Inhibiting p90 Ribosomal S6 Kinase Prevents Na+-H+ Exchanger–Mediated Cardiac Ischemia-Reperfusion Injury
Background— Pharmacological and genetic studies indicate that the Na+-H+ exchanger isoform 1 (NHE1) plays a critical role in myocardial ischemia and reperfusion (I/R) injury. We found that p90 ribosomal S6 kinase (RSK) phosphorylated serine 703 of NHE1, stimulating 14–3–3 binding and NHE1 activity. Therefore, we hypothesized that inhibiting RSK in cardiomyocytes would prevent NHE1 activation and decrease I/R-mediated injury.
Methods and Results— To examine the role of RSK in vivo, we generated transgenic mice with cardiac-specific overexpression of dominant negative RSK (DN-RSK-TG). DN-RSK-TG hearts demonstrated normal basal cardiac function and morphology. However, myocardial infarction (left coronary artery occlusion for 45 minutes) in DN-RSK-TG hearts was significantly reduced at 24 hours of reperfusion from 46.9±5.6% area at risk in nontransgenic littermate controls to 26.0±4.2% in DN-RSK-TG (P<0.01). Cardiomyocyte apoptosis was significantly reduced after I/R in DN-RSK (0.9±0.2%) compared with nontransgenic littermate controls (6.2±2.6%). Importantly, activation of RSK and interaction of 14–3–3 with NHE1, necessary for agonist-stimulated NHE1 activity, were increased by I/R and inhibited by 70% in DN-RSK-TG (P<0.01). Next, we transduced rat neonatal cardiomyocytes with adenovirus-expressing DN-RSK (Ad.DN-RSK) and measured NHE1 activity. The baseline rate of pH recovery in acid-loaded cells was equal in cells expressing LacZ or DN-RSK. However, NHE1 activation by 100 μmol/L H2O2 was significantly inhibited in cells expressing DN-RSK (0.16±0.02 pH units/min) compared with Ad.LacZ (0.49±0.13 pH units/min). Apoptosis induced by 12 hours of anoxia followed by 24 hours’ reoxygenation was significantly reduced in cells expressing Ad.DN-RSK (18.6±2.0%) compared with Ad.LacZ (29.3±5.4%).
Conclusions— In summary, RSK is a novel regulator of cardiac NHE1 activity by phosphorylating NHE1 serine 703 and a new pathological mediator of I/R injury in the heart.
Received December 22, 2004; de novo received May 20, 2005; revision received March 16, 2006; accepted March 17, 2006.
The sodium/hydrogen exchanger (NHE) family regulates intracellular pH (pHi). Among the plasma membrane isoforms, only NHE1 is expressed at significant levels in heart. Numerous experimental studies show that NHE1 activity plays a critical role in acute cardiac ischemia and reperfusion (I/R) injury. Pharmacological strategies that inhibit NHE1 activity dramatically reduce infarct size (IS) and improve cardiac function.1 Several compounds, including amiloride, eniporide (EMD-85131), and cariporide (HOE642), are well known as specific NHE1 inhibitors. When cariporide was used in an experimental I/R model, IS was reduced and cardiac cell death was improved.2,3 This evidence led to the clinical testing of highly selective pharmacological inhibitors of NHE1 as potential therapeutic agents for cardioprotection in acute coronary syndromes and after myocardial infarction. Unfortunately, no clinical benefit was observed in 2 large clinical trials,4,5 although a smaller trial in patients with acute anterior myocardial infarction who underwent PTCA demonstrated benefit.6 One reason may be that the basal, acid-stimulated homeostatic function of NHE1 is impaired by cariporide and zoniporide, and this function probably is important in cell survival. These data suggest that strategies focused on inhibiting agonist-mediated (eg, reactive oxygen species in I/R) NHE1 activation while preserving basal function may represent an attractive therapeutic approach.
Clinical Perspective p 2523
We previously reported that transfection of HEK 293 cells with wild-type (WT)-RSK enhanced NHE1 phosphorylation and activity, whereas dominant negative RSK (DN-RSK) reduced NHE1 phosphorylation and activity.7 Furthermore, we found that RSK phosphorylated S703 on the C-terminus of NHE1 and the adapter protein 14–3–3 bound to phospho-S703, which increased NHE1 activity.8,9
Takeishi et al10 reported that RSK and ERK1/2 were activated in patients with late-phase dilated cardiomyopathy. In addition, Seko et al11 reported that RSK and ERK1/2 were activated by Raf-1-MAPK cascade in neonatal rat cardiomyocytes stimulated by VEGF. Additionally, RSK and ERK1/2 were activated by Raf-1 stimulation after hypoxia/reoxygenation in neonatal rat cardiomyocytes.12 From these reports, we propose that NHE1 is activated in the myocardium after I/R by a cascade including ERK1/2, RSK, and NHE1. To verify our hypothesis, we generated DN-RSK transgenic (TG) mice and performed left anterior descending coronary artery occlusion and reperfusion. Here, we demonstrate that DN-RSK-TG mice exhibit reduced IS, decreased cardiomyocyte apoptosis, and partially preserved cardiac function after I/R. The fact that cardiomyocytes transduced with DN-RSK exhibited decreased agonist-stimulated NHE1 activity and improved survival from I/R supports the concept that RSK is a pathological mediator of I/R injury.
Generation of Cardiac-Specific DN-RSK-TG Mice
Rat RSK-1 (Genebank NM031107) was mutated to K94A/K447A to create a kinase dead protein13 with the QuikChange site-directed mutagenesis kit (STRATAGENE, La Jolla, Calif). The DN-RSK gene was cloned into the α-myosin heavy chain promoter.14 DNA was injected into fertilized oocytes derived from FVB mice by the Transgenic Facility at the University of Rochester. Mice were maintained by breeding to FVB F1 animals (Jackson Laboratory, Bar Harbor, Me). Polymerase chain reaction was used to identify transgenic mice to detect the DN-RSK with α-myosin heavy chain promoter constructs. Integration of the transgene used the following primer set to amplify the DN-RSK gene: forward, 5′-TTAGCAAACCTCAGGCACCCTTACCCCACATA-3′; and reverse, 5′-TCCAGCTTCTTCCCCGAAGCCTGTCCATT-3′. All mice were used in accordance with Guidelines for the Care and Use of Laboratory Animals of the National Institutes of Health. All procedures were approved by the University of Rochester Animal Care Committee.
Nontransgenic littermate control (NLC) mice lacking the DN-RSK gene were used as controls. DN-RSK-TG (TG) and NLC male mice at 10 to 14 weeks of age were used. Mice were anesthetized with 2% halothane and 40% oxygen and maintained with 0.5% halothane and 40% oxygen during the chest opening surgery. Tracheotomy was performed to provide artificial ventilation (0.3-mL tidal volume, 120 breaths/min), and the left anterior descending coronary artery was ligated with 8-0 nylon surgical suture 2.0 mm distal from the tip of the left auricle.15
Measurements of Infarct Area and Area at Risk
After 45 minutes of ligation and reperfusion, the left anterior descending coronary artery was reoccluded at the same location, and Evans’ blue dye was perfused from the left ventricular (LV) cavity. The heart was removed and cut transversely into 5 sections, which were incubated in 1.0% 2,3,5-triphenyltetrazolium chloride (TTC; Sigma, St Louis, Mo) for 20 minutes at 37°C. The area at risk (AAR) and IS correspond to the area unstained with Evans’ blue dye and the area unstained with TTC solution, respectively. The ratios of AAR to LV and IS to LV of each slice were determined with NIH Image version 1.63.
Protein Extraction From Heart Tissue
Mouse hearts were washed with 10 mL cold PBS. Ischemic area and nonischemic area were identified by Evans’ blue staining, and the isolated ischemic tissues were frozen in liquid nitrogen and homogenized with 0.5 mL lysis buffer (10 mmol/L Tris-HCl, pH 7.4, 0.15 mol/L NaCl, 0.05% Triton X-100, 0.05% NP-40) containing 2 mmol/L sodium orthovanadate and protease inhibitor cocktail (Sigma). Protein concentration was determined with the Bradford protein assay (Bio-Rad). Protein (30 μg) was separated on SDS-PAGE gels and transferred to nitrocellulose membranes.
Western Blot Analysis
We purchased phospho-p90RSK (Thr359/Ser363) and p90RSK (695–708 of mouse Rsk), phospho-ERK1/2 (Thr202/Tyr204), and JNK antibodies from Cell Signaling Corp (Beverly, Mass). Active JNK (Thr183/Tyr185) antibody was purchased from Promega (Madison, Wis). ERK1/2 and 14–3–3 β antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, Calif). The NHE1 antibody was purchased from Chemicon (Temecula, Calif).
In Vitro Kinase Assay
Protein lysates from the ischemic area were used for in vitro kinase assay. Total protein (1 mg) was immunoprecipitated with RSK antibody (Cell Signaling Corp) and incubated with reaction buffer (25 mmol/L HEPES, 10 mmol/L MgCl2, 10 mmol/L MnCl2, 10 mmol/L ATP), 32P-γ-ATP, and RSK peptide (Upstate, Chicago, Ill). Samples were blotted on filter paper and washed with 0.75% phosphoric acid 3 times. Radioactivity was measured by liquid scintillation.
Preparation of Rat Neonatal Cardiomyocytes and Adenoviral Transfection
For adenovirus preparation, the DN-RSK construct was cloned into the AdEasy-CMV system (QBIOGene, Carlsbad, Calif) with SalI and HindIII restriction enzymes. Primary cultures of cardiac myocytes were prepared from ventricles of 1- to 3-day-old neonatal Wistar rats.16 Briefly, cells were dissociated by collagenase II (Worthington Biochem, NJ) from the ventricles and plated at a density of 1×105 cells/cm2 on 25-mm collagen-coated coverslips in Dulbecco’s modified Eagle’s medium (DMEM) culture medium (DMEM with 10% FBS and 10% horse serum). After 6 hours of plating the isolated cardiomyocytes, 10 μmol/L Ara C was added, and cells were cultured for 24 hours. Then, the culture medium was changed to the DMEM medium with 10 μmol/L Ara C in 10% FBS.
Measurement of NHE1 Activity in Neonatal Rat Cardiac Myocytes
Isolated neonatal cardiomyocytes were cultured on 25 mm glass coverslips. The pHi indicator BCECF-AM was incubated with DMEM without FBS for 30 minutes at 37°C.17 The glass coverslips were mounted into a modified Sykes-Moore chamber (Bellco, Vineland, NJ) with Tris-buffered saline solution (130 mm NaCl, 5 mm KCl, 1.5 mm CaCl2, 1.0 mm MgCl2, 20 mm HEPES, pH 7.4) at room temperature. For acid loading, 20 mmol/L NH4Cl was added before recording. After 2 to 3 minutes of acid loading, cells were washed with Tris-buffered saline solution. The recording chamber was placed on an inverted microscope (Nikon Diaphot) equipped with epifluorescence. The field of interest was reduced to the area of a single cardiomyocyte by the viewfinder placed between the microscope and the photon multiplier tube (R928, Hamamatsu, Japan). BCECF-AM was excited at 490 and 440 nm, and the emission fluorescence was recorded at 500 nm.18
Cell Death Detection In Vitro
Ad.LacZ and Ad.DN-RSK were transduced into neonatal rat cardiomyocytes at varying multiplicity of infection (MOI) (Data Supplement Figure I). There was a concentration-dependent expression of DN-RSK (Data Supplement Figure I) with expression greater than endogenous RSK at 100 MOI. WT-RSK, WT-NHE1, and NHE S703A cDNAs were inserted into pLL3.7-IRES-EGFP (kindly provided by Dr Jay Yang, Columbia University) to make a pLL3.7-WT-RSK-IRES-EGFP expression vector. These vectors were transfected into H9c2 rat embryonic myoblasts using lipofectamine 2000 (Invitrogen). Cells were cultured for 24 hours to allow sufficient protein expression; then, cells were exposed to anoxia. Cells were placed for 12 hours in the anoxia chamber (5% CO2, 95% N2), and after 24 hours, reoxygenation was performed by changing the medium and placing cells in an air incubator (5% CO2, 95% air). After 24 hours, cell death was detected by TUNEL and by a cell death detection ELISA kit (Roche Applied Science, Indianapolis, Ind). Only transfected cells identified by EGFP expression were counted to compare the effects of vector alone (pLL3.7-IRES-EGFP) with WT-RSK, WT-NHE1, and MHE1-S703A (pLL3.7-WT-RSK-IRES-EGFP, pLL3.7-WT-NHE1-IRES-EGFP, and pLL3.7-NHE1-S703A-IRES-EGFP).
NLC and TG hearts were removed and fixed by 4% formaldehyde. The fixed hearts were washed 3 times with 70% ethanol and then embedded in paraffin, sectioned (5 μm thick), and stained with hematoxylin and eosin or Masson’s trichrome stain. The fibrotic area was measured by NIH image version 1.63. LV area was calculated as the surface area of the LV at the widest section. Cardiomyocyte apoptosis was measured by TUNEL (Figure II in the Data Supplement).
M-mode echocardiographic analysis was performed in unanesthetized mice with an Acuson Sequoia C236 echocardiography machine equipped with a 15-MHz–frequency probe (Siemens Medical Solutions, Malvern, Pa). LV function was measured in the short-axis view at midlevel. Percent fractional shortening (%FS) was assessed by measurement of the end-diastolic and end-systolic diameters: end-diastolic diameter minus end-systolic diameter divided by end-diastolic diameter×100%.
All values presented are mean±SE. One-way ANOVA completed by Fisher test was used when appropriate. Values of P<0.05 were considered statistically significant.
The authors had full access to the data and take full responsibility for their integrity. All authors have read and agree to the manuscript as written.
I/R Infarct Area
To determine the effect of inhibiting RSK on I/R injury in vivo, we generated DN-RSK-TG mice. We confirmed cardiac specific overexpression of DN-RSK in TG mice by Western blotting (Figure 1A). No difference in RSK expression was found in kidney (not shown). In the TG mouse, cardiac DN-RSK expression was 13 times higher than endogenous RSK in the NLC heart. Under basal conditions, TG mice displayed no apparent cardiac phenotype compared with NLCs (values similar to sham; see the Table). There were no significant differences between males and females. To assess the effect of DN-RSK on I/R injury, mice underwent 45 minutes of ischemia and 24 hours of reperfusion as described in Materials and Methods. IS, measured by TTC staining, was clearly greater in NLC than TG hearts (Figure 1B). Quantification of the ratio of IS to AAR is summarized in Figure 1C and shows that IS was significantly reduced in TG compared with NLC hearts (46.9±5.6% in NLG versus 26.0±4.2% in TG; P<0.05; n=11). The ratio of AAR to LV did not differ significantly between NLC and TG mice (NLC, 62.5±2.9%; TG, 61.9±2.5%). To evaluate the effects of DN-RSK on cardiomyocyte apoptosis, histopathology and immunohistochemistry for apoptosis (TUNEL) were performed (Data Supplement Figure II). Increased numbers of TUNEL-positive cells were clearly present in NLC hearts compared with DN-RSK hearts, with apoptotic rates of 6.2±2.6% in NLC compared with 0.9±0.2% in DN-RSK (P<0.05; Data Supplement Figure III).
Cardiac RSK Expression and RSK Phosphorylation
We next determined the effect of I/R on RSK phosphorylation as a measure of RSK activity. There was a low basal level of phosphorylation in the absence of I/R (Figure 2). After 45 minutes of ischemia, p-RSK did not change (Figure 2, lane 2). However, after 45 minutes of ischemia and 20 minutes of reperfusion, endogenous p-RSK phosphorylation increased 4-fold (Figure 2, lane 3). p-RSK returned to basal levels within 40 minutes of reperfusion. These data show that endogenous RSK is rapidly and transiently activated by I/R (peak at 20 minutes’ reperfusion).
NHE1 Binding to 14–3–3 Increases After Cardiac I/R
We previously showed that RSK stimulated NHE1 activity by phosphorylating serine 703 (S703) and increasing binding of 14–3–3.7,8,19 To relate NHE1 activity to RSK activity, we measured binding of 14–3–3 to NHE1. We performed immunoprecipitation of 14–3–3 followed by immunoblotting for NHE1 to assay their interaction (Figure 3A and 3B). In mice subjected to a sham procedure, binding of NHE1 to 14–3–3 was not detected in either TG or NLC heart tissue lysates. After I/R (45 minutes/20 minutes), 14–3–3 binding to NHE1 increased by 6.5±0.6-fold in NLC mice compared with TG mice (Figure 3B). In contrast, there was markedly reduced 14–3–3 binding to NHE1 in TG hearts (P<0.05 versus NLC; Figure 3B). To prove that DN-RSK inhibited endogenous RSK activity after I/R, we performed an in vitro kinase assay (Figure 3C). Hearts were exposed to I/R (45 minutes/20 minutes), and RSK was immunoprecipitated from lysates. Activity was measured by 32P incorporation into a synthetic RSK substrate peptide. RSK kinase activity increased ≈4-fold in NLC hearts after I/R but was completely inhibited in DN-RSK-TG hearts (Figure 3C). Therefore, DN-RSK prevents binding of 14–3–3 to NHE1 by inhibiting endogenous RSK in hearts exposed to I/R.
NHE1 Activity in Neonatal Rat Cardiomyocytes
To prove the essential role of RSK as a regulator of NHE1 activity in the heart, we transduced neonatal rat cardiomyocytes with Ad.DN-RSK and Ad.LacZ (500 MOI) and measured NHE1 activity (Figure 4). In response to 100 μmol/L H2O2, NHE1 activity increased 3-fold in LacZ-expressing cardiomyocytes (0.16±0.02 to 0.49±0.13 pHi/min) (Figure 4A). In contrast, in cardiomyocytes expressing DN-RSK, H2O2 did not significantly stimulate NHE1 (0.17±0.08 to 0.14±0.02 pHi/min; Figure 4B). The difference in rate of pHi recovery was highly significant (P<0.05; Figure 4C).
To show the difference in pHi recovery when NHE1 was inhibited by DN-RSK compared with pharmacological antagonism of transport, we used the potent NHE1 inhibitor EIPA (Figure 4B). Pretreatment with 5 μmol/L EIPA decreased pHi recovery to a much greater extent than DN-RSK (0.012±0.0001 pHi/min), significantly below acid-stimulated recovery (Figure 4A). Because NHE1 phosphorylation changes the affinity for H+, we also calculated H+ efflux. There was a dramatic decrease in H+ efflux in DN-RSK–expressing cells over the pH range 6.8 to 7.2, suggesting a primary effect of DN-RSK on affinity NHE1 for H+ (Figure 4D). Western blotting for NHE1 showed no change in expression. These data show that DN-RSK prevents agonist-mediated activation of NHE1.
Effect of DN-RSK and WT-RSK on Cardiomyocyte Cell Death
To provide further evidence for the importance of RSK-mediated activation of NHE1, we studied the effect of altering RSK activity on cardiomyocyte apoptosis induced by anoxia for 12 hours followed by reoxygenation for varying times (A/R). Phosphorylation of endogenous RSK was significantly increased (2.3±0.4-fold; P<0.05) after A/R (12 hours/10 minutes) (Data Supplement Figure IV). We next studied the effect of overexpressing Ad.DN-RSK on rat neonatal cardiomyocyte death induced by A/R. Cells were treated with A/R (12 hours/24 hours) (Figure 5). A/R significantly increased both TUNEL-positive cells (10±2.8% to 32±3.1%; P<0.01) and DNA fragmentation (0.18±0.01 to 0.78±0.09; P<0.01). Transduction with Ad.LacZ or Ad.DN-RSK alone had no effect on apoptosis in the absence of A/R. However, DN-RSK–transduced cardiomyocytes exhibited significantly decreased apoptosis compared with LacZ-transduced cells (A/R Ad.LacZ: TUNEL, 29.3±5.4%; ELISA, 0.63±0.08; A/R Ad.DN-RSK: TUNEL, 18.6±2.0%; ELISA, 0.27±0.06; P<0.05).
To provide further support that RSK-mediated phosphorylation of NHE1 S703 was responsible for the protective effect of DN-RSK, we performed 2 additional experiments (Figure 6): overexpression of WT-RSK and/or NHE1-S703A. Because of technical issues related to transfection efficiency, we used H9c2 cells. In H9c2 cells exposed to A/R, apoptosis was 44.4±3.4% and was not significantly increased after transduction with pLL3.7-IRES-EGFP (51±8.1%; A/R+pLL; Figure 6). In contrast, there was a significant increase in apoptosis in cells transduced with WT-RSK to 77.5±4.6% (A/R+WT-RSK, P<0.05). Transfection of NHE1-WT caused a small increase in apoptosis above that observed with A/R alone (A/R+NHE-WT, 61±4%). However, transfection of NHE1-S703A significantly decreased apoptosis compared with transfection with NHE1-WT. In fact, apoptosis of cells transfected with NHE1-S703A was significantly less than both controls (A/R control and A/R+EGFP). These data suggest that NHE1-S703A acts as a dominant negative for the signal events induced by A/R. A critical role for NHE1 activity in the proapoptotic effect of WT-RSK was shown by 2 findings. First, A/R-induced apoptosis was significantly reduced in H9c2 cells cotransfected with WT-RSK and NHE1-S703A. In these cells, the increase in apoptosis stimulated by WT-RSK was significantly inhibited (to 30±5%, a 60% inhibition). Second, the increase in apoptosis stimulated by WT-RSK was significantly reduced in the presence of the NHE1 inhibitor EIPA compared with untreated cells (EIPA+A/R+WT-RSK, 29.9±5.2%; Figure 6). The magnitude of inhibition by NHE1-S703A was similar to that observed with EIPA (A/R+NHE-WT+EIPA, 39±4%). In summary, these data show that WT-RSK promotes H9c2 apoptosis induced by A/R and that the apoptosis is decreased by inhibiting NHE1 function pharmacologically (EIPA) or genetically (transduction of NHE1-S703A).
Effect of DN-RSK on Functional Recovery 2 Weeks After Reperfusion
To determine the effects of DN-RSK on long-term LV functional recovery, we studied mice after 45 minutes of ischemia and 2 weeks of reperfusion (Figure 7 and the Table, n=11). There were no significant differences in body weight (BW) or heart rate between DN-RSK-TG and NLC mice after sham operation or after 2 weeks of I/R (the Table). There was a significant 21% increase in heart weight (HW) to BW in the NLC mice that reflected an enlarged LV in NLC mice. In contrast, there was a much smaller increase of 8% in HW to BW in the TG mice that was statistically less than in NLC mice (the Table). Morphological measures of ischemic damage were also significantly less in TG mice with increased LV free wall thickness and decreased LV area (a measure of LV dilation). Histological analysis (Masson’s trichrome stain) showed that TG hearts exhibited markedly less fibrosis 2 weeks after reperfusion (Figure V), with a reduction in fibrotic area from 18.2±1.7% in NLC hearts to 6.7±0.9% in DN-RSK-TG hearts (Figure 7A).
Echocardiographic analysis showed that LV dimension at diastole and systole and %FS (the Table) did not differ between NLC and TG sham mice (Figure 7B). However, LV dimension at diastole and systole were significantly smaller in TG than NLC hearts, consistent with the histological measurements (n=11; P<0.05). There was a highly significant improvement in %FS in TG hearts (n=11; P<0.05), consistent with improved systolic function in TG versus NLC.
The major findings of this study are that p90RSK is the primary regulator of NHE1 activity in cardiomyocytes exposed to I/R and that cardiomyocyte-specific expression of DN-RSK in a transgenic mouse decreased the extent of myocardial infarction, prevented cardiomyocyte apoptosis, and improved function after I/R. The mechanisms for the cardioprotective effect of DN-RSK were related to inhibiting NHE1 activity as demonstrated by several experiments. First, there was decreased NHE1 activity after I/R in DN-RSK–expressing hearts as measured by 14–3–3 binding. Second, there was improved functional recovery 2 weeks after I/R in DN-RSK–expressing hearts compared with NLCs. Third, there was increased apoptosis in H9c2 cells expressing WT-RSK that was inhibited by the NHE1 blocker EIPA. Fourth, apoptosis was reduced in H9c2 cells that expressed NHE1-S703A, a mutant lacking the RSK phosphorylation site. These results are consistent with our previous findings that 14–3–3 bound to NHE1 via phosphoserine 703 and increased NHE1 activity.8,9
Importantly, inhibition of NHE1 by blocking RSK decreases agonist-activated NHE1 function without inhibiting basal, homeostatic NHE1 function. This result suggests that blocking RSK may be a better therapeutic strategy than NHE1 inhibitors (such as cariporide and zoniporide) that completely block ion transport as a mechanism to decrease sodium-hydrogen exchange and calcium overload during ischemia. Although RSK has multiple cellular substrates, it appears that NHE1 is the critical substrate for the protective effect of DN-RSK on the basis of 3 experiments. First, we showed that WT-RSK overexpression in H9c2 cells stimulated apoptosis and that an NHE1 inhibitor could reverse the increase in apoptosis. Mechanistically, we showed that DN-RSK inhibits phosphorylation of S703 and binding of 14–3–3, an event we previously showed was required for activation of NHE1. Second, we found that DN-RSK inhibited cardiomyocyte apoptosis induced by A/R in culture. Third, we demonstrated that transduction of NHE1-S703A acted as a dominant negative for Na/H exchange and diminished apoptosis caused by A/R and by WT-RSK. A caveat is that we have not shown that decreased phosphorylation of S703 is the only mechanism by which DN-RSK inhibits NHE1 activity and apoptosis; it is formally possible that alterations in other substrates and/or gene transcription may contribute to the protective effects.
NHE1 is regulated by multiple mechanisms in a tissue- and stimulus-specific manner. Work from our laboratory and others has identified 4 kinases that are putative NHE1 kinases: ERK1/2,20,21 NIK,22 RSK,7,8 and p160ROCK.23 Several groups have characterized kinases activated in hearts exposed to I/R or cardiomyocytes exposed to H2O2.18,24–26 All groups found that both ERK1/2 and RSK were activated under these conditions. It was concluded that the upstream signaling pathway involved MEK1/2 because pretreatment of neonatal rat cardiomyocytes with 2 structurally distinct inhibitors, PD98059 or UO126, inhibited activation of ERK1/2 and RSK and abolished stimulation of NHE activity by I/R or H2O2.18,24,25 Importantly, Rothstein et al27 suggested that H2O2–induced calcium overload was partially mediated by NHE1 activation secondary to phosphorylation of NHE1. The present study is the first to show that RSK activity is specifically required for NHE1 activation in cardiomyocytes in response to I/R and H2O2.
RSK consists of 3 isoforms—RSK1, RSK2, and RSK3—that show the same overall structure consisting of 2 kinase domains, a linker region, and short N-terminal and C-terminal tails. The N-terminal kinase belongs to the AGC group of kinases, which include PKA and PKC. The N-terminal kinase phosphorylates the known substrates of RSK.28 The C-terminal kinase belongs to the calcium/calmodulin-dependent kinase (CaMK) group. The only known function of the C-terminal kinase is regulation of the activity of the N-terminal kinase. Richards and colleagues29 showed that the individual RSK1 kinase domains were under separate regulatory control: ERK1/2 phosphorylates RSK within the C-terminal kinase domain, and phosphoinositide-dependent kinase 1 (PDK1) phosphorylates RSK1 within the N-terminal kinase domain. In addition, our laboratory showed that 14–3–3 is a negative regulator of RSK and that agonist-mediated RSK activation requires dissociation of 14–3–3.9 The individual roles of 14–3–3, PDK1, and ERK1/2 in regulating RSK activation by I/R remain unknown. However, the present study clearly establishes RSK as the primary regulator of NHE1 activation by H2O2 and I/R on the basis of both in vivo and in vitro results with DN-RSK-TG mice and DN-RSK adenovirus. The finding that NHE1-S703A apparently functions as a dominant negative suggests that phosphorylation of S703 may be necessary to stabilize NHE1 in an active state, perhaps via recruitment of other proteins.
Inhibition of NHE1 has been proposed as a therapeutic strategy for cardioprotection because both pharmacological and molecular approaches that inhibit NHE1 are associated with reduced I/R injury. For example, the NHE1 inhibitors cariporide and zoniporide reduced I/R injury and improved recovery of heart function after I/R.2,3 In NHE1-null mice, there was also reduced I/R injury and improved functional recovery.30 In the present study, we found that cardiac-specific DN-RSK overexpression improved LV function 2 weeks after I/R as assessed by LV systolic dimensions and %FS. There was a significant decrease in HW/BW in the TG mice compared with NLC mice (the Table) that reflects a decrease in LV cavity size. Future studies are necessary to elucidate the molecular mechanisms for changes in LV function and remodeling. However, in the large clinical trials that used the NHE1 inhibitors cariporide and eniporide (GUARDIAN and ESCAMI) to assess whether there was a benefit in patients experiencing myocardial infarction, no significant reduction in mortality was observed.4,6 Of interest, in the subgroup of patients who underwent CABG, there was a 25% improvement in LV function with cariporide. A similar result was observed in a small trial of 100 patients with acute anterior myocardial infarction who received cariporide before PTCA,6 suggesting that timing of drug administration and/or nature of ischemia and reperfusion are critical determinants for clinical outcome. The failure of NHE1 inhibitors to improve outcome also may be related to the fact that these inhibitors block the homeostatic functions of NHE1, which may lead to intracellular acidosis and cell death. The present study suggests that targeted inhibition of RSK and reduction of NHE1 activity in response to agonists such as H2O2 (with preservation of NHE1 homeostatic function) is a novel strategy to treat cardiac I/R injury.
Sources of Funding
This work was supported by National Institutes of Health grants HL44721 (to Dr Berk) and HL66919 and GM71985 (to Dr Abe) and American Heart Association grant 0435437T (to Dr Blaxall).
Theroux P, Chaitman BR, Danchin N, Erhardt L, Meinertz T, Schroeder JS, Tognoni G, White HD, Willerson JT, Jessel A. Inhibition of the sodium-hydrogen exchanger with cariporide to prevent myocardial infarction in high-risk ischemic situations: main results of the GUARDIAN trial: Guard During Ischemia Against Necrosis (GUARDIAN) Investigators. Circulation. 2000; 102: 3032–3038.
Zeymer U, Suryapranata H, Monassier JP, Opolski G, Davies J, Rasmanis G, Linssen G, Tebbe U, Schroder R, Tiemann R, Machnig T, Neuhaus KL. The Na+/H+ exchange inhibitor eniporide as an adjunct to early reperfusion therapy for acute myocardial infarction: results of the evaluation of the safety and cardioprotective effects of eniporide in acute myocardial infarction (ESCAMI) trial. J Am Coll Cardiol. 2001; 38: 1644–1650.
Rupprecht HJ, vom Dahl J, Terres W, Seyfarth KM, Richardt G, Schultheibeta HP, Buerke M, Sheehan FH, Drexler H. Cardioprotective effects of the Na(+)/H(+) exchange inhibitor cariporide in patients with acute anterior myocardial infarction undergoing direct PTCA. Circulation. 2000; 101: 2902–2908.
Takahashi E, Abe J, Gallis B, Aebersold R, Spring DJ, Krebs EG, Berk BC. p90RSK is a serum-stimulated NHE1 kinase: regulatory phosphorylation of serine 703 of Na+/H+ exchanger isoform-1. J Biol Chem. 1999; 274: 20206–20214.
Lehoux S, Abe J, Florian JA, Berk BC. 14–3–3 Binding to Na+/H+ exchanger isoform-1 is associated with serum- dependent activation of Na+/H+ exchange. J Biol Chem. 2001; 276: 15794–15800.
Cavet ME, Lehoux S, Berk BC. 14–3–3beta is a p90 ribosomal S6 kinase (RSK) isoform 1-binding protein that negatively regulates RSK kinase activity. J Biol Chem. 2003; 278: 18376–18383.
Takeishi Y, Huang Q, Abe J, Che W, Lee JD, Kawakatsu H, Hoit BD, Berk BC, Walsh RA. Activation of mitogen-activated protein kinases and p90 ribosomal S6 kinase in failing human hearts with dilated cardiomyopathy. Cardiovasc Res. 2002; 53: 131–137.
Seko Y, Tobe K, Ueki K, Kadowaki T, Yazaki Y. Hypoxia and hypoxia/reoxygenation activate Raf-1, mitogen-activated protein kinase, mitogen-activated protein kinases, and S6 kinase in cultured rat cardiac myocytes. Circ Res. 1996; 78: 82–90.
Bjorbaek C, Zhao Y, Moller DE. Divergent functional roles for p90rsk kinase domains. J Biol Chem. 1995; 270: 18848–18852.
Gulick J, Subramaniam A, Neumann J, Robbins J. Isolation and characterization of the mouse cardiac myosin heavy chain genes. J Biol Chem. 1991; 266: 9180–9185.
Akimoto H, Ito H, Tanaka M, Adachi S, Hata M, Lin M, Fujisaki H, Marumo F, Hiroe M. Heparin and heparan sulfate block angiotensin II–induced hypertrophy in cultured neonatal rat cardiomyocytes: a possible role of intrinsic heparin-like molecules in regulation of cardiomyocyte hypertrophy. Circulation. 1996; 93: 810–816.
Sabri A, Byron KL, Samarel AM, Bell J, Lucchesi PA. Hydrogen peroxide activates mitogen-activated protein kinases and Na+-H+ exchange in neonatal rat cardiac myocytes. Circ Res. 1998; 82: 1053–1062.
Lehoux S, Abe J, Berk BC. 14–3β–3 is a serum stimulated Na+/H+ exchanger binding protein. Circulation. 1999; 100: I-280–1-281.
Bianchini L, L’Allemain G, Pouysségur J. The p42/p44 mitogen-activated protein kinase cascade is determinant in mediating activation of the Na+/H+ exchanger (NHE1 isoform) in response to growth factors. J Biol Chem. 1997; 272: 271–279.
Yan W, Nehrke K, Choi J, Barber DL. The nck-interacting kinase (NIK) phosphorylates the Na+-H+ exchanger NHE1 and regulates NHE1 activation by platelet-derived growth factor. J Biol Chem. 2001; 276: 31349–31356.
Tominaga T, Ishizaki T, Narumiya S, Barber DL. p160ROCK mediates RhoA activation of Na-H exchange. Embo J. 1998; 17: 4712–4722.
Haworth RS, McCann C, Snabaitis AK, Roberts NA, Avkiran M. Stimulation of the plasma membrane Na+/H+ exchanger NHE1 by sustained intracellular acidosis: evidence for a novel mechanism mediated by the ERK pathway. J Biol Chem. 2003; 278: 31676–31684.
Moor AN, Gan XT, Karmazyn M, Fliegel L. Activation of Na+/H+ exchanger-directed protein kinases in the ischemic and ischemic-reperfused rat myocardium. J Biol Chem. 2001; 276: 16113–16122.
Wei S, Rothstein EC, Fliegel L, Dell’Italia LJ, Lucchesi PA. Differential MAP kinase activation and Na+/H+ exchanger phosphorylation by H2O2 in rat cardiac myocytes. Am J Physiol Cell Physiol. 2001; 281: C1542–C1550.
Rothstein EC, Byron KL, Reed RE, Fliegel L, Lucchesi PA. H2O2-induced Ca2+ overload in NRVM involves ERK1/2 MAP kinases: role for an NHE1-dependent pathway. Am J Physiol Heart Circ Physiol. 2002; 283: H598–H605.
Leighton IA, Dalby KN, Caudwell FB, Cohen PT, Cohen P. Comparison of the specificities of p70 S6 kinase and MAPKAP kinase-1 identifies a relatively specific substrate for p70 S6 kinase: the N-terminal kinase domain of MAPKAP kinase-1 is essential for peptide phosphorylation. FEBS Lett. 1995; 375: 289–293.
Richards SA, Fu J, Romanelli A, Shimamura A, Blenis J. Ribosomal S6 kinase 1 (RSK1) activation requires signals dependent on and independent of the MAP kinase ERK. Curr Biol. 1999; 12: 810–820.
Wang Y, Meyer JW, Ashraf M, Shull GE. Mice with a null mutation in the NHE1 Na+-H+ exchanger are resistant to cardiac ischemia-reperfusion injury. Circ Res. 2003; 93: 776–782.
The Na+-H+ exchanger isoform 1 (NHE1) has been implicated in cardiac dysfunction after myocardial ischemia and reperfusion (I/R) injury by virtue of its ability to promote calcium overload. Biochemical and genetic approaches that blocked NHE1 activity showed cardioprotection after I/R. However, several large clinical trials that used pharmacological NHE1 inhibitors failed to show substantial clinical benefits. One explanation for these results is that both modes of NHE1 regulation (homeostatic [directly through protons] and agonist mediated [indirectly through phosphorylation by kinases]) are blocked by these pharmacological inhibitors. Recently, we found that p90RSK was the predominant kinase responsible for NHE1 phosphorylation and activation by oxidative stress as may occur in I/R. Here, we investigated the role of p90 ribosomal S6 kinase (p90RSK) in I/R injury, focusing on NHE1 activation. Our results show that p90RSK was rapidly activated in hearts exposed to I/R and that cardiomyocyte-specific expression of a dominant negative p90RSK decreased the extent of myocardial infarction and improved long-term left ventricular function after I/R. Our findings suggest that p90RSK activation in ischemic myocardium is pathological and that inhibiting p90RSK and its activation of the NHE1 represents a novel approach to limit the cardiac dysfunction that occurs in ischemic myocardium and diabetic cardiomyopathy.
↵*Drs Maekawa and Abe contributed equally to this study.
The online-only Data Supplement can be found at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.105.563486/DC1.