ischemic preconditioning involves dual cardio-protective axes with p38mapk as upstream target

Journal of Molecular and Cellular Cardiology 42 (2007) 981–990www.elsevier.com/locate/yjmcc

Original article

Ischemic preconditioning involves dual cardio-protective axes withp38MAPK as upstream target

Norbert Nagy a,d, Keisuke Shiroto a, Gautam Malik a, Chi-Kuang Huang a, Mathias Gaestel c,Maha Abdellatif b, Arpad Tosaki d, Nilanjana Maulik a, Dipak K. Das a,⁎

a FAHA Cardiovascular Research Center, University of Connecticut School of Medicine Farmington, Cardiovascular Research Institute, CT 06030-1110, USAb University of Medicine and Dentistry of New Jersey, Newark, USA

c Martin Luther University, Hallem, Germanyd University of Debrecen, Debrecen, Hungary

Received 5 December 2006; received in revised form 19 January 2007; accepted 14 February 2007Available online 24 February 2007 ED

Abstract

The existing literature indicates a crucial role of p38 MAP (mitogen-activated protein) kinase (p38MAPK) and its downstream targetMAPKAP kinase 2 (MK2) in ischemic preconditioning (IPC). Accordingly, deletion of MK2 gene should abolish the cardioprotective ability ofIPC. Interestingly, we were able to partially precondition the hearts from MK2−/− knockout mice suggesting the existence of an as yet unknownalternative downstream target of p38MAPK. A recent study from our laboratory also determined a crucial role of CREB (cyclic AMP responseelement binding protein) in IPC. Since CREB is a downstream target of MSK-1 (mitogen- and stress-activated protein kinase-1) situated at thecrossroad of ERK (extracellular receptor kinase) and p38MAPK signaling pathways, we reasoned that MSK-1 could be a downstream moleculartarget for p38MAPK and ERK signaling in the IPC hearts. To test this hypothesis, the rat hearts were subjected to IPC by four cyclic episodes of5 min ischemia and 10 min reperfusion. As expected, IPC induced the activation of ERK1/2, p38MAPK, MK2 and HSP (heat shock protein) 27 asevidenced by their increased phosphorylation; and the inhibition of p38MAPK with SB203580 almost completely, and the inhibition of ERK1/2with PD098059 partially, abolished cardioprotective effects of IPC. Inhibition of MSK-1 with short hairpin RNA (shRNA) also abolished the IPC-induced cardioprotection. SB203580 partially blocked the effects of MSK-1 suggesting that MSK-1 sits downstream of p38MAPK. shRNA-MSK-1 blocked the contribution of both p38MAPK and ERK1/2 as it is uniquely situated at the downstream crossroad of both of these MAP kinases.Although MSK-1 sits downstream of both ERK1/2 and p38MAPK, ERK1/2 activation appears to play less significant role compared top38MAPK, since its inhibition blocked MSK activation only partially. Consistent with these results, shRNA-MSK-1 blocked the partial PC inMK2−/− hearts, and in combination with SB203580, completely abolished the PC effects in the wild-type hearts. The IPC-induced survivalsignaling was almost completely inhibited with SB203580, and only partially with PD 098059 as evidenced from the inhibition patterns of IPCinduced activation of CREB, Akt and Bcl-2. Again SB203580 alone or in combination with shRNA-MSK-1 inhibited IPC induced survival signalcomparatively, suggesting that MSK-1 exists downstream of p38MAPK. Taken together, these results indicate for the first time MSK-1 as analternative (other than MK2) downstream target for p38MAPK, which also transmits survival signal through the activation of CREB.© 2007 Elsevier Inc. All rights reserved.

TRACT

E
Keywords: p38MAP kinase; MAPKAP kinase 2; MSK-1; CREB; Ischemia/reperfusion
1. Introduction

Ischemic preconditioning (IPC) is a phenomenon in whichcyclic episodes of brief ischemia–reperfusion protect myocar-dium against subsequent lethal ischemic injury [1–3]. IPC

R
⁎ Corresponding author. Tel.: +1 860 679 3687; fax: +1 860 6799 4606.E-mail address: [email protected] (D.K. Das).
0022-2828/$ - see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.yjmcc.2007.02.010

remains the state-of-the art technique of cardioprotection, whichresults in the reduction of myocardial infarction by reducingboth necrotic and apoptotic cell death [4,5]. IPC occurs throughdiverse mechanisms that include multiple kinases and involvemitogen-activated kinases (MAPK) as upstream signalingmolecules [6–10].

Stress-activated protein kinases (SAPKs) including extra-cellular signal-regulated kinase (ERK), JNK and p38 mitogen-

mailto:[email protected]

http://dx.doi.org/10.1016/j.yjmcc.2007.02.010

982 N. Nagy et al. / Journal of Molecular and Cellular Cardiology 42 (2007) 981–990

A

activated kinase (p38MAPK) play a crucial role in cardiopro-tection achieved by IPC [11,12]. All SAPKs become activatedduring IPC, and such activation appears to be obligatory for theIPC to be effective [11–13]. Among the SAPKs, p38MAPK hasbeen extensively studied as an effector for IPC. While ischemia/reperfusion-mediated activation of p38MAPK causes cellularinjury and apoptotic cell death, its activation appears to be anobligatory step for IPC [9]. IPC leads to the translocation ofp38MAPK followed by its phosphorylation resulting in theactivation of its downstream target MAPKAP kinase 2 (MK2)and phosphorylation of HSP27 [10]. Cardioprotective effects ofIPC are abolished if p38MAPK is blocked with its specificinhibitor SB203580 [9–12]. Thus, p38MAPK-MAPKAPkinase 2-HSP27 axis has been recognized to play an importantrole in IPC.

Consistent with the previous reports, a recent study deter-mined that the hearts of the MK2−/− gene knockout mice wereresistant to myocardial ischemic reperfusion injury [13,14]. Tofurther test the importance of p38MAPK-MK2-HSP27 axis, weattempted to precondition the MK-/- mouse hearts. Interestinglyenough, IPC of these mouse hearts provided partial cardio-protection, suggesting the existence of an alternate signalingpathway or downstream target other than MK2.

A recent study documented a novel signaling pathwayinvolving p38MAPK-CREB during pharmacological precon-ditioning with resveratrol [15]. CREB is activated by MSK,which is situated downstream of both ERK1/2 and p38MAPK.In order to determine if MSK-1 is also a downstream target forp38MAPK during IPC, we pretreated the heart with shRNAagainst MSK-1 prior to IPC. Cardioprotective ability of IPCwas partially abolished with shRNA against MSK-1 and com-pletely with p38MAPK inhibitor SB203508 in rats suggesting

Fig. 1. Experime

RETR

that both MSK-1 and MK2 are downstream targets forp38MAPK. Additionally, MSK-1 shRNA also completelyabolished partial cardioprotection in MK2−/− mice further sup-porting MSK-1 as another downstream target of p38MAPK inaddition to MK2.

2. Materials and methods

2.1. Animals

All animals used in this study received humane care incompliance with the principles of the laboratory animal careformulated by the National Society for Medical Research andGuide for the Care and Use of Laboratory Animals preparedby the National Academy of Sciences and published by theNational Institutes of Health (Publication Number NIH 85-23,revised 1985). Sprague–Dawley male rats weighing between250 and 300 gm were fed ad libitum regular rat chow withfree access to water until the start of the experimentalprocedure. The rats were randomly assigned to one of thefollowing groups (Fig. 1): perfused for 15 min with (i) KHBbuffer only (control); (ii) an ERK1/2 inhibitor PD098059(20 μM); (iii) a p38 MAPK inhibitor SB 202190 (10 μM);(iv) shRNA-MSK-1 (hearts were injected 48 h prior toexperiment as described below); (v) the hearts were subjectedto ischemic preconditioning (IPC) by four cyclic episodes of5 min ischemia and 10 min reperfusion; (vi) IPC+20 μM PD098,059; (vii) IPC+10 μM SB 202190; (viii) scrambledsequence of shRNA-MSK-1+IPC; (ix) IPC+ shRNA-MSK-1; (x) IPC+shRNA-MSK-1+10 μM SB 203580. All thehearts were then subjected to 30 min ischemia followed by2 h reperfusion.

CTED

ntal model.

983N. Nagy et al. / Journal of Molecular and Cellular Cardiology 42 (2007) 981–990

A

2.2. Construction of short hairpin RNA (shRNA) adenoviralexpression vectors

pSilencer 1.0-U6 expression vector was purchased fromAmbion. The U6 RNA polymerase III promoter and thepolycloning region were subcloned into the adenoviral shuttlevector pDC311 (Microbix). The hairpin forming oligo, corre-sponding to bases 2556–2575 (51CTTTGGCTAAGAGGAG-GAAATTCAAGAGATTTCCTCCTCTTAGCCAAAGT-TTTTT) of the mouse MSK-1 cDNA (accession # NM 153587),and its antisense with ApaI and HindIII overhangs, weresynthesized, annealed and subcloned distal to the U6 promoter.The loop sequence is shadowed. This vector was used togenerate a recombinant adenovirus, using homologous recom-bination in 293.

For the delivery of shRNA-MSK-1, the rats were anesthe-tized with ketamine (100 mg/kg)/xylazine (10 mg/kg, i.p.) inconjunction with buprenorphine (0.5–2.5 mg/kg, s/c, b.d.),intubated and ventilated at a rate of 60 breaths/min with a PEEPof 1 cm H2O. A left lateral thoracotomy was performed underpainless and aseptic conditions. After the chest was opened, thepericardium was cut and 100 μl of shRNA against MSK-1 wasinjected at three different sites in the anterior wall of the leftventricle, approximately 3 mm below the auricle of the leftatrium between the LAD and the first diagonal branch. Theinjection needle was introduced 1 mm into the myocardium atan angle (10°–20°) in a cranial direction. Slight bulging andblanching of the epicardial surface verified the deposition of thesuspension. After injection, the chest was closed with 4-0 nylonsutures and the ventilation rate was reduced to 40 breaths/minuntil the spontaneous respiration starts. The rats were thenextubated and kept in a temperature-controlled environment toprevent hypothermia where the rats were continuouslymonitored. After 48 h of surgery, the rats were re-anesthetizedwith sodium pentobarbital (80 mg/kg, i.p.), the hearts excised toprepare isolated working hearts as described in the subsequentsection.

2.3. Development of MK2−/− mouse

The MK2−/− mice were generated at Martin LutherUniversity, Hallem Germany [14]. Breeding pairs of MK2−/−

mice were transferred to the University of Connecticut HealthCenter where they were available for our use.

MK2−/− or wild-type (with same genetic background) micewere randomly assigned to one of the four groups: (i) control;(ii) ischemia/reperfusion by subjecting the hearts to 30 minischemia and 2 h reperfusion; (iii) IPC by subjecting the heartsto four cyclic episodes of 5 min ischemia and 10 minreperfusion; (iv) by preperfusion w-t hearts with SB203580+shRNA-MSK-1 (as described below) and MK2−/− hearts withshRNA-MSK-1 only.

2.4. Isolated working rat and mouse hearts preparation

Male Sprang–Dawley rat and MK2−/− and its correspondingwild-type mice were used for this study. All animals received

RETR

humane care in compliance with the “Principles of LaboratoryAnimal Care” formulated by the National Society for MedicalResearch and the Guide for the Care and Use of LaboratoryAnimals prepared by the National Academy of Sciences andpublished by the National Institute of Health (NIH PublicationNo. 86-23, revised 1985). The mice (25–34 g) and rats (250–300 g) were anesthetized with sodium pentobarbital (80 mg/kgb/w) (Abbott Laboratories, North Chicago, IL, USA) i.p.injection in case of mice, 80 mg/kg b/w i.p. injection in case ofrat and anticoagulated with heparin sodium (500 IU/kg b/w i.p)(Elkin-Sinn Inc., Cherry Hill, NJ, USA) injection.

After ensuring sufficient depth of anesthesia, thoracotomywas performed and aorta of heart was identified. Excising ofthe heart from chest by the aorta, the lung and fat tissue wasremoved and whole heart transferred to ice-cold (4 °C)modified Krebs–Henseleit bicarbonate solution (KHS), whichcontaining (in mM): NaCl 118; KCl 4.7; CaCl2 1.7; NaHCO3

24; KH2PO4 1.2; MgSO4 12; Glucose 10) until contractionhad ceased. Both the aorta and pulmonary vein werecannulated as quickly as possible and perfused in retrogradeLangendorff mode against at constant perfusion pressure of100 cm of water (10 kPa) for standardization period in case ofrat16. For mouse, a perfusion pressure of 50 cm of water(5 kPa) was used.17 Immediate start of retrograde perfusionhelped to wash blood and its component from the vascularsystem. Perfusate (KHS) temperature was maintained at 37 °Cand saturated with 95% O and 5% CO2 gas mixture in entirethe experiment. The duration of the retrograde perfusion was10 min, after this procedure the heart was switched to inantegrade perfusion mode. In the antegrade perfusion modethe buffer enters the cannulated left atrium at pressureequivalent to 17 cm water (1.7 kPa) in rat and passed to theleft ventricle, from which it is spontaneously ejected throughthe aortic cannula against at pressure equivalent to 100 cmwater (10.0 kPa). In case of mice the preload pressure wasequivalent to 10 cm of water (1 kPa). Control drug-freemeasurements of heart rate, coronary flow, aortic flow, leftventricular end-diastolic pressure, left ventricular developedpressure and its first derivative were recorded in this period.After 30 min stabilization period, coronary perfusate wascollected. After this period, the antegrade perfusion line wasclosed, and the heart was subjected to 30 min ischemia. Beforethe initiation of 2 h reperfusion the heart was perfused inretrograde mode to avoid the development of high incidenceventricular fibrillation. The measurements of the cardiacfunction were carried out at 15, 30, 60 and 120.min of the2 h reperfusion. Any heart that showed any cardiacdisturbance (ventricle arrhythmia, and fibrillation) during theentire experiment was excluded from this study. A total of 12rats (control 2, PD098059 1, SB203580 2, shRNA-MSK-1 3,IPC+shRNA MSK-1 2, IPCshRNA MSK-1+ SB203580 2)were excluded from our study.

2.5. Measurement of the cardiac function

The heart rate, left ventricular develop pressure (the dif-ference between the maximum systolic and diastolic pressure),

CTED

Fig. 2. Primary culture of cardiomyocytes from rat or mouse was infected withMSK-shRNA virus and after 48 h. Western blot analysis was performed usingantibody against MSK-1 (Santa Cruz, CA). Scrambled sequence of shRNA-MSK-1 was used as control.


A

left ventricular end-diastolic pressure, and the first derivative ofdevelop pressure were recorded by Gould p23XL transducer(Gould Instrument System Inc., Valley View, OH). A smallopening (as small as possible) was created for the insertion of abent cannula to the pulmonary vein. All the recordings weretaken when the hearts were in the working condition, that is thebuffer is coming through the pulmonary vein to the left atriumand from there into the left ventricle and finally through theaortic valve into the aorta and entering into the system. Apressure transducer (Gould P23X, Gould Instrument SystemsInc., Valley View, OH) was connected to a side arm of the aorticcannula. The signal was amplified by using Gould 6600 seriessignal conditioner (Gould Instrument System Inc., Valley View,OH) and monitored on Cordat II real-time acquisition system(Triton technologies, San Diego, CA) [16,17]. The aortic flowwas measured by flow meter. The coronary flow was measuredby time-collection of the coronary effluent dripping from theheart.

2.6. Measurements of the infarct size

After the global ischemic procedure, the heart was infusedwith 10% solution of the Triphenyl tetrazolium (TTC) inphosphate buffer through the aortic cannula for 20 min [9,17].The left ventricle was removed and sliced into 1-mm thicknessof cross-sectional pieces and weight. Each slice was scannedwith computer-assisted scanner (Scanjet 5370C). The risk areaof the whole myocardium was stained in red by TTC while theinfarct zone remained in unstained by TTC. These weremeasured by using of computerized software (Scion Image)and these areas were multiplied by the weight of the eachsection, and these results summed up to obtain the total of therisk zone and a infarct zone. The infarct size was expressed asthe ratio of the infarct zone to the risk zone.

2.7. TUNEL Assay for assessment of Apoptotic Cell Death

Immunohistochemical detection of apoptotic cells wascarried out using TUNEL in which residues of digoxigenin-labeled dUTP are catalytically incorporated into the DNA byterminal deoxynucleotidyl transferase II, an enzyme whichcatalyzes a template-independent addition of nucleotidetriphosphate to the 3′-OH ends of double- or single-strandedDNA [4,17]. The incorporated nucleotide was incubated witha sheep polyclonal anti-digoxigenin antibody followed by anFITC-conjugated rabbit anti-sheep IgG as a secondaryantibody as described by the manufacturer (Apop Tag Plus,Oncor Inc., Gaithersburg, MD). The sections were washed inPBS three times, blocked with normal rabbit serum andincubated with mouse monoclonal antibody recognizingcardiac myosin heavy chain (Biogenesis Ltd., Poole, U.K.)followed by staining with TRIRC-conjugated rabbit anti-mouse IgG (200:1 dilution, Dako Japan, Tokyo, Japan). Thefluorescence staining was viewed with a confocal laser micro-scope (Olympus Co., Tokyo, Japan). The number of apoptoticcells was counted and expressed as a percent of total myocytepopulation.

RETR

2.8. Western blot Analysis

Left ventricles from the hearts were homogenized in abuffer containing 25 mM Tris–HCl, 25 mM NaCl, 1 mMorthovanadate, 10 mM NaF, 10 mM pyrophosphate, 10 mMokadaic acid, 0.5 mM EDTA and 1 mM PMSF [6,9]. 100 μg ofprotein of each heart homogenate was incubated with 1 μg ofantibody against the phospho-CREB, p38MAPK, ERK1/2,MSK-1, MAPKAP kinase 2, HSP-27 or Akt (Santa CruzBiotechnology, Inc., Santa Cruz, CA) for 1 h at 4 °C. Theimmune complexes were precipitated with protein A Sephar-ose, immunoprecipitates separated by SDS-PAGE and immo-bilized on polyvinylidene difluoride membrane. The membranewas stripped and reblotted with specific antibodies againstCREB, p38MAPK, ERK1/2, MSK-1, MAPKAP kinase 2,HSP27, Akt or bcl-2. The resulting blots were digitized andsubjected to densitometric scanning using a standard NIHimage program.

2.9. Statistical analysis

The values for myocardial functional parameters, total andinfarct volumes and infarct sizes and cardiomyocyte apoptosisare all expressed as the mean±standard error of mean (SEM).Analysis of variance test was first carried out to test for anydifferences between the mean values of all groups. Ifdifferences between established, the values of the treatedgroups were compared with those of the control group by amodified t-test. The results were considered significant ifp<0.05.

3. Results

3.1. Experiments with isolated rat hearts

We first wished to examine the effects of MSK-1 inhibitionon cardioprotection induced by ischemic preconditioning. Theeffects of MSK-1 inhibition were compared with inhibition ofp38MAPK in order to determine if MSK-1 and p38MAPKfunction in different signal transduction pathways. Since nospecific inhibitors for MSK-1 is commercially available, weused short hairpin shRNA to block MSK induction. Theefficacy of shRNA to block MSK-1 is shown in Fig. 2. Asshown in the figure, MSK-1 from both rat and mouse heartswere inhibited with shRNA. Scrambled (sr) sequence of shRNAwas used as control. The rats were anesthetized; the hearts

CTED


exposed and shRNA-MSK-1 was injected into three differentplaces of the left ventricle. The chest was closed and the ratswere allowed to recover. After 48 h, the rats were re-anesthetized to prepare isolated working heart. As shown inFig. 3, inhibition of MEK1/2, p38MAPK or MSK bythemselves had no effect on the function of the hearts sincethe results were identical to the control. As expected IPCimproved left ventricular function as evidenced by improvedaortic flow, LVDP and LVdp/dt. Inhibition of p38MAPK andMSK-1 (almost completely) and ERK1/2 (only partially)blocked the cardioprotective ability of IPC. No additionalinhibition of IPC-induced cardioprotection was observed whenboth p38MAPK and MSK-1 were used simultaneously

Fig. 3. Effects of IPC and inhibition of MAP kinases as well as MSK-1 on post-ischemperfused via working mode were made globally ischemic for 30 min followed by 2ischemia by four brief cyclic episodes of ischemia and reperfusion. Results are expres

RETRA

suggesting that MSK-1 and p38MAPK function in the samesignaling pathway.

3.1.1. Effects of MSK-1 inhibition on the reduction ofmyocardial infarct size and cardiomyocyte apoptosis inducedby IPC

IPC reduced myocardial infarct size (Fig. 4, top) andcardiomyocyte death due to apoptosis (Fig. 4, bottom)compared to all four controls. Inhibition of ERK1/2 withPD098059 had only partial effect on IPC, but both SB203580and shRNA-MSK-1 almost completely abolished the infarctsize and apoptosis lowering abilities of IPC. Again, thecombination of SB203580 and shRNA-MSK-1 abolished the

ic ventricular recovery. Nine groups of rats were anesthetized, and isolated heartsh of reperfusion. Five groups of rats were subjected to preconditioning prior tosed as means±SEM of six rats per group. *p<0.05 vs. control; †p<0.05 vs. PC.

CTED

Fig. 4. Effects of IPC and inhibition of MAP kinases as well as MSK-1 onmyocardial infarct size (top) and cardiomyocyte apoptosis (bottom). Ninegroups of rats were anesthetized, and isolated hearts perfused via working modewere made globally ischemic for 30 min followed by 2 h of reperfusion. Fivegroups of rats were subjected to preconditioning prior to ischemia by four briefcyclic episodes of ischemia and reperfusion. Results are expressed as means±SEM of six rats per group. *p<0.05 vs. control; †p<0.05 vs. PC.

Fig. 5. Effects of IPC and inhibition of MSK-1 with shRNA on post-ischemicventricular recovery. Four groups (two wild-type; two MK2−/−) of mice wereanesthetized, and isolated hearts perfused via working mode were made globallyischemic for 30 min followed by 2 h of reperfusion. One from each of wild-typeand MK2−/− mice were subjected to preconditioning prior to ischemia by fourbrief cyclic episodes of ischemia and reperfusion. Results are expressed asmeans±SEM of six mice per group. *p<0.05 vs. non-IPC.


ACT
cardioprotective ability of IPC to a similar degree as SB203508alone.
3.2. Experiments with isolated mouse hearts

3.2.1. Effects of MK2 gene deletion on the ventricular recoveryinduced by IPC

Two groups of mice were studied, MK2−/− and wild-typemice of the same genetic background. The hearts from bothgroups of mice were perfused via working mode, precondi-tioned and then subjected to 30 min ischemia and 2 h ofreperfusion. The results were compared with those that werenot subjected to preconditioning. As shown in Fig. 5, bothwild-type and MK2−/− hearts showed post-ischemic ventri-cular recovery by IPC, although the recovery was greater forwild-type hearts. The results suggested that the hearts of theMK2−/− mice could be partially preconditioned suggesting theexistence of an alternative signaling pathway downstream ofp38MAPK.

3.2.2. Effects of MK2 gene deletion on the reduction ofmyocardial infarct size and apoptotic cardiomyocyte deathinduced by IPC

Similar to the results with rat hearts, IPC reduced themyocardial infarct size (Fig. 6, top) and cardiomyocyteapoptosis (Fig. 6, bottom) in the hearts of the wild-typemice. Interestingly enough, IPC also lowered the myocardialinfarct size and cardiomyocyte death due to apoptosis tosome extent again suggesting that the hearts of MK-/-

RETR

knockout mice could be partially preconditioned. In one ofthe groups of MK2−/− and wild-type hearts, shRNA-MSK-1was injected (as described for rat hearts) 48 h prior toexperiment. Inhibition of MSK-1 blocked the cardioprotec-tive ability of IPC in the MK2−/− hearts completely andwild-type hearts partially confirming the previous results forthe existence of an alternative signaling pathway (other thanMK2) downstream of p38MAPK.

3.2.3. Effects of MSK-1 inhibition on the induction of ERK1/2,p38MAPK, MAKKAPK2 (MK2), HSP27 and MSK and theirphosphorylation

Having established the existence of an alternative (other thanMK2) downstream molecular target for p38MAPK, weattempted to determine the identity of the signaling moleculesby Western blot analysis. The results showed increasedphosphorylation of MAPK signaling targets by IPC includingERK1/2, p38MAPPK, MAPKAPK2, HSP27 and MSK-1(Fig. 7). The activation (phosphorylated/non-phosphorylated)of these targets is shown in Fig. 8. The activation of ERK1/2 was partially blocked by PD 098059 while the activationof p38MAPK, MAPKAPK2 and HSP27 was completely

ED

Fig. 6. Effects of IPC and inhibition of MSK-1 with shRNA on myocardialinfarct size (top) and cardiomyocyte apoptosis (bottom). Four groups (two wild-type; two MK2−/−) of mice were anesthetized, and isolated hearts perfused viaworking mode were made globally ischemic for 30 min followed by 2 h ofreperfusion. One from each of wild-type and MK2−/− mice were subjected topreconditioning prior to ischemia by four brief cyclic episodes of ischemia andreperfusion. Results are expressed as means±SEM of six mice per group.*p<0.05 vs. control; †p<0.05 vs. I/R; #p<0.05 vs. IPC.

Fig. 7. (A) Representative Western blot analysis of the proteins obtained fromthe left ventricle at the end of each experiment. (B) Densitometric scanning ofthe blots normalized against non-phosphorylated products. Results are shown asmeans±SEM of three hearts per group.


RACTE

abolished by SB203580 suggesting that MAPKAPK2 andHSP27 as downstream molecular targets for p38MAPK. PD098059 blocked the activation of MSK-1 only partially. Incontrast, SB203580 blocked the activation of MSK-1 almostcompletely suggesting not only MSK-1 as downstream targetfor p38MAPK but also the relatively more importance ofp38MAPK in IPC-induced cardioprotection. The combina-tion of shRNA-MSK-1 and SB203580 was effectively andcomparatively blocked the activation of MSK-1.

3.2.4. Effects of MSK-1 inhibition on the activation of CREB,Akt and Bcl-2

We next attempted to determine the role of MSK-1 on thesurvival signal generated through IPC. The role of CREB,Akt and Bcl-2 on IPC-induced cell survival has already beenestablished. Consistent with previous reports, IPC increasedthe phosphorylation of CREB and Akt and induced theexpression of Bcl-2 (Fig. 9). The activation of CREB and Akt(normalized against non-phosphorylated products) and Bcl-2is shown in Fig. 9. IPC-induced activation of Bcl-2, CREBand Akt was almost completely blocked with SB203580,

RET

shRNA-MSK-1, but not with PD 098059, which onlypartially inhibited the activation of CREB and Akt. Again,the combination of shRNA-MSK-1 and SB203580 inhibitedthe activation of CREB, Akt and Bcl-2 comparativelysuggesting that p38MAPK and MSK-1 share the commonsignaling pathway.

4. Discussion

The most important finding of this study is the identificationof MSK-1 as downstream molecular target of p38MAPK inaddition to the known target MAPKAPK2. MSK-1 is situated atthe crossroad of ERK1/2 and p38MAPK signaling pathway.IPC activated p38MAPK and ERK1/2 as expected; however,inhibition of ERK1/2 with PD 098059 only partially blockedthe cardioprotective ability of IPC and did not inhibit MSK-1D

Fig. 8. (A) Representative Western blot analysis of the proteins obtained fromthe left ventricle at the end of each experiment. (B) Densitometric scanning ofthe blots normalized against non-phosphorylated products. Results are shown asmeans±SEM of three hearts per group.

Fig. 9. Proposed model for the signal transduction pathway triggered by IPCshowing dual cardioprotective axes involving p38MAPK-MSK-1-CREB-Akt-Bcl-2 and p3MAPK- MAPKAPK2-HSP27-Akt-Bcl-2.


RA
induction significantly. In contrast, inhibition of p38MAPKwith SB203580 almost completely abolished IPC-inducedcardioprotection. In concert, SB203580 blocked the phosphor-ylation of MAPKAPK2 and its downstream target HSP27 aswell as phosphorylation of MSK-1 and its downstream targetCREB. The results thus indicate a crucial role of p38MAPK inIPC, which regulates cardioprotection through dual signalingpathways p38MAPK-MK2-HSP27 and p38MAPK-MSK-1-CREB. In consistent with these results, the survival signalgenerated through IPC was also almost completely blocked witheither SB203580 or shRNA-MSK-1, but only partially PD098059 as evidenced from the results of the activation of CREB,Akt and Bcl-2.
These results were further substantiated with the resultsobtained from MK2−/− mice. The hearts of the MK2−/− micecould be partially preconditioned (compared to wild-typemouse hearts) suggesting the existence of an alternativesignaling pathway in addition to p38MAPK-MK2- HAP27. IfMK2 is the only downstream target of p38MAPK, MK2−/−

mouse hearts could not be preconditioned. These results were

RET

further confirmed from the results that shRNA-MSK-1 couldabolish cardioprotective effects of IPC in the MK2−/− mousehearts and a combination of SB203580+shRNA-MSK-1 couldsimilarly block IPC-induced cardioprotection in the wild-typemouse hearts.

IPC has been known to modulate MAP kinase signaling.Among the three MAP kinases, ERK1/2 is involved in cellproliferation, while p38MAPK and JNK are activated inresponse to environmental stress. MSK-1, a mitogen- andstress-activated protein kinase-1, is situated downstream ofERK1/2 and p38MAPK. MSK-1 that belongs to AGC family ofkinases and is related in structure to ribosomal p70 S6 subfamilycan be activated by both ERK1/2 and p38MAPK. MSK-1 aswell as MSK-2 can be directly activated both in vitro and invivo by p42/44 ERK and p38MAPK in cultured cells [18]. In arecent study, MSK-1 and MSK-2 activities were increased 400–500% and 200–300%, respectively, in exercised muscle alongwith an increase in MAPKAP kinase 2 [19]. In another study,ERK1/2 phosphorylation increased 7.8-fold and p38 MAPKphosphorylation increased 4.4-fold after the exercise [20]. Theactivity of MAPKAP kinase 2, the downstream target of p38MAPK, increased 3.1-fold while MSK-1, downstream of bothERK1/2 and p38MAPK increased 2.4-fold at the same time. Inthe present study, IPC- mediated increase in MSK-1 appears tobe the result of the activation of both p38MAPK and ERK1/2,because inhibition of either p38MAPK or ERK1/2 resulted in ldownregulation of MSK-1. It should be noted, however, thatERK1/2 had only a little contribution compared to p38MAPKas its inhibition only partially blocked the activation of MSK-1.

As mentioned earlier, MSK-1 (mitogen- and stress-activatedprotein kinase) is a kinase activated in cells situated at the

CTED


A

downstream crossroad of ERK1/2 and p38MAPK. MSK-1 andits closely related isoform MSK2 are activated by both ERK1/2and p38MAPK in cells and are involved in the regulation oftranscription downstream of ERK1/2 and p38MAPK [18].MSKs are homologous to the RSK family kinases, as like RSK,they also contain two kinase domains joined by a short linkerregion [21].

MSK-1 is required for cyclic AMP response element (CRE)-binding protein (CREB) and the closely related activatingtranscription factor (ATF1) activation after mutagenic or stressstimuli. Upon phosphorylation, they recruit the co-activatorCREB binding protein thereby effecting phosphorylation atSer133. The phosphorylation of CREB at Ser133, however, isnot catalyzed by MAPK family members, but by RSK andMSK. The overexpression of MSK2 stimulates CREB-depen-dent reporter gene transcription in transfected cells [22].Recently, pharmacological preconditioning with resveratrolwas found to phosphorylate CREB via adenosine A1 and A3receptors through the activation of Akt survival pathway14.Another related study demonstrated the activation of CREB byresveratrol though Akt-dependent as well as Akt-independentpathways [23]. Several distinct pathways can induce CREB,which is an important nuclear factor for cell survival. Forexample, growth factors and stress can induce CREBphosphorylation through the activation of downstream targetsof MAP kinase signaling pathways including classical ERKpathway and stress-activated p38MAPK pathway [24]. Recentstudies determined that MSKs are the major growth factor-regulated CREB kinase [25]. Previous studies demonstrated theinvolvement of CREB in transmitting resveratrol-mediatedsurvival signal through the activation of Bcl-2 [14,23]suggesting that resveratrol activates the cell survival proteinBcl-2 through the phosphorylation of CREB. The present studydocuments that similar to pharmacological preconditioning withresveratrol, IPC also activates Bcl-2 through CREB, which inturn becomes activated through the phosphorylation of itsupstream signaling molecule MSK-1.

Phosphorylation of P38MAPK has been found to play anessential role in IPC [9–12]; however, the exact mechanism bywhich activation of this pro-apoptotic MAP kinase generatessurvival signal remains under considerable debate. It was notedthat IPC-mediated activation of p38MAPK prevents furtheractivation during the subsequent ischemia/reperfusion [9].Thus, the amount of p38MAPK remains lower after ischemia/reperfusion in the preconditioned heart compared to non-preconditioned heart, which leads to the reduction of the deathsignal. Subsequently, it was found that among the four differentisomers of p38MAPK, p38MAPKα is proapoptotic whereasp38MAPKβ is antiapoptotic. Nevertheless, both α and βisomers are equally activated by IPC thereby generating bothdeath and survival signals. More recently, α and β isomers ofp38MAPK have been found to be differentially associated withthe lipid rafts during ischemia/reperfusion and IPC. This studyfound that the more p38MAPKb becomes associated withcaveolin-1 during ischemia/reperfusion resulting in a deathsignal, the more p38MAPKa becomes associated with caveolin-3 during IPC resulting in a survival signal. MAPKAP kinase 2 is

RETR

the well-known downstream target for p38MAPK. A largenumber of reports exist in the literature indicating thatMAPKAP kinase 2 also plays a crucial role in IPC [12].Preconditioning potentiates the activation of p38MAPK leadingto the phosphorylation of MAPKAP kinase 2, which in turnupregulates heat shock protein 27 (HSP 27) [26]. Consistentwith the previous reports, the present study also demonstratesthat IPC results in the activation of p38MAPK-MK2-HSP27signaling pathway.

As expected, IPC triggered the enhancement of thephosphorylation of the survival proteins, Akt and Bcl-2,confirming many previous reports [27,28]. In addition, IPCalso increased another anti-death transcription factor CREB,activated by its upstream target MSK-1 as the inhibition ofMSK-1 with shRNA abolished the phosphorylation of CREB.Phosphorylation of CREB by MAPK and Akt/PKB has beenimplicated as important for cellular survival in cultured cells[29]. CREB functions as an antiapoptotic factor in sympatheticneurons [30]. In this study, CREB was found to mediateneurotrophin-dependent survival via induction of antiapoptoticBcl-2. The results of our study also indicate simultaneousactivation of CREB and Bcl-2 along with Akt since inhibition ofMSK-1 with shRNA blocked the activation of all of thesesurvival factors.

In summary, the results of the present study showed for thefirst time that IPC could potentiate two different survivalpathways with p38MAPK as the principal regulator. To the bestof our knowledge, we demonstrate for the first time IPC-mediated activation of its downstream target MSK-1, which inturn phosphorylates CREB and transmits survival signal.

Acknowledgments

This study was supported in part by NIH HL 22559, HL33889 and HL 56803.

CTED

References

[1] Das DK, Maulik N. Cardiac genomic response following preconditioningstimulus. Cardiovasc Res 2006;70:254–63.

[2] Shintani-Ishida K, Nakajima M, Uemura K, Yoshida K. Ischemicpreconditioning protects cardiomyocytes against ischemic injury byinducing GRP78. Biochem Biophys Res Commun 2006;345:1600–5.

[3] Halestrap AP. Mitochondria and preconditioning: a connexin connection?Circ Res 2006;99:10–2.

[4] Das S, Engelman RM, Maulik N, Das DK. Angiotensin preconditioning ofthe heart: evidence for redox signaling. Cell Biochem Biophys 2006;44:103–10.

[5] Suzuki YJ, Nagase H, Day RM, Das DK. GATA-4 regulation of myo-cardial survival in the preconditioned heart. J Mol Cell Cardiol 2004;37:1195–203.

[6] Das S, Otani H, Maulik N, Das DK. Redox regulation of angiotensin IIpreconditioning of the myocardium requires MAP kinase signaling. J MolCell Cardiol 2006;41:248–55.

[7] Otani H. Reactive oxygen species as mediators of signal transduction inischemic preconditioning. Antioxid Redox Signal 2004;6:449–69.

[8] Takeishi Y, HuangQ,Wang T, GlassmanM, YoshizumiM, Baines CP, et al.Src family kinase and adenosine differentially regulate multiple MAPkinases in ischemicmyocardium: modulation ofMAP kinases activation byischemic preconditioning. J Mol Cell Cardiol 2001;33:1989–2005.


[9] Sato M, Cordis GA, Maulik N, Das DK. SAPKs regulation of ischemicpreconditioning. Am J Physiol: Heart Circ Physiol 2000;279:H901–7.

[10] Maulik N, Yoshida T, Zu YL, Sato M, Banerjee A, Das DK. Ischemicpreconditioning triggers tyrosine kinase signaling: a potential role forMAPKAP kinase 2. Am J Physiol 1998;275:H1857–64.

[11] Maulik N, Sato M, Price BD, Das DK. An essential role of NFkappaB intyrosine kinase signaling of p38 MAP kinase regulation of myocardialadaptation to ischemia. FEBS Lett 1998;429:365–9.

[12] Maulik N, Watanabe M, Zu YL, Huang CK, Cordis GA, Schley JA, et al.Ischemic preconditioning triggers the activation of MAP kinases andMAPKAP kinase 2 in rat hearts. FEBS Lett 1996;396:233–7.

[13] Nakamura Y, Miura T, Nakano A, Ichikawa Y, Yano T, Kobayashi H, et al.Role of microtubules in ischemic preconditioning against myocardialinfarction. Cardiovasc Res 2004;64:322–30.

[14] Shiroto K, Otani H, Yamamoto F, Huang C-K, Maulik N, Das DK. MK2−/−

gene knockout mouse hearts carry anti-apoptotic signal and are resistant toischemia reperfusion injury. J Mol Cell Cardiol 2005;38:93–7.

[15] Das S, Tosaki A, Bagchi D, Maulik N, Das DK. Resveratrol-mediatedactivation of cAMP response element-binding protein through adenosineA3 receptor by Akt-dependent and -independent pathways. J PharmacolExp Ther 2005;314:762–9.

[16] Engelman DT, Watanabe M, Engelman RM, Rousou JA, Kisin E, KaganVE, et al. Hypoxic preconditioning preserves antioxidant reserve in theworking rat heart. Cardiovasc Res 1995;29:133–40.

[17] Ray PS, Martin JL, Swanson EA, Otani H, Dillmann WH, Das DK.Transgenic overexpression of α B-crystallin confers simultaneous protec-tion against cardiac myocyte apoptosis and necrosis during myocardialischemia reperfusion. FASEB J 2001;15:393–402.

[18] Deak M, Clifton A, Lucocq L, Alessi D. Mitogen- and stress-activatedprotein kinase-1 (MSK-1) is directly activated by MAPK and SAPK/p38,and may mediate activation of CREB. EMBO J 1998;15:4426–41.

[19] Krook A, Widegren U, Jiang XJ, Henriksson J, Wallberg-Henriksson H,Alessi D, et al. Effects of exercise on mitogen- and stress-activated kinasesignal transduction in human skeletal muscle. Am J Physiol 2000;279:R1716–21.

RETRA

[20] YuM, Blomstrand E, Chibalin AV, Krook A, Zierath JR. Marathon runningincreases ERK1/2 and p38MAP kinase signalling to downstream targets inhuman skeletal muscle. J Physiol 2001;536(Pt 1): 273–82.

[21] Blenis J. Signal transduction via the MAP kinases: proceed at your ownRSK. Proc Natl Acad Sci U S A 1993;90:5889–92.

[22] Pierrat BJ, Corria JS, Mary JL, Tomas-Zuber M, Lesslauer W. RSK-B, anovel ribosomal S6 kinase family member, is a CREB kinase underdominant control of p38alpha mitogen-activated protein kinase(p38αMAPK). J Biol Chem 1998;273:29661–71.

[23] Das S, Tosaki A, Bagchi D, Maulik N, Das DK. Resveratrol-mediatedactivation of cAMP response element-binding protein through adenosineA3 receptor by Akt-dependent and -independent pathways. J PharmacolExp Ther 2005;314:762–9.

[24] Shaywitz AJ, Greenberg ME. CREB: a stimulus-induced transcriptionfactor activated by a diverse array of extracellular signals. Annu RevBiochem 1999;68:821–61.

[25] Wiggin GR, Soloaga A, Foster JM, Murray-Tait V, Cohen P, Arthur JS.MSK1 and MSK2 are required for the mitogen- and stress-inducedphosphorylation of CREB and ATF1 in fibroblasts. Mol Cell Biol2002;22:2871–81.

[26] Chevalier D, Allen BG. Two distinct forms of MAPKAP kinase-2 in adultcardiac ventricular myocytes. Biochemistry 2000;39:6145–56.

[27] Maulik N, Engleman RM, Rouson JA, Flack JE, Deaton D, Das DK.Ischemia preconditioning reduces apoptosis by regulating anti death geneBcl-2. Circulation 1999;100:II369–75.

[28] Uchiyama T, Engleman RM, Maulik N, Das DK. Role of AKT signaling inmitochondrial survival pathway triggered by hypoxic preconditioning.Circulation 2004;109(24):3042–9.

[29] Bonni A, Brunet A, West AE, Datta SR, Takasu MA, Greenberg ME.Cell survival promoted by the RAS-MAPK signaling pathway bytranscription-dependent and independent mechanisms. Science 1999;286:1358–62.

[30] Riccio A, Ahn S, Davenport CM, Blendy JA, Ginty DD. Mediation by aCREB family transcription factor of NGF-dependent survival of sym-pathetic neurons. Science 1999;286:2358–61.

CTED

ischemic preconditioning involves dual cardio-protective axes with p38mapk as upstream target

Documents