Soluble epoxide hydrolase (sEH) metabolizes epoxyeicosatrienoic acids (EETs), primarily 14,15-EET. EETs are derived from arachidonic acid via P-450 epoxygenases and are cardioprotective. We tested the hypothesis that sEH deficiency and pharmacological inhibition elicit tolerance to ischemia via EET-mediated STAT3 signaling in vitro and in vivo. In addition, the relevance of single nucleotide polymorphisms (SNPs) of EPHX2 (the gene encoding sEH) on tolerance to oxygen and glucose deprivation and reoxygenation and glucose repletion (OGD/RGR) was assessed in male C57BL\6J (WT) or sEH knockout (sEHKO) cardiomyocytes by using transactivator of transcription (TAT)-mediated transduction with sEH mutant proteins. Cell death and hydrolase activity was lower in Arg287Gln EPHX2 mutants vs. nontransduced controls. Excess 14,15-EET and SEH inhibition did not improve cell survival in Arg287Gln mutants. In WT cells, the putative EET receptor antagonist, 14,15-EEZE, abolished the effect of 14,15-EET and sEH inhibition. Cotreatment with 14,15-EET and SEH inhibition did not provide increased protection. In vitro, STAT3 inhibition blocked 14,15-EET cytoprotection, but not the effect of SEH inhibition. However, STAT3 small interfering RNA (siRNA) abolished cytoprotection by 14,15-EET and sEH inhibition, but cells pretreated with JAK2 siRNA remained protected. In vivo, STAT3 inhibition abolished 14,15-EET-mediated infarct size reduction. In summary, the Arg287Gln mutation is associated with improved tolerance against ischemia in vitro, and inhibition of sEH preserves cardiomyocyte viability following OGD/RGR via an EET-dependent mechanism. In vivo and in vitro, 14,15-EET-mediated protection is mediated in part by STAT3.
- oxygen and glucose deprivation
- reoxygenation and glucose repletion
- N-adamantanyl-N′-dodecanoic acid urea
p-450 epoxygenase metabolizes arachidonic acid (AA) into epoxyeicosatrienoic acids (EETs) (20). EETs are converted into dihydroxyeicosatrienic acids (DHET) by soluble epoxide hydrolase (sEH) (32). EPHX2 encodes sEH, the enzyme that breaks down EETs. More than 10 genetic variants of the human EPHX2 gene have been identified (27); some of these polymorphisms have been implicated in susceptibility to cardiovascular disease (7, 8, 14, 19). Previously, we demonstrated that EPHX2 polymorphisms influence neuronal survival in vitro (17). In addition, we have recently shown that inhibition or deletion of sEH, and exogenous 14,15-EET administration, decreases myocardial infarct size after regional ischemia-reperfusion in mice (22). The cardioprotective effect of EETs is linked to activation of known cell survival pathways like the reperfusion injury salvage kinase (RISK) involving PI3K/Akt, as well as MAPK, and KATP channels (13, 26). However, cardioprotective signaling elicited by sEH inhibition/endogenous EET augmentation is not fully known. Recently, it has become appreciated that JAK/STAT signaling participates in acquired ischemic tolerance elicited by pharmacological or ischemic preconditioning (1, 6, 15).
MATERIAL AND METHODS
Experiments were conducted in isolated cardiomyocytes from wild-type (WT) and sEH knockout (sEHKO) mice and in adult WT male mice. Animals used in these studies were allowed access to food (no. 2014, Harlan Teklad, Madison, WI) and water ad libitum until induction of anesthesia. Under local Institutional Animal Care and Use Committee approval, all animals received treatment in compliance with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Research, National Research Council; National Academy Press, 1996). The sEHKO strain was obtained from Dr. Frank Gonzalez at the National Institutes of Health. Homozygous mice are viable, fertile, and normal in size and do not display any gross physical or behavioral abnormalities (25). EPHX2 deletion was confirmed by PCR. sEHKO mice were backcrossed on a C57BL\6J for at least seven generations and compared with WT C57BL\6J mice. The time table and experimental design are summarized in Fig. 1. All drugs were dissolved in DMSO (1:1,000) unless otherwise stated.
Ventricular cardiomyocytes from neonatal mice [postnatal day (PND) 7–8] were prepared (4) and plated at a density of 1.0 × 105 cells/cm2 and incubated for 24 h. To simulate ischemia-reperfusion in vitro, cardiomyocytes were subjected to oxygen-glucose deprivation (OGD, 90 min) followed by reoxygenation in glucose-replete medium (RGR, 180 min), and cell death was quantified by propidium iodine (5 μM) staining (21). Western blotting for sEH and STAT3 was performed as described in the online supplement. DHET ELISAs were performed following the manufacturer's instructions (Detroit R&D). To elicit acute myocardial ischemia-reperfusion in vivo, the left coronary artery (LCA) was occluded for 40 min followed by 2 h of reperfusion in male WT (16–24 wk), and infarct size expressed as a percentage of area at risk as previously published (22).
We assessed whether variation in sEH activity due to EPHX2 polymorphisms alters cardiomyocyte survival after OGD. Cultured cardiomyocytes were transduced with five human EPHX2 variants linked to the protein transduction domain TAT, and post-OGD viability was assessed. Six TAT-human recombinant sEH (hr-sEH) fusion proteins (one WT and five EPHX2 mutants: Lys55Arg, Arg103Cys, Cys154Tyr, Arg287Gln, and Arg103Cys/Arg287Gln) were expressed and purified as previously reported (17). TAT-hr-sEH fusion proteins were added to the medium of cultured cardiomyocytes at a final concentration of 100 nM. Successful transduction was confirmed by Western blot and immunohistochemistry ( Fig. 2, A and B). sEH activity was estimated by 14,15-DHET production in TAT-hr-sEH-WT, TAT-hr-sEH-Lys55Arg, and TAT-hr-sEH-Arg287Gln, following incubation with 1 μM 14,15-EET for 4.5 h. This time point was chosen to match the viability assessment time point (1.5 h OGD plus 3 h RGR). For viability experiments, cells were incubated at 37°C for 24 h before initiation of 90 min OGD and 180 min RGR with or without the sEH substrate 14,15-EET (1 μM) or the sEH inhibitor 4-phenylchalcone oxide (4-PCO, 2 μM) (n = 5–9 replicates). A subset of experiments was performed in cardiomyocytes derived from male sEHKO mice (n = 3 replicates).
To determine whether 14,15-EET directly protects cardiac cells from injury due to simulated ischemia, and to verify that the cytoprotection elicited is due to EET-mediated signaling, two approaches were taken. First, cultured cardiomyocytes were treated with exogenous 14,15-EET (1 μM), 4-PCO (2 μM), or both for 60 min prior to 90 min OGD followed by 180 min RGR. Secondly, cardiomyocytes were pretreated with 14,15-EET (1 μM), and sEH inhibitors N-adamantanyl-N′-dodecanoic acid urea (AUDA, 2 μM) or 4-PCO (2 μM), in the presence or absence of putative EET receptor antagonist 14,15-epoxyeicosa-5(Z)-enoic acid (EEZE, 1 μM) (n = 3–5 replicates). In a subset of cells treated with 14,15-EET, viability was assessed up to 6 h RGR (n = 3 replicates) to determine sustainability of cytoprotection. In addition, an additive effect of sEH inhibition and 14,15-EET was tested.
sEH is a bifunctional enzyme, with the COOH-terminal domain acting as hydrolase and the NH2-terminal domain as phosphatase (16, 17). Although pharmacological inhibition of sEH with either 4-PCO or AUDA only eliminates the hydrolase activity, genetic deletion of the enzyme eradicates both hydrolase and phosphatase activities. To determine whether inhibition of the hydrolase is sufficient to induce cytoprotection, we compared post-OGD cardiomyocyte viability in cultured WT cells treated with AUDA (2 μM), in the presence and absence of the EET receptor antagonist EEZE (1 μM) (n = 4–5 replicates). In addition, cardiomyocytes from sEHKO mice, with and without concomitant sEH inhibition with AUDA or exogenous 14,15-EET, were also examined (n = 5 replicates).
To determine the role of STAT3 signaling in the cytoprotection elicited by 14,15-EET and sEH inhibition, cultured cardiomyocytes were treated with either 14,15-EET (1 μM) or AUDA (2 μM) for 1 h before initiation of OGD, with and without the selective STAT3 inhibitors Stattic (Stat three inhibitor compound, 6-nitrobenzo[b]thiophene 1,1-dioxide; Tocris, 10 μM) or STAT3 inhibitor VI (S31-201; EMD-Calbiochem, 100 μM), which were administered to cardiomyocytes 5 min before treatment with 14,15-EET or AUDA and subsequent OGD-RGR (n = 3–5 replicates). STAT3 activation (phosphorylation on tyrosine 705 or serine 727) was assessed by Western blot (n = 5–6 replicates). In addition, small interfering RNA (siRNA) for STAT3 and JAK2 was used in isolated cardiomyocytes to further dissect the role of STAT3 signaling. Cells were pretreated with either STAT3, JAK2, or control siRNA for 48 h followed by 14,15-EET (1 μM) or AUDA (2 μM) 1 h before initiation of OGD. Cell death was assessed 180 min after RGR (n = 4–5 replicas). The siRNA effect on protein expression after 48 h was assessed by Western blot.
To corroborate the role of STAT3 signaling in 14,15-EET-mediated cardioprotection, WT male mice were treated with 14,15-EET (2.5 μg/g iv) 15 min prior to 40 min of LCA occlusion followed by 2 h of reperfusion in the presence and absence of STAT3 Inhibitor VI (S31–201; EMD-Calbiochem, 6 μg/g dissolved in 10% DMSO and 90% PBS at pH 8.5), which was given intravenously 5 min prior to 14,15-EET.
Data analysis was performed with a PC-based software package (Prism 5.0). Differences between groups over time were analyzed by two-way ANOVA with Bonferroni post hoc tests to compare replicate means and at end-reoxygenation by one-way ANOVA and a Student-Newman-Keuls post hoc test. For other variables, differences between groups were assessed by ANOVA with repeated measures and a Student-Newman-Keuls post hoc test as appropriate. Statistical significance was assumed for P values ≤ 0.05. Because the amount of left ventricular myocardium that progresses to infarction depends on the size of the area at risk (AAR), infarct size was normalized as a percentage of the AAR.
Transduction over time (30 min to 24 h) showed a time-dependent increase in sEH levels following transduction with TAT-hr-sEH-WT fusion protein, reaching a plateau between 8 and 24 h (Fig. 2A). For subsequent experiments the 24-h transduction time was used. Intracellular presence of TAT fusion proteins was confirmed by immunohistochemistry staining with anti-His-tag antibody (Fig. 2B). sEH activity was estimated by measuring the conversion rate to the 14,15-EET metabolite 14,15-DHET in TAT-transduced human WT and EPHX2 mutants Lys55Arg and Arg287Gln and was compared with that in naive cardiomyocytes. Arg287Gln showed reduced conversion rate compared with WT and Lys55Arg (Fig. 2E). There was no difference in cell death between naive cardiomyocytes and human EPHX2 WT, whereas the Arg287Gln mutant showed improved cell survival (Fig. 2, C and D, solid bars). Excess supply of 14,15-EET, the main substrate for sEH, improved cell survival in WT and mutants but did not confer additive protection against OGD to the already protected Arg287Gln mutants (Fig. 2C). Similarly, inhibition of sEH with 2 μM 4-PCO improved cell survival but did not confer additive protection to the Arg287Gln mutants (Fig. 2D). Transduction of cardiomyocytes from sEHKO mice with hr-sEH protein increased cell death, but the Arg287Gln mutant protein had no effect on cell viability (Fig. 2F).
As seen in Fig. 3A, the protection conferred by 14,15-EET is long lasting and can be observed up to 6 h after OGD. As shown in Fig. 3B, exogenous 14,15-EET increased cell survival following OGD. Inhibition of sEH, the main degradation pathway for 14,15-EET, with 4-PCO also improved cardiomyocyte viability. However, concomitant treatment with 4-PCO, to augment endogenous 14,15-EET levels, plus exogenous 14,15-EET, did not result in additive protection. As seen in Fig. 3C, treatment with 14,15-EET, AUDA, and 4-PCO alone reduced cell death ∼33% (P < 0.001 vs. vehicle control), and 14,15-EEZE abolished this salutary effect for all three agents. As seen in Fig. 3D, pretreatment of cells subjected to OGD/RGR with AUDA and 14,15-EET resulted in greater viability of WT cardiomyocytes compared with untreated cells, with 14,15-EET being more protective than AUDA. In sEHKO cardiomyocytes, AUDA increased cell death whereas exogenous 14,15-EET resulted in reduced cell death. Untreated sEHKO cells showed improved viability compared with WT cells. In addition, as shown in Fig. 3E, lower doses of 14,15-EET (0.1 and 0.3 μM) alone had no protective effect on cell death, and AUDA showed no additive effect with all three 14,15-EET doses tested.
Both 14,15-EET and AUDA rapidly elicited phosphorylation of STAT3 at residue 705Tyr (Fig. 4, A and B) within 5 min of incubation. 14,15-EET had no effect on the alternative phosphorylation site at residue 727Ser. Figure 5 shows the effect of STAT3 inhibition on cardiomyocyte viability following OGD. Figure 5A shows the cell death over the complete observation time period, with viability assessed at baseline, beginning of RGR, and at 60, 120, and 180 min of RGR. Figure 5B shows the percentage of cell death corrected to the oxygenated control at 180 min of RGR. Finally, Fig. 5C shows the percentage of dead cells compared with its vehicle control to account for the increased cell death observed with each STAT3 inhibitor alone. STAT3 inhibition abolishes the cytoprotection of 14,15-EET but has less effect on AUDA-induced enhancement of survival. Figure 6 shows the effect of STAT3 siRNA and JAK2 siRNA on cardiomyocyte viability. STAT3 knockdown completely abolished cytoprotection by 14,15-EET and AUDA, whereas JAK2 knockdown had no effect.
As shown in Fig. 7A, 14,15-EET significantly reduced infarct size after 40 min of LCA occlusion and 120 min of reperfusion. Pretreatment with the STAT3 inhibitor VI abolished this cardioprotective effect. Ischemic AAR relative to left ventricular volume did not differ between groups (Fig. 7B). There was no difference in heart rate within groups at all assessed time points, and body weight was similar among groups tested. Body temperature was similar between groups at all assessed time points but was different at baseline between vehicle group (36.6 ± 0°C) and 14,15-EET-treated group (36.9 ± 0.1°C), between vehicle and STAT3 inhibitor alone (36.8 ± 0°C) groups, and between STAT3 inhibitor plus 14,15-EET (36.6 ± 0.1°C) and 14,15-EET treated (36.9 ± 0.1°C) groups.
The primary findings of the present study are that 1) one single polymorphism in EPHX2 that results in reduced sEH enzymatic activity is associated with tolerance of cultured cardiomyocytes to OGD; 2) pharmacological inhibition of sEH increases cardiomyocyte survival following OGD via an EET-dependent mechanism; 3) exogenous administration of the main substrate of sEH, 14,15-EET, confers tolerance to OGD; and 4) 14,15-EET-induced and AUDA-induced protection is mediated at least in part by STAT3 residue prosurvival signaling independent of JAK2.
AA is liberated from membrane phospholipids in response to a variety of stimuli, including myocardial ischemia. Free AA is metabolized by three different enzymatic pathways: cyclooxygenase (COX), lipoxygenase (LOX), or cytochrome P-450 monooxygenases. P-450 epoxygenases add an epoxide moiety to AA to form EETs. Although the biological actions of EETs in the vasculature has been extensively studied and well described, and the role of LOX and COX metabolites in acquired ischemic tolerance has been known for some time (2, 23, 26), the role of EETs in myocardial protection has only recently been appreciated.
EETs possess several properties that are beneficial in combating the deleterious effects of myocardial ischemia, such as vasodilation (18), angiogenesis (32), and suppression of inflammation (25), and the salutary effects of EETs that have been observed in vivo may be in part due to these actions. EETs have also been linked to activation of cell survival pathways, including opening of mitochondrial ATP-dependent potassium channel and activation phosphatidylinositol 3-kinase/Akt and mitogen-activated protein kinase pathways (13, 29), which raises the possibility of EETs providing tolerance to ischemia-hypoxia directly to cardiomyocytes.
The present study demonstrates that the Arg287Gln human SNP of the EPHX2 gene results in decreased cardiomyocyte sEH activity and confers tolerance to OGD. This finding is consistent with those of others who found reduced hydrolase activity with the Arg287Gln polymorphism (17, 19). Interestingly, reduced sEH activity in cardiomyocytes transfected with protein possessing the Arg287Gln polymorphism occurred in the presence of murine WT sEH. Arg287Gln results in reduced dimerization (31), which could have a negative dominant effect on endogenous murine sEH dimerization. The increased cell death after transduction with the human WT TAT-hr-sEH protein but not with the Arg287Gln mutant TAT-hr-sEH protein, on a sEH-depleted background (sEHKO mice), supports the physiological relevance of sEH polymorphisms. These data also support the concept that sEH inhibition is protective against ischemic insults and provide a potential explanation for variation in ischemic tolerance that is observed in both humans and experimental animals. Genetic association studies have linked the Arg287Gln SNP with increased risk for coronary artery disease in African Americans (7, 34). Interestingly, comorbid disease states may negate or mask the effect of the Arg287GLN SNP, since this SNP was not independently associated with an increased risk of coronary arteriosclerosis in a cohort of Type 2 diabetic subjects (3). Of course, the link between increased cardiovascular risk and Arg287Gln does not necessarily indicate causality by itself. In addition, increased cardiovascular risk is different from tolerance against ischemia-reperfusion injury. No studies have evaluated the effect of EPHX2 SNPs on susceptibility of the intact myocardium to ischemia-reperfusion. Additional studies are needed to delineate the relevance of this SNP in the context of in vivo myocardial ischemia-reperfusion.
The present study expands on our and others' previous findings of reduced infarct size following inhibition of sEH or administration of exogenous 14,15-EET in vivo (12, 13, 22, 24) and in isolated heart experiments (28). We now show direct cardiomyocyte protection following OGD via inhibition of sEH, sEH gene deletion, or administration of its main substrate, 14,15-EET. Similar to previously published isolated heart and in vivo studies, the present study demonstrates that sEH inhibition and exogenous EET administration both provide tolerance to simulated ischemic stress via a 14,15-EET-dependent mechanism. However, whereas Gross and colleagues (12 found that inhibition of sEH in conjunction with EET administration elicited greater infarct limitation than either treatment alone, the present isolated cardiomyocyte studies show that at a cellular level combined sEH inhibition and EET was not more efficacious than either treatment alone. This suggests that the additive effects of these interventions in vivo may be due in part to indirect effects that provide ischemic tolerance, such as a reduction in inflammatory mediators or augmentation of regional myocardial blood flow (18, 25), whereas in the present in vitro studies the effects of sEH inhibition and EET were restricted to direct cardiomyocyte effects. Interestingly, in cardiomyocytes from sEHKO mice, the sEH inhibitor AUDA resulted in increased cell death compared with untreated cardiomyocytes from sEHKO mice, but exogenous 14,15-EET produced additional cytoprotection against OGD/RGR. This disparity may reflect the fact that AUDA inhibits the hydrolase activity of sEH but has no effect on the phosphatase activity. In contrast, EPHX2 gene deletion in sEHKO mice inactivates both enzyme functions. Thus AUDA may possess non-sEH-specific effects that inhibit other protective pathways or elicit cell damage by a yet unknown mechanism. Alternatively, inhibition of the phosphatase may also confer ischemic tolerance.
14,15-EET induced cytoprotection was long-lasting (up to 6 h). The cell death vs. time curve appears to reach a near asymptote between 3 and 4 h. It is likely that the observed salutary effect would also be observed at time points exceeding 6 h post-OGD, suggesting that 14,15-EET mediated protection is not only delaying injury but truly reducing cell death following OGD/RGR.
Protective cellular signaling pathways utilized by EETs have not been fully investigated. Recently, it has become apparent that ischemic preconditioning relies in part on JAK and STAT signaling (1, 16), as does insulin- and opioid-induced cardioprotection (9, 11). It was recently reported that, in human dermal microvascular endothelial cells, 14,15-EET stimulates tyrosine phosphorylation of STAT3 and its translocation from the cytoplasm to nucleus (5). Thus we focused on investigating the role of STAT3 in vitro and in vivo. Our data support the hypothesis that STAT3 phosphorylation on the 705Tyr residue plays an important role in mediating 14,15-EET induced tolerance against ischemic injury. Recently, it was postulated that STAT3 activation is upstream of the RISK pathway (10). STAT3 contains a 727Ser residue that can be phosphorylated; however, we did not observe phosphorylation at this site following 14,15-EET incubation. 727Ser-P seems to play a role in maintaining the mitochondrial electron transport chain (33). Similar to our findings, Gross et al. (11) did not show 727Ser phosphorylation 5 min after reperfusion in an isolated rat heart model following opioid-induced cardioprotection, but rather phosphorylation of the 705Tyr site. Similarly, we show that inhibition of 705Tyr phosphorylation abolished 14,15-EET mediated protection in vitro and in vivo. In addition, our data suggest that STAT3 phosphorylation mediated by 14,15-EET occurs independent of JAK2 activation. Interestingly, whereas neither STAT3 inhibitor abolished AUDA-mediated protection in vitro, STAT3-specific knockdown with siRNA completely abolished AUDA-induced protection. This might be the result of a nonspecific interaction between the different inhibitors. Alternatively, this raises the possibility that the STAT3 pharmacological inhibitors are not fully selective for STAT3 and that AUDA modulates this unknown target. As noted in Fig. 3, treatment of sEHKO cardiomyocytes with AUDA exacerbated cell death (as compared with sEHKO alone, with both sEHKO and sEHKO+AUDA less than WT), whereas treatment of WT cardiomyocytes with AUDA improved viability, suggesting that AUDA has both a slight proinjury effect and a predominant protective effect. Thus the present results could be observed if STAT3 inhibitors blocked both STAT3 and the (proinjury) target of AUDA, whereas STAT3 siRNA blocked only STAT3. Regardless, since directly targeting STAT3 expression abolishes AUDA-mediated protection, this suggests that AUDA-mediated protection occurs via a STAT3-dependent mechanism.
In the present study, neonatal cardiomyocytes were used rather than adult cardiomyocytes; thus there is a possibility that adult tissue would behave differently from what was observed. However, murine cardiomyocyte hyperplasia is largely complete by PND 7 (the age of myocytes used in the present study), with cardiac DNA labeling with tritiated thymidine < 10% at PND 7 and 0% at PND 10, and GAX (growth arrest homeobox) gene expression absent in embryonic cardiac tissue but expressed at adult levels by PND 7 (30). Thus the PND 7–8 cardiomyocytes used were likely comprised of almost exclusively binucleated and terminally differentiated cells.
In summary, the present study shows that reduced or inhibited sEH activity improves tolerance against ischemia-reperfusion in vitro and in vivo via 14,15-EET- mediated activation of the STAT3 signaling pathway.
This work was supported by an RO1 NS44313 (N. J. Alkayed); VA Merit Review grant 317 (Medical Research Service, Department of Veterans Affairs; D. M. Van Winkle) and Oregon Health and Science University, Department of Anesthesiology and Perioperative Medicine, Anesthesiology Research and Education Foundation (M. J. Merkel).
No conflicts of interest are declared by the author(s).
The authors thank Dr. Ines Koerner for helpful instructions in completing the in vitro experiments involving TAT-sEH transfection experiments with EPHX2 SNPs and acknowledge the excellent service and care provided by the Department of Anesthesiology and Perioperative Medicine Mouse Colony Core, which oversaw management of the sEHKO mouse breeding colonies.
- Copyright © 2010 the American Physiological Society