced Ca²⁺ store depletion in the absence of extracellular Ca²⁺ and have not assessed or compared the contribution of store-operated Ca²⁺ influx with different physiological stimuli under more biologically relevant conditions. Finally, the store-operated Ca²⁺ channel inhibitors used in previous studies (2-APB and SKF96365) were not those that are more selective for this Ca²⁺ influx pathway [23]. Hence, in these studies, we sought to compare the nature of Ca²⁺ influx induced by different stimuli under more physiological conditions and the sensitivity of this Ca²⁺ influx to the store-operated Ca²⁺ entry inhibitors Synta66 and YM58483, in a different basal breast cancer cell model, MDA-MB-468 triple negative breast cancer cells. MDA-MB-468 cells are PTEN mutant with high levels of EGFR expression [24–26], two features that are also associated with metastasis and poor patient survival [27, 28]. These cells also possess epithelial–mesenchymal plasticity features [29], a characteristic that is a critical regulator of cancer progression [30]. This work provides new insights into how different stimuli may engage store-operated Ca²⁺ influx and point to differences between Synta66 and YM58483 in the regulation of specific aspects important in breast cancer progression.

Materials and methods

Cell culture

The MDA-MB-468 cell line was obtained from The Brisbane Breast Bank, The University of Queensland Centre for Clinical Research (UQCCR), Brisbane, QLD, Australia. Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) containing 1000 mg/L of glucose (D6546, Sigma-Aldrich, St Louise, MO, USA), supplemented with 10% foetal bovine serum (FBS) and 4 mM l-glutamine, and maintained in a humidified incubator at 37 °C in a 5% CO₂ atmosphere. Cells were tested every 6 months for mycoplasma using MycoAlert kit (Lonza, Basel, Switzerland), and authenticated by short tandem repeat (STR) profiling using the GenePrint 10 System (Promega, Madison, WA, USA) at QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia.

Measurement of intracellular Ca²⁺

Measurement of cytosolic free Ca²⁺ ([Ca²⁺]_CYT) was performed using a fluorometric Imaging Plate Reader (FLIPR^TETRA, Molecular Devices) with either a PBX no-wash Ca²⁺ Assay Kit (640175, BD Biosciences) in 384-well black plates (in case of initial characterisation of agonist-induced Ca²⁺ increases shown in Fig. 1) as previously described [23, 31], or Fluo-4 loaded cells in 96-well black plates (in case of Ca²⁺ measurements in Figs. 2, 3) as previously described [32]. Data were analysed using ScreenWorks Software (v2.0.0.27, molecular devices), and response over baseline was assessed as a relative measure of [Ca²⁺]_CYT.

Fig. 1.

Characterisation of EGF, ATP and Trypsin-induced Ca²⁺ increases in MDA-MB-468 breast cancer cells. Mean [Ca²⁺]_CYT levels during addition of EGF (50 ng/mL), ATP (100 µM) and Trypsin (10 nM) in the presence of 1.8 mM extracellular Ca²⁺ (a, e, i) or the extracellular Ca²⁺ chelator, BAPTA (500 µM) in the absence of added extracellular Ca²⁺ (b, f, j). Quantification of relative peak [Ca²⁺]_CYT (c, g, k), and relative [Ca²⁺]_CYT at 800 s (d, h, l) is shown for each treatment group (mean ± SD, n = 3). ns not significant (P > 0.05), *P < 0.05, **P < 0.01 (paired t test)

Fig. 2.

Inhibition of CPA, EGF and trypsin-induced store-operated Ca²⁺ entry by Synta66 and YM58483. Mean relative [Ca²⁺]_CYT levels and quantification of relative peak [Ca²⁺]_CYT and relative [Ca²⁺]_CYT at 1600 s during addition of CPA (10 µM) (a–c), and at 600 s during addition of EGF (50 ng/mL) (d–f) and Trypsin (15 nM) (g–i) in the presence of 10 µM Synta66 or YM58483 compared with the DMSO control (0.1%). Cells were pre-treated with Synta66, YM58483 or DMSO control for 15 min prior to addition of agonists (containing the same concentration of Synta66, YM58483 and DMSO control). All the Ca²⁺ traces and bar graphs presented are the mean of three independent experiments ± SD and for bar graphs, ns not significant (P > 0.05), *P < 0.05, **P < 0.01 (paired t test compared to the DMSO control)

Fig. 3.

Inhibition of ATP-induced store-operated Ca²⁺ entry by Synta66 and YM58483. Mean [Ca²⁺]_CYT levels (a, b) and quantification of relative peak [Ca²⁺]_CYT and relative [Ca²⁺]_CYT at 600 s during addition of 0.5 and 10 µM ATP (c, d) in the presence of 10 µM Synta66 or YM58483 compared with the DMSO control (0.1%). Cells were pre-treated with Synta66, YM58483 or DMSO control for 15 min prior to addition of ATP solution (containing the same concentration of compounds and DMSO control). All the Ca²⁺ traces and bar graphs presented are the mean of three independent experiments and for bar graphs ± SD, ns not significant (P > 0.05), **P < 0.01 (paired t test compared to the DMSO control)

Cell migration assay

Cell motility was measured using a collagen-based single cell migration assay and a live cell imaging system. 96-well plates were coated with 50 µL of collagen solution containing sterile-filtered 10× phosphate-buffered saline (PBS) (8% v/v), cell culture media (24% v/v), type I collagen from bovine skin, buffered to physiological pH with 1 M NaOH. After incubating plates at 37 °C and 5% CO₂ in humidified incubator for 1 h to allow collagen gel to solidify, cells were plated on top of the gel (500 cells, 100 µL volume per well). Post seeding (24 h), cell culture media were replaced with serum-reduced media (SRM, 0.5% FBS) containing Synta66 (10 µM) or YM58483 (10 µM) or DMSO control (0.1%) for 24 h. Media were replaced with either complete media (CM, 10% FBS) or SRM with EGF (50 ng/mL), containing DMSO control, Synta66 or YM58483. Plates were then placed in a JuLi™ Stage Live Cell Imaging System (NanoEnTek Inc. Seoul, South Korea) housed in a 37 °C and 5% CO₂ humidified incubator for 24 h with bright-field images acquired every 15 min. Single cell tracking was performed on images acquired 12 h after addition of EGF or complete media for a period of 12 h. ImageJ 1.49q software (NIH, Bethesda, MD, https://imagej.nih.gov/ij/) was used to track cells. Accumulated distance travelled by each cell was calculated and illustrated by Chemotaxis and Migration Tool V2.0 (Ibidi, Munich, Germany). Strict criteria were employed regarding exclusion of cells from analysis. Cells which died (loss of plasma membrane integrity or formation of apoptotic bodies), or moved into or out of the field of view during the period of migration assessment was excluded from analysis. Cells which clumped together were also excluded as only single cell motility was assessed.

Cell proliferation assay

Cells were seeded at a density of 6 × 10³ per well in 96-well plates. Post seeding (24 h), cells were treated with vehicle (0.1% DMSO) or 10 µM Synta66 or YM58483 for 72 h. Fresh DMSO control or SOCE inhibitor containing media was added every 24 h. Cell proliferation was assessed using the colorimetric MTS assay (CellTiter 96 AQueous One Solution Cell Proliferation Assay; Promega, Madison, WA, USA). CellTiter 96^® AQueous One Solution Reagent (20 µL) was added into each well containing 100 µL of culture medium. Plates were incubated at 37 °C and 5% CO₂ in a humidified incubator for 2 h. Subsequently, absorbance was measured at 490 nm using a model 550 microplate Reader (Bio-Rad, Hercules, CA, USA) and numbers subtracted from the background “no cell” absorbance.

siRNA-mediated silencing

MDA-MB-468 cells were seeded at 15 × 10³ cells per well in 96-well plates. 24 h post seeding, DharmaFECT4 Transfection Reagent (0.1 μL per well) and Dharmacon ON-TARGETplus SMARTpool siRNAs (Dharmacon Inc, USA), were used for silencing Orai1 (L-014998-00) and TRPC3 (L-006509-00). Non-targeting siRNA (siNT; D-001810-10) was used as control treatment. RNA was isolated 48 h post-transfection and silencing efficiency was confirmed using real-time RT-PCR.

Real-time RT-PCR

RNA was isolated and purified using the Qiagen RNeasy Mini Kit (Qiagen, Hilden, Germany). cDNA was synthesized from isolated RNA using the Omniscript Reverse Transcription Kits (Qiagen) and amplified using the TaqMan Fast Universal PCR Master Mix (Applied Biosystems, CA, USA). Real-time PCR were performed using a StepOne Plus Real-Time PCR System (Applied Biosystems) under universal cycling conditions. TaqMan gene expression assay IDs used were: ORAI1 (Hs03046013_m1) and TRPC3 (Hs00162985_m1). The comparative C_T method was used to quantify the relative expression of assessed genes normalised to 18 s ribosomal RNA gene.

Statistical analysis

Statistical analyses were performed using GraphPad Prism Version 7.00 for Windows, GraphPad Software (La Jolla, CA, USA, www.graphpad.com). Specific statistical tests for each experiment are stated in the corresponding figure legend.

Results

To compare the degree of sustained Ca²⁺ influx in MDA-MB-468 triple negative breast cancer cells initiated by different stimuli, we compared [Ca²⁺]_CYT transients in the presence and absence of extracellular free Ca²⁺ initiated by epidermal growth factor (EGF) (50 ng/mL), adenosine triphosphate (ATP) (100 µM) and trypsin (10 nM) (Fig. 1). EGF is an activator of the EGF receptor (EGFR), a tyrosine kinase receptor that can couple to phospholipase C (PLC)β, and induce Ca²⁺ store release via generation of inositol 1,4,5-trisphosphate (IP₃) [33]. As can be seen in Fig. 1a–d, and consistent with the pronounced expression of EGFR in MDA-MB-468 cells [24], EGFR produced a gradual increase in [Ca²⁺]_CYT with a sustained phase of elevated [Ca²⁺]_CYT that was abolished by the removal of extracellular Ca²⁺. ATP is an activator of various P2Y G-protein coupled receptors, which can couple to PLCβ to generate IP₃ and then cause intracellular Ca²⁺ store release [34]. ATP can also activate various P2X receptors that are ligand gated Ca²⁺ permeable ion channels [34]. ATP produced a rapid increase in [Ca²⁺]_CYT with a very pronounced sustained phase of [Ca²⁺]_CYT that was entirely dependent of Ca²⁺ influx (Fig. 1e–h). Similarly, the G-protein coupled receptor protease-activated receptor 2 (PAR2) activator, trypsin, produced Ca²⁺ store release but with a less sustained phase of [Ca²⁺]_CYT elevation, although this sustained phase was also abolished by the removal of extracellular Ca²⁺ (Fig. 1i–l). Hence, all of the activators assessed had two components of increase in [Ca²⁺]_CYT in MDA-MB-468 breast cancer cells. The first phase was Ca²⁺ store release, which as expected, was not dependent on Ca²⁺ influx (Fig. 1c, g, k), and a second sustained phase of elevated [Ca²⁺]_CYT, which was due to the influx of extracellular Ca²⁺ (Fig. 1d, h, l). We then sought to assess the sensitivity of this Ca²⁺ influx to two pharmacological inhibitors of store-operated Ca²⁺ entry, Synta66 and YM58483 at 10 µM, a concentration selected from previous studies that showed this concentration is above the IC₅₀ in cell line models [23, 35–37].

Most studies assessing store-operated Ca²⁺ entry have used a protocol involving removal of extracellular Ca²⁺ followed by pronounced Ca²⁺ store depletion with sarco/endoplasmic reticulum Ca²⁺-ATPase inhibitors [e.g. cyclopiazonic acid (CPA)], and then the re-addition of extracellular Ca²⁺. Consistent with the promotion of store-operated Ca²⁺ entry induced by CPA in MDA-MB-468 cells under these conditions [38], CPA induced a Ca²⁺ influx phase that was sensitive to the store-operated Ca²⁺ entry inhibitors Synta66 and YM58483 (15 min pre-incubation) (Fig. 2a–c), with only a modest effect on maximal [Ca²⁺]_CYT, due to concurrent Ca²⁺ influx during the peak, which was a statistically significant effect for Synta66. The [Ca²⁺]_CYT transient initiated by EGF was totally remodelled by Synta66 and YM58483 (Fig. 2d), with no effect on the peak [Ca²⁺]_CYT initiated by Ca²⁺ store release (Fig. 2e), but a complete attenuation of the very sustained phase of Ca²⁺ influx, which was observed after 600 s of EGF addition (Fig. 2f). Although trypsin (15 nM), did not produce the same degree of sustained Ca²⁺ influx as EGF (Fig. 2g), it similarly had a peak [Ca²⁺]_CYT increase that was insensitive to Synta66 and YM58483 (Fig. 2h), but a sustained Ca²⁺ influx that was sensitive to these store-operated Ca²⁺ entry inhibitors (Fig. 2i). Assessment of increases in [Ca²⁺]_CYT initiated by ATP under these conditions indicated that the contribution of store-operated Ca²⁺ entry varied with the concentration of ATP (Fig. 3a, b). Although peak [Ca²⁺]_CYT initiated by both submaximal (0.5 µM) and maximal (100 µM) ATP were insensitive to Synta66 and YM58483 (15 min pre-incubation) (Fig. 3c), the sustained phase of [Ca²⁺]_CYT elevation with 100 µM ATP was far more sensitive to Synta66 and YM58483 (Fig. 3d). These studies collectively indicate that many stimuli can promote store-operated Ca²⁺ entry that is sensitive to Synta66 and YM58483 in MDA-MB-468 breast cancer cells. We, therefore, assessed the effects of these agents on the migration and proliferation [19, 22] of this breast cancer cell line.

Cell migration of MDA-MB-468 breast cancer cells was assessed in control, Synta66 and YM58483 treated wells using live cell imaging. In the case of cell migration induced by the addition of complete media (including 10% FBS), single cell analysis (Fig. 4a–c) demonstrated that there was no significant effect of Synta66 on mean accumulated distance travelled during the assessed period, but YM58483 significantly reduced mean accumulated distance travelled (Fig. 4d). However, it was clear that the proportion of cells with little to no movement was increased by both Synta66 and YM58483 (Fig. 4e) and this was significant for both inhibitors (Fig. 4f). EGF was also assessed (Fig. 5a–c), and EGF was a more effective stimulus for MDA-MB-468 breast cancer cell migration, with a mean accumulated distance of around 270 µm compared to that for complete medium-induced migration at 160 µm (Fig. 4d vs 5d, and also Fig. S3). YM58483, but not Synta66, significantly reduced the mean accumulated distance of cell migration observed in the presence of EGF (Fig. 5d). In contrast to complete media, there was no promotion of the population of cells that have little to no migration, as neither Synta66 nor YM58483 had a significant effect on this population of cells (Fig. 5e, f). The differential effects of Synta66 and YM58483 were also apparent when proliferation was compared between MDA-MB-468 breast cancer cells cultured in the presence and absence of YM58483 and Synta66 for 3 days, where YM58483, but not Synta66, had a significant effect (Fig. 6a, b). Although further study is required, this difference could be related to the nature of inhibition of these two agents, as evident by the clear difference in time dependence of SOCE inhibition, where prolonging incubation with the inhibitor (24 h vs 15 min) promoted SOCE inhibition more with YM58483 than Synta66 (Fig. S1). The anti-proliferative effects of YM58483 were evident in MTS metabolism assays and decreased nuclei numbers (using Hoechst 33342 staining) (Fig. 6d, e). These decreases were not due to cell death as indicated by the absence of apoptotic morphology in bright field imaging (Fig. 6a) and apoptotic/condensed nuclei (Fig. 6d).

Fig. 4.

Assessment of the effect of Synta66 and YM58483 on a complete media (CM)-model of cell migration. Cells were serum-deprived (0.5% FBS) for 24 h prior to addition of complete media (10% FBS) containing DMSO (0.1%), Synta66 (10 µM) or YM58483 (10 µM) for 24 h. Cell migration was analysed from 12 h to 24 h after addition of complete media. Representative spatial plots of DMSO control (a), Synta66 (b) and YM58483 (c) of 15 randomised cells from one experiment. d Quantitative analysis of the median accumulated distance travelled by all the cells from three independent experiment (duplicate wells) treated with DMSO control (total 152 cells), Synta66 (total 126 cells) or YM58483 (total number of 156 cells). ns not significant (P > 0.05), **P < 0.01 (nonparametric Mann–Whitney test). e Histograms of accumulated distances travelled by cells binned at 10 µm intervals. f Percent of cells with accumulated distance travelled of less than 10 µm. Data are the mean ± SD from three independent experiments (n = 3). ns not significant (P > 0.05), *P < 0.05, **P < 0.01 (unpaired t test compared to the DMSO control)

Fig. 5.

Assessment of the effect of Synta66 and YM58483 on an EGF-model of cell migration. Cells were serum-deprived (0.5% FBS) for 24 h prior to addition of EGF (50 ng/mL) containing DMSO (0.1%), Synta66 (10 µM) or YM58483 (10 µM) for 24 h. Cell migration was analysed from 12 h to 24 h after EGF addition. Representative spatial plots of DMSO control (a), Synta66 (b) and YM58483 (c) of 15 randomised cells from one experiment. d Quantitative analysis of the median accumulated distance travelled by cells from three independent experiment (duplicate wells) treated with DMSO control (total number of 185 cells), Synta66 (total number of 176 cells) or YM58483 (total number of 198 cells). ns not significant (P > 0.05), ****P < 0.0001 (nonparametric Mann–Whitney test). e Histograms of accumulated distances travelled by cells binned at 10 µm intervals. f Percent of cells with accumulated distance travelled of less than 10 µm. Data are the mean ± SD from three independent experiments (n = 3). ns not significant (P > 0.05), ***P < 0.0001 (unpaired t test compared to the DMSO control)

Fig. 6.

Assessment of cell viability with Synta66 and YM58483 treatment. a Representative pictures of MDA-MB-468 breast cancer cells after treatment with DMSO control (0.1%) or 10 µM Synta66 or YM58483 for 72 h. b, c Percent inhibition of growth of MDA-MB-468 cells treated with different concentrations of Synta66 or YM58483 (0.03, 0.1, 0.3, 1, 3, 10 µM) for 72 h normalised to the DMSO control group. Data are the mean ± SD from three independent experiments (n = 3). *P < 0.01 (unpaired t test compared to the DMSO control). Scale bar = 500 µm. d Representative images from cell staining with Hoechst 33342 in cells treated with Synta66 or YM58483 for 72 h and a 24 h treatment with staurosporine (1 µM) as a positive control for apoptotic nuclei (4 regions per well, in triplicate wells). Scale bar = 250 µM. e Nucleus number of cells treated with DMSO control, or different concentrations of Synta66 or YM58483 (0.03, 0.1, 1, 3, 10 µM) for 72 h (n = 3, ± SEM). *** P < 0.001 (two-way ANOVA with Bonferroni’s multiple comparison)

Discussion

The remodelling of Orai1 mediated Ca²⁺ influx in breast cancer is particularly associated with breast cancers of the basal molecular phenotype, which are associated with a poorer prognosis. Basal breast cancers have higher levels of Orai1 than non-basal breast cancers [19]. Moreover, breast cancers of the basal molecular phenotype that have high STIM1 but low STIM2, are associated with a poor prognosis [19]. This association is not evident in breast cancers that are not of the basal molecular subtype [19]. The importance of Orai1 in basal breast cancers is also reflected in the anti-proliferative and anti-migratory effects of Orai1 silencing in the MDA-MB-231 basal breast cancer cell line [19, 22]. MDA-MB-231 breast cancer cells are of the Basal B molecular classification [39]. By comparison there has been limited study of breast cancer cell lines of the basal A subtype such as MDA-MB-468 cells, which are of particular interest because of their cellular plasticity and sensitivity to undergo epithelial–mesenchymal transition, a process important in breast cancer progression [40, 41]. Although Orai1-mediated store-operated Ca²⁺ entry has been demonstrated using Orai1 silencing in MDA-MB-468 breast cancer cells [42], these like most studies in cancer cells, have assessed store-operated Ca²⁺ entry under conditions that promote substantial and sustained Ca²⁺ store depletion, i.e. by including experimental periods where extracellular Ca²⁺ was excluded. Other limitations in the study of store-operated Ca²⁺ entry in basal breast cancer cells include using pharmacological modulators to inhibit store-operated Ca²⁺ entry, that we now know have significant off target effects [11, 19, 22, 23]. The research conducted here, therefore, sought to address some of these knowledge gaps.

MDA-MB-468 basal A breast cancer cells are clearly capable of sustained Ca²⁺ influx as exemplified by our comparison of the [Ca²⁺]_CYT changes elicited by ATP, EGF and trypsin. However, the magnitude of this Ca²⁺ influx phase was dependent on the stimuli, with ATP (100 µM) resulting in the most pronounced Ca²⁺ influx phase. The store-operated Ca²⁺ entry inhibitors Synta66 and YM58483 had a dramatic effect on [Ca²⁺]_CYT changes associated with the SERCA inhibitor, CPA. The abolition of the CPA-induced sustained phase of Ca²⁺ influx by Synta66 and YM58483 is consistent with Orai1 silencing studies in this cell line [42] and indicative that this Ca²⁺ influx pathway is important in this basal A breast cancer cell line. In contrast to the other activators assessed in these studies, there was a modest, and in the case of Synta66, a significant effect of store-operated Ca²⁺ entry inhibitors on peak [Ca²⁺]_CYT levels induced by CPA. This is likely because the other agents assessed in this study cause reductions in endoplasmic reticulum Ca²⁺ levels (the driving force of store-operated Ca²⁺ influx), via the rapid generation of IP₃ and release of Ca²⁺ via IP₃ gated endoplasmic Ca²⁺ channels, whereas SERCA inhibition by CPA depletes endoplasmic Ca²⁺ levels, via stopping the re-sequestration of Ca²⁺ slowly leaking from the endoplasmic reticulum [43]. Hence, in the case of CPA, store-operated Ca²⁺ entry is activated more gradually, and therefore, it would also contribute to the maximal increase in [Ca²⁺]_CYT seen with CPA. EGF and trypsin also produced Ca²⁺ influx that was sensitive to store-operated Ca²⁺ entry inhibitors.

The contribution of store-operated Ca²⁺ entry to global sustained increases in [Ca²⁺]_CYT was dependent (as would be predicted), on the degree of Ca²⁺ store depletion, as exemplified by a major effect of Synta66 and YM58483 on a maximal but not a submaximal concentration of ATP. However, the absence of pronounced effects on global [Ca²⁺]_CYT at lower concentrations of activators does not discount a potentially important role for store-operated Ca²⁺ entry during these more modest stimuli. Local increases around Orai1 channels are a regulator of NFAT1-mediated gene transcription; indeed Orai1 has been described as having a private line of communication to NFAT1 [10]. Hence more modest activation of Orai1 may remodel signalling and processes in breast cancer cells using this or similar mechanisms. Despite the presence of purinergic ATP activated Ca²⁺ permeable ion channels (P2X channels) in MDA-MB-468 breast cancer cells [42, 44], the sustained phase of Ca²⁺ influx induced by ATP (100 µM) was completely abolished by inhibitors of store-operated Ca²⁺ entry. This defines store-operated Ca²⁺ entry as either the major trigger and/or the major mediator of Ca²⁺ influx induced by some stimuli in MDA-MB-468 basal A breast cancer cells. In this context, the consequences of pharmacological inhibitors on two key processes in breast cancer progression (migration and proliferation) was assessed.

Consistent with studies using Orai1 silencing and less specific pharmacological inhibitors in MDA-MB-231 basal B breast cancer cells, Synta66 and YM58483 reduced the migration of MDA-MB-468 basal A breast cancer cells, as reflected in a significant increase in the percentage of cells with levels of migration less than 10 µm with complete media re-addition. However, effects on migration were not universal. The effects of Synta66 and YM58483 were not as evident when migration was induced by EGF. The degree of migration by EGF was greater than complete media as reflected in a higher mean accumulated distance, and this may indicate that other pathways that are initiated by EGF are less dependent on store-operated Ca²⁺ entry. EGF activated pathways in this cell line include STAT3, AKT, ERK1/2 [45, 46], which may be not be completely dependent on store-operated Ca²⁺ entry. However, under these conditions YM58483, but not Synta66, did significantly reduce accumulated distance observed with EGF stimulated migration. This suggests that Synta66 and YM58483 may not be totally equivalent and not interchangeable despite their similar effects on store-operated Ca²⁺ entry in various model systems [11, 23]. Differences were also clear in our studies on proliferation with YM58483, but not Synta66, having a significant effect on this parameter. Future studies could explore the potential impact of SOCE inhibitors on the modification of Ca²⁺ signals induced by ATP and trypsin and on other cellular processes in this model.

The differences between Synta66 and YM58483 in EGF-induced migration and cellular proliferation could be due to a variety of factors. These factors include differential degrees of store-operated Ca²⁺ entry inhibition, different effects on other ion channels and different mechanisms of action that may influence the nature of Orai1 inhibition. In the context of different degrees of store-operated Ca²⁺ entry inhibition, this was not evident in our Ca²⁺ assays using a variety of stimuli. Distinctive off targets effects may contribute to differential effects, indeed YM58483 inhibits TRPC3 [47, 48], which is expressed in MDA-MB-468 cells [38]. However, this is unlikely given the consistent reports of the inhibitory effects of Orai1 silencing on the proliferation and migration of many cancer cell lines [19, 22, 49, 50], and the lack of inhibition of SOCE with TRPC3 silencing in the MDA-MB-468 cells used in this study (Fig. S2). Our previous studies also showed that silencing of TRPC1, that is also expressed in MDA-MB-468 cells, did not inhibit SOCE in these cells [38]. Differential mechanisms of inhibition may have a greater impact on the effectiveness of store-operated Ca²⁺ entry inhibitors in different models; this could be through the duration of channel inhibition or what stimuli may be most affected by these inhibitors. Although YM58483 has been postulated to act via an extracellular site [51], this has not been fully explored and the mechanism of Synta66 is still unknown or at least unpublished [11]. As studies begin to further define the mechanism of action for these agents, the reasons for the differential effects of YM58483 and Synta66 observed in this study may become clearer. Unfortunately, very few studies have directly compared different store-operated Ca²⁺ entry inhibitors in the same model system as has been done in these studies. Our experiments did demonstrate that the SOCE inhibition mediated by YM58483 is enhanced to a much greater extent with increased pre-incubation time (as previously reported [52]) than Synta66, which to our knowledge has not been previously described (Fig. S1). Our work, therefore, demonstrates the importance of using multiple SOCE inhibitors in pharmacological studies. Such an approach should be used more widely at least until the mechanisms of these and other store-operated Ca²⁺ entry inhibitors are better understood.

In conclusion, these studies implicate store-operated Ca²⁺ entry as a differential contributor to global sustained increases in [Ca²⁺]_CYT induced by different stimuli in MDA-MB-468 basal A breast cancer cells. The use of more selective pharmacological inhibitors of store-operated Ca²⁺ entry than previously used in MDA-MB-231 basal B breast cancer cells, supports a role for such agents in the control of breast cancer cell migration in MDA-MB-468 breast cancer with some stimuli. However, these studies have also identified differences between YM58483 and Synta66 in EGF-induced migration and proliferation that require further investigation in other basal breast cancer cell lines and other model systems.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

ATP

Adenosine triphosphate

CM

Complete media

CPA

Cyclopiazonic acid

DMEM

Dulbecco’s modified Eagle’s medium

EGF

Epidermal growth factor

EGFR

Epidermal growth factor receptor

ER

Endoplasmic reticulum

FBS

Foetal bovine serum

IP₃

Inositol 1,4,5-trisphosphate

PAR2

Protease-activated receptor 2

PBS

Phosphate-buffered saline

PLC

Phospholipase C

SOCE

Store-operated Ca²⁺ entry

SRM

Serum-reduced media

STIM1

Stromal interaction molecule 1

STR

Short tandem repeat

Compliance with ethical standards

Conflict of interest

GR.M. and S.J.R.-T. are associated with QUE Oncology Inc.