HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 293/2/C509    most recent 00098.2007v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (16) Citing Articles via Google Scholar Google Scholar Articles by Schoner, W. Articles by Scheiner-Bobis, G. Search for Related Content PubMed PubMed Citation Articles by Schoner, W. Articles by Scheiner-Bobis, G.

INVITED REVIEW

Endogenous and exogenous cardiac glycosides: their roles in hypertension, salt metabolism, and cell growth

Institute of Biochemistry and Endocrinology, Justus Liebig University Giessen, Giessen, Germany

	ABSTRACT

TOP ABSTRACT CHEMICAL NATURE OF ENDOGENOUS... MECHANISMS OF SIGNAL... EFFECT OF ENDOGENOUS AND... PHYSIOLOGY AND PATHOPHYSIOLOGY... ENDOGENOUS CARDIAC GLYCOSIDES IN... GENERAL CONCLUSIONS FUTURE PERSPECTIVES GRANTS REFERENCES

 
Cardiotonic steroids (CTS), long used to treat heart failure, are endogenously produced in mammals. Among them are the hydrophilic cardenolide ouabain and the more hydrophobic cardenolide digoxin, as well as the bufadienolides marinobufagenin and telecinobufagin. The physiological effects of endogenous ouabain on blood pressure and cardiac activity are consistent with the "Na⁺-lag" hypothesis. This hypothesis assumes that, in cardiac and arterial myocytes, a CTS-induced local increase of Na⁺ concentration due to inhibition of Na⁺/K⁺-ATPase leads to an increase of intracellular Ca²⁺ concentration ([Ca²⁺]_i) via a backward-running Na⁺/Ca²⁺ exchanger. The increase in [Ca²⁺]_i then activates muscle contraction. The Na⁺-lag hypothesis may best explain short-term and inotropic actions of CTS. Yet all data on the CTS-induced alteration of gene expression are consistent with another hypothesis, based on the Na⁺/K⁺-ATPase "signalosome," that describes the interaction of cardiac glycosides with the Na⁺ pump as machinery activating various signaling pathways via intramembrane and cytosolic protein-protein interactions. These pathways, which may be activated simultaneously or selectively, elevate [Ca²⁺]_i, activate Src and the ERK1/2 kinase pathways, and activate phosphoinositide 3-kinase and protein kinase B (Akt), NF-B, and reactive oxygen species. A recent development indicates that new pharmaceuticals with antihypertensive and anticancer activities may be found among CTS and their derivatives: the antihypertensive rostafuroxin suppresses Na⁺ resorption and the Src-epidermal growth factor receptor-ERK pathway in kidney tubule cells. It may be the parent compound of a new principle of antihypertensive therapy. Bufalin and oleandrin or the cardenolide analog UNBS-1450 block tumor cell proliferation and induce apoptosis at low concentrations in tumors with constitutive activation of NF-B.

endogenous cardiotonic steroids; ouabain; marinobufagenin; rostafuroxin; bufalin; oleandrin; sodium pump; sodium/potassium-adenosinetriphosphatase; arterial hypertension; sodium metabolism; cell proliferation; cancer therapy

Heart failure therapy with cardiac glycosides, however, is based on increasing cardiac output (40). It is mostly explained by the Na⁺-lag hypothesis (37), which postulates that inhibition of myocardial Na⁺/K⁺-ATPase activity by cardiotonic steroids (CTS) leads to a local rise of intracellular Na⁺ concentration ([Na⁺]_i), which in turn increases [Ca²⁺]_i and, thereby, results in a positive inotropic action on the cardiac muscle. This therapeutic concept seems to contradict modern concepts of heart failure therapy that are based on avoiding a destructive rise of [Ca²⁺]_i that leads to further heart failure via altered protein expression and apoptosis. Yet one may not exclude that digoxin, the prominent cardiac glycoside used in therapy, also acts indirectly on the failing heart by depressing the activated neuroendocrine system and, hence, the adrenergic and renin-angiotensin system (109). Thus both therapeutic approaches may protect against stress. This apparently contradictory role of [Ca²⁺]_i in the two therapeutic approaches merits further investigation, especially since a recent clinical post hoc reanalysis of the beneficial effects of digitalis therapy (4, 5) contradicts the report of the Digitalis Investigation Group (360) and reestablishes digoxin as an important remedy in the treatment of heart failure (42). According to this study, treatment with low digoxin concentrations significantly reduces mortality and hospitalizations in patients with ambulatory chronic systolic and diastolic heart failure (4, 5).

The impressive benefit of digitalis therapy in the last century led to the postulate that an "endogenous digitalis" might exist (312, 354). Such endogenous inhibitors of the Na⁺ pump might circulate in essential hypertension in higher concentrations in blood plasma and induce natriuresis and, thereby, decrease the fluid volume (35, 53, 126). Following this concept, Hamlyn et al. (131) were the first to demonstrate that an endogenous inhibitor of the Na⁺ pump circulates in human blood plasma and that its concentration correlates with the blood pressure of the donors. This finding was soon confirmed (262). These observations paved the way for the identification of a number of endogenous CTS as a new type of steroid hormone belonging to the group of cardenolides and bufadienolides (Fig. 1). The Na⁺ pump, which exists in all cells, acts as a hormone receptor for these substances. Consequently, endogenous CTS may control not only cardiac and kidney function, salt metabolism, and hypertension but, also, cell proliferation, cell half-life, and, more generally, cell function. Hence, this review discusses 1) the chemical nature of endogenous cardiac glycosides, 2) the mechanism of signal transduction via the Na⁺ pump, 3) the regulatory role of endogenous and exogenous cardiac glycosides in tissue proliferation and apoptosis, especially in cancer cells, 4) the physiology and pathophysiology of endogenous cardiac glycosides in the circulatory system and its role in hypertension, and 5) endogenous cardiac glycosides in diabetes mellitus.

View larger version (15K): [in this window] [in a new window]   Fig. 1. Structures of the endogenous cardiac glycosides which were identified on the search for "endogenous digitalis" from mammalian tissues and fluids.

	CHEMICAL NATURE OF ENDOGENOUS CARDIAC GLYCOSIDES

TOP ABSTRACT CHEMICAL NATURE OF ENDOGENOUS... MECHANISMS OF SIGNAL... EFFECT OF ENDOGENOUS AND... PHYSIOLOGY AND PATHOPHYSIOLOGY... ENDOGENOUS CARDIAC GLYCOSIDES IN... GENERAL CONCLUSIONS FUTURE PERSPECTIVES GRANTS REFERENCES

Endogenous Ouabain

Endogenous ouabain has been isolated from human plasma (127, 247), bovine adrenal gland (333), bovine hypothalamus (177), and the supernatant of rat pheochromocytoma (PC-12) cells (193). Endogenous ouabain has been shown by ¹H-NMR (177, 333) and liquid chromatography-electrospray ionization-mass spectrometry (193) to be identical to the plant-derived steroid. However, there is some speculation that an 11-isomer of ouabain may circulate in human blood (128). Ouabain is able to form complexes under physiological conditions, for instance, with borate. This may explain some reports of an increased sensitivity of some preparations of endogenous ouabain compared with ouabain alone (12, 177, 178).

Adrenal gland as a source of ouabain. Orally and parenterally administered ouabain is selectively taken up by adrenal glands (187), but its intestinal uptake accounts for only 3–5% of the orally administered substance (122). Hence, it is essential to exclude the possibility that ouabain isolated from mammalian tissues does not simply represent resorbed ouabain from food. Conscious dogs release ouabain from their adrenal glands (38). Consistent with the assumption that the adrenal gland is a site of synthesis and/or storage of ouabain, adrenalectomy leads to a decline of ouabain levels in plasma (127, 246). Most likely, the zona glomerulosa and zona fasciculata of the adrenal cortex store and/or synthesize endogenous ouabain (203). The adrenal cortex contains more ouabain than the medulla (214), and plasma concentrations of ouabain have not been shown to be lower in medullectomized rats than in sham-operated controls (129). Overproduction of ouabain by adrenal tumors has been reported, and excision of these tumors lowered the elevated blood pressure (194, 233).

Biosynthesis of ouabain in adrenal and hypothalamic cells. De novo synthesis of ouabain and dihydroouabain has been demonstrated in tissue culture experiments (294, 301). The amount of ouabain secreted by bovine adrenocortical cells in vitro is up to 10-fold greater than the cellular content (69, 202, 294). The biosynthesis occurs in zona fasciculata cells. Pregnenolone and progesterone are precursors of endogenous ouabain (129, 193, 294). Inhibition of the conversion of pregnenolone to progesterone by trilostane, an inhibitor of 3-hydroxysteroid dehydrogenase, inhibits ouabain synthesis (294). When [7-³H]pregnenolone is added to primary cultures of rat adrenal cells, radioactivity is found in a fraction that has digitalis-like activity but does not contain ouabain (218). It is possible that the mechanism leading to 5-hydroxylation in ouabain (but not in digoxin) and the A/B conformation of the steroid backbone eliminates the ³H atom in position 7. The sugar [¹⁴C]rhamnose, which is part of the ouabain molecule, is synthesized in mammals (230), readily enters adrenocortical cells, and increases the biosynthesis of endogenous ouabain (294).

Since ouabain is synthesized by bovine adrenocortical cells in tissue culture, one may ask how its release is controlled hormonally. ACTH, ₁-adrenergic receptor agonists, and angiotensin II stimulate ouabain's release from bovine adrenocortical cells (203–205). Yet the release of ouabain from human CLR7050 cells (an adrenal cortex-derived cell line) is insensitive to ACTH and angiotensin II but sensitive to arginine vasopressin and phenylephrine (205). In bovine adrenocortical cells, angiotensin II acts via the angiotensin type 2 (AT₂) receptor, since the AT₂ receptor agonist CGP-42112 stimulates the release of ouabain and the AT₂ receptor antagonist PD-123319 inhibits it (203). The phenylephrine-dependent release of ouabain from human CRL7050 and bovine adrenocortical cells in culture is blocked by the ₁-adrenergic receptor antagonist doxazosin. This was interpreted to indicate that the sympathetic nervous system is involved in regulation of the release of this hormone to the bloodstream (204). In fact, evidence has been obtained in studies with salt-hypertensive rats that high Na⁺ concentrations in plasma and cerebrospinal fluid may stimulate the release of endogenous ouabain via an Na⁺ sensor (143) or epithelial Na⁺ channels (ENaC) in the brain (376). This may then lead to a sympathetic hyperactivity and an elevation of brain angiotensin levels (87, 146).

When endogenous ouabain can be isolated from the hypothalamus (177), it may be synthesized there. A marked upregulation of genes coding for P-450 side-chain cleavage of cholesterol and ⁵,3-hydroxysteroid dehydrogenase/⁵,₄-isomerase (catalyzing the synthesis of progesterone from pregnenolone) was seen in the hypothalamus of hypertensive compared with normotensive Milan rats, but not in adrenal cells. Knockdown of the latter enzyme decreased the production of endogenous ouabain in neural tissue (270).

Endogenous Digoxin

There is much evidence that mammalian cells synthesize digoxin as well. A substance indistinguishable from digoxin was isolated from human urine and identified with fast atom bombardment-mass spectrometry, proton NMR, and several different HPLC systems (116). Deglycosylated and reduced forms of the immunoreactive digoxin-like factor were identified in bovine adrenal gland (118, 302). It was found in blood plasma, urine, adrenal gland, and breast cyst fluid (117, 334). Since digoxin is taken up from the gut at a much higher rate than ouabain, demonstration of digoxin's biosynthesis is a prerequisite to acceptance of the physiological role of this CTS. Qazzaz et al. (300) demonstrated that Y-1 murine adrenocortical tumor cells can use [1,2-¹⁴C]acetate and [4-¹⁴C]cholesterol as precursors for the synthesis of a [¹⁴C]digoxin-like substance. Its synthesis from acetate was inhibited by the 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor mevastatin. [7-³H]pregnenolone seems to be a precursor for digoxin in bovine adrenal cells (218). Hence, the unsaturated lactone ring in digoxin is not formed from the isoprenoid side chain, a conclusion supported also by the finding that radioactivity from [26,27-³H]25-cholesterol is not present in the CTS (218). Evidence for the existence of digitoxose sugars in mammals has been reported recently (303).

Endogenous Marinobufagenin and Other Bufadienolides

Marinobufagenin (3,5-dihydroxy-14,14-epoxybufadienolide; Fig. 1), originally discovered in amphibians, was isolated and identified from the urine of patients with myocardial infarction (18). Telocinobufagin, the reduced form of marinobufagenin, was identified, by high-resolution mass spectrometry and NMR, as a constituent of human plasma (192). Its plasma concentration is higher than that of marinobufagenin (192). The compound is synthesized from cholesterol in the adrenal cortex by a pathway that is independent of the cholesterol side-chain cleavage (65). 19-Norbufalin and its Thr-Gly-Ala tripeptide derivative were isolated from human cataractous lenses in an investigation of greater immunoreactivities against bufalin and ouabain in cataractous than in normal lenses (217).

Additional endogenous bufadienolides may circulate in mammalian blood. Sich et al. (343) demonstrated a correlation between the concentration of a proscillaridin A-immunoreactive substance with a polarity similar to that of ouabain with systolic blood pressure. A bufalin-like immunoreactivity also correlated with systolic blood pressure (282).

	MECHANISMS OF SIGNAL TRANSDUCTION OF CTS VIA THE SODIUM PUMP

TOP ABSTRACT CHEMICAL NATURE OF ENDOGENOUS... MECHANISMS OF SIGNAL... EFFECT OF ENDOGENOUS AND... PHYSIOLOGY AND PATHOPHYSIOLOGY... ENDOGENOUS CARDIAC GLYCOSIDES IN... GENERAL CONCLUSIONS FUTURE PERSPECTIVES GRANTS REFERENCES

 
CTS bind to Na⁺/K⁺-ATPase at a site formed in the extracellular part of the catalytic -subunit by the H₁-H₂, H₃-H₄, and H₅-H₆ loops. Tentative three-dimensional structures of the cardiac glycoside binding site have been proposed (181, 304). Reversible interaction of CTS with this site may induce a conformational change of the Na⁺/K⁺-ATPase protein. In the actively pumping Na⁺ pump, CTS are fixed by tight binding in the E₂ conformational state, a process leading to the enzyme's inactivation. Since the E₂-phosphoenzyme-CTS complex formed is almost irreversible (84), it is likely that, under cellular conditions, the ouabain-pump complex is internalized and degraded. K_d values of the CTS complexes vary with the form of the isozyme and the animal (33). Although K_d values of human ₁-, ₂-, and ₃-isoforms range from 10^–8 to 10^–9 mol/l (51, 265, 378), rodents exhibit an ouabain-insensitive ₁-isozyme (33). Since endogenous ouabain circulates in blood plasma of resting humans in the subnanomolar-to-nanomolar concentration range (132) and noninhibitory subnanomolar concentrations of ouabain stimulate the proliferation of smooth muscle (1, 17), endothelial (329), and kidney tubule cells in culture (66, 184), several mechanisms may transduce the docking of endogenous ouabain at the cell surface to the cell interior with or without inhibition of the Na⁺ pump. Two different mechanisms of signal transduction have been proposed (see below).

Na⁺-Lag Hypothesis and the PlasmERosome

The Na⁺-lag hypothesis assumes that a partial inhibition of the Na⁺ pump by cardiac glycosides leads to a transient increase of [Na⁺]_i, which in turn increases [Ca²⁺]_i via the Na⁺/Ca²⁺ exchange system (NCX1) running in a reverse mode (311). In cardiac and vascular smooth muscle cells, this exchanger is thought to contribute to Ca²⁺ extrusion from the cytosol in the relaxation process (36).

In smooth muscle cells, astrocytes, and hippocampal neurons of rodents, minimal inhibitory effects of CTS on the Na⁺ pump seem to be amplified by a special arrangement of the -isoforms in those cells (34, 37, 383). Immunochemical studies in rat arteries revealed that NCX1 and the ouabain-sensitive ₂- and ₃-isoforms (but not the rather insensitive ₁-isoform) of Na⁺/K⁺-ATPase reside in plasma membrane regions adjacent to the sarcoplasmic reticulum (SR) (113, 172, 173, 261, 411). The ₂-isoform of Na⁺/K⁺-ATPase is targeted and tethered to this site by Leu²⁷ and Ala³⁵ in its NH₂-terminal segment (346). This special arrangement may lead to a fortification of inhibitory effects of endogenous ouabain on ₂- and ₃-isozymes. In fact, enhanced Ca²⁺ transients evoked by the vasoconstrictors angiotensin II and phenylephrine have been recorded when the Na⁺ pump was inhibited, and opening of store-operated Ca²⁺ channels led to coentry of Na⁺ and Ca²⁺ (14). Since cytosolic Na⁺ levels did not increase (15), a local transient rise of [Na⁺]_i may occur in a subplasmalemmal space known as the plasmERosome (14, 209). In addition, a local increase of [Ca²⁺]_i may occur via the NCX1. The transiently increased Ca²⁺ in the plasmERosome, as well as in the bulk cytoplasm, is thought to be taken up by the SR via Ca²⁺-ATPase (SERCA). Because of the elevated Ca²⁺ content in the SR, more Ca²⁺ can be released when myocytes of vascular smooth muscle cells, astrocytes, or hippocampal neurons are stimulated by a rise of [Ca²⁺]_i in the plasmERosome. This occurs presumably via an activation of the inositol (1,4,5)-trisphosphate (IP₃) receptor (IP₃R) and leads to an augmented force of contraction (34, 37, 383) (Fig. 2). The specific subcellular localization of Na⁺ pump isoforms and the NCX1 protein is consistent with this concept: mice with an ouabain-insensitive mutant ₂-isoform of the Na⁺ pump failed to show cardiac inotropy upon ouabain treatment and did not develop ouabain-induced hypertension (71, 72). However, lowering of expression of the wild-type ₂-isoform increased blood pressure (411). The inhibition of Ca²⁺ entry via NCX1 by the inhibitor SEA-0400 lowers arterial blood pressure in salt-dependent hypertensive rat models. Heterozygous NCX1-deficient mice have low salt sensitivity. Transgenic mice with the NCX1.3 variant in their smooth muscle cells are salt hypersensitive (165, 166), but transgenic mice expressing the transgenic NCX1.1 are salt insensitive, and their blood pressure did not respond to SEA-0400 (166). These findings are consistent with Blaustein's plasmERosome hypothesis, which may explain, in part, the development of arterial hypertension at sustained elevated concentrations of endogenous ouabain. However, other mechanisms must exist as well, since, especially at subnanomolar ouabain concentrations, the Na⁺ pump (23, 108, 329) and cell proliferation (1, 17, 329) are stimulated. Furthermore, there is no strict correlation between the hypertensive action of CTS and inhibition of the Na⁺ pump (237).

View larger version (32K): [in this window] [in a new window]   Fig. 2. Na+-lag hypothesis of the action of endogenous ouabain (EO). The existence of a subplasmalemmal space, called the plasmERosome, is assumed. In this compartment, a transient local increase of intracellular Na+ concentration ([Na+]i) caused by stimulation of store-operated channels (SOC) or by ouabain (O)-dependent inhibition of 2- and 3-subunits of the Na+ pump (SP) may lead to a local rise of intracellular Ca2+ concentration ([Ca2+]i). This signal is amplified by a greater amount of Ca2+ in the sarco/endoplasmic reticulum (SR/ER) from sarcoplasmic Ca2+-ATPase (SERCA). Any agonist (angiotensin II or epinephrine)-stimulated Ca2+ release leads to an amplified Ca2+ release via the inositol trisphosphate receptor (IP3R), which stimulates opening of the ryanodine Ca2+ channel (RyR) and, thus, results in a higher degree of vasoconstriction. In a noninhibited Na+/K+-ATPase, Ca2+ in the plasmERosome will be extruded by the Na+/Ca2+ exchanger (NCX1) and the plasmalemmal Ca2+-ATPase (PMCA). Inhibition of NCX1 by SEA-0400 will suppress the effect of endogenous ouabain, as does the ouabain antagonist PST-2238.

Na⁺/K⁺-ATPase Signalosome

A large number of experiments support the hypothesis that Na⁺ pump inhibition is not necessary for the inotropic effect of cardiac glycosides in the myocardium (276). Because of the pioneering work of Xie and colleagues (396, 397), we are now starting to understand how nano- and subnanomolar concentrations of endogenous and exogenous cardiac glycosides may lead to increased inotropy of cardiac muscle and to hypertension, proliferation, differentiation, and altered cell life span (37, 68, 108, 317, 327, 418) (Fig. 3, see Fig. 5). In contrast to the Na⁺-lag hypothesis, the Na⁺/K⁺-ATPase signalosome uses all -isoforms to transduce the information of ouabain's binding from the Na⁺ pump to the cell interior and nucleus (7, 221, 223, 256, 329, 414). The signalosome is located in caveolar structures (56, 57, 98, 371, 397, 408) and may transfer signals to the cell interior, even when the pump is unable to work (216), and affect membrane recycling and trafficking, as well as cell-cell interactions (Fig. 3).

View larger version (34K): [in this window] [in a new window]   Fig. 3. Na+/K+-ATPase acting as a signalosome. Signaling reactions in various cells triggered by the interaction of ouabain with the Na+ pump are shown. Endogenous or exogenous cardiotonic steroids (CTS) affect the processes in various signal transducing pathways. AP-1, activator protein-1; ASK-1, apoptosis signal-regulating kinase-1; FAK, focal adhesion kinase; GSK-3, glycogen synthase kinase-3; IP3, inositol 1,4,5-trisphosphate; IP3R, IP3 receptor; MEK, mitogen-activated ERK-activating kinase (an MAPK kinase); NCX1, Na+/Ca2+ exchanger; PI3K, phosphoinositide 3'-kinase; Raf, an MAPK kinase kinase kinase; Ras, an MAPK kinase kinase; ROS, reactive oxygen species; Src, sarcoma kinase; EGFR, epithelial growth factor receptor.

View larger version (16K): [in this window] [in a new window]   Fig. 4. Cardiac glycosides and their derivatives with anticancer action.

View larger version (26K): [in this window] [in a new window]   Fig. 5. Effects of CTS on cell proliferation and apoptosis. Effects favoring proliferation/differentiation are indicated by solid arrows and those favoring apoptosis/cell death with dashed arrows. CDK, cyclin-dependent kinase; TNFR-1, TNF receptor type 1; Cyt c, cytochrome c; RSK, ribosomal S6 kinase.

Ouabain-Na⁺/K⁺-ATPase complex in caveolae stimulates different pathways of signal transduction. Interaction of endogenous and exogenous cardiac glycosides with the cardiac glycoside receptor of Na⁺/K⁺-ATPase, even at nanomolar concentrations of the drug, may lead to conformational changes that are recognized by neighboring proteins (396). Both processes lead to the activation of various signal transduction pathways.

CALCIUM AS SECOND MESSENGER. Changes of [Ca²⁺]_i upon cardiac glycoside binding are well known (256, 327, 329, 363, 414). The rise of [Ca²⁺]_i caused by nanomolar concentrations of ouabain may result from a stimulation of T- and L-type currents (212), an increase of voltage-dependent Ca²⁺ influx (244), an activation of promiscuous Na⁺ channels (327), or an activation of Ca²⁺ release channels of the SR (325). In cardiac myocytes, ouabain binding may lead to protein-protein interactions of the Na⁺ pump with Na⁺ channels, inducing the latter to become promiscuous in their ion selectivity and, thereby, allow Ca²⁺ influx from the extracellular fluid (327). This event may be connected to an activation of the ₂- and ₃-isoforms of the Na⁺ pump (108). Clearly, the Na⁺-lag hypothesis does not apply to such a condition (see above). Additional pathways can lead to a rise of [Ca²⁺]_i. An ouabain-induced conformational change of the Na⁺ pump may be transmitted via the -subunit's NH₂-terminal end to the IP₃R of the endoplasmic reticulum/SR (256, 414). This complex also contains PLC-₁, the Src-dependent phosphorylation of which may increase IP₃ formation, which in turn enhances Ca²⁺ release from intracellular stores (257, 258, 408). Additionally, activation of PLC-₁ stimulates diacylglycerol formation and, thus, PKC activity (120). It may also affect, by phosphorylation, the activities of L-type voltage-dependent Ca²⁺ channels (82, 103, 269, 396), the Na⁺/Ca²⁺ exchanger, and even the Na⁺/K⁺-ATPase itself (which may increase [Ca²⁺]_i further by Na⁺ pump inhibition) (257, 363) (Figs. 2 and 3). Nanomolar concentrations of ouabain have been reported to induce Ca²⁺ oscillations in hippocampal astrocytes and kidney tubule cells, as well as endothelial cells, but toxic ouabain concentrations lead to constantly elevated [Ca²⁺]_i (7, 224, 329). [Ca²⁺]_i may, furthermore, increase as a result of ouabain exposure via the Na⁺/K⁺-ATPase-Src-epidermal growth factor receptor (EGFR) complex, which activates the Ras-MEK-ERK1/2 pathway. This pathway may lead to the activation and phosphorylation of Ca²⁺ channels and/or the Na⁺/Ca²⁺ exchanger (257, 363) and, thereby, increase the heart's positive inotropy as well (257, 258) (Fig. 3). The frequency of Ca²⁺ oscillations seems to determine whether a cell undergoes proliferation or apoptosis (414). Low-frequency Ca²⁺ oscillations are known to improve the specificity and efficiency of gene expression (67). Ca²⁺ may stimulate protein biosynthesis via activation of PKC (160, 258, 292) and NF-B (7, 256, 414) or, in the heart, by the loosening of cell-cell contacts and myofibril disassembly and, hence, promote cardiac remodeling via activation of calpain (105, 136, 207). Low concentrations of ouabain triggering low-frequency Ca²⁺ oscillations activate NF-B and protect cells from apoptosis (414) (Fig. 3). Finally, activation of PKC and Ca²⁺-calmodulin kinase may activate the expression of early-response genes, such as c-fos and c-jun, leading to formation of the transcription factor AP-1 (292). These events also may result in expression of late growth-related genes, such as skeletal -actin, atrial natriuretic factor, myosin light chain 2, and transforming growth factor-1, in the heart (160) and may differentially regulate the expression of the isoforms of the - and -subunits of Na⁺/K⁺-ATPase (159, 191, 377).

ACTIVATION OF PHOSPHATIDYLINOSITIDE 3'-KINASE AND AKT (PROTEIN KINASE B). The NH₂-terminal end of the catalytic -subunit of Na⁺/K⁺-ATPase contains a binding motif for PI3K (397, 409). Its ouabain-induced conformational change was shown to activate cell proliferation in kidney proximal tubule cells via a Ca²⁺-dependent phosphorylation of Akt (PKB) (184). In the heart, this conformational change may lead to cardiac hypertrophy (142) and metabolic alterations (83). Akt-1-knockout mice show a reduced heart size and body size and develop cardiac dilation and dysfunction. Activated Akt may induce hypertrophy and cardiac dysfunction over time in some models (142). Ouabain-induced activation of Akt was reported to show an antiapoptotic action (366). Hence, Akt plays an important role in ouabain-induced signal transduction.

CAVEOLIN-SUPPORTED PATHWAYS. In rat cardiac myocytes, 20–30% of cellular ₁- and ₂-isoforms of Na⁺/K⁺-ATPase are in caveolae, along with most of the cellular caveolin-3. Caveolin-3 and the isoforms are located in peripheral sarcolemma and T tubules. Exposure of the contracting heart to ouabain led to a two- to threefold increase in activation/phosphorylation of ERK1/2 and a 50–60% increase in caveolar Src and the ₂-isoform of Na⁺/K⁺-ATPase, suggesting an ouabain-induced recruitment of Src and ₂-isoforms to caveolae as a prerequisite for the manifestation of ouabain's inotropic response (223). Also, kidney tubule cells form an ouabain-induced Na⁺/K⁺-ATPase-EGFR-Src-caveolin-1 complex, leading to an increased tyrosine phosphorylation of caveolin-1 and the subsequent activation of the Ras-Raf-ERK cascade, with changes in [Ca²⁺]_i and gene activation, cell proliferation, and hypertrophy (98, 184). Abolition of ouabain-induced recruitment of Src and the depletion of cellular caveolin-1 blocked this cascade (371). Another pathway resulting in endocytosis and the formation of clathrin coats, as well as early and late endosomes, may be triggered by cardiac glycosides from this complex (221). The clathrin coat also contains AP-2, PI3K, and clathrin heavy chain (220, 221). Formation of a complex of PI3K with the proline-rich motif of the NH₂-terminal sequence of the Na⁺ pump -subunit and its phosphorylation at Ser¹¹ and Ser¹⁸ by PKC [resulting in a critical conformational change (80)], as well as the binding of AP-2 to this NH₂-terminal domain, seems to regulate the trafficking of the Na⁺ pump (46, 183, 409). Intracellular Na⁺ modulates the phosphorylation of the -subunit of Na⁺/K⁺-ATPase by PKC (163), which seems to counteract the internalization and activation of the Na⁺ pump. This is also the case with the hypertension-linked mutation of adducin, which shows impaired trafficking in response to dopamine as a result of a higher phosphorylation state of the AP2-µ2 protein (81). A bufalin-induced stimulation of the formation of large vesicles adjacent to the nucleus (containing transferrin, low-density lipoprotein, and the Rab protein) was also seen in human NT2 (neuronal precursor) cells and interpreted to indicate stimulation of plasma membrane recycling (317). There have been reports that inhibition of endocytosis protects against the toxicity of cardiac glycosides (280) and that internalized hydrophobic cardiac glycosides may increase [Ca²⁺]_i and heart contractility via direct interaction with the RyR (281, 325).

SRC-EGFR-RAS-RAF-ERK CASCADE. Ouabain binding to the Na⁺ pump stimulates Src kinase, which in turn phosphorylates the EGFR, leading to activation of the Ras-Raf-MEK-ERK pathway (Fig. 3). Fluorescence resonance energy transfer studies revealed a close proximity of Na⁺/K⁺-ATPase to Src at the plasma membrane (362). Binding of Src to Na⁺/K⁺-ATPase is independent of the Na⁺ pump's catalytic activity (216). Binding of the SH2 domain of Src to cytosolic domains 2 and 3 of the Na⁺/K⁺-ATPase -subunit inhibits Src activity if no CST is bound to the Na⁺ pump (362). Apparently and as a consequence of an ouabain-induced conformational change of the Na⁺ pump, Src is released from the Na⁺/K⁺-ATPase-Src complex and activated by phosphorylation at Tyr⁴¹⁸ (124, 362). Consistent with an ouabain-induced release of Src from the Src-Na⁺/K⁺-ATPase complex, a knockout of the ₁-isoform of Na⁺/K⁺-ATPase in kidney epithelial LLC-PK1 cells led to an increased basal Src activity and also an increased tyrosine phosphorylation of focal adhesion kinase (216). As a consequence, dissociation of Src from the Src-Na⁺/K⁺ATPase complex leads to an increased tyrosine phosphorylation of EGFR (which is not identical to epidermal growth factor-induced autophosphorylation at Tyr¹¹⁷³) and to recruitment and phosphorylation of the adaptor protein Shc. This results in binding of the adaptor protein Grb2 to the Src-EGFR complex and, subsequently, activation of the p42/44 MAPK (123, 189). This signaling pathway seems to be activated with, as well as without, inhibition of the Na⁺ pump. In cardiac A7r5 and kidney LLC-PK1 cells, inhibition of Na⁺/K⁺-ATPase by ouabain clearly correlated with ouabain-induced activation of Src and MAPK. Evidently, ouabain inhibition of ₁- and ₂-isoforms of Na⁺/K⁺-ATPase transactivates the EGFR and, subsequently, stimulates the Ras-MAPK cascade (124). Yet, in cultures of vascular and prostate smooth muscle cells as well as in renal epithelial cells, stimulation of proliferation and activation of MAPK and ERK1/2 by ouabain or marinobufagenin did not correlate with Na⁺/K⁺-ATPase inhibition (1, 17, 66, 112). This is important, because the Na⁺/K⁺-ATPase-Src complex is the only known receptor for ouabain and other CTS that stimulates the protein kinase cascades (216).

Ras activation leads to further activation of three different branches of the signal transduction cascades. One of the pathways communicates with the mitochondria: the activation of MAPK and increase in [Ca²⁺]_i result in opening of mitochondrial ATP-sensitive K⁺ channels (364) and generation of mitochondrial reactive oxygen species (ROS) (363, 398). Both processes cooperate to increase cardiac muscular contraction (258, 363, 364). ROS subsequently activate NF-B (256, 398) and slow [Ca²⁺]_i oscillations at nanomolar ouabain concentrations (7). NF-B is involved as a transcription factor in processes related to growth, differentiation, and inflammatory responses (7) and inhibits apoptosis (111, 213, 366, 414) (Fig. 3). The second pathway, the Ras-Raf-MEK-ERK1/2 cascade (159, 189, 398), leads to gene activation when ERK1/2 is activated in cooperation with PLC in the presence of Ca²⁺ (257) and ROS (222). Interestingly, ERK1/2 controls surface expression of Na⁺/K⁺-ATPase in kidney tubule and muscle cells (10, 253). JNK might be a substrate of ERK (190), as well as the JNK kinase SEK1 (215). A third pathway of ouabain-dependent Ras activation in muscle cells results in activation of p90 ribosomal S6 kinase and inactivation of glycogen synthase kinase (GSK-3/) via activation of ERK1/2 and phosphorylation; this process stimulates glycogen synthesis (196) (Fig. 3). Since GSK-3 is a master switch regulating cell-fate specificity and tumorigenesis (185), inhibition or suppression of GSK-3 may also affect transcription factors and increase cardiac hypertrophy (133).

Ouabain, Na⁺ pump, and cell-cell interaction. A number of bidirectional interactions between the cytoskeleton and the Na⁺/K⁺-ATPase signaling pathway have been reported (31, 48–50, 342). The c-Src, p190^Rho-GAP, and ERK1/2 signaling pathways are known to regulate cell adhesion through specific attachment molecules and to modify the interaction with the extracellular matrix (50). Stimulation of a complex signaling cascade by CTS may promote cell-cell communication by means of gap junctions in epithelial cells by specifically enhancing connexin32 expression (206). The opposite may occur, however, on blockade of the Na⁺ pump: prolonged ouabain blockade of Na⁺/K⁺-ATPase in Madin-Darby canine kidney cells leads, via an increase in [Ca²⁺]_i, to a detachment of cells from each other, an increase in MAPK activity, and a redistribution of molecules involved in cell attachment (occludin, ZO-1, desmoplakin, cytokeratin, -actinin, vinculin, and actin). Inhibition of protein tyrosine kinases and MAPK kinase, respectively, blocks this detachment. The content of p190^Rho-GAP, a GTPase-activating protein of the Rho small G protein subfamily, is increased by ouabain (48), suggesting that the Rho/Rac and MAPK pathways are involved (50). However, in the MA104 rhesus monkey epithelial cell line, in which ouabain binding to Na⁺/K⁺-ATPase is of high affinity (K_m 4 x 10^–8 M), blockade of the Na⁺ pump failed to modify phosphorylation, as well as the pattern of distribution of associated molecules (50). Furthermore, other inhibitors of the Na⁺ pump were also without effect (48). This was interpreted to mean that, in MA104 cells, the signaling sequence is faulty and, in normal cells, inhibition of the Na⁺ pump affects their retrieval (perhaps in plasmalemma turnover) and, eventually, leads to a loss of cell-cell contact by - subunit interaction (110, 268, 332, 342) (Fig. 3).

	EFFECT OF ENDOGENOUS AND EXOGENOUS CARDIAC GLYCOSIDES ON CELL PROLIFERATION AND DEATH OF NORMAL AND CANCER CELLS

TOP ABSTRACT CHEMICAL NATURE OF ENDOGENOUS... MECHANISMS OF SIGNAL... EFFECT OF ENDOGENOUS AND... PHYSIOLOGY AND PATHOPHYSIOLOGY... ENDOGENOUS CARDIAC GLYCOSIDES IN... GENERAL CONCLUSIONS FUTURE PERSPECTIVES GRANTS REFERENCES

 
It is evident from the literature cited above that CTS may influence cell proliferation, differentiation, and, eventually, cell death via the Na⁺/K⁺-ATPase signalosome pathways (Fig. 3). It is unclear whether endogenous cardiac glycosides may affect tumor growth. In a preliminary study, Weidemann (388) found lowered plasma concentrations of endogenous digitalis immunoreactivity in 74% of patients with breast cancer but markedly increased concentrations in 10.8% of these patients. Since tumor cells may show abnormal Na⁺/K⁺-ATPase activity, which is explained by the "Na⁺/K⁺-leakage theory" (176, 388), clinical studies on the therapeutic effects of CTS seem reasonable. In fact, a recent clinical study by Stenkvist (349) reported that cardiac glycosides (digoxin and digitoxin) may have therapeutic effects on cancer: a group of 175 patients with breast cancer, 32 of which were on digitalis therapy when the disease was diagnosed, were studied over >22 yr. The death rate was significantly lower in patients treated with digitalis glycosides than in those not treated with digitalis (6% vs. 34%). In vitro studies show that a number of different cancer cell types are blocked in the G₂/M phase of the cell cycle (140, 249). Yet, in another careful study of 9,271 patients, Haux et al. (138) were unable to confirm Stenkvist's observation. However, fewer cases of leukemia were diagnosed in a group of patients treated for cardiac failure with digitoxin than in a control group of patients, who, after cancer diagnosis, required digitoxin treatment. There might be a cancer-protective trend for higher concentrations of digitoxin for lymphoma/leukemia and kidney/urinary organ cancers (138). Interestingly, studies with immune-compromised mice seem to indicate that failure of the adrenal glands to release endogenous digitalis-like immunoreactivity may facilitate the establishment of tumors (389). Bufalin derived from the toad skin has been used for several hundred years in traditional Chinese medicine to treat malignant diseases such as hepatocellular and hematologic cancers (169, 170, 404). Oleandrin, another hydrophobic cardenolide, is in phase I trials in the United States as an anticancer remedy for refractory human cancers (275). Since the therapeutic window for cardiac glycosides is very narrow (188), there is a need for cardiac glycoside derivatives without cardiotonic action, but with cytotoxic action, to overcome this problem. Fortunately, such compounds have been detected recently (Fig. 4, see Fig. 7) (54, 255, 368). One of these compounds, UNBS-1450, entered the candidate drug development stage of phase I clinical trials in Belgium in 2006 (255).

View larger version (10K): [in this window] [in a new window]   Fig. 6. Mechanism for discrimination of bufalin in human monocytic THP-1 cells between differentiation and apoptosis. nPKC and cPKC, novel and conventional PKC. [Modified from Kurosawa et al. (198).]

View larger version (8K): [in this window] [in a new window]   Fig. 7. Bufadienolide derivatives with tumor-specific cytotoxicity but without cardiac activity (54). N8, 3,11-diacetoxy-12-keto-5,14-aetianic acid methyl ester; N34, 3,16-diacetoxy-14,15-epoxyaetianic acid methyl ester; N83, 3-acetoxy-14,15-epoxyaetianic acid.

In vitro observations that leukemia cells undergo apoptosis in the presence of bufalin and oleandrin, but normal leukocytes do not, are consistent with the hypothesis of a therapeutic effect of CTS on tumor growth (169, 179, 245, 348). An endogenous bufalin-like factor in human blood may induce differentiation of leukemia cells (277). Hence, studies on the effects of various CTS on normal and tumor cell growth are published more frequently. A survey of the data reveals that exogenous CTS in the nanomolar concentration range and, certainly, endogenous CTS as well may stimulate cell proliferation and differentiation (1, 17, 47, 98, 112, 184, 306, 309) and protect normal cells and some cancer cells from apoptosis (213, 366), but there are also reports that hydrophobic CTS at rather low concentrations may induce apoptosis or death of cancer cells (47, 190, 219, 272, 278, 306, 405, 406) (Table 1).

Tumor cells are more sensitive to hydrophobic CTS, such as bufalin and oleandrin (Fig. 4), which may induce differentiation (279, 412) and/or apoptosis in a number of human myeloid leukemia cells, including HL-60, U937, and K562 cells (231, 384–386) (Table 1). Lower (nanomolar) concentrations of cardiac glycosides may stimulate differentiation (54, 169, 180, 264, 341, 359, 384, 412) or proliferation (306), but at >10 nM they induce cell death (306). In the picomolar-to-nanomolar concentration range, bufalin strongly induces differentiation of human promyelocytic HL-60, monoblastic U937, and myeloblastic ML1 cells, whereas other CTS, such as cinobufagin, ouabain, and digitoxigenin, had no effect or only weak effects (54, 169, 180, 264, 341, 359, 384, 412). The murine leukemia cell line Ml-T22 was insensitive to bufadienolides (279).

Higher concentrations of bufalin and oleandrin, however, induce apoptosis in leukemic and other tumor cells (169, 198, 232, 249, 272, 275, 278, 345, 384–386, 406) (Table 1). Induction of apoptosis by bufalin in human tumor cells is associated with an increase in [Na⁺]_i, i.e., an inhibition of the Na⁺ pump (179). The apparent bufalin specificity of the apoptotic effect in K562 human erythroleukemia cells is due to the higher affinity of Na⁺/K⁺-ATPase for bufalin, but other cardiac glycosides at higher concentrations may induce apoptosis as well (279). Human leukemic cell lines such as Jurkat (T cell leukemia) and Daudi (B cell leukemia) cells are among the most susceptible to induction of apoptosis by digitoxin treatment in vitro (139). Digitoxin may inhibit proliferation and promote apoptosis at concentrations commonly found in patients on digitalis therapy (140, 226).

Cytotoxic effects of cardiac glycosides in vitro have also been reported for a number of additional tumor cells, including mammary tumor (190), lung cancer (254), prostate cancer (140, 219, 249, 405), renal cancer (226), malignant melanoma cells (226, 413), and human skin squamous cell carcinoma (9).

Mechanisms of the Anticancer Effect of Cardiac Glycosides

We are far from understanding how cells, on interaction with CTS, can undergo responses such as proliferation, differentiation, cell cycle arrest, and apoptosis, which are so opposed in nature. Some intracellular signals for differentiation and apoptosis are activated simultaneously (197, 198, 385, 386). It might be the orchestration of various intracellular signals and the duration of activation that decide the path that is finally taken. A rough scheme incorporating most of the information is shown in Fig. 5.

Differentiation vs. apoptosis. Regulation of the cell cycle is an important target of intracellular signaling. At low nanomolar concentrations, the hydrophobic cardiac glycoside bufalin arrests the cell cycle of ovarian endometrial cyst stromal cells in the G₀/G₁ phase (272) and of human leukemia ML 1 cells and several prostate cancer cells in the G₂/M phase (140, 170) (Fig. 5). In ovarian endometrial cyst stromal cells, bufalin (10^–9 M) downregulates the expression of cyclin A, which is important for entering the S phase (DNA replication). The arrest of the cell cycle in the G₁/S phase is further supported by the upregulation of p21 (a suppressing cofactor of G₁S-Cdk). In human leukemia ML 1 cells, bufalin (10^–9 M) leads to changes in cyclin-dependent protein kinase (Cdk-2) and the Cdk inhibitor protein CKII, which controls transition from the G₂ to the M phase. Additionally, activities of PKC, PKA, and DNA topoisomerases I and II are inhibited (32, 170). Cell cycle arrest and activation of the non-cell cycle-dependent Cdk-5 and p25 formation (via p35 cleavage) by digoxin may induce apoptosis (219) (Fig. 5).

The mechanism by which a normal cell or a tumor cell enters the pathway of differentiation and proliferation or apoptosis seems to be essentially the same for normal and malignant cells. Nanomolar concentrations of CTS leading to low-frequency [Ca²⁺]_i oscillations and activation of NF-B protect cells from apoptosis (Fig. 3), whereas higher concentrations lead to a sustained increase of [Ca²⁺]_i and apoptosis (7). Protection from apoptosis by cardiac glycosides is achieved by activation of Akt (184, 417), which inactivates the killer protein Bad by phosphorylation (58, 184) (Fig. 5). Bufalin-induced differentiation is associated with activation of the conventional PKC subfamily (dependent on Ca²⁺ and diacylglycerol) (197). However, at higher concentrations (100 nM) of bufalin (which certainly lead to an inhibition of Na⁺/K⁺-ATPase and a rise of [Ca²⁺]_i), THP-1 cells undergo apoptosis via the activation of novel PKC, which is activated by diacylglycerol but does not require Ca²⁺ (198) (Fig. 6). An essential role of PKC in inducing apoptosis is also evident from the strong resistance of cells defective in conventional PKC- and novel PKC- isoforms to the bufalin-induced DNA ladder formation (198). Treatment of human leukemia THP-1 cells with bufalin sequentially induces c-fos and expression of inflammatory cytokine IL-1 and TNF- genes before the appearance of mature phenotypes of monocytic cells (197). Induction of differentiation of human monocytic leukemia THP-1 cells by bufalin into macrophage-like cells is characterized by loss of proliferation, cell adherence, increased ability to reduce nitro blue tetrazolium, and increased expression of IL-1. During this process, c-myb and c-myc expression is downregulated and c-fos and Egr-1 transcripts are induced. Furthermore, bufalin fails to induce c-fos expression and downregulate c-myb transcripts in low-Na⁺ medium. These findings indicate the importance of [Na⁺]_i handling in triggering the change in protooncogene expression and differentiation (200, 278, 283), inasmuch as it is the "Na⁺ cycle" for cell proliferation (324).

It is puzzling that the ERK cascade seems to control bufalin-induced cell differentiation and apoptosis simultaneously (197, 385, 386). Experiments in PC-12 cells suggest that transient activation of ERK promotes proliferation, whereas sustained ERK activation promotes differentiation (350). Possibly, distinct preceding signals (perhaps [Ca²⁺]_i) modulate the ERK cascade. It has been suggested that scaffold or adaptor proteins may participate in such a regulation (391, 399). Additionally, differences in the subcellular localization and substrate specificity of PKC isozymes (316), as well in the ERK cascade, seem to be important. ERK activation has been shown to play a critical role in the bufalin-induced differentiation of human monocytic leukemia THP-1 cells. p38 MAPKs and their downstream mediators may modulate ERK activity and, eventually, cell differentiation (197).

Sustained increase in [Ca²⁺]_i by cardiac glycosides proves to be an important factor in cell death in cardiac glycoside-triggered apoptosis of tumor cells (219, 226, 249, 279, 392, 405, 406). The bufalin-induced apoptosis in endometrial cells in the G₀/G₁ cell cycle is connected to the downregulation of cyclin A, Bcl-2, and Bcl-xL expression and the simultaneous upregulation of p21 and Bax expression and caspase-9 activation (272) (Fig. 5). In the estrogen receptor-negative human breast cancer cell line MDA-MB-435s, ouabain lowers cell proliferation by cell cycle arrest and activation of ERK1/2, leading to an increased expression of the cell cycle inhibitor p21^CIP1 but a decreased expression of the tumor suppressor protein p53 and activation of JNK (190). Oleandrin, another potent carcinostatic cardiac glycoside (Fig. 4), suppresses the activation of NF-B, AP-1, and the associated JNK in lymphoma cell lines (231, 348). It also activates the expression of Fas, which is considered to be responsible for apoptosis. Such a Fas induction is not seen in primary cells (348). Persistent activation of the MAPK pathway by bufalin (386) and altered expression of c-myc and bcl-2 genes are involved in apoptosis (179, 180, 245, 272, 385, 386, 392), as well as in the activation of Rac1, p21-activated kinase, and JNK pathways (180, 231). Overexpression of the survival protein Bcl-2 has an antiapoptotic effect (392). Oleandrin stimulates the formation of ROS in melanoma cells, which leads to a release of cytochrome c from mitochondria. This release in turn activates caspases, leading to apoptosis (275). Additionally, activation of apoptosis signal-regulating kinase 1 (ASK-1) may occur (Fig. 3), resulting in activation of JNK and apoptosis, as shown with palytoxin, a very potent ligand at the cardiac glycoside binding site of Na⁺/K⁺-ATPase (387). In the prostate cell lines PC-3, LNCaP, and DU145, the CTS digoxin, ouabain, and bufalin may induce apoptosis specifically via activation of caspase-3, -8, and -9 (249, 406), early cytochrome c release from mitochondria, and ROS generation without damaging other cells (161, 249, 405, 406). This is also true for the cytotoxic effects of bufadienolides and their derivatives on malignant T lymphoblasts that occur via the classical caspase-dependent pathway with damage to mitochondria and internucleosomal DNA fragmentation (54). In endometriotic stromal cells, bufalin also supports apoptosis by activation of caspase-9 (272) (Fig. 5). The Ca²⁺-activated protease calpain is known to be activated by ouabain in myocardial cells (136). This activation of protease activity by phosphorylation through ERK (104, 105) may occur also in tumor cells. Activated calpain may release cytochrome c from mitochondria and inactivate antiapoptotic proteins such as Bcl-2 and Bcl-xL by proteolysis, thereby promoting cell death (134, 307) (Fig. 5).

Search for More Sensitive Antitumor Derivatives of Cardiac Glycosides

The cardiotoxic effects of cardiac glycosides represent a major obstacle for their use in cancer therapy. Hence, Daniel et al. (54) searched for bufadienolide derivatives that induce apoptosis without affecting the heart (Fig. 7). Such compounds contain the configuration of the specific bufadienolide steroid backbone with A/B and C/D cis configurations. Additionally, a cardenolide derivative from the root bark of Calotropis procera, 2'''-oxovoruscharin, and its derivative UNBS-1450 (Fig. 4) are potent drugs with antitumor activity at nanomolar concentrations in a panel of 57 human cancer cell lines (368). UNBS-1450 had the maximally tolerated dose (i.e., the highest tolerated daily administered intraperitoneal dose for 28 days) of 80 mg/kg in mice, which is 24 times higher than that of ouabain (5 mg/kg) and 12 times higher than that of the parent compound oxovoruscharin (368). In non-small lung cancers (NSCLC), which are associated with a very dismal prognosis, chronic in vivo intraperitoneal and oral UNBS-1450 treatment of immunodeficient mice with metastases into the brain and liver had very significant therapeutic effects (255). Human A549 NSCLC cells showed highly activated cytoprotective NF-B signaling pathways. UNBS-1450 affected the expression and activation status of different constituents of the NF-B pathways in A549 tumor cells. The modifications induced by UNBS-1450 led to a decrease in the DNA binding capacity of the p65 subunit and the NF-B transcriptional activity (255). Furthermore, UNBS-1450 leads to the induction of nonapoptotic cell death and decreases heat shock protein 70 at the mRNA and protein levels and downregulates nuclear factor of activated T cells/tonicity-responsive enhancer-binding protein (a factor responsible for the transcriptional control of heat shock protein 70). It also induces an increase in permeability of lysosomal membranes of NSCLC cells (254). Since UNBS-1450 showed effective in vivo antitumor activity in nude mice carrying subcutaneous xenografts of human NCl-H727 and A549 cells, it is entering phase I clinical trials in Belgium (254, 255).

	PHYSIOLOGY AND PATHOPHYSIOLOGY OF ENDOGENOUS CARDIAC GLYCOSIDES IN THE CIRCULATORY SYSTEM

TOP ABSTRACT CHEMICAL NATURE OF ENDOGENOUS... MECHANISMS OF SIGNAL... EFFECT OF ENDOGENOUS AND... PHYSIOLOGY AND PATHOPHYSIOLOGY... ENDOGENOUS CARDIAC GLYCOSIDES IN... GENERAL CONCLUSIONS FUTURE PERSPECTIVES GRANTS REFERENCES

 
Because of the existence of manifold CTS-induced intracellular signaling pathways in a great number of cells, it is likely that the various endogenous cardiac glycosides act as a new class of steroid hormones. In fact, the quite substantial information on the role of the different endogenous CTS in cardiac and kidney function and the regulation of salt and mineral metabolism strengthens the concept that endogenous cardiac glycosides play an essential role in the physiology and pathophysiology of blood pressure regulation and development of arterial hypertension. It is probable that the various endogenous CTS with different functional and tissue specificity cooperate in the physiological tasks. Differences in the specificity of endogenous ouabain and marinobufagenin for kidney and heart function, as well as for the circulatory system, have been described (235). On a molecular level, such differences may be caused by variations in the affinities of various Na⁺/K⁺-ATPase isoforms for specific CTS, by altered interaction times of the cardiac glycoside receptor site with hydrophobic vs. hydrophilic CTS, by a slightly different conformation of the -subunit upon CTS binding due to an induced-fit mechanism, and, certainly, also by variations of the gene expression pattern of the various intracellular signaling proteins and other proteins in the target cells.

Endogenous Ouabain, a Blood Pressure-Modulating Hormone?

Ouabain and hemodynamic effects. Endogenous ouabain is structurally identical to plant-derived ouabain (127, 177, 193, 247, 333), and they have identical cardiotonic and vasotonic actions in guinea pig left atria and aortic rings (39). Plasma concentration of endogenous ouabain is mostly determined after a specific cleanup process by cross-reaction with ouabain-specific antibodies (26, 96, 135). Endogenous ouabain increases in blood plasma in stress situations demanding acutely increased blood supply of organs in humans and dogs running on a treadmill (26) or in swimming rats (119). Ouabain rises rapidly and concomitantly with epinephrine (119) when physical exercise starts and declines quickly upon rest. It is likely that the substantial increases in endogenous ouabain induce an inotropic effect on heart function. Pretreatment of dogs with the -blocker atenolol, as well as with the angiotensin-converting enzyme (ACE) inhibitor benazepril, abolished the exercise-dependent rise in endogenous ouabain levels, indicating that the release of ouabain in dogs is controlled by epinephrine and angiotensin II (26, 146). Also, ACTH raises the plasma concentration of endogenous ouabain in humans and rats (347, 402). Since plasma concentration of endogenous ouabain correlates with systolic and mean arterial blood pressure, the behavior of endogenous ouabain is similar to that of a blood pressure-modulating factor (235, 242, 297, 379). Most likely, the increased concentration of endogenous ouabain in the cord blood of newborns (64) is the result of stress during delivery. Intra-arterial in