Partial inhibition of Na+/K+-ATPase by ouabain induces the Ca2+-dependent expressions of early-response genes in cardiac myocytes.

Exposure of neonatal rat cardiac myocytes to ouabain concentrations that caused partial inhibition of Na+/K+-ATPase but no loss of viability, increased c-fos and c-jun mRNAs and the transcription factor AP-1. The increased mRNAs were proportional to the extent of inhibition of Na+/K+-ATPase and the resulting rise in steady state intracellular Ca2+ concentration. The rapid and sustained increase of c-fos mRNA was shown to be due to increased transcriptional rate. Induction of c-fos by ouabain was prevented when either extracellular or intracellular Ca2+ was lowered and was attenuated by pretreatment of myocytes with a phorbol ester under conditions known to down-regulate protein kinase C. Exposure to ouabain for 24-48 h also increased total transcriptional activity and protein content of myocytes. The findings suggest that the same signal responsible for the positive inotropic action of ouabain, i.e. net influx of Ca2+ caused by partial inhibition of Na+/K+-ATPase, also initiates the rapid protein kinase C-dependent inductions of the early-response genes, the subsequent regulations of other cardiac genes by the resulting transcription factors, and stimulation of myocyte growth. Whether these hitherto unrecognized effects of cardiac glycosides are obtained in the intact heart and their relevance to the therapeutic uses of these drugs remain to be determined.

Ouabain and related cardiac glycosides are highly specific inhibitors of Na ϩ /K ϩ -ATPase. This enzyme (the sodium pump) catalyzes the coupled active transport of Na ϩ and K ϩ across the plasma membranes of most animal cells (1). It is now well established that the positive inotropic effect of a cardiac glycoside on the myocardium is due to the partial inhibition of the cardiac Na ϩ /K ϩ -ATPase, causing a small increase in intracellular Na ϩ , which in turn affects the sarcolemmal Na ϩ /Ca 2ϩ exchanger, leading to an increase in intracellular Ca 2ϩ and in the force of contraction (2)(3)(4). This effect on cardiac contractility is the basis of the major therapeutic use of these drugs in the treatment of congestive heart failure (2)(3)(4). There is also evidence to suggest that ouabain or related cardiac glycosidelike substances may be paracrine hormones in higher animals (5).
In several cell types other than cardiac myocytes, inhibition of Na ϩ /K ϩ -ATPase, either by ouabain or by low extracellular K ϩ , has been shown to affect the expression of proto-oncogenes c-fos and c-jun (6 -8). Because the inductions of such earlyresponse genes have been implicated in cardiac growth and hypertrophy (9,10), and since the hypertrophied failing heart is the primary target of therapy with cardiac glycosides, it was of interest to explore the effects of these drugs on the expressions of cardiac early-response genes. Here, we report the results of our initial studies in this direction, using the primary cultures of neonatal rat cardiac myocytes as a model.

EXPERIMENTAL PROCEDURES
Neonatal ventricular myocytes were prepared and cultured as described before (11,12). Briefly, myocytes were isolated from ventricles of 1-day-old Sprague-Dawley rats and purified by centrifugation on Percoll gradients. Myocytes were then cultured at a density of 5 ϫ 10 4 cells/cm 2 in a medium containing four parts of DME 1 and one part Medium 199 (Life Technologies, Inc.), penicillin (100 units/ml), streptomycin (100 g/ml), and 10% fetal bovine serum. After 24 h of incubation at 37°C in humidified air with 5% CO 2 , the medium was changed to one with the same composition as above, but without the serum. Unless specified otherwise, all experiments were done at 37°C after 24 h of additional incubation under serum-free conditions. These cultures contained more than 95% myocytes as assessed by immunofluorescence staining with a myosin heavy chain antibody (11). For experiments on myocyte growth (Table I) culture media also contained 0.1 mM bromodeoxyuridine during the first 48 h. The great majority of the serumstarved myocytes were quiescent or contracted infrequently (13). Treatment with ouabain increased the number of myocytes beating regularly. No attempt was made to quantitate these effects on beating or on the contractile force. For some experiments with myocytes, the nominally Ca 2ϩ -free DME/F-12 base medium (Sigma) was used. Incubations of myocytes in this medium for the indicated short durations did not affect myocyte viability. Rat2 cells (CRL 1764, American Type Culture Collection) and HeLa cells were cultured in DME medium supplemented with 10% fetal bovine serum. The cells (about 90% confluence) were serum-starved for 24 h before use in experiments.
For Northern blot analysis, total RNA was isolated using TRI reagent (Molecular Research Center, Inc.) as recommended by the manufacturer. Routinely, 10 -15 g of total RNA was subjected to gel electrophoresis, transferred to a Nytran membrane, UV-immobilized, and hybridized to 32 P-labeled probes. The probes used for c-fos, c-jun, and GAPDH were prepared as described before (11). Autoradiograms obtained at Ϫ70°C were scanned with a Bio-Rad densitometer. Multiple exposures were analyzed to ensure that the signals were within the linear range of the film. The relative amount of RNA in each sample was normalized to that of GAPDH mRNA to correct for differences in sample loading and transfer.
For nuclear run-on assays, myocyte nuclei were isolated as described by Lee et al. (14) and counted in 0.04% trypan blue using a hemocytometer. To label the nascent RNA transcripts, the nuclei (3 ϫ 10 5 ) were * This work was supported by National Institutes of Health Grant HL-36573 awarded by NHLBI, United States Public Health Service, Department of Health and Human Services. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  incubated in 0.14 M KCl, 10 mM MgCl 2 , 1 mM MnCl 2 , 14 mM 2-mercaptoethanol, 20% glycerol, 0.2 M Tris (pH 8.0), 0.1 mg/ml creatine kinase, 10 mM phosphocreatine, 1 mM each of ATP, GTP, CTP, 0.03 M UTP, and 100 Ci of [␣-32 P]UTP (DuPont NEN, 3,000 Ci/mmol) for 15 min at 30°C. The nuclei were collected, lysed, and digested with RNase-free DNase (Worthington). Fifty g of carrier yeast tRNA were added, and the 32 P-labeled run-on RNA produced was then isolated using TRI reagent (Molecular Research Center, Inc.). Purified [ 32 P]RNA was counted as an index of total transcriptional activity. Equal counts of 32 P-labeled RNA from different groups were used for hybridization. The probes were applied to Nytran membrane through dot-blot apparatus, denatured, and immobilized by UV-cross-linking.
Electrophoretic mobility shift assay for AP-1 binding activity was done as described previously (11). In brief, after myocyte nuclei were isolated, they were extracted in a solution containing 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 26% glycerol (v/v), 5 mM Hepes (pH 7.9), and 0.3 M NaCl. Nuclear protein (3-5 g) was used in the binding reaction; the reaction mixture (20 l) contained 50 mM NaCl, 1 mM EDTA, 1 g of acetylated bovine serum albumin, 2 g of poly(dI-dC), 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10% glycerol (v/v), 5 mM MgCl 2 , 10 mM Tris-HCl (pH 7.5), and about 30,000 cpm of 32 P-labeled AP-1 probe. Control reaction mixtures for the detection of specific binding contained 100fold excess of unlabeled probe. DNA-protein complexes were separated on nondenaturing 4.5% polyacrylamide gels in 90 mM Tris borate, 2 mM EDTA (pH 8.0). Gels were vacuum-dried and exposed to x-ray film at Ϫ70°C for 12 to 24 h. The oligonucleotide probe for AP-1 was synthesized according to the sequences as described by Meyer et al. (15) and labeled by T4 DNA kinase using [␥-32 P]ATP. Intracellular Ca 2ϩ concentration was determined using the Ca 2ϩsensitive fluorescent probe fura-2 as we have previously described (16). Briefly, the myocytes were cultured on glass coverslips and loaded with 5 M fura-2/AM for 30 min at room temperature. The coverslip was affixed to a culture chamber and perfused for 15 min with the medium containing the indicated ouabain concentration before measurements were made on about 20 single cells. A single cell was isolated from the surrounding area using a pinhole aperture under Nikon microscope, and single cell fluorescence was recorded at an emission wavelength of 505 nm and excitation wavelengths of 340/380 nm using Spex Fluoro II dual beam spectrofluorometer (Spex Industries, Inc.). Measurement of time-averaged signals on each single cell was completed in 30 s. The relationship between the fluorescence ratio (340/380) and steady state Ca 2ϩ concentration was established by calibration procedures and calculations described before (16,23). Because of the well-known uncertainties associated with such calibrations, however, the indicated values should be considered as relative changes in intracellular Ca 2ϩ concentrations.
For the assay of 86 Rb ϩ uptake, myocytes cultured in 12-well plates were washed and incubated in 1.0 ml of uptake medium containing 150 mM NaCl, 5 mM RbCl, 10 mM Hepes (pH 7.0), 5 mM MgCl 2 , 0.5 mM CaCl 2 , and 10 mM glucose. After 10 min of incubation with different concentrations of ouabain, 86 Rb ϩ uptake was started by the addition of 1 Ci of 86 Rb ϩ , and the reaction was stopped after 10 min by washing four times with 3 ml of ice-cold 0.1 M MgCl 2 . The cells were then treated with 10% trichloroacetic acid at 4°C for 60 min, and trichloroacetic acid-soluble 86 Rb ϩ was counted. Trichloroacetic acid-precipitated cellular protein was dissolved in 0.1 N NaOH, and protein content was determined (17), using bovine serum albumin as standard. DNA was assayed fluorometrically (48) using calf thymus DNA as standard. Myocyte viability was assayed by the extent of release of LDH into the medium (11).

RESULTS
Ouabain Effects on Myocyte Viability, Rb ϩ Uptake, and Intracellular Ca 2ϩ -Time courses of the effects of different ouabain concentrations on the viability of the serum-starved myocytes were studied in experiments, the results of which are summarized in Fig. 1. At concentrations of 0.1 mM or less, ouabain did not affect viability up to 24 h of incubation. At higher ouabain concentrations, there was time-dependent loss of viable cells. These data were used as a guide for the remaining experiments on myocytes; i.e. when ouabain concentrations higher than 0.1 mM were used, exposure times did not exceed 45 min. Hence, there was no significant ouabain-induced decrease in the number of viable cells in any of the experiments presented below.
Inhibitory effects of different ouabain concentrations on 86 Rb ϩ uptake, and on intracellular Ca 2ϩ concentrations, were determined in experiments shown in Fig. 2. Our findings on inhibition of 86 Rb ϩ uptake were in general agreement with previous observations on dose-dependent inhibitory effects of ouabain on pump fluxes in rat neonatal cardiac myocytes (18,19) and on Na ϩ /K ϩ -ATPase activities of these myocytes and the rat heart (2,20).
As expected from a large body of previous work relating ouabain-induced inhibition of the cardiac sodium pump to intracellular Ca 2ϩ (2)(3)(4)(21)(22)(23), our data showed significant increases in intracellular Ca 2ϩ concentrations of these myocytes in parallel with ouabain-induced inhibition of 86 Rb ϩ uptake (Fig. 2).
The following aspects of the data of Figs. 1 and 2 are worthy of note at this point. 1) That complete inhibition of pumpmediated 86 Rb ϩ uptake requires 1 mM or higher concentrations of ouabain ( Fig. 2) is consistent with the fact that of the two isoforms (␣ 1 and ␣ 3 ) of Na ϩ /K ϩ -ATPase present in neonatal rat cardiac myocytes, the predominant one is the relatively ouabain insensitive ␣ 1 isoform (24, 25). 2) The observation that 0.1 mM ouabain causes significant inhibition of the pump (Fig. 2) without affecting myocyte viability ( Fig. 1) is consistent with the well-established fact that the heart tolerates up to about 50% inhibition of Na ϩ /K ϩ -ATPase without overt toxicity (3).
Inductions of Myocyte c-fos and c-jun by Ouabain-Effects of the exposure of a myocyte preparation to different ouabain concentrations for 45 min on c-fos and c-jun mRNAs are shown in Fig. 3A. Experiments similar to this were repeated in several different preparations, and the results were combined as shown in Fig. 3B. The data clearly indicate that ouabain increased both messages in a dose-dependent manner. These dose-response curves are in good agreement with ouabain's dose-response curves on rat cardiac contractility (2), rat cardiac sarcolemmal Na ϩ /K ϩ -ATPase (2), and Na ϩ /K ϩ -ATPase of rat neonatal cardiac myocytes (20). That ouabain's effect on c-fos was more pronounced than that on c-jun (Fig. 3B) is similar to the differential effects of other stimuli (e.g. Ref. 11) on c-fos and c-jun expressions in cardiac myocytes. When time-dependent changes in c-fos mRNA in response to 0.1 mM ouabain were measured (Fig. 4), a significant increase was noted within 15 min after exposure, the maximal increase was obtained at 3 h, and the level gradually declined to that of control after 24 h of continuous exposure to ouabain.
It was of interest to know if the above ouabain effects on serum-starved myocytes were also observed in the presence of serum. Comparison of ouabain effects on c-fos mRNA in cells cultured for 48 h in the presence of serum, with the effects in our standard myocyte preparation (24 h in serum-containing medium, followed by 24 h in serum-free medium) showed nearly identical results (Fig. 5). When serum-starved cells were exposed for 45 min to 0.1 mM ouabain, or 10% fetal bovine serum, or the combination of the two, ouabain-induced increase in c-fos mRNA was greater than that caused by serum, and the combined effects of the two agents were not greater than that of ouabain alone (data not shown). All remaining experiments were done with the serum-starved myocytes to minimize contamination by other cell types (see "Experimental Procedures").
The nuclear run-on experiments of Fig. 6 showed that ouabain-induced increase in steady state concentrations of c-fos mRNA was due, at least in part, to increase in transcriptional rate of c-fos.
To determine if the above inductions were accompanied by the formation of AP-1, containing Fos and Jun and capable of binding to DNA, nuclear extracts of control and ouabain-exposed myocytes were subjected to mobility shift assays. The results (Fig. 7) showed that ouabain increased a nuclear protein that could bind to a probe containing the AP-1 binding site. In experiments the results of which are not shown, we also noted that the nuclei of myocytes treated with ouabain contained significantly more Fos and Jun proteins than control nuclei, as determined by immunostaining procedures that we have used before to show the inductions of these proteins by other stimuli (11).
Dependence of Ouabain-induced c-fos Expression on Ca 2ϩ and PKC-Ouabain effect on c-fos mRNA was blocked when myocytes were pretreated with cell permeant BAPTA-AM to buffer intracellular Ca 2ϩ (Fig. 8). Pretreat with BAPTA-AM, which has been shown to block phenylephrine's effect on c-fos expression (28), also blocked the effect of PMA on c-fos expression (Fig. 8). This suggests a role of intracellular Ca 2ϩ in the pathways of the effects of ouabain, phenylephrine, and PMA on c-fos. When myocytes were incubated in a nominally Ca 2ϩ -free medium, ouabain had no effect on expression of c-fos mRNA, whereas phenylephrine and PMA, two well-established inducers of c-fos (12,26), remained effective (Fig. 9). Addition of Ca 2ϩ to the medium restored the effect of ouabain (Fig. 9), clearly establishing the necessity of ouabain-induced net influx of extracellular Ca 2ϩ for ouabain's effect on c-fos. In similar experiments (not shown), it was established that repletion of medium Ca 2ϩ did not have a significant effect on phenylephrine's or PMA's effect on c-fos expression. Evidently, the effects of these stimuli do not require the influx of extracellular Ca 2ϩ .
Preincubation of myocytes with PMA for 24 h, which is known to down-regulate PKC in these cells (27,28), suppressed ouabain-induced c-fos expression (Fig. 10), indicating an important role of one or more isoforms of PKC in the pathway of response to ouabain.
Ouabain Effects on Total Myocyte Protein and Transcrip- tional Activity-Since many of the stimuli that induce earlyresponse genes also cause hypertrophic growth of cardiac myocytes (10,29,30), we compared the effects of ouabain with those of phenylephrine and PMA on myocyte protein content ( Table  I). The large increases caused by phenylephrine and PMA were in agreement with previous observations (26). The increase in total protein caused by ouabain was significant but smaller than those caused by phenylephrine and PMA. Experiments of Fig. 11 showed that total transcriptional activity was also increased significantly after 24 h of exposure of myocytes to 0.1 mM ouabain.
Ouabain Effects on c-fos Expression in HeLa and Rat2 Cells-Previous studies of Nakagawa et al. (6) showed that ouabain-induced increases in c-fos and c-jun transcriptions in several human cell lines (e.g. HeLa cells), and in mouse fibroblasts, required ouabain concentrations that cause complete inhibition of Na ϩ /K ϩ -ATPase. Our experiments on c-fos expression in HeLa cells confirmed their findings. A ouabain-induced increase in c-fos mRNA required 3-4 h of exposure to 1 M or higher concentrations of ouabain, and the maximal level of mRNA was about 15-fold greater than the control level. HeLa cells contain a highly ouabain-sensitive isoform of Na ϩ /K ϩ -ATPase that is almost completely inhibited at 1 M ouabain (31). We also did limited studies to compare ouabain's effects on cardiac myocytes with those on Rat2 fibroblasts. This cell line's only Na ϩ /K ϩ -ATPase isoform is the same relatively ouabainresistant ␣ 1 that exists in rat cardiac myocytes. 2 The results of these experiments showed that (a) ouabain concentrations of 0.5 mM or higher were required to note an increase in c-fos mRNA and (b) the maximal level of c-fos mRNA, about 5 times that of control, was obtained after 15 min of exposure to ouabain and rapidly declined to control level after about 1 h of exposure. Thus, ouabain effects on c-fos of these fibroblasts differed greatly, both in magnitude and duration, from those noted in myocytes (Figs. 3 and 4). DISCUSSION The data presented here show that in neonatal rat cardiac myocytes: (a) the proto-oncogenes c-fos and c-jun are rapidly induced by ouabain concentrations that inhibit the sodium pump partially and raise intracellular Ca 2ϩ , but do not cause loss of viability, (b) the sustained induction of c-fos is dependent on extracellular Ca 2ϩ and its net influx is due to ouabaininduced inhibition of the pump, (c) PKC is likely to be involved in this pathway of ouabain-induced gene regulation, and (d) along with the above effects, ouabain stimulates the growth of these myocytes.
Cardiac Specificity of Ouabain Effects on Early-response Genes-Nakagawa et al. (6) were the first to report the transcriptional regulation of early-response genes by ouabain. Their studies on HeLa and some other cell lines showed that increased expressions of c-fos and c-jun required several hours of exposure to ouabain concentrations that cause near complete inhibition of Na ϩ /K ϩ -ATPase. Our limited experiments on HeLa cells (see "Results") have confirmed their findings. Nakagawa et al. (6) also noted that in these cell lines, ouabaininduced expression of c-fos was neither Ca 2ϩ -dependent nor affected by depletion of PKC. It is clear, therefore, that the slow induction of the genes by the toxic concentrations of ouabain in the cells used by these investigators occur through mechanisms different from the ouabain-induced pathways in cardiac myocytes. More recently, Golomb et al. (7) reported that in vascular smooth muscle cells, ouabain did not affect the expression of c-fos in the absence of serum, but that the mitogenic effects of serum, including the inductions of early-response genes, were enhanced by ouabain. In cardiac myocytes, however, ouabaininduced regulation of c-fos is independent of serum (Fig. 5), and no synergistic effects of serum and ouabain are apparent (see "Results"). To our knowledge, the only other prior report relating the early-response genes to the sodium pump is that of Cayanis et al. (8), in which the inhibition of sodium pump by   48 h in the absence of bromodeoxyuridine (Experiment 1) or in its presence (Experiments 2 and 3) as described under "Experimental Procedures." The serum-starved cells were then incubated for an additional 48 h in the presence of ouabain, phenylephrine, PMA, and without these agents (control), before they were washed and harvested for assays. No significant loss of viability was noted either with the ouabain concentration used (Fig. 1) or with other treatments. In each experiment, 6 dishes were used for each indicated condition, all dishes containing equal numbers of cells (5 ϫ 10 4 /cm 2 ). In Experiments 1 and 2, the total protein of each dish was determined; the indicated results are means Ϯ S.E. of these determinations for each condition expressed as percent of protein contents of control dishes. In Experiment 3, protein and DNA contents of each dish were determined; the indicated results are means Ϯ S.E. of protein/DNA ratios for each condition as percent of the same ratio for control dishes. Trichloroacetic acid-precipitated protein and DNA were assayed as indicated under "Experimental Procedures." All treatments increased protein significantly (p Ͻ 0.05).  11. Effect of ouabain on total transcriptional activity of myocytes. Cells were incubated in the presence of 0.1 mM ouabain and in its absence for 24 h. Nuclei were prepared and subjected to run-on transcriptional analysis as described under "Experimental Procedures." Values are mean Ϯ S.E. of 5 experiments (p Ͻ 0.05). low extracellular K ϩ was shown to cause a modest, rapid, and transient increase in c-fos mRNA in a liver cell line. This pattern, and a similar one that we observed in Rat2 fibroblasts (see "Results"), also differ from the large and sustained ouabain-induced increases in c-fos mRNA in cardiac myocytes (Figs. 3 and 4). Clearly, based on available evidence, there seem to be cardiac-specific regulatory effects of ouabain on expressions of proto-oncogenes.
Comparison of Ouabain with Other Inducers of Cardiac Early-response Genes-A number of stimuli induce early-response genes in cultured cardiac myocytes (9,10), including ␣and ␤-adrenergic agonists (12,30), angiotensin II (32), stretch (33), endothelin (34), purinergic receptor agonists (35), phorbol esters (26), hypoxia (36), and oxidants (11). Examination of this literature shows differences between ouabain effects and those of any other stimulus. The evident differences are either quantitative, or qualitative, or both. For example, the magnitude of the increase in c-fos mRNA induced by ouabain in neonatal myocytes is matched only by that of PMA ( Fig. 9 and Ref. 26), and, while the effects of ouabain and phenylephrine on c-fos are regulated by intracellular Ca 2ϩ concentrations and PKC (Figs. 8 and 10, and Ref. 28), only ouabain's effect, but not that of phenylephrine, is dependent on the net influx of extracellular Ca 2ϩ ( Fig. 9 and "Results"). Interestingly, ATP, which also causes a net influx of Ca 2ϩ , albeit by a mechanism different from that of ouabain (37), has a much smaller effect on c-fos mRNA (35) than that of ouabain. These and other similarities and differences between the various regulators of cardiac earlyresponse genes point to the existence of both conserved and divergent pathways of early-response gene regulation and the need for the identification of control mechanisms specific to each stimulus.
Mechanisms of Ouabain Effects on c-fos Expression-Although ouabain's effects are evident on both c-fos and c-jun, we have focused most of our initial studies on c-fos due to the wealth of knowledge on the regulatory mechanisms of this gene and because the magnitude of ouabain effect on c-fos mRNA is about 10-fold greater than that on c-jun mRNA (Fig. 3).
An important question is whether the rise in intracellular Ca 2ϩ that is the well-established result of the partial inhibition of cardiac Na ϩ /K ϩ -ATPase (2)(3)(4) can account for the observed induction of c-fos by ouabain. The c-fos promoter contains two well-characterized elements, SRE and Ca 2ϩ response element/ cAMP response element, one or both of which are known to be regulated by rise in intracellular Ca 2ϩ in various cells including neonatal rat cardiac myocytes (33,38,39). Since induction of c-fos by ouabain involves PKC (Fig. 10), and SRE is known to be stimulated by PKC (38,40), it seems that ouabain induction of c-fos may be, at least in part, through SRE. However, because of the complex and multiple mechanisms of Ca 2ϩ effects on the activations of the above two elements of the c-fos gene (38,39,41), our present data are insufficient either to establish the definite involvement of SRE or to rule out the role of Ca 2ϩ response element/cAMP response element in c-fos induction by ouabain. Studies aimed at clarification of these issues are in progress.
There is uncertainty regarding the mechanism of PKC activation in the course of induction of c-fos by ouabain. Although increased intracellular Ca 2ϩ may cause direct activation or redistribution of PKC isoforms (28,42), it is not clear that ouabain-induced increase in myocyte Ca 2ϩ is responsible for PKC activation. Previous studies have demonstrated that both inotropic and toxic concentrations of ouabain enhance phosphoinositide turnover, increase diacylglycerol content, and activate PKC in the myocardium (43,44). While these effects may be the consequence of activation of phospholipase C by oua-bain-induced increases in intracellular Ca 2ϩ (43), there is also some evidence to suggest that activation of PKC by ouabain may occur by unidentified mechanisms that are independent of a ouabain-induced rise in intracellular Ca 2ϩ (44).
Based on the above considerations, it is reasonable to conclude that the same signal (increased net influx of Ca 2ϩ ) that accounts for the classical effect of ouabain on contractility is also responsible for the initiation of ouabain effect on c-fos expression, but that additional pathways independent of the rise in intracellular Ca 2ϩ may also be involved in response of the c-fos to ouabain.
Ouabain's Effect on Myocyte Growth-Our findings (Table I  and Fig. 11) clearly show that ouabain, like most other inducers of cardiac early-response genes, increases protein content and stimulates myocyte growth. At first sight, these findings may seem to be in conflict with some previous observations. There is a large body of older literature (45) showing inhibitory effects of ouabain on transformation and proliferative growth of several mammalian cell types and indicating the important role of the sodium pump in these processes. Examination of this literature reveals that these effects of ouabain are accompanied by significant dissipation of the normal Na ϩ and K ϩ gradients and are most likely due to the well-established inhibitory effects of low intracellular K ϩ on protein synthesis (45). There are also limited data showing that in ferret papillary muscle, ouabain reduced protein synthesis, in contrast to several other inotropic agents that stimulated protein synthesis (46). The ferret heart contains nearly equal amounts of two isoforms of the sodium pump, one of which is highly sensitive to ouabain (47). Based on the ouabain sensitivity of total Na ϩ /K ϩ -ATPase of ferret heart (47), it seems likely that the noted ouabain inhibition of protein synthesis (46) was also due to large changes in intracellular Na ϩ and K ϩ . It is well established that when the myocardium is exposed to inotropic but nontoxic ouabain concentrations, a significant increase in intracellular Ca 2ϩ is obtained without large changes in intracellular Na ϩ and K ϩ concentrations (2)(3)(4)(21)(22)(23). We suggest that the primary effect of such lower concentrations of ouabain on protein synthesis and myocyte growth will be stimulatory as in experiments of Table I. Extensive studies on hypertrophic growth of cardiac myocytes have shown that contraction stimulates growth (48,49), but that for some stimuli induction of growth is independent of the enhancement of contractility (30,49). It is also well-established that cardiac hypertrophy caused by many stimuli is associated with the selective induction of the fetal gene program (29,30). In these contexts, we do not have sufficient data to compare ouabain with other hypertrophic stimuli. In our experiments, ouabain clearly affected myocyte contractility as expected (see "Experimental Procedures"), although we did not quantitate these effects. It remains to be seen if ouabain's effect on growth can be distinguished from its effects on contractility. Studies are also in progress to determine whether or not ouabain exerts selective effects on the cardiac fetal gene program.
Physiological and Therapeutic Implications-Studies similar to those reported here need to be done with the heart, because there are significant differences between the responses of cultured myocytes and isolated heart preparations to some, but not all, inducers of cardiac early-response genes (50). Assuming that the present findings may be extrapolated to the intact heart, however, ouabain's rapid effects on the early-response genes, and the transcription factor AP-1, suggest that soon after its effect on contractility, ouabain begins to exert potentially profound effects on the expressions of a multitude of cardiac proteins. AP-1 and related proto-oncogene products have been shown to regulate several cardiac genes, including those of contractile proteins (51-53). The sodium pump genes also contain AP-1 or related binding elements (54,55), and these have been implicated in the control of pump expression in liver cells (56). Since partial inhibition of the sodium pump by ouabain is known to cause a subsequent increase in functional pump sites in cardiac myocytes (57,58), it is reasonable to think that the adaptive induction of cardiac sodium pump may also be a consequence of ouabain-induced stimulation of AP-1. Ouabain's apparent effect on myocyte growth raises intriguing questions about the chronic treatment of the hypertrophied failing heart with a hypertrophic stimulus. Until the nature of these previously unrecognized effects of cardiac glycosides are clarified, it is not possible to say whether they are a part of the beneficial or the undesirable effects of these drugs. These questions may be pertinent to the current efforts to reassess the value of cardiac glycosides and related drugs for the treatment of congestive heart failure (59).