Cardiotonic Steroids Differentially Affect Intracellular Na (cid:1) and [Na (cid:1) ] i /[K (cid:1) ] i -independent Signaling in C7-MDCK Cells*

Recently, we reported that ouabain kills renal epithelial and vascular endothelial cells independently of elevation of the [Na (cid:1) ] i /[K (cid:1) ] i ratio. These observations raised the possibility of finding cardiotonic steroids (CTS) that inhibit the Na (cid:1) ,K (cid:1) pump without attenuating cell survival and vice versa. To test this hypothesis, we compared CTS action on Na (cid:1) ,K (cid:1) pump, [Na (cid:1) ] i content, and survival of Madin-Darby canine kidney cells. At a concentration of 1 (cid:2) M , ouabain and other tested cardenolides, as well as bufadienolides such as bufalin, cinobufagin, cinobufotalin, and telobufotoxin, led to (cid:1) 10-fold inhibition of the Na (cid:1) ,K (cid:1) pump, a 2–3-fold decrease

Na ϩ /K ϩ -ATPase is an ubiquitous plasma membrane heterodimer detected in all types of animal cells. Its larger ␣-subunit (ϳ110 kDa) provides ATP hydrolysis coupled to the transport of three Na ϩ and two K ϩ ions against their electrochemical gradients, whereas its ␤-subunit (ϳ35 kDa) is mainly involved in plasma membrane delivery and assembly of the enzyme (1,2). In addition to its unique enzymatic properties, the Na ϩ /K ϩ -ATPase ␣-subunit is the only known target of cardiotonic steroids (CTS) 1 initially extracted from the leaves of Digitalis purpurea and Digitalis lanata plants and termed as digitalis. Since the report of Sir William Withering in 1789, glycosides have played a prominent role in the treatment of congestive heart failure (3). Later on, ouabain was extracted from Strophanthus gratus, and because of much higher water solubility, it was employed in an overwhelming number of animal and in vitro studies (for more details, see Ref. 4). Sideby-side with plant-derived cardenolides, other members of the CTS superfamily, bufadienolides, have been isolated from amphibians (5). Moreover, augmented levels of CTS indistinguishable from cardenolides (ouabain and digoxin) and bufadienolides (marinobufagenin (MBG)) have been found in the plasma of hypertensive subjects and animals with volume-expanded hypertension (6 -8).
In myocytes, inhibition of the electrogenic Na ϩ /K ϩ pump by CTS leads to depolarization, elevation of [Na ϩ ] i , activation of electrogenic [Ca 2ϩ ] o /3[Na ϩ ] i exchanger, and the opening of voltage-gated Ca 2ϩ channels that results in elevation of [Ca 2ϩ ] i , contraction of cardiomyocytes, and augments the sensitivity of vascular smooth muscle cells to endogenous constrictors (9,10). Recently, several research teams proposed that CTS are also involved in Na ϩ -independent signaling. This hypothesis was based on data showing that at lower concentrations CTS augment cell proliferation (11,12), DNA synthesis (11), mitogen-activated protein kinase activity (11,13,14), and the production of reactive oxygen species (15,16) without significant inhibition of the Na ϩ ,K ϩ pump and elevation of [Na ϩ ] i .
More direct evidence of [Na ϩ ] i -independent signaling was obtained in a study of epithelial cells from the Madin-Darby canine kidney (MDCK). In these cells, long term exposure to ouabain at concentrations sufficient to completely inhibit Na ϩ, K ϩ -ATPase resulted in attenuated staining with dimethylthiazol-diphenyltetrazolium (MTT) (17), commonly used to estimate the relative number of alive cells and their redox state (18), and massive detachment of dead cells (19). The death of ouabain-treated C7-MDCK cells shares markers with apoptosis (caspase-3 activation) and necrosis (cell swelling and smear pattern of DNA degradation) and, in contrast to "classic" apoptosis, is resistant to the pan-caspase inhibitor N-benzyloxycarbonyl-Val-Ala-Asp (17). Importantly, both a decline of MTT staining and death were triggered by ouabain in the absence of  a transmembrane gradient of monovalent cations (17). Similar results were obtained under analysis of the ouabain action on primary cultures of endothelial cells from the porcine aorta (20). This intriguing discovery raises the possibility of finding CTS that trigger [Na ϩ ] i /[K ϩ ] i -independent signaling without significant modulation of intracellular ion homeostasis and vice versa. Here, we confirm the hypothesis and demonstrate that MBG and marinobufotoxin do not affect MTT staining and survival of C7-MDCK cells at concentrations sufficient to inhibit Na ϩ ,K ϩ -ATPase and augment the [Na ϩ ] i /[K ϩ ] i ratio to the same extent as ouabain and other CTS under investigation.

MATERIALS AND METHODS
Cell Culture-C7-MDCK cells resembling principal cells from collecting ducts (21) were kindly provided by Dr. Michel Gekle (University of Warzburg, Warzburg, Germany) and maintained in culture as de-scribed previously (17). Before the experiments, the cells were seeded in 12-, 24-, or 96-well plates and grown until confluency in the presence of Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum, penicillin (100 units/ml), and streptomycin (100 g/ml). Cell morphology was evaluated by phase contrast microscopy at ϫ100 magnification without preliminary fixation. To estimate cell attachment, cells growing in 24-well plates in the absence or presence of CTS were washed four times with medium W containing 100 mM MgCl 2 and 10 mM HEPES-Tris (pH 7.4), and the protein content of attached cells was measured by the modified Lowry method (22). Staining with MTT was performed in cells seeded in 96-well microplates and grown to confluence as described in detail previously (17). The optical density of the resulting solution was measured at 590 nm. Caspase-3 activity in cells growing in 6-well plates was measured as the rate of the caspase-3 inhibitor (Ac-DEVD-CHO)-sensitive component of caspase-3 fluorescent substrate (DEVD-AMC, N-acetyl-Asp-Glu-Val-Asp-AMC) hydrolysis according to a previously described protocol (23).
Na ϩ ,K ϩ Pump Activity-Na ϩ ,K ϩ pump activity was studied as a CTS-sensitive component of the rate of 86 Rb influx. To measure maximal Na ϩ ,K ϩ pump activity, the cells were preloaded with Na ϩ in accordance with the protocol described previously (17). Briefly, C7-MDCK cells seeded in 24-well plates were washed twice with 2-ml aliquots of phosphate-buffered saline and incubated for 2 h at 37°C in K ϩ -free medium A containing 140 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 5 mM D-glucose, and 20 mM HEPES-Tris buffer (pH 7.4). Isotope uptake was measured in medium A containing 0.5% bovine serum albumin, 5 or 2 mM KCl, and 1 Ci/ml 86 Rb with or without the CTS. To inhibit 86 Rb influx mediated by Na ϩ ,K ϩ ,Cl Ϫ cotransport, bumetanide was added at a final concentration of 10 M. To measure base-line Na ϩ ,K ϩ -ATPase activity, preincubation in K ϩ -free medium was omitted. In these experiments, the cells were incubated with 86 Rb in DMEM-like medium containing 104.4 mM NaCl, 5 mM KCl, 1.8 mM CaCl 2 , 0.8 mM MgSO 4 , 29.8 mM NaHCO 3 , 0.9 mM NaH 2 PO 4 , 8.4 mM HEPES (pH 7.4), 5 mM glucose, and vitamins and amino acids at concentrations indicated in the DMEM recipe. In the absence of CTS, the kinetics of isotope uptake were linear up to 20 -25 min of incubation (data not presented). Keeping this in mind, we limited incubation time to 15 min. Then the cells were transferred onto ice, washed four times with 2-ml aliquots of medium W, and lysed with 0.5 ml of 1% SDS, 4 mM EDTA mixture. Radioactivity of the incubation medium and cell lysate was measured with a liquid scintillation analyzer, and the rate of 86 Rb influx was calculated as V ϭ A/amt, where A was the radioactivity of the samples (cpm), a was the specific radioactivity of 86 Rb calculated from the total concentration of K (up to 80 mM) and Rb (up to 10 M) (cpm/nmol), m was the protein content (mg), and t was incubation time with the isotope.
Intracellular Content of Exchangeable Na ϩ -Intracellular content of exchangeable Na ϩ was measured as the steady-state distribution of extracellular and intracellular 22 Na (17). In these experiments, the cells were incubated for 5 h in physiologically balanced medium containing 2 mM KCl and 2 Ci/ml 22 NaCl with or without CTS. Then the cells were transferred onto ice, washed four times with 2 ml of ice-cold medium W, and lysed with SDS/EDTA mixture. The radioactivity of the incubation medium and cell lysate was measured, and intracellular cation content was calculated as A/am, where A was the radioactivity of the samples (cpm), a was the specific radioactivity of Na ϩ in the medium (cpm/ nmol), and m was the protein content (mg).
[ 3 H]Ouabain Binding-Cells grown in 24-well plates were incubated for 3 h in complete or K ϩ -free DMEM-like medium in the presence of [ 3 H]ouabain at concentrations Ͻ0.5 Ci/ml and other compounds mentioned in the table and figure legends. Then the cells were transferred on ice, washed four times with 2 ml of ice-cold medium W, and lysed for the measurement of radioactivity as indicated above. At a concentration of 3 mM, [ 3 H]ouabain binding was not affected by [K ϩ ] o elevation from 0 to 80 mM, indicating nonspecific (Na ϩ ,K ϩ -ATPase-independent) binding. These values (dpm/mg protein) were subtracted from [ 3 H]ouabain binding measured at concentrations below 1 M.
Chemicals-Ouabain was from ICN (Irvine, CA); digoxin, digitoxin, digoxigenin, digitoxigenin, strophanthidin, bufalin, cinobufotalin and cinobufagin, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were from Sigma. Marinobufotoxin and telobufotoxin were kind gifts from Dr. V. P. Butler (Columbia University, New York, NY). Cortexolone, hydrocortisone, prednisolone, dihydrodeoxycotricosterone, and acetylandrosol were synthesized at the Institute of Organic Chemistry (Russian Academy of Science, Moscow, Russia). MBG was purified from the venom of Bufo marinus as described in detail elsewhere (24). The chemical structures of these compounds are shown in Fig Rb) influx, cell attachment, and MTT staining of C7-MDCK cells Maximal Na ϩ ,K ϩ pump activity was estimated as the rate of K ϩ ( 86 Rb) influx in Na ϩ -loaded cells. In these experiments, the cells were preincubated for 2 hr at 37°C in 0.25 ml of K ϩ -free medium A, and then 0.25 ml of medium A containing 10 mM KCl, 1% bovine serum albumin, 20 M bumetanide, 2 Ci/ml 86 Rb, and CTS at concentrations twice higher than indicated in the left column was added for the next 15 min. The values of the 86 Rb influx rate inhibited by 10 M ouabain varying in the range from 756 to 611 nmol (mg protein) Ϫ1 15 min Ϫ1 were taken as 100%. MTT staining and cell attachment were measured after 6 and 24 h of incubation, respectively, in DMEM. The values obtained in the absence of CTS were taken as 100%. The means Ϯ S.E. from experiments performed in quadruplicate ( 86 Rb influx and protein content) or octaplicate (MTT staining) are given.

RESULTS
Screening of CTS-The addition of cardenolides at a concentration of 1 M resulted in 2-4-fold inhibition of maximal Na ϩ ,K ϩ -ATPase pump activity measured as the rate of bumetanide-resistant 86 Rb influx in Na ϩ -loaded cells (Table I). About the same Na ϩ ,K ϩ pump inhibition was detected in cells treated with 1 M of marinobufotoxin, MBG, cinobufotalin or telobufotoxin, whereas bufalin and cinobufagin suppressed its activity by 10-fold (Table I). Fig. 2A shows that in the presence of 1 M ouabain accumulation of 22 Na reached a plateau in 2 h of ouabain addition, whereas the death of C7-MDCK cells, indicated by detach-ment and the appearance of rounded, floating cells (Fig. 3), was detected after 10 h of ouabain addition. In contrast to the delayed kinetics of cell detachment seen, significant attenuation of MTT staining was observed in 30 min with a 2-3-fold decrease of optical density in 6 h (Fig. 2B). The second phase of MTT staining decline, detected in 10 h of ouabain addition, was in proportion to the reduced number of attached cells. Based on these results, we compared the potency of CTS as modulators of MTT staining and triggers of cell detachment after 6 and 24 h, respectively.
Similarly to ouabain, the addition of other cardenolides as well as bufalin, cinobufagin, cinobufotalin, and telobufotoxin led to a 2-3-fold decrease of MTT staining and cell detachment (Table I). In contrast, neither MBG nor marinobufotoxin affected MTT staining and attachment of C7-MDCK cells (Table  I and (Fig. 4, A and B). Neither MTT staining nor cell survival was significantly affected by MBG at concentrations under 3 M (Figs. 3 and 4, C and D). At 10 M, MBG decreased MTT staining and cell attachment by 20 -25% only (Fig. 4, C and D). The low potency of MBG and marinobufotoxin in the triggering of death signal was further confirmed by measurement of caspase-3 activity (Table II).
Effect of Extracellular K ϩ -In K ϩ -free medium containing 4.1 M Rb, the rate of ouabain-sensitive 86 Rb influx was decreased by ϳ30-fold compared with medium containing 5 mM KCl (Fig. 5A). Na ϩ ,K ϩ pump inhibition seen in K ϩ -free medium resulted in ϳ10-fold elevation of [Na ϩ ] i content (Fig. 5B). About the same [Na ϩ ] i elevation was observed in cells treated  with 1 M ouabain in the presence of [K ϩ ] o . However, in contrast to ouabain, Na ϩ ,K ϩ pump inhibition in K ϩ -free medium did not affect cell survival estimated by phase contrast micros-copy (Fig. 6), cell attachment (Fig. 5D), caspase-3 activity (Table II), and slightly augmented rather than attenuated MTT staining (Fig. 5C). It is well documented that affinity of the Na ϩ /K ϩ pump for ouabain is increased in K ϩ -depleted medium (25)(26)(27). Indeed, in K ϩ -free medium, ouabain and MBG inhibited 86 Rb influx with ID 50 values of 0.04 and 0.2 M, respectively (Fig. 7A). In this medium, a half-maximal decline of MTT staining and cell attachment were revealed at ouabain concentrations of ϳ0.004 and 0.03 M, respectively (Fig. 7, B and C). Elevation of MBG in the range from 0.3 to 10 M resulted in a modest reduction of MTT staining (Fig. 7B) and sharp cell detachment (Fig. 7C) compared with K ϩ -containing medium (Fig. 4D). The distinct effect of 10 M MBG on attachment of cells in control and K ϩ -free medium was confirmed by phase contrast microscopy (Fig. 6).
Effect of MBG on Ouabain Binding and Signaling-Scatchard plot analysis showed that at [K ϩ ] o ϭ 5 mM, ouabain binding was fit by a single component of hyperbolic saturation with K d and B max values of ϳ40 nM and 3.5 pmol/mg protein, respectively (Fig. 8B). Incubation in K ϩ -free medium increased ouabain affinity (K d ϭ ϳ4 nM) but did not affect the number of binding sites (Fig. 8).
Recently, Ward et al. (28) reported that side by side with "classic" extracellular K ϩ -inhibited sites, bovine adrenocortical cells exhibit high affinity ouabain binding sites in the presence of 20 mM KCl, i.e. under conditions when its binding with Na ϩ ,K ϩ -ATPase is negligible. Keeping these data in mind, we examined the effect of [K ϩ ] o on ouabain binding in the range from 0.6 to 10 nM (Fig. 9). We observed that [K ϩ ] o elevation up to 20 mM led to further suppression of ouabain binding compared with control and K ϩ -free medium. These results are in contrast to the presence of high affinity ouabain-binding sites distinct from the Na ϩ ,K ϩ -ATPase ␣-subunit in C7-MDCK cells.
MBG dose-dependently inhibited ouabain binding (Fig. 10). The inhibitory action of MBG was in inverse proportion to ouabain concentration, indicating their competition for the same binding sites. Indeed, binding of 0.1 M ouabain was suppressed by 10 M MBG by ϳ10-fold, whereas at 1 M ouabain MBG attenuated its binding by only 2-fold (Fig. 10B). In contrast to CTS, we observed slight inhibition of ouabain binding and 86 Rb influx by 100 M acetylandrosol, whereas the other tested steroids were completely inactive (Table III).
In the presence of 5 mM K ϩ , ouabain triggered a full scale decline of MTT staining and cell detachment at concentrations of 0.1 and 1 M, respectively (Fig. 4). Table IV shows that both ouabain-induced signals were attenuated in the presence of 10 M MBG. Neither base-line MTT staining and cell attachment nor the decline of these parameters triggered by ouabain was affected by steroids distinct of CTS and listed in Table II (data  not shown). DISCUSSION The data obtained in the present study reveal that, similarly to ouabain, the addition of other tested cardenolides as well as bufalin, cinobufagin, cinobufotalin, and telobufotoxin resulted in decreased MTT staining and death of C7-MDCK cells. Both the decline of MTT staining and cell detachment of CTStreated cells were independent of the inhibition of Na ϩ ,K ϩ pump-mediated ion fluxes and elevation of the ATPase ␣-subunit. However, the potency of CTS in triggering of [Na ϩ ] i ,[K ϩ ] i -independent signaling is not proportional to their potency in inhibiting the Na ϩ ,K ϩ pump.
[Na ϩ ] i ,[K ϩ ] i -independent mechanisms of decreased staining with MTT and death of CTS-treated cells are supported by several observations. First, in K ϩ -free medium, Na ϩ ,K ϩ pump activity measured as the rate of ouabain-sensitive 86 Rb influx was ϳ2 orders of magnitude lower than in control medium containing 5 mM KCl (Fig. 5A). Similarly to ouabain, 5 h of incubation in K ϩ -free medium resulted in ϳ6-fold elevation of [Na ϩ ] i content (Fig. 5B) and massive loss of intracellular K ϩ (20). However, neither decreased MTT staining nor cell detachment was detected in K ϩ -free medium (Fig. 5, C and D). Second, at [K ϩ ] o ϭ 5 mM, maximal decrement of MTT staining was seen at 0.1 M ouabain (Fig. 4C). At this concentration, ouabain did not significantly affect [Na ϩ ] i content (Fig. 4B). Third, MBG and marinobufotoxin caused about the same Na ϩ ,K ϩ pump inhibition as was detected with other tested CTS (Table   I and Fig. 4A). At a concentration of 3 M, MBG led to ϳ6-fold elevation of [Na ϩ ] i content; the same effect was observed with 1 M ouabain. (Fig. 4B). However, in contrast to cardenolides and other bufadienolides, MBG and marinobufotoxin did not decrease MTT staining and only slightly affected cell attachment (Table I and Figs. 3

and 4, C and D).
During the last two decades, the MTT assay was widely employed to assess the redox state as a marker of mitochondrial function and/or cell viability (18). Keeping these data in mind, decreased MTT staining in CTS-treated cells might be considered as an early marker of induction of the cell death machinery. It should be underlined, however, that in control as well as in K ϩ -free medium, maximal decrement of MTT staining occurred at lower concentrations of ouabain than triggering of cell detachment (Figs. 4 and 7). In contrast, in K ϩ -free medium, 3 M marinobufagenin evoked detachment of cells without significant modulation of MTT staining (Fig. 7). Thus, additional experiments should be performed to study the mechanism of reduced MTT staining in CTS-treated cells and its relevance to the triggering of death signal.   It may be proposed that [Na ϩ ] i ,[K ϩ ] i -independent signaling triggered by CTS is mediated by a target distinct from the Na ϩ ,K ϩ -ATPase ␣-subunit. This hypothesis contradicts several observations. First, Scatchard analyses revealed the single class of [ 3 H]ouabain-binding sites (Fig. 8). Second, both ouabain binding (Fig. 8) and Na ϩ ,K ϩ pump inhibition by ouabain (Figs. 4A and 7A) were suppressed under [K ϩ ] o elevation, which is consistent with previously reported data (25)(26)(27). The same left-hand shift was detected in the effect of ouabain and MBG on cell detachment and MTT staining (Figs. 4, 6,  and 7). Third, neither cell attachment nor MTT staining was affected by steroids distinct from cardenolides and bufadienolides. This observation is consistent with the negligible effect of these compounds on ouabain binding and 86 Rb influx (Table III).
Our data show that being potent inhibitors of Na ϩ ,K ϩ -ATPase, MBG, and marinobufotoxin are much less potent inducers of [Na ϩ ] i ,[K ϩ ] i -independent cell death machinery and reduced MTT staining than the other CTS tested. Previously, it was shown that substitution of the amino by the hydroxy group at the 14␤ position of the ouabain analogue, compound LDN-623, sharply increased its ability to inhibit Na ϩ ,K ϩ -ATPase but attenuated its cytotoxic action (29). Daniel et al. (30) reported that incorporation of the 14,15-epoxy group in bufadienolide derivatives reduced their cytotoxicity. This observation is consistent with our data disclosing that the presence of the 14,15epoxy group in the absence of the 16-acetoxy group (Fig. 1) determines the attenuated cytotoxicity of MBG and marinobufotoxin.
Several crucial questions should be answered to clarify the novel function of Na ϩ ,K ϩ -ATPase as a receptor differentially affected by CTS. We do not know whether or not [Na ϩ ] i ,[K ϩ ] iindependent actions of CTS are limited to their interaction with the ␣1 Na ϩ ,K ϩ -ATPase subunit, i.e. the only isoform detected in the renal epithelium, or whether the cell death signal can be also generated by cell type-specific ␣2-␣4 subunits. Moreover, we do not know whether or not sharp differences in the efficiency of CTS, as triggers of these signaling pathways revealed in C7-MDCK cells, are applicable to other cell types abundant with ␣1 Na ϩ ,K ϩ -ATPase. Indeed, the cell type-specific effect of CTS on Na ϩ ,K ϩ pump activity and cell survival has been detected in comparison of wild type and mutated cell lines (31). Apart from cell type-and tissue-specific differences, CTS interaction with the Na ϩ ,K ϩ pump is affected by hypertension-induced cardiovascular remodeling (32) and extracellular stimuli, such as dietary salt (33) and modulators of protein kinase C activity (33,34). The role of these stimuli in the regulation of the [Na ϩ ] i /[K ϩ ] i -mediated and -independent signaling triggered by CTS should be examined further.