Báo cáo khóa học: Ouabain stimulates endothelin release and expression in human endothelial cells without inhibiting the sodium pump Robert Saunders and Georgios Scheiner-Bobis pptx

Ouabain stimulates endothelin release and expression in human endothelial cells without inhibiting the sodium pump Robert Saunders and Georgios Scheiner-Bobis Institut fu ¨ r Biochemie und Endokrinologie, Fachbereich Veterina ¨ rmedizin, Justus-Liebig-Universita ¨ t Giessen, Germany Ouabain, a sodium pump (Na + /K + -ATPase) inhibitor, has been shown to act as a hormone and is possibly involved in the pathogenesis of hypertension. The mechanism by which ouabain may act was investigated using primary cultures of human umbilical artery endothelial cells (HUAECs), which are known to express and release the vasoconstrictive hor- mone endothelin (ET-1). Five minutes after application, low concentrations of ouabain induced Ca 2+ oscillations and stimulated ET-1 release from endothelial cells into the medium. To investigate whether the observed effects were due to inhibition of the sodium pump, the effects of ouabain on the uptake of 86 Rb + by HUAECs were examined. Unexpectedly, ouabain concentrations below 10 n M stimu- lated 86 Rb + uptake by 15–20%, and in some experiments by 50%, results that are consistent with a stimulation of the pump. Within the concentration range 1–10 n M , ouabain induced a 2.5-fold stimulation (phosphorylation) of mito- gen-activated protein kinase (MAP kinase). After incuba- tion of HUAECs with ouabain for 12 h, the glycoside stimulated cell growth by 49 ± 4%, as measured by cell number, with a maximum response at 5 n M . At similar concentrations, ouabain also increased ET-1 mRNA 1 abun- dance by 19.5 ± 3.1%. The results indicate that, by influ- encing ET-1 expression and release, ouabain may contribute to the regulation of vascular tone. The data also confirm that it is not a global inhibition of the sodium pump that is involved in the mechanism of action of this cardiac glycoside. Keywords: endothelin; human umbilical cord endothelial cell (HUAEC); mitogen-activated protein kinase (MAP kinase); sodium pump; ouabain 2 . Na + /K + -ATPase (the sodium pump) is a membrane- embedded protein in all animal cells that couples ATP hydrolysis to a vectorial transport of Na + and K + ions against their electrochemical gradient. For each ATP hydrolyzed, three Na + ions are moved out of the cytosol and two K + ions are taken up from the environment, resulting in the formation and maintenance of a negative membrane potential. The sodium pump is specifically inhibited by a series of naturally occurring steroids, termed cardiac steroids or cardiac glycosides, such as ouabain or digitalis glycosides such as digoxin or digitoxin [1]. Inhibition of the sodium pump by cardiac steroids is of clinical use, as application of these substances, especially digitalis and its congeners, helps to increase muscular contractility of the failing heart [2]. In recent years, various research groups succeeded not only in isolating circulating factors that interact with the sodium pump and inhibit 86 Rb + uptake (Rb + is a surrogate for K + ) but also in identifying several of them as ouabain or its isomer [3–6] or as its congeners, such as digoxin [7,8], proscillaridin A [9], 19-norbufalin [10] and marinobufagenin [11]. In addition, evidence was provided in several investi- gations that the level of so-called endogenous ouabain increases in the plasma upon excessive work and is present in higher concentrations in the serum of hypertensive patients [12]. All of these data indicate that ouabain may be directly or indirectly involved in the regulation of vascular tone and possibly also in the pathogenesis of hypertension. To address mechanisms that may be involved in the generation of hypertension, we investigated the effects of ouabain on human endothelial cells in culture. Experimental procedures Isolation and culture of human umbilical cord endothelial cells (HUAECs) Human umbilical cords were collected within 2 h of birth and kept on ice in wash buffer [Hanks buffered salt solution (HBSS) containing 20 m M Hepes] until ready for cell isolation. An artery was cannulated, washed with the above solution, filled with collagenase (CLS 2; Worthing- ton) in Pucks saline solution (Seromed, Berlin, Germany) containing 20 m M Hepes, and incubated at 37 °Cfor 20 min to detach the endothelium. The cells were washed out using 20–50 mL wash buffer containing 10% fetal bovine serum and centrifuged at 50 g,4°C for 10 min. The cell pellet was suspended in 10 mL endothelial cell Correspondence to G. Scheiner-Bobis, Institut fu ¨ r Biochemie und Endokrinologie, Fachbereich Veterina ¨ rmedizin, Justus-Liebig- Universita ¨ t Giessen, Frankfurter Str. 100, D-35392 Giessen, Germany. Fax: + 49 641 9938189, Tel.: + 49 641 9938180, E-mail: Georgios.Scheiner-Bobis@vetmed.uni-giessen.de Abbreviations: ET-1, endothelin; HUAEC, human umbilical cord endothelial cell; HBSS, Hanks buffered salt solution; ECGM, endothelial cell growth medium; MAP kinase, mitogen-activated protein kinase. (Received 17 October 2003, revised 16 January 2004, accepted 26 January 2004) Eur. J. Biochem. 271, 1054–1062 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04012.x growth medium (ECGM; Promocell, Heidelberg, Ger- many) and transferred to a gelatin-coated, 94-mm cell culture dish. After 4 h, the medium and any nonadherent cells were aspirated and the medium replaced. During culture, the medium was replaced every 48 h. After the first passage, ECGM was mixed with M199 (Gibco, Eggenstein, Germany) to give a 2 : 1 ratio (v/v), with an additional 2% fetal bovine serum supplement (ECGM-2). Measurement of Ca 2+ oscillations induced by ouabain Relative changes in intracellular Ca 2+ concentration were measured using the Ca 2+ -sensitive fluorescent dye, fura-2. The cells were loaded as described below with the membrane-permeable acetoxymethylester form of the dye (fura-2 AM; Molecular Probes, Leiden, the Netherlands) 3 , which is then converted into nonpermeable fura-2 by intracellular esterase activity. Glass coverslips of 18-mm diameter were coated with 0.1 gÆL )1 poly( L -lysine) (Seromed) for 30 min at 4 °Cin 12-well plates before a wash with phosphate-buffered saline (NaCl/P i ). Thereafter, 1.5 · 10 5 HUAECs were pipetted on to each coverslip in 300 lL ECGM-2 and allowed to adhere for 1 h at 37 °C under 5% CO 2 before the addition of a further 300 lL medium. The medium was changed every 48 h thereafter. After 4 days of culture on the coverslips, the cells were incubated for 1 h in 600 lL ECGM-2 containing 2.5 l M fura-2 AM, 0.01% (w/v) Pluronic F-127 (Molecular Probes) at 37 °C, 5% CO 2 . Then the incubation medium was carefully removed by a pipette and cells were washed once with HBSS/20 m M Hepes. During microscopy, the cells were maintained in fresh HBSS/20 m M Hepes. Imaging was carried out on an inverted microscope (Olympus IX-50) equipped with an epifluorescence set-up and an image analysis system (Till Photonics, Martinsried, Germany). The emission above 470 nm was measured from several regions of interest, each approximately the size of one cell. The cells were excited alternately at 340 nm and 380 nm, and the ratio of the emission signal at the two excitation wavelengths was calculated. 86 Rb + uptake at various extracellular ouabain concentrations HUAECs were plated at a density of 2 · 10 4 cells per well of a 12-well plate (Greiner, Frickenhausen, Germany) precoated with 1% gelatin (Bio-Rad, Munich, Germany) and grown to confluency. The cells were washed three times in a potassium-free uptake medium containing (in m M ) NaCl 150, Hepes 10, glucose 10, RbCl 5.0, MgCl 2 5.0, and CaCl 2 0.5, pH 7.0, and equilibrated in uptake medium with various concentrations of ouabain for 30 min at 37 °C. Then, 1 lCi 86 Rb was added to each well, and incubation was continued for an additional 60 min. Afterwards, the medium was aspirated and the cells washed three times with ice-cold 0.1 M MgCl 2 to stop pump activity and remove excess 86 Rb + . The washed cells were then disrupted by treatment with 10% (w/v) trichloroacetic acid at 4 °Cfor 1 h to release intracellular 86 Rb + , and the radioactivity in the lysate was measured by liquid-scintillation counting. The 86 Rb + activity was normalized against the amount of protein [13] per well. 86 Rb + uptake was also investigated after preincubating the cells for 15 min with 1 l M protein kinase C inhibitor Ro-31-8425, 5 l M Na + -channel inhibitor tetrodotoxin, 500 l M mitochondrial ATP-sensitive K + [mitoK(ATP)] channel inhibitor 5-hydroxydecanoate, or with 50 l M Na + /K + /2Cl – cotransporter inhibitor bumetanide. Binding of ouabain to sodium pumps on the surface of the HUAEC plasma membrane This experiment was performed to investigate whether ouabain treatment influences the sodium pump number on the cell surface. Theexperimental conditions were the same as for the Rb + -uptake experiment. After incubation for 30 min in the uptake medium containing either no ouabain or 1 n M or 5 n M [ 3 H]ouabain (6.7 · 4 10 5 MBqÆmmol )1 ; Amersham- Pharmacia, Freiburg, Germany), [ 3 H]ouabain was added instead of radioactive Rb + to a final concentration of 100 n M and incubation was continued for another 30 min. After- wards, the medium was removed by aspiration, and cells were washed twice in 500 lL ice-cold water. Then cells were dissolved by incubation for 20 min at 70 °Cin500lL1 M NaOH. Radioactivity was counted in a liquid-scintillation counter after neutralizing 250 lL of the solution with 250 lL 1 M HCl and adding 3 mL liquid-scintillation fluid. Assay of mitogen-activated protein kinase (MAP kinase) activation HUAECs were plated at a density of 1.5 · 10 4 cells per well of a 24-well plate and grown to confluency. The cells were then serum-starved (i.e. 0.5% serum) in a 2 : 1 M 199/ ECGM (basal medium) mix (ECGM-SF) for 48 h before the start of the experiment. Serum starvation was necessary because otherwise serum components might induce MAP kinase activation independently of ouabain. Ouabain was added to the cells at various concentra- tions in ECGM-SF, and the cells were further incubated at 37 °C for 30 min. The cells were then lysed using a commercially available cell lysis buffer (Cell Signaling Technology, Frankfurt, Germany). The lysates were screened for MAP kinase activation using SDS/PAGE (10% acrylamide, 0.3% N,N¢-methylenebisacrylamide) [14] followed by Western blotting with an antibody against phospho-p44/42 (Cell Signaling Technology), and the resulting signal was visualized using the luminescent ECL system (Amersham Pharmacia). In all cases, the protocols of the providers were followed. A positive control derived from HEK cells after stimulation with serum (Cell Signaling Technology) was also run in parallel. Pre-stained proteins (Cell Signaling Technology) were used as molecular mass markers. The bands were relatively quantified using a digital documentation system (Biostep, Jahnsdorf, Germany) and Phoretix TotalLab gel image analysis software (Biostep). To ensure that changes in phosphorylation are not due to changes in the overall content of MAP kinase, parallel samples were probed following the above protocols with the only exception that in this case an antibody against the overall (phosphorylated and nonphosphorylated) MAP Ó FEBS 2004 Ouabain stimulation of ET-1 release and expression (Eur. J. Biochem. 271) 1055 kinase (Cell Signaling Technology) was used in place of the antibody against the phosphorylated forms of MAP kinase. The same cell extracts were also used here as a positive control. Effect of ouabain on cell number HUAECs were plated at a density of 2 · 10 4 cells per well of a 12-well plate precoated with 1% gelatin (Bio-Rad) and grown to confluency. The cells were then incubated for 12 h in a 2 : 1 M 199/ECGM mix with 0.5% fetal bovine serum (ECGM-0.5) to minimize serum-related growth signals before the start of the ouabain treatment. Ouabain was added to the cells at various concentrations in fresh ECGM- 0.5, and the cells incubated for an additional 12 h. For counting, cells were detached and dispersed in trypsin/ EDTA (Gibco). The trypsin was then neutralized using 2 : 1 M199/ECGM with 50% fetal bovine serum and placed on ice. Suspended cells were counted using a Neubauer haemocytometer. Reverse transcription Total cellular RNA was isolated from HUAECs using the RNeasy kit (Qiagen, Hilden, Germany). Then, cDNA synthesis with Moloney murine leukemia reverse transcrip- tase was carried out by following the protocol of the enzyme provider (Promega, Mannheim, Germany). PCR amplification Changes in the expression of endothelin (ET-1)mRNA transcripts were measured using PCR. The 421-bp region between bases 646 and 1067 of NM_001955.1 (GenBank; NIH, Bethesda, MD, USA), the human prepro-endothelin-1 gene, was amplified with the primers: 5¢-GACCGTGA GAATAGATGCCAATGTGCT-3¢ and 5¢-CTCCTGCT CTGATCCCAGCCAG-3¢. The sequences of primers used for the detection of the sodium pump a1-mRNA, a2-mRNA and a3-mRNA have been published [15]. Normalization was performed by comparison with amplif- icates of the housekeeper gene, glyceraldehyde-3-phosphate dehydrogenase, which was amplified in parallel using the primers 5¢-TGGGGAAGGTGAAGGTCGGAGTCAA-3¢ (ET-1 FORW) and 5¢-TAAGCAGTTGGTGGTGCAG GAGGCA-3¢ (ET-1 REV) to amplify the 469-bp region between bases 62 and 531 of NM_002046.1 (GenBank), the human glyceraldehyde-3-phosphate dehydrogenase gene. PCR amplification was performed in a gradient PCR ÔMaster CyclerÕ (Eppendorf, Hamburg, Germany) following published protocols [15]. After 24 amplification cycles, the amplified DNA was separated by electrophoresis on a 1% agarose gel in 40 m M Tris/acetate/2 m M EDTA buffer, pH 8.5, visualized by ethidium bromide, and quantified using a digital gel documentation system and the PHORETIX TOTALLAB gel image analysis software. Dot-blot measurement of ET-1 protein A total of 300 lL cell culture supernatant was blotted under vacuum through a dot-blot apparatus on to a nitrocellulose membrane presoaked in Tris/NaCl. The membrane was then blocked in Tris/NaCl containing 5% (w/v) skimmed milk for 1 h before three washes in Tris/NaCl containing 0.1% (v/v) Tween 20. This was followed by incubation with a mouse primary monoclonal antibody to ET-1 (MA3-005; Dianova, Hamburg, Germany) diluted 1 : 2000 in Tris/ NaCl containing 0.1% (v/v) Tween 20 for 1 h before another wash as above. The membrane was then incubated with the secondary antibody, an anti-mouse IgG (NIF-825; Amersham-Pharmacia) diluted 1 : 2500 in Tris/NaCl con- taining 0.1% Tween 20. The membrane was washed four times in the same buffer before luminescent detection of ET-1 dots using the ECL system (Amersham-Pharmacia). Dots were relatively quantified using the Phoretix TotalLab array measurement. Under the conditions used here, the signals follow a linear dose/response relation in the range 0.1–32 ng ET-1 per vial. This was established in experiments using ET-1 protein as a standard (QBiogene-Alexis, Gru ¨ nberg, Germany), which was dissolved in cell culture medium before being investi- gated under otherwise identical conditions. Statistical analysis Statistical analysis of the results was carried out by an unpaired, two-tailed t test. P < 0.05 indicates that results are significantly different from each other. Results Immediate, nongenomic effects Effect of ouabain on Ca 2 + concentrations in endothelial cells. As low concentrations of ouabain have been shown to cause low frequency Ca 2+ oscillations in rat tubule cells [16], we were interested in investigating the effects of ouabain on cytosolic [Ca 2+ ] of HUAECs. As shown in Fig. 1B, ouabain at the low concentration of 1 n M induced clear Ca 2+ oscillations in these cells with a period of about 4–8 min (Fig. 1B), sometimes even greater (Fig. 1C). First oscillations were observed after 4 min (Fig. 1C). The cells showing oscillations did not appear to be synchronized. Nevertheless, not all cells investigated displayed such oscillations. In the best case, four of the six cells observed displayed Ca 2+ oscillations. The average over three inde- pendent experiments was, however, about 38%. In the absence of ouabain, however, control cells never showed any calcium oscillations (Fig. 1A), indicating that the observed slow Ca 2+ oscillations were specifically induced by ouabain. Effects of ouabain on ET-1 release. ET-1 is synthesized as a prepro-hormone and is stored in vesicles. As vesicular exocytosis is induced by a rise in intracellular [Ca 2+ ], we were interested in determining whether ouabain added to endothelial cells in culture might induce such a response. Various concentrations of ouabain were added to the cell culture wells, and, after 10 min of incubation, the super- natant was collected and analyzed for its ET-1 content using the dot-blot method. The data in Fig. 2A show a stimula- tion of ET-1 release even at 1 n M ouabain. The rise in ET-1 release follows a hyperbolic dose/response relation and reaches a more than twofold stimulation at 50 n M ouabain (Fig. 2A). 1056 R. Saunders and G. Scheiner-Bobis (Eur. J. Biochem. 271) Ó FEBS 2004 When extracellular Ca 2+ was omitted by replacing the HBSS/20 m M Hepes with NaCl/P i , ET-1 in the supernatant was considerably reduced (Fig. 2B). The same was observed when the Ca 2+ channel blockers NiCl 2 (1 m M )andCdCl 2 (1 m M )andtheNa + /Ca 2+ -exchanger-specific inhibitor 2¢,4¢-dichlorobenzamil (0.1 m M ) were included in HBSS/ 20 m M Hepes and allowed to act on the HUAECs for 20 min before the addition of ouabain (Fig. 2B). Under all these conditions, ET-1 in the supernatant was about half of the amount detected in the controls and only 25% of the ET-1 secreted after ouabain stimulation. Effects of ouabain on 86 Rb + uptake. To investigate whether the observations made thus far are based on a global inhibition of the sodium pump, the uptake of 86 Rb + into endothelial cells was determined as a function of the ouabain concentration. Rubidium is recognized by the sodium pump and its uptake can easily be inhibited by ouabain and several of its congeners. This experiment, however, produced a most unexpected result. Ouabain at low concentrations not only failed to inhibit 86 Rb + uptake Fig. 1. Ouabain-induced Ca 2+ oscillations in HUAECs. (A) Without ouabain, (B) with 1 n M ouabain (arrowhead), or (C) with 10 n M ouabain. Cells were cultured on coverslips and loaded with the Ca 2+ - sensitive dye fura-2. Traces shown are for individual cells that were representative of most of the cells observed in a field of view. Ouabain or control treatment started at the arrowhead. See Experimental procedures for details. (D) False-color image of endothelial cells loa- ded with fura-2. The picture was taken at an excitation of 340 nm. The scale on the right shows a relative loading condition with Ca 2+ . Fig. 2. Effects of ouabain and extracellular Ca 2+ on ET-1 release from HUAECs. (A) Accumulation of ET-1 in the medium of HUAECs was measured by the dot-blot immunological method. Medium was col- lected after 10 min of incubation with the various concentrations of ouabain shown (bars represent ± SEM; n ¼ 6). (B) In the absence of extracellular Ca 2+ , ET-1 in the supernatant is reduced by 50% when compared with the control and is only 25% of the amount secreted after stimulation by 10 n M ouabain. A mixture of the Ca 2+ channel blockers NiCl 2 (1 m M ), CdCl 2 (1 m M )andtheNa + /Ca 2+ -exchanger inhibitor 2¢,4¢-dichlorobenzamil (0.1 m M ) added before ouabain have a similar effect. Ó FEBS 2004 Ouabain stimulation of ET-1 release and expression (Eur. J. Biochem. 271) 1057 by the endothelial cells, but with all HUAEC preparations from various umbilical cords a stimulation of 86 Rb + uptake was observed at low ouabain concentrations. Although in most cases the stimulation observed was 15–20% above the control without ouabain, in a series of measurements with HUAECs prepared from umbilical cord number 4, stimu- lation of 86 Rb + uptake reached 50 ± 22% at 0.1 n M ouabain and 49 ± 2% at 1 n M ouabain over the control without ouabain (Fig. 3A). To investigate the mechanism for the observed stimula- tion of 86 Rb + uptake by ouabain, the same experiment was carried out after preincubation of the cells for 15 min with various substances known to be specific inhibitors of cellular components. Thus, tetrodotoxin, which specifically inhibits Na + channels, was used to determine whether the observed stimulation was due to a secondary stimulation of the sodium pump by Na + cations that enter the cell via these channels. The protein kinase C inhibitor Ro-31-8425 [17] was used to see whether the observed stimulation of the 86 Rb + uptake was the result of Na + /K + -ATPase phos- phorylation by this kinase, which in the past has been repoted to activate or inactivate the sodium pump. The 5-hydroxydecanoate inhibitor of the mitochondrial ATP- sensitive K + [mitoK(ATP)] channels [18] was used to investigate whether the increased accumulation of 86 Rb + was the result of increased transportation of the cation into the mitochondria, and finally, bumetanide, the specific inhibitor of Na + /K + /2Cl – cotransporters [19], was used to investigate whether the observed stimulation of Rb + uptake was due to the stimulation of this uptake system. As shown in Fig. 3B, however, none of these substances had any effect on the stimulation of 86 Rb + uptake by ouabain when used at concentrations reported to affect the various channels and enzymes described above. Finally, ouabain-binding experiments with whole cells were carried out to investigate whether the observed stimu- lation of 86 Rb + uptake is due to a translocation of sodium pumps from cytosolic compartments to the surface of the plasma membrane. The experiments, however, did not indicate any differences in [ 3 H]ouabain binding to the membrane surface after the cells were preincubated with either 1 or 5 n M [ 3 H]ouabain. Whereas cell membranes that were not preincubated with ouabain bound 685 ± 58 fmolÆmg )1 protein, ouabain binding after preincubation with either 1 or 5 n M ouabain was 638 ± 68 or 702 ± 54 fmolÆmg )1 , respectively (all values are mean ± SEM; n ¼ 6). MAP kinase activation by ouabain In rat cardiomyocytes, ouabain has been shown to stimulate the MAP kinase reaction cascade [20,21]. To investigate a similar mechanism in endothelial cells, MAP kinase stimulation on ouabain exposure was investigated by Western blotting using an antibody against the phos- phorylated form of p44/p42 MAP kinase. Using this method, MAP kinase activation was clearly detectable in HUAECs after 30 min of stimulation (Fig. 4A). The increase in the phosphorylated MAP kinase forms is not due to an increase in overall MAP kinase in the various preparations, as shown by an antibody to MAP kinase that does not distinguish between phosphorylated and unphos- phorylated forms of the enzyme (Fig. 4B). Notably, MAP kinase stimulation was detectable at the low ouabain concentration of 1 n M . Measured over the ouabain concentration range 1 n M to 1 l M , stimulation was 2 to 2.5-fold over control (Fig. 4C). Delayed effects Cell proliferation in response to ouabain. Ouabain has been described as a mitogen that induces cell proliferation in the upper micromolar and low millimolar range. After 12 h of incubation, the cell number of HUAECs increased linearly with increasing ouabain concentrations (from 0.1 to 5n M ), peaking at 5 n M (Fig. 5). At this concentration, the relative increase was 49 ± 4% (P<0.05) above the proliferation observed in the absence of the glycoside. Cell Fig. 3. Stimulation of Rb + uptake by ouabain and effects of various inhibitors. The accumulation of 86 Rb + during 1 h of treatment of HUAECs was measured as described under Experimental procedures. (A) Cells were treated during the 86 Rb + incubation with the indicated concentrations of ouabain or control buffer. (B) Preincubation of HUAECs with 5 l M tetrodotoxin (TTX), 1 l M Ro-31-8425, 500 l M 5-hydroxydecanoate (5-HDA) 6 ,or50l M bumetanide was carried out for 15 min before addition of control buffer or 1 or 10 n M ouabain. The two experiments (A and B) were carried out with primary cell cultures prepared from two different umbilical cords. While the stimulation of up to 50% above background was only observed in the series of experiments shown here (A), most commonly the stimulation of 86 Rb + uptake ranged between 15% and 20% above control, as shown in B (bars represent ± SEM, n ¼ 3–6; *P <0.05). 1058 R. Saunders and G. Scheiner-Bobis (Eur. J. Biochem. 271) Ó FEBS 2004 numbers declined in response to higher ouabain concentra- tions, such that 20 n M ouabain inhibited proliferation relative to that of untreated cells (14%). Effects of ouabain on ET-1 mRNA. The human ET-1 gene promoter region contains two active transcription regula- tory sites: an AP-1-binding site (TGACTAA) at )108 to )102 bp from transcription initiation and a GATA-2- binding site (TTATCT) at )136 to )131 bp [22]. As signal- transduction pathways that activate MAP kinase also activate various genes through these promoters, it was important to address the question of whether ouabain could cause a long-term upregulation in ET-1 mRNA biosyn- thesis. As shown in Fig. 6, ouabain at 10 n M stimulated ET-1 mRNA concentrations in HUAECs by 19.5 ± 3.1% (mean ± SEM; n ¼ 8) after 12 h, as determined by semiquantitative RT-PCR. Discussion Cardiac steroids have a positive inotropic effect on the heart muscle. The mechanism is thought to involve inhibition of the sodium pump, which results in a reduction in the sodium gradient. This in turn has an impact on the transport activity of the Na + /Ca 2+ exchanger, which does not transport Ca 2+ ions out of the cytosol as effectively. Thus, the resulting increased cellular concentration of Ca 2+ is thought Fig. 5. Effect of ouabain on HUAEC cell number. Cells growing on 12-well plates were treated with the indicated concentrations of oua- bainfor12h,afterwhichcellnumberwasassessedasdescribedunder Experimental procedures (bars represent ± SEM; n ¼ 6). Fig. 6. Effect of 12 h exposure to various ouabain concentrations on ET-1 mRNA in HUAECs. HUAECs were incubated for 12 h with the various concentrations of ouabain shown. Then, mRNA isolated from HUAECs was transcribed into cDNA by a reverse transcriptase step, and, by using ET-1-specific primers, the abundance of the latter was analyzed by semiquantitative PCR (bars represent ± SEM; n ¼ 8; *P<0.05). Fig. 4. MAP kinase activation of endothelial cells by ouabain. (A) Cells were incubated with the ouabain concentrations shown for 30 min as described under Experimental procedures. Thereafter, 20 lg protein was separated by SDS/PAGE and probed by an antibody to phospho- p44/42. The resulting signal shown was obtained by the luminescent ECL system. A positive control, commercially available phospho-p44/ 42, was run in parallel (lane 5). Relative amounts of phosphoproteins were analyzed with a digital documentation system and a gel image analysis software. (B) The conditions were the same as in (A) except an antibody to the total (nonphosphorylated) MAP kinase was used. (C) Relative amounts of MAP kinase activated by 1 n M ouabain. Data are derived from experiments similar to those described in (A) (bars represent ± SEM; n ¼ 5). Ó FEBS 2004 Ouabain stimulation of ET-1 release and expression (Eur. J. Biochem. 271) 1059 to stimulate the contractile elements of the heart or vascular muscle and increase contractility. Although thus far plausible, the model implies that the increase in contractility by cardiac steroids is associated with inhibition of the pump and that the rise in cytosolic Na + concentrations occurs before the rise in cytosolic Ca 2+ . Several investigations, however, do not appear to support this mechanism. Not only is inhibition of the pump not a prerequisite for the positive inotropic effect [23], but ouabain was shown to stimulate Ca 2+ transients in arterial smooth muscle without raising cytosolic [Na + ][24]. Our observations shown in Fig. 3 are consistent with these results. 86 Rb + uptake by HUAECs was not inhibited by ouabain, but at 1 n M or 10 n M concentrations of the steroid, the uptake was stimulated by at least 15% (Fig. 3B) and in some cases up to 50% over control uptake (Fig. 3A). Although the molecular basis for this observation is not yet clear, effects of ouabain at low concentrations that do not correlate with sodium pump inhibition have been described in several investigations, including induction of positive inotropic effects [23] or cytosolic [Ca 2+ ] elevation [24]. The stimulation of Rb + uptake that we observed, however, was not simply the result of cell swelling (data not shown) or sodium pump recruitment to the plasma membrane from cytosolic stores. It was also not affected by the Na + /K + /2Cl – cotransporter inhibitor bumetanide or by the mitoK(ATP) channel-specific blocker 5-hydroxydecanoate. As these two major K + uptake systems are not involved in the observed stimulation of Rb + uptake, the sodium pump is the most likely route for 86 Rb + uptake. Stimulation of the sodium pump by protein kinase C [25,26], however, can be excluded, as no effect was observed in the presence of the protein kinase C-specific inhibitor Ro-31-8425. As a direct activation of the sodium pump at very low concentrations of ouabain has been shown in a recent investigation [27], this possibility is currently the most likely explanation for the ability of ouabain to stimulate 86 Rb + uptake. Several reports [20,21,28,29] show that ouabain induces signaling cascades, resulting in both nongenomic and genomic effects. The most apparent of the nongenomic effects described thus far are the rise in cytosolic Ca 2+ concentration associated with slow Ca 2+ oscillations, activation of NF-jB, activation of ERK1/2 (MAP kinase p42/p44) by interactions of the sodium pump with the epidermal growth factor (EGF) receptor, or the release of reactive oxygen species from mitochondria [28]. Consistent with these results, we determined that ouabain at low concentrations stimulates MAP kinase p42/p44 phosphorylation (activation) by 2 to 2.5-fold. Within the same concentration and time range, it also induces Ca 2+ oscillations of an approximate frequency of 8 min and a duration lasting for more than 40 min (until the experiment was terminated; Fig. 1). This phenomenon, which was first described for rat renal proximal tubule cells [16], is clearly confirmed here for human endothelial cells. Whether such Ca 2+ oscillations are also responsible for the induction of the positive inotropic effect on the heart has still to be investigated. In smooth muscle cells of rat mesenteric arteries, Ca 2+ transients were induced at low (3–100 n M ) concentrations of ouabain [24]. Here, however, it was thought that these effects are mediated by the a3 isoform of the sodium pump a subunit, which binds ouabain with high affinity. Nevertheless, a3 and a2 mRNA were not detectable in HUAECs by RT-PCR (not shown), leading to the con- clusion that in the experiments described here the abundant a1 isoform is the most likely mediator of the effects of ouabain at low concentrations. HUAECs, like all endothelial cells, regulate the muscular tone of the underlying arteries by releasing either the vasoconstrictive ET-1 or the vasorelaxant NO. Taking into consideration the fact that ouabain in animal models applied over a long period of time causes hypertension [30], we investigated here its effects on ET-1 release and expression in HUAECs. ET-1, the only member of the endothelin peptide family produced by endothelial cells, is stored in vesicles before it is released towards the underlying smooth muscle cells of the artery. As this substance is a potent vasoconstrictor and growth promoter of vascular smooth muscle cells via the ET A receptor, and because vesicular fusion and release has been shown in numerous cases to be triggered by Ca 2+ ,we were interested in investigating whether ET-1 release from HUAECs to the medium might be stimulated by ouabain. Indeed, a few minutes after application, ouabain causes ET-1 release into the medium (Fig. 2A). This release was depending on extracellular Ca 2+ , as its absence, or the presence of Ca 2+ channel inhibitors such as Ni 2+ or Cd 2+ and the Na + /Ca 2+ -exchanger blocker 2¢,4¢-dichlorobenz- amil considerably reduced ET-1 release (Fig. 2B). This finding is in good agreement with earlier reports showing that ET-1 release depends on extracellular Ca 2+ [31,32]. We do not know yet, however, whether or not ouabain stimulates ET-1 secretion also in vivo. It is possible that ouabain – should it indeed be an endogenously produced hormone, as proposed by many – influences vascular tone by such a mechanism and regulates blood pressure. Elevated ET-1 production has been shown in human vessels subjec- ted to increased pressure and shear stress [33], and increased levels of circulating ET-1 are associated with pulmonary and essential hypertension [34,35]. Besides these immediate, nongenomic effects, ouabain has been shown previously to increase mitotic activity [36–38], a result complemented by the newer findings of ouabain-induced MAP kinase activation and induction of Ca 2+ oscillations followed by the translocation of the transcriptional factor NF-jB into the nucleus [16]. In agreement with these results, we also showed that ouabain at low concentrations stimulates the growth of HUAECs by 49 ± 4% as measured by cell number (Fig. 5). This stimulation was observed after 12 h of incubation. At the same time, mRNA coding for ET-1 is also increased, possibly because of ouabain stimulation of MAP kinase, which is known to stimulate the Fos and Jun transcription factors to form activator protein-1 (AP-1) heterodimers. As the human ET-1 gene promoter region contains an AP-1-binding site regulating transcription [22], this may be responsible for the apparent increased ET-1 gene transcrip- tion we observed. In addition, the bovine ET-1 gene promoter region is also known to contain an NF-jB- responsive region [39]. If this site were also to be present in the human ET-1 promoter region, the Ca 2+ oscillations we show to be induced by ouabain may also contribute to the observed ET-1 upregulation. Although the increase in ET-1 mRNA concentrations appears to be rather modest 1060 R. Saunders and G. Scheiner-Bobis (Eur. J. Biochem. 271) Ó FEBS 2004 (19.5 ± 3.1%), ET-1 in the serum of hypertensive patients is no more than 13% above normal [40]. Thus, an increase in ET-1 by this margin could be one of the reasons for the observed induction of hypertension by ouabain in animal experiments [30]. In conclusion, the results presented here clearly show that ouabain can act in a hormone-like manner on endothelial cells. It becomes effective at concentrations that are within the same range as the effective concentrations of other hormones, including the steroid hormones, and, like the latter, it also induces both genomic and nongenomic effects that are independent of its ability to inhibit the sodium pump at higher concentrations. Assuming that ouabain might be endogenously produced, as has been shown in some instances [9,41], it could act as a hormone involved in blood pressure regulation by fine-tuning and controlling ET-1 release and expression. Acknowledgements R. S. was supported through the Giessen Graduiertenkolleg Molekulare Veterina ¨ rmedizin; G. S. B. is supported through a grant from the Deutsche Forschungsgemeinschaft (DFG), SCHE 307/5-1. References 1. Hansen, O. (1984) Interaction of cardiac glycosides with (Na + ,K + )-activated ATPase. A biochemical link to digitalis- induced inotropy. Pharmacol. Rev. 36, 143–163. 2. Schwinger, R.H., Bundgaard, H., Muller-Ehmsen, J. & Kjeldsen, K. (2003) The Na,K-ATPase in the failing human heart. Cardio- vasc. Res. 57, 913–920. 3. Mathews, W.R., DuCharme, D.W., Hamlyn, J.M., Harris, D.W., Mandel, F., Clark, M.A. & Ludens, J.H. (1991) Mass spectral characterization of an endogenous digitalislike factor from human plasma. Hypertension 17, 930–935. 4. Schneider, R., Wray, V., Nimtz, M., Lehmann, W.D., Kirch, U., Antolovic, R. & Schoner, W. (1998) Bovine adrenals contain, in addition to ouabain, a second inhibitor of the sodium pump. J. Biol. Chem. 273, 784–792. 5. Hamlyn, J.M., Blaustein, M.P., Bova, S., DuCharme, D.W., Harris, D.W., Mandel, F., Mathews, W.R. & Ludens, J.H. (1991) Identification and characterization of a ouabain-like compound from human plasma. Proc. Natl Acad. Sci. USA 88, 6259–6263. 6. Kawamura, A., Guo, J., Itagaki, Y., Bell, C., Wang, Y., Haupert, G.T. Jr, Magil, S., Gallagher, R.T., Berova, N. & Nakanishi, K. (1999) On the structure of endogenous ouabain. Proc.NatlAcad. Sci. USA 96, 6654–6659. 7. Goto, A. & Yamada, K. (1998) Purification of endogenous digi- talis-like factors from normal human urine. Clin.Exp.Hypertens. 20, 551–556. 8. Goto, A., Ishiguro, T., Yamada, K., Ishii, M., Yoshioka, M., Eguchi, C., Shimora, M. & Sugimoto, T. (1990) Isolation of a urinary digitalis-like factor indistinguishable from digoxin. Biochem. Biophys. Res. Commun. 173, 1093–1101. 9. Schneider, R., Antolovic, R., Kost, H., Sich, B., Kirch, U., Tepel, M., Zidek, W. & Schoner, W. (1998) Proscillaridin A immuno- reactivity: its purification, transport in blood by a specific binding protein and its correlation with blood pressure. Clin. Exp. Hypertens. 20, 593–599. 10. Lichtstein, D., Gati, I., Samuelov, S., Berson, D., Rozenman, Y., Landau, L. & Deutsch, J. (1993) Identification of digitalis-like compounds in human cataractous lenses. Eur. J. Biochem. 216, 261–268. 11. Bagrov, A.Y., Fedorova, O.V., Dmitrieva, R.I., Howald, W.N., Hunter, A.P., Kuznetsova, E.A. & Shpen, V.M. (1998) Char- acterization of a urinary bufodienolide Na + ,K + -ATPase inhi- bitor in patients after acute myocardial infarction. Hypertension 31, 1097–1103. 12. Schoner, W., Bauer, N., Mu ¨ ller-Ehmsen, J., Kra ¨ mer, U., Hambarchian, N., Schwinger, R., Mo ¨ ller, H., Kost, H., Weitk- amp, C., Schweitzer, T., Kirch, U., Neu, H. & Gru ¨ nbaum, E H. (2003) Ouabain as a mammalian hormone. Ann. NY Acad. Sci. 986, 1–7. 13. Lowry, O.H., Rosenbrough, N.J., Farr, A.L. & Randall, R.J. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem. 193, 265–275. 14. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 15. Jager, H., Wozniak, G., Akinturk, I.H., Hehrlein, F.W. & Schei- ner-Bobis, G. (2001) Expression of sodium pump isoforms and other sodium or calcium ion transporters in the heart of hyper- tensive patients. Biochim. Biophys. Acta 1513, 149–159. 16. Aizman, O., Uhlen, P., Lal, M., Brismar, H. & Aperia, A. (2001) Ouabain, a steroid hormone that signals with slow calcium oscil- lations. Proc. Natl Acad. Sci. USA 98, 13420–13424. 17. Nixon, J.S., Bishop, J., Bradshaw, D., Davis, P.D., Hill, C.H., Elliott, L.H., Kumar, H., Lawton, G., Lewis, E.J., Mulqueen, M., 5 Westmacott, D., Wadsworth, J. & Wilkinson, S.E. (1991) Novel, potent and selective inhibitors of protein kinase C show oral anti- inflammatory activity. Drugs Exp. Clin. Res. 17, 389–393. 18. Jaburek, M., Yarov-Yarovoy, V., Paucek, P. & Garlid, K.D. (1998) State-dependent inhibition of the mitochondrial KATP channel by glyburide and 5-hydroxydecanoate. J. Biol. Chem. 273, 13578–13582. 19. Haas, M. & Forbush, B. 3rd (1998) The Na-K-Cl cotransporters. J. Bioenerg. Biomembr. 30, 161–172. 20. Kometiani, P., Li, J., Gnudi, L., Kahn, B.B., Askari, A. & Xie, Z. (1998) Multiple signal transduction pathways link Na + /K + - ATPase to growth-related genes in cardiac myocytes. The roles of Ras and mitogen-activated protein kinases. J. Biol. Chem. 273, 15249–15256. 21. Scheiner-Bobis, G. & Schoner, W. (2001) A fresh facet for ouabain action. Nat. Med. 7, 1288–1289. 22. Kawana,M.,Lee,M.E.,Quertermous,E.E.&Quertermous,T. (1995) Cooperative interaction of GATA-2 and AP1 regulates transcription of the endothelin-1 gene. Mol. Cell Biol. 15, 4225– 4231. 23. Godfraind, T. & Ghysel-Burton, J. (1980) Independence of the positive inotropic effect of ouabain from the inhibition of the heart Na + /K + pump. Proc.NatlAcad.Sci.USA77, 3067–3069. 24. Arnon, A., Hamlyn, J.M. & Blaustein, M.P. (2000) Ouabain augments Ca 2+ transients in arterial smooth muscle without raising cytosolic Na + . Am. J. Physiol. Heart Circ. Physiol. 279, H679–H691. 25. Pedemonte, C.H., Pressley, T.A., Lokhandwala, M.F. & Cinelli, A.R. (1997) Regulation of Na,K-ATPase transport activity by protein kinase C. J. Membr. Biol. 155, 219–227. 26. Vasilets, L.A., Fotis, H. & Gartner, E.M. (1997) Regulatory phosphorylation of the Na + /K + -ATPase from mammalian kid- neys and Xenopus oocytes by protein kinases. Characterization of the phosphorylation site for PKC. Ann. NY Acad. Sci. 834, 585–587. 27. Gao, J., Wymore, R.S., Wang, Y., Gaudette, G.R., Krukenkamp, I.B., Cohen, I.S. & Mathias, R.T. (2002) Isoform-specific stimu- lation of cardiac Na/K pumps by nanomolar concentrations of glycosides. J. Gen. Physiol. 119, 297–312. 28. Xie, Z. & Askari, A. (2002) Na + /K + -ATPase as a signal trans- ducer. Eur. J. Biochem. 269, 2434–2439. Ó FEBS 2004 Ouabain stimulation of ET-1 release and expression (Eur. J. Biochem. 271) 1061 29. Nakagawa, Y., Rivera, V. & Larner, A.C. (1992) A role for the Na/K-ATPase in the control of human c-fos and c-jun tran- scription. J. Biol. Chem. 267, 8785–8788. 30. Manunta, P., Rogowski, A.C., Hamilton, B.P. & Hamlyn, J.M. (1994) Ouabain-induced hypertension in the rat: relationships among plasma and tissue ouabain and blood pressure. J. Hypertens. 12, 549–560. 31. Brunner, F., Stessel, H. & Graier, W.F. (1995) The effects of intracellular Ca 2+ concentration and hypoxia on basal endothe- lin-1 secretion by cultured porcine aortic endothelial cells. Agents Actions Supplement 45, 269–273. 32. Carlini, R.G., Gupta, A., Liapis, H. & Rothstein, M. (1995) Endothelin-1 release by erythropoietin involves calcium signaling in endothelial cells. J. Cardiovasc Pharmacol. 26, 889–892. 33. Gan, L.M., Selin-Sjogren, L., Doroudi, R. & Jern, S. (2000) Temporal regulation of endothelial ET-1 and eNOS expression in intact human conduit vessels exposed to different intraluminal pressure levels at physiological shear stress. Cardiovasc. Res. 48, 168–177. 34. Li-Saw-Hee, F.L., Blann, A.D. & Lip, G.Y. (1997) The detection of endothelial dysfunction in patients with essential hypertension. Int. J. Cardiol. 61, 171–174. 35. Veyssier-Belot, C. & Cacoub, P. (1999) Role of endothelial and smooth muscle cells in the physiopathology and treatment management of pulmonary hypertension. Cardiovasc. Res. 44, 274–282. 36. Murata, Y., Matsuda, T., Tamada, K., Hosoi, R., Asano, S., Takuma, K., Tanaka, K. & Baba, A. (1996) Ouabain-induced cell proliferationinculturedratastrocytes.Jpn. J. Pharmacol. 72, 347– 353. 37. Aydemir-Koksoy, A. & Allen, J.C. (2001) Low concentrations of ouabain induce vascular smooth muscle cell proliferation. Cell Mol. Biol. 47, 341–345. 38. Chueh, S.C., Guh, J.H., Chen, J., Lai, M.K. & Teng, C.M. (2001) Dual effects of ouabain on the regulation of proliferation and apoptosis in human prostatic smooth muscle cells. J. Urol. 166, 347–353. 39. Quehenberger, P., Bierhaus, A., Fasching, P., Muellner, C., Klevesath,M.,Hong,M.,Stier,G.,Sattler,M.,Schleicher,E., Speiser, W. & Nawroth, P.P. (2000) Endothelin 1 transcription is controlled by nuclear factor-kappaB in AGE-stimulated cultured endothelial cells. Diabetes 49, 1561–1570. 40. Ooi, B.S., Papademetrious, V. & Cohen, D.J. (1996) Demonstra- tion of endothelial-activating properties of hypertensive sera. Am. J. Hypertens. 9, 1232–1235. 41. Hamlyn, J.M., Lu, Z.R., Manunta, P., Ludens, J.H., Kimura, K., Shah, J.R., Laredo, J., Hamilton, J.P., Hamilton, M.J. & Hamilton, B.P. (1998) Observations on the nature, biosynthesis, secretion and significance of endogenous ouabain. Clin. Exp. Hypertens. 20, 523–533. 1062 R. Saunders and G. Scheiner-Bobis (Eur. J. Biochem. 271) Ó FEBS 2004 . Ouabain stimulates endothelin release and expression in human endothelial cells without inhibiting the sodium pump Robert Saunders and Georgios Scheiner-Bobis Institut fu ¨ r. l M Na + /K + /2Cl – cotransporter inhibitor bumetanide. Binding of ouabain to sodium pumps on the surface of the HUAEC plasma membrane This experiment was performed to investigate whether ouabain treatment in uences the sodium. determining whether ouabain added to endothelial cells in culture might induce such a response. Various concentrations of ouabain were added to the cell culture wells, and, after 10 min of incubation,