Oxygen-induced regulation of Na,K-ATPase was studied in rat myocardium. In rat heart, Na,K-ATPase responded to hypoxia with a dose-dependent inhibition in hydrolytic activity. Inhibition of Na,K-ATPase in hypoxic rat heart was associated with decrease in nitric oxide (NO) production and progressive oxidative stress. Accumulation of oxidized glutathione (GSSG) and decrease in NO availability in hypoxic rat heart were followed by a decrease in S-nitrosylation and upregulation of S-glutathionylation of the catalytic α-subunit of the Na,K-ATPase. Induction of S-glutathionylation of the α-subunit by treatment of tissue homogenate with GSSG resulted in complete inhibition of the enzyme in rat a myocardial tissue homogenate. Inhibitory effect of GSSG in rat sarcolemma could be significantly decreased upon activation of NO synthases. We have further tested whether oxidative stress and suppression of the Na,K-ATPase activity are observed in hypoxic heart of two subterranean hypoxia-tolerant blind mole species (Spalax galili and Spalax judaei). In both hypoxia-tolerant Spalax species activity of the enzyme and tissue redox state were maintained under hypoxic conditions. However, localization of cysteines within the catalytic subunit of the Na,K-ATPase was preserved and induction of S-glutathionylation by GSSG in tissue homogenate inhibited the Spalax ATPase as efficiently as in rat heart. The obtained data indicate that oxygen-induced regulation of the Na,K-ATPase in the heart is mediated by a switch between S-glutathionylation and S-nitrosylation of the regulatory thiol groups localized at the catalytic subunit of the enzyme.
- redox stat
local or global, hypoxia causes acute changes in the myocardial function. Hypoxic responses include adjustments of cardiac output mediated by vagal control, humoral factors, and by autonomous heart oxygen sensors. At the cellular level, these responses are mirrored by coordinated regulation of electric, mechanical, and metabolic processes. Thereby decrease in energy demand is achieved to match the changes in ATP supply (19).
Oxygen-induced regulation of the Na,K-ATPase activity in cardiomyocytes is of particular importance for successful adaptation to hypoxia. Responsible for about 20% of total ATP expenditure in the heart (51), this enzyme sustains the transmembrane Na+/K+ gradients that are used for action potential generation. Furthermore, Na,K-ATPase actively participates in regulation of contractile force as it is functionally coupled to the Na+/Ca2+ exchanger (57, 62). Inhibitors of Na,K-ATPase, known as cardiac glycosides, have been used to treat congestive heart failure for over 200 years (50). Suppression of the Na,K-ATPase activity reported in hypoxic mammalian heart is a result of several converging processes at systemic, organ, and cellular levels. Systemic hypoxia stimulates release of endogenous inhibitors of Na,K-ATPase, ouabain-like factors, into the circulation (14). Hypoxia-induced ATP depletion may also contribute to enzyme inactivation at later stages. However, induction of hypoxic response of Na,K-ATPase occurs long before intracellular ATP drops to the submillimolar levels, which may result in shortage of substrate supply (23, 29, 64). Indirect evidence suggests that the enzyme responds not to the changes in O2 availability but to resulting shifts in the tissue redox state and nitric oxide (NO) production (10, 11, 47). Labile cytosolic inhibitory factor causing the enzyme inactivation in ischemic rat heart was reported to be redox sensitive (23). Apart from hypoxia or ischemia, loading of cells with reduced glutathione or using glutathione depleting agents was shown to cause alterations in Na,K-ATPase activity (45). Accumulating evidence links hypoxia-induced changes in Na,K-ATPase function to the changes in NO production (9, 20, 44). In ischemic myocardium Na,K-ATPase activity could be rescued by NO donors (61). Na,K-ATPase in the hearts of transgenic mice lacking NO synthases (eNOS−/−, nNOS, and double eNOS-nNOS knockout animals) is suppressed (63).
These single observations did not provide a molecular mechanism of redox- or oxygen-induced regulation of the enzyme until more recently. Oxidative thiol modifications (S-nitrosylation and S-glutathionylation) of Cys46 in the β-subunit of Na,K-ATPase were recently reported to cause a 20% decrease in the enzyme activity upon exposure of isolated cardiomyocytes to hypoxia (8, 20). S-nitrosylation was shown to be necessary as an intermediate step in the induction of regulatory S-glutathionylation of the regulatory β-subunit of the enzyme. However, inhibition of the Na,K-ATPase in ischemic heart was much more extensive than that shown in cardiomyocytes exposed to low oxygen in which S-glutathionylation of the β-subunit was shown (23, 61).
At present there are no reports in the literature on hypoxia-induced reversible thiol modifications in the catalytic α-subunit. The rodent Na,K-ATPase α-subunit contains over 20 cysteine residues, most of which are localized in cytosolic loops forming nucleotide- and ion-binding sites (10). Localization of cysteine residues within the protein sequence is largely conserved, suggesting that they play an important role in enzyme function. Our recent findings indicated that S-glutathionylation of the cysteine residues of large and small cytosolic loops of the Na,K-ATPase α-subunit occurs when the isolated enzyme is incubated with oxidized glutathione (46). Interaction of glutathione with several cysteine residues of the large cytosolic loop occurs only under conditions of mild ATP deprivation. Upon S-glutathionylation of these few cysteines, the ATP binding site of the enzyme becomes inaccessible for ATP and the enzyme is completely inhibited (46). Deglutathionylation occurring in the presence of glutaredoxin and NADPH is associated with restoration of the enzyme activity (46).
The present study was designed to explore the possible interplay between S-nitrosylation and S-glutathionylation in control of function of Na,K-ATPase in hypoxic myocardium and its role in the adaptation success of rodents to oxygen deprivation. We have monitored the changes in enzyme activity in rat heart as a function of oxygen availability and related alterations in phospholemman phosphorylation state, and S-nitrosylation and S-glutathionylation of cysteine residues in the Na,K-ATPase catalytic subunit. The shifts in tissue reduced and oxidized glutathione and nitrite production associated with reversible thiol modifications were assessed. Furthermore, we have compared oxygen-sensitivity of Na,K-ATPase in the myocardium of hypoxia-sensitive Wistar rats (Rattus norvegicus) with that of two species of the blind subterranean mole rat (Spalax ehrenbergi): Spalax galili and Spalax judaei. These subterranean rodents share suborder (Myomorpha) with rats, but show profound hypoxia-tolerance (41, 42).
Our results revealed that Na,K-ATPase in Wistar rat myocardium shows a profound sensitivity to oxygen availability. A switch from S-nitrosylation to S-glutathionylation of cysteine residues within the catalytic α-subunit plays a key role in oxygen-induced regulation of the Na,K-ATPase in Wistar rat heart. In parallel, hypoxia induces a shift in phosphorylation of phospholemman. Na,K-ATPase function in Spalax myocardium is preserved during hypoxic exposure. Lack of inhibitory response of Na,K-ATPase to hypoxia in Spalax heart was associated with the absence of hypoxia-induced changes in redox state.
MATERIALS AND METHODS
Rat myocardial tissue isolation and handling.
Animal handling and experimentation was approved by the Swiss Federal Veterinary Office and performed in accordance with Swiss animal protection laws and institutional guidelines that comply with those of the Institute for Laboratory Animal Research.
Heparin (100 μl of 10,000 units/ml heparin; Braun, Grenchen, Switzerland) was injected into the caudal vena cava of anesthetized (3% isoflurane in a 1:1 mixture of O2 and N2O) Wistar male rats (300–400 g) purchased from Janvier (Le Genest, St. Isle, France) via abdominal incision, and 8–10 ml of blood were collected. The heart was excised and cooled down in heparin-containing isotonic phosphate buffer (in mM) containing 120 NaCl, 25 NaHCO3, 1 CaCl2, 0.15 MgCl2, 10 glucose, 0.1 l-arginine, 10 Tris·HCl (pH 7.4).
A perfusion circuit including a hollow fiber mini-oxygenator was filled with the collected heparinized blood, and the heart was mounted on a cannulae and perfused via aorta with autologous blood equilibrated with humidified gas phase (20% O2, 5% CO2, and 75% N2 PanGas; Basel, Switzerland; 37°C) for 20 min (29). After the restitution period, oxygen concentration in gas phase was retained at 20% (normoxia) or readjusted to 15%, 10%, 5%, and 3% depending on scientific protocol and perfused for 40 min. Glucose consumption by erythrocytes and water loss from the organ chamber were compensated for by supplementation of 1.1 mmol/l glucose every 20 min.
Aqueous solution of l-NAME (to a final concentration 300 μM) was administrated to blood in the perfusion circuit after the restitution period, according to protocol.
Aliquots of blood (100 μl) were collected at the end of the perfusion for plasma nitrite and nitrate analysis.
At the end of the blood perfusion protocol, the heart was chilled and perfused with ice-cold sucrose-buffer solution containing 300 mM sucrose in 20 mM of a HEPES·Tris buffer (pH 7.4 at 0°C) to remove blood from coronary vessels. Ventricular tissue was frozen in liquid nitrogen and later analyzed for reduced (GSH) and oxidized (GSSG) glutathione content, Na,K-ATPase activity, and used for preparation of sarcolemmal fraction, immunoblotting and biotin-switch.
Spalax myocardial tissue isolation and handling.
Superspecies Spalax ehrenbergi includes four distinct biological species and is located in Israel (42). The most extreme differences in ecological conditions exist between regions inhabited by Spalax galili in the northern cool-humid Upper Galilee Mountains, and Spalax judaei in the southern warm-dry xeric regions of Israel. Differences in environment are reflected in differential adaptations to hypoxia. An improved hypoxic adaptation of S. galili has been established with higher hematocrit and hemoglobin levels as compared with S. judaei (3, 4). In laboratory experiments the lowest levels of oxygen concentrations tolerated by S. galili (2.6 ± 0.4%) were significantly lower than S. judaei (3.7 ± 0.9%) (5). Using field measurements, we recorded 6% O2 and 7% CO2 in S. ehrenbergi burrows in flooded heavy soils during the Mediterranean rainy season (55).
All animal handling protocols were approved by the Haifa University Committee for Ethics on Animal Subject Research, permit No. 193/10, and approved by the Israel Ministry of Health. Six animals of S. galili and six further animals of S. judaei of both sexes 2–4 years old were captured in the field and housed under ambient conditions in individual cages. Three out of six animals of each Spalax species were randomly assigned into the hypoxic group. These animals were placed in a 70 × 70 × 50 cm chamber divided into separate cells flushed with gas mixture at 5 l/min. Oxygen concentration in the gas mixture was gradually decreased for 45–60 min and finally adjusted at 6% O2. This oxygen concentration was chosen as the lowest oxygen levels recorded in S. ehrenbergi tunnels after rainstorms (55). Animals were exposed to hypoxia for 6 h. Animals in normoxic groups were exposed to ambient air. After hypoxic exposure the animals were administered Ketaset CIII at 5 mg/kg of body weight and heart muscle tissue was harvested and immediately frozen in liquid nitrogen.
Nitrite assessment in blood plasma.
One of the widely used markers for NO production in biological systems is nitrite (NO2−) concentration because its accumulation is proportional to the NO production (28). Blood aliquots collected during perfusion were pelleted by centrifugation (16,000 g), and the supernatant (blood plasma) was collected. Nitrate was reduced to nitrite using Cd coated with Cu and (NO2− + NO3−) assessed by a chemiluminiscence detector CLD 88 (Eco Medics AG, Switzerland) as described elsewhere (39). Briefly, a sample aliquot was deproteinized by addition of methanol (100%) and centrifuged (13,200 g, 10 min). Supernatant (50 μl) was injected into the preheated (65°C) reaction chamber containing acidic triiodide (I3−) reagent. The reagent was prepared fresh before the measurements (1.65 g KI, 0.57 g I2, 15 ml H2O, 200 ml glacial CH3COOH). The reaction chamber was purged with helium and released NO was detected using CLD-88 analyzer (ECO MEDICS, Durnten, Switzerland). The signal was processed using PowerChrom 280 system (eDAQ Pty; Spechbach, Germany).
GSH and GSSG levels in myocardium.
Tissue GSH and GSSG were assessed in blood-free ventricular tissue preparations. Frozen ventricular fragments (0.1 g) were homogenized on ice in KCl-MOPS buffer (100 mM KCl and 10 mM MOPS, pH 7.4) and deproteinized with 5% trichloracetic acid. GSH and GSSG were assessed in the protein-free supernatant using Ellmann's reagent and normalized to sample wet weight (58). Briefly, GSH concentration was determined in deproteinized samples using 5,5′-dithiobis (2-nitrobenzoic acid) (Ellman's reagent). Optical density of the colored complex was measured photometrically at 412 nm. Simultaneously, aliquots from the same samples were incubated in the presence of glutathione reductase and NADPH for reduction of GSSG to GSH, and total glutathione levels (GSH + GSSG) were determined. The half-cell reduction potential (Ehc) was then calculated for GSH/GSSG couple (54).
Isolation of sarcolemmal fraction from rat myocardium.
Sarcolemmal fraction of rat myocardium was performed as described elsewhere (27). Briefly, the ventricle tissue samples were washed, minced, and homogenized in 0.6 M sucrose and 10 mM imidazole-HCl (pH 7.0, 3.5 ml/g tissue) with a Polytron PT-3000 (9,000 rpm). The resulting homogenate was centrifuged at 12,000 g for 30 min, and the pellet was discarded. After dilution (5 ml/g tissue) with 140 mM KCl-MOPS buffer containing 140 mM KCl and 10 mM MOPS (pH 7.4), the supernatant was centrifuged (95,000 g, 60 min). The resulting pellet was resuspended in the KCl-MOPS buffer and layered over a 30% sucrose solution, 0.3 M KCl with 50 mM Na4P207 in 0.1 M Tris·HCl (pH 8.3). After centrifugation (95,000 g, 90 min), the band at the sucrose-buffer interface was taken and diluted with 3 vol of KCl-MOPS solution. A final centrifugation (95,000 g, 30 min) resulted in a pellet rich in sarcolemma. The pellet was resuspended in KCl-MOPS buffer.
Protein concentration was assayed in all experiments by the Lowry method (34).
Ouabain-sensitive (1 mM) inorganic phosphate production (49) was used to assess Na,K-ATPase hydrolytic activity. Activity of the enzyme was measured in the presence of saturating concentrations of ligands and substrate of (in mM) 130 NaCl, 20 KCl, 3 MgCl2, 3 ATP, and 30 imidazole (pH 7.30 at 37°C) as described elsewhere (45). Briefly, protein samples were mixed with media containing (in mM) 130 NaCl, 20 KCl, 3 MgCl2, and 30 ATP with or without 1 ouabain according to the experimental protocol at 37°C for 10 min. The reaction was then stopped by ice-cold 4% formaldehyde in 1.3 M sodium acetate buffer (pH 4.3). Accumulated inorganic phosphate was determined using the Rathbun-Betlach method (49).
In vitro treatment of isolated sarcolemmal fraction with oxidized glutathione.
The inhibitory action of GSSG on the Na,K-ATPase activity was assessed on sarcolemal fraction and ventricular homogenates of all species used in this study. The sarcolemmal membrane fraction and ventricular tissue homogenate were exposed to 300 μM GSSG for 10 min (37°C). Na,K-ATPase activity was then assessed as described above.
In vitro activation of nitric oxide synthase.
Substrates and cofactors for NOS were administrated to rat sarcolemmal fractions to a final concentration of 0.1 mM l-arginine, 0.5 mM CaCl2, 10 μg/ml calmoduline, 0.2 mM NADPH, 10 μM biopterin, 5 μM FMN, and 5 μM FAD, and the samples were incubated for 10 min at room temperature. After incubation the sample was used in the Na,K-ATPase activity assay with or without treatment with GSSG as described above. Nitrite accumulation in the incubation medium was used as a marker of NO generation.
Western blot analysis.
Tissue homogenates and isolated sarcolemmal fraction from the in vitro studies were separated by 7.5%, 10%, and 12.5% SDS-PAGE with Tris-glycine (for resolution of high molecular weight proteins) or Tris-tricine (for resolution of low molecular weight proteins) buffer systems and transferred to Protean BA83 nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany). Protein transfer was controlled by Ponceau red staining. Membranes were blocked for 1 h at room temperature and incubated overnight at 4°C with the appropriate primary antibodies.
Primary antibodies kindly provided by M. J. Shattok and W. Fuller (35) were used to assess Ser68 phosphorylation of the FXYD1 subunit of Na,K-ATPase (phospholemman) and determine total and phosphorylated forms. The S-glutathionylation state of the α1-subunits of Na,K-ATPase in sarcolemmal preparations, as well as in crude ventricular homogenates, was assessed using monoclonal anti-glutathione antibody (Chemicon Millipore, MAB5310). The membranes were then stripped and mouse monoclonal anti-Na,K-ATPase α1-antibody clone C464-6 (Upstate Millipore) were applied for detection of the total amount of α1-subunit. Appropriate horseradish peroxidase-conjugated secondary antibodies were applied and enhanced chemiluminescent detection system (Fujifilm LAS-3000 System; Fujifilm Life Science) was used for detection. Antinitotyrosine antobodies (Antinitrotyrosine, clone 1A6, HRP conjugate from Upstate; cat No. 16-207; 1:1,000 dilution) were used to assess tyrosine nitration. As a negative control for the staining nitrocellulose membrane after the transfer was incubated for 1 min (room temperature) in 100 mM Na2S2O4 (in 100mM Na2B4O7, pH 9). Afterward it was washed and probed with antinitrotyrosine antibody according to the protocol. No staining was detected after Na2S2O4 pretreatment indicating that the signal from the antibodies was specific for nitro-tyrosine. ImageJ software was used for quantification of the recorded signals. Total-actin (Sigma; A2066) was used as a loading control.
S-nitrosylation of the α1-subunit of Na,K-ATPase in the rat ventricular homogenate was assessed using the S-nitrosylated Protein Detection Kit (Cayman Chemical Ann Arbor, MI). Briefly, in the first step free thiols are blocked by incubation with the thiol-specific methylthiolating agent methyl methanethiosulfonate (MMTS). After block of free thiols, nitrosothiol bonds are selectively decomposed with ascorbate, which results in the reduction of nitrosothiols to thiols. In the last step, the newly formed thiols are reacted with N-[6-(biotinamido)hexyl]-3′-(2′-pyridyldithio)propionamide (biotin-HPDP), a sulfhydryl-specific biotinylating reagent. Labeled proteins can were detected by immunoblotting with antibiotin antibodies, after SDS-PAGE (26). In parallel, we have used anti-S-nitrosocysteine antibody [S-nitroso-Cys (SNO-Cys), Alpha Diagnostic, cat No. NISC11-A, dilution 1:1,000] and were able to reproduce the findings obtained using biotin switch assay.
Oxygen-dependence of Na,K-ATPase function.
Perfusion of isolated rat heart with autologous blood equilibrated with 20%, 15%, 10%, 5%, or 3% oxygen in gas phase for 45 min resulted in decrease of autonomous heart rate to 42 ± 4% of the normoxic value of 254 ± 50 beats/min (n = 16). The time course of the response is reflected by a decrease in heart rate when Po2 in the gas phase was decreased from 20% to 5% or 3% (Table 1 and Fig. 1A). Furthermore, the ECG showed inversion of the T wave (Fig. 1B) characteristic of ischemic myocardium in hearts perfused with blood equilibrated to 5% or 3% O2 (30). Hypoxia-induced bradycardia was associated with arrhythmia in some of the hearts. These responses developed within 1–20 min after decrease of hemoglobin oxygen saturation.
These changes in myocardial function occurred along with a dose-dependent decrease in activity of Na,K-ATPase in ventricular tissue homogenates from 6.4 ± 3.3 µmol Pi/(mgprot·h) in normoxic myocardium to 1.25 ± 0.18 µmol Pi/(mgprot·h) in tissue exposed to 3% O2 (hemoglobin oxygen saturation SO2 of 16%; Fig. 2A).
Decrease in oxygen levels in gas phase to which the blood was equilibrated from 20% to 10%, and the resulting decrease in hemoglobin oxygen saturation SO2 from 95.8% to 85% was associated with a twofold decrease in the activity of Na,K-ATPase.
Similar profound inactivation of the Na,K-ATPase was previously reported to occur in ventricular tissue homogenates prepared from rat heart exposed to no-flow ischemia (22). Inhibition of the enzyme in ischemic tissue homogenate was associated with an increase in Na,K-ATPase activity in sarcolemmal membranes. The stimulatory effect of ischemia at the sarcolemmal membrane level was caused by phosphorylation of the regulatory FXYD1 subunit of the Na,K-ATPase, phospholemman at Ser68 (22). Similar to that in ischemic myocardium, in which hypoxia occurred along with acidosis and aglycemia, exposure of rat heart to hypoxia alone was also associated with phosphorylation of phospholemman (Fig. 2B) and with a dose-dependent increase in Na,K-ATPase activity in sarcolemmal membranes from 16.5 ± 2.7 μmol Pi/(mgprot·h) to 25.0 ± 2.2 μmol Pi/(mgprot·h) (Fig. 2A). Mechanisms of activation of the enzyme in sarcolemmal fraction of ischemic heart and the role of phospholemman have been addressed earlier (22). We have therefore concentrated on characterization of the molecular mechanisms behind the inhibitory action of hypoxia on the Na,K-ATPase in ventricular homogenate. Hypoxic conditions for further experiments were defined as perfusion with blood equilibrated with 5% O2 in gas phase (SO2 = 35%). Hypoxia-induced changes were related to the values obtained in hearts perfused with oxygen-saturated blood (blood equilibrated with gas phase containing 20% O2, SO2 = 95.8%). Exposure of isolated blood-perfused rat heart to hypoxia for 40 min was associated with a sixfold suppression in activity of Na,K-ATPase in ventricular tissue homogenate (Fig. 2A).
Redox state, NO production, and function of Na,K-ATPase in hypoxic rat myocardium.
We have assessed hypoxia-induced changes in redox state and NO production in isolated blood-perfused rat heart. Perfusion of hearts with hypoxic blood for 40 min resulted in a shift in half-cell redox potential for GSH/GSSG couple oxidation and in a decrease in NO2− levels reflecting NO production (Fig. 3). Inhibition in nitric oxide synthase (NOS) activity was, to a large extent, the cause of oxidative stress. Perfusion of hearts with normoxic blood containing 300 μM l-NAME induced oxidation of GSH (expressed as Ehc, a measure of GSH:GSSG), and no further pro-oxidative effect of hypoxia was observed in l-NAME-treated myocardium (Fig. 3). The changes in actual tissue GSH and GSSG concentrations are presented in Table 2. The onset of bradycardia was elevated in the hypoxic hearts treated with l-NAME, whereas l-NAME had little effect on heart rate in the normoxic hearts (Table 1 and Fig. 1). Exposure to l-NAME induced arrhythmia in some of the hearts even under normoxic conditions (Table 1). Oxygen sensitivity of Na,K-ATPase was lost in hearts perfused with blood containing 300 μM l-NAME, in which the enzyme was preserved under hypoxic conditions (Fig. 4). Taken together these data indicated that, apart from gradual ATP deprivation, hypoxia-induced responses of the Na,K-ATPase were closely associated with the changes in GSSG and NO levels.
Thiol modifications of Na,K-ATPase α-subunit are dependent on hypoxic stress and NO production.
We have assessed S-glutathionylation and S-nitrosylation of cysteines in the catalytic α-subunit of the Na,K-ATPase in normoxic and hypoxic myocardium using immunoprecipitation, Western blotting, and Biotin-Switch techniques. Immunoprecipitation was carried out using anti-GSH antibodies and was then followed by Western blotting against α1 and α2 isoforms of the catalytic subunit, revealing the presence of both in immunoprecipitate (data not shown).
The presence of S-glutathionylated thiols in the α1-subunit in crude homogenate prepared from ventricular tissue was confirmed using immunoblotting. Basal S-glutathionylation was detected in normoxic tissue samples (Fig. 5). The number of S-glutathionylated cysteines in the α1-subunit increased in response to l-NAME treatment under normoxic conditions as well as in response to hypoxia without l-NAME. A combination of hypoxic conditions and l-NAME exposure was not associated with an increase in S-glutathionylation of the α1-subunit compared with normoxic control.
S-nitrosylation of the α1-subunit was high in normoxic samples and decreased in response to hypoxia and l-NAME treatment (Fig. 6A). Decrease in the number of S-nitrosylated cysteines in the α1-subunit in the hypoxic heart was accompanied by pronounced increase in nitrated tyrosine residues. Exposure to l-NAME suppressed hypoxia-induced nitrotyrosine formation (Fig. 6B).
We have further explored the effect of S-glutathionylation on Na,K-ATPase activity. Exposure of sarcolemmal membranes isolated from normoxic rat heart to 250 μM GSSG-induced S-glutathionylation of the α1-subunit was associated with a marked suppression of the enzyme hydrolytic activity (Fig. 7). Activation of NO production in suspension of sarcolemmal microsomes was induced by supplementation of NOS substrates and coactivators. NO production was monitored as NO2− accumulation in the incubation medium. Nitrite accumulation was found to be 0.7 nmol/(μg protein·30 min). Calculations based on the assumption of the 2.5% of cysteine occurrence frequency in mammalian protein sequences reported earlier (40) and an average molecular weight of an amino acid of 110 g/mol give an estimation of 11 nmol of cysteines in our samples containing 50 μg of protein. The amount of NO transferred to NO2− and not bound to protein thiols in a sample containing 50 μg protein is 35 nmol. This amount exceeds the number of potential binding sites by 3.5-fold and suggests that NO produced by microvesicles upon activation of NO synthases was sufficient to cause S-nitrosylation of at least some thiol groups. Both S-glutathionylation and inhibition of the enzyme by GSSG were prevented if NOSes were activated at the moment of administration of 250 μM GSSG (Fig. 7).
Lack of oxygen-sensitivity of Na,K-ATPase in spalax heart.
The impact of hypoxic exposure on Na,K-ATPase activity in the heart was explored in hypoxia-tolerant Spalax genus, S. galili and S. judaei, blind fossorial mole rats. Whereas normoxic animals were exposed to ambient air, hypoxic individuals were exposed to hypoxic atmosphere containing 6% O2-94% N2 for 6 h. Activity of Na,K-ATPase was then assessed in ventricular tissue homogenate. The absolute values of enzyme activity as well as the α1 abundance in normoxic Spalax hearts were decreased by approximately threefold compared with that in Wistar rat (R. norvegicus) hearts (Fig. 8). In contrast with that in R. norvegicus hearts, Na,K-ATPase in the myocardium of both Spalax species was not suppressed in response to hypoxic exposure (Fig. 8).
Is Spalax Na,K-ATPase redox-sensitive?
We have further tested if the apparent insensitivity of Spalax myocardial Na,K-ATPase to hypoxia was stemming from the lack of regulatory cysteine residues, which were present in rat enzyme. To do so, the α1-subunit of Na,K-ATPase of Spalax was cloned and sequenced and localization of cysteine residues within the sequence was compared with that of R. norvegicus. Sequence alignment presented as supplementary data set showed strong similarity of both proteins and particularly conserved localization of all cysteine residues within the sequences. The ability of the enzyme to respond to GSSG-induced S-glutathionylation has been confirmed for S. galili and S. judaei. Treatment of crude ventricular tissue homogenates with 300 μM of GSSG for 10 min caused complete inhibition of the enzyme in both Spalax and R. norvegicus species (Fig. 9). Thus all rodents possess redox-sensitive Na,K-ATPase in the heart. However, only R. norvegicus responded to systemic hypoxia (29) or deoxygenation of isolated blood-perfused heart with glutathione oxidation. Neither of the mole rat species tested showed any signs of oxidative stress in the heart in response to hypoxia (Fig. 10, actual tissue GSH and GSSG levels in Table 2). Moreover, Ech (GSH:GSSG) in normoxic myocardium of S. galili and S. judaei was significantly more reduced than that in R. norvegicus hearts.
Our findings indicate that the changes of Na,K-ATPase activity in the heart is very tightly controlled by reversible thiol modifications in the catalytic α-subunit, which occur to even slight changes in oxygen availability. NO availability as well as maintenance of local ATP levels in premembrane space protect regulatory cysteines from inhibition, which follows their S-glutathionylation (46). Accumulation of GSSG and decrease in NO production occur in hypoxic rat myocardium (Figs. 2 and 3) along with gradual ATP deprivation (29) making the enzyme oxygen sensitive. At the molecular level changes in activity of Na,K-ATPase are associated with modifications of protein cysteine residues. Plasticity of the enzyme function appears to be tightly coupled to the reversible transition for SH groups to switch between S-nitrosylated to S-glutathionylated forms. The present study revealed that Na,K-ATPase is one more element contributing to the complex responses of cardiomyocytes to changes in redox state (53). Whether the enzyme will respond to the changes in oxygen supply depends entirely on shifts in redox state and NO production in the heart and ATP availability.
Progressive development of oxidative stress in hypoxic myocardium was suggested to reflect uncoupling in electron transduction in the mitochondria (33). Our data suggest that suppression in NO production in hypoxic heart contributes to this process as well (Figs. 3 and 4). Indeed, oxygen affinities of inducible and neuronal NO sythases (Kd of 7.8 kPa and 5.9 kPa, respectively) are significantly lower than those of superoxide generating enzymes including Kd of ∼2 kPa NOX2 and 4 and even higher affinities of mitochondrial cytochromes (10, 17). NO interacts with superoxide anion four orders of magnitude faster than superoxide dismutase (rate constants being 7·109 vs. 105 M−1·s−1, respectively) (43). The imbalance between O2·− and NO production towards superoxide will result in an increased ONOO− production and accumulation of H2O2 generated in the SOD-catalysed reaction (43). An increase in nitrotyrosine levels along with decrease in NO2− levels in hypoxic rat myocardium indicate that this shift indeed occurs in hypoxic heart (Fig. 6, A and B). Accumulation of GSSG following hypoxic exposure mirrors an increase in H2O2 and ONOO− (Fig. 3A). These processes do not occur in Spalax heart exposed to hypoxia as follows from the lack of changes in Ehc for the GSH/GSSG couple (Fig. 10). Our findings do not allow any speculations on the molecular mechanisms of resistance of Spalax myocardium to hypoxia-driven oxidation. Among the possible contributors are maintenance of NO production (downregulation of NO2− levels (21), high activity of eNOS), lower O2·− production rates, and/or more efficient H2O2 processing enzymes.
Determinants of oxygen sensitivity of Na,K-ATPase.
Multiple processes are known to mediate the changes in activity of Na,K-ATPase in response to hypoxia. Among them are phosphorylation, internalization of the enzyme, and S-glutathionylation of β- and FXYD subunits of the enzyme (8, 10, 20). Some of these regulatory mechanisms function as on-off switches; others mediate fine-tuning of the enzyme resulting in partial suppression or activation. Internalization of the enzyme in clathrin-coated vesicles (12) and complete inhibition of the enzyme caused by S-glutathionylation of the α-subunit (Fig. 9) (46) represent the on-off mechanisms, which dominate over the fine-tuning regulatory mechanisms. In a separate study we have shown that complete inactivation of the Na,K-ATPase is caused by binding of glutathione to a cysteine residue within an adenine nucleotide binding site, making it inaccessible to ATP (46).
Fine-tuning of the enzyme activity in hypoxic tissue is achieved by S-glutathionylation of the β- and FXYD1 (phospholemman) subunits (8, 20) as well as by phosphorylation of phospholemman at Ser68 (Fig. 2) (22, 23, 56), resulting in modest suppression or stimulation of Na,K-ATPase in the heart.
NO and S-glutathionylation of the catalytic α-subunit.
S-glutathionylation of thiols by thiol/disulfide exchange with GSSG requires dissociation of the target thiol. Alternative pathways include binding of GSH to the protein thiol groups that have undergone S-nitrosylation or oxidation to sulfenic anion (−SO−) (13, 38). For many, but not all proteins including the β-subunit of Na,K-ATPase (20) and SERCA2A (1), S-nitrosylation of a thiol is a necessary intermediate step precluding S-glutathionylation. Nitrosothiols are formed in reaction with N2O3, an adduct of NO and O2 (37, 43). Hypoxic conditions in the heart do not support S-nitrosylation as NO is converted to ONOO− instead of N2O3 (Fig. 3). The resulting ONOO−- and H2O2-induced oxidation of thiols to thiyl radicals and sulfenic anions generation as well as GSSG accumulation promotes thiol S-glutathionylation (24, 37).
S-glutathionylation of the α1-subunit thiols in hypoxic myocardium occurred in parallel to the decrease of the number of S-nitrosylated cysteine residues (Figs. 5 and 6). However, due to the fact that hypoxic conditions were associated with glutathione oxidation, S-glutathionylation most likely was not mediated by interaction of GSH with nitrosylated thiols of the α-subunit, but via formation of thiyl and sulfenic derivatives (37).
Acute response of the heart to hypoxia: the role of S-glutathionylation.
Rat heart responds to hypoxia with an acute bout of centrally driven tachycardia followed by autonomous bradycardia along with induction of arrhythmias in vivo (36, 52) and ex vivo (Fig. 1 and Table 1). Wistar rats are very sensitive to hypoxia and acute decrease in the atmospheric O2 levels below 10% renders them unconscious. When compared with R. norvegicus, Spalax survive at lower O2 and higher CO2 levels for longer periods of time (5, 6). Spalax live predominately in underground tunnels in which the oxygen tension is often very low (41). Spalax can conduct aerobic work under low O2 pressures due to adaptations in the structural design of skeletal muscles and the cardiorespiratory system, allowing better oxygenation and more efficient perfusion of tissues in the course of hypoxic exposure (6, 60). These animals respond to hypoxia with tachycardia and maintained stroke volume, which altogether results in increased in cardiac output under hypoxic conditions (2, 18). Arrhythmias reported in mole rats under normoxic conditions were diminished during hypoxia-induced increase in heart rate (2).
Comparison of hypoxic responses of the Na,K-ATPase in myocardial tissue of mole rats and Wistar rats provided further support for the importance of the redox state in control of enzyme function. Lack of oxidative stress in Spalax hearts exposed to hypoxia is associated with insensitivity of Na,K-ATPase to hypoxia. However, complete inhibition of the enzyme may be induced by adding of GSSG to ventricular tissue homogenates in all species studied. Of note, mole rats survive hypoxic periods with their hearts-on. This is in marked contrast with the majority of hibernating and hypoxia-sensitive species (19).
Na,K-ATPase is an active player in control of heart rhythm, excitation propagation, and contractile force (64). Hypoxia-induced inactivation of the Na,K-ATPase in rat heart contributes to an decrease in heart rate and scooping of the ST interval (Fig. 1), which has previously been reported in digitalis-treated hearts (25). Some of these changes in ECG are observed in ischemic myocardium along with dose-dependent suppression of Na,K-ATPase triggered by oxygen deprivation (30). Na,K-ATPase is not only a mediator of transmembrane Na/K gradients but also actively participates in Ca2+ handling in the heart (7). Recently, S-glutathionylation was proposed as a universal mechanism regulating all Ca2+ handling systems in cardiomyocytes including ryanodine receptors, SERCAs, L-type Ca2+ channels, and Na/Ca exchange (53, 65). Inhibition of the Na,K-ATPase occurs in parallel with activation of ryanodine receptors and SERCA in the heart as these ion transporters also possess sites of regulatory S-glutathionylation (16, 31, 32). Thus S-glutathionylation allows coordinated regulation of several ion transport systems with the associated increase in intracellular calcium stores in cardiomyocytes. Hypoxia-induced S-glutathionylation promotes fast Ca2+ release and pumping back into the sarcoplasmic reticulum as well as facilitation of intracellular Ca2+ accumulation. Together, these alterations result in an increase in contractile force. However, extensive calcium accumulation elevates the danger of necrotic tissue damage (15).
Our data reveal that hypoxia-induced inhibition of the Na,K-ATPase in rat heart shares similar mechanisms with those in ischemic myocardium. Oxygen deprivation is, hence, decisive in regulation of the enzyme, dominating over acidosis and the lack of glucose. Suppression of Na,K-ATPase activity in hypoxic myocardium is mediated by S-glutathionylation of the catalytic α-subunit of the enzyme, which occurs with gradual ATP deprivation. S-glutathionylation coordinates the activity of a number of ion transport systems in control of contractile function of the heart in response to the changes in redox state and NO production. This regulatory mechanism is conserved in Na,K-ATPase of hypoxia-sensitive and hypoxia-tolerant rodent species. Na,K-ATPase in hypoxia-tolerant Spalax myocardium remains active under hypoxic condition due to the preservation of redox state in heart tissue.
This work has been supported by the Swiss National Academy of Sciences Grants SNF No. 112449 and No.310030_124970 (to A. Bogdanova), Forschungskredit UZH Candoc Grant No. 7082 (to S. Yakushev), and by the United States-Israel Binational Science Foundation Grant No. 2005346 (to A. Avivi and M. Band).
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: S.Y., M.B., and A.A. performed experiments; S.Y., M.B., A.A., and A.B. analyzed data; S.Y., M.B., M.T.v.P., M.G., A.A., and A.B. interpreted results of experiments; S.Y., M.B., and A.B. prepared figures; S.Y., M.B., M.T.v.P., M.G., A.A., and A.B. edited and revised manuscript; S.Y., M.B., M.T.v.P., M.G., A.A., and A.B. approved final version of manuscript; M.T.v.P. and A.B. drafted manuscript; A.B. conception and design of research.
We thank M. J. Shattok and W. Fuller for primary antibodies to total and phosphorylated forms of phospholemman.
- Copyright © 2012 the American Physiological Society