OUP user menu

Protein kinase activation and myocardial ischemia/reperfusion injury

Stephen C Armstrong
DOI: http://dx.doi.org/10.1016/j.cardiores.2003.09.031 427-436 First published online: 15 February 2004

Abstract

Myocardial ischemia and ischemia/reperfusion activate several protein kinase pathways. Protein kinase activation potentially regulates the onset of myocardial cell injury and the reduction of this injury by ischemic and pharmacologic preconditioning. The primary protein kinase pathways that are potentially activated by myocardial ischemia/reperfusion include: the MAP kinases, ERK 1/2, JNK 1/2, p38 MAPKα/β; the cell survival kinase, Akt; and the sodium–hydrogen exchanger (NHE) kinase, p90RSK. The literature does not support a role for ischemia/reperfusion in the activation of the tyrosine kinases, Src and Lck, or the translocation and activation of PKC. This review will detail the role of these protein kinases in the onset of myocardial cell death by necrosis and apoptosis and the reduction of this injury by preconditioning.

Keywords
  • Protein kinases
  • Cardiomyocytes
  • Ischemia
  • Necrosis
  • Apoptosis

1 Introduction

The events that occur subsequent to coronary artery occlusion are termed myocardial ischemia/reperfusion denoting that these are distinct phases of cellular injury with ATP depletion, lactate accumulation and acidosis observed during ischemia and the production of reactive oxygen and nitrogen species during reperfusion.

The primary purpose of this review is to describe the protein kinase pathways that are activated by myocardial ischemia/reperfusion, regulating the onset, extent and mode of cell death. Myocardial ischemia and ischemia/reperfusion activates several protein kinase families. The primary protein kinase pathways that have been implicated in myocardial ischemia (detailed in Fig. 1) include: the PKC isoforms; the MAP kinases, ERK 1/2, JNK 1/2 and p38MAPKα/β; PI-3 kinase/Akt; and the tyrosine kinases, Src and Lck.

Fig 1

Protein kinase activation pathways. (A) The association of Ras/Raf with a receptor tyrosine kinase (RTK) and the MEK1 mediated activation of ERK permits the phosphorylation of its substrates. (B) The association of G-protein coupled receptors (GPCR) with phospholipase-C (PLC) and production of diacylglycerol (DAG) initiates PKC-ε translocation to the membrane, activating the MAPK pathways. (C) The Src initiated association of PI-3 kinase (PI-3K) with epidermal growth factor receptor (EGFR) induces the production of 3-phosphatidyl-inositol-triphosphate (PIP3), the activation of phosphoinositide dependent kinases (PDK) 1/2, the phosphorylation of Akt Thr308/Ser473 and its substrates. (D) TNF receptor (TNF-R) activates apoptosis signal regulating kinase (ASK1) or mixed lineage kinase (MLK), initiating the phosphorylation of MKK4/7 and MKK3/6 with the downstream activation of JNK and p38MAPK, respectively.

This review will examine the role of these protein kinases in myocardial cell death by necrosis and apoptosis and the reduction of cell death by preconditioning. Ischemic preconditioning consists of brief cycles of ischemia and reperfusion and the protein kinase activation events that are initiated by this cardioprotective protocol are representative of myocardial ischemia/reperfusion events. The controversies on the role of PKC and p38MAPK in preconditioning will be briefly discussed in that this topic has been described in other publications [1,2]. In vivo and perfused heart models will be examined along with in vitro cellular models. Finally, the effect of individual components of myocardial ischemia/reperfusion (I/R), reactive oxygen species (ROS) and reactive nitrogen species (RNS), on protein kinase activation will be discussed. Elucidation of the protein kinase pathways that are activated by ischemia/reperfusion will permit the rational development of pharmacologic approaches that delay myocardial ischemic cell death.

2 Protein kinase C and ischemia/reperfusion

Protein kinase C was the first kinase examined in detail in ischemia/reperfusion and ischemic preconditioning models. Protein kinase C isoforms are grouped as: (1) Classical, Ca2+-dependent PKC α, β and γ; (2) Novel, Ca2+-independent PKC η, θ, δ and ε and (3) Atypical, diacylglycerol (DAG) independent PKC λ, μ, ζ, ι. The translocation of PKC to the membrane and activation by ischemia was initially determined by a protein kinase activity assay (utilizing histone as a substrate) of the cytosol and membrane fractions of hearts subjected to ischemia or ischemia/reperfusion [3]. Ischemia (30 min) increased the H-7 and CGS934B inhibitable protein kinase activities in the membrane fraction. This ischemic translocation was not altered by reperfusion. Ischemic PKC α, δ and ε translocation to the membrane and nuclear/myofibril fractions was reported in rat hearts by immuno-blotting techniques [4]. The translocation of PKC α, ε and ι by global ischemia in a perfused rat heart model was also observed [5]. However, no translocation of PKC δ was observed. These translocation events were not altered by reperfusion.

3 PKC and preconditioning

The role of PKC in ischemic preconditioning is controversial [2]. The role of PKC in preconditioning was initially indicated by the loss of cardioprotection by pharmacologic PKC inhibition [6]. Recent studies utilizing molecular biology techniques have clearly demonstrated the contribution of PKC epsilon to preconditioning. Isolated perfused hearts of PKC-ε−/− knock-out mice lost the reduction of infarct size observed with sibling heterozygous PKC-ε+/− mice [7]. The contribution of multiple protein kinase C isoforms to myocardial ischemic injury was demonstrated by the perfusion of isolated rat hearts with separate peptides that activate PKC-ε (added prior to ischemia) and inhibit PKC-δ (added during reperfusion). These hearts were protected from ischemia/reperfusion as demonstrated by an increase in cardiac function and a decrease in CPK release and infarct size [8]. The pro- and anti-ischemic functions of PKC-ε and PKC-δ, respectively, demonstrate the distinct functions of protein kinase isoforms that will be described for p38MAPK. The regulation of PKC by these peptides is a model for the development of pharmacologic approaches to cardioprotection.

Ischemic preconditioning increased PKC translocation in a conscious rabbit model [9]. PKC-ε translocation was significantly and incrementally increased by one, four and six cycles of ischemia (4 min)–reperfusion (4 min) as compared to control. However, 4 min of ischemia alone also increased PKC-ε translocation, similar to a single cycle of ischemia/reperfusion. Thus it appears that the reperfusion component of the ischemic preconditioning protocol does not contribute to the enhancement of PKC-ε translocation. This is in agreement with the study noted above that showed that reperfusion does not increase the ischemic translocation of PKC-α and -ε [5].

4 Mitogen activated protein kinases and ischemia/reperfusion

In contrast to the heterogeneity of the protein kinase C isoforms, the MAP kinases share a common Thr–X–Tyr site in the activation loop and Thr–Tyr phosphorylation initiates kinase activity. MAP kinases are classified as: (1) Extracellular signal regulated kinases, ERK 1/2 (44/42 kDa), which are activated by growth hormone receptors via a Ras/Raf pathway; and (2) stress activated kinases-Jun-NH2-terminal kinases 1/2 (JNK 1/2) and p38MAPK α/β that are activated, via MKK4/MKK7 and MKK3/MKK6 pathways, respectively, by cellular stresses including: reactive oxygen species, heat shock, inflammatory cytokines, and ischemia.

MAP kinase activation was initially demonstrated by in-gel protein kinase activity assays in perfused rat hearts subjected to ischemia and ischemia/reperfusion [10]. Ischemia activated p38 MAPK and this activation was increased and accelerated by ischemia/reperfusion. JNK was not activated by ischemia alone. However, JNK activation was observed in the ischemia/reperfusion group. ERK was not activated by ischemia or ischemia/reperfusion.

In contrast, another report in perfused rat hearts, utilizing myelin basic protein and Jun substrate kinase assays, observed that ERK and JNK, respectively, were not activated by ischemia [11]. ERK and JNK were rapidly (10 min) activated by ischemia/reperfusion and this was maintained and abrogated, respectively, by 60 min of reperfusion.

4.1 Extracellular signal regulated kinase (ERK)

In a perfused rat heart model, p42ERK was not activated in the nuclear fraction by global ischemia alone. However, ischemia/reperfusion increased ERK dual phosphorylation and kinase activity in the nuclear fraction, as determined by Elk-1 phosphorylation [12]. ERK activation subsequent to reperfusion was preceded by the nuclear translocation of inactive ERK. The activation of MEK-2, an upstream ERK kinase, was also increased by ischemia/reperfusion in the nuclear fraction.

In an in vivo rat model of ischemia/reperfusion, p44ERK was not activated by ischemia alone but was activated by 5–30 min reperfusion, as determined by in-gel kinase assays [13]. Ischemia decreased p42ERK activity while reperfusion increased p42ERK activity as compared to the ischemic time control, although it was not increased above the 0 min control. In contrast, p55 and p46JNK were activated by ischemia alone and this was not increased by 5–15 min reperfusion, requiring 30 min of reperfusion to increase JNK activity.

4.2 Jun-kinase (JNK)

In perfused rat heart and in vivo canine models, ischemia did not activate p55JNK. However, reperfusion for 5–30 min gave a dramatic increase in activation [14]. No activation of p46JNK was observed. Similar findings were observed for p38MAPK in the rat model.

In a perfused rat heart model, global ischemia initiated the translocation of JNK1 to the nuclear fraction [15]. This translocation was followed by the activation of JNK-1, in a reperfusion dependent manner, as demonstrated by in-gel kinase assays. The phosphorylation of SEK1, an upstream activator of JNK1, was increased in the nuclear fraction subsequent to ischemia/reperfusion. In an in vivo rabbit model, JNK kinase activity in cytosol and mitochondrial fractions was not activated by ischemia alone but was activated by ischemia/reperfusion [16].

An overall comparison of PKC and the MAP kinases indicates that PKC translocation and activation is not increased by reperfusion while the MAP kinases, ERK, p38 MAPK and JNK, can be activated by reperfusion (Table 1). This is presumably related to the established activation of MAP kinases by ROS while the activation of PKC by ROS is not clearly defined. The role of ROS in protein kinase activation will be discussed in detail in a subsequent section.

View this table:
Table 1

Protein kinase activation by myocardial ischemia/reperfusion

Protein kinaseIschemiaIsch/RxnPCReferences
ERK-heartsAct=Act++NDBogoyevitch, 1996 [10]
Act=Act++NDKnight, 1996 [11]
PO4/Act=PO4/Act+NDMizukami, 1997a [12]
PO4=PO4+NDTakeshi, 1999 [40]
Act=Act++NDOmura, 1999 [13]
NDNDAct+Ping, 1999a [28]
PO4=PO4+PO4+Fryer, 2001 [26]
PO4=PO4+PO4=Takeshi, 2001 [41]
NDNDPO4+Mocanu, 2002 [27]
CardiomyocytesPO4+PO4+NDPunn, 2000 [49]
Act+Act+NDYue, 2000 [47]
JNK-heartsAct=Act++NDBogoyevitch, 1996 [10]
Act=Act++NDKnight, 1996 [11]
Act=Act++NDYin, 1997 [14]
Act=Act++NDMizukami, 1997b [15]
Act=Act++NDClerk, 1998 [57]
Act=Act++NDHe, 1999 [16]
Act+Act+NDOmura, 1999 [13]
Act+Act+Act+Ping, 1999b [29]
Act=Act++NDMa, 1999 [43]
NDPO4+PO4+Sato, 2000 [25]
Act−NDAct=Nakano, 2000 [20]
PO4+PO4++PO4+Fryer, 2001 [24]
CardiomyocytesAct+Act=Act+Takeshi, 2001 [41]
Act=Act++NDLaderoute, 1997 [51]
Act=Act+NDHe, 1999 [16]
Act+Act+NDYue, 2000 [47]
p38MAPK-heartsAct+Act++NDBogoyevitch, 1996 [10]
NDAct+Act+Maulik, 1996 [17]
PO4=PO4+NDYin, 1997 [14]
PO4=NDPO4+Weinbrenner, 1997 [18]
Act+Act=NDClerk, 1998 [57]
Act+Act−NDPing, 1999b [29]
Act+Act++NDMa, 1999 [43]
NDPO4+PO4+Sato, 2000 [25]
PO4+PO4+PO4−Marais, 2001 [19]
PO4=PO4=PO4=Fryer, 2001 [24]
PO4+PO4=PO4+Takeshi, 2001 [41]
PO4+PO4++NDOtsu, 2003 [23]
CardiomyocytesPO4+NDPO4+Armstrong, 1999 [45]
Act+Act+NDYue, 2000 [47]
MAPKAPK-2Act+Act=NDClerk, 1998 [57]
Act+Act=NDPing, 1999b [29]
Act=NDAct+Nakano, 2000 [20]
Akt-heartsNDNDPO4+Tong, 2000 [32]
PO4=NDPO4+Krieg, 2002 [34]
CardiomyocytesNDNDPO4+Mocanu, 2002 [27]
PO4=PO4+NDMockbridge, 2000 [50]
SrcNDNDAct+Ping, 1999c [36]
PO4+PO4PO4+Takeshi, 2001 [41]
  • The activation levels of each individual protein kinase family in hearts or cardiomyocytes is indicated by the assay employed—phosphorylation or kinase activity. The symbols utilized are: (=) no change; (+) or (++) an increase; and (−) a decrease. These changes are in comparison to: (1) The oxygenated control for the ischemia column; (2) the ischemia control for the ischemia/reperfusion (Isch/Rxn) column; (3) the appropriate oxygenated or ischemic control for the preconditioned (PC) column. ND denotes not determined in the cited reference.

5 MAP kinases and preconditioning

In addition to the role of PKC-ε in ischemic preconditioning, MAP kinases also contribute to this endogenous, cardioprotective mechanism. The first indication that MAP kinases contribute to the ischemic preconditioning was provided by studies utilizing MBP and glycogen synthase peptide substrate kinase assays for MAP kinase and MAPK activated protein kinase 2 (MAPKAPK-2), respectively [17]. The activation of these kinases was increased by ischemic preconditioning in a perfused rat heart model.

5.1 p38MAPK and preconditioning

The specific activation of p38MAPK was determined by immuno-blotting for phospho-Tyr182 p38MAPK in rabbit hearts [18]. The phosphorylation of p38MAPK was significantly increased at 10–20 min prolonged ischemia, subsequent to ischemic preconditioning. This increase was abrogated by the adenosine receptor antagonist, SPT.

In a perfused rat heart model, an initial 5 min episode of preconditioning (global ischemia) gave a dramatic increase in p38MAPK dual phosphorylation that was decreased by 5 min of reperfusion [19]. However, second and third cycles of global ischemia gave incrementally diminished increases in p38MAPK activity, with further decreases by reperfusion. In this model, the ischemic preconditioning protocol decreased the subsequent activation of p38MAPK during a sustained (5–25 min) period of ischemia and during reperfusion (10–30 min).

The activity of the p38MAPK substrate, MAPKAPK-2, was increased by preconditioning in perfused rabbit hearts subjected to global ischemia, as determined by a kinase activity assay of FPLC fractions [20]. This increase in MAPKAPK-2 activity was observed in two separate peaks, corresponding to the 50 and 60 kDa isoforms of MAPKAPK-2, and was abrogated by the adenosine receptor antagonist, SPT and the p38MAPK inhibitor, SB203580. Interestingly, pharmacologic preconditioning with the A1 adenosine receptor agonist, PIA or the p38MAPK/JNK activator, anisomycin, only increased the activity of the 60 kDa peak. JNK activity was decreased by global ischemia.

The mRNA expression of p38MAPKα and β isoforms has been detected in the heart. The differential functions of these p38MAPK isoforms is supported by a report, in a neonatal rat ventricular myocyte (NRVM) model, in which co-expression of active MKK3 and p38MAPKβ enhanced hypertrophy while co-expression of active MKK3 and p38MAPKα increased apoptosis [21]. However, the expression of p38 MAPKβ in the heart, at the protein level, has recently been questioned [22].

In an in vivo mouse model of ischemia/reperfusion, p38MAPKα specific activity was determined by immuno-precipitation of p38MAPKα and kinase activity assays with an ATF-2 substrate [23]. The activation of p38MAPKα by ischemia alone was markedly increased by reperfusion for 10–20 min with a loss of activity by 30 min reperfusion. The deletion of a single copy of p38MAPK α+/− markedly decreased the activation of p38MAPK in this mouse model. This decrease in p38MAPKα activation decreased infarct size, to a degree comparable with preconditioning. The contribution of p38MAPKα to apoptosis could mediate this pro-ischemic function of p38MAPKα.

5.2 Jun kinase and preconditioning

In an in vivo rat model, ischemic preconditioning and pharmacologic preconditioning with the opioid receptor agonist, TAN-67 did not increase p38MAPK dual phosphorylation [24]. However, these preconditioning protocols significantly increased p46/p54JNK dual phosphorylation, subsequent to ischemia and reperfusion (5–30 min), with a loss of this increase subsequent to 60 min reperfusion.

In a perfused rat heart model, phospho-specific antibodies were utilized to determine JNK and p38MAPK activation [25]. Ischemia/reperfusion increased stress kinase dual phosphorylation as compared to control, oxygenated hearts. Ischemic preconditioning also increased kinase activation as compared to the control, oxygenated group. Ischemia/reperfusion of the preconditioned hearts decreased stress kinase dual phosphorylation as compared to the preconditioned, oxygenated group. JNK and p38MAPK phosphorylation was decreased by the inhibitors curcumin and SB203580, respectively, with a corresponding attenuation of the cardioprotection provided by ischemic preconditioning.

5.3 Extracellular signal regulated kinase and preconditioning

In an in vivo rat model, ischemia alone did not increase the dual phosphorylation of p42/p44ERK in the nuclear fraction [26]. However, reperfusion increased p42/p44ERK activation. IPC and pharmacologic preconditioning with TAN-67 significantly increased ERK activation, subsequent to ischemia/reperfusion. ERK activation in nuclear fractions was not blocked by the ERK inhibitor, PD98059. In cytosolic fractions, p44ERK activation was increased by ischemia/reperfusion and this was enhanced by IPC and TAN-67. The cytosolic activation of ERK and the cardioprotection that is mediated by preconditioning was attenuated by PD98059. This indicates that cytosolic ERK activation is an important component of preconditioning. This contrasts with a report in a perfused rat heart model in which IPC increased ERK dual phosphorylation. However, the decrease in infarct size by IPC was not reversed by PD98059, although this inhibitor abrogated ERK1/2 dual phosphorylation [27].

An overall assessment of the literature supports a potential role for ERK, p38MAPK and JNK in preconditioning. However, contrary reports are available suggesting a contribution of different experimental models to these discrepancies. The utilization of the MAPK inhibitors, SB203580 and PD98059, must be considered in the evaluation of these studies.

6 Protein kinase C and MAPK activation

In addition to the distinct activation of the PKC and MAPK pathways, there are cross-interactions between these pathways. The role of PKC in myocardial MAPK activation was demonstrated by studies in which ERK kinase activity was increased by an ischemic preconditioning protocol—six cycles of ischemia (4 min) and reperfusion (4 min) in conscious, instrumented rabbits [28]. This increased ERK activity was: (1) Reduced by PKC inhibition; (2) localized to the nuclear fraction; (3) observed in both p42 (210%) and p44ERK (97%); and (4) accompanied by an increase of total p42/p44 protein in the nuclear fraction. The effect of PKC inhibition upon ERK nuclear translocation was not determined in this study. Ischemic preconditioning also increased the cytosolic activity of MEK1/2 kinase, the upstream activator of ERK. The role of PKC in ERK activation was further indicated by an increase of p42/p44 activity in adult cardiomyocytes over-expressing PKC-ε. This increase was abrogated by PKC inhibition. Interestingly, in this model p42 and p44 JNK activities were increased approximately 50% and 325%, respectively; the inverse of the pattern observed with ischemic preconditioning.

The role of PKC in p46/p54 JNK and p38MAPK activation was also examined in the conscious rabbit model [29]. Ischemic preconditioning stimulated an increase in the activities of p42 and p54JNK, in the nuclear and cytosolic fractions, respectively, as determined by in-gel kinase assays. This increase in JNK activity was abrogated by PKC inhibition. A single brief ischemic event significantly increased the activity of p46JNK in the cytosolic and nuclear fraction and this increase was not altered by a single brief reperfusion event. The activation of p54JNK was not increased by a single ischemic event in either fraction. However, a significant increase in the activity of cytosolic p54JNK was observed with a single ischemia/reperfusion cycle. Thus there is a differential response of these JNK isoforms to ischemia/reperfusion. In the cytosolic fraction, p38MAPK; its upstream activator, MKK3/6; and its downstream substrate, MAPKAPK-2, were all activated by a single ischemic event. This ischemic activation of the p38MAPK pathway was attenuated by a single brief reperfusion period. The differential response of these two stress kinase pathways was further indicated by the observation that an over-expression of PKC-ε increased p46/p54JNK kinase activity, in the absence of p38MAPK activation.

The molecular mechanism that mediates the activation of ERK1/2 by PKC-ε is presumably the established activation of Ras and Raf, upstream of MEK1/2 and ERK1/2. In an NRVM model, PKC-ε overexpression activated p42ERK, with minimal activation observed subsequent to PKC-δ overexpression [30]. In contrast, PKC-δ overexpression activated the stress kinases, p46/p54JNK and p38 MAPK while PKC-ε overexpression only minimally activated the stress kinases. The PKCδ mediated activation of the stress kinases initiated apoptosis, as indicated by the rounding and detachment of cardiomyocytes and DNA fragmentation, as determined by DNA laddering and TUNEL staining. This PKC-δ mediated initiation of apoptosis and stress kinase activation might relate to the aforementioned cardioprotection by the specific inhibition of PKC-δ by peptides [8]. The role of stress kinases in apoptosis will be subsequently described in detail. The discrepancy in JNK activation by PKC-ε is possibly explained by the use of adult [29] and neonatal [30] cardiomyocytes in these studies. This interaction of the PKC and the MAP kinase pathways is an example of the general phenomenon of cross talk between distinct protein kinase pathways, as will be noted in subsequent discussions of the Akt and Src pathways.

7 PI-3 kinase/AKT pathway

The PI-3 Kinase/Akt pathway is a cell survival mechanism that attenuates apoptosis and regulates glycogen synthesis and glucose transport. Akt is activated subsequent to the production of PIP3 by phosphatidylinositol-3-kinase (PI-3K). Akt is a serine/threonine kinase that mediates several functions through the phosphorylation and inactivation of the pro-apoptotic kinase, glycogen synthase kinase-3 (GSK-3 α/β).

Akt, phosphorylation and activation has been associated with ischemia/reperfusion and ischemic preconditioning. This line of research was supported by studies in a perfused rat heart model and an in vivo rabbit model in which the PI-3K inhibitors, LY294002 and wortmannin, respectively, reversed the decrease in infarct size by ischemic preconditioning [27,31]. Ischemic preconditioning increased post-ischemic functional recovery and this was reversed by wortmannin and LY294002 in a perfused rat heart model [32]. This was associated with an increase in Akt phosphorylation that was reversed by wortmannin. Akt activation by ischemic preconditioning resulted in the Ser9 phosphorylation and inactivation of GSK-3β [33]. An increase in Akt Ser473 phosphorylation was mediated by pharmacologic preconditioning with acetylcholine during global ischemia in a perfused rabbit heart model as compared to the low levels of Akt phosphorylation observed in untreated hearts [34]. Akt Ser473 phosphorylation was blocked by the Src inhibitor, PP2, indicating a role for this tyrosine kinase in the upstream activation of PI-3K.

Akt can also modulate the p38MAPK and JNK pathways by phosphorylating and inhibiting their upstream activator, apoptosis signal regulating kinase 1 (ASK1) at Ser83 [35]. This blocks the contribution of p38MAPK and JNK to the apoptotic cell death pathway. The contribution of Src to Akt activation and the regulation of stress kinases by Akt are further examples of the cross-interactions that operate between protein kinase pathways.

8 Tyrosine kinases and preconditioning

In contrast to the abundant literature on the serine/threonine kinases, PKC and MAPK, the role of tyrosine kinases in myocardial ischemia/reperfusion injury has not been extensively examined. The role of the Src family of tyrosine kinases in ischemia/reperfusion was initially examined in the conscious rabbit model that was employed in studies of PKC and MAPK [36]. This report was derived from evidence that tyrosine kinase inhibition blocks the cardioprotection provided by ischemic preconditioning [37]. The activities of the myocardial members of the Src tyrosine kinase family were quantitated in the particulate and cytosolic fractions of hearts subjected to multiple cycles of ischemia/reperfusion. A significant increase of Src activity in the cytosolic fraction, and a dramatic increase in the particulate fraction was observed in the group subjected to the preconditioning protocol and a 30-min period of reperfusion. This increase in Src activation was abrogated by tyrosine kinase and PKC inhibitors. This is further evidence of the cross talk that occurs between PKC and other protein kinase pathways.

In contrast, an increase of Lck activity was observed in the particulate fraction immediately after the preconditioning protocol, with an additional increase in the cytosolic fraction after the 30-min period of reperfusion. Lck activation was also reversed by tyrosine kinase and PKC inhibitors, although the mechanisms of Lck and Src activation are probably not identical. The PKC-ε mediated phosphorylation of Lck at Ser166 presumably mediates the role of PKC in Lck activation [38]. However, the molecular mechanisms that mediate the PKC initiated activation of Src are unclear. A potential mechanism is the PKC facilitated activation of protein tyrosine phosphatases and the dephosphorylation of the Src negative regulatory site, Tyr530 [39].

In a guinea pig model, Src was activated by ischemia as determined by Tyr530 dephosphorylation. Src activity was decreased by ischemia/reperfusion [40] and accelerated by ischemic preconditioning [41]. Src is involved in MAPK activation in that the kinase activities of JNK and Big MAP Kinase 1 (ERK5) were significantly reduced by the Src inhibitor, PP2. The dual phosphorylation of p38MAPK was not altered by PP2 [41].

9 Myocardial ischemic apoptosis and MAPK activation

This review has focused upon the role of protein kinase activation in the necrotic mode of myocardial ischemic cell injury, as determined by infarct size. The role of apoptosis in myocardial ischemic cell death has been described and this role is controversial [42].

JNK and p38MAPK are the primary kinases involved in the execution of myocardial ischemic apoptosis. In a perfused rabbit heart model, global ischemia activated p38MAPK at 15 min with a loss of activation by 30 min [43]. Reperfusion for 10–20 min markedly re-activated p38MAPK in the 30 min ischemic hearts, with a loss of activation by 30–60 min reperfusion. Inhibition of p38MAPK activity with SB203580 significantly reduced apoptosis as determined by DNA laddering and TUNEL assays. Apoptosis was presumably mediated by p38MAPKα. In this model, JNK was not activated by ischemia but was markedly activated by 10–30 min reperfusion with a decrease by 60 min reperfusion. This activation was not altered by SB203580.

10 Cellular models of ischemia/reperfusion and protein kinase activation

This review has focused upon in vivo and perfused heart models of ischemia/reperfusion. Cellular models have been developed to examine the role of protein kinase activation in ischemic injury. These models have been employed to elucidate the necrotic and apoptotic cell death pathways. Cellular models offer the advantage that protein samples are derived solely from cardiomyocytes that are more accessible to molecular biology studies.

An in vitro adult cardiomyocyte pellet model of ischemia was developed that provides the two primary components of ischemia–hypoxia and metabolite accumulation—in the absence of any exogenous metabolic inhibitors or metabolites [44]. This model was employed to assess p38MAPK activation as determined by immuno-blotting with phospho-specific Thr180/Tyr182 antibodies [45]. A significant activation of p38MAPK was noted by 30 min in vitro ischemia and this activation was significantly increased by in vitro ischemic preconditioning. The effect of reperfusion on p38MAPK activation was not determined in this model. The role of ischemic p38MAPK activation in the phosphorylation of the small heat shock protein, HSP27, was indicated by a significant acceleration of the ischemic dephosphorylation of HSP27 by SB203580. In addition, p38MAPK dual phosphorylation was significantly decreased by SB203580. This was one of the initial reports of this observation and it was unexpected in that SB203580 blocks the ATP binding site of p38MAPK. This inhibition of p38MAPK dual phosphorylation by SB203580 was associated with a significant decrease in the ischemic loss of p38MAPK from cytosol. This ischemic loss presumably reflects a nuclear translocation.

The observation that SB203580 decreases p38MAPK dual phosphorylation has been attributed to the MKK3/6 independent, auto-phosphorylation of p38MAPKα, subsequent to the formation of a complex with Tab1 [46]. This p38MAPK auto-phosphorylation event has recently been reported in a myocardial ischemia/reperfusion model [22] and could be an alternative pathway for the initiation of p38MAPKα mediated apoptosis.

Cellular models of ischemia/reperfusion primarily utilize cultured cardiomyocytes, predominantly neonatal rat ventricular myocytes (NRVM) or cell lines. Simulated ischemia (SI) is generally initiated by metabolic inhibitors (deoxyglucose and dithionite) and metabolites (high potassium, lactate, low pH) or by hypoxia in an anaerobic chamber. Reperfusion is simulated by resuspension in an oxygenated buffer.

In an NRVM model, serum free/glucose free hypoxia rapidly and transiently (10–30 min) increased the activity of ERK, JNK and p38MAPK [47]. Reoxygenation restored the activation of these MAP kinases. These kinase activity assays were validated by dual phosphorylation studies. This ischemia/reoxygenation model initiated apoptosis as determined by DNA laddering and TUNEL assays. ERK inhibition with PD98059 accentuated this apoptotic process. ERK inhibition also increased p38MAPK and JNK activation subsequent to ischemia/reoxygenation, indicating the inverse relationship that exists between ERK and stress kinase activation. It is proposed that this cross-inhibition is due to MAPK phosphatase activation or by direct interactions between p38MAPK and ERK [48].

Ischemia was also simulated in an NRVM model by the metabolic inhibitors and metabolites, described above [49]. ERK1/2 dual phosphorylation was increased at 10–20 min with a loss of activation by 40 min ischemia. Dual phosphorylation was restored by 10 min reoxygenation. The dual phosphorylation of p38MAPK was similarly increased and decreased during early and prolonged ischemia with a weak re-activation during reoxygenation. No activation of p46/p55JNK by ischemia or ischemia/reoxygenation was observed. No Akt Ser473 phosphorylation was observed during ischemia. However, reoxygenation activated Akt in a PI-3K dependent manner as determined by inhibition with wortmannin. Akt Ser473 and Thr308 phosphorylation, subsequent to simulated ischemia/reoxygenation was blocked by Src inhibition [50].

In an adult rabbit cardiomyocyte model, simulated ischemia/reperfusion increased JNK activity [16]. This activation was attenuated by transfection of cardiomyocytes with a dominant negative mutant of the upstream activator of JNK, JNK kinase-2 or JNK interacting protein-1 (JIP-1).

In an NRVM model, JNK was not activated by hypoxia in an anaerobic chamber but was activated by reoxygenation with a corresponding induction of apoptosis [51]. JNK activation was blocked by transfection with dominant negative JNK. JNK inhibition increased apoptosis, indicating that JNK can function as an anti-apoptotic kinase [52]. JNK was pro-apoptotic during hypoxia/reoxygenation at low levels of glucose and ATP and anti-apoptotic during hypoxia/reoxygenation at high levels of glucose and ATP [53]. Thus JNK acts as a “bioenergetic sensor” that determines the fate of a cell by its energetic state subsequent to hypoxia/reoxygenation.

These cellular models support the conclusions drawn from in vivo and perfused heart models and facilitate an elucidation of the molecular mechanisms operative in protein kinase activation.

11 Oxidative stress and protein kinase activation

A reductionist approach to the functional components in protein kinase activation, subsequent to ischemia/reperfusion, would focus upon reactive oxygen species (ROS) and reactive nitrogen species (RNS) in that they are the primary components that have been examined for protein kinase activation.

11.1 Oxidative stress and MAPK/PKC activation

Peroxide initiated rapid ERK and p38MAPK activation in an H9c2 cell line, while JNK activation was weak and delayed, as determined by dual phosphorylation and kinase assays [54]. The superoxide generator, menadione, initiated ERK, JNK and p38MAPK dual phosphorylation at 30–60 min. This corresponded to the activation of ERK, JNK and p38MAPK by 15 min of simulated ischemia (serum free/glucose free hypoxia)/reperfusion. The menadione mediated activation of JNK (at 10–20 μM) was associated with apoptosis, as determined by DNA laddering; while ERK and p38MAPK activation (at 30–50 μM) was correlated with necrosis, as determined by trypan blue exclusion.

The nitric oxide donor, SNAP, increased the translocation and kinase activity of PKC-ε in adult rabbit cardiomyocytes. PKCε activation was associated with the tyrosine nitration of PKC-ε [55]. This mechanism potentially contributes to the cardioprotection that is mediated by nitric oxide.

In an NRVM model, the nitric oxide (NO) donor S-nitrosoglutathione activated JNK activity at 10 μM but inhibited JNK at 100 μM [56]. JNK inhibition correlated to an increase in apoptosis as previously described in an NRVM model of hypoxia/reoxygenation [52]. Nitric oxide also increased ERK1/2 activation at 10–1000 μM. The activation of p38MAPK was unaltered by NO. The activation of JNK and ERK by NO (10 μM) was a rapid (10 min) and delayed (120 min) response, respectively.

In a perfused rat heart model, ischemia did not activate p46/p54JNK but a dramatic activation of p46/p54JNK was observed subsequent to ischemia/reperfusion [57]. Peroxide activated JNK, p38MAPK and MAPKAPK-2 to a similar degree as ischemia/reperfusion.

In an NRVM model, peroxide induced a rapid and transient (6–15 min) phosphorylation of Akt Ser473 that was inhibited by the PI-3 kinase inhibitor, LY294002 [58]. In this NRVM model, peroxide also induced a rapid (5 min) activation of p38MAPK and MAPKAPK-2 and a delayed (30 min) activation of ERK1/2, [59]. MAPKAPK-2 activation mediated the phosphorylation of its substrate, HSP27.

These studies indicate that the protein kinase activation events that are initiated by ischemia/reperfusion can be reproduced by oxidative stress, demonstrating that reactive oxygen and nitrogen species are primarily responsible for the protein kinase activation observed subsequent to ischemia/reperfusion.

11.2 Ischemic/oxidative stress and p90RSK

ERK activation by peroxide is dependent upon PKC and tyrosine kinases and is blocked by catalase [60]. In this model, sodium–hydrogen exchanger (NHE) activity is increased by oxidative stress. This increase is abrogated by ERK inhibition with PD98059. This reflects the role of ERK in the phosphorylation of the serine/threonine kinase, p90 ribosomal S6 kinase (p90RSK) at Ser369 and Thr577 [61]. p90RSK activity (as determined by in-gel kinase assays with an NHE substrate) is rapidly and transiently activated by global ischemia in a perfused guinea pig heart model [40]. This activity is restored by reperfusion. ERK and p90RSK activation during reperfusion is dependent upon oxidant production in that MPG abrogated this activation. The ischemic activation of p90RSK is not altered by ischemic preconditioning while p90RSK activation, subsequent to reperfusion, is decreased by preconditioning [41]. Thus decreased p90RSK activation by ischemia/reperfusion, subsequent to preconditioning, would decrease NHE phosphorylation, inhibit NHE and mediate the cardioprotection provided by NHE inhibition [62]. ERK is not activated by ischemia in this model although ERK is activated by ischemia/reperfusion [40]. However, it is difficult to conclude that the ERK mediated activation of p90RSK during ischemia/reperfusion is decreased by preconditioning, in that ERK activation by reperfusion is not altered by preconditioning. It is possible that additional p90RSK kinases are regulated by preconditioning. A possible candidate is 3-phosphoinositide dependent kinase, which is responsible for p90RSK Ser227 phosphorylation in one of the activation loops of p90RSK [63].

A perfused rat heart model was examined with an in-gel kinase assay utilizing NHE as the substrate [64]. This study identified four potential NHE kinases at: 90, 55, 44 and 40 kDa. The activation of the 55, 44 and 40 kDa NHE kinases were increased by ischemia while the activation of the 90, 55 and 44 kDa NHE kinases were increased by ischemia/reperfusion. The ischemic activation of the 40 kDa NHE kinase was decreased by reperfusion. Immuno-precipitation of p90RSK and ERK1 and NHE kinase assays demonstrated that the 90 kDa protein is p90RSK while the 44 kDa protein is partially accounted for by ERK1. p90RSK and ERK activation is inhibited by PD98059. No candidate was offered in this study for the 55 kDa protein. However, PDK1 migrates at 55–60 kDa, suggesting a potential role for PDK1 as an upstream mediator of NHE phosphorylation that is decreased by preconditioning.

12 Summary

The purpose of this review was to define the protein kinases that are activated by ischemia and to examine the impact of reperfusion upon this ischemic protein kinase activation. A summary is presented in Table 1 and it supports the role of ischemia/reperfusion in the activation and dual phosphorylation of ERK, p38MAPK and JNK. The ischemic activation of the p38MAPK substrate, MAPKAPK-2, is not enhanced by reperfusion. Akt activation is potentially dependent upon reperfusion, in that the simulated ischemia studies with cardiomyocytes are supported by ROS studies. The available data support a role for reperfusion in the activation of p90RSK and NHE phosphorylation. The ischemic translocation of PKC is not increased by reperfusion and the ischemic increases in Src and BMK1 activities are negatively regulated by reperfusion.

The preconditioning studies described in this review were utilized to illustrate the protein kinase activation events that can occur subsequent to brief cycles of ischemia/reperfusion. The purpose of this review was not to resolve the controversial roles of p38MAPK and PKC in the cardioprotection provided by preconditioning [1,2]. However, a primary source of this controversy is possibly the various isoforms of PKC and p38MAPK that are active in cardiomyocytes and the alternate and opposing functions of these isoforms. Specifically, the aforementioned distinct roles of p38MAPKα/β and PKCδ/ε in ischemic injury potentially contribute to the conflicting results presented in the literature.

A critical question is the relative contributions of the substrates that are phosphorylated by these protein kinases to ischemia/reperfusion injury and myocardial preconditioning. The potential end-effectors of the protein kinase pathways are: transcription factors—Elk-1 and c-Myc (ERK), Jun and ATF-2 (JNK), and Elk-1, ATF-2, MEF-2 and CREB (p38 MAPK); heat shock proteins-HSP27 (p38MAPK) and alpha B-crystallin (p38MAPK/ERK) and an NHE kinase, p90RSK (ERK/PDK1). Akt initiates the serine/threonine phosphorylation of several relevant substrates including GSK-3, BAD and ASK1. The phosphorylation of these substrates is illustrated in Fig. 1. The goal of future studies will be to establish the functions that these protein kinase substrates perform during myocardial ischemia/reperfusion, subsequent to phosphorylation. Another goal is the development of biocompatible and specific p38 MAPKα and JNK inhibitors, to attenuate myocardial ischemic apoptosis, analogous to the PKC inhibitory peptides [8].

Footnotes

  • Time for primary review 20 days

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
View Abstract