The aim of this study was to examine the hypothesis that δ-opioid receptor activation before ischemia suppresses gap junction (GJ) permeability by PKC-mediated connexin 43 (Cx43) modulation, which contributes to infarct size limitation afforded by the δ-opioid receptor activation. A δ-opioid receptor agonist, [d-Ala2,d-Leu5]-enkephalin acetate (DADLE, 300 nM), was used in place of preconditioning (PC) ischemia to trigger PC mechanisms in rat hearts. GJ permeability during ischemia, which was assessed by Lucifer yellow, was reduced by DADLE to 47% of the control level, and this effect of DADLE was almost abolished by a PKC-ε inhibitor [PKC-ε translocation inhibitory peptide (PKC-ε-TIP)] but was not affected by a PKC-δ inhibitor (rottlerin). After DADLE infusion, PKC-ε, but not PKC-δ, was coimmunoprecipitated with Cx43, and the level of phosphorylation of Cx43 at a PKC-dependent site (Ser368) was significantly elevated during ischemia. DADLE reduced infarct size after 35 min of ischemia followed by 2 h of reperfusion by 69%, and PKC-ε-TIP and rottlerin eliminated 48% and 63%, respectively, of the infarct size-limiting effect of DADLE. Infusion of a GJ blocker, heptanol, before reperfusion reduced infarct size by 36%, and this protection was not enhanced by preischemic infusion of rottlerin + DADLE, which allows PKC-ε activation by DADLE. These results suggest that phosphorylation of Cx43 by PKC-ε plays a crucial role in δ-opioid-induced suppression of GJ permeability in ischemic myocardium and that this modulation of the GJ is possibly an adjunct mechanism of infarct size limitation afforded by preischemic δ-opioid receptor activation.
- infarct size
preconditioning (PC) with brief ischemia substantially salvages the myocardium from necrosis after sustained ischemia and reperfusion (22, 35). This cardioprotective effect of PC has been demonstrated not only in intact hearts, but also in isolated cardiomyocytes subjected to simulated ischemia, indicating that PC can directly protect each cardiomyocyte under ischemia-reperfusion. However, the extent of protection by PC is generally less in isolated cell preparations than in whole hearts (1, 19, 22, 35). Furthermore, there is also evidence to suggest that a part of myocardial protection by PC in whole hearts is achieved by chemical uncoupling of cardiomyocytes during ischemia-reperfusion, resulting in suppressed extension of injury within the area at risk (8, 20, 23, 28). In our recent studies (20, 23), both ischemic PC and an opener of the mitochondrial ATP-sensitive K+ (mKATP) channel, which is involved in the PC mechanism, significantly reduced gap junction permeability to a chemical tracer in the ischemic rabbit myocardium. Inhibition of gap junction communication by pharmacological gap junction blockers infused during ischemia or upon reperfusion has been shown to reduce infarct size in pig, rabbit, and rat hearts, as does ischemic PC (8, 20, 28). These findings support the notion that PC-induced suppression of gap junction permeability contributes to myocardial protection from ischemia-reperfusion injury. However, mechanisms by which PC inhibits gap junction communication and the relative importance of this gap junction-mediated mechanism in overall protection afforded by PC remain unclear.
In the present study, we examined the hypothesis that phosphorylation of connexin 43 (Cx43) by PKC-ε is a primary mechanism of PC-induced suppression of gap junction permeability during ischemia. Rationales for postulating the involvement of PKC-ε in PC-induced gap junction modulation are twofold: 1) the results of our previous study (20) showed that a PKC inhibitor abrogated PC-induced suppression of Cx43 dephosphorylation in the ischemic myocardium, and 2) results of another study (5) showed that PKC-ε physically interacts with Cx43 in response to fibroblast growth factor type 2 receptor activation. Since PC ischemia could dephosphorylate phosphorylated sites in Cx43 as an epiphenomenon (20), we used an agonist of the δ-opioid receptor, a major receptor involved in ischemic PC, to trigger PC mechanisms in the present study.
We also aimed to obtain insight into the contribution of this gap junction-mediated mechanism to infarct size limitation by PC. It is difficult to separately assess each of the contributions of gap junction-mediated and -independent mechanisms to PC, since no agent can selectively reverse chemical uncoupling of the gap junction. Thus we postulated that use of a δ-opioid receptor agonist to trigger the PC mechanism might provide an experimental condition under which we could separate a part of cardioprotection afforded by mechanisms independent of gap junction suppression. PKC-δ has been reported to play a major role in cardioprotection by δ-opioid receptor activation (7), whereas roles of PKC isoforms in ischemic PC are possibly more complicated, because several classes of G protein receptors are simultaneously activated by PC ischemia (3, 17, 31, 34). Thus, if PKC-δ is not responsible for gap junction modulation by PC with δ-opioid receptor activation, protection related and unrelated to modulation of gap junction permeability might be estimated by PKC isoform-selective inhibitors. We used [d-Ala2,d-Leu5]-enkephalin acetate (DADLE), a δ-opioid-selective agonist, to examine these possibilities in isolated buffer-perfused rat hearts.
The results of the present study indicate that δ-opioid receptor activation induces phosphorylation of Cx43 directly by PKC-ε, leading to suppression of gap junction permeability. Suppression of gap junction-mediated propagation of injury was suggested to possibly contribute to a maximum of ∼35% of infarct size limitation afforded by activation of the δ-opioid receptor before ischemia.
This study was conducted in strict accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996) and was approved by the Animal Use Committee of Sapporo Medical University.
Experiment 1: Assessment of Gap Junction Permeability in the Myocardium
Preparation of isolated perfused rat hearts.
Sprague-Dawley rats were anesthetized with pentobarbital sodium (80 mg/kg ip) and ventilated with a rodent respirator using room air and supplemental oxygen. Hearts were isolated and perfused with modified Krebs-Henseleit buffer, and gap junction permeability in the ischemic myocardium was assessed according to the method described by Ruiz-Meana et al. (30), as described in our previous studies (20, 23).
Before global ischemia, hearts received no pretreatment, heptanol (a gap junction blocker), DADLE, a PKC inhibitor + DADLE, or a PKC inhibitor alone (Fig. 1). Heptanol (1 mM) was infused for 5 min before ischemia. DADLE (300 nM) was administered as two 5-min infusions, separated by 5 min, to mimic PC with two cycles of 5 min of ischemia followed by 5 min of reperfusion. As PKC inhibitors, 200 nM calphostin C (a nonselective PKC inhibitor), 100 nM PKC-ε translocation inhibitory peptide (PKC-ε TIP; Calbiochem/EMD Biosciences, Darmstadt, Germany), and 0.1 μM rottlerin (a PKC-δ-selective inhibitor) were used. This dose of rottlerin was selected to avoid substantial hemodynamic effects by higher doses in the present preparation (unpublished data). To exclude the possibility that nonspecific effects of TIP affect the results, we additionally used scrambled TIP, which is a scrambled peptide with an amino acid composition identical to that of PKC-ε-TIP (Calbiochem/EMD), as a negative control for PKC-ε-TIP. These PKC inhibitors and a control peptide were infused for 25 min before ischemia. At 5 min after the onset of global ischemia, a sharp transmural incision was made in the ventricular wall along the long axis with use of a surgical blade, and the ventricle was incubated in anoxic PBS containing 2.5 mg/ml of Lucifer yellow (LY) and 2.5 mg/ml of rhodamine-conjugated dextran (RD) at 37°C for 25 min. At the end of incubation, the ventricular tissue was fixed with 1% glutaraldehyde-4% formaldehyde in 0.2 M cacodylate buffer (pH 7.4).
Microscopy and image analysis.
Fixed hearts were sectioned into four slices parallel to the atrioventricular groove, and slides were prepared for histology according to standard techniques. Images of LY and RD were obtained by a confocal laser microscope, and the areas stained with LY and RD in each section were determined using NIH Image. Since RD staining indicates areas with disrupted sarcolemma, the area of gap junction communication was determined as the area of LY staining without RD staining, as described in previous studies (20, 23, 30). Data for gap junction communication were obtained from four heart slices and averaged for each heart.
Experiment 2: Immunoblotting for Cx43 and PKC
Tissue sampling protocol.
Rat hearts were perfused as described in experiment 1 and divided into three groups: nonpretreated (control), DADLE before global ischemia, and DADLE + PKC-ε-TIP before ischemia. Doses and times of DADLE and PKC-ε-TIP treatment were the same as those described for experiment 1. Ventricular tissues for immunoblotting were sampled under baseline conditions, immediately before ischemia (after pretreatment), and 10 min after ischemia in each heart. Ventricular tissues from separate groups of hearts were sampled after DADLE treatment or after the time control period for immunoprecipitation experiments (see below). Tissues were frozen in liquid nitrogen immediately after sampling and stored at −80°C until analysis.
Immunoblotting for phosphorylation of Cx43 at Ser368.
Intercalated disk-rich fractions were separated from tissue samples as previously reported (20, 23) and electrophoresed on 12.5% polyacrylamide gel and then electroblotted onto polyvinylidene difluoride membranes (Millipore, Bedford, MA). After they were blocked, the blots were incubated with antibodies against Cx43 phosphorylated at Ser368 (Cell Signaling Technologies, Beverly, MA). As a secondary antibody, peroxidase-linked anti-rabbit IgG F(ab′)2 fragments (Amersham Biosciences, Buckinghamshire, UK) were used, and signals of phosphorylated Cx43 were visualized using an ECL Western blotting detection kit (Amersham Biosciences) and quantified using SigmaGel (SPSS, Chicago, IL). Then a Re-blot Western blot recycling kit (Chemicon, Temecula, CA) was used to strip bound antibodies from the polyvinylidene difluoride membranes, and the membranes were blotted with anti-Cx43 antibody (Sigma, St. Louis, MO) for determination of total Cx43 level. The level of Cx43 phosphorylation at Ser368 was normalized to the densitometric value of the corresponding immunoblots for total Cx43.
Immunoprecipitation of Cx43 and immunoblotting for PKC.
Each intercalated disk-rich fraction (500 μg) was solubilized and immunoprecipitated with anti-Cx43 antibody, as previously reported (23). Cx43 immunoprecipitates were used for immunoblotting using anti-PKC-ε and anti-PKC-δ antibodies (BD Biosciences, San Jose, CA).
Experiment 3: Infarct Size
Preparation of isolated perfused rat hearts.
Hearts were perfused as described in experiment 1, and a snare was prepared around the left main coronary artery, as previously reported (24).
Three protocols were used in this series of experiments to examine the effects of isoform-selective inhibition of PKC on infarct size limitation by DADLE and the effect of DADLE-induced PKC-ε activation on myocardial protection afforded by a direct blocker of the gap junction. Myocardial infarction was induced by 35 min of coronary occlusion followed by 2 h of reperfusion. The effects of rottlerin on pretreatment with DADLE were first examined in protocol 1 (Fig. 2A) to confirm reported findings (7), and on the basis of the results of protocol 1, the effects of PKC-ε-TIP and PKC-ε-TIP + rottlerin were examined in protocol 2 (Fig. 2A). Doses of DADLE and the PKC inhibitors were the same as those used in experiment 1. In protocol 3 (Fig. 2B), hearts received no pretreatment (control) or were infused with heptanol (1 mM) from 5 min before reperfusion to 10 min after reperfusion with or without infusion of both rottlerin and DADLE before ischemia. After 2 h of reperfusion, infarct size and size of area at risk were determined as previously reported (24).
Differences in the parameters within and between study groups were analyzed by two-way repeated-measures ANOVA. Intergroup differences in infarct size data were tested by one-way ANOVA. When overall ANOVA indicated a significant difference, multiple comparisons were conducted by the Student-Newman-Keuls post hoc test. P < 0.05 was considered to be statistically significant.
Figure 3 shows representative samples from an untreated control and a DADLE-treated heart and summarized data of gap junction communication in the study groups. Areas stained with LY, but not RD, indicate areas where LY was transported via gap junctions. Size of this area was used as an index of gap junction communication, as in previous studies (20, 23, 30). Reliability of this index of gap junction communication was confirmed by its significant reduction by heptanol, a gap junction blocker (Fig. 3). Similar to ischemic PC in our previous study (20), DADLE significantly reduced the area of gap junction communication by 53%, and this effect of DADLE was almost abolished by PKC-ε-TIP and calphostin C, but not by rottlerin or scrambled TIP (Fig. 3). Neither PKC-ε-TIP alone nor rottlerin alone significantly modified gap junction permeability in ischemic myocardium.
There was no significant difference between total Cx43 levels before and after pretreatment (or its time control period) and after 10 min of ischemia (data not shown). As shown in Fig. 4, a weak signal of Ser368-phosphorylated Cx43 was detected under baseline conditions. Although the level of Ser368-phosphorylated Cx43 was slightly increased after ischemia in untreated controls, it was significantly higher in the DADLE-pretreated group after 10 min of ischemia. PKC-ε-TIP suppressed this DADLE-induced increase in Ser368-phosphorylated Cx43. Using a small number of hearts (n = 3 in each group), we additionally examined the level of Ser368-phosphorylated Cx43 after 35 min of ischemia; Ser368-phosphorylated Cx43 (%baseline) tended to be higher in the DADLE-treated group than in untreated controls (201 ± 34 vs. 162 ± 43%). Consistent with these results for Ser368-phosphorylated Cx43, PKC-ε coimmunoprecipitated with Cx43 after DADLE treatment was almost twofold higher than in untreated controls (Fig. 5). In contrast with PKC-ε, PKC-δ signal was not detected in Cx43 immunoprecipitates from DADLE-treated myocardium.
Baseline levels of heart rate, left ventricular developed pressure (LVDP), and coronary flow were comparable between study groups in each protocol (Table 1). There was no significant intergroup difference in coronary flow reduction during ischemia in each protocol. LVDP was reduced by coronary occlusion and also by infusion of heptanol, and LVDP after reperfusion was slightly lower in PKC-ε-TIP-treated groups than in the control group.
Infarct size data.
Heart weight and size of the area at risk were comparable in all treatment groups (Table 2). DADLE significantly reduced infarct size in protocols 1 and 2. As shown in Fig. 6A, the infarct size-limiting effect of DADLE was largely eliminated by pretreatment with rottlerin, and the residual protection was 37% of that afforded by DADLE alone. Treatment with rottlerin alone did not modify infarct size. Pretreatment with PKC-ε-TIP also significantly blunted infarct size limitation by DADLE (Fig. 6B), but its effect was modest compared with that of rottlerin, leaving 52% of DADLE-induced protection. Rottlerin + PKC-ε-TIP completely inhibited myocardial protection by DADLE. In protocol 3, heptanol infusion initiated during the early reperfusion period reduced infarct size. Preischemic infusion of rottlerin + DADLE, which should have suppressed gap junction permeability by PKC-ε, did not further protect hearts that received heptanol upon reperfusion (Fig. 7).
Role of PKC-ε in δ-Opioid-Induced Suppression of Gap Junction Permeability
Gap junction-mediated electrical coupling is lost within 10–20 min of ischemia. However, chemical coupling of cardiomyocytes by gap junctions has been shown to persist for ≥45 min after the onset of ischemia in rat hearts (30). We previously showed that ischemic PC suppresses gap junction permeability to LY in the ischemic myocardium (20), whereas either the opposite effect or no significant effect of PC has been reported for electrical coupling (14, 26). In the present study, preischemic activation of the δ-opioid receptor reduced gap junction permeability in the ischemic myocardium by ∼50%, which is equivalent to the effect of ischemic PC in our previous study (20).
Pretreatment with DADLE induced complex formation of Cx43 with PKC-ε, but not with PKC-δ (Fig. 5). Ischemia per se slightly increased phosphorylation of Cx43 at Ser368, a PKC-dependent site, as recently reported by Ek-Vitorin et al. (6). However, pretreatment with DADLE significantly increased phosphorylation of Cx43 at Ser368 after ischemia by >50% (Fig. 4). This phosphorylation of Cx43 at Ser368 by DADLE was inhibited by PKC-ε-TIP, indicating involvement of PKC-ε. These results are consistent with our previous finding that calphostin C inhibited PC-induced suppression of Cx43 dephosphorylation during sustained ischemia (20). Although there are two PKC phosphorylation sites in Cx43, recent studies using deletion mutants of Cx43 (2, 6) demonstrated that phosphorylation of Cx43 at Ser368 is responsible for reduction of gap junction permeability by PKC. Furthermore, suppression of gap junction permeability by PC with DADLE was almost abolished not only by calphostin C, but also by PKC-ε-TIP, whereas rottlerin and a negative control for PKC-ε-TIP (scrambled TIP) had no inhibitory effect (Fig. 3). Taken together, these findings indicate that direct phosphorylation of Cx43 by PKC-ε is a primary mechanism of chemical uncoupling of the gap junction by preischemic activation of the δ-opioid receptor, whereas PKC-δ does not play a role in that gap junction modulation.
PKC-ε-TIP and calphostin C prevented 70% of the inhibitory effect of DADLE on the gap junction permeability, leaving ∼30% of the effect uninhibited. A possible explanation for this incomplete blockade by the PKC-ε inhibitors is mKATP channel-mediated regulation of the gap junction. Stimulation of the δ-opioid receptor or other Gi protein-coupled receptors has been shown to activate the mKATP channel in a PKC-independent manner, and opening of this channel induces production of reactive oxygen species (ROS), as signaling molecules, in the mitochondria (4, 16). Interestingly, a determinant of this mKATP channel-mediated ROS production was recently found to be Cx43 protein in the mitochondrial inner membrane (11). Our recent study (23) indicated that ROS produced by mKATP channel opening activates ERK1/2, which induces suppression of gap junction permeability by phosphorylation of Cx43 at ERK-dependent sites, and this mechanism could be operative as a PKC-independent mechanism downstream from the δ-opioid receptor.
Contribution of Suppression of Gap Junction-Meditated Injury to Protection Afforded by δ-Opioid Receptor Activation
Since no agent can directly reverse suppression of gap junction communication, the extent of the contribution of gap junction suppression to infarct size limitation by preischemic activation of the δ-opioid receptor is difficult to estimate. Thus we assessed loss of the infarct size-limiting effect of DADLE when the effect on gap junction permeability was eliminated. The results of the present study suggest relationships between gap junction-dependent and -independent mechanisms in myocardial protection afforded by pretreatment with DADLE (Fig. 8). PKC-ε-TIP, which abolished DADLE-induced suppression of gap junction permeability (Fig. 3), eliminated 48% of cardioprotection afforded by DADLE (Fig. 6B). This is a maximal estimate of the contribution of gap junction suppression to DADLE-induced protection, since some PKC-ε target molecules, other than Cx43, may also play roles in myocardial protection (25). A more conservative estimate is obtained from data from protocol 1. Pretreatment with rottlerin, which did not affect the suppression of gap junction communication by pretreatment with DADLE (Fig. 3), left ∼35% of the infarct size-liming effect of DADLE unblocked (Fig. 6A). Taken together, the present results suggest that ≤35% of myocardial salvage by pretreatment with DADLE is attributable to suppression of gap junction permeability.
Earlier studies (8, 20, 28) and the present experiments have confirmed that pharmacological gap junction blockers infused during ischemia or at the time of reperfusion significantly limited infarct size. Infarct size-limiting effects of these gap junction blockers were not less than those of ischemic PC and pretreatment with DADLE. In a study by Rodriguez-Sinovas et al. (28), lactate dehydrogenase release in isolated rat hearts after reoxygenation following 35 min of hypoxia was reduced by >50% when gap junction blockers (heptanol, 18α-glycyrrhetinic acid, and palmitoleic acid) were infused during the hypoxic period. Interestingly, this protective effect of gap junction blockers was not accompanied by alterations in indexes of electrical coupling of cardiomyocytes, which is consistent with our hypothesis that chemical, but not electrical, uncoupling of cardiomyocytes during ischemia-reperfusion affords protection (20, 23). The findings that the gap junction blockers, except heptanol, did not attenuate elevation of intracellular Ca2+ during simulated ischemia in isolated cardiomyocytes (28) argued against the involvement of suppressed Ca2+ overload during hypoxia in the protection.
A possible explanation for myocardial protection by blockade of the gap junction during ischemia-reperfusion is suppression of propagation of Na+ overload during ischemia and reperfusion, which would reduce Ca2+ overload by Na+/Ca2+ exchange upon reperfusion. Ruiz-Meana et al. (29) showed that the gap junction provides a route for propagation of Na+ overload between adjacent cardiomyocytes. Ischemia increases intracellular Na+ in cardiomyocytes by reduced Na+ efflux via the Na+-K+ pump, persistent influx via the Na+ channel, and increased Na+ influx via the Na+/H+ exchanger (10, 13, 32, 33). Reperfusion is also thought to increase Na+ influx by washing out extracellular protons, which reactivates Na+/H+ exchange, which has been attenuated by accumulation of extracellular protons after long-sustained ischemia. The contribution of these mechanisms of Na+ overload to myocardial necrosis has been supported by the findings that inhibition of the Na+/H+ exchanger or Na+ channel during ischemia-reperfusion significantly limits myocardial infarct size (9, 18).
Gap Junction-Independent Mechanisms in δ-Opioid-Induced Protection
An important role of PKC-δ in myocardial protection by δ-opioid receptor activation was previously suggested by a finding that 70% of the infarct size-limiting effect of TAN-67, a δ-opioid agonist, was lost by pretreatment with rottlerin in rat hearts in situ (7). Consistent with this finding, rottlerin inhibited 63% of cardioprotection afforded by DADLE in the present study, and the PKC-δ-mediated protection did not require suppression of gap junction permeability. How PKC-δ increases anti-infarct tolerance remains unclear, but recent studies (15, 27) have suggested that PKC-δ activates ERK1/2 and the mKATP channel, which are potentially cytoprotective. In a recent study (12), we found that activation of the δ-opioid receptor induced not only ERK activation, but also suppression of calcineurin, which also leads to myocardial resistance to ischemia-reperfusion injury. It is unclear, however, whether PKC-δ is involved in this suppression of calcineurin. Nevertheless, we speculate that PKC-δ activity during activation of the δ-opioid receptor and/or during ischemia is important for myocardial protection, since activation of PKC-δ upon reperfusion has been shown to be rather detrimental to cell survival (21).
This study was supported by Japanese Society for the Promotion of Science Grants-in-Aid for Scientific Research 16590702, 18590781, and 18590781 and a grant from the Sapporo Medical University Academic Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by the American Physiological Society