Functional characterization and anti-cancer action of the clinical phase II cardiac Na+/K+ ATPase inhibitor istaroxime: in vitro and in vivo properties and cross talk with the membrane androgen receptor

Sodium potassium pump (Na+/K+ ATPase) is a validated pharmacological target for the treatment of various cardiac conditions. Recent published data with Na+/K+ ATPase inhibitors suggest a potent anti-cancer action of these agents in multiple indications. In the present study, we focus on istaroxime, a Na+/K+ ATPase inhibitor that has shown favorable safety and efficacy properties in cardiac phase II clinical trials. Our experiments in 22 cancer cell lines and in prostate tumors in vivo proved the strong anti-cancer action of this compound. Istaroxime induced apoptosis, affected the key proliferative and apoptotic mediators c-Myc and caspase-3 and modified actin cystoskeleton dynamics and RhoA activity in prostate cancer cells. Interestingly, istaroxime was capable of binding to mAR, a membrane receptor mediating rapid, non-genomic actions of steroids in prostate and other cells. These results support a multi-level action of Na+/K+ ATPase inhibitors in cancer cells and collectively validate istaroxime as a strong re-purposing candidate for further cancer drug development.


INTRODUCTION
Na + /K + ATPase is a well-established pharmacological target acting as a receptor for cardiotonic steroids (CTS) such as digoxin and digitoxin [1,2]. Most recently, a new generation of non-sugar containing steroidal enzyme inhibitors with improved therapeutic indices (i.e. better ratio of therapeutic activity versus toxicity) has been reported [3][4][5][6][7]. Istaroxime represents the lead candidate from this class and has successfully concluded phase II clinical trials in cardiac failure patients [8][9][10]. Na + /K + ATPase is also emerging as a novel anti-cancer target. Overall, four layers of evidence support this role (reviewed in [11]). First, aberrant expression of enzyme subunits has been observed in a growing number of cancers. Second, several different CTS have shown outstanding activities in chemical and functional screens. Third, data from multiple epidemiological studies involving CTS-treated cardiac patients have shown reduced cancer development risk, incidence or mortality in certain cancer indications. Finally, clinical trials with various CTS have proven the promising anti-cancer potential of these inhibitors.
Based on the promising anti-cancer properties described above, our group previously assessed the anticancer potential of novel Na + /K + ATPase inhibitors [3][4][5][6][7], showing strong activity of these compounds in multiple cancer cell lines [12,13]. Moreover, 3-R-POD, the most active derivative of this series, exhibited dose dependent tumor inhibition in prostate and lung cancer xenografts in vivo [12,13]. In the current study, we focused on istaroxime, the lead inhibitor of this class. Specifically, we show that istaroxime is active in 22 different cancer cell lines derived from 9 tumor panels in vitro as well as in prostate cancer xenografts in vivo. We demonstrate novel mechanistic insights on the signaling of this compound in prostate cancer cells. Moreover, we link for the first time Na + /K + ATPase and mAR, a membrane androgen receptor mediating rapid, non-genomic anti-cancer effects of androgens in multiple cancer cells [14,15]. Our results provide novel insights into the anti-cancer properties of istaroxime further supporting development of this agent as a novel anti-cancer drug candidate.

In vitro anti-cancer activity of istaroxime in multiple cell lines
Having recently characterized 17 cardiac enzyme inhibitors in anti-cancer assays [12,13], we tested the anti-cancer activity of istaroxime, the prototype cardiac inhibitor of this class ( Figure 1A, Na + /K + ATPase IC 50 : 407.5 nM). Specifically, we determined GI 50 , TGI and LC 50 values (see SRB assays, Methods) of the compound in 22 different cancer cell lines from 9 tumor panels (lung, melanoma, ovarian, renal, CNS, breast, pancreas, colon and prostate). Istaroxime exhibited GI 50 and LC 50 values in the low micromolar range in all cell lines; PC-3 and DU145 prostate cancer cells were among the most sensitive cells to the action of the compound (Table 1). Although some anti-proliferative activity was observed in normal fibroblasts, TGI and LC 50 values of istaroxime were significantly higher in comparison to values in cancer cells (Table 1). Interestingly, and as shown previously for other inhibitors of the same class [12], istaroxime exhibited comparable anti-cancer activity in multi-drug resistant NCI/ADRRES cells [16,17]. Similar results were observed with MTT assays in DU145 and CAKI-1 cells (Table 1) further confirming istaroxime's anti-cancer action.

Istaroxime shows anti-cancer activity in PC-3 prostate cancer xenografts
To further characterize the anti-cancer properties of istaroxime, we have performed experiments in PC-3 prostate xenografts in vivo. Initially, we determined the compound's maximum tolerated dose (MTD) by injecting increasing doses in NOD/SCID mice via intraperitoneal (IP) injection. These studies defined the acute MTD (single dose) at 200 mg/kg. Daily administration of 40-50 mg/kg over a period of several days (chronic administration) was also well tolerated (data not shown). Taken together this dosing information and the short half-life of the compound (< 1 hour; [18]), we have selected a dose of 22.5 mg/kg administered IP twice daily (at 12 hours intervals) for prostate xenograft experiments. Docetaxel, an anticancer drug approved for metastatic prostate cancer was also included as a positive control in our assays (dosed intravenously at 12 mg/kg, once weekly). As shown in Figure 1B, istaroxime showed statistically significant tumor growth inhibition against PC-3 xenografts (p < 0.05, days 7, 10, 17 and 24). DT/DC values ranged between 41-67% throughout the experiment ( Figure 1; Panel C), whereas Tumor Growth Inhibition (TGI) at day 24 was 43.1%. Neither istaroxime nor docetaxel significantly modified body weight as indicator of toxicity ( Figure 1; Panel D). Similar results were obtained in a separate xenograft experiment employing a dose of 40 mg/kg injected IP once daily (four daily treatments followed by three days of rest for three weeks; data not shown). Altogether, these results confirm the anti-cancer activity of istaroxime in prostate cancer xenografts in vivo.

Istaroxime induces apoptosis and caspase-3 activation in prostate cancer cells
We further analyzed the apoptosis of DU145 prostate cancer cells treated with 5 μM of the compound over a period of 24 hours. As shown by FACS analysis (Figure 2, Panels A, B), istaroxime increased the number of apoptotic cells from 9.48% in control samples to 46.54% following treatment ( Figure 2, Panel C). In agreement with the FACS data, istaroxime induced a modest, yet reproducible increase in caspase-3 activity peaking at 24 h post treatment initiation (Figure 2; Panel D).

Istaroxime reduces c-Myc expression and induces actin cytoskeleton re-organization and RhoA activation in prostate cancer cells
Recent studies with other CTS have recently reported a reduction of c-Myc oncoprotein expression and actincytoskeleton re-arrangements in prostate and lung cancer cells [19,20]. To assess whether istaroxime induced similar effects, we performed c-Myc Western blot analysis in DU145 prostate cells treated with 5 μΜ of the compound in a time course extending up to 6 hours. In agreement with previous observations, istaroxime significantly down-regulated c-Myc protein levels ( Figure 3). c-Myc mRNA levels remained unchanged at the same time (data not shown), indicating that istaroxime-mediated effects most likely occurred through protein destabilization. Interestingly, the compound modulated actin polymerization dynamics by inducing rapid -within 30 minutes-F/G-actin ratio increase; actin polymerization persisted for at least 120 minutes ( Figure 4A). This finding was supported by laser scanning microscopy analysis ( Figure 4B), showing a reorganization of the actin network with formation of stress fibers ( Figure 4B arrows) in treated cells.
The Rho family of small GTPases holds a prominent role in regulating rapid actin reorganization induced by various effectors [21][22][23]. Thus, we performed affinity precipitation assays with a GST-fusion protein comprising the rhotekin Rho-binding domain (GST-RBD) to assess RhoA GTP loading [24]. As shown in Figure 4C, treatment with a positive control (PC; 10 −7 M testosterone-HSA conjugate) induced a rapid and moderate 1.5 fold activation of RhoA as reported previously [14,25]. Interestingly, istaroxime was also able to induce early and robust RhoA activation (3.2 fold); this result was in agreement with its capacity to induce rapid actin polymerization and the reported role of RhoA in this process [23,26].

Istaroxime precludes binding of testosterone conjugates to the membrane androgen receptor
Results reported above indicated a major role of istaroxime in actin cytoskeleton re-organization and rapid RhoA activation in prostate cancer cells. Since: i) similar properties are shared by ligands of the membrane androgen receptor (mAR) (e.g. Testosterone-Human Serum Albumin conjugates; Testo-HSA conjugates, TAC), [14,15] and ii) istaroxime contains the typical steroid A-D ring structure similar to testosterone-HSA conjugates ( Figure 1A), we tested the capacity of this compound to preclude binding of a fluorescent Testosterone-HSA conjugate (TAC-FITC) to mAR in fluorescent-based assays as described previously [27]. For these assays, we used LNCaP prostate cancer cells due to their optimal mAR detection properties and similarities in growth responses to istaroxime in comparison to DU145 or PC-3 cells (data not shown). In agreement with published results [27], TAC-FITC conjugates showed clear membrane staining, (

Istaroxime displaces radioactive testosterone bound to the membrane androgen receptor
To further characterize the putative interaction of istaroxime with mAR, we performed binding experiments using radiolabeled testosterone in membrane preparations of DU145 cells. As shown in Figure 6, membrane preparations incubated with [ 3 H]-testosterone in the presence of increasing concentrations of Dihydrotestosterone (DHT) (10 -12 to 10 -5 M) revealed a displacement of radiolabeled testosterone by DHT. In agreement with fluorescent assay results, istaroxime displaced radiolabeled testosterone and was more potent than DHT in this assay. Displacement of radiolabeled testosterone seemed to be more pronounced when both DHT and istaroxime were combined. However, differences between groups were not statistically significant. These results indicated direct binding of istaroxime to the mAR. Of note, similar results were obtained with digoxin (data not shown).

DISCUSSION
Recent preclinical and clinical data point to a potential anti-cancer effect of Na + /K + ATPase inhibition in various indications [28]. For example, several cancer types harboring isoform overexpression or holoenzyme subunit alterations have been identified (reviewed in [11]; references herein). In addition, a series of screening and epidemiological studies qualify CTS as highly active anticancer compounds with the potential to reduce cancer risk [29][30][31][32][33]. In agreement with these observations, plant derived inhibitors such as PBI-05204/oleandrin or HuaChanShu have shown anti-cancer activity in early clinical trials [34][35][36].
In this study, we focused on istaroxime, a Na + /K + ATPase inhibitor designed to overcome known proarrhythmic and other limitations of CTS [6,18]. Having shown favorable properties in cardiac phase II clinical trials [8][9][10], istaroxime seemed an ideal cancer repurposing candidate. Our growth assays in multiple cell lines (Table 1) together with efficacy experiments in PC-3 prostate cancer xenografts in vivo (Figure 1) confirmed the purported anti-cancer effects of this compound. Analyzing istaroxime's mechanism of action, we have shown that the compound induced apoptosis and caspase-3 activation in prostate cancer cells (Figure 2). Similar to what was shown for other Na + /K + ATPase inhibitors, istaroxime suppressed c-Myc oncoprotein expression ( Figure 3) and induced actin cytoskeleton re-organization in prostate cancer cells (Figure 4). Finally, in agreement with its actin re-modeling role, istaroxime treatment resulted in rapid activation of RhoA signaling ( Figure 4C). We believe that the activation of RhoA induced by istaroxime is an early and transient stimulus of actin polymerization, rather than associated with enhanced invasiveness of the cells [37]. In line with this, a wound healing assay showed clear inhibition of cell migration in DU145 cells treated by istaroxime as compared to untreated cells (data not shown). Further studies are now needed to address this issue in more detail.
Searching for additional targets and/or downstream effectors of istaroxime, we have also tested the potential functional crosstalk of Na + /K + ATPase with mAR, a membrane receptor mediating rapid, non-genomic anticancer effects of androgens [14,15]. Our hypothesis was triggered by the observation that mAR activation shared several similarities to Na + /K + ATPase inhibition, namely induction of rapid actin cytoskeleton reorganization and rapid RhoA activation (Figure 4; [15]). Using both fluorescent binding exclusion ( Figure 5) and displacement assays ( Figure 6), we have shown binding of istaroxime to the mAR. This indicated functional crosstalk of this receptor with the Na + /K + ATPase.
These results initially supported the idea that the membrane pool of Na + /K + ATPase may actually serve as the elusive mAR, a hypothesis also put forward more than 20 years ago by Farnsworth [38]. However, several lines of evidence argue against this hypothesis. First, we have been unable to show direct inhibition of Na + /K + ATPase activity by testosterone, DHT, testosterone-albumin conjugates or testosterone 3-CMO (a selective mAR ligand [39]) in assays using purified Na + /K + ATPase (data not shown). Similarly, assays in Xenopus laevis oocytes failed to show any influence of testosterone on Na + /K + ATPase currents (data not shown). Third, neither istaroxime nor digoxin or ouabain were capable of blocking mAR dependent apoptosis triggered by testosterone-albumin conjugates in prostate cancer cells and vice versa (data not shown). These results together with observations that mAR is a G-protein coupled receptor (reviewed in [40]) further support the notion that Na + /K + ATPase and mAR may not be identical. Indeed, the finding that istaroxime together with DHT was more effective than each agent alone in displacement assays performed with membranes of DU145 cells (Figure 6), support the notion that Na + /K + ATPase and mAR may either share a common subunit, be part of a bigger membrane complex or be functionally interlinked. For example, it is possible that istaroxime binds to a Na + /K + ATPase accessory factor increasing its affinity to mAR or a relevant complex. This may result in mAR activation (explaining the results on rapid RhoA activation and actin re-organization, Figure 4). Alternatively, membrane calcium channels could be formed on the membrane linking Na + /K + ATPase and mAR [41]. This could explain the observations that both, Na + /K + ATPase inhibition and mAR activation result in an increase in intracellular Ca 2+ levels [42][43][44].
The pore forming Ca 2+ release activated Ca 2+ channel (CRAC) subunit Orai, shown to participate in both mAR signaling and regulation of Na + /K + ATPase expression in various tumors [45][46][47][48] may play as well a role. The same applies for Na + /H + exchanger and glucocorticoid inducible kinase-1 (SGK1), both shown to be functionally interlinked with mAR and Na + /K + ATPase [49][50][51][52]. Additional candidates include GPRC6A, a Ca 2+ activated G-protein coupled receptor proposed to be identical to mAR [53,54] or the zinc transporter ZIP9 also proposed to fulfill the role of mAR in cancer cells [55]. Finally, it is worth noting that potential involvement of the intracellular androgen receptor (iAR) to Na + /K + ATPase is highly unlikely at this point since i) iAR is functionally discrete to mAR as shown by multiple publications (reviewed by [15]) and ii) istaroxime does not bind the iAR [6]. On the other hand, published observations that progesterone may directly bind the Na + /K + ATPase complex clearly point to the necessity of performing a more detailed functional characterization of the interplay of Na + /K + ATPase with mAR, iAR or other steroid receptors.
In conclusion, istaroxime, a clinically validated cardiac Na + /K + ATPase inhibitor, exhibits strong anticancer potential in vitro and in in vivo. Our results and reported mechanisms of action support a drug re-purposing potential of the compound against prostate and other tumors. (D) Caspase-3 activity was measured at 405 nm in lysates derived from cells exposed to 5 μΜ istaroxime for the indicated time periods and then incubated with the caspase-3 substrate DEVD conjugated to the chromophore pNA as described in Methods. The relative caspase-3 activity is expressed as percentage with that of serum cultured cells taken as 100%. Data presented in bars are mean values ± SE of n = 6 independent experiments (*P < 0.05). www.impactjournals.com/oncotarget

Ethical statement
Investigation has been conducted in accordance with the ethical standards and according to the Declaration of Helsinki and according to national and international guidelines and has been approved by the authors' institutional review board.

Na + /K + ATPase assays
Istaroxime's inhibitory effect on ATPase activity was assessed in vitro using the Adenosine 5´-Triphosphatase Enzymatic Assay of Sigma (St.Louis, MO) according to the manufacturer's instructions as reported previously [12]. This assay utilizes enzyme isolated from porcine cerebral cortex.

MTT and SRB assays
Cell proliferation/viability was assessed by MTT [3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide] assays (Sigma, St.Louis, MO) done as previously described [12]. Sulforhodamine B (SRB) assays were performed according to NCI guidelines for anti-cancer drug screening (https://dtp.cancer.gov/ discovery_development/nci-60/default.htm) and as previously published [12,57]. In SRB assays, three dose response parameters were calculated for each agent. Growth inhibition of 50% (GI 50 ): drug concentration resulting in a 50% reduction of growth; Total growth inhibition (TGI): drug concentration resulting in a 100% reduction of growth and lethal concentration of 50% (LC 50 ): concentration of drug resulting in a 50% reduction of cell viability. Typically, all assays were done in triplicates in 2-3 repetitions for each compound. This is deemed sufficient to extrapolate on the activity of each compound. Note that MTT-calculated IC 50 values may differ from SRB-assay calculated GI 50 values due to inherent differences in assay methodologies.

Maximum tolerated dose (MTD) assays
Female NOD/scid mice, 8 weeks old, weighing ~21 g were used for the MTD studies according to NCI guidelines (https://dtp.cancer.gov/organization/btb/acute_ toxicity.htm). Briefly, one animal per individual dose was injected with a single dose of 200/100/50/25/12.5 mg/kg of instaroxime respectively, in a volume of 20 μL/g of weight. Animals were weighed prior to each administration and volumes/dose administered was adjusted according to body weights. The animal that received 200 mg/kg suffered from sedation but recovered within 1-2 hours; all other animals showed no side effects. In a second round of experiments, animals received once-daily doses of 40 and 50 mg/kg for several consecutive days; these treatments were also well tolerated. Based on these studies, we concluded that the compound's acute MTD was 200 mg/kg whereas the chronic MTD was > 50 mg/kg.

PC-3 xenograft studies
Xenografts were generated by subcutaneously injecting exponentially growing cultures of ~2 × 10 6 PC-3 cells (in Matrigel in 0.1 ml PBS) at the right flank of 6-8 weeks old male Balb/c nude mice. Following development of palpable tumors (150-200 mm 3 ) and group randomization (10 animals per group), the following treatments were administered: Group A: Water for Injection (WFI), IP, twice daily for 23 days at 12 hours intervals.
Group B: Istaroxime at 22.5 mg/kg, IP, twice daily for 23 days at 12 hours intervals.
Group C: Docetaxel at 12 mg/kg, IV, once weekly for 23 days.
Tumor volumes were measured twice weekly in two dimensions using a caliper according to the formula V = 1/2 × a × b 2 where a and b are the long and short diameters of the tumor respectively. %DT/DC values were also calculated, where DT = T − Do and DC = C − Do (Do is the average tumor volume at the beginning of the treatment; T and C are the volumes of treated and untreated tumors, respectively, at a specified day). Tumor Growth Inhibition was calculated as the percentage of tumor volume versus vehicle control in a given date. Losses of weight, neurological disorders, behavioral and dietary changes were also recorded as indicators of toxicity (side effects). Experiments were terminated when tumors in control animals reached a volume of ~2000-2500 mm 3 (about 11% of the body weight). Animals used in xenograft studies were treated according to the Guide for the Care and Use of Laboratory Animals, Institute of Laboratory Animal Resources, National Academy Press, Washington, 1996. Specific conditions regarding handling of moribund animals as determined by the veterinary staff of the test facility were explicitly defined in the study protocol according to international guidelines including euthanasia for humane reasons. This included carbon dioxide inhalation followed by exsanguination. Final disposition of all animals placed on study was documented in all study records.

TUNEL apoptosis assay and FACS analysis
DU145 cells were cultured in serum containing medium in the absence or presence of 5 μM istaroxime for 24 hours. At the end of the treatment, cells were harvested in PBS and apoptosis was assessed using the APO-BrdU ™ TUNEL Assay kit (Molecular Probes, Eugene, OR) according to the manufacturer's instructions. The APO-BrdU ™ TUNEL Assay kit labelled DNA strand breaks for the detection of apoptotic cells through addition of BrdU at the 3´-hydroxyl ends by terminal deoxynucleotide transferase. Final detection of BrdU incorporation at DNA break sites was detected using an Alexa Fluor488-labeled anti-BrdU antibody. Propidium Iodide was used to stain the total cell population. Flow cytometry was performed with FACSArray apparatus (BD Biosciences) and the results were analyzed by the CellQuest software (BD Biosciences).

Caspase-3 assay
The activity of caspase-3 was measured in whole cell lysates pretreated or not with 5 μM istaroxime for the time periods indicated in the Figure legends, using the Clontech ApoAlert ® Caspase Colorimetric Assay kit according to the manufacturer's instructions. Caspase-3 activity was determined by incubating lysates with a caspase-3 substrate (the peptide DEVD conjugated to the chromophor p-nitroaniline) for 2 h at 37 o C. The absorbance of each sample was measured at 405 nm by using a 96-well colorimetric plate reader.

Measurement of F/G actin ratio by Triton X-100 fractionation
The Triton X-100 soluble G-actin and Triton X-100 insoluble F-actin containing fractions of cells exposed to 5 μM istaroxime for the time periods indicated in Figure legend, were prepared as previously described [58]. An increase of the triton-insoluble (F-) to the triton-soluble (G-) actin ratio is indicative of actin polymerization.

Confocal laser scanning microscopy
Cells were cultured on glass coverslips with 5 μM istaroxime for the time points indicated in the Figure legend. For direct fluorescence microscopy of F-actin, cells were fixed with 3% paraformaldehyde in PBS for 30 min, permeabilized with 0.5% Triton X-100 in PBS (10 min) and incubated with rhodamine-phalloidin (Molecular Probes, Eugene, OR, 1:100 dilution) for 40 min in the dark. Confocal microscopy was performed with a Zeiss LSM 5 EXCITER confocal laser-scanning module (Carl Zeiss) and images were analyzed with the instrument's software.

Membrane androgen receptor competition assays
To determine the capacity of a given compound to preclude binding of fluorescent testosterone-HSA conjugates (TAC-FITC) to the mAR, starved LNCaP cells grown on coverslips were pre-treated with 40 μΜ of the indicated drug for 30 minutes prior to the addition of testosterone-HSA-FITC. At the end of the incubation period, cells were washed twice with PBS and 40 μΜ of testosterone-HSA-FITC was added. Specimens were prepared and analyzed as described above. Compounds binding to the mAR preclude binding of fluorescent testosterone-HSA conjugates to their target and abolish membrane-specific fluorescence.

Detection of membrane androgen receptors and competition assays
For membrane preparation DU145 cells cultured in five 75 cm 2 flasks, were washed twice with PBS, removed by scraping, and centrifuged at 1, 500 g for 5 min. Pelleted cells were homogenized by sonication in 50 mM Tris-HCl buffer, pH 7.4, containing freshly added protease inhibitors (10 μg/ml PMSF and Roche complete protease inhibitor tablets). Unbroken cells were removed by centrifugation at 2, 500 g for 15 min. Membranes were obtained by centrifugation at 20, 000 g for 1 h and washed once with the same buffer. Protein concentration was measured by the method of Bradford using reagents from Bio-Rad (Hercules, CA). Displacement binding experiments were performed as previously described [59]. In brief, cell membrane preparations at a final concentration of 1.0 mg/ ml were incubated with 5 nM [ 3 H] testosterone in the absence or in the presence of different concentrations of unlabeled steroid (istaroxime, DHT, istaroxime and DHT), ranging from 10 -12 to 10 -5 M. Non-specific binding was estimated in the presence of 5 μM istaroxime. In both types of binding experiments, after an overnight incubation at 4°C, bound radioactivity was separated by filtration under reduced pressure through GF/A filters previously soaked in 0.5% polyethylenimine (PEI) in water and rinsed three times with ice-cold Tris-HCl buffer. Filters were mixed with 10 ml scintillation cocktail (03999, Fluka), and bound radioactivity was counted in a scintillation counter (TriCarb ® 2900TR Liquid Scintillation Analyzer, PerkinElmer).