OUP user menu

Signal transduction of ischemic preconditioning

Rainer Schulz, Michael V Cohen, Matthias Behrends, James M Downey, Gerd Heusch
DOI: http://dx.doi.org/10.1016/S0008-6363(01)00384-4 181-198 First published online: 1 November 2001
  • Ischemia
  • Preconditioning
  • Signal transduction

Time for primary review 20 days.

1 Ischemic preconditioning and its endpoints

Brief episodes of ischemia/reperfusion protect the myocardium from the damage induced by subsequent more prolonged ischemia. When first described by Murry et al. [1] such ischemic preconditioning was elicited by brief coronary occlusion, and the endpoint was reduced infarct size. Soon, however, a variety of preconditioning stimuli were uncovered including hypoxia [2], rapid cardiac pacing [3,4], thermal stress [5], stretch [6,7] and various pharmacological agents (for review, see Refs. [8,9]). Also, various endpoints of ischemic preconditioning have been used: ischemic preconditioning protects against infarction in all species tested so far (for review, see Ref. [8]), and there is also evidence that it might be operative in man (for review, see Ref. [10]). Ischemic preconditioning also reduces the extent of apoptosis [11–13]. Other studies have used recovery of contractile function as an end-point of ischemic preconditioning. Although it appears logical that less infarcted myocardium in preconditioned hearts should result in improved regional and subsequently global function, ischemic preconditioning does not improve regional myocardial function within the first hours of reperfusion (thus it does not attenuate stunning) in rabbits, dogs and pigs in vivo [14–18] (for discussion, see Ref. [19]). However, with longer reperfusion ischemic preconditioning diminishes adverse left ventricular remodeling following infarction and improves long-term functional recovery in chronically instrumented rabbits [20]. Therefore, lack of functional recovery during short periods of reperfusion can not be used to refute preconditioning’s morphological protection, while on the other hand caution should be used when extrapolating findings from short-term recovery of contractile function to reduction of infarct size.

Ischemic preconditioning protects against arrhythmias in mice [21], rats [22–24], rabbits [25] and dogs [26]. In pigs, however [27–30], ischemic preconditioning not only fails to reduce the incidence of ventricular fibrillation during ischemia/reperfusion, but even accelerates the onset of ventricular fibrillation during sustained ischemia and decreases the ventricular fibrillation threshold [27]. These pro-arrhythmic effects cannot be blocked by glibenclamide, indicating that the acceleration of ventricular fibrillation by preconditioning is not a consequence of KATP channel activation [29]. Therefore, the present review will mainly concentrate on the signal transduction of the infarct-reducing effect of ischemic preconditioning.

2 Temporal limits of ischemic preconditioning

Not all combinations and durations of ischemia and reperfusion will trigger the preconditioning phenomenon and protect ischemic myocardium. There appears to be a critical threshold. A preconditioning regime of only 1 or 2 min of ischemia with subsequent reperfusion prior to the index ischemia has no protective effect in rabbits [31], pigs [32] and humans [33]. Above this threshold, a controversy exists on whether or not the protection conferred by ischemic preconditioning is a graded phenomenon that depends on the intensity of the preconditioning stimulus. While in one study in rabbits [34], increasing the intensity of the preconditioning stimulus by increasing the number of preconditioning cycles of ischemia/reperfusion resulted in greater infarct size reduction, this was not confirmed in dogs [35]. On the other hand, increasing the intensity of the preconditioning stimulus by increasing the duration of the preconditioning ischemia from a single cycle of 3 min low-flow ischemia and subsequent 15 min reperfusion to a single cycle of 10 min low-flow ischemia followed by 15 min reperfusion results in a greater infarct size reduction in anesthetized pigs [32].

Apart from the number of preconditioning cycles and the duration of ischemia, the duration of the intermittent reperfusion determines the protection achieved by ischemic preconditioning. In rats, protection is still evident when the reperfusion period is shortened to 1 min, although there is no protection if reperfusion is only 30 s in duration [38]. Typically, a 5-min period of ischemia followed by up to 60 min of reperfusion before the index ischemia results in salvage. However, if the reperfusion interval is prolonged beyond 1 h in anesthetized pigs [39], 1–2 h in anesthetized rabbits [31] or dogs [40], or 2–4 h in conscious rabbits [41] protection is no longer evident. While protection can be reinstituted with a second cycle of preconditioning ischemia/reperfusion in rabbits [42] and dogs [43], this is not the case in pigs [39]. Full reperfusion is not mandatory for ischemic preconditioning’s protection [36,37].

Thus there is a definite and fairly rigid time frame for ischemic preconditioning. Somehow the myocardium ‘remembers’ that it has been preconditioned by a brief ischemic period which has occurred up to several hours before the index ischemia. The exact nature and location of this memory is one of the great unsolved mysteries of ischemic preconditioning.

Fig. 1

Schematic diagram of current ideas on the signal transduction of ischemic preconditioning.

3 Signal transduction in the early phase of ischemic preconditioning

3.1 Receptor-dependent endogenous triggers

3.1.1 Adenosine

The first signal transduction element identified as part of preconditioning’s mechanism (Fig. 1) was the adenosine receptor [44]. An increase in interstitial adenosine concentration during the preconditioning ischemia occurs in rats [45], rabbits [46–48], dogs [49] and pigs [32,50–52]. Attenuation of the increase in interstitial adenosine concentration in pigs [50] or blockade of adenosine A1- and A3-, but not A2-receptors [44,53–56] almost completely blocks the infarct size reduction achieved by ischemic preconditioning in rabbits. Conversely, an increase in the interstitial adenosine concentration with uptake inhibition by dipyridamole lowers the threshold for ischemic preconditioning in anesthetized dogs [57] and rabbits [58].

3.1.2 Bradykinin

Bradykinin also contributes to infarct size reduction by ischemic preconditioning [32,59,60]. The increase in the interstitial bradykinin concentration occurs earlier than the increase in the interstitial adenosine concentration in pigs [32], and its interstitial release during the index ischemia is enhanced by ischemic preconditioning [61]. Bradykinin is a necessary trigger only during a less intense preconditioning stimulus, i.e., a shorter duration of preconditioning ischemia in pigs [32] or a single cycle of ischemia/reperfusion in rabbits [60], while during a more intense preconditioning stimulus, i.e., a more prolonged period of preconditioning ischemia [32] or repeated cycles of ischemia/reperfusion [60], adenosine is more important. From their experiments in rabbits, Goto et al. [60] have proposed a model of additive interaction of triggers that contribute to the initiation of ischemic preconditioning. Also, in pigs only blockade of the bradykinin B2-receptor combined with increased breakdown of endogenous adenosine by adenosine deaminase completely abolishes the infarct size reduction achieved by an intense preconditioning stimulus [32].

3.1.3 Opioids

Blockade of opioid receptors with naloxone abolishes the infarct size reduction achieved by ischemic preconditioning in rats [62–64], rabbits [65,66] and pigs [67]. The important opioid receptor for such cardioprotection has been identified as the δ-receptor [68,69], while the importance of activation of the κ-opioid receptor is still controversial [24,69]. In one study in isolated rat hearts, blockade of κ-opioid receptor abolished ischemic preconditioning [24], suggesting a beneficial role of κ-receptor activation in ischemia/reperfusion. In contrast, in another study using the same animal model, activation of the κ-opioid receptor increased infarct size [69] with the possibility of putting the heart into an ‘anti-preconditioning’ state [70].

Both adenosine and opioids appear to trigger ischemic preconditioning in an interactive fashion in several species [62–64,67]. Whether both signals are oriented in parallel or back-to-back remains to be established. The latter explanation has recently been suggested by studies in isolated rat hearts in which the fentanyl-induced increase in contractile function following ischemia/reperfusion was abolished by pretreatment with an adenosine-receptor antagonist [71].

Of course, the importance of different triggers varies among different species; for example, in the rat opioid receptors appear to be the principal ones involved in ischemic preconditioning [68], and it is difficult to show a role for adenosine [72], while in rabbits and pigs adenosine, bradykinin and opioids each are of major importance.

3.1.4 Prostaglandins, norepinephrine, angiotensin and endothelin

Blockade of cyclooxygenase by aspirin in rats [73] or rabbits [74] did not interfere with the infarct size reduction achieved by ischemic preconditioning, suggesting that endogenous prostaglandins are not involved in preconditioning’s cardioprotection. However, there is evidence that lipoxygenase products may be involved. The improvement in postischemic contractile function in preconditioned hearts was completely abolished by a 12-lipoxygenase inhibitor in rats [75] as well as in 12-lipoxygenase-deficient mice [76].

There are other neurohumoral agonists which can precondition the heart when administered exogenously but which are not released in sufficient quantity by ischemic myocardium to trigger protection endogenously, such as norepinephrine, endothelin and angiotensin. Therefore, administration of antagonists to α-adrenergic [77–80], angiotensin [80,81], or endothelin [82] receptors has no effect on the ability of brief ischemia to precondition the heart.

3.2 Receptor-independent endogenous triggers

3.2.1 Free radicals

Mitochondrial respiration and oxidative phosphorylation are gradually uncoupled during hypoxia or ischemia/reperfusion ([83,84]; for review, see Ref. [85]). An immediate consequence of such gradual impairment of respiratory function is the increase in the production of reactive oxygen species and free radicals in the mitochondria [84]. Also, activated neutrophils can produce large quantities of free radicals [86]. At high concentrations, the highly reactive free radicals can induce lipid peroxidation of membranes, altering their integrity and increasing their fluidity and permeability. A second important site of free radical attack are the membrane proteins involved in the transport of ions and the maintenance of cellular ionic homeostasis. This is especially true for proteins containing sulfhydryl groups (for review, see Ref. [87]). The myocardium, however, has a series of defense mechanisms including the enzymes superoxide dismutase, catalase, and glutathione peroxidase in addition to endogenous antioxidants such as vitamin E, ascorbic acid, and cysteine to protect the cell against cytotoxic oxygen metabolites.

Free radicals, however, not only damage cardiac myocytes but also act as triggers of ischemic preconditioning. Infusion of N-2-mercaptopropionyl glycine, a diffusible antioxidant [88,89], or dimethylthiourea, a radical scavenger [90], blocks the protection of ischemic preconditioning. The mechanism by which oxygen radicals precondition the myocardium is less clear, but free radicals can activate G-proteins [91], protein kinases [90] and ATP-dependent potassium channels [92–94].

3.2.2 Nitric oxide (NO)

The role of endogenous NO in ischemic preconditioning is currently unclear. Blockade of nitric oxide synthase attenuates the increased functional recovery following ischemia/reperfusion induced by ischemic preconditioning in rats [95]. In contrast, NO is not a trigger or mediator of the early phase of ischemic preconditioning against infarction in either rabbits [96] or pigs [97].

3.2.3 Calcium

The flux of calcium through L-type calcium channels has been discussed as a trigger of ischemic preconditioning, since atrial muscle taken from patients with chronic blockade of L-type calcium channels could not be preconditioned by ischemia [98]. However, the calcium antagonist nisoldipine does not attenuate ischemic preconditioning in an in situ model of regional ischemia in anesthetized pigs, thereby challenging a major role of endogenous calcium as a trigger of ischemic preconditioning in vivo [99].

3.3 Exogenous triggers

All of the above substances can put the heart into a preconditioned state when given exogenously by infusion. This includes the endogenous substances that act through receptor activation (adenosine, bradykinin, opioids).

Exposure of rabbit hearts to a low flux of oxygen radicals generated by purine/xanthine oxidase for 5 min, followed by a 15-min washout, puts the heart into a preconditioned state [89,100]. Exogenous NO donors decrease infarct size [96,101], potentially by acting through free radicals [96] or by activating ATP-dependent potassium channels [102] (for review, see Ref. [103]). Also, activation of the prostaglandin EP3 receptor [104–106], the α1-adrenoceptor [77–80], the endothelin receptor [82] or the angiotensin receptor [80,81], by exogenous administration of the respective agonists can put the heart into a preconditioned state. Similarly, repeated intracoronary calcium infusions in anesthetized dogs mimic ischemic preconditioning [107–109].

3.4 G proteins and phospholipases

Adenosine [110–112] and opioid receptors [113] are coupled to inhibitory G proteins. Therefore, chronic treatment with pertussis toxin which blocks inhibitory G proteins abolishes preconditioning’s protection in the rabbit [112]. The adenosine receptor is coupled through Gi to, among others, phospholipases C (PLC) and D (PLD). PLC catalyzes the hydrolysis of membrane inositol-containing phospholipids into inositol trisphosphate and diacylglycerol (DAG). PLD degrades other membrane phospholipids, including phosphatidylcholine. The immediate degradation products are choline and phosphatidic acid. The latter is metabolized further by a phosphohydrolase to DAG [114]. DAG stimulates the translocation and activation of protein kinase C (PKC). The onset of the PLC reaction is typically very rapid, and DAG production is short-lived, peaking at 30 s, whereas the PLD reaction is delayed but accounts for a more prolonged production of DAG, associated with prolonged activation of PKC [115]. Indeed increased PLD activity accompanies preconditioning and PLD blockade with high-dose propranolol attenuates its protection [116].Thus both pathways are thought to be critical components of PKC activation.

3.5 Protein kinases

3.5.1 Protein kinase C

Activation of PKC is physiologically mediated by liberation of DAG from membrane lipids by phospholipases. DAG binds to the regulatory subunit of PKC resulting in a conformational change which causes the molecule to attach to its RACK binding sites (see below) and also uncovers the ATP binding site. Activation of PKC first requires a transphosphorylation of a threonine residue at its activation loop which is followed by dual autophosphorylation at serine and threonine residues at its C-terminus [117,118]. The initial transphosphorylation depends on a phosphoinositide (PI)-dependent kinase (PDK) 1 [119,120]. Whether phosphorylation of PKC is also involved in ischemic preconditioning is unknown. PI3 kinase is activated during ischemia, and the prevention of its activation by wortmannin or LY294002 abolishes the improvement in functional recovery by ischemic preconditioning following 20 min of global ischemia in isolated rat hearts [121]; wortmannin also attenuates the reduction of infarct size by ischemic preconditioning in isolated rabbit hearts [122].

PKC isozymes are not randomly distributed within the cell, but are associated with specific compartments or structures. A peculiar trait of PKC isozymes is that they physically move within the cell when activated. Mochly-Rosen [123,124] demonstrated with immunofluorescent techniques that inactive PKC isozymes are found in the nucleus and perinuclear region, and that they translocate to cross-striated structures (possibly the contractile elements) and cell–cell contact regions following activation. RACKs (receptors for activated C kinase) are located on various intracellular structures and form docking sites for the activated isozymes [124,125]. The RACKs were previously thought to be highly isozyme-specific so that the activated isozyme would bind only to its RACK and to no other; for PKCε it was thought to be RACK 2. However, using PKCε transgenic mice Pass et al. recently demonstrated that at high levels of expression PKCε can interact with both RACK 2 and RACK 1, arguing against the complete selectivity of PKC–RACK interactions [126]. Nevertheless a peptide antagonist for RACK 2 [127] blocked preconditioning’s ability to protect both rat [125] and rabbit [128] cardiomyocytes, suggesting that PKCε is the isoform involved in preconditioning [129].

Ping et al. [130] identified 11 PKC isozymes in the rabbit heart. Using specific antibodies to each isozyme they demonstrated that a series of four cycles of 4 min of coronary occlusion/6 min of reperfusion caused translocation of only two novel isozymes, ε and η, suggesting that it was one of these two isozymes translocated during brief ischemia that contributed to cardioprotection. Studies in rat hearts confirmed that specific isozyme translocation occurs in preconditioned hearts, although the important isoforms appeared to be α, δ, and ε [131]. Finally in dogs, only PKCα was translocated in response to ischemic preconditioning [132]. Using isozyme-specific peptide antagonists [127], it was suggested that only PKCε is capable of putting rat [125] or rabbit [128] cardiomyocytes into a preconditioned state.

In vivo experimental data strongly support the notion that PKC plays an important role in eliciting preconditioning’s protection. In rabbits, blockade of PKC alone completely abolishes the infarct size reduction triggered by ischemic preconditioning [133–135]. However, in the hearts of rats [131,136,137] and dogs [138,139] the results of pharmacological blockade of PKC are controversial. In pigs [140] blockade of PKC with staurosporine alone does not interfere with the infarct size reduction achieved by ischemic preconditioning; this is in part because of an alternate signaling pathway utilizing activation of a protein tyrosine kinase rather than PKC (see below).

3.5.2 Tyrosine kinases

Ischemic preconditioning’s protection in the rabbit can also be aborted by administration of either genistein, a relatively selective tyrosine kinase antagonist by virtue of competitive inhibition of the enzyme’s ATP binding site, or the more selective antagonist lavendustin A, a noncompetitive inhibitor at the ATP binding site as well as a noncompetitive inhibitor at the substrate binding site [141]. Because both genistein and lavendustin A can block the protection induced by PMA, a direct activator of PKC, the involved tyrosine kinase is unlikely to be part of a surface receptor, but rather appears to be downstream of PKC [141].

In the rabbit, PKC and tyrosine kinases appear indeed to be in series. In other species, however, tyrosine kinases may also be present in a second pathway that bypasses PKC. In the pig heart [140,142] antagonists of PKC or tyrosine kinase alone could not block protection from ischemic preconditioning. Yet if they were combined, protection was completely eliminated [142], suggesting the existence of a second signal transduction pathway which parallels PKC and contains at least one tyrosine kinase. The rat heart may act similarly [80,137], since multiple cycles of preconditioning can overcome the abrogation of protection by PKC blockers, suggesting the presence of a similar bypass pathway in that species [137]. Also in dog hearts, blockade of tyrosine kinase alone failed to abolish infarct size reduction by ischemic preconditioning [143]; however, recalling the observations in pigs, this finding does not exclude the involvement of tyrosine kinase in ischemic preconditioning in dogs.

3.5.3 Mitogen-activated protein (MAP) kinases

Potential downstream targets of PKC and tyrosine kinases are the mitogen-activated protein kinases (MAPK). Each subfamily of the MAPK family, ERK, JNK and p38, has been suggested to play a role in the cardioprotection achieved by ischemic preconditioning (for review, see Refs. [144,145]). All of the MAP kinases are activated by dual phosphorylation of a serine and a threonine residue by a MAP kinase kinase. Upon activation, MAP kinases — like PKC — physically move within the cell, and for example translocate into the nucleus (p38, ERK) (for review, see Ref. [145]). The MAP kinase kinase is a tyrosine kinase and at least that for p38 MAPK can be blocked by genistein [146]. Thus MAP kinase kinase has been proposed to be the tyrosine kinase in preconditioning’s signaling pathway [147]. ERK

In isolated rabbit hearts ERK1 activity increases only in ischemically preconditioned myocardium [148], but no difference in ERK1 and ERK2 phosphorylation between non-preconditioned and preconditioned myocardium is detectable in pigs in vivo [149]. Like p38 (see below), ERK can activate MAPKAP kinase 2 α and β, leading to phosphorylation of the small heat shock protein 27 [150].

A causal role of ERK activation in the cardioprotection achieved by ischemic preconditioning is controversial; while an inhibitor of ERK (PD 98059) fails to block the infarct size reduction seen after ischemic preconditioning in isolated rabbit hearts [148], its intramyocardial infusion abolishes ischemic preconditioning’s protection in pigs in vivo [151]. JNK

Both JNK46 and JNK54 are present in the heart [152] and are strongly activated during reperfusion following ischemia. Their activation/phosphorylation during ischemia has been suggested by some studies [149] (for review, see Ref. [145]), but in one report a reduction in JNK phosphorylation during ischemia has been measured [146]. JNK46 activation during no-flow ischemia is most likely mediated by PKC, since activation of JNK46 is completely blocked by chelerythrine, a PKC inhibitor [153].

Intramyocardial infusion of anisomycin reduces infarct size in rabbits [154] and pigs [155] and is associated with activation of JNK. In isolated rat hearts, blockade of JNK46 with curcumin blocks the infarct size reduction of ischemic preconditioning to a similar extent as blockade of p38 using SB203580 [156]. p38

Most attention has focused on the p38 MAPK cascade. At least five isoforms of p38 MAPK have been identified, although only p38 α and β are expressed to any degree within the heart [157]. Different isoforms of p38 appear to mediate different biological functions (for review, see Ref. [145]). In neonatal rat cardiomyocytes, p38α mediates apoptosis whereas p38β is anti-apoptotic [158]. p38 MAPK, like all of the MAP kinases, has two activation sites which must be phosphorylated for activation: a threonine residue at amino acid 180 and a tyrosine residue at site 182. This kinase is activated by a MAP kinase kinase (MEK 3 or MEK 6) [145] which itself is activated by a MAP kinase kinase kinase. p38 is phosphorylated within minutes during global or regional no-flow ischemia in isolated rat hearts [159–164] as well as in rat [162,165], dog [166] and pig hearts in vivo [149,167].

With prolongation of ischemia, however, the phosphorylation of p38 may be reduced towards preischemic values, whereas phosphorylation is once again increased upon reperfusion [162,165]. The transient p38 activation during prolonged ischemia might be related to decreased phosphorylation or increased dephosphorylation by phosphatases (for review, see Ref. [168]). In contrast to the above studies, no activation of p38 by ischemia per se is seen in isolated rabbit hearts [146–148,169].

Following ischemic preconditioning, phosphorylation of p38 during the index ischemia is increased in isolated rat and rabbit hearts [146,147,160,161,170], unaltered in pig hearts in vivo [149], or even decreased in isolated rat hearts [164] and dog hearts in vivo [166]. In vitro, ischemic preconditioning primarily prevents the ischemia-induced activation of p38α in p38α-transfected rat cardiomyocytes, while p38β is downregulated during ischemia in control and preconditioned cells [157].

The importance of p38 activation for cardioprotection thus remains controversial. Protection in isolated cell models is determined from reduced trypan blue uptake and/or enzyme release (lactate dehydrogenase, creatine kinase), while in isolated hearts or hearts in vivo the extent of necrosis (TTC staining, histology) and/or apoptosis is quantified. Rat cardiomyocytes transfected with a dominant negative p38α isoform, which prevents ischemia-induced p38 activation, are more resistant to lethal simulated ischemia [157]. Similarly, in isolated rat hearts [171] and pig hearts in vivo [167], blockade of p38 with SB 203580 does not affect the infarct size reduction achieved by ischemic preconditioning, and even reduces infarct size per se. In contrast, SB 203580 effectively blocks protection in other cell models of ischemic preconditioning [147,172,173], and abolishes the infarct size reduction of ischemic preconditioning in isolated rat hearts [160,174], rabbit hearts [175] and dog hearts in vivo [166].

Explanations for the controversial findings might relate to the relative balance between different isoforms of p38 in different species and experimental models as well as the selectivity of different inhibitors in a given dose range. SB 203580, for example, not only inhibits p38 MAP kinase [176], but also dose-dependently inhibits JNK [177], tyrosine kinases such as p56 lck and c-src [178] and cyclooxygenase [179], and it activates c-raf [180].

3.5.4 Timing of protein kinase activation

The signal transduction cascade of ischemic preconditioning can be divided into triggers and mediators. Triggers are important during the episode of preconditioning ischemia and reperfusion, and blockade of a trigger’s action during this preconditioning phase will abolish its cardioprotective effects. Mediators are important during the prolonged index ischemia, and blockade of a mediator’s action during the sustained ischemia will abolish preconditioning’s protection.

The timing of protein kinase activation appears to be critical for the cardioprotection achieved by ischemic preconditioning. The rabbit heart can still be ischemically preconditioned in the presence of staurosporine, a PKC antagonist which blocks PKC’s ability to phosphorylate substrates, as long as staurosporine is washed out prior to the index ischemia. However, introduction of staurosporine just prior to the index ischemia is sufficient to completely abort preconditioning’s protection against infarction [135]. Thus it would appear that phosphorylation by PKC is required only during the index ischemia. That finding would eliminate phosphorylation of a PKC substrate as the memory step in preconditioning.

A similar behavior is seen for tyrosine kinase activation in rabbit hearts. Neither of the tyrosine kinase inhibitors genistein or lavendustin A alone had any effect on infarct size in rabbits, and when infused to bracket the preconditioning ischemia each failed to block protection. However, when present at the onset of the index ischemia both genistein and lavendustin A completely aborted preconditioning’s protection [141]. Thus both PKC and protein tyrosine kinase can be defined as mediators and not triggers of cardioprotection.

The time for activation of p38 is not clear. Timing studies in isolated rat hearts indicate that the index ischemia rather than the preconditioning ischemia is the critical time for SB203580 to block protection [174]. Studies in isolated myocytes also suggest that PKC activation precedes p38 activation [172]. However, Maulik et al. [160] noted that pretreatment of rat hearts with SB 203580 before the preconditioning ischemia effectively blocked protection. A recent study in dogs concluded that activation of p38 is important only during the preconditioning ischemic period. In that study blockade of p38 with SB 203580 abolished cardioprotection only when given prior to the preconditioning, but not when given prior to the index ischemia [166]. According to these results, activation of p38 during ischemic preconditioning would act as a trigger rather than mediator, and would be expected to be located upstream of PKC.

4 Potential end-effectors of ischemic preconditioning

4.1 Energetic and substrate metabolism

Apart from infarct size reduction, ischemic preconditioning reduces energy demand during the index ischemia in dogs [181]. Slowing of ATP consumption in preconditioned hearts was confirmed by NMR studies in pig hearts [15,182]. Studies in rat hearts, however, yielded diverse results, with preconditioning either slowing [183–186], accelerating [187–190] or having no effect [191,192] on the rate of ATP consumption during the index ischemia. The differences between these studies in rats relate to the preconditioning procedure. In studies with one or two cycles of preconditioning ischemia, ATP consumption is slowed [183–186], while more than two preconditioning cycles abolish the effect on ATP consumption during the index ischemia. Regardless of changes in the rate of ATP consumption during the index ischemia, left ventricular developed pressure recovered faster in preconditioned than control hearts. These data suggest that preservation of high energy stores during the index ischemia is not a prerequisite for the cardioprotection achieved by ischemic preconditioning.

Also, a slower accumulation of ischemic catabolites, such as lactate, has been demonstrated in preconditioned hearts ([181,185,193,194]; for review, see Ref. [195]). Most importantly, glycogen depletion by the preconditioning ischemia has been proposed to be involved in cardioprotection [196,197]. However, other reports have cautioned that glycogen depletion alone did not protect the heart against ischemic injury [189,198]. Furthermore, pharmacological preconditioning of the heart gives full protection with no glycogen depletion [198].

Apart from the alterations in energy and substrate metabolism, ischemic preconditioning reduces intracellular acidification during the index ischemia. This reduced intracellular acidification most likely reflects decreased anaerobic glycolysis, since blockade of the increased proton efflux by a sodium–proton exchange inhibitor did not alter ischemic acidification [199]. Also, ischemic preconditioning delays the detrimental rise in intracellular sodium [200] and free calcium [200–203] caused by subsequent index ischemia. Both reduced acidification and the slower rise in intracellular sodium and calcium could contribute to the protection achieved by ischemic preconditioning, insofar as activation of potentially detrimental enzymes depends on the intracellular pH and calcium concentration [204,205]. Also, the resolution of ionic disturbances during the reperfusion period following the index ischemia was accelerated in pharmacologically preconditoned rat hearts [206].

4.2 Protein synthesis

Whether or not protein synthesis is required to mediate ischemic preconditioning’s protection is controversial. In the first description in anesthetized rabbits, protein synthesis, either at the translational level (mRNA–protein) or at the transcriptional level (DNA–mRNA), was blocked with cycloheximide and actinomycin D, respectively. Neither substance given intravenously 30 min prior to the preconditioning ischemia altered infarct size reduction triggered by ischemic preconditioning [207]. Subsequent studies in isolated rat [208] and rabbit hearts [209] also failed to demonstrate an effect of transcriptional blockade with actinomycin D on preconditioning’s protection, but clearly demonstrated that blockade at the translational level with cycloheximide abolished ischemic preconditioning’s salutary effect on infarct size reduction. The only striking difference between the two latter studies and the initial report is the 10-fold lower concentration of cycloheximide used in the two later investigations.

4.3 KATP channels

Two distinct populations of KATP channels exist in cardiomyocytes. The sarcolemmal and mitochondrial channels have different reactivities and different properties. Whereas pinacidil opens and glibenclamide closes both channels at the concentrations generally employed, nicorandil [210] and diazoxide [211,212] are believed to primarily open mitochondrial channels. 5-hydroxydecanoate (5HD) is a potent closer of mitochondrial channels ([213]; for review, see Ref. [214]), while HMR 1098 or 1883 closes only surface channels [215–217]. These drugs have been used as tools to explore the role of KATP channels, specifically mitochondrial channels, in preconditioning, and conclusions are necessarily dependent on the avowed specificity of the agents.

It is still a matter of debate which channel is important for the cardioprotection achieved by ischemic preconditioning (for review, see Refs. [214,218–221]). In some studies in rat and rabbit hearts [213,222], pretreatment with diazoxide protected the myocardium from ischemia, while in another study in rabbits diazoxide failed to reduce infarct size [223]. 5HD [213,224] but not HMR 1883 [225] blocks protection by ischemic preconditioning in rats and rabbits, suggesting that the mitochondrial but not sarcolemmal KATP channel is responsible for preconditioning’s protection. In dogs, blockade of both channels is required to completely abolish the infarct size reduction of ischemic preconditioning [226].

Some of the controversy may be the result of incorrect assumptions about the specificity of the channel openers and closers. In fact, recent data indicate that both diazoxide and 5HD may not be as selective as previously thought. Diazoxide, the alleged mitochondrial channel opener at the concentrations employed in most studies, can open sarcolemmal KATP channels in the presence of elevated ADP [227], and diazoxide-mediated protection can be prevented by either HMR 1098 [218] or HMR 1883 [228], putative sarcolemmal channel closers. Furthermore, 5HD, a drug heretofore felt to close only mitochondrial KATP channels, prevented the ischemia-induced shortening of action potential duration which is associated with opening of sarcolemmal KATP channels [229].

Another ongoing debate relates to the question of whether activation of the mitochondrial KATP channel is a trigger rather than an end-effector of preconditioning’s protection [222,230]. In rabbit hearts, 5HD blocks protection induced by either ischemia or diazoxide only when it is present during the preconditioning ischemic period or the diazoxide infusion. When 5HD is present only during the index ischemia, protection persists. Both genistein and an antioxidant block protection from diazoxide, suggesting that mitochondrial KATP channel opening might be acting as a signal transduction element rather than serving as end-effector [222]. In support of this hypothesis, diazoxide-induced opening of KATP channels increased the formation of reactive oxygen species in isolated adult rat cardiomyocytes [231]. Again in support of this hypothesis, PKC antagonists reportedly block diazoxide’s protection in the isolated rat heart [232]. Although these data would support the role of the mitochondrial KATP channel as trigger of preconditioning, other compelling data do not support this hypothesis: the protection induced by the PKC activator PMA in isolated rabbit hearts was abolished by 5HD [233]. Also, 5HD blocked the protective effect of the MAPK activator anisomycin in isolated rabbit hearts [154]. Finally, the KATP channel blocker glibenclamide infused after preconditioning but before the index ischemia eliminated protection in the dog heart [234]. More recently, Fryer and coworkers found attenuated infarct size reduction by ischemic preconditioning in anesthetized rats, when 5HD was given during the index ischemia only, but also when given only during the preconditioning stimulus, suggesting that KATP channels serve both a trigger and effector function [235].

Finally, the preeminent question of how opening of either the sarcolemmal or the mitochondrial KATP channel could directly protect the heart remains. Opening of sarcolemmal KATP channels increases the outward potassium current, resulting in a decrease in the duration of the action potential [236,237] which, in turn, might decrease the inward calcium current through the voltage-dependent calcium channels [238]. Decreased intracellular calcium overload then would reduce ischemic injury [239]. However, when changes in action potential duration were blocked with dofetilide, an antagonist of yet another potassium channel, preconditioning’s protection was not affected [240].

Two hypotheses have been proposed to explain how the opening of mitochondrial KATP channels could be potentially cardioprotective. The first hypothesis refers to mitochondrial calcium handling, insofar as opening of the mitochondrial KATP channels would decrease mitochondrial calcium overload, by reducing calcium uptake and increasing calcium release, thereby preserving the integrity of mitochondria (for review, see Ref. [220]). The second hypothesis relates to changes in mitochondrial volume [241], which is regulated by opening of mitochondrial KATP channels. These volume changes are important, insofar as they affect energy flow through the electron transport system and the architecture of the intermembrane space, thereby influencing efficient energy transfer between mitochondria and cellular ATPases [241,242]. Because the mitochondrial KATP channel may well not be the end-effector at all but rather a signal transduction element (see above), one must be receptive to other suggestions.

4.4 Sodium–proton exchanger

Xaio and Allen [243] have suggested that the sodium–proton exchanger might be the end-effector of ischemic preconditioning, and that its inhibition might lead to protection of the ischemic heart. They noted that the sodium–proton exchanger appeared to be blocked at reperfusion only when rat hearts had been preconditioned. Addition of HOE 642, a highly selective blocker of the sodium–proton exchanger, shortly before reperfusion to a non-preconditioned heart preserved postischemic function by an amount equal to that achieved by ischemic preconditioning. HOE 642 had no additive effect when combined with ischemic preconditioning, further suggesting a common mechanism. Similarly, 5-(N-ethyl-N-isopropyl)amiloride, another blocker of the sodium–proton exchanger, limited infarct size in rabbits by an amount equal to that of preconditioning [244], and amiloride could not augment preconditioning’s anti-infarct effect [245]. Neither kinase inhibitors nor 5HD could abolish amiloride’s protection, suggesting that the sodium–proton exchanger must be operative at a very distal point in the signaling pathway. Adenosine is reported to block the exchanger [246], and this is consistent with the previously demonstrated ability of adenosine to pharmacologically precondition the heart (see above). Finally in dogs, blockade of the sodium–proton exchanger limits infarct size [247], and this reduction in infarct size is not attenuated by either blockade of KATP channels or an adenosine receptor antagonist [248].

In dogs, ischemic preconditioning and blockade of the sodium–proton exchanger are equally effective in reducing infarct size following an index ischemia of 60 min duration, while with prolongation of the index ischemia to 90 min, blockade of the sodium–proton exchanger reduces infarct size to a greater extent than ischemic preconditioning [249]. Also, activation of PKC, which is involved in preconditioning (see above), activates rather than blocks the sodium–proton exchanger [250]. These data do not support a role of the sodium–proton exchanger as the end-effector of ischemic preconditioning’s protection, but rather suggest distinct pathways and mechanisms (for review, see Ref. [251]).

4.5 Cytoskeleton

MAPKAP kinase 2 phosphorylates the small heat shock protein HSP27 [252–254], which appears to be regulated by its phosphorylation state. HSP27 acts as a molecular chaperone and binds unfolded proteins, preventing nonspecific aggregation of proteins [255]. HSP27 also regulates actin filament organization. In HSP27-transfected cells heat shock-induced disassembly of actin filaments is prevented [256]. Furthermore, HSP27 stabilizes actin filaments during oxidative stress [257].

Prolonged ischemia is known to cause cytoskeletal disruption [258]. Overexpression of HSP27 confers protection against ischemia in isolated rat cardiac myocytes [256]. In Langendorff-perfused rat hearts, ischemic preconditioning induces redistribution of HSP27 from the cytosol to the sarcomere, and the recovery of global contractile function following 40 min of global ischemia is significantly enhanced [259]. Similarly, ischemic preconditioning induces redistribution of HSP27 from the cytosol to the sarcomere in dog hearts in vivo [166]. This redistribution of HSP27 is abolished by pretreatment with SB 203580, supporting involvement of p38 activation [260]. Furthermore, in isolated cardiomyocytes subjected to osmotic stress and hypoxia, cytochalasin D, a drug known to disrupt the cytoskeleton, blocks the protective effect of ischemic preconditioning or diazoxide against increasing osmotic fragility [154].

Therefore, redistribution of heat shock proteins from the cytosol to the sarcomere and the subsequent alteration in the stability of the cytoskeleton appear to be important for ischemic preconditioning’s cardioprotection, although further studies in vivo using more specific drugs than cytochalasin D are required to corroborate this concept.

4.6 Volume regulation

The osmolarity of body fluids is strictly controlled so that cells do not experience changes in osmotic pressure under normal conditions. Osmotic balance is maintained by matching the osmotic pull of proteins and nucleotides within the cell by sodium outside the cell. Because the conductance of the sarcolemma to sodium is very low extracellular sodium is an efficient osmolyte. During ischemia ATP is broken down to AMP and two inorganic phosphates thus tripling the osmotic pull. Similarly, failure of the sodium/potassium pumps leads to buildup of intracellular sodium and thus a collapse of the sodium gradient. Each increase in osmolyte concentration of 1 mM above the baseline intracellular osmotic pressure of 5430 mmHg exerts an additional pull of 19 mmHg [261]. After 30 min of ischemia myocardial cells of an isolated rat heart accumulate about 20 mM of sodium [262]. That translates to an increase in transmembrane osmotic pressure from the sodium alone of 380 mmHg. Breakdown of ATP and creatine phosphate also would contribute. This osmotic imbalance results in swelling of the ischemic cells. This swelling may be exacerbated further on reperfusion when the hyperosmotic extracellular milieu is replaced by isoosmotic perfusate. In the end the cell bursts from osmotic forces. Preconditioning makes cardiomyocytes very resistant to membrane failure when they are challenged with hypotonic media [263]. In ischemically preconditoned rat [264] and pig [265] hearts, the extent of myocardial edema formation — along with infarct size — is reduced. Alterations in channels involved in cell volume regulation therefore might be involved in the cardioprotection achieved by ischemic preconditioning. Chloride channels are involved in moment to moment volume regulation, and Diaz et al. [266] hypothesized from experiments in rabbit cardiomyocytes and isolated hearts that opening of swelling-induced chloride channels was responsible for the protection achieved by ischemic preconditioning. However, their electrophysiological and infarct size data could not be confirmed [267]. There are also organic osmolytes, such as taurine or sorbitol, which are involved in ischemia/reperfusion damage. In anesthetized rats, taurine depletion indeed reduces infarct size following ischemia/reperfusion, with the relationship between myocardial taurine content and infarct size being almost linear [268]. Similarly, depletion of myocardial sorbitol content by inhibition of aldose reductase attenuates infarct size in isolated rabbit hearts [269]. Whether ischemic preconditioning alters myocardial taurine or sorbitol contents is unclear at present.

The sodium–hydrogen exchanger may also provide an avenue for sodium entry into the ischemic cell. Although most theories suggest that calcium entry might be the ultimate consequence of sodium entry (via the Na+/Ca2+ exchanger) the osmotic effects of sodium entry should not be overlooked. If the sodium–hydrogen exchanger turns out to be the end-effector of preconditioning and its inhibition the final step in the cascade leading to protection, then the principal effect of its inhibition may be prevention of swelling rather than calcium entry.

Connexins are distributed throughout the heart [270], and connexin 43 forms channels which are involved in volume regulation during changes in the intra/extracellular milieu [271]. During acute [272] and chronic [273] myocardial ischemia, as well as during inflammation [274], heterogeneous loss of connexin 43 occurs, thereby potentially affecting cell volume regulation during osmotic stress. However, whether ischemic preconditioning alters myocardial connexin concentration or its distribution is unknown at present. Taken together, an alteration in the osmotic load might be one potential way of reducing myocardial ischemia/reperfusion injury. The importance of osmotic load for ischemic preconditioning, however, remains to be established.

4.7 Tumor necrosis factor (TNF) α

TNFα concentrations are increased during hypoxia [275] and ischemia [276,277]. In patients, monoclonal antibodies against TNFα reduce the extent of irreversible tissue damage during acute myocardial infarction, suggesting a deleterious role of TNFα in the scenario of ischemia/reperfusion [278]. Ischemic preconditioning attenuates infarct size and the ischemia-induced increases in the serum and myocardial TNFα-concentration in rabbits in vitro [276] and in vivo [277]. Therefore, cardioprotection by ischemic preconditioning might involve downregulation of TNFα, although a cause–effect relationship remains to be established in future experiments.

5 Differences between the early and the late phase of ischemic preconditioning

While the first window or early phase of protection is short-lived, a second window or late phase of protection appears about 24 h after ischemic preconditioning [8,279,280], and this latter anti-infarct effect is believed to last for up to 3 days [281]. In contrast to the early phase of ischemic preconditioning, the late phase of protection includes a strong anti-stunning effect [279]. Although the signal transduction pathway leading to protection in the late phase shares many of the steps identified for early phase protection, there are certain differences between both phenomena (for review, see Refs. [279,280]). For example, endogenous nitric oxide is a key trigger and mediator of the late phase of ischemic preconditioning [282], and the late phase of protection clearly results from altered expression of protective proteins.

6 Potential reasons for current controversies

The issue of endpoints was already discussed in Section 1.

6.1 Species

Ischemic preconditioning protects against infarction in all species tested so far (for review, see Ref. [8]); however, the signal transduction — as clarified so far — varies among different species. For example the functional role for PKC isozymes differs among species, with PKCε being most important in rabbits, PKCδ in rats and PKCα in dogs (see above). The question of which species is the closest to humans can obviously not be answered, although it appears that large mammals may have an advantage.

6.2 Models

A variety of different models has been used to study preconditioning’s effect, such as neonatal and adult cardiomyocytes, in vitro isolated, buffer- or blood-perfused hearts or in vivo anesthetized or conscious animals. The signaling system is inherently different among those models due to the specific conditions of the preparations. For example, free radicals are certainly less important in blood-perfused in situ systems than in oxygenated buffer-perfused preparations.

Also, age appears to be an important issue. During postnatal development the expression of protein kinase C declines [283], MAP kinases remain unaltered [284], and free radical production is increased [285]. Therefore, the sensitivity towards a certain signaling cascade will change during development. Indeed, rat hearts become more vulnerable to ischemia with advancing age and the beneficial effect of ischemic preconditioning, which was seen in hearts of young rats, appears to be reversed in middle-aged rat hearts; i.e., creatine kinase release increased rather then decreased in aged preconditioned hearts [286].

6.3 Protocols

A variety of protocols has been used to precondition the heart, including single vs. multiple cycles of ischemia/reperfusion, short vs. long durations of the preconditioning ischemia, low-flow vs. no-flow ischemia and regional vs. global ischemia. Thus, the intensity of the preconditioning stimulus varies among most experimental studies, which in itself will lead to differences in the triggers involved (e.g., bradykinin vs. adenosine, see above). We now know that there is a large degree of redundancy in the preconditioning system. For example, at least three receptors are involved as preconditioning triggers. Antagonists to a single receptor do not abolish protection, but rather raise the threshold for a protective effect. Thus one investigator preconditioning hearts with multiple cycles might report that a given receptor blocker fails to abolish protection, while another using a single cycle of ischemia to precondition might see abolition of protection with the same drug. Both observations are valid, and confusion is unavoidable. The same seems to occur with the kinases. Activation of kinases appears to be very protocol-dependent. With multiple cycles of ischemia/reperfusion PKC is further upregulated, while p38 MAP kinase is decreased [287], and the relative importance of tyrosine kinases is increased (see above).

None of the studies so far has completely analyzed all identified targets of preconditioning’s signal transduction cascade within a given experimental protocol.

6.4 Inhibitors and activators

Pharmacological agents are valuable tools in delineating the mechanism(s) of ischemic preconditioning. However, the results obtained are influenced by the effectiveness and specificity of the drugs for a given target at a given concentration. As outlined above, most pharmacological agents lack such specificity (e.g., diazoxide, SB203580), especially when given at high enough concentrations to be effective. Furthermore, most kinase inhibitors lack specificity for a certain isoform which may have opposing functional roles (e.g., apoptotic effect of p38α and p38β, see above).

6.5 Transgenic and transfection approaches

With transgenic and transfection approaches, intra-species differences must be considered. For example, infarct size following 30 min of ischemia and 24 h of reperfusion is highly strain-dependent in mice [288] and rats [289]. Therefore, transgenic and transfected mice must always be compared to controls from the respective wild-type strain.

Apart from the problems of quantifying the rate of transfection, overexpression of a certain protein might cause loss of its selectivity. For example, the RACK for PKCε was thought to be RACK 2. However, at high levels of expression PKCε interacts with both RACK 2 and RACK 1 [126].

Another emerging facet of signaling systems is the crosstalk between different pathways [290], suggesting that myocardial signaling systems, rather than following the linear signaling paradigm postulated for ischemic preconditioning, are instead highly interactive. This is especially of importance in transgenic animals, as the knockout of one gene might up-regulate another gene, or overexpression of one protein (by transfection) might downregulate others.

7 Summary

Following 15 years of research, parts of the signal transduction cascade of ischemic preconditioning have been identified, and there is good agreement on the major endogenous triggers (adenosine, bradykinin, opioids, free radicals) involved in ischemic preconditioning, although, of course, the importance of different triggers varies among different species. Agreement exists also on the involvement of certain protein kinases, such as protein kinase C or protein tyrosine kinases, although the sequence of their activation (in series or parallel) appears to be once more species- and model-dependent. However, the importance of other elements of the downstream signal cascade, such as the MAP kinases and the ATP-dependent potassium channels, is still incompletely understood and the end-effector of preconditioning’s protection is still unknown. Potential candidates of interest already exist (cytoskeleton or volume control), and the importance of novel proteins [291], such as desmin and villin, or kinases, such as Akt [291], in cardioprotection needs to be defined. A better understanding of the different elements within the signal transduction cascade and their hierarchic order is of utmost importance to potentially utilize the phenomenon of preconditioning in the clinical setting in the future.


  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]
  65. [65]
  66. [66]
  67. [67]
  68. [68]
  69. [69]
  70. [70]
  71. [71]
  72. [72]
  73. [73]
  74. [74]
  75. [75]
  76. [76]
  77. [77]
  78. [78]
  79. [79]
  80. [80]
  81. [81]
  82. [82]
  83. [83]
  84. [84]
  85. [85]
  86. [86]
  87. [87]
  88. [88]
  89. [89]
  90. [90]
  91. [91]
  92. [92]
  93. [93]
  94. [94]
  95. [95]
  96. [96]
  97. [97]
  98. [98]
  99. [99]
  100. [100]
  101. [101]
  102. [102]
  103. [103]
  104. [104]
  105. [105]
  106. [106]
  107. [107]
  108. [108]
  109. [109]
  110. [110]
  111. [111]
  112. [112]
  113. [113]
  114. [114]
  115. [115]
  116. [116]
  117. [117]
  118. [118]
  119. [119]
  120. [120]
  121. [121]
  122. [122]
  123. [123]
  124. [124]
  125. [125]
  126. [126]
  127. [127]
  128. [128]
  129. [129]
  130. [130]
  131. [131]
  132. [132]
  133. [133]
  134. [134]
  135. [135]
  136. [136]
  137. [137]
  138. [138]
  139. [139]
  140. [140]
  141. [141]
  142. [142]
  143. [143]
  144. [144]
  145. [145]
  146. [146]
  147. [147]
  148. [148]
  149. [149]
  150. [150]
  151. [151]
  152. [152]
  153. [153]
  154. [154]
  155. [155]
  156. [156]
  157. [157]
  158. [158]
  159. [159]
  160. [160]
  161. [161]
  162. [162]
  163. [163]
  164. [164]
  165. [165]
  166. [166]
  167. [167]
  168. [168]
  169. [169]
  170. [170]
  171. [171]
  172. [172]
  173. [173]
  174. [174]
  175. [175]
  176. [176]
  177. [177]
  178. [178]
  179. [179]
  180. [180]
  181. [181]
  182. [182]
  183. [183]
  184. [184]
  185. [185]
  186. [186]
  187. [187]
  188. [188]
  189. [189]
  190. [190]
  191. [191]
  192. [192]
  193. [193]
  194. [194]
  195. [195]
  196. [196]
  197. [197]
  198. [198]
  199. [199]
  200. [200]
  201. [201]
  202. [202]
  203. [203]
  204. [204]
  205. [205]
  206. [206]
  207. [207]
  208. [208]
  209. [209]
  210. [210]
  211. [211]
  212. [212]
  213. [213]
  214. [214]
  215. [215]
  216. [216]
  217. [217]
  218. [218]
  219. [219]
  220. [220]
  221. [221]
  222. [222]
  223. [223]
  224. [224]
  225. [225]
  226. [226]
  227. [227]
  228. [228]
  229. [229]
  230. [230]
  231. [231]
  232. [232]
  233. [233]
  234. [234]
  235. [235]
  236. [236]
  237. [237]
  238. [238]
  239. [239]
  240. [240]
  241. [241]
  242. [242]
  243. [243]
  244. [244]
  245. [245]
  246. [246]
  247. [247]
  248. [248]
  249. [249]
  250. [250]
  251. [251]
  252. [252]
  253. [253]
  254. [254]
  255. [255]
  256. [256]
  257. [257]
  258. [258]
  259. [259]
  260. [260]
  261. [261]
  262. [262]
  263. [263]
  264. [264]
  265. [265]
  266. [266]
  267. [267]
  268. [268]
  269. [269]
  270. [270]
  271. [271]
  272. [272]
  273. [273]
  274. [274]
  275. [275]
  276. [276]
  277. [277]
  278. [278]
  279. [279]
  280. [280]
  281. [281]
  282. [282]
  283. [283]
  284. [284]
  285. [285]
  286. [286]
  287. [287]
  288. [288]
  289. [289]
  290. [290]
  291. [291]
View Abstract