1 Introduction

In 1997 we [2] as well as Tritto et al. [25] reported that reactive oxygen species (ROS) were involved in the triggering of ischemic preconditioning (IPC). Later it was found that ROS were generated as a result of opening of mitochondrial ATP-sensitive potassium channels (mKATP) and acted as second messengers in a pathway ending in activation of PKC, which triggered the preconditioned state [21]. In those studies ROS were recognized as part of the trigger pathway that preconditions the heart prior to the onset of the index ischemia. Several years ago it was discovered that IPC actually exerts its protective effect in the first minutes of reperfusion [9] presumably by inhibiting the formation of mitochondrial permeability transition pores (mPTP) [11]. Then it was found that redox signaling is part of the pathway that mediates IPC’s protection at the time of reperfusion as well. Penna et al. [22] reported that ROS generation was required for ischemic postconditioning to be protective. The finding that ROS could carry a protective signal at the same time that ROS were thought to be inducing reperfusion injury was very surprising [7]. But then Hausenloy et al. [10] reported that a ROS scavenger given immediately following the index ischemia also blocked IPC’s protection. These observations again illustrated similarities of the mechanisms of protection of IPC and postconditioning.

It is unknown how redox signaling acts to mediate protection at reperfusion. A number of signal transduction components have been identified in IPC’s mediator pathway including phosphatidylinositol (PI) 3-kinase and extracellular signal-regulated kinase (ERK) [9]. Additionally PKC must be activated at reperfusion [10, 16, 22] and adenosine A2b receptors must be occupied [16]. Hausenloy and colleagues [10] have speculated that the order of these components was first PKC activation followed in order by occupancy of adenosine receptors, activation of the survival kinases PI3-kinase and ERK, opening of mKATP, and generation of ROS and finally ending with inhibition of mPTP formation. In the present study we tested that order by directly activating PKC with phorbol 12-myristate 13-acetate (PMA), which we have previously shown mimics IPC when given at reperfusion [16] in a PKC-dependent fashion [23]. If redox signaling is downstream of PKC as Hausenloy et al. [10] have proposed, then a ROS scavenger should block protection resulting from PKC activation.

2 Methods

2.1 Surgical preparation

All animal care satisfied published guidelines [20] and procedures were approved by institutional committees. Briefly, New Zealand White rabbits were anesthetized with pentobarbital sodium (30 mg/kg i.v.) and ventilated with 100% oxygen. The heart was exposed and a suture was passed around a coronary arterial branch. The heart was excised and perfused on a Langendorff apparatus with Krebs-Henseleit bicarbonate buffer bubbled with 95% O2/5% CO2 to a pH of 7.35–7.45 at 38°C. A fluid-filled latex balloon measured pressure in the left ventricle.

2.2 Experimental protocol

Hearts of 6 experimental groups underwent 30-min coronary branch occlusion/2-h reperfusion. Control hearts had no other intervention. IPC hearts were preconditioned with 5-min global ischemia/10-min reperfusion prior to the index ischemia. In group 3 IPC hearts received N-(2-mercaptopropionyl) glycine (MPG, 300 µM) for 20 min starting 5 min before reperfusion. In the fourth group MPG was given in the same dose and schedule to non-preconditioned hearts. In the fifth group PMA (0.05 nM) was given from 1 min before to 5 min after reperfusion, and in group 6 PMA was similarly administered in addition to MPG as in groups 3 and 4. The dose of PMA was selected empirically as one that was known to confer protection [18]; and because that protection could be blocked by chelerythrine [16] the dependence of PMA’s protective effect on PKC was assured. It should be noted that group 4, a control group of MPG alone, was not contemporary but rather was from a previous study in our laboratory [5].

2.3 Measurement of risk zone and infarct size

At the end of the experiments the coronary artery was reoccluded, and 2–9 µm fluorescent microspheres (Microgenics Corp., Freemont, CA) were infused to demarcate the non-fluorescent risk zone. The heart was cut into 2-mm-thick slices that were incubated in 1% triphenyltetrazolium chloride to stain noninfarcted tissue. Areas of infarct and risk zone were determined by planimetry of each slice and volumes calculated. Infarction is expressed as a percentage of the region at risk.

2.4 Statistics

Infarct sizes were compared with one-way ANOVA, and post hoc pairwise comparisons done with Student–Newman–Keuls test. P < 0.05 was considered to be significant.

3 Results

Hemodynamics are presented in Table 1. There were no significant baseline group differences. Coronary occlusion caused significant decreases in left ventricular developed pressure and coronary flow in all groups. There was partial recovery following reperfusion.

Table 1 Hemodynamic data
Full size table

Infarct size data appear in Fig. 1. Thirty minutes of regional ischemia and 2 h of reperfusion resulted in 17.6 ± 1.7% infarction of the risk zone in IPC hearts, which was significantly smaller than 36.4 ± 1.7% infarction seen in non-preconditioned hearts (P < 0.001). When IPC hearts were reperfused in the presence of MPG, infarct size increased to 32.6 ± 3.4% of the risk zone indicating that IPC’s protection had been blocked. MPG in a non-preconditioned heart had no effect on infarct size (37.0 ± 2.3% infarction). The pulse of PMA fully protected the hearts (6.8 ± 1.0% infarction) and MPG did not block PMA’s protection as infarct size was only 9.2 ± 1.5% of the risk zone. Interestingly infarct size in the two PMA groups was significantly smaller than that in the IPC group (P < 0.002).

Fig. 1
figure 1

Infarct size as a percentage of risk zone for individual hearts and group means with SEM. Abbreviations: IPC = ischemic preconditioning; MPG = N-(2-mercaptopropionyl) glycine; PMA = phorbol 12-myristate 13-acetate *P < 0.001 versus control

Full size image

4 Discussion

We were able to confirm the observation of Hausenloy et al. [10] that ROS production early in reperfusion is required for IPC’s protection in our rabbit model. Our data also indicate that the protection from a direct PKC activator is independent of ROS production suggesting that the ROS step occurs upstream of the PKC step. Through the use of inhibitors PKC activity early in reperfusion has been shown to be required for protection against infarction from both IPC [10, 16] and postconditioning [22, 23]. Furthermore, the direct PKC activator PMA given at reperfusion mimics IPC’s protection [16, 23] and this protection is blocked by chelerythrine verifying that PMA protected in a PKC-dependent manner [23].

Our previous data indicate that the adenosine A2b receptor resides downstream of PKC in the mediator pathway because at reperfusion an A2b receptor antagonist blocks protection from PMA but the PKC blocker chelerythrine does not block protection from an A2b agonist [16, 23]. We found that PKC activity lowers the threshold for A2b agonists to activate the survival kinases suggesting that in the IPC heart the increased sensitivity allows endogenous adenosine to activate protective signaling through the normally low-affinity A2b receptors [16]. Accordingly we were able to demonstrate that PI3-kinase and ERK activity is dependent on population of adenosine receptors since receptor inhibition at reperfusion blocked survival kinase activation in IPC hearts [24]. Also, protection from PMA at reperfusion could be blocked by the PI3-kinase blocker wortmannin [23]. Taken together this information now allows us to propose a revised arrangement for the sequence of signal transduction elements for the mediator pathway. In our new order we start with ROS followed in sequence by PKC, A2b receptors, and survival kinases leading to inhibition of mPTP (Fig. 2).

Fig. 2
figure 2

Flowchart originally presented by Hausenloy et al. [10] and our revision based on our new data in Fig. 1

Full size image

Costa et al. [6] found that mKATP openers protected isolated mitochondria from Ca++-induced mPTP formation. That protection was both ROS- and PKC-sensitive. Furthermore they could show that ROS protected by activation of PKC. Because these were isolated mitochondria they concluded that once the mKATP opened all of the signal transduction elements required to inhibit mPTP formation were present within the mitochondria. The present findings are in agreement with some of the observations made by Costa et al. [6] Protection from PMA does not require ROS and is therefore downstream of the ROS signal. However, PMA’s protection requires both adenosine receptor occupancy as well as PI3-kinase activation [23] indicating that in a whole heart protection from a direct PKC activator still requires signals that arise from outside the mitochondria. We have no explanation for this apparent discrepancy.

The protection from PMA was significantly more potent than that from IPC. While IPC is thought to selectively activate a single isoform of PKC (perhaps ɛ), PMA is not selective for any isoform and may have activated additional isoforms to provide more protection. Alternatively the single IPC protocol we used may not have fully preconditioned our hearts. PMA also reduced coronary flow and contractility. The reduced coronary flow could have ischemically preconditioned the hearts, but if that were the case MPG should have blocked it. Although PMA was present in the perfusate for about a minute before reperfusion it would not have appreciably entered the ischemic tissue until the artery was reperfused because of sparse collateralization in the rabbit heart.

Our data do not provide information about the location of the mKATP channel in the signaling cascade. In a recent study Gross et al. [8] gave the mKATP opener BMS-191095 just prior to reperfusion in open-chest rats and were able to mimic IPC’s anti-infarct effect. Moreover they could block BMS-191095’s protection with the PI3-kinase inhibitor wortmannin thus putting mKATP ahead of PI3-kinase. The most likely place for mKATP to reside would be ahead of the ROS step. IPC’s trigger pathway has been extensively studied and in that pathway we [15, 21] and others [1] have found that potassium entry from mKATP opening results in ROS production by the mitochondria. Thus the mediator pathway could simply look like IPC’s trigger pathway with mKATP opening resulting in ROS production. This suggestion does not fit all of the data, however. In a cardiomyocyte model Lebuffe et al. [17] found that they could protect cells against simulated ischemia with either PMA or hydrogen peroxide and that the mKATP blocker 5-hydroxydecanoate blocked protection from both. That would put the mKATP channel downstream of ROS and PKC. Juhaszova et al. [13] also were confronted with discrepant data about the location of mKATP in their flowchart and decided that the data warranted a dual location, one upstream of the ROS step and another downstream near mPTP, and we have also done that in Fig 2.

How do we explain the apparent coexistence of a protective free radical signal and the very injurious free radical burst that is supposed to occur in the reperfused heart [3]? Most observers have noted that IPC does not increase but actually reduces overall ROS production in reperfused hearts [26]. The most likely explanation is that the ROS signal is carried by one specific radical species and/or in one intracellular compartment. MPG obviously is able to scavenge the radical species that signals the protection. MPG is a very selective scavenger for hydroxyl radical and peroxynitrite and does not affect superoxide or hydrogen peroxide [4]. We also have unpublished measurements showing that MPG does not scavenge nitric oxide. In addition MPG reportedly concentrates 500-fold in mitochondria making it highly effective against mitochondria-generated ROS [19]. Protection from preconditioning with a ROS generator is known to be PKC-dependent [2, 25], and indeed Korichneva et al. [14] found that ROS could directly activate PKC in vitro by reacting with its thiol groups. Thus hydroxyl radical made by mitochondria could directly activate local PKC independent of the other injurious ROS species generated during reperfusion. The simplest explanation for our observation is that hydroxyl radical made by the IPC heart’s mitochondria early in reperfusion protects by activating PKC. It is currently unknown why that activation fails to occur in the non-IPC heart.