Elevation of Cytoplasmic Calcium Suppresses Microtentacle Formation and Function in Breast Tumor Cells

Simple Summary Calcium is a versatile and ubiquitous signaling molecule that long-term dysregulation can increase the spread of cancer to various parts of the body but that short-term effects are understudied. Disseminated cancer cells in circulation have distinct extensions or protrusions, called microtentacles, that enhance their ability to attach to surfaces or other cells. In this study, we show rapidly increasing cellular calcium with the compounds of Ionomycin and Thapsigargin decreases the microtentacle frequency and clustering functions on cancer cells in a detached and suspended environment. Acute calcium-induced signaling events promoted changes to actin contraction and rearrangement responsible for suppressing microtentacles. The results from this study support clinical trial data from Thapsigargin derivatives, suggesting Ca2+ modulating therapies can potentially be used to promote cellular shape and structure changes in free-floating tumor cells to reduce metastasis. Abstract Cytoskeletal remodeling in circulating tumor cells (CTCs) facilitates metastatic spread. Previous oncology studies examine sustained aberrant calcium (Ca2+) signaling and cytoskeletal remodeling scrutinizing long-term phenotypes such as tumorigenesis and metastasis. The significance of acute Ca2+ signaling in tumor cells that occur within seconds to minutes is overlooked. This study investigates rapid cytoplasmic Ca2+ elevation in suspended cells on actin and tubulin cytoskeletal rearrangements and the metastatic microtentacle (McTN) phenotype. The compounds Ionomycin and Thapsigargin acutely increase cytoplasmic Ca2+, suppressing McTNs in the metastatic breast cancer cell lines MDA-MB-231 and MDA-MB-436. Functional decreases in McTN-mediated reattachment and cell clustering during the first 24 h of treatment are not attributed to cytotoxicity. Rapid cytoplasmic Ca2+ elevation was correlated to Ca2+-induced actin cortex contraction and rearrangement via myosin light chain 2 and cofilin activity, while the inhibition of actin polymerization with Latrunculin A reversed Ca2+-mediated McTN suppression. Preclinical and phase 1 and 2 clinical trial data have established Thapsigargin derivatives as cytotoxic anticancer agents. The results from this study suggest an alternative molecular mechanism by which these compounds act, and proof-of-principle Ca2+-modulating compounds can rapidly induce morphological changes in free-floating tumor cells to reduce metastatic phenotypes.


Kinetic Calcium Assay
A collagen layer was coated rotating overnight at 4 • C on a clear bottom 96-well black plate (Corning 3603) as previously described [21]. Cells were then plated in the prepared plate at 50,000 cells per well for a confluent monolayer. Cells were then loaded with 4 µM Fluo4-AM (Life Technologies F14201) and prepared as previously described [21] prior to the addition of compound. The final volume per well was 100 µL of Hank's Balanced Salt Solution (HBSS + Ca 2+ ; Gibco 14025-092).
Dilutions of compounds were prepared in a separate plate at 5× concentration for the robotic addition of 25 µL to cells by the FLEX Station 3 Multi-Mode plate reader (Molecular Devices). The FLEX Station 3 reads Relative Fluorescence Units (RFUs) measured every 1.28 s for a period of 300 s after compound addition. The results were either plotted as the maximum-minimum for each compound concentration over the time period or ∆F/F 0 as previously calculated [21] with the initial 30 s reading as a baseline value prior to compound addition. All values shown are mean ± SD of triplicate samples.

Cell Viability
White 96-well tissue culture plates (Greiner Bio-One 655180) were PEM coated with a single bilayer and formaldehyde crosslinkedas previously described [24]. Cells were seeded in triplicate at 5000 cells per well in culture media in the presence of DMSO, Staurosporine (Sigma S6942 (Stauro)) or select concentrations of Iono, or Tg for 0, 6, 24 h. After respective drug treatment times, Promega Cell titer Glo (G7571), was added. Cells were prepared following the manufacturer's instructions, and luminescence was read on the FLEX Station 3 (Molecular Devices). Viability was measured as a percentage of ATP production at time 0 ([luminescence at hour x ÷ average luminescence at time 0] × 100). All values shown are mean ± SD of triplicate samples.

McTNs Scoring and Analysis
Cells were trypsinized and suspended in DMEM in the presence of DMSO, Iono or Tg for 30 min while in suspension, and then tethered, fixed, and, stained on our TetherChip Technology as previously described [25,26]. For zero Ca 2+ conditions, cells were initially washed in zero Ca 2+ Hank's Balanced Salt Solution (HBSS−Ca 2+ ; Gibco 1424-092) then suspended in HBSS−Ca 2+ supplemented with 100 µM ethylene glycol-bis(β-aminoethylether)-N,N,N',N'-tetra acetic acid (EGTA) (Sigma E3889) alone or with Iono or Tg. For inhibition of actin polymerization experiments, detached and suspended cells were subjected to an initial 15 min pretreatment of 5 µM Latrunculin A (LA) (Emdmillipore 428021) prior to the addition of DMSO, Iono or Tg. Samples were treated for 30 min in suspension then tethered, fixed, and stained as previously described [25,26]. Blind scoring for the presence of McTNs in a given population of 100 cells per channel was conducted. Cell positivity was defined by a blinded individual observer as to whether the suspended cells are producing at least two membrane protrusions greater than the radius of the cell body [25]. Additionally, single whole cell perimeter analysis was performed, as previously described [25,26].

Cellular Clustering Assay
In the presence of Iono, Tg or DMSO, cells were allowed to aggregate for 6 h in a 96-well low-attached plate (50,000 cells/well) then transferred to our TetherChip Technology [26]. Cells were fixed using 4% formaldehyde and stained with Hoechst 3325 (1:1000). Images were acquired using the Nikon Eclipse Ti2-E inverted microscope with a 4× air objective. Images were auto-contrasted in Nikon Elements software before being analyzed as previously described (https://github.com/ScientistRachel/CellAggregationAnalysis) accessed on 13 December 2021 [27].

Cell-Electrode Impedance Reattachment Assay
Real-time monitoring of cellular reattachment of suspended cells was measured using the xCelligence RTCA-DP real-time sensing device (Agilent Technologies, Santa Clara, CA, USA) to compare attachment rates of cells treated in DMSO, Iono or Tg.
MDA-MB-231 cells and MDA-MB-436 cells were grown to 80% confluency in a 10 cm tissue culture dish or a 6-well tissue culture plate. Both cell lines were then detached and seeded at 40,000 cells per well. Cell impedance was recorded every 5 min over a 24 h time course.
Live cell imaging of precipitation, attachment and spreading of cells in suspension attaching to the bottom of a 12-well tissue culture-treated plate was taken using the Nikon Eclipse Ti2-E inverted microscope with a Tokai-Hit stage top incubation chamber. Population images were collected every hour over a 17 h time course at a 10× air objective with phase contrast.

Confocal Microscopy
All confocal imaging was conducted on tethered and fixed cells using an Olympus IX81 microscope with a FV-1000 confocal laser scanning system with a 60× oil emersion magnification and objective lens numerical aperture of 1.42. Z-stack sliced images were taken at 0.5 µm slice along the entire thickness of each cell.

Statistical Analysis
Statistical analyses were conducted using either a t-test, or one-way ANOVA with Bonferroni's multiple comparison test in GraphPad Prism 9.0 software; p < 0.05 was considered significant. Outlier data in the clustering efficiency analysis were identified using the ROUT method with Q = 1% in GraphPad Prism 9.0 software. Biological replicates with an identified outlier were not included in the paired t-test for cellular clustering efficiency analysis.

Increasing Concentrations of Ionomycin or Thapsigargin Induces an Elevation of Cytoplasmic Calcium in Breast Cancer Cells
To begin to understand how cytoplasmic Ca 2+ affects McTN production, we initially validated whether the compounds Ionomycin (Iono) and Thapsigargin (Tg), a, induce an elevation in cytoplasmic Ca 2+ in the metastatic breast cancer cell lines, MDA-MB-231 and MDA-MB-436. Iono facilitates the movement of extracellular Ca 2+ ions across the plasma membrane into the cytoplasm, while Tg diffuses across the plasma membrane to inhibit the reuptake of Ca 2+ into the endoplasmic reticulum thereby increasing cytoplasmic Ca 2+ [22,23,29]. Using both compounds to increase cytoplasmic Ca 2+ concentrations, we can further test the impact of Ca 2+ entry from different sources on cancer cell functions and phenotypes in detached and suspended conditions in future assays.
A standard Ca 2+ dye assay was used to measure real-time changes in fluorescence measured as a Relative Fluorescence Unit (RFU). ∆F/F 0 was calculated from the recorded RFU values and shown as ∆F/F 0 traces over a 300 s time course. After an initial 30 s baseline reading, increasing concentrations of Iono or Tg were added to an adherent monolayer of cells. Increasing concentrations of Iono added to MDA-MB-231 cells caused a sharp and sustained increase in ∆F/F 0 in a dose-dependent manner ( Figure 1A). The difference between the minimum RFU value recorded and the maximum RFU value was used to create the EC50 Ca 2+ response curve. In the MDA-MB-231 cells, 5 µM Iono was the lowest concentration to elicit a maximal Ca 2+ response ( Figure 1B). Since the minimal or maximal reading is not necessarily seen at the beginning or end of each reading, the terminal ∆F/F 0 value was analyzed to determine potential differences between the concentrations. While 20 µM Iono had the greatest terminal ∆F/F, when compared to the 5 µM Iono, there was no statistical difference ( Figure 1C).
While McTN extension can be measured in less than an hour, the phenotypic consequences of initial McTN extensions are measured with longer-term assays such as cell clustering (6 h) or cell reattachment (24 h). A cell viability study was conducted to determine the potential toxicity of compounding variables, such as the nonadherent environment and the various compound concentrations capable of triggering a cytoplasmic Ca 2+ increase. Cell viability was calculated by measuring luminescence at 0, 6, and 24 h time points. MDA-MB-231 cells were seeded into increasing concentrations of 1, 5, and 20 µM Iono. Treatment with 20 µM of Iono was shown to be extremely cytotoxic at 6 and 24 h, while 1 µM Iono treatment had limited cytotoxic effects at the time points of interest ( Figure 1D). The concentration of interest, 5 µM Iono, also showed limited cytotoxic effects at 6 and 24 hours ( Figure 1D). Therefore, the optimal concentration for Iono used for further investigation was 5 µM.
The Ca 2+ assay and cytotoxicity experiments were repeated to determine the optimal concentration of Tg needed to trigger an increase in cytoplasmic Ca 2+ at low cytotoxic levels. Using the same Ca 2+ dye-based assay, increasing concentrations of Tg were added to MDA-MB-231 cells. Graphical ∆F/F0 traces illustrate a gradual rise in cytoplasmic Ca 2+ over time. Increasing additions of Tg show that 2 µM and 10 µM Tg have overlapping ∆F/F0 traces that indicate both concentrations achieve a maximal response ( Figure 1E). The EC50 Ca 2+ response curve also shows the addition of 2 µM Tg was sufficient to elicit a maximal Ca 2+ response similar that of 10 µM Tg ( Figure 1F). Additional analysis of the calculated end ∆F/F values found no statistical difference between 10 µM Tg and 2 µM Tg ( Figure 1G). Cytotoxicity experiments showed increased concentrations of Tg had limited cytotoxic effects at 6 and 24 hours ( Figure 1H). Collectively, these data indicate 2 µM Tg is the optimal concentration to use for subsequent studies.
Breast cancer is a heterogeneous disease with diverse genetic backgrounds, resulting in differing functional and biological effects. Therefore, a second metastatic breast cancer cell line, MDA-MB-436 cells, was tested to determine the minimal doses of Iono and Tg to elicit the maximum cytoplasmic Ca 2+ increase without affecting cell viability. The MDA-MB-436 cells demonstrated different Ca 2+ kinetic responses to the addition of Iono and Tg from that

Distinct Calcium Entry Sources Result in an Inhibited Reattachment Response to Surfaces
We have previously demonstrated that elevated levels of McTNs in free-floating cells enhance reattachment to surfaces [7,28,30]. Therefore, we hypothesized the suppression of

Increasing Cytoplasmic Calcium Decreases Homotypic Cellular Clustering
The ability of cancer cells to cluster/aggregate together plays a vital role in metastatic progression. Therefore, to determine whether clustering was impacted by the increase in cytoplasmic Ca 2+ , cells were directly plated in a nonadherent environment with either 5 µM Iono, 2 µM Tg, or vehicle control and allowed to cluster for 6 h. Quantification of homotypic clustering efficiency was determined using a custom MATLAB script optimized to the area of a single nuclei of 75 µM 2 [27]. Clustering efficiency was determined by the number of nuclei aggregates detected at time 0 divided by the number of nuclei aggregates at 6 h. MDA-MB-231 cells treated with 5 µM Iono showed a reduction in clustering efficiency, but do not reach statistical significance (p = 0.0547). However, treatment with 2 µM Tg did significantly reduce clustering efficiency ( Figure 5A,B). Representative images of MDA-MB-231 cells stained with nuclear staining visualize the initial and final seeding density and clusters comparing the vehicle control group and the treatment group at the end time points (t = 0H and t = 6H) are shown in Figure 5C,D ( Figure S1). In the MDA-MB-436 cells, 5 µM Iono and 2 µM Tg treatment significantly decreased clustering efficiency over the 6 h time course (p = 0.0053 and p = 0.0009) ( Figure 5E,F). Images of MDA-MB-436 cells stained with Hoechst illustrate the initial and final seeding density and clusters of the treatment group and the vehicle control group at the time points t = 0H and t = 6H ( Figures 5G,H and S2). As previously shown, a decrease in McTN formation and the ability of cells to reattach to surfaces over time with Iono or Tg treatment, both treatments are effective in reducing the ability of detached MDA-MB-231 and MDA-MB-436 cells to cluster over time.

Elevated Cytoplasmic Calcium Concentration Induces Actin Contraction and Rearrangement
The specific mechanisms that underlie cytoplasmic Ca 2+ -mediated McTN suppression was next investigated. McTNs are microtubule-based structures that are stabilized by their post-translational modifications (PTM) of acetylation at lysine 40 and detyrosination on α-tubulin. We initially investigated whether changes in expression of these PTMs contributed to McTN suppression. Protein analysis of acetylation and detyrosination in cells treated with either 5 µM Iono or 2 µM Tg did not reach statistical significance ( Figures 6A,B, S3A,B and S4A,B). These results suggest the loss of tubulin PTMs known to support McTNs is not the mechanism of McTN suppression by elevated cytoplasmic Ca 2+ .   by their post-translational modifications (PTM) of acetylation at lysine 40 and detyrosination on α-tubulin. We initially investigated whether changes in expression of these PTMs contributed to McTN suppression. Protein analysis of acetylation and detyrosination in cells treated with either 5 μM Iono or 2 μM Tg did not reach statistical significance (Figures 6A,B, S3A,B and S4A,B). These results suggest the loss of tubulin PTMs known to support McTNs is not the mechanism of McTN suppression by elevated cytoplasmic Ca 2+ .  Given that the PTMs of tubulin remained unchanged by Iono or Tg treatment, we next investigated the role of the actin network. However, evidence from the literature suggests that calcium-calmodulin dependent myosin light chain kinase (MLCK) activity and expression regulates actin rearrangement. To determine whether this mechanism is conserved in breast cancer cells, we first verified the basal expression of MLCK, myosin light chain 2 (MLC2), and phospho-myosin light chain 2 (p-MLC2). We next probed for p-MLC2 at serine 19 (S19), an indicator of actin cortex contraction, to interrogate if increasing cytoplasmic Ca 2+ stabilized the actin cortex. An initial time course was performed to determine the maximal effect seen. Five minutes of either compound treatment was sufficient to show the greatest change in expression. Immunoblot images of MDA-MB-231treated cells with either 5 µM Iono or 2 µM Tg show an increase in phosphorylation at S19 for MLC2 in comparison to the vehicle control, however, densitometry analysis of three biological replicates do not reach statistical significance (Figures 6C and S3C). Immunoblot images of MDA-MB-436 cells samples treated with either 5 µM Iono or 2 µM Tg show an increase in phosphorylation on MLC2 on S19; however, a variation of the total MLC2 between each sample within a biological replicate challenges any conclusion that can be drawn from these results (Figures 6D and S4C). On the other hand, actin cortex contractility is also regulated through myosin phosphatase1 (MYPT1) activity. Phosphorylation at threonine 853 (T853) on MYPT1 is the inactive state of MYPT1 that is an additional indicator of actin cortex contraction. The immunoblot images of three biological replicates of MDA-MB-231 and MDA-MB-436 cells show a trending increase in phosphorylation of MYPT1 with Iono or Tg treatment in comparison to the vehicle, but densitometry analysis does not reach statistical significance (Figures 6C,D, S3D and S4D).

Calcium-Induced Microtentacle Suppression Requires Actin Polymerization
Given the rapid phosphorylation events regulating actin turnover observed in the immunoblot analysis, we next assessed the necessity of filamentous actin polymerization for Ca 2+    Orthogonal whole cell perimeter measurements from confocal images showed MDA-MB-231 cells treated with 5 µM LA have a significantly larger perimeter than the vehicle ( Figure 7D). 5 µM Iono or 2 µM Tg treatment in the MDA-MB-231 cells also duplicated previous results showing a significant decrease in the perimeter compared to the vehicle ( Figure 7D). MDA-MB-231 cells initially treated with 5 µM LA before Ca 2+ flux stimulation by either 5 µM Iono or 2 µM Tg increased whole cell perimeter, but only 2 µM Tg treatment after 5 µM LA pretreatment achieved statistical significance ( Figure 7D). The overall results for whole cell perimeter measurements in the MDA-MB-436 cells trended in the same directions under each condition but did not reach statistical significance ( Figure 7F).

Discussion
The impact of acute Ca 2+ -mediated signaling pathways is well established in many organ systems and cell types, however, its role in tumor biology remains a knowledge gap. Early work using two-dimensional adherent cell culture models showed that chemical and mechanical-induced rapid Ca 2+ signaling differs between breast epithelial and breast cancer cells [19][20][21], suggesting dysregulation of acute Ca 2+ signaling mechanisms in cancer. These studies generated new questions about the contributions of acute Ca 2+ to cancer cell morphologies and phenotypes. Of notable interest is the role of Ca 2+ signaling in the dynamic arrangement of actin and tubulin in the nonadherent environment that can produce the McTN metastatic phenotype.
Currently, little is known about the physiological relevance of Ca 2+ signaling on the cytoskeleton in nonadherent models. Our lab has previously demonstrated the mechanisms of McTN formation through the examination of tubulin and actin dynamics in breast cancer cells in a nonadherent state. We have shown an increased frequency of McTNs through actin depolymerization with Cytochalasin-D or Latrunculin A, while McTN frequency decreased after treatment with tubulin depolymerizers such as Colchicine or Vinorelbine [7,27]. Additionally, our lab has shown inhibition of the upstream effectors of actomyosin contractility, such as Rho-associated kinase (ROCK), destabilizes the actin cortex, and increases the formation of microtubule based McTNs [28]. This current work shows that treatment of Iono or Tg induces an elevation of cytoplasmic Ca 2+ (Figures 1 and 2) and suppresses McTN formation in the presence of extracellular Ca 2+ (Figure 3). Additionally, the ablation of extracellular Ca 2+ alone with a Ca 2+ chelator leads to an enhancement of McTN levels, while the co-treatment of compound and zero Ca 2+ conditions shows no difference ( Figure 3). The data may suggest that the presence or absence of McTN relies on a necessary balance of spatiotemporal Ca 2+ signaling and provides an intriguing avenue for future studies.
Cytoskeletal plasticity and coordinated remodeling play an essential role in metastatic dissemination and progression. Actin filaments are composed of monomeric globular subunits, which make up part of the cellular cytoskeleton, and is responsible for maintaining cellular morphology [2,3]. Maintenance of the actin cytoskeleton is a highly dynamic process modulated by rapid signaling cascades and the rapid recruitment of accessory proteins to mediate the organization, polymerization, and depolymerization of filamentous actin from pools of globular actin and vice versa [2,3].
Research within the muscle field has established a conserved signaling pathway by which Ca 2+ signaling can induce cellular and actin contraction [11,31,32]. The conserved mechanism of cellular and actin contraction via Ca 2+ signaling in epithelial cells is through non-muscle myosin II. The role of non-muscle myosin II is primarily regulated via the phosphorylation of S19 MLC2 by MLCK in a calcium-calmodulin-dependent manner [33]. MLCK-dependent phosphorylation of MLC2 leads to an unfolding of non-muscle-myosin II and a 1000-fold increase in ATPase activity that promotes motor activity on actin filaments and contraction of the actin cortex [34] [4]. Furthermore, MYPT1 dephosphorylates MLC2, resulting in the relaxation of the actin cortex [28,35]. Investigation into the regulation of the actomyosin cortex was also examined through cofilin activity.
Cofilin plays an essential role in tumor cell motility and has been shown to regulate McTN formation [3,36]. Our immunoblot data show increases in phosphorylated MLC2 at S19 and MYPT1 at T853, indicating contraction of the actin cortex, with simultaneous decreases in phosphorylated cofilin to increase actin severing and rearrangement through cofilin activity ( Figure 6). Additionally, inhibition of actin polymerization prior to the elevation of cytoplasmic Ca 2+ abrogates Ca 2+ -mediated McTN suppression (Figure 7). The summation of these phenotypic observations and immunoblot analysis indicates dynamic actin polymerization and turnover are necessary for Ca 2+ -mediated McTN suppression. These results suggest a concurrent molecular mechanism between actin contraction and dynamic actin polymerization and depolymerization for Ca 2+ -mediated McTN suppression. The dual molecular roles of myosin light chain II and cofilin support the nonlinear mechanical response of individual actin filaments contracting and buckling, which drive cortical actomyosin contractility and polymerization dynamics [37]. Ca 2+ -dependent activation of myosin light chain kinase stimulates actin contraction while concurrently, cofilin activation mediates filamentous actin depolymerization to suppress McTNs over an acute time course [38][39][40][41].
While cytoplasmic Ca 2+ entry from either internal or external Ca 2+ sources gives rise to similar phenotypic suppression of McTNs (Figure 3), further scrutiny into reattachment at later time points in the 24 h time frame begins to yield observable differences in cellular spreading (Figure 4). From previous live cell time course imaging observations and xCelligence reattachment data, the initial attachment was estimated to begin immediately after seeding to around 5 h post seeding, while cellular spreading was visualized as early as 4 h over the 24 h time course [27]. While Iono and Tg treatments acutely decrease the McTN phenotype, these observations do not correlate with the hypothesis of a decrease in electrical impedance reattachment trend from initial seeding to post 5 h after seeding compared to the vehicle control ( Figure 4). However, visualization of cellular spreading at 8 h and 16 h illustrated Iono-treated cells have similar cellular spreading patterns to the bottom of a tissue culture plate to the vehicle control, while Tg treatment revealed a decrease in cellular spreading at 8 h and 16 h compared to the vehicle control ( Figure 4). These observed differences in the cellular spreading are potentially a result of differential signaling cascades based on the entrance of extracellular Ca 2+ or the release of intracellular Ca 2+ stores. Different genetic backgrounds between cell lines can account for differential functional responses. Yet, functional differences observed within a cell line suggest the initiation of distinct signaling cascades that are dependent on the location of Ca 2+ entrance ( Figure 4). For example, extracellular Ca 2+ can increase migration and invasion preferentially to the bone through Ca 2+ -mediated cytoskeletal rearrangement in breast cancer cells [42,43]. Furthermore, sustained extracellular Ca 2+ signaling can additionally cross talk with other oncogenic signaling pathways to promote cell survival, migration, invasion, and enhanced epithelial-mesenchymal transition (EMT) [33,[44][45][46]. In contrast, Ca 2+ signaling through the release of internal stores into the cytoplasm can trigger different signaling cascades some of which result in apoptosis or autophagy over longer time periods, i.e., 48 to 72 h to days or weeks [23,47,48]. While cell death, tumor growth and migration and invasion are established phenotypes and experimental endpoints in the study of metastatic progression and tumor biology, their measurements are not within the time frame of this study.
Targeting CTCs that have shed from the primary tumor remains challenging. These CTCs that aggregate into homotypic and heterotypic clusters in circulation have increased metastatic potential [49][50][51]. Clustering together allows CTCs to survive the hostile environment of shear forces within the circulatory and lymphatic systems to metastasize to distal regions within the body [50]. A recent study in animal models showed cancer cells treatment with Ouabain or Digitoxin, cardiac glycosides, decreased CTC clustering through increasing intracellular Ca 2+ levels, and disruption of cell-cell junctions [49]. Furthermore, inhibiting McTNs with the microtubule depolymerizer Vinorelbine was also recently shown to reduce homotypic McTN-mediated clustering and significantly delayed lung metastasis in mouse models [27]. Our results provide additional evidence linking these previous phenomena by demonstrating the utility of increasing cytoplasmic Ca 2+ to decreases McTN-mediated clustering over a short time course, i.e., 6 h ( Figure 5). Drugs that induce this rapid change of free-floating cancer cell morphology could be leveraged as possible adjuvant treatments immediately post-surgery to reduce the metastatic potential of tumor cells that have shed into the blood stream [52].
Current Ca 2+ -mediating therapies, such as Digitoxin, used in the clinic for the treatment of cardiac disease have shown promise in vitro studies to have a synergistic cytotoxic effect when used in combination with chemotherapeutic agents such as Paclitaxel [23,53]. Tg is a known cytotoxic agent that has demonstrated potential as both a single and combination anticancer agent in vitro [23,47,53]. The results from phase 1 and 2 clinical trials of its prodrug derivative, Mipsigargin, for metastatic disease have shown a favorable pharmacokinetic profile, with dosages that are well tolerated by patients and have been shown to prolong disease stabilization [38,39]. It was also observed in patients with advanced hepatocellular carcinoma who had progressed from sorafenib treatment, Mipsigargin treatment reduced blood flow to hepatic lesions [39]. The results of our study in conjunction with the clinical trial results highlight the potential utility of reducing the metastatic potential of circulating tumor cells through an alternative cytoskeletal mechanism of action.
Close examination of the acute effects of treatment are often overlooked due to a lack of a distinct change in phenotype, i.e., change in tumor size. In cancer biology, we often use the reduction of tumor size as a common endpoint measurement in animal and human studies to determine whether experimental treatments are effective. However, this reduction in primary tumor size occurs days to weeks after treatment is administered, leaving an overlooked time frame immediately following the administration of treatment. The primary tumor is estimated to shed 3.2 × 10 6 cells per gram of tumor tissue per day with the majority of cells quickly dying [31,51]. An increasing number of published studies highlight the elevation of the dissemination of cancer cells from the primary tumor after insults to primary tumors such as tumor biopsies and surgical or pharmaceutical interventions such as neoadjuvant chemotherapy [32,34,37,40,41,52,54]. The emerging data suggest targeting these subpopulations with current chemotherapies, such as Paclitaxel, can impact their metastatic potential. Previous high throughput screening and global gene analysis of selective inhibitors targeting breast cancer stem cells identified HMLER breast cancer cells treated with Paclitaxel have enriched expression of cancer stem cell genes [5,55]. Karaginannis and colleagues later showed patient-derived xenografts treated with Paclitaxel that demonstrated pro-metastatic changes within the tumor microenvironment of metastasis, which increased the dissemination and intravasation of cancer cells [56]. Currently, the monitoring of CTCs serves as a prognostic biomarker in cancer treatments. They are used to determine the efficacy of treatment or disease progression by measuring increases or decreases in CTCs found in the patient's blood samples over time [57]. However, these enumeration studies rarely examination of the phenotype or morphology of each CTC at the time of collection. The short-term impacts of drug treatment on CTC enumeration and phenotypes highlight the gap in the knowledge of molecular mechanisms in this subpopulation and emphasize the importance of appropriate time frames for the use of pharmaceuticals in both adjuvant and neoadjuvant settings. By identifying and understanding the various phenotypes and molecular targets at specific time points within the metastatic cascade will lead to the development of personalized treatments.

Conclusions
The complex 3-dimensional environment that CTCs survive encompasses a vast array of factors acting on the cell, including fluid shear stress mechanical signals, other circulating cell types, and various soluble signaling factors. Recent studies have begun to elucidate CTC sensitivity to fluid shear stress and have suggested increased CTC stiffness leads to an increase in cell death [51,58,59]. Our current data supports clinical trial data indicating a rationale for the use of Ca 2+ modulators as a potential therapeutic strategy for preventing metastasis [38,39]. Our study shows proof-of-principle that viability-independent transient increases in cytoplasmic Ca 2+ with Iono or Tg yield rapid morphological changes to tumor cells in suspension that reflect a less advantageous phenotype for metastatic behaviors.