Distinct Regulation of Cytoplasmic Calcium Signals and Cell Death Pathways by Different Plasma Membrane Calcium ATPase Isoforms in MDA-MB-231 Breast Cancer Cells*

Background: The roles of different PMCA isoforms are not fully understood particularly in cell death. Results: PMCA4 and PMCA1 silencing has differential effects on Ca2+ signaling and caspase-dependent and -independent cell death in MDA-MB-231 cells. Conclusion: PMCA isoforms have distinct roles in the control of cell death pathways. Significance: Inhibition of PMCA4 may increase the effectiveness of some cancer therapies. Plasma membrane calcium ATPases (PMCAs) actively extrude Ca2+ from the cell and are essential components in maintaining intracellular Ca2+ homeostasis. There are four PMCA isoforms (PMCA1–4), and alternative splicing of the PMCA genes creates a suite of calcium efflux pumps. The role of these different PMCA isoforms in the control of calcium-regulated cell death pathways and the significance of the expression of multiple isoforms of PMCA in the same cell type are not well understood. In these studies, we assessed the impact of PMCA1 and PMCA4 silencing on cytoplasmic free Ca2+ signals and cell viability in MDA-MB-231 breast cancer cells. The PMCA1 isoform was the predominant regulator of global Ca2+ signals in MDA-MB-231 cells. PMCA4 played only a minor role in the regulation of bulk cytosolic Ca2+, which was more evident at higher Ca2+ loads. Although PMCA1 or PMCA4 knockdown alone had no effect on MDA-MB-231 cell viability, silencing of these isoforms had distinct consequences on caspase-independent (ionomycin) and -dependent (ABT-263) cell death. PMCA1 knockdown augmented necrosis mediated by the Ca2+ ionophore ionomycin, whereas apoptosis mediated by the Bcl-2 inhibitor ABT-263 was enhanced by PMCA4 silencing. PMCA4 silencing was also associated with an inhibition of NFκB nuclear translocation, and an NFκB inhibitor phenocopied the effects of PMCA4 silencing in promoting ABT-263-induced cell death. This study demonstrates distinct roles for PMCA1 and PMCA4 in the regulation of calcium signaling and cell death pathways despite the widespread distribution of these two isoforms. The targeting of some PMCA isoforms may enhance the effectiveness of therapies that act through the promotion of cell death pathways in cancer cells.

Plasma membrane calcium (Ca 2ϩ ) ATPases (PMCAs) 2 actively remove cytoplasmic free Ca 2ϩ ([Ca 2ϩ ] CYT ) and are essential regulators of basal Ca 2ϩ (1). PMCAs also contribute to the duration and amplitude of Ca 2ϩ transients evoked by many physiological stimuli (1,2) and participate in the decoding of the Ca 2ϩ signal into a multitude of cellular processes (3).
Separate genes encode for each of the four PMCA isoforms (PMCA1-4), with alternative splicing of PMCA transcripts acting to further enhance their biological diversity (1,4). The identification of more than 30 PMCA variants (4) and studies in mice null for specific PMCA isoforms support distinct roles for specific PMCA isoforms. For example, PMCA2 null mice have hearing and balance disorders (5) and produce milk with reduced Ca 2ϩ levels (6). Differences in the consequences of PMCA knockdown are apparent even between PMCA isoforms with almost ubiquitous tissue distribution such as PMCA1 and PMCA4. PMCA1 knock-out is embryonically lethal in mice (5), whereas PMCA4 knock-out mice, although associated with male infertility, are generally healthy (5,7,8).
Studies have started to identify some of the potential mechanisms that may underpin the functions predicted for specific PMCA isoforms. One such example is the differential association of PMCA1, PMCA2, and PMCA4 with calcineurin in MCF-7 cells (9). The functional significance of this interaction has not been investigated in MCF-7 cells; however, in cardiac cells, PMCA4 regulates nuclear factor of activated T-cells (NFAT), a calcineurin-regulated transcription factor, and cardiac hypertrophy (10).
One process where the relative contribution of specific PMCA isoforms may be particularly important but has not been fully addressed is cell death, where the duration and amplitude of increases in [Ca 2ϩ ] CYT are critical (11,12). Initiators of cell death include Ca 2ϩ ionophores that produce sus-tained (M) increases in [Ca 2ϩ ] CYT that can induce necrotic cell death (13) and inhibitors of Bcl-2 protein (B-cell lymphoma-2) (14 -16). Bcl-2 inhibitors promote apoptosis by blocking the prosurvival activity of Bcl-2 proteins (14 -16) and are undergoing clinical trial assessment for the treatment of specific types of lymphomas and solid lung cancers (17).
A few studies have assessed the consequences of PMCA expression on cell death. In HeLa cells, PMCA overexpression protects from ceramide-induced cell death (18). The authors attributed this to alterations in intracellular Ca 2ϩ stores, given that PMCA overexpression in CHO cells reduces Ca 2ϩ levels within the endoplasmic reticulum and mitochondria (19). In the context of breast cancer, where cells acquire the ability to evade cell death (20), assessment of PMCA has been restricted to overexpression studies in T47D breast cancer cells (21). Recombinant PMCA2 expression attenuates increases in [Ca 2ϩ ] CYT associated with ionomycin-induced cell death (21).
Although PMCA2 knockdown has been assessed in spinal cord neurons (22) and SH-SY5Y neuroblastoma cells (23), and PMCA4 knockdown evaluated in HT-29 colon cancer cells (24), isoform-specific PMCA knockdown has not yet been studied in breast cancer cells. In addition, the differential role of PMCA isoforms in the regulation of distinct cell death mechanisms, within the same cell type has not been evaluated. In the present study we compared siRNA-mediated knockdown effects of PMCA1 and PMCA4 in the regulation of global Ca 2ϩ signals and assessed the effects on the viability of MDA-MB-231 breast cancer cells, in the absence and presence of the cell death initiators, ionomycin (Ca 2ϩ ionophore) and ABT-263 (Bcl-2 inhibitor). The basal breast cancer molecular subtype is associated with a particularly poor prognosis; therefore, MDA-MB-231 breast cancer cells were selected for these studies because they are a widely used model of the basal-like phenotype.

EXPERIMENTAL PROCEDURES
Cell Culture-The human breast cancer cell line MDA-MB-231 (American Type Culture Collection) was cultured in high glucose DMEM (Sigma Aldrich) supplemented with 10% FBS and 4 mM L-glutamine (Invitrogen) at 37°C/5% CO 2 in a humidified air incubator. All cultures were periodically screened for mycoplasma infection.
Real Time RT-PCR-Cells were plated at 5.0 ϫ 10 3 cells per well into 96-well plates and siRNA-transfected. Total RNA was isolated (RNeasy Plus mini kit; Qiagen) 24, 48, or 120 h post-transfection and then reverse transcribed using the Omniscript RT kit (Qiagen) according to the manufacturer's instructions. Target cDNA were amplified using the TaqMan Fast Universal PCR Master Mix (Applied Biosystems) and TaqMan Gene Expression assays (PMCA4, Hs00608066_m1; PMCA1, Hs00155949_m1) and standard cycling conditions with a StepOnePlus real time PCR system (Applied Biosystems). PMCA4 and PMCA1 mRNA levels were quantified by the comparative Ct method as described previously (27), normalizing to 18 S rRNA (4319413E) and presented relative to the siNT control.
Cytoplasmic Free Ca 2ϩ Measurements-MDA-MB-231 cells were plated at 7.5 ϫ 10 3 cells per well into 96-well plates and siRNA-transfected. Transfection media was removed 72 h post-siRNA treatment, and cells were loaded with culture medium containing Fluo-4 AM (4 M; Molecular Probes) or Fluo-4FF AM (4 M; Molecular Probes) and incubated at 37°C for 30 min in a 5% CO 2 -humidified air incubator. Loading solution was replaced with physiological salt solution (5.9 mM KCl, 1.4 mM MgCl 2 , 10 mM HEPES, 1.2 mM NaH 2 PO 4 , 5 mM NaHCO 3 , 140 mM NaCl, 11.5 mM glucose, 1.8 mM CaCl 2 ), cells were equilibrated to room temperature and then washed twice with a physiological salt solution containing bovine serum albumin (BSA, 0.3%). [Ca 2ϩ ] CYT was measured in Ca 2ϩ probe loaded cells using a fluorescence imaging plate reader (28) (FLIPR TETRA , Molecular Devices Corporation) using an excitation intensity of 470 -495 nm and a 515-575 nm emission filter. Fluorescence was normalized to base-line values to assess relative [Ca 2ϩ ] CYT .
Immunoblotting-Cells were plated at 5.0 ϫ 10 3 cells per well into 96-well plates and siRNA-transfected. Cell extracts were harvested 72 h post-transfection in protein lysis buffer supplemented with protease inhibitor mixture (Roche Applied Science) as described previously (24). Proteins were separated using gel electrophoresis and transferred onto a polyvinylidene fluoride membrane as described previously (29). Membranes were probed with monoclonal anti-PMCA4 antibody (1:1000, JA9, Pierce Antibodies), monoclonal anti-PMCA antibody (1:2000, 5F10, Pierce Antibodies) and monoclonal anti-␤-actin antibody (1:10,000, Sigma Aldrich). Anti-mouse horseradish peroxidase-conjugated secondary antibody (1:10000, Bio-Rad) was used to visualize protein bands by chemiluminescence using Super-Signal West Dura (Pierce). All antibodies were diluted in PBST (0.1% Tween 20 in PBS) with skim milk powder (2.5%). Images were acquired and analyzed by densitometry using a VersaDoc instrument and Quantity One Analysis software (Bio-Rad), respectively. All data were normalized to the ␤-actin loading control and are presented relative to siNT.
Assessment of Cell Viability-MDA-MB-231 cells were plated at 5.0 ϫ 10 3 cells per well into 96-well plates and siRNA-transfected. At 72 h post-transfection, medium was removed, and cells were treated with ABT-263 (Selleckchem), ionomycin (Enzo Life Sciences), Z-VAD-FMK (Enzo Life Sciences), IMD-0354 (Sigma-Aldrich), or dimethyl sulfoxide (up to 1%) for 48 h in phenol red-free DMEM containing FBS (8%). Live cells were stained at 37°C for 15 min with Hoechst 33342 (10 g/ml, Invitrogen) and propidium iodide (1 g/ml, Invitrogen). Images were acquired using an ImageXpress micro auto-mated epifluorescence microscope (Molecular Devices Corp.) with a 10ϫ objective, a system previously used for cell death assays (30 -32). Four images were automatically acquired per well in the DAPI and Cy3 channel for Hoechst 33342 and propidium iodide, respectively. The multiwavelength cell scoring application module (MetaXpress, version 3.1.0.83; Molecular Devices Corp.) was used to segment cell nuclei to calculate the average Hoechst 33342 integrated intensity and the average propidium iodide intensity emitted per cell. Criteria for viable, apoptotic and necrotic (and secondary apoptosis) are shown graphically in figures and validated using the caspase inhibitor Z-VAD-FMK.
Analysis of PMCA1 and PMCA4 Levels in Published Data Sets-Gene expression data for breast cancer cell lines in threedimensional culture (36) or for breast tumors (37) were obtained from online repositories (ArrayExpress accession no. E-TABM-244 (cell lines) and NCBI GEO accession no. GSE2034 (tumors)). Affymetrix probes for PMCA1 (215716_s_at) and PMCA4 (212136_at) were shared between these data sets. Samples were stratified based on their reported annotations (basal versus luminal for cell lines, and ERϪ versus ERϩ for tumors). Normalized gene expression levels (log 2 ) were plotted, and median levels of each gene were compared between groups (Mann-Whitney test). For relapse-free survival analysis, tumors were stratified into groups corresponding to the upper and lower quartiles and the interquartile range for Kaplan-Meier analysis.
Statistical Analysis-All statistical tests were performed as described in the figure legends using GraphPad Prism (version 5.04) for Windows.

RESULTS
siRNA-mediated Knockdown of Specific PMCA Isoforms in MDA-MB-231 Breast Cancer Cells-There was a significant (p Ͻ 0.05) reduction in PMCA4 (Fig. 1A) and PMCA1 (Fig. 1B) mRNA levels in cells transfected with PMCA4 siRNA or PMCA1 siRNA, respectively, compared with the siNT control. PMCA4 protein expression was also significantly (p Ͻ 0.05) reduced by PMCA4 siRNA (Fig. 1, C and D) but was not associated with significant changes in total PMCA protein expression ( Fig. 1, C and E). Consistent with high levels of PMCA1, total PMCA protein levels were reduced by PMCA1 siRNA (Fig. 1, C and E; p Ͻ 0.05). This was not associated with changes in PMCA4 expression (Fig. 1, C and D).
Consequences of PMCA1 or PMCA4 Knockdown on [Ca 2ϩ ] CYT in MDA-MB-231 Breast Cancer Cells-In the absence of extracellular Ca 2ϩ , increases in [Ca 2ϩ ] CYT elicited by the sarcoplasmic-endoplasmic reticulum Ca 2ϩ -ATPase inhibitor cyclopiazonic acid (CPA, 10 M) were markedly altered with siRNA-mediated knockdown of PMCA1 but not by PMCA4 knockdown (Fig. 2, A-D). PMCA1 silencing significantly (p Ͻ 0.05) delayed the time to reach the maximal Ca 2ϩ response (Fig. 2B), increased the half-peak decay time (Fig. 2C) and increased the area under the curve (Fig. 2D) compared with siNT. The effect was isoform-specific, as PMCA4 knockdown did not alter the nature of the CPA-induced Ca 2ϩ response when compared with siNT (Fig. 2, A-D).  (Fig. 2F), a delay in half-peak decay time (Fig. 2G) and an increase in the area under the curve (Fig. 2H). These effects were isoformspecific as similar changes were not seen with PMCA4 silencing (Fig. 2, E-H). To assess the consequences of isoform-specific PMCA knockdown with [Ca 2ϩ ] CYT increases of a greater magnitude, we compared the effects of PMCA4 or PMCA1 siRNA on [Ca 2ϩ ] CYT changes mediated by the Ca 2ϩ ionophore ionomycin in the presence of extracellular Ca 2ϩ (1.8 mM).
Increases in [Ca 2ϩ ] CYT mediated by ionomycin (3 M) were significantly (p Ͻ 0.05) increased with PMCA1 or PMCA4 knockdown compared with siNT (Fig. 3, A and B). These effects were more pronounced (p Ͻ 0.05) with PMCA1 knockdown compared with PMCA4 knockdown (Fig. 3, A and B). Silencing of PMCA1 or PMCA4 also augmented increases in [Ca 2ϩ ] CYT mediatedby 10 M ionomycin; however, in contrast to 3 M ionomycin, there were no differences in effect between the two isoforms (Fig. 3, C and D).
Assessment of PMCA1 or PMCA4 Knockdown on Cell Viability in MDA-MB-231 Breast Cancer Cells-No effect on cell viability was detected with knockdown of PMCA4 (Fig. 4, A, B, and D) or PMCA1 (Fig. 4, A, C, and D) compared with siNT. PMCA1 and PMCA4 were effectively silenced at the time of the assessment of cell viability (Fig. 4, E and F). Reduced Ca 2ϩ efflux mediated by either PMCA1 or PMCA4 silencing was therefore insufficient for the activation of cell death pathways in MDA-MB-231 cells.
Mechanism of Ionomycin and ABT-263 Activated Cell Death in MDA-MB-231 Cells-To assess whether reduced PMCA1 or PMCA4 expression could alter the responsiveness of breast cancer cells to different cell death mechanisms, we first characterized cell death in MDA-MB-231 cells initiated by the Ca 2ϩ ionophore ionomycin (Fig. 5, A and C) and the Bcl-2 inhibitor, ABT-263 (Fig. 5, A and E). Ionomycin (10 M) and ABT-263 (10 M) both initiated significant (p Ͻ 0.05) increases in cell death compared with control (Fig. 5, G and H). The nature of the initiated cell death was probed using the pan-caspase inhibitor, Z-VAD-FMK (50 M). Consistent with a necrotic mechanism (13) ionomycin-induced (10 M) cell death was not affected by Z-VAD-FMK (Fig. 5, C, D, and G), whereas cell death initiated by ABT-263 was almost entirely prevented by caspase inhibition (Fig. 5, E, F, and H), indicating an apoptotic pathway (38).
Consequences of PMCA1 or PMCA4 Silencing on Ionomycininduced Necrosis in MDA-MB-231 Breast Cancer Cells-Necrosis mediated by submaximal ionomycin (3 M) was not significantly affected by PMCA4 knockdown (Fig. 6, A, B, and G); however, it was significantly (p Ͻ 0.05) augmented with PMCA1 silencing (Fig. 6, A, C, and G). Additionally, the level of necrosis was significantly (p Ͻ 0.05) higher in cells treated with PMCA1 siRNA compared with PMCA4 siRNA (Fig. 6G) at 3 M.
For the higher ionomycin concentration (10 M), knockdown of both PMCA4 (Fig. 6, D, E, and G) and PMCA1 (Fig. 6,  D, F, and G) significantly (p Ͻ 0.05) elevated the proportion of necrotic cells compared with siNT. However, similar to their effects on [Ca 2ϩ ] CYT , the isoform-specific differences seen with submaximal ionomycin concentration (3 M) were not detected at higher Ca 2ϩ loads associated with 10 M ionomycin (Fig. 6G).

Consequences of PMCA1 or PMCA4 Silencing on ABT-263induced Apoptosis in MDA-MB-231 Breast Cancer Cells-Sub-
maximal ABT-263-induced (3 M) apoptosis was significantly (p Ͻ 0.05) augmented by PMCA4 siRNA compared with siNT ( Fig. 7, A, B, and G). This effect was isoform-specific, as PMCA1 knockdown did not alter apoptosis compared with siNT (Fig. 7,  A, C, and G). In contrast, ABT-263-mediated (10 M) cell death was not modulated by either PMCA4 siRNA or PMCA1 siRNA compared with siNT (Fig. 7, D-G). (39 -41), which is implicated in breast cancer progression (42) and the resistance to cancer therapies (43,44). PMCA4 is linked to the regulation of Ca 2ϩ -dependent transcription factors such as NFAT, independent of bulk [Ca 2ϩ ] CYT changes (10). We assessed the effects of PMCA4 and PMCA1 silencing on nuclear translocation of NFB induced by PMA (50 nM) (33). PMA-induced NFB nuclear translocation (Fig. 8, A and B) was not affected by PMCA1 silencing but was significantly (p Ͻ 0.05) reduced by PMCA4 siRNA relative to siNT.
PMCA1 and PMCA4 in Clinical Breast Cancers-To determine the distribution of PMCA1 and PMCA4 in breast cancer, we first examined their expression in a panel of breast cancer cell lines grown in three-dimensional culture (36). Expression varied widely between cell lines, with variation between the highest and lowest measurements being 18.9-and 12.8-fold for PMCA1 and PMCA4, respectively (Fig. 9, A and B). PMCA1 expression measurements were found throughout this range in both luminal and basal cell lines (Fig. 9A); however, PMCA4 was most highly expressed in the basal cell lines (Fig. 9B, p ϭ 0.007). We next evaluated the level of these genes in a cohort of 286 node-negative breast cancer patients described by Wang and co-workers (37). Consistent with the non-significant trend suggested in the cell line data (Fig. 9A), PMCA1 was expressed at higher levels in the ERϩ tumors (Fig. 9C, p ϭ 0.007). Median PMCA4 levels were higher in the ER-negative tumors (Fig. 9D, p ϭ 0.028). No differences were observed in patient outcomes when tumors were stratified by PMCA1 (Fig. 9E, p ϭ 0.92) or PMCA4 (Fig. 9F, p ϭ 0.76) levels.

DISCUSSION
Despite the proposed differential contributions of PMCA isoforms in Ca 2ϩ -dependent cellular processes, few studies have evaluated the specific roles of different PMCA isoforms in shaping intracellular Ca 2ϩ signals and cellular responses in the same cell types. PMCA1 and PMCA4 have a broad tissue distribution and frequently co-express within the same cell (46). We compared the consequences of knockdown of these two PMCA isoforms on [Ca 2ϩ ] CYT signals generated by various Ca 2ϩ -mobilizing agents. Pronounced effects with PMCA1 knockdown but not with PMCA4 knockdown were seen on the nature of CPA-and ATP-induced [Ca 2ϩ ] CYT transients. PMCA1 siRNA reduced the rate of [Ca 2ϩ ] CYT decay and prolonged the ATP Ca 2ϩ response. Previous studies overexpressing PMCA isoforms in CHO cells found that some variants not only alter Ca 2ϩ homeostasis within the cytoplasm but also within subcellular compartments such as the endoplasmic reticulum (2). Increases in [Ca 2ϩ ] CYT mediated by sarcoplasmic endoplasmic reticulum Ca 2ϩ -ATPase inhibitors have been reported to correlate with the calcium content of the endoplasmic reticulum (47). Hence, increases in the CPA-mediated [Ca 2ϩ ] CYT transient upon PMCA1 silencing, in addition to being a consequence of reduced Ca 2ϩ efflux, may also involve increases in endoplasmic reticulum Ca 2ϩ levels. Our study suggests that in MDA-MB-231 breast cancer cells, the PMCA1 isoform is the major regulator of global [Ca 2ϩ ] CYT increases, such as those generated by phosphatidylinositol 1,4,5-trisphosphate-mediated Ca 2ϩ release after G proteincoupled receptor activation. The significance of PMCA4 in MDA-MB-231 cells may become apparent during very high Ca 2ϩ loads. Indeed, during the high magnitude increases in [Ca 2ϩ ] CYT initiated by 3 M ionomycin, a contribution for PMCA4 was identified. However, this PMCA4 siRNAmediated effect was not as pronounced as seen with PMCA1 knockdown. This finding is also consistent with a predominant role for PMCA1 in the regulation of global [Ca 2ϩ ] CYT signals in this cell type and is consistent with previous studies, demonstrating distinct phenotypes in PMCA1 and PMCA4 knock-out animals (5, 7, 8). PMCA1 ablation is lethal during embryogenesis, indicative of its vital role in the maintenance of [Ca 2ϩ ] CYT  homeostasis in an array of cell types (5). Although PMCA4 has a broad tissue distribution, PMCA4 null mice reach adulthood and exhibit tissue-specific phenotypes such as male infertility (5,7,8). These differences suggest that PMCA1 adopts a vital housekeeping function by regulating global Ca 2ϩ homeostasis and supports the involvement of PMCA4 in Ca 2ϩ -dependent signal transduction and cell processes by shaping of Ca 2ϩ signals within subcellular domains (5,7,8).
We extended our study to examine the consequences of PMCA4 and PMCA1 siRNA on cell death in the presence of cell death stimuli. Initial studies, in the absence of any stimuli, showed that the viability of MDA-MB-231 breast cancer cells was not affected by PMCA1 or PMCA4 knockdown. In contrast, PMCA2 knockdown produces spinal cord neuronal cell death (22,48) in the absence of external stimuli. Assessment of PMCA1 and PMCA4 knockdown on cell death initiated by ionomycin or ABT-263 demonstrated distinct roles for PMCA1 and PMCA4 in the regulation of caspase-independent and -dependent cell death pathways, respectively. Consistent with the effects seen with ionomycin-induced [Ca 2ϩ ] CYT transients, PMCA1 knockdown had a more pronounced effect on ionomycin-induced cellular necrosis than PMCA4 knockdown. This result further underscores the more subtle nature of the change in calcium handling in the cell with PMCA4 knockdown. Previous results examining PMCA4 as a sensitizer of cell death in HT-29 colon cancer cells showed that PMCA4 silencing does not alter the sensitivity of HT-29 cells to tumor necrosis factorrelated apoptosis inducing ligand (TRAIL) or carbonyl cyanide 3-chlorophenylhydrazone (CCCP)-induced cell death (24). Reduced expression of another PMCA isoform, PMCA2, augments ionomycin-mediated cell death in SH-SY5Y neuroblastoma cells (23), whereas its overexpression in T47D breast cancer cells bestows resistance to ionomycin-mediated cell death through the attenuation of [Ca 2ϩ ] CYT responses (21). The  potential significance of this survival advantage is reflected in the poorer prognosis of breast cancer patients with elevated expression of PMCA2 (21). Our silencing studies suggest that modulators of PMCA isoforms involved in global [Ca 2ϩ ] CYT may sensitize cells to cell death stimuli.
The Bcl-2 inhibitor (ABT-263) used in this study to induce apoptotic cell death in MDA-MB-231 breast cancer cells is progressing through clinical trials (17). A structurally related analog (ABT-737) sensitizes Bcl-2-expressing breast cancers to chemotherapies (49), highlighting the potential for Bcl-2 inhibitors as a therapeutic option for breast cancer. ABT-263-induced cell death was not affected by PMCA1 knockdown but instead was augmented upon reduced expression of PMCA4. These distinct differences in the consequences of isoform-specific PMCA knockdown on ionomycin and ABT-263 initiated cell death is not totally unexpected, considering these agents activate cell death by distinct mechanisms and have markedly different effects on Ca 2ϩ signals. Ionomycin produces sustained global increases in [Ca 2ϩ ] CYT , augmented by PMCA1 siRNA, resulting in Ca 2ϩ overload and cell necrosis. Bcl-2 inhibitors such as ABT-263 initiate caspase-dependent apoptosis by binding to and blocking the prosurvival activity of BH3 domains present within Bcl-2 proteins (14 -16). Increases in intracellular Ca 2ϩ signals, particularly within subcellular compartments, such as the endoplasmic reticulum and mitochondria can modulate Bcl-2-mediated survival pathways (18,50,51). Our identification of PMCA4 as a modulator of ABT-263mediated cell death may be analogous to the characterized role of PMCA4 in cardiac cells, where altered PMCA4 expression appears to play little role in shaping global [Ca 2ϩ ] CYT increases (10,52) yet is an important regulator of the Ca 2ϩ -dependent transcription factor NFAT influencing outcomes such as cardiac hypertrophy (9,10). Ca 2ϩ -dependent gene transcription critically depends on localization of Ca 2ϩ signals as well as Ca 2ϩ oscillation frequency and amplitude (39,40). Thus, PMCA4mediated regulation of transcription may be mediated through the fine-tuning of Ca 2ϩ signals within localized subcellular domains or through alterations in oscillation frequency. Context-dependent modulation of apoptosis by PMCAs is reflected in PMCA4 knock-out mice studies. Smooth muscle cells from the portal veins of PMCA4 knock-out mice on a mixed 129/SvJ and Black Swiss background display features of apoptosis during in vitro contraction studies (7).
Our study identified PMCA4 siRNA-mediated inhibition of NFB nuclear translocation. Pharmacological inhibition of NFB phenocopied the augmentation of ABT-263-mediated apoptosis produced by PMCA4 silencing, suggesting that PMCA4 siRNA augmentation of apoptosis may be due to modulation of NFB. The ability of PMCA4 silencing to inhibit NFB is significant. The activity of NFB is governed by the nature of Ca 2ϩ signals (39 -41), and Ca 2ϩ influx mediated by the calcium channel TRPC1 can inhibit NFB activity in an intestinal epithelial cell line (53). Breast cancers with a poor prognosis are associated with elevated constitutive activity of NFB (43,44). Agents that inhibit NFB are themselves promising anti-tumor modulators for the treatment of breast cancers (54, 55) and can enhance the effects of Bcl-2 inhibitors (56). Our studies comparing PMCA1 and PMCA4 in clinical samples suggest that although inhibition of PMCA4 in basal breast cancer may be an effective way to augment responsiveness to Bcl-2 inhibitors in breast cancer therapy, increases in PMCA4 (or PMCA1) expression are not a cause of general apoptotic resistance in breast cancer, as there was no association between PMCA4 (or PMCA1) levels and prognosis.
In summary, we show that PMCA1 is a major regulator of global Ca 2ϩ homeostasis in MDA-MB-231 breast cancer cells and this is associated with an ability of PMCA1 silencing to augment necrotic cell death generated by high Ca 2ϩ loads. Although, PMCA4 is not a key regulator of global changes in [Ca 2ϩ ] CYT associated with many stimuli, our study demon-strates a novel relationship between PMCA4 and NFB. Our study highlights isoform diversity between the two almost ubiquitously expressed PMCA isoforms (PMCA1 and PMCA4) and identifies PMCA4 as a potential anti-tumor modulator. Inhibitors of PMCA4 (57) may be novel therapeutics to sensitize some cancer cells to apoptotic stimuli.