Ouabain Increases Sarcoplasmic Reticulum Calcium Release in Cardiac Myocytes

The role of Na⁺ pump inhibition in the inotropic and toxic actions of cardiac glycosides has been known for nearly 40 years. However, there have been other reports suggesting that additional actions might contribute to their cellular effects in vivo and in vitro, possibly involving direct actions of glycosides to increase SR Ca²⁺ release (Dutta et al., 1968; Besch and Watanabe, 1978; Fujino and Fujino, 1982; Isenberg, 1984). This suggestion is consistent with recent observations that low concentrations of glycosides activate the cardiac but not the skeletal isoform of the SR Ca²⁺ release channel (Rardon and Wasserstrom, 1990; McGarry and Williams, 1993; Sagawa et al., 2002). However, there has been no direct demonstration that glycosides might alter SR function and Ca²⁺ release under conditions where their well known actions on the sarcolemmal Na⁺ pump and on Na-Ca exchange are excluded.

The purpose of this study was to investigate the effects of ouabain and its analogs on SR Ca²⁺ release under experimental conditions designed to exclude the involvement of the Na⁺ pump. We studied the effects on Ca²⁺ sparks and waves in intact myocytes under Na⁺-free conditions where no Na⁺ is available to produce the accumulation necessary for pump inhibition to alter intracellular ionic balances. We also used saponin-permeabilized myocytes to “clamp” cytoplasmic [Ca²⁺] and [Na⁺] so that ionic accumulation could not contribute to any alterations in cell or SR function so that any change in intracellular Ca²⁺ handling was likely to be the result of direct actions on SR Ca²⁺ release.

Materials and Methods

Cat ventricular myocytes were obtained using collagenase perfusion of the coronary arteries of cat hearts obtained under pentobarbital anesthesia (25 mg/kg i.v.). Rat myocytes were obtained using a similar procedure after anesthetization using 35 mg/kg pentobarbital i.p. All procedures involving animals were performed using protocols approved by the Institutional Animal Care and Use Committee. Cells were placed in an experimental chamber on the stage of a laser scanning confocal microscope (LSM 510 system: Carl Zeiss, Thornwood, NY). A 25-mW argon laser was used to excite fluo-4 at 488 nm, and emitted light was collected at wavelengths >505 nm. Ca²⁺ sparks and waves were recorded in line scan mode using 1.5-2-ms scan rates for 8-s periods. The objective was a 40× water immersion lens (P-Apochromat, numerical aperture 1.2). Automated spark analysis was performed using a modification of a program developed by Cheng et al. (1999) based on IDL software (Research Systems, Inc., Boulder, CO). Spark magnitude had to exceed resting fluorescence by 3.2 standard deviation units to be included in the analysis.

SR Ca²⁺Release in Intact Cells under Na⁺-Free Conditions. Myocytes were loaded with fluo-4AM (5 μM for 30 min at room temperature) then washed twice in normal Tyrode's solution before being placed in the experimental chamber and allowed to settle to the laminin-coated coverslip floor. Superfusion was then initiated with Na⁺-free solution (Li⁺ replacement) which had a low CaCl₂ to prevent cell death due to Ca²⁺ overload (Nishio et al., 2002). It is not known whether these conditions are sufficient to ensure that cells are completely depleted of Na⁺; however, even if some Na⁺ remains in the cell, there is still no external source of Na⁺ that could allow the increase in internal Na⁺ necessary for Na⁺ pump inhibition to contribute to the cellular actions of ouabain. After 10 min of superfusion with Na⁺-free solution, the scan line was set, and control recordings were obtained. The superfusate was then switched to one containing ouabain (3 μM), and line scan images were recorded every minute for 5 min. In some experiments, caffeine (2-4 mM) was then added to the superfusate and recordings repeated.

Nominally Na⁺-free Tyrode's solution contained 140 mM LiCl, 5.4 mM KCl, 0.5 mM CaCl₂, 0.5 mM MgCl₂, 0.4 mM NaH₂PO₄, 10 mM HEPES, 5 mM glucose, and 20 mM 2,3-butanedione monoxime, pH 7.4 (LiOH).

Measurement of Cytoplasmic [Ca²⁺] in Saponin-Permeabilized Myocytes. Myocytes were allowed to settle on the floor of the chamber in normal Tyrode's solution. The solution was then replaced twice with Ca²⁺-free solution then with a permeabilization solution containing 0.01% saponin (30 s). The saponin-containing solution was then replaced with 0.5 ml of experimental solution containing 50 nM [Ca²⁺]. This approach has been used successfully to investigate Ca²⁺-induced Ca²⁺ release in cardiac myocytes with little evidence for direct effects on SR Ca²⁺ release (Lukyanenko and Györke, 1999). After confirmation of cell viability (maintenance of cross-striations and cylindrical shape), fluo-4 (5K⁺-salt, 20 μM) was added to the experimental solution. Line scan images then confirmed the presence of Ca²⁺ sparks. [Ca²⁺] was then maintained at 50 nM or adjusted 275 nM after which control recordings were obtained. Ouabain (20 nM) was then added to the experimental solution and recordings repeated at 2 and 3 min, at which point drug concentration was increased cumulatively up to 100 nM. Caffeine (2 mM) was then added to the solution and recordings repeated after 1 and 2 min.

Permeabilization solution contained 120 mM K-aspartate, 3 mM MgATP, 0.1 mM EGTA, 10 mM phosphocreatine (disodium), 20 mM HEPES, 5 U/ml creatine phosphokinase, 2% (w/v) dextran, and 0.01% saponin. Experimental solution contained 120 mM K-aspartate, 3 mM MgATP, 0.5 mM EGTA, and 0.114 mM CaCl₂, 10 mM phosphocreatine, 20 mM HEPES, 0.02 mM fluo-4 (5K⁺-salt), and 5 U/ml creatine phosphokinase.

Both protocols were also performed on rat myocytes to compare ouabain effects in a glycoside-insensitive species. Rat myocytes under Na⁺-free conditions were exposed to 100 μM ouabain, whereas saponin-permeabilized myocytes were exposed to 100 to 1000 nM ouabain before application of caffeine. All other experimental conditions were identical to those described for cat myocytes.

Statistics. All data are presented as mean ± S.E.M. Significance between data sets was evaluated with a one-way analysis of variance followed by a Newman-Keul's test (multiple comparisons), or the Student's paired t test (single comparison) where appropriate. Differences between means were considered significant if p < 0.05, unless specified otherwise.

Results

Effects of Ouabain on Ca²⁺Sparks under Na⁺-Free Conditions. A representative line scan image of a single cat ventricular myocyte in 0 Na⁺ conditions is shown in Fig. 1. Under these conditions, Ca²⁺ waves were commonly observed and were interspersed with numerous sparks. After a 5- to 10-min equilibration period in 0 Na⁺, ouabain (3 μM) was added to the superfusate. Within 2 min, spark and wave frequency began to decline and by 4 min (Fig. 1, middle), spark frequency was diminished and spark duration was prolonged. After 5 min, sparks and waves were nearly completely abolished (Fig. 1, bottom, left of arrow). Note that the increase in steady-state cytoplasmic fluorescence that was observed seemed to coincide with the loss of spark activity. The maintained SR Ca²⁺ release in the presence of ouabain was then overwhelmed by the massive release induced by caffeine (2 mM; Fig. 1, bottom, added at arrow). The caffeine response demonstrates that SR Ca²⁺ stores were well maintained during glycoside exposure, and therefore SR depletion was not the cause of the time-dependent decrease in Ca²⁺ spark frequency seen with maintained cardiac glycoside exposure.

Virtually identical results were obtained with dihydroouabain and ouabagenin (individual data not shown), so pooled results from all experiments are summarized in Fig. 2. Spark amplitude (Fig. 2A) was unchanged and duration was increased (Fig. 2C), whereas width (Fig. 2B) and frequency (Fig. 2D) were diminished. It should be noted, however, that the fact that statistical significance was achieved with the large sample size should not be taken to indicate that the very small changes induced by ouabain are in fact physiologically significant. In contrast to the small changes in spark characteristics, cardiac glycosides caused a reduction in Ca²⁺ wave frequency (Fig. 2E). In addition, when the steady-state fluorescence intensity was normalized to control resting fluorescence (excluding waves and sparks), there was an increase in intensity during exposure to ouabain (Fig. 3). This increase in resting fluorescence suggests that there is an increase in cytoplasmic [Ca²⁺] induced by ouabain, possibly occurring as a result of increased SR Ca²⁺ release. The subsequent increase in intensity by caffeine demonstrates that SR Ca²⁺ stores are well maintained during exposure to glycoside.

These results demonstrate that ouabain has a significant and rapid effect on intracellular SR Ca²⁺ release in the form of sparks and waves when added to the extracellular solution. However, it is not known whether these events are activated by Ca²⁺ influx across the sarcolemma via L-type Ca²⁺ channels or by local elevation in cytoplasmic [Ca²⁺] due to inhibition of Na⁺-Ca²⁺ exchange in the absence of Na⁺. To further study these potential mechanisms, effects of ouabain on SR Ca²⁺ release were studied in permeabilized myocytes where transsarcolemmal ion gradients were eliminated.

[Ca²⁺] Dependence of Ouabain Effects in Saponin-Permeabilized Cat Myocytes. We then investigated the effects of ouabain on sparks activated by cytoplasmic [Ca²⁺] rather than by transsarcolemmal Ca²⁺ influx. This was accomplished by permeabilization of the sarcolemma by saponin and exposure to an experimental solution containing known concentrations of free [Ca²⁺].

Figure 4 shows a line scan image of Ca²⁺ sparks in a cat ventricular myocyte when [Ca²⁺] = 275 nM. Sparks were numerous under control conditions (top recording). Two minutes after addition of ouabain (20 nM) to the experimental solution, spark frequency declined despite little change in peak amplitude, width, and duration. As ouabain concentration was increased cumulatively to 50 and 100 nM, sparks were abolished and there was a concentration-dependent increase in whole cell fluorescence.

The effects of ouabain to activate SR Ca²⁺ release were pronounced when [Ca²⁺] was higher than the normal resting concentration of about 100 nM. We then investigated the effects of ouabain on Ca²⁺ release when [Ca²⁺] was lower than normal to determine whether there was dependence of ouabain action on [Ca²⁺]. Figure 5 shows that, at [Ca²⁺] = 50 nM, there is a relatively low frequency of Ca²⁺ sparks in control, presumably because both load and trigger Ca²⁺ are much less than at 275 nM. When ouabain (20 nM) was added to the experimental chamber, there were few changes in spark characteristics and in whole cell fluorescence. The main change occurred in spark frequency, which was reduced at 20 nM ouabain, nearly abolished at 50 nM ouabain, and absent at 100 nM.

The effects of ouabain on spark frequency and steady-state fluorescence (normalized to control) are summarized in Figs. 6A and 7. There was little change in spark amplitude, width, and duration before they were abolished (Table 1). In terms of frequency, there is significantly less reduction in spark frequency at 20 nM ouabain at the [Ca²⁺] of 50 nM than at 275 nM (Fig. 6). This difference disappeared at 100 nM ouabain where sparks were almost entirely suppressed at both [Ca²⁺] tested. In contrast to the results at 275 nM Ca²⁺, however, when ouabain concentration was increased to 100 nM, whole cell fluorescence was unchanged (Fig. 6A and summary in Fig. 7). In contrast, spark frequency declined markedly at 20 nM (Fig. 6, open columns), and the steady-state fluorescence became brighter at cumulatively increasing ouabain concentrations (Fig. 7), just as occurred in intact cells (Fig. 3).

If ouabain is inducing SR Ca²⁺ release in a manner similar to the effects of caffeine, then addition of caffeine after ouabain exposure should have little additional effect on SR Ca²⁺ release. Figure 4, bottom, shows the results of an experiment in a saponin-permeabilized cat myocyte in which ouabain increased steady-state fluorescence compared with control ([Ca²⁺] = 275 nM). Subsequent addition of 2 mM caffeine caused no additional increase in fluorescence. When the same protocol is followed in an experiment at lower SR load ([Ca²⁺] = 50 nM), once again there is no additional increase in fluorescence. These results are summarized in Fig. 7 and confirm that the level of SR Ca²⁺ load is determined by free [Ca²⁺] in the solution as expected. Thus, there is indeed a difference in SR load at different [Ca²⁺] even in the presence of ouabain.

These data demonstrate that the ability of ouabain to activate SR Ca²⁺ release is dependent on the extent of SR Ca²⁺ loading. This increased release may take the form not of altered spark properties but rather as a suppression of spark frequency and increased SR Ca²⁺ leak. Sparks are suppressed by ouabain at both low and high load but a more general effect to enhance Ca²⁺ release or leak occurs only at the elevated SR load. In contrast, effects on spark frequency occur at both levels of SR load, but spark incidence is more sensitive to suppression by ouabain at higher SR Ca²⁺ load levels.

Effects of Ouabain on SR Ca²⁺Release in Rat Ventricular Myocytes. We also investigated the effects of ouabain on spark characteristics in rat ventricular myocytes, because this species is known to be less sensitive to the effects of cardiac glycosides than cat. In addition, we have also recently reported that single channel activity of Ca²⁺ release channels is activated only at extremely high concentrations of glycoside compared with sensitive species (Sagawa et al., 2002). The experiment illustrated in Fig. 8 shows that, like cat myocytes, intact rat cells also demonstrate a fairly high incidence of Ca²⁺ sparks and waves under control conditions. When ouabain (100 μM) was added to the experimental solution, there was little change in spark characteristics, frequency, steady-state fluorescence intensity, or wave frequency (Fig. 8). The summarized data in Fig. 9 and Table 2 show that, unlike in cat myocytes, Ca²⁺ sparks and waves are very insensitive to the effects of ouabain. Steady-state fluorescence was also unchanged after addition of ouabain to the superfusate (19 ± 18% increase above control, not significantly different, in seven myocytes from three rats).

In saponin-permeabilized rat myocytes, on the other hand, a high concentration of ouabain (100 nM) caused a slight decrease in spark frequency with only modest changes in steady-state fluorescence (Fig. 10). Only at 1 μM was there a clear decrease in spark frequency and an increase in steady-state fluorescence. Addition of caffeine caused a further increase in steady-state fluorescence beyond that induced by 1 μM ouabain. The summarized data in Fig. 6B show that there was significantly less suppression of spark frequency at 100 nM ouabain compared with that in cat myocytes under the same experimental conditions (p < 0.01). In contrast, 1 μM ouabain in permeabilized rat myocytes caused a similar degree of reduction in spark frequency to that observed with 100 nM ouabain in an identical preparation of cat myocytes (no significant difference).

When changes in steady-state fluorescence was measured during exposure to ouabain, it was only at the higher concentration of ouabain (1 μM) that there was a significant increase in intensity over control (Fig. 11). These summary data demonstrate that 1 μM ouabain produces an increase in fluorescence intensity in rat that is comparable with that induced by 50 nM ouabain in cat myocytes (a 20-fold decrease in sensitivity). In addition, caffeine causes a significant increase in steady-state fluorescence compared with 100 nM but not to 1 μM ouabain, suggesting that SR release at the higher concentration is of comparable magnitude to that of caffeine. Thus, it is possible to produce activation of SR Ca²⁺ release by ouabain in this relatively glycoside-insensitive species but only at much higher concentrations than in cat.

Discussion

Effects of Ouabain on SR Ca²⁺Release. Over the years, there has been extensive evidence presented that cardiac glycosides inhibit the sarcolemmal Na⁺ pump, which increases [Na⁺] and therefore [Ca²⁺] through the Na⁺-Ca²⁺ exchanger (Lee, 1985; Levi et al., 1994). There is little doubt that this action probably contributes both to the inotropic and the toxic (Ca²⁺ overload) effects of glycosides. In fact, a recent report demonstrated that the Na⁺-Ca²⁺ exchanger was essential to increasing intracellular Ca²⁺ transients in cardiac myotubes because its absence in a mice knockout was associated with the loss of response to ouabain (Reuter et al., 2002). However, there have also been numerous reports that are inconsistent with a single mechanism of action for both the inotropic and toxic effects for all such agents under all experimental conditions. These inconsistencies have been interpreted as suggesting that either or both actions might be influenced by additional cellular mechanisms independent of glycoside actions to inhibit the Na⁺ pump.

One of the most compelling suggestions for secondary actions has been an effect to activate the SR Ca²⁺ release channel (or ryanodine receptors). An intracellular site of action was first suggested by specific binding of different glycosides to SR fractions (Dutta et al., 1968) and increased ⁴⁵Ca²⁺ release from cardiac SR by ouabain (Fujino and Fujino, 1982). We then reported a direct SR action by several glycosides to increase single channel open probability (P_o) of canine crude cardiac release channels inserted into artificial lipid bilayers (Rardon and Wasserstrom, 1990). McGarry and Williams (1993) confirmed that digoxin (1-3 nM) increased P_o by increasing the number of openings with no effect on skeletal SR Ca²⁺ release channel activity. The effect of digoxin was the result of a sensitization of the channel to Ca²⁺. These authors also reported direct [³H]digoxin binding to SR vesicles, with both high (K_i = 10 nM) and low (K_i = 3.5 μM) affinity binding (McGarry et al., 1995).

In fact, there is additional pharmacological and physiological evidence suggesting the possibility of an intracellular site of action for cardiac glycosides (Park and Vincenzi, 1975; Fujino and Fujino, 1982; Nuñez-Duran et al., 1988; Sagawa et al., 2002). Isenberg (1984) found that intracellular injection of nanomolar concentrations of ouabain and digoxin caused positive inotropic effects in isolated bovine ventricular myocytes, even in the complete absence of Na⁺ or the presence of digoxin-specific antibodies outside the cell. We have also observed that, in Na⁺-free solutions, positive inotropic effects of ouabain persist in cat (but not rat) ventricular myocytes and that aftercontraction frequency is increased by ouabain (Nishio et al., 2002). Thus, there is some very compelling evidence for additional cellular actions of glycosides that might involve direct effects on SR function, possibly through direct activation of SR Ca²⁺ release channels.

In the current study, we found that ouabain altered SR Ca²⁺ release in the form of both Ca²⁺ sparks and waves in intact cat myocytes as well as producing a general increase in SR release in the form of enhanced leak, again in the absence of Na⁺. Whether the small but statistically significant changes in spark characteristics are also physiologically significant remains to be determined. However, ouabain and its analogs had rapid effects to decrease the incidence of both Ca²⁺ sparks and waves as well as to increase whole cell fluorescence. These data suggest that ouabain might have actions that do not involve intracellular Na⁺ accumulation resulting from Na⁺ pump inhibition. Because these observations were made in intact cells, it was not possible to distinguish between intracellular and sarcolemmal actions. However, we also observed that ouabain increased SR Ca²⁺ release after membrane permeabilization where sarcolemmal mechanisms are unlikely to contribute to alterations in SR function. Although we cannot be certain as to the mechanism by which ouabain caused SR Ca²⁺ release under these conditions, the effects of glycoside on SR Ca²⁺ release are consistent with the notion that glycosides may have direct actions on Ca²⁺ release channel behavior at low drug concentrations (Rardon and Wasserstrom, 1990; McGarry and Williams, 1993; Sagawa et al., 2002).

In addition, these data suggest that ouabain induces a release of SR Ca²⁺ stores that is reminiscent of the effects of caffeine. Consequently, exposure to caffeine after exposure to ouabain had little additional effect on SR Ca²⁺ release, possibly because release channels were already activated. The sustained elevation in fluorescence could then be the result of continuous cycling of SR Ca²⁺ that occurs in intact cells where trans-sarcolemmal removal of Ca²⁺ from the cytoplasm is prevented by the absence of Na⁺-Ca²⁺ exchange. In permeabilized myocytes, a similar net result could occur because an equilibrium exists between cytoplasmic, SR, and bulk [Ca²⁺] that may be capable of sustaining continuous SR Ca²⁺ cycling. Although it is difficult to relate these observed changes in SR Ca²⁺ release directly to inotropic or toxic effects of digitalis, the data indicate a direct action of glycosides on SR function that may contribute to alterations in contractility and/or development of toxicity. The nonphysiological conditions under which these data were obtained make it difficult to extrapolate to the normal changes in cell function ordinarily observed with glycosides. Nevertheless, the demonstration of direct changes in SR function observed here join an expanding list of studies suggesting that intracellular effects of digitalis might contribute to its actions in the intact heart.

It should be pointed out that we cannot exclude an effect of ouabain on IP₃ receptors in cardiac cells. The generalized response to increase whole cell SR Ca²⁺ release occurred within a few minutes of exposure to ouabain in the saponin-permeabilized myocytes. It is therefore unlikely that this effect of ouabain involves a sarcolemmal receptor for drug action but might be the result of a direct effect to activate IP₃ receptors themselves. The result might be expected to resemble and be indistinguishable from that of SR Ca²⁺ release channel (ryanodine receptor) activation. Further experimentation is required to determine whether the cardiac isoform of IP₃ receptors is activated by ouabain and participates in the increased SR Ca²⁺ release observed under these experimental conditions.

It is also possible that cardiac glycosides induce a nonspecific leak of Ca²⁺ from the SR, independent of an effect on RyR function. This seems unlikely because we would expect nonspecific actions to induce SR Ca²⁺ leak at drug concentrations far in excess of those we have previously found to activate RyR channel (<1 nM). Thus, although we cannot exclude the possibility of a nonspecific action to promote SR Ca²⁺ leak, we must await experimental verification of such an action before we can consider this as a possible explanation for our observations.

Mechanisms of Increased SR Ca²⁺Release by Ouabain: Lumenal Ca²⁺Sensitivity. One of the most compelling observations in this study was that low concentrations of ouabain increased SR Ca²⁺ release at high cytoplasmic [Ca²⁺] and therefore, presumably, at high SR Ca²⁺ load. Conversely, ouabain had little effect on SR Ca²⁺ release at low levels of SR Ca²⁺ load, both in terms of spark properties and whole cell fluorescence. These observations were predicted based on the fact that glycosides fail to activate single SR Ca²⁺ release channels at low physiological [Ca²⁺] in the SR lumen but become progressively more effective in increasing single channel open probability as lumenal [Ca²⁺] increases (Sagawa et al., 2002). As a result of these observations, we recently proposed a model (Sagawa et al., 2002) in which there was a synergistic interaction between pump inhibition and direct activation of Ca²⁺ release channels; pump inhibition increased SR load, which in turn increased the effectiveness of glycoside ability to stimulate single release channel activity. In the current study, we found that low cytoplasmic [Ca²⁺], which leads to reduced SR Ca²⁺ loading, produced little SR Ca²⁺ release by ouabain at concentrations up to 100 nM. In contrast, lower concentrations (50 nM ouabain) were able to induce a generalized increase in Ca²⁺ release at elevated SR load, which achieved statistical significance at 100 nM (Fig. 7). These observations in intact and permeabilized myocytes are consistent with the implications of the single channel data and underscore the importance of SR load on the cellular actions of glycosides.

It should also be pointed out that it is not yet clear whether the effects on SR Ca²⁺ release reported here contribute to either or both the inotropic and toxic effects of glycosides. The data suggest that the decline in sparks is associated with an increase in steady-state SR Ca²⁺ release, even though there is no evidence that this leads to a depletion of SR stores. If anything, the result seems to set up a cycle of release and reuptake that results in an overall increase in resting calcium. The physiological outcome of these effects, however, is not clear. A general Ca²⁺ release could raise [Ca²⁺] in the vicinity of the release channels, thus sensitizing the release channels to calcium by reducing the magnitude of the trigger required to activate release. The resulting reduction in threshold could promote toxicity by encouraging the phenomena associated with Ca²⁺ overload. It is also possible that this sensitization to a trigger for excitation-contraction coupling might also contribute to the positive inotropic effects of glycosides. Further study is required to determine how and whether these mechanisms might contribute to the inotropic and/or toxic actions of glycosides in the heart.

Species Differences in Sensitivity of SR Ca²⁺Release by Ouabain. The difference in species sensitivity to glycosides has been known for decades; rat is less sensitive to the effects of ouabain than cat, guinea pig, rabbit, dog, and human, among other species (Langer et al., 1975). The basis for this difference is unclear, but it has generally been assumed that it is the result of the fact that the α₁-isoform of the Na,K-ATPase, the predominant isoform in rat ventricle, has a lower affinity for glycoside binding than the α₃ isoform found primarily in the ventricles of sensitive species (McDonough et al., 1995). However, we have also found that the cardiac isoform of the release channel (RyR2) in rat is much less sensitive to activation by glycosides than in sensitive species (Sagawa et al., 2002). The implication of this result is that the low sensitivity to the inotropic and/or toxic effects of glycosides in rat might be in part a result of the reduced sensitivity of rat RyR2 to activation by glycoside compared with sensitive species.

Because the complete sequence of the rat and cat RyR2 proteins have not yet been reported, it is not known whether differences in primary and/or secondary structure might be responsible for the difference in binding affinity of glycoside to a putative high-affinity receptor. There is a very high sequence homology between rabbit and human RyR2 (98.6%; Tunwell et al., 1996) both of which are glycoside-sensitive species. Because we cannot compare sequence conservation of either of these with rat RyR2, it is difficult to predict how similar the structures of these proteins are in glycoside-sensitive and -insensitive species. However, even in the case of a very high level of sequence homology, the enormous size of the protein (∼565 kDa/monomer) could still allow important differences in ligand binding sensitivities. Even very minor differences in primary structure may reduce glycoside binding to a site on the rat protein or reduce access of drug to a putative site by altering protein folding. Alternatively, it is also possible that access to a glycoside binding site lies on an ancillary protein, whose affinity for the drug might be reduced in the rat as a result of different channel assembly in this species compared with cat. The precise location and affinities of a possible glycoside binding site on the RyR2 itself or on associated proteins in the complete channel assembly (e.g., FKBP12.6, calmodulin, junction, traidin, and calsequestrin) in different species remain to be investigated. The location of a glycoside binding site on ancillary proteins is not at all unreasonable, given the fact of a lack of activation of the purified RyR2 protein by glycoside compared the high-affinity activation of single channel protein derived from crude SR vesicles (Sagawa et al., 2002).

These conclusions are consistent with the current results where activation of SR Ca²⁺ release in rat was accomplished at higher concentrations than in cat. This difference in sensitivity between the two species could be the result of the lower sensitivity of rat Ca²⁺ release channel to activation by ouabain compared with the high sensitivity in cat. Thus, at a similar SR Ca²⁺ load, ouabain is more likely to increase SR Ca²⁺ release in cat than in rat. This effect can be overcome in rat only when ouabain concentration is increased to concentrations demonstrated to allow activation of SR Ca²⁺ release channels (100-1000 nM; Sagawa et al., 2002).

Access of Glycosides to SR Ca²⁺Release Channels. One of the questions raised in response to the possibility that ouabain affects SR Ca²⁺ release through a direct intracellular action concerns the ability of the drug to gain access to the intracellular receptor. Ouabain is considered to be one of the most hydrophilic glycosides and should therefore be unable to cross the lipid bilayer of the sarcolemma. However, despite its renown for high aqueous solubility, ouabain is also easily dissolved in ethanol and is slightly soluble in chloroform and must therefore have a measurable and distinct, albeit limited, solubility in lipid. Cohnen et al. (1978) measured the octanol/water partition coefficients for many glycosides and found that the value for ouabain was 0.01; thus, ∼1% of ouabain in a lipid/water mixture will be dissolved in the lipid. This fact provides a likely explanation for how micromolar external concentrations could cause accumulation of ouabain in the cytoplasm; if 1% of the external concentration becomes dissolved in the sarcolemma, it is then free to accumulate in the aqueous cytosolic compartment, at concentrations that far exceed that 1% present in the lipid membrane.

In fact, some studies have measured the rate and/or amount of [³H]ouabain accumulation in the cytoplasm. Ong et al. (1982) reported that [³H]ouabain accumulated in the cytoplasm of hepatic cells after i.v. injection with steady-state achieved within 0.5 min. The hepatic cell membrane served only as a minimal barrier for ouabain entry into the cell. Dutta et al. (1968a,b) reported that [³H]ouabain accumulated in cardiac SR; however, subsequent identification of specific binding to the Na⁺ pump caused many investigators to discard these observations as inconsequential to glycoside actions. Thus, it is likely that ouabain, in light of its physicochemical properties and despite its reputation as highly hydrophilic, is capable of rapidly crossing the sarcolemma, allowing intracellular concentrations that might affect intracellular targets, possibly including the SR Ca²⁺ release channel.