Activated protein C modulates cardiac metabolism and augments autophagy in the ischemic heart

Authors


  • Present address: Children's Hospital Boston, Harvard Medical School, Boston, MA 02115, USA

Alireza R. Rezaie, Department of Biochemistry and Molecular Biology, Saint Louis University School of Medicine, Saint Louis, MO 63104, USA.
Tel.: +1 314 977 9240; fax: +1 314 977 9205.
E-mail: rezaiear@slu.edu
Correspondence: Ji Li, Department of Pharmacology and Toxicology, University at Buffalo-SUNY, Buffalo, NY 14214, USA.
Tel.: +1 716 829 5711; fax: +1 716 829 2801.
E-mail: jli23@buffalo.edu

Abstract

Summary.  Background:  Modulation of energy substrate metabolism may constitute a novel therapeutic intervention against ischemia/reperfusion (I/R) injury. AMP-activated protein kinase (AMPK) has emerged as a key regulator of favorable metabolic signaling pathways in response to myocardial ischemia. Recently, we demonstrated that activated protein C (APC) is cardioprotective against ischemia/reperfusion (I/R) injury by augmenting AMPK signaling.

Objectives:  The objective of this study was to determine whether the APC modulation of substrate metabolism contributes to its cardioprotective effect against I/R injury.

Methods:  An ex vivo working mouse heart perfusion system was used to characterize the effect of wild-type APC and its signaling-proficient mutant, APC-2Cys (which has dramatically reduced anticoagulant activity), on glucose transport in the ischemic heart.

Results:  Both APC and APC-2Cys (0.2 μg g−1) augment the ischemic stress-induced translocation of the glucose transporter (GLUT4) to the myocardial cell membrane, leading to increased glucose uptake and glucose oxidation in the ischemic heart (< 0.05 vs. vehicle). Both APC derivatives increased the autophagic flux in the heart following I/R. The activity of APC-2Cys in modulating these metabolic pathways was significantly higher than APC during I/R (< 0.05). Intriguingly, APC-2Cys, but not wild-type APC, attenuated the I/R-initiated fatty acid oxidation by 80% (< 0.01 vs. vehicle).

Conclusions:  APC exerts a cardioprotective effect against I/R injury by preferentially enhancing the oxidation of glucose over fatty acids as energy substrates in the ischemic heart. Given its significantly higher beneficial metabolic modulatory effect, APC-2Cys may be developed as a potential therapeutic drug for treating ischemic heart disease without risk of bleeding.

Introduction

Myocardial ischemia is a condition that occurs when blood flow to the myocardium is reduced because of coronary atherosclerosis, coronary thrombosis, and narrowing of arterioles in the heart. Current treatments are largely targeted at immediate restoration of blood flow to the heart by recanalization of the occluded coronary artery via the use of percutaneous coronary intervention, thrombolytics, and anticoagulants. These interventions, however, have the risk of exacerbating bleeding in patients who may already be at increased risk because of other medications or conditions [1]. Although the process of reperfusion of the heart aids in reducing the mortality rate of myocardial ischemia by up to 50% [2], the rapid restoration of coronary blood flow following myocardial ischemia can paradoxically induce the death of cardiac myocytes, a pathologic condition known as reperfusion injury [3]. Reperfusion injury has been shown to largely arise from the oxidative stress, and, to date, no true therapies to alleviate it have been developed [4]. Therefore, there is an urgent need for novel therapeutic strategies that can limit myocardial ischemia/reperfusion (I/R) injury without increasing the risk of bleeding [5].

We recently demonstrated that recombinant activated protein C (APC) exerts a potent cardioprotective effect in I/R injury through activation of AMP-activated protein kinase (AMPK) [6]. APC is a vitamin K-dependent serine protease that inhibits blood clotting through proteolytic degradation of procoagulant factors Va and VIIIa, which are required cofactors for thrombin generation in plasma during the blood coagulation process [7,8]. In addition to its anticoagulant function, APC also elicits anti-inflammatory and cytoprotective signaling responses when it binds to endothelial protein C receptor (EPCR) to activate protease-activated receptor-1 (PAR-1) on the vascular endothelium [9–11]. The anti-inflammatory and cytoprotective signaling properties of APC have been shown to be independent of its anticoagulant activity [11]. Thus, in a recent study, we demonstrated that a signaling-proficient APC mutant (APC-2Cys), which has dramatically reduced anticoagulant activity, ameliorated cardiac dysfunction under I/R conditions to a similar or even greater extent as wild-type APC [6].

AMPK signaling has been shown to protect against myocardial ischemic injury through the regulation of metabolism in the heart, by balancing the energy demand and supply in response to ischemic stress [3,12]. Activated AMPK can phosphorylate acetyl-CoA carboxylase, thereby inhibiting its activity in the fatty acid synthesis pathway [13]. Other downstream effects of AMPK pathways include glucose uptake [14,15], glycolysis [14], and fatty acid oxidation [16], all of which favor ATP production in order to supply sufficient energy for cell survival under stress conditions. Recently, AMPK-mediated autophagy has also been shown to play a possible protective role during myocardial ischemia [17]. Autophagy has emerged as an important mediator in the regulation of glucose homeostasis [18]. There is increasing evidence that enhancing glucose oxidation and inhibiting fatty acid oxidation in the ischemic heart has a beneficial effect in maintaining cardiac efficiency [19,20]. For this reason, novel therapeutics that can selectively target the upregulation of glucose metabolism in the ischemic heart may be of great value. Whether the cardioprotective signaling function of APC contributes to modulation of energy substrate metabolism in the ischemic heart has never been investigated. We undertook this study to address this important question.

Materials and methods

Animals

Male C57BL/6 mice, 8–12 weeks of age, were purchased from Jackson Laboratory (Bar Harbor, ME, USA). Mice were maintained on a 12-h light/dark cycle in a controlled environment with water ad libitum. All animal protocols were approved by the Institutional Animal Care and Use Committee of State University of New York (SUNY) at Buffalo.

Ex vivo heart perfusion

Mice were anesthetized with intraperitoneal pentobarbital sodium (60 mg kg−1) and heparinized (100 units intraperitoneally). Hearts were excised and retroperfused (4 mL min−1) in the perfusion heart system (Radnoti Glass Technology, Monrovia, CA, USA) with 95% O2 and 5% CO2 in equilibrated Krebs–Henseleit buffer (KHB) containing 7 mm glucose, 1% bovine serum albumin, and 0.4 mm oleate. For the ex vivo ischemic model, the buffer flow was stopped for 10 min, at which point hearts were reperfused with the same flow rate and buffer containing APC (20 nm), APC-2Cys (20 nm) [21], APC-E170A (20 nm) [22], or protein C-2Cys (PC-2Cys) (20 nm) [21]. LabChart7 software from ADInstruments (Colorado Springs, CO, USA) was used to monitor the heart rate and left ventricle (LV) pressure.

Cell surface GLUT4 labeling by Bio-LC-ATB-BGPA

Cell membrane glucose transporter type 4 (GLUT4) labeling with 4,4′-O-[2-[2-[2-[2-[2-[6-(biotinylamino)hexanoyl]amino]ethoxy]ethoxy]ethoxy]-4-(1-azi-2,2,2,-trifluoroethyl)benzoyl]amino-1,3-propanediyl]bis-d-glucose (bio-LC-ATB-BGPA) was performed as previously described [23]. After perfusion, mouse hearts were flushed by aortic cannulation with 1 mL of ice-cold glucose-free KHB, and then perfused with the same buffer containing 300 μm bio-LC-ATB-BGPA. Hearts infused with bio-LC-ATB-BGPA were incubated at 4 °C for 15 min. To enhance crosslinking between bio-LC-ATB-BGPA and cell surface GLUT4, the LV and right ventricle were cut sagittally, and reactions were exposed to UV irradiation twice for 5 min each. Heart tissues were then freeze-clamped and stored at − 80 °C until further analysis.

For isolation of cell surface GLUT4, photolabeled cardiac tissues were homogenized in 250 μL of Hepes–EDTA–sucrose buffer containing 20 mm Hepes, 5 mm Na-EDTA, 255 mm sucrose, and protease inhibitor cocktail (Hoffmann-La Roche, Indianapolis, IN, USA). Tissue homogenates were added to 250 μL of 4% Thesit/phosphate-buffered saline (PBS), incubated on ice for 15 min, and then kept at 4 °C for another 15 min. Tissue homogenates were centrifuged at 20 000 × g at 4 °C for 30 min, and pellets were discarded. Ten microliters of the supernatant was taken for measurement of the total protein concentrations. For isolation of the photolabeled GLUT4, 400 μg of total membrane protein was incubated with 100 μL of streptavidin bound to 6% agarose beads (Pierce, Rockford, IL, USA) overnight at 4 °C. The streptavidin–agarose-isolated labeled fraction of GLUT4 was washed extensively with PBS containing decreasing concentrations of Thesit (1%, 0.1%, and 0%). The labeled GLUT4 was then dissociated from streptavidin by boiling in the loading buffer for 30 min prior to analysis by SDS-PAGE.

Measurement of glucose uptake

Glucose uptake was analyzed in the Langendorff heart perfusion mode by measuring the production of 3H2O from d-[2-3H]glucose, as previously described [12]. Briefly, the KHB containing d-[2-3H]glucose (50 μCi L−1) was perfused into the isolated heart, and the coronary effluent was sampled every 5 min. For separation of the non-metabolized d-[2-3H]glucose from 3H2O, ion exchange chromatography (Bio-Rad AG1–8X resin; Bio-Rad, Hercules, CA, USA) was conducted by activating the resin with 1 m sodium hydroxide. Resin columns were extensively washed with deionized H2O (H2O) to ensure that the pH was < 8. A 500-μL coronary effluent sample was added to each column to allow binding of glucose to the resin. 3H2O was washed out by flushing the column with 2.5 mL of dH2O. All flow was collected in scintillation vials, which were then subjected to radioactive counting. The rate of glucose uptake was calculated from the amount of 3H2O production.

Fatty acid/glucose oxidation analysis

Cardiac substrate metabolism was determined in the working heart model as previously described [12,24]. The working heart preload was set up at 15 cm H2O, and the afterload at 80 cm H2O. The flow rate was kept at 15 mL min−1. The mouse heart was first aortically cannulated in order to initiate Langendorff perfusion. The pulmonary vein was then cannulated, and the working heart mode was started with the perfusion of [9,10-3H]oleate and [U-14C]glucose. Heart function was monitored with a pressure transducer connected to an aortic outflow.

Fatty acid oxidation was determined from the production of 3H2O from [9,10-3H]oleate. For separation of 3H2O from [9,10-3H]oleate, perfusate samples were filtered through the anion exchange resin (200-400 Bio-Rad AG1-2X resin; Bio-Rad) pretreated with 1 m sodium hydroxide. Resin columns were extensively washed with dH2O to lower the pH below 8. A 400-μL sample was loaded onto the column and eluted into scintillation vials with 2.5 mL of dH2O. Ten milliliters of scintillation fluid was added to each vial, and the samples were read. Glucose oxidation was determined from the production of 14CO2 from [U-14C]glucose. The 14CO2 was captured with hyamine hydroxide, and placed in scintillation vials; 13 mL of scintillation fluid was then added to each vial, and the radioactive signal was read with a liquid scintillation counter.

Immunoblotting analysis

Immunoblotting was performed as previously described [6]. Heart homogenates were resolved by SDS-PAGE, and proteins were transferred onto poly(vinylidene difluoride) membranes. For reprobing, membranes were stripped with 50 mm Tris-HCl, 2% SDS, and 0.1 mβ-mercaptoethanol (pH 6.8). Rabbit polyclonal antibodies against phospho-Akt (Ser473), GLUT4 and LC3 were purchased from Cell Signaling Technology (Danvers, MA, USA). Rabbit polyclonal antibodies against total Akt were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-rabbit secondary antibodies were purchased from Cell Signaling Technology.

Cellular redox measurement

The production of reactive oxygen species (ROS) by ex vivo hearts subjected to 10 min of global ischemia and 20 min of reperfusion was inferred through measurement of the reduced glutathione (GSH)/oxidized glutathione (GSSG) ratio in the heart tissue with a glutathione detection kit (Enzo Life Sciences, Farmingdale, NY, USA) and a plate reader, according to the manufacturer’s instruction.

Measurement of autophagy

Autophagy was investigated by analysis of the intracellular localization and processing of the microtubule-associated protein 1 light chain 3 (LC3) in the heart following I/R, as previously described [25]. LC3 is synthesized as pro-LC3, which is cleaved by Atg4B to form LC3-I with the C-terminal Gly exposed [25]. LC3-I is activated by Atg7 and conjugated to phosphatidylethanolamine, and this form has been designated LC3-II [25]. LC3-II is used as a marker of autophagy, because its lipidation and specific recruitment to autophagosomes increases its electrophoretic mobility on the gels as compared with LC3-I [25]. For determination of the amount of LC3-II, immunoblotting analysis was performed with an antibody against LC3, and the densities of LC3-II and LC3-I were measured with the NIH imaging software.

Statistical analysis

Data were expressed as means ± standard error. Data were analyzed with one-way anova for measurement of statistical significance. For single-factor and multi-factor analyses, appropriate post hoc test(s) were performed to measure individual group differences of interest. A P-value of < 0.05 was considered to be statistically significant.

Results

APC and APC-2Cys modulate GLUT4 translocation to the membrane

AMPK activation is known to protect against I/R injury by modulating energy substrate metabolism [26,27]. In light of our recent finding that the signaling function of APC elicits potent cardioprotective activity through activation of AMPK [6], we decided to investigate whether APC and the anticoagulant-defective mutant APC-2Cys can modulate glucose metabolism during I/R. Hearts from C57BL/6 mice were isolated and subjected to 10 min of global ischemia followed by 20 min of reperfusion. Isolated hearts were then infused with bio-LC-ATB-BGPA in order to label the pool of GLUT4 localized to the cell membrane. This strategy allowed quantification of the amount of GLUT4 translocated from intracellular vesicles to the cell surface during ischemic stress. The results presented in Fig. 1A showed that both APC and APC-2Cys markedly increased I/R-stimulated GLUT4 membrane accumulation as compared with the I/R vehicle group (< 0.05).

Figure 1.

 Activated protein C (APC) increases glucose uptake during ischemia/reperfusion (I/R). (A) APC modulates glucose transporter type 4 (GLUT4) translocation to the membrane. Immunoblotting analysis of cell membrane-bound and total GLUT4 in the heart tissues. Isolated C57BL/6 mouse hearts were subjected to 10 min of global ischemia followed by 20 min of reperfusion in the ex vivo working heart perfused system. Cell surface GLUT4 was labeled with the cell membrane-impermeable compound bio-LC-ATB-BGPA. (B) C57BL/6 mouse hearts were isolated and perfused with d-[2-3H]glucose-labeled perfusion buffer in the ex vivo working heart perfused system. Isolated hearts were subjected to 10 min of global ischemia followed by 20 min of reperfusion. Perfusates were collected at 5-min intervals during reperfusion, and the production of 3H2O from d-[2-3H]glucose was measured with a scintillation counter. Values are means ± standard error, n = 6 per group, *< 0.05 vs. control, †< 0.05 vs. I/R vehicle. RLU, relative light units.

APC increases glucose uptake during I/R

Next, we investigated the functional significance of APC-mediated and APC-2Cys-mediated enhancement of the membrane accumulation of GLUT4 during I/R. Glucose uptake was analyzed by collecting samples every 5 min and measuring the production of 3H2O from d-[2-3H]glucose in the perfusate of an ex vivo working heart perfusion system. Both APC and APC-2Cys significantly enhanced I/R-induced glucose uptake as compared with the I/R vehicle groups (Fig. 1B). Intriguingly, APC-2Cys showed significantly stronger augmentation of glucose uptake than APC (< 0.05) (Fig. 1B). The capacity of APC-2Cys to mediate GLUT4 translocation to the cell surface was also slightly higher than that of wild-type APC, although the difference was not statistically significant (Fig. 1A).

Akt signaling is not involved in APC-mediated glucose transport during I/R

To understand the molecular basis of APC-mediated glucose uptake during I/R, we examined the role of Akt signaling in I/R-induced GLUT4 translocation and glucose uptake [28]. Akt is at the center of the insulin cascade, and is required for insulin-mediated glucose uptake by mediating GLUT4 translocation [29]. Isolated hearts were subjected to 10 min of ischemia followed by 20 min of reperfusion in the ex vivo working heart perfusion system, and phospho-Akt (Ser473) levels were then assessed by immunoblotting. The results presented in Fig. 2 indicate that neither APC nor APC-2Cys activated Akt during I/R, suggesting that Akt is not involved in APC modulation of glucose transport in the heart.

Figure 2.

 Activated protein C (APC) inhibits Akt signaling pathway in the heart during ischemia/reperfusion (I/R). Isolated C57BL/6 mouse hearts were subjected to 10 min of global ischemia followed by 20 min of reperfusion in the ex vivo working heart perfused system. Immunoblotting analysis of Akt phosphorylation in the heart tissue homogenates was conducted with an antibody that recognizes levels of Akt phosphorylated at Ser473 in the heart. Values are means ± standard error, n = 5–7 per group, *< 0.05 vs. control, †< 0.05 vs. I/R vehicle. RLU, relative light units.

Both APC and APC-2Cys modulate glucose oxidation during I/R

With APC increasing both GLUT4 translocation and glucose uptake in the heart during I/R, the next question was whether or not this process is correlated with an increase in cardiac glucose oxidation postreperfusion. Glucose oxidation was analyzed by measuring [14C]glucose incorporation into 14CO2 released from the ex vivo working heart after 10 min of ischemia and 20 min of reperfusion. The results demonstrated that both APC and APC-2Cys markedly increased the rates of glucose oxidation in the I/R-stressed hearts (< 0.01) (Fig. 3A). Consistent with more pronounced glucose uptake, APC-2Cys was more efficient than APC in glucose oxidation in the heart during I/R (< 0.01) (Fig. 3A). However, neither APC-E170A, which lacks cytoprotective signaling activities, nor PC-2Cys, which has no catalytic activity, had an effect on ischemic glucose oxidation, suggesting that the active site-dependent signaling function of APC is involved in modulating this metabolic process (Fig. 3A).

Figure 3.

 Activated protein C (APC) augments glucose oxidation in the heart during ischemia/reperfusion (I/R). Glucose oxidation was analyzed by measuring [14C]glucose incorporation into 14CO2 in ex vivo C57BL/6 mouse hearts subjected to 10 min of ischemia and 20 min of reperfusion. Oleate oxidation was analyzed by measuring the incorporation of [9,10-3H]oleate into 3H2O. Values are means ± standard error, n = 5–6 per group, *< 0.05 vs. control, †< 0.05 vs. I/R vehicle, #< 0.01 vs. I/R APC. PC-2Cys, protein C-2Cys.

Effects of APC derivatives on fatty acid oxidation

In light of the significant effect of APC and APC-2Cys on glucose metabolism, we decided to investigate their impact on cardiac fatty acid oxidation, the major energy source for normal cardiac metabolism [30]. Fatty acid oxidation was measured by the production of 3H2O from [9,10-3H]oleate in the ex vivo working hearts. Isolated hearts were subjected to 10 min of ischemia and 20 min of reperfusion. As shown in Fig. 3B, oleate oxidation was significantly upregulated in the heart during I/R. Interestingly, only APC-2Cys significantly reduced the rate of oleate oxidation during I/R (Fig. 3B). None of the other APC derivatives, including wild-type APC, the signaling-defective APC-E170A, and the catalytically inactive zymogen PC-2Cys, altered the level of oleate oxidation in the heart during I/R.

APC-2Cys improves the intracellular redox status in the heart during I/R

The I/R-induced acceleration of cardiac fatty acid oxidation has been shown to create more ROS than glucose oxidation [31]. Therefore, we reasoned that a reduction in oleate oxidation may lead to a decrease in ROS generation, thereby improving the intracellular redox status in the heart during I/R. The GSH/GSSG ratio has been used as an indicator of the intracellular redox status in heart tissue [32]. Thus, we measured the GSH/GSSG ratio to assess the redox status of hearts subjected to 10 min of global ischemia and 20 min of reperfusion in the ex vivo working heart perfusion system. The results showed that I/R stress impaired the intracellular redox status by decreasing the cardiac GSH/GSSG ratio (Fig. 4). APC-2Cys significantly improved the intracellular redox status of the ischemic heart, as indicated by an increase in the GSH/GSSG ratio as compared with the I/R vehicle group (Fig. 4). In contrast, neither APC nor APC-E170A had an effect on the GSH/GSSG ratio (Fig. 4). We also determined the effects of APC derivatives on the intracellular redox status of isolated cardiomyocytes. The results showed that none of the APC derivatives altered the intracellular redox status under basal conditions (data not shown).

Figure 4.

 Activated protein C (APC)-2Cys improves intracellular redox status in the heart during ischemia/reperfusion (I/R). The reduced glutathione (GSH)/oxidized glutathione (GSSG) ratio was calculated as an index of intracellular redox status. GSH/GSSG ratios were measured with a glutathione detection kit. Values are means ± standard error, n = 11 per group, *< 0.01 vs. control, †< 0.01 vs. I/R vehicle. RLU, relative light units.

APC and APC-2Cys modulate autophagy in the heart during I/R

The importance of autophagy in the heart during I/R has been shown by several studies [17]. More recently, evidence has been provided that autophagy is involved in the modulation of glucose homeostasis in the skeletal muscle in response to energy stress such as exercise [18]. Moreover, it has been demonstrated that AMPK signaling regulates autophagy [34–36]. Therefore, we tested the ability of APC to modulate autophagic flux during I/R by measuring the LC3-II/LC3-I ratio as an indicator of autophagy [18]. The results showed that I/R induced autophagy, and that both APC and APC-2Cys augmented this process (Fig. 5). Intriguingly, APC-2Cys had a more pronounced effect in modulating autophagy, which may explain the observed higher activity of APC-2Cys in regulating cardiac glucose metabolism during I/R.

Figure 5.

 Activated protein C (APC) augments ischemia/reperfusion (I/R)-triggered autophagy in the heart. Isolated C57BL/6 mouse hearts were subjected to 10 min of global ischemia followed by 20 min of reperfusion in the ex vivo working heart perfused system. Autophagy was monitored by immunoblotting analysis of microtubule-associated protein 1 light chain 3 (LC3)-II and LC3-I in the heart tissue homogenates, and defined as the LC3-II/LC3-I ratio in the heart. Values are means ± standard error, n = 6 per group, *< 0.05 vs. control, †< 0.05 vs. I/R vehicle. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RLU, relative light units.

Discussion

Previous studies have established a protective role for APC in limiting myocardial I/R injury [37–40]. We recently showed that APC exerts its EPCR-dependent and PAR-1-dependent cardioprotective effects through upregulation of AMPK signaling, largely independently of its anticoagulant function [6]. In the present study, we have shown that APC increases glucose uptake and upregulates GLUT4 translocation to the cell membrane, thereby modulating energy substrate metabolism in the ischemic heart in a beneficial manner, so that ATP is generated primarily through glucose oxidation during reperfusion. Interestingly, APC also increased the rate of autophagy and improved the intracellular redox status of the I/R-stressed heart. These results support recent interesting findings that autophagy contributes to intracellular glucose homeostasis under short-term stress conditions [18].

Activated AMPK, through phosphorylation of downstream substrates, initiates a number of biological events that culminate in the modulation of energy substrate metabolism and subsequent restoration of cardiomyocyte ATP levels [3,41]. In the ischemic heart, AMPK enhances glucose uptake by mediating the translocation of GLUT4 from storage vesicles to the cell surface [12,42]. AMPK also stimulates glycolysis by directly activating phosphofructokinase 2 (PFK-2) through phosphorylation at Ser466, which further increases the production of fructose 2,6-bisphophate, an allosteric activator of PFK-1 in the glycolytic pathway [43]. AMPK activation maintains enhanced glucose uptake during the initial reperfusion stage to improve cardiac contractile functions, as shown by transgenic mice expressing a kinase dead mutation exhibiting impaired glucose uptake and post-ischemic contractile function [12]. The results of this study now demonstrate that both APC and APC-2Cys increase the cell surface accumulation of GLUT4, thereby leading to augmentation of glucose uptake and glucose oxidation during I/R. These results, together with our previous findings, indicate that both APC and APC-2Cys modulate cardiac glucose metabolism through activation of the AMPK signaling pathway [6].

It has been demonstrated that insulin can also exert a cardioprotective effect during ischemic stress [15,44]. In this case, however, the cardioprotective effect of insulin is mediated through the phosphoinositide 3-kinase-dependent activation of Akt [15,44]. Insulin-dependent Akt phosphorylation has been shown to mediate the translocation of GLUT4 to the plasma membrane, thereby leading to the enhancement of glucose uptake and its oxidation by the I/R-stressed heart. The finding in this study that APC inhibited the phosphorylation of Akt suggests that APC-mediated GLUT4 translocation in the ischemic heart is mediated by a different mechanism that is independent of Akt signaling. Actually, an inverse correlation has been observed between Akt and AMPK signaling in the ischemic heart [44]. APC exerts its cardioprotective and anti-inflammatory activities through the Gla-dependent interaction with EPCR [9], followed by its activation of PAR-1 [10,11], localized to membrane lipid rafts [45]. Thrombin can also activate PAR-1 [11]. In this case, however, activation of PAR-1 initiates a proinflammatory response [11]. We have demonstrated that the interaction of the Gla-domain of APC with EPCR in the lipid raft microenvironment switches the signaling specificity of PAR-1 from a proinflammatory to a protective response [46]. Given that both APC and APC-2Cys have normal affinity for EPCR, and that both proteases cleave PAR-1 with similar efficiency [40], the molecular basis of the significantly higher modulatory activity of APC-2Cys towards glucose metabolism remains unknown. Nevertheless, the active site-dependent signaling function of the protease was required for its ability to modulate glucose metabolism in the heart, as neither the zymogen PC-2Cys nor the signaling-defective APC-E170A mutant exhibited a glucose modulatory effect in the I/R-stressed heart. It is worth noting that, in addition to EPCR and PAR-1, the protective signaling activity of APC also requires receptor crosstalk and/or protease interaction with a number of other G-protein-coupled (i.e. protease-activated receptor-3 [47] and sphingosine 1-phosphate receptor 1 [11]) and non-G-protein-coupled (i.e. apolipoprotein E receptor 2 [48] and Tie2 [49]) receptors. Thus, further studies will be required to determine whether possible differences in the binding affinities of APC and APC-2Cys for any one of these receptors contribute to their differential modulatory effects on substrate metabolism in the I/R-stressed heart.

It was interesting to note that, unlike its enhanced activity in modulating glucose metabolism, APC-2Cys markedly inhibited the oxidation of oleate in the heart during I/R. This function of APC-2Cys may be beneficial for the ischemic heart, because, during reperfusion, a sudden increase in the oxygen level can lead to enhanced fatty acid oxidation and subsequent generation of more ROS, which can cause greater cardiac damage [50]. There are reports suggesting that cardiac AMPK activation accelerates fatty acid oxidation in the heart [13,51]. Similarly, the results obtained with the ex vivo working heart model in this study showed an increase in oleate oxidation during reperfusion after ischemia. Interestingly, whereas wild-type APC did not significantly affect oleate oxidation, APC-2Cys dramatically inhibited cardiac oleate oxidation during I/R. The molecular basis of the APC-2Cys-mediated inhibition of oleate oxidation may be its upregulation of glucose oxidation, which can lead to decreased fatty acid β-oxidation via the feedback inhibition of 3-ketoacyl-CoA thiolase and 3-hydroxyacyl-CoA dehydrogenase [19]. Thus, the APC-2Cys-mediated increase in glucose oxidation can improve the coupling of glucose metabolism, thereby decreasing proton production and improving cardiac efficiency [19]. In line with the beneficial effect of enhanced glucose metabolism during I/R, several fatty acid β-oxidation inhibitors have been developed as therapeutic drugs and proven to be highly effective against ischemic heart disease. Two such drugs, trimetazidine and ranolazine, are both partial fatty acid β-oxidation inhibitors that reciprocally increase glucose oxidation [52,53]. Therefore, not withstanding the controversial nature of the question as to whether or not elevated glucose oxidation at the expense of reduced fatty acid oxidation in the ischemic heart is beneficial [54], APC-2Cys has potential therapeutic utility as a fatty acid oxidation inhibitor for the treatment of ischemic heart disease without increasing the risk of bleeding that may be associated with the use of wild-type APC.

APC is a metal ion-binding protease that requires both Ca2+ and Na+ for its normal catalytic function [21]. APC-2Cys is a mutant in which the Ca2+-binding loop of the protease in the catalytic domain has been stabilized by an engineered disulfide bond [21]. Given that the Ca2+-binding and Na+-binding loops of APC are allosterically linked, this construction strategy abrogates the requirement for the metal ions by the protease, and also eliminates the anticoagulant function of APC without significantly affecting its cytoprotective signaling properties [21]. In light of the observation that APC-2Cys promoted cardiac glucose metabolism and inhibited fatty acid oxidation during reperfusion, we postulated that this property of the mutant may arise from a small fraction of the protease potentially existing in equilibrium with the reduced form of the protease, in which the engineered cysteines are not engaged in a disulfide bond. In this case, this small fraction of free cysteine-containing protease could function as a sink for the ROS produced during I/R, thus leading to alterations in the lipid raft redox signaling mechanism in the membrane of cardiomyocytes [55]. This possibility was tested by measuring glucose and oleate oxidation in the working heart model with a reperfusion buffer containing 20 nm GSSG, which is sufficient to quench the potential trace amount of free cysteines in the mutant protease. However, we found that addition of GSSG to the buffer does not affect either oleate or glucose oxidation rates of the APC-2Cys groups, ruling out the possibility that free cysteines, if there are any, in APC-2Cys contribute to its unique metabolic modulatory properties. Further support for this hypothesis was provided by the observation that the zymogen form of the mutant (PC-2Cys) had no modulatory effect on either glucose or fatty acid oxidation. Nevertheless, APC-2Cys did demonstrate stronger antioxidative activity, as shown by the increased GSH/GSSG ratio during I/R, which could somehow contribute to modulation of the lipid raft redox signaling system of the cell membrane, thereby differentially affecting glucose and fatty acid oxidation by the mutant protease.

Finally, autophagy, defined as ‘self-eating’, has been recognized as an important process in modulating glucose homeostasis under stress conditions [18,56]. Therefore, we tested the effect of APC and APC-2Cys on the stimulation of autophagy in the heart during I/R. Both APC and APC-2Cys were found to significantly augment I/R-triggered autophagy in the heart, with APC-2Cys demonstrating a significantly greater potency in modulating I/R-induced autophagy. This property of APC-2Cys may further contribute to the mutant protease’s stronger modulatory activity in glucose metabolism in the ischemic heart. Although autophagy is thought to be protective in the ischemic heart, it has recently been shown that excessive autophagy can also be detrimental [57], so further studies are required to understand the mechanisms by which APC regulates autophagy in the ischemic heart and also to decipher the importance of autophagy in general during I/R injury.

Taken together, the results of this study suggest that APC modulation of energy substrate metabolism may exert a cardioprotective effect in the ischemic heart, thereby reducing the incidence of myocardial infarction. This observation may potentially lead to the development of APC-based novel therapeutics for better treatment of myocardial I/R injury in humans.

Acknowledgements

We would like to thank A. Rezaie for proofreading the manuscript. This work was supported by grants awarded by the National Heart, Lung, and Blood Institute of the National Institutes of Health (HL 101917 and HL 68571; A. R. Rezaie), and by grants awarded by the American Heart Association (0835169N and 12GRNT11620029) and the American Diabetes Association (1-11-BS-92) (J. Li).

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.

Ancillary