Cardiovascular Research

Cardiovascular Research 1998 37(2):335-345; doi:10.1016/S0008-6363(97)00261-7
© 1998 by European Society of Cardiology

This Article Extract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Schmidt, T. A Articles by Kjeldsen, K. Search for Related Content PubMed PubMed Citation Articles by Schmidt, T. A Articles by Kjeldsen, K. Social Bookmarking          What's this?

Copyright © 1998, European Society of Cardiology

Human myocardial Na,K-ATPase — quantification, regulation and relation to Ca

Thomas A Schmidt^* and Keld Kjeldsen

Department of Medicine B 2142, The Heart Center, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark

^* Corresponding author. Tel. (+45) 3545 2871; Fax: (+45) 3138 3186.

Received 14 August 1997; accepted 17 October 1997

	1 The Na,K-ATPase

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
The Na,K-ATPase (sodium, potassium adenosine triphosphatase) or the Na,K-pump is an ubiquitous enzyme which performs the active transport of Na and K across the cell membrane [1]. Thus, by hydrolysing ATP the Na,K-ATPase pumps 3 Na ions out of the cell and 2 K ions into the cell. Hence, Na,K-ATPase is of importance for the characteristic distribution of these cations across the cell membrane and for generation of the membrane potential; thus optimum Na,K-ATPase function is required to ensure excitability and conduction. By generating the Na gradient across the sarcolemma, Na,K-ATPase creates a driving force for active Ca extrusion via the Na — Ca exchanger embedded in the membrane. Together with the Ca-ATPase located in the sarcolemma and in the membrane of the sarcoplasmic reticulum, the Na-Ca exchanger is of importance for maintaining low diastolic Ca concentration [2]. When muscular cells are depolarized intracellular Ca rises as much as 10 fold to around 1 µmolxlitre^–1 [3]. Subsequently Ca binds to troponin C triggering the force generating interaction between actin and myosin. Thus, Na,K-ATPase has the capacity for influencing muscular force generation. It might be noted that whereas the Na,K-pump utilizes around 4% of myocardial energy expenditure to maintain Na,K gradients, as much as 30% is used to fuel the Ca-ATPase [4].

Cardiac glycosides bind to the Na,K-ATPase in a 1: 1 ratio [5], and myocardial Na,K-ATPase is generally accepted to be the target for its positive inotropic action eliciting an increase in the availability of Ca to the contractile proteins in heart cells [6]. Cardiac glycosides bind with high affinity and specificity to an inhibitory site on the subunit of Na,K-ATPase facing the outside of the cell, resulting in the complete inhibition of Na and K transport by the enzymatic unit so long as the cardiac glycoside molecule remains attached to the site. In myocardial plasma membranes this inhibition reduces the driving force for Na entry and coupled Ca extrusion via the Na-Ca exchanger leading to increased cytosolic Ca and contractile force. Thus, general consensus has emerged supporting the role of the Na-Ca exchanger in exerting positive inotropy as a result of cardiac glycoside binding to Na,K-ATPase [6, 7]. Additional cellular mechanisms are of interests in relation to the positive inotropic effect of digitalis. Thus, Marban and Tien propose that a small increase in intracellular Ca concentration due to the Na-Ca exchange mechanism may act as a positive feedback signal to increase Ca entry through slow Ca channels [8]. Additionally, a report by McGarry and Williams suggests that digoxin may also contribute to positive inotropy via activation of sarcoplasmic reticulum Ca release channels, a Na,K-ATPase independent mechanism [9].

It is important that the amount of free cytoplasmic Ca is closely regulated not only to ensure force generation, but also to protect against cellular degradation. Thus, cellular Ca accumulation may activate Ca dependent proteases, which in turn may initiate protein degradation leading to dystrophic calcification and cell death [10]. Furthermore it has recently been reported that Ca overloading may play an important role in the mechanism of myocardial stunning [11]. Thus, the Na,K-pump has the capacity for influencing cellular proteins as well as their functions. Moreover myocardial Na,K-ATPase is of importance for regulation of interstitial K concentration among myocytes, and skeletal muscular Na,K-ATPase is of importance for plasma K regulation. During excitation K leaks out of the cytosol increasing extracellular K. Subsequently K is transported back into the cell by the Na,K-pump. High extracellular K rises have been associated with reduced muscular force in the myocardium [12]as well as in skeletal muscles [13]. Furthermore major changes in K concentration in the extracellular phase of the myocardium as well as in plasma K are well recognized to be of major importance for generation of lethal arrhythmias [14].

Na,K-ATPase may undergo regulation of its pump activity as well as its concentration mediated on a short and a long term basis, respectively [15, 16]. Thus, apart from regulation by ions and excitation, major acute stimulators of muscular Na,K-ATPase activity are catecholamines [17, 18]and insulin [19]. From a clinical point of view these mechanisms are of importance in for instance catecholamine induced hypokalemia and treatment of hyperkalemia with insulin [20]. With respect to the long term regulation, decreases in human skeletal muscular Na,K-ATPase concentration in the range of 18–50% has been observed in K/Mg depletion [21], heart failure [22]and hypothyroidism [23], while increases in muscular Na,K-ATPase concentration in the range of 14–68% has been observed following short term cycle endurance training [24], intensified running schedule [25], cycle sprint training [26], physically active life style [27], insulin resistance (NIDDM) [28], insulin treatment (treated IDDM) [28]and hyperthyroidism [23]. The most pronounced regulatory changes in muscular Na,K-ATPase concentration were observed following altered thyroid function. These observations are of clinical importance in for instance hypokalemia and/or K depletion which often are induced by diuretic treatment in patients with cardiovascular diseases [29], for the hypokalemic attacks that may occur in thyrotoxicosis [30], and for exercise during which excessive K changes have been associated with sudden unexpected death [31]. The changes in human myocardial and skeletal muscular Na,K-ATPase concentration with heart disease will be reviewed below.

	2 Na,K-ATPase quantification

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
The major problem regarding Na,K-ATPase quantification in myocardial and skeletal musculature is the presence of large quantities of nonspecific ATPases in these tissues contrary to for instance renal cortex. This necessitates purification procedures prior to traditional Na,K-ATPase activity measurements, and accordingly membrane material is invariably lost. Thus, it has been calculated that the recovery of Na,K-ATPase in measurements of Na,K-ATPase in skeletal muscle is only 2–18% [32], and in myocardial tissue a similar problem presents itself [33–35]. As is evident complete recovery may be regarded as a prerequisite for evaluations regarding quantitative aspects of Na,K-ATPase, and furthermore incomplete recovery of enzyme additionally means that it cannot be ensured that a recovered fraction of Na,K-ATPase is representative of the in life situation. Thus, it is not feasible to compare Na,K-ATPase activity measurements on purified preparations of enzyme between experimental and control musculature, because it is impossible to ensure comparability of the recovered material.

In contrast a method for quantifying Na,K-ATPase has been available since 1974 using -ouabain binding to skeletal muscle fibres in rats [36]. In 1979 it became possible to measure -ouabain binding sites in small nonpurified samples of skeletal muscle utilizing the properties of vanadate to stabilize the Na,K-ATPase in a conformation in which digitalis glycosides can bind irrespectively of the presence of ATP [37–39]. The errors of the method of vanadate facilitated -ouabain binding to intact samples have been identified, i.e. nonspecific uptake and retention of the isotope, incomplete saturation, loss of specifically bound -ouabain during washout and impurity of the isotope. Following correction of these errors the total Na,K-ATPase concentration has been determined in both myocardial and skeletal muscle, not only in experimental animals [32, 40], but also in humans [23, 41]. These measurements could furthermore be performed on samples harvested postmortem [42, 43], because -ouabain binding sites proved to degrade only slowly postmortem, i.e between 0.5–1% per h, respectively [44, 45]. A problem with studies of -ouabain binding to tissue samples from heart failure patients, however, is that such patients often have received cardiac glycoside therapy. Thus, quantification using standard -ouabain binding would lead to an underestimation of the Na,K-ATPase concentration. However, this problem has been overcome by development of a method in which 95–97% clearance of digoxin from the Na,K-ATPase in myocardial as well as skeletal muscular tissue can be performed by washing samples in buffer containing excess digoxin antibody fragments (F_ab) before -ouabain binding [46].

-ouabain binding to intact tissue samples is generally expressed relative to tissue wet weight (pmol/g wet wt.), because it has proved to be a relevant parameter and simple to determine [37]. However water content of samples could at times be expected to differ among groups, e.g. in case of edema, which thus would affect -ouabain binding measurements resulting in a reduced binding site concentration. Such a problem may be evaluated by determining the water content of representative samples, thus allowing the use of dry weight as reference for -ouabain binding results. In general simple variations in tissue water content have not been the cause of observed differences in -ouabain binding capacity determined per g wet weight [28, 47–50]. It is also feasible that tissue exposed to vascular insults and subsequent healing with fibrous scar tissue formation, thus would not exhibit -ouabain binding to the same extent as control tissue. Such problems may be evaluated histologically or by determination of hydroxyproline content. In experimental heart failure caused by rapid ventricular pacing in dogs, it was established that decreased left ventricular -ouabain binding site concentration (16%, p<0.05) was selective, because when -ouabain binding capacity was related to protein, adrenoreceptor or dihydropyridine receptor concentrations, a significant decrease of 16% and 19% compared to control values, respectively were observed [51]. In human dilated cardiomyopathy no correlation between volume fraction of collagen tissue and Na,K-pump concentration (n=24, p>0.05, r=–0.08) was found [52]. In heart failure predominantly as a result of ischemic heart disease in humans, no major difference in fibrous tissue was observed as evaluated by light microscopy in left ventricular samples from patients with ischemic heart disease and the age matched controls in spite of a tendency to a reduction in ventricular Na,K-ATPase concentration of 14% [47]. As determined by computer imaging and hydroxyproline determinations it has been established that there was no increase in left ventricular connective tissue content in pacing induced heart failure in dogs [51]. Conclusively, wet wt. of samples seems to be an adequate reference for -ouabain binding when sources of error are adequately evaluated and taken into account.

In crude homogenates, i.e. homogenates of tissue unexposed to purification procedures, it is possible to quantify Na,K-ATPase using K-dependent 3-O-methylfluorescein phosphatase (3-OMFPase) [53], or K-dependent paranitrophenyl phosphatase activity (pNPPase) [54]determinations. These methods have given results which are in very good accord with vanadate facilitated -ouabain binding to intact samples [55]. As compared to -ouabain binding an advantage of K-dependent 3-O-MFPase as well as pNPPase measurements in crude homogenates is that both methods allow Na,K-ATPase quantification in tissues where Na,K-ATPase has a low affinity for binding cardiac glycosides, e.g. rat myocardium. A drawback with regard to these methods, however, is that they both require homogenization of rather large tissue quantities (>25 mg wet wt.), which in turn renders measurements on human tissues difficult. In measurements of K activated 3-O-MFPase activity in crude homogenates, detergent treatment of homogenates is used in the form of sodium deoxycholate (DOC) to unmask latent activity [56]. Also in measurements of K activated pNPPase activity in crude homogenates, DOC treatment of the homogenate is generally used to unmask latent activity. Recently it was found, however that DOC only unmasked latent K activated pNPPase activity insignificantly [54].

In addition a correlation between skeletal muscular Na,K-ATPase concentration as quantified by -ouabain binding and the maximum ouabain suppressible uptake has been reported (r=0.95, p<0.001, n=7) [57], bearing in mind that uptake is a reliable method for measurements Na,K-ATPase mediated K transport as determined both in cardiomyocytes [58]and in Langendorff perfused hearts [59, 60]. Unfortunately, neither maximum ouabain suppressible nor uptake is simply applicable for measurements on human tissue, because of the need for undisrupted, vital tissue fibres, which additionally must be perfused or superfused to provide adequate oxygenation.

While various measurements of Na,K-ATPase using northern and western blotting might be of interest regarding aspects of molecular biology or isoform abundance, such measurements are not quantitative. Thus, they are often performed on purified membranes; furthermore a constant relationship between mRNA and protein level cannot be ensured under different conditions using northern blotting; and finally there may be uncertainties in affinity of antibodies used in western blotting techniques.

Since also muscular Ca-ATPase quantification may be affected with methodological problems, it is of interest that it recently has been established that Ca-ATPase can be quantified in crude homogenates using Ca activated pNPPase measurements [50, 54, 61]. In Ca activated pNPPase activity measurements on crude homogenates Triton X-100 is used as detergent to reveal latent Ca dependent activity in membrane vesicles [50, 54]. In the rat myocardium Ca-ATPase concentration has been found to be around 5 times greater than the Na,K-ATPase concentration [54]. This may agree with the observation that Ca-ATPase uses around 30%, whereas Na,K-ATPase uses only around 4% of the total basal myocardial energy expenditure [4].

	3 Normal human myocardial Na,K-ATPase concentration

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
Published values of -ouabain binding site concentration in normal human myocardium are ranked in Fig. 1, upper panel. Values obtained using material obtained in life are illustrated by open circles while values obtained using material harvested postmortem are illustrated by filled circles. -ouabain binding capacity fell into two groups as evaluated in human, vital control myocardium (Fig. 1, upper panel, open circles) i.e 728 pmolxg^–1 wet wt. [41]which is greater (p<0.05) than a group of values which do not differ statistically from each other (p>0.70) ranging from 485–559 pmolxg^–1 wet wt. [22, 41, 52, 62–64]. There are a number of plausible reasons why the value observed by Schmidt et al., 1993 [41]is greater than the other vital control values reported for myocardial -ouabain binding capacity. Thus, it is the only -ouabain binding site concentration observed in normal left ventricular tissue harvested from normal hearts intended for heart transplantation. Apart from the value of 507 pmolxg^–1 wet wt. [64], the reports ranging from 485–559 pmolxg^–1 wet wt. are not the outcome of measurements on normal left ventricular tissue. Thus, the lower -ouabain binding site concentrations may be explained by the fact that all these control subjects were designated as such only because they displayed normal left ventricular ejection fraction (EF). In fact these individuals had undergone left sided cardiac catheterization including cardiac biopsy, because they either displayed enlarged cardiac silhouette on chest roentgenograms, arrhythmias, or both — and indeed as such were not normal. Furthermore a number of these individuals were digitalized [22], which was not accounted for in -ouabain binding measurements. It is furthermore of interest that the ranges in -ouabain binding concentrations were rather large, i.e. 70–682 pmolxg^–1 wet wt. [62]and 89–637 pmolxg^–1 wet wt. [22]. The value of 507 pmolxg^–1 wet wt. [64]differing from 728 pmolxg^–1 wet wt. can be the outcome of methodological differences, since binding was carried out on homogenates, albeit determined as crude, as well as the fact that Shamraj et al. sampled the intraventricular septum as well as the right and left ventricle of patients, who were presumed to be normal but did not have EF determined. Although animal experiments have not shown significant differences in -ouabain binding site concentration between these areas of the heart [40], the binding results cannot unambiguously represent left ventricular -ouabain binding. In conclusion 728±58 (n=5) pmolxg^–1 wet wt. [41]may be deemed the most accurate value for normal, vital left ventricular -ouabain binding site concentration in humans.

View larger version (20K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 1 Published myocardial -ouabain binding capacities observed in persons without (control) and with heart disease. The symbols represent mean observations with bars denoting SEM where these exceed the size of the symbols, and are ranked in a successive decreasing order of magnitude. Besides impression by appearance one way analysis of variance was performed on -ouabain binding site concentrations (see text) to assess overall significance and followed by a Tukey post test procedure to locate possible differences. When only 2 values were compared statistical significance was ascertained by Student's two tailed t-test for unpaired observations. A probability level of p<0.05 was considered statistically significant. In perpendicular order given from above references are: (upper panel, open circles) (41), (52), (64), (63), (22), (62); (upper panel, filled circles) (65), (45), (46), (47); (lower panel, open circles) (41), (22), (52), (63), (62), (65), (62); (lower panel, filled circles) (65), (47).

 
As may be observed in Fig. 1, upper panel, filled circles, postmortem evaluations of human left ventricular -ouabain binding site concentration in the normal heart seem to fall into 3 separate groups, which by analysis of variance differ significantly, i.e. 598 [65], 413 [45]and 283–300 pmolxg^–1 wet wt. [46, 47]. The postmortem myocardial -ouabain binding capacity of 598 pmolxg^–1 wet wt. [65]seems relatively high compared to the value of 728 pmolxg^–1 wet wt. which was observed using fresh samples [41]. Even though the postmortem tissue was harvested 18 h after death and considering that postmortem receptor degradation must have occurred, but which does not appear to have been taken into consideration. Additionally one of the included subjects had been treated with digoxin. That postmortem left ventricular -ouabain binding capacity amounting to 300 pmol/g wet wt. [47]is smaller than the value amounting to 413 pmolxg^–1 wet wt. observed by Nørgaard et al., 1986 [45]may be explained in part by the fact that the tissue in the first case had undergone longer postmortem degradation. Additionally it might be added that differences in storage of cadavers may have influenced receptor degeneration in the studies carried out on tissue harvested after death, and thus may have affected the results. Finally it may be added that the reported range in -ouabain binding capacity was rather large, i.e. 223–682 pmolxg^–1 wet wt. Thus it may be concluded that 300–600 pmolxg^–1 wet wt. would appear to be the level of myocardial -ouabain binding capacity in tissue harvested postmortem from patients with no known cardiac disease.

	4 Human myocardial Na,K-ATPase concentration changes in heart disease

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
As may be observed in Fig. 1, lower panel, open circles, myocardial -ouabain binding capacities as evaluated in vital tissue from cardiac patients appear to be divided into a main group ranging from 285–340 pmolxg^–1 wet wt. [22, 52, 62–64]and a value of 467 pmolxg^–1 wet wt. [41]. In case analysis of variance is applied, in which high values are included in the analysis before low values 2 groups of values appear which are significantly different. Thus, the lowest value observed amounting to 285 pmolxg^–1 wet wt. [62]discriminates itself from the main group ranging from 293–467 pmolxg^–1 wet wt. [22, 41, 52, 62–64]. Aspects regarding the use of endomyocardial biopsies, including measurements performed on tissues from the intraventricular septum and the right ventricle may have been of importance as described above. Conclusively, 467±55 (n=6) pmolxg^–1 wet wt. [41]would appear to be the most valid value for -ouabain binding in vital, human myocardial tissue from patients with cardiac disease.

As may be observed in Fig. 1, lower panel, filled circles, there are two published values for myocardial -ouabain binding capacities as evaluated in samples obtained postmortem from cardiac patients, i.e. 449 pmolxg^–1 wet wt. [65]and 179 pmolxg^–1 wet wt. [47]. As previously described under myocardial -ouabain binding capacities observed in control subjects, Ellingsen and coworkers [65]appear to have published a relatively high value. Thus, 449 pmolxg^–1 wet wt. is in the same order of magnitude as the highest -ouabain binding value observed in vital biopsies from cardiac patients (446 pmolxg^–1 wet wt.) even though only 50% had heart failure. The value of 174±10 (n=11) pmolxg^–1 wet wt. [47]seems to be a reasonable value for the functional -ouabain binding taking into account that all patients were digitalized. Thus, following wash of samples in digoxin antibody -ouabain binding increased to a value of 265 pmol/g wet wt. [47]. In general both in experimental models as well as in human beings heart failure has been found to be associated with around 30–40% decreased left ventricular Na,K-ATPase concentration [55]. Cardiomyopathy has been reported to be associated with the most pronounced reduction in left ventricular Na,K-ATPase: Human heart failure patients with ischemic heart disease and dilated cardiomyopathy showed significant reductions of 34% [62]and 41% [52], respectively in left ventricular Na,K-ATPase concentration, which seems to be supported by more recent studies of human ischemic heart disease showing reductions of 14% [47]and 19% [41]. Thus, a 33% lower K dependent 3-O-methylfluorescein phosphatase (3-O-MFPase) activity was observed in crude homogenates from the ventricle of 17 month old cardiomyopathic hamsters [56]. Furthermore it has recently been established that heart failure in the hereditary cardiomyopathic hamster model is associated with a significant reduction in left ventricular Ca-ATPase of 26% which was selective to overall protein content, and thus not the mere outcome of increased myocardial mass [50].

	5 Normal human skeletal muscle Na,K-ATPase concentration

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
In Fig. 2, upper panel, a rundown is given over published results of human skeletal muscle -ouabain binding capacities from normal control subjects [55]. As may be observed from the upper panel, open circles skeletal muscular -ouabain binding capacities as evaluated in samples obtained in fresh human skeletal muscle from normal individuals appear to span homogenously from 223–360 pmolxg^–1 wet wt., since there are no obvious outliers. All but 2 of the 12 reports are the results of -ouabain binding measurements performed on vastus lateralis samples, in which samples of the quadriceps femoris (not additionally specified), rectus abdominis and gluteus maximus were used in one case [66], and nonspecified musculature harvested during hip surgery was used in the second case [44]. By analysis of variance -ouabain binding capacities are divided into 2 groups (p<0.05), i.e. 223–278 pmolxg^–1 wet wt. [21, 27, 28, 44, 67]and 281–360 pmolxg^–1 wet wt. [22, 24–26, 66, 68, 69]. It should be noted that the two reports carried out on muscular samples other than those harvested from vastus lateralis were divided into each group of -ouabain binding results differing significantly as determined by analysis of variance, i.e. 223–278 pmolxg^–1 wet wt. included the report by Nørgaard et al. [44], and 281–360 pmolxg^–1 wet wt. included the report by Benders et al. [66]. Thus, it is unlikely that the use of the muscle samples in question should have influenced the overall data analysis to any major extent. The main reason for the predominant use of vastus lateralis as a source of skeletal muscle is that biopsies are rather easily obtainable from the vastus lateralis, consistently taken 10 cm above the knee following application of local anesthesia, using the biopsy technique first described by Bergström [70]. Secondly, it could be perceived as an advantage to sample muscle with a mixed fibre composition so as to reflect skeletal muscle in general. With regard to fibre type composition of the vastus lateralis, it has been reported that differences in fibre type proportions within fascicles are caused by local factors in the muscle, secondary to overall age related functional demands put on the fibre population [71]. Thus, it cannot be ruled out that variations in -ouabain binding within the controls and the heart failure patients (see below) may be influenced by fibre type predominance. Notwithstanding, the group of values spanning from 281–360 pmolxg^–1 wet wt. primarily differs from the group of values spanning from 223–278 pmolxg^–1 wet wt. by being the result of studies evaluating effects of physical activity [24–26, 69], while the group of values spanning from 223–278 pmolxg^–1 wet wt. only contains a single study on effects of physical activity [27]. It seems reasonable to surmise that people who participate in studies regarding effects of physical activity might be interested in aspects of training, and thus such controls could be presumed to be more fit than people who do not enjoy physical exertion. This in part could explain the larger capacity for skeletal muscular -ouabain binding exhibited by this group. In conclusion 223–278 pmolxg^–1 wet wt. is the normal level for skeletal muscular -ouabain binding in human normal, vital skeletal muscle.

View larger version (16K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 2 Published skeletal muscular -ouabain binding capacities observed in persons without and with heart failure. Out of a total of the 24 reports, all but 2 reports are the results of measurements performed on vastus lateralis samples. The two reports in question made use of quadriceps femoris (not specified), rectus abdominis and gluteus maximus samples (66), and nonspecified musculature harvested during hip surgery (44). Both of the papers in which vastus lateralis was not the source of the employed muscular samples were reports on -ouabain binding capacity in normal, vital human muscle. All current values are ranked and compared as are the -ouabain binding capacities obtained in human myocardium (Fig. 1.). In perpendicular order given from above references are: (upper panel, open circles) (66), (24), (26), (69), (22), (25), (23), (44), (27), (67), (21), (28); (upper panel, filled circles) (72), (72), (46), (45), (49), (72); (lower panel, open circles) (73), (22), (67), (21), (67); (lower panel, filled circles) (49).

 
As may be observed in Fig. 2, upper panel, closed circles, skeletal muscular -ouabain binding capacities as evaluated in vastus lateralis samples obtained postmortem from humans without heart disease appear to span homogenously from 216–305 pmolxg^–1 wet wt., since there are no obvious outliers [45, 46, 49, 72]. This was confirmed by analysis of variance which showed no significant difference of values. In conclusion 216–305 pmolxg^–1 wet wt. seems to be the level for skeletal muscular -ouabain binding in normal, human skeletal muscle obtained postmortem. In general it may appear that the use of tissue harvested postmortem is associated with more dispersion of -ouabain binding results as compared to the results obtained using vital tissue, and more importantly also here a problem regarding recovery of Na,K-ATPase may present itself.

	6 Human skeletal muscle Na,K-ATPase concentration changes in heart disease

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
As may be observed in Fig. 2, lower panel, open circles, skeletal muscular -ouabain binding capacities as evaluated in fresh samples from the vastus lateris from heart patients appear to span homogenously from 195–239 pmolxg^–1 wet wt. [21, 22, 73]since there are no obvious outliers. By analysis of variance no significant difference was found among the values. In conclusion 195–239 pmolxg^–1 wet wt. is the level for skeletal muscular -ouabain binding in fresh human skeletal muscle from patients with cardiac disease.

Concerning human skeletal muscular -ouabain binding site concentration as observed in postmortem tissue harvested from the vastus lateris from heart failure patients (Fig. 2, lower panel, filled circles), there appears to be only one reported value of 150 pmolxg^–1 wet wt. [49]. This value seems to be a reasonable value for the functional -ouabain binding when taking into account that all the patients were digitalized. Thus, following wash of samples in digoxin antibody -ouabain binding increased to a value of 173±13 (n=11) pmol/g wet wt. [49]. In general heart failure seems to be associated with a significant decrease in skeletal muscular -ouabain binding site concentration, indicating that cardiac failure is not only a disease afflicting the heart, but also a disease afflicting skeletal muscle.

	7 Digitalization and muscular Na,K-ATPase

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
Specific left ventricular receptor (Na,K-ATPase) occupancy with digoxin as determined postmortem, i.e. 15 h after death, amounted to 34% (p<0.001) [47], and as determined in life without putative confounding effects of postmortem receptor degradation it amounted to 24% (p<0.05) [41]. The specific left ventricular receptor occupancies with digoxin determined after death and in life are in good accord with the report of a 38% reduction in maximum hyperpolarization of resting membrane potential in human atrial tissue as a result of digitalization [74]. Specific skeletal muscular receptor occupancy with digoxin as determined postmortem amounted to 13% (p<0.02) [49]. Using vital muscular samples from digitalized patients, muscular receptor occupancy with digoxin as determined in life without putative confounding effects of postmortem receptor degradation amounted to 9% (p<0.05) [73]. Thus occupancy of receptors with digoxin in the heart was 2.6–2.7 fold higher than in skeletal muscle. This seems to concur with the higher affinity of human left ventricular Na,K-ATPase for vanadate facilitated -ouabain binding as determined by Scatchard type plot analysis compared to corresponding measurements performed on human vastus lateralis samples [44, 45]. These skeletal muscular digitalis glycoside receptor occupancies seem to accord with the report of 9% lower standard -ouabain binding to muscular samples from digitalized as compared to nondigitalized patients [67]. However, a problem with the evaluation by Dørup and coworkers might be that the occupancy was obtained comparing standard -ouabain binding in a group of digitalized patients to that of another group of non digitalized subjects. If the digoxin receptors had been downregulated due to heart failure in the digitalized subjects, this may have caused an overestimation of digoxin occupancy. Additionally, it would not have been possible to reveal a putative concealed upregulation of skeletal muscular cardiac glycoside receptor concentration in response to digoxin exposure, which might have resulted in an underestimation of receptor occupancy with digoxin. Studies in man of skeletal muscular content of digoxin as a result of digitalization have hitherto been carried out by way of radioimmunoassay. These studies reported digoxin concentrations in skeletal muscular samples obtained postmortem of 20–32 pmolxg^–1 wet wt. [75–77]. A receptor occupancy by digoxin of 9% in human skeletal muscle corresponds to around 24 pmolxg^–1 wet wt. of receptors being occupied by digoxin in response to digitalization. Compared with the radioimmunoassay studies this would seem to imply that a major part of the digoxin located in skeletal muscle is specifically bound to its receptor. In clinical medicine there are various indications of the importance of skeletal muscular Na,K-ATPase for digoxin distribution. Thus, it has been reported that in humans plasma digoxin decreases 9–26% during exercise and that the concentration of digoxin in skeletal muscle increases depending on exercise level [78, 79].

Cardiac glycoside exposure of animals without heart failure has been reported to increase myocardial Na,K-ATPase activity around 13–47% [80–82], and likewise peripheral blood cells from patients reportedly showed 35–64% increases in -ouabain binding [83–86]. As a consequence of cardiac glycoside exposure studies on aspects of Na,K-ATPase have reported receptor upregulation also in various in vitro systems. There are, however, questions which could be raised regarding peripheral blood cells as well as in vitro systems such as cultured HeLa cells [87–90], Girardi cells [87, 88], chick heart cells [91, 92], rat skeletal muscular myotubes [93]and kidney tubules [94, 95]showing an upregulation of Na,K-ATPase following exposure to cardiac glycosides. Primarily, it is of concern that peripheral blood cells and cultured cells may have regulatory mechanisms different from those of intact tissues. Thus, for instance in response to K depletion peripheral blood cells and cultured cells have been shown to regulate Na,K-ATPase in a manner which is different from that of intact muscular tissue: Hence peripheral blood cells have been shown to upregulate Na,K-ATPase in response to hypokalemia [85, 96, 97], and growth of HeLa cells [87, 88, 90]and chick heart cells [98]in medium containing a low concentration of K was also found to cause an upregulation of Na,K-ATPase. Conversely, measurements of -ouabain binding to intact muscular samples have in animals [99, 100]as well as in human subjects [21]shown downregulation in response to K depletion. It is also of concern that in several of the studies performed on cultured cells, concentrations of ouabain have been used which would be toxic in the human organism, i.e. 1x10^–6 [89], 2x10^–6 [91], 0.3–3x10^–6 [92], 1x10^–5 [101]and 1x10^–4 mol/l [94, 95]. However, by comparing measurements of -ouabain binding following prolonged wash in excess digoxin F_ab in samples from digitalized patients with heart failure and from non digitalized individuals without heart failure, putative regulation of receptor concentration as a result of in life digitalis therapy was evaluated. Thus, there appeared to be a trend to a 14% (p>0.10) lower -ouabain binding site concentration in left ventricular samples obtained after death from digitalized as compared with non digitalized patients [47]. Using human, vital left ventricular samples harvested at heart transplantation, a trend was observed to a 19% (p<0.08) lower -ouabain binding site concentration in left ventricular samples from digitalized as compared with non digitalized subjects [49]. The observations were unaffected by occupancy as well as by duration of treatment [55]. Neither did evaluations of skeletal muscular digitalis glycoside receptor upregulation in response to digoxin treatment give any indications of Na,K-ATPase upregulation in response to cardiac glycoside treatment. Thus, there was evidence of 37% (p<0.005) lower digitalis glycoside receptor concentration in the vastus lateralis of digitalized patients with heart failure compared with non digitalized subjects without heart failure [49]. From animal studies it is known that muscles composed predominantly of type 2 fibres exhibit higher -ouabain binding capacity than type 1 fibres. Hence extensor digitorum longus (predominantly type 2 fibres) exhibits 29% (p<0.003, n=6) higher -ouabain binding capacity than soleus (predominantly type 1 fibres) in rats [102]. The lower -ouabain binding capacity observed in vastus lateralis samples from heart failure patients compared to controls without heart failure is not likely the outcome of differences in vastus lateralis fibre type composition, because heart failure has been reported to be associated with a significantly reduced proportion of type 1 fibres and an increased proportion of type 2 fibres [103]. It is conceivable that upregulation of left ventricular and skeletal muscular cardiac glycoside receptor concentration was not seen after exposure to digoxin, because the patients suffered from heart failure. Hence, on the basis of the current evaluations it might not be ruled out that a normal subject would respond to digoxin exposure by receptor upregulation. Such an observation, although it might be of potential physiological interest, could however not be ascribed any therapeutic importance for digoxin treatment of heart failure patients.

It may be noted that a reduced capacity for active K uptake by skeletal Na,K-ATPase was observed as a result of digitalization [73]. Thus, following clinical digitalization of 10 patients with moderate heart failure the femoral veno arterial (v-a) difference in plasma K increased 50–100% (p<0.05) during exercise, and decreased 66–75% during early recovery. Taken together evidence is increasing that skeletal muscular Na,K-ATPase concentration, i.e. active skeletal muscular cation transport capacity, is of importance for extrarenal K homeostasis in man.

	8 Conclusions and clinical perspectives

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 
Taken together it has become evident that myocardial Na,K-ATPase concentration is reduced in heart failure. Furthermore also skeletal muscular Na,K-ATPase concentration is reduced in heart diseases. This supports the evolving concept that heart failure is not only a disease affecting the myocardium, but is a generalized disease also affecting skeletal muscle [104, 105]. Digitalization induces a functional reduction in muscular Na,K-ATPase. As is illustrated in Fig. 3 this may increase intracellular Ca concentration which initially may induce inotropy in the myocardium. Furthermore, the initial reduction of myocardial Na,K-ATPase concentration seen in heart failure patients could be perceived as an adaptive digitalization and the administration of cardiac glycosides to heart failure patients may add to this phenomenon. The decreased Na,K-ATPase concentration in heart failure may also affect K homeostasis: In the myocardium extracellular K concentration may rise to higher values, and plasma K changes during physical activity may be more pronounced rendering increased susceptibility to arrhythmias. At later stages when heart failure has progressed from compensated to a decompensated condition, the exacerbation will result in more pronounced increases in intracellular Ca: High intracellular Ca levels may activate proteases inducing cellular calcification and degeneration, and a vicious circle promoting further Na,K-ATPase degeneration will thus, have been initiated. Force of contraction would decrease further while susceptability to arrhythmia would increase. It might be noted that indications from experimental animal studies indicate that also other important transport proteins such as the myocardial Ca-ATPase is downregulated in heart disease [50]. Thus, caution is due when extrapolating results of studies of only a single transport protein to a clinical setting. Future studies need to engage into the interaction between regulations of various transport proteins, channels and receptors. It is however, at present of interest that the clinical perspectives and implications of the results outlined above on the effect of digoxin treatment on cardiac glycoside receptors concur with recent clinical studies of digitalization of heart failure patients [106–108], which report a beneficial effect of digitalization on symptoms and need for hospitalization of heart failure patients. Additionally digitalis treatment has been shown not to affect overall mortality [108]. Interestingly, and also in agreement with the clinical implications made on the basis of the cardiac glycoside receptor studies, a tendency to a lower risk of death of terminal heart failure appeared to be cancelled by a tendency to an increased risk of lethal arrhythmias [108]. Conversely, it might be ventured that future clinical trials evaluating digoxin treatment of heart failure could gain considerable impact by taking into account the current basic knowledge concerning Na,K-ATPase regulation. This would allow approaching the role of digoxin therapy in heart failure on a more rational basis than merely defining simple clinical endpoints.

View larger version (16K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 3 Relationships between muscular Na,K-ATPase concentration ([Na,K-ATPase]), digitalization, intracellular Ca concentration ([Cai2+]), extracellular K concentration ([Ke+]) changes, force of contraction, arrhythmia and cellular degeneration in early compensated and late decompensated stages of heart failure.

 
Time for primary review 26 days.

	Acknowledgements

 
The present study has been supported by the Danish Heart Foundation.

	References

Top 1 The Na,K-ATPase 2 Na,K-ATPase quantification 3 Normal human myocardial... 4 Human myocardial Na,K-ATPase... 5 Normal human skeletal... 6 Human skeletal muscle... 7 Digitalization and muscular... 8 Conclusions and clinical... References

 

1. Skou J.C. Enzymatic basis for active transport of Na and K across cell membrane. Physiol Rev (1965) 45:596–617.[Free Full Text]
2. Carafoli E. The calcium pumping ATPase of the plasma membrane. Annu Rev Physiol (1991) 53:531–547.[CrossRef][ISI][Medline]
3. Reiter M. Calcium mobilization and cardiac inotropic mechanisms. Pharmacol Rev (1988) 40:189–217.[ISI][Medline]
4. Clausen T., Van Hardeveld C., Everts M.E. Significance of cation transport in control of energy metabolism and thermogenesis. Physiol Rev (1991) 71:733–774.[Free Full Text]
5. Matsui H., Schwartz A. Mechanism of cardiac glycoside inhibition of the Na,K-dependent ATPase from cardiac tissue. Biochim Biophys Acta (1968) 151:655–663.[Medline]
6. Smith T.W. Digitalis. Mechanisms of action and clinical use. N Engl J Med (1988) 318:358–365.[Medline]
7. Rose A.M., Valdes R. Jr. Understanding the sodium pump and its relevance to disease. Clin Chem (1994) 40:1674–1685.[Abstract/Free Full Text]
8. Marban E., Tsien R.W. Enhancement of calcium current during digitalis inotropy in mammalian heart: Positive feedback regulation by intracellular calcium? J Physiol Lond (1982) 329:589–614.[Abstract/Free Full Text]
9. McGarry S.J., Williams A.J. Digoxin activates sarcoplasmic reticulum Ca release channels: A possible role in cardiac inotropy. Br J Pharmacol (1993) 108:1045–1050.
10. Weishaar R.E., et al. Protection of the failing heart. J Appl Cardiol (1987) 2:339–360.[ISI]
11. Gao W.D., Liu Y., Marban E. Selective effects on oxygen free radicals on excitation-contraction coupling in ventricular muscle. Implications for the mechanism of stunned myocardium. Circulation (1996) 94:2597–2604.[Abstract/Free Full Text]
12. Ryan D.M., Paterson D.J. Restoration of cardiac contraction by angiotensin II during raised K in the rabbit. Acta Physiol Scand (1996) 156:419–427.[CrossRef][ISI][Medline]
13. Lindinger M.I., Sjögaard G. Potassium regulation during exercise and recovery. Sports Med (1991) 11:382–401.[ISI][Medline]
14. Gettes L.S. Electrolyte abnormalities underlying lethal and ventricular arrhythmias. Circulation (1992) 85(Suppl. 1):I70–I77.[Medline]
15. Clausen T. Regulation of active Na,K transport in skeletal muscle. Physiol Rev (1986) 66:542–580.[Free Full Text]
16. Clausen T. Long- and short-term regulation on the Na,K-pump in skeletal muscle. News Physiol Sci (1996) 11:24–30.[Abstract/Free Full Text]
17. Clausen T., Flatman J.A. The effect of catecholamines on Na-K transport and membrane potential in rat soleus muscle. J Physiol Lond (1977) 270:383–414.[Abstract/Free Full Text]
18. Ellingsen O., Sejersted O.M., Vengen O.A., Ilebekk A. Frequency dependent myocardial potassium fluxes during beta adrenergic stimulation of intact pig hearts. Cardiovasc Res (1991) 25:364–370.[ISI][Medline]
19. Flatman J.A., Clausen T. Combined effects of adrenaline and insulin on active electrogenic Na,K-transport in rat soleus muscle. Nature (1979) 281:580–581.[CrossRef][Medline]
20. DeFronzo R.A., Ferrannini E. Insulin resistance. A multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care (1991) 14:173–194.[Abstract]
21. Dørup I., Skajaa K., Clausen T., Kjeldsen K. Reduced concentrations of potassium, magnesium, and sodium-potassium pumps in human skeletal muscle during treatment with diuretics. Br Med J (1988) 296:455–458.[ISI][Medline]
22. Nørgaard A., Bjerregaard P., Baandrup U., Kjeldsen K., Reske-Nielsen E., Thomsen P.E. The concentration of the Na,K-pump in skeletal and heart muscle in congestive heart failure. Int J Cardiol (1990) 26:185–190.[CrossRef][ISI][Medline]
23. Kjeldsen K., Nørgaard A., Gøtzsche C.O., Thomassen A., Clausen T. Effect of thyroid function on number of Na-K pumps in human skeletal muscle. Lancet (1984) 2:8–10.[CrossRef][ISI][Medline]
24. Green H.J., Chin E.R., Ball Burnett M., Ranney D. Increases in human skeletal muscle Na,K-ATPase concentration with short-term training. Am J Physiol (1993) 264:C1538–C1541.[ISI][Medline]
25. Madsen K., Franch J., Clausen T. Effects of intensified endurance training on the concentration of Na,K-ATPase and Ca-ATPase in human skeletal muscle. Acta Physiol Scand (1994) 150:251–258.[ISI][Medline]
26. McKenna M.J., Schmidt T.A., Hargreaves M., Cameron L., Skinner S.L., Kjeldsen K. Sprint training increases human skeletal muscle Na,K-ATPase concentration and improves K regulation. J Appl Physiol (1993) 75:173–180.[Abstract/Free Full Text]
27. Klitgaard H., Clausen T. Increased total concentration of Na-K pumps in vastus lateralis muscle of old trained human subjects. J Appl Physiol (1989) 67:2491–2494.[Abstract/Free Full Text]
28. Schmidt T.A., Hasselbalch S., Farrell P.A., Vestergaard H., Kjeldsen K. Human and rodent muscle Na,K-ATPase in diabetes related to insulin, starvation and training. J Appl Physiol (1994) 76:2140–2146.[Abstract/Free Full Text]
29. Nørgaard A., Kjeldsen K. Interrelation of hypokalemia and potassium depletion and its implications: a re-evaluation based on studies of the skeletal muscle sodium, potassium-pump. Clin Sci (1991) 81:449–455.[ISI][Medline]
30. Ali K. Hypokalaemic periodic paralysis complicating thyrotoxicosis. Br Med J (1975) 4:503–504.[Medline]
31. Lindinger M.I. Potassium regulation during exercise and recovery in humans: Implications for skeletal and cardiac muscle. J Mol Cell Cardiol (1995) 4:1011–1022.
32. Kjeldsen K. Complete quantification of the total concentration of rat skeletal-muscle Na,K-dependent ATPase by measurements of -ouabain binding. Biochem J (1986) 240:725–730.[ISI][Medline]
33. Hansen O., Clausen T. Quantitative determination of Na,K-ATPase and other sarcolemmal components in muscle cells. Am J Physiol (1988) 254:C1–C7.[ISI][Medline]
34. Jones L.R., Besch H.R. Isolation of canine cardiac sarcolemmal vesicles. Meth Pharmacol (1984) 5:1–12.
35. Matsui H., Schwartz A. Purification and properties of a highly active ouabain-sensitive Na, K-dependent adenosinetriphosphatase from cardiac tissue. Biochim Biophys Acta (1966) 128:380–390.[Medline]
36. Clausen T., Hansen O. Ouabain binding and Na-K-transport in rat muscle cells and adipocytes. Biochim Biophys Acta (1974) 345:387–404.
37. Nørgaard A., Kjeldsen K., Hansen O., Clausen T. A simple and rapid method for the determination of the number of -ouabain binding sites in biopsies of skeletal muscle. Biochem Biophys Res Commun (1983) 111:319–325.[CrossRef][ISI][Medline]
38. Hansen O. Facilitation of ouabain binding to Na,K-ATPase by vanadate at in vivo concentrations. Biochim Biophys Acta (1979) 568:265–269.[Medline]
39. Hansen O. Interaction of cardiac glycosides with Na, K-activated ATPase. A biochemical link to digitalis-induced inotropy. Pharmacol Rev (1984) 36:143–163.[ISI][Medline]
40. Schmidt T.A., Svendsen J.H., Haunso S., Kjeldsen K. Quantification of the total Na,K-ATPase concentration in atria and ventricles from mammalian species by measuring -ouabain binding to intact myocardial samples. Stability to short term ischemia reperfusion. Basic Res Cardiol (1990) 85:411–427.[CrossRef][ISI][Medline]
41. Schmidt T.A., Allen P.D., Colluci W.S., Marsh J.D., Kjeldsen K. No adaption to digitalization as evaluated by digitalis receptor (Na,K-ATPase) quantification in explanted hearts from donors without heart disease and from digitalized patients in endstage heart failure. Am J Cardiol (1993) 70:110–114.
42. Kjeldsen K., Grøn P. Skeletal muscle Na,K-pump concentration in children and its relationship to cardiac glycoside distribution. J Pharmacol Exp Ther (1989) 250:721–725.[Abstract/Free Full Text]
43. Kjeldsen K., Grøn P. Age-dependent change in myocardial cardiac glycoside receptor (Na,K-pump) concentration in children. J Cardiovasc Pharmacol (1990) 15:332–337.[ISI][Medline]
44. Nørgaard A., Kjeldsen K., Clausen T. A method for the determination of the total number of -ouabain binding sites in biopsies of human skeletal muscle. Scand J Clin Lab Invest (1984) 44:509–518.[ISI][Medline]
45. Nørgaard A., Kjeldsen K., Hansen O., Clausen T., Larsen C.G., Larsen F.G. Quantification of the -ouabain binding site concentration in human myocardium: a postmortem study. Cardiovasc Res (1986) 20:428–435.[Abstract/Free Full Text]
46. Schmidt T.A., Kjeldsen K. Enhanced clearance of specifically bound digoxin from human myocardial and skeletal muscle samples by specific digoxin antibody fragments. Subsequent complete digitalis glycoside receptor (Na,K-ATPase) quantification. J Cardiovasc Pharmacol (1991) 17:670–677.[ISI][Medline]
47. Schmidt T.A., Holm Nielsen P., Kjeldsen K. No upregulation of digitalis glycoside receptor (Na,K-ATPase) concentration in human heart left ventricle samples obtained at necropsy after long term digitalisation. Cardiovasc Res (1991) 25:684–691.[ISI][Medline]
48. Schmidt T.A., Larsen J.S., Kjeldsen K. Quantification of rat cerebral cortex Na,K-ATPase: Effect of age and potassium depletion. J Neurochem (1992) 59:2094–2104.[ISI][Medline]
49. Schmidt T.A., Holm Nielsen P., Kjeldsen K. Human skeletal muscle digitalis glycoside receptors (Na,K-ATPase) — importance during digitalization. Cardiovasc Drugs Ther (1993) 7:175–181.[CrossRef][ISI][Medline]
50. Larsen J.S., Schmidt T.A., Bundgaard H., Kjeldsen K. Reduced concentration of myocardial Na,K-ATPase in human aortic valve disease as well as of Na,K- and Ca-ATPase in rodents with hypertrophy. Mol Cell Biochem (1997) 169:85–93.[CrossRef][ISI][Medline]
51. Schmidt T.A., Larsen J.S., Shannon R.P., Komamura K., Vatner D.E., Kjeldsen K. Reduced -ouabain binding site (Na,K-ATPase) concentration in ventricular myocardium of dogs with tachycardia induced heart failure. Basic Res Cardiol (1993) 88:607–620.[CrossRef][ISI][Medline]
52. Nørgaard A., Bagger J.P., Bjerregaard P., Baandrup U., Kjeldsen K., Thomsen P.E. Relation of left ventricular function and Na,K-pump concentration in suspected idiopathic dilated cardiomyopathy. Am J Cardiol (1988) 61:1312–1315.[CrossRef][ISI][Medline]
53. Nørgaard A., Kjeldsen K., Hansen O. K-dependent 3-O-methylfluorescein phosphatase activity in crude homogenate of rodent heart ventricle: effect of K depletion and changes in thyroid status. Eur J Pharmacol (1985) 113:373–382.[CrossRef][ISI]