Activation of Pak1/Akt/eNOS signaling following sphingosine-1-phosphate release as part of a mechanism protecting cardiomyocytes against ischemic cell injury

Abstract

We investigated whether plasma long-chain sphingoid base (LCSB) concentrations are altered by transient cardiac ischemia during percutaneous coronary intervention (PCI) in humans and examined the signaling through the sphingosine-1-phosphate (S1P) cascade as a mechanism underlying the S1P cardioprotective effect in cardiac myocytes. Venous samples were collected from either the coronary sinus (n = 7) or femoral vein (n = 24) of 31 patients at 1 and 5 min and 12 h, following induction of transient myocardial ischemia during elective PCI. Coronary sinus levels of LCSB were increased by 1,072% at 1 min and 941% at 5 min (n = 7), while peripheral blood levels of LCSB were increased by 579% at 1 min, 617% at 5 min, and 436% at 12 h (n = 24). In cultured cardiac myocytes, S1P, sphingosine (SPH), and FTY720, a sphingolipid drug candidate, showed protective effects against CoCl induced hypoxia/ischemic cell injury by reducing lactate dehydrogenase activity. Twenty-five nanomolars of FTY720 significantly increased phospho-Pak1 and phospho-Akt levels by 56 and 65.6% in cells treated with this drug for 15 min. Further experiments demonstrated that FTY720 triggered nitric oxide release from cardiac myocytes is through pertussis toxin-sensitive phosphatidylinositol 3-kinase/Akt/endothelial nitric oxide synthase signaling. In ex vivo hearts, ischemic preconditioning was cardioprotective in wild-type control mice (Pak1^f/f), but this protection appeared to be ineffective in cardiomyocyte-specific Pak1 knockout (Pak1^cko) hearts. The present study provides the first direct evidence of the behavior of plasma sphingolipids following transient cardiac ischemia with dramatic and early increases in LCSB in humans. We also demonstrated that S1P, SPH, and FTY720 have protective effects against hypoxic/ischemic cell injury, likely a Pak1/Akt1 signaling cascade and nitric oxide release. Further study on a mouse model of cardiac specific deletion of Pak1 demonstrates a crucial role of Pak1 in cardiac protection against ischemia/reperfusion injury.

sphingolipids are biologically active lipids that play important roles in various cellular biological processes (12). Recent studies indicate that myocardial sphingosine (SPH) levels are elevated in animal models of myocardial infarction (47, 52) and have suggested that SPH have an important role in protection against ischemic injury (19). Sphingosine-1-phosphate (S1P) has also been shown to be an important mediator of cardiac ischemic pre- and postconditioning in both pharmacological and knockout animal studies (15, 26). FTY720 (fingolimod), a sphingolipid drug candidate displaying structural similarity to S1P, demonstrated a protective effect in preventing organ ischemia/reperfusion (I/R) injury in an animal model (7, 11). Thus it appears highly likely that these sphingolipid molecules and their analogs have the potential to act as therapeutic modulators of cardiac responses to myocardial injury. The underlying detailed key mechanism(s) and signaling pathway(s) for their cardioprotective effect, however, remain poorly understood (18).

A recent study (11) suggests that the activation of Akt underlies the protective effects of S1P receptor agonist treatment after myocardial I/R, which opens the door for understanding the key signaling mechanism(s) for S1P cardioprotection. Another significant clue as to the mechanism came from experiments in a mammalian cell line, which demonstrated that p21-activated kinase (Pak1), a Ser/Thr kinase downstream of small G proteins, is activated by sphingosine and several related long-chain sphingoid bases (LCSBs) in a time- and dose-dependent manner (2). We (7) recently also showed that FTY720 prevents arrhythmias induced by I/R injury by activation of signaling through the Pak1 and Akt cascade. However, to date, neither the signaling pathways mediating these effects nor an in vivo metabolism for endogenously released S1P has been established during acute cardiac ischemic conditions. Accordingly, we designed both an in vivo study to determine whether plasma sphingolipids measured as LCSB concentrations were altered by transient cardiac ischemia following temporary coronary artery occlusion during percutaneous coronary intervention (PCI) in humans and in vitro experiments to examine the cell-survival pathways regulated by S1P agonist FTY720 in cardiomyocytes.

METHODS AND MATERIALS

Human subjects study protocol.

Blood samples were obtained from 31 patients aged 40 to 73 yr old undergoing elective PCI to native coronary arteries at the Manchester Heart Centre (Manchester, UK). The study complies with the Declaration of Helsinki in that the Manchester Royal Infirmary Ethics Committee approved the research protocol and that informed consent was obtained from the subjects. Demographic data were as follows: men (97%), Caucasions (87%), normal left ventricular (LV) function (100%), and normal renal function (100%). Incidence of risk factors was as follows: hypertension (53%), diabetes (15%), hyperlipidaemia (85%), smokers (39%), and body mass index (27.4 ± 4.8 kg/m²). Medication was as follows: antiplatelet therapy (100%), β-adrenoceptor blockers (82%), angiotensin-converting enzyme inhibitors (65%), statins (100%), nitrates (22.5%), and Ca²⁺-channel antagonists (29.5%). Only patients with angiographic single vessel disease undergoing elective PCI were used in this study who had documented normal left ventricular and renal function. Patients with a previous history of coronary artery bypass graft, valvular heart disease, or recent myocardial infarction/acute coronary syndrome were excluded as were PCI procedures in patients with chronic total occlusions.

Procedures were performed via the femoral artery through standard 6-Fr sheaths and peripheral venous samples were collected through a 6-Fr femoral venous sheath. Coronary sinus (CS) sampling was performed using a 6-Fr Amplatz left-1 catheter (AL-1) during PCI. Predilation of the target lesions was performed with angioplasty balloons inflated between 14 to 22 atmospheres (mean 15 atmospheres) for a period of between 28 to 40 s (mean 31.1 s). Control venous blood samples were obtained either from the CS (7 patients) or via the femoral venous sheath (24 patients) once the guide catheter and guide wire were in position prior to the PCI procedure. Balloon inflations of between 30 s and 1 min were performed to predilate the target lesions. Serial venous samples were then collected from either the CS or femoral vein at 1 min and 5 min postballoon inflation. PCI was then completed as per routine at our center. Twelve hours postprocedure, samples were taken from a peripheral vein to measure LCSB levels. Blood samples were immediately dispensed into 3-ml EDTA tubes with 2-chloroadenosine (0.05 mmol/l) and procaine hydrochloride (0.154 mol/l) and equilibrated at 4°C.

Only patients with angiographic single vessel disease undergoing elective PCI were used in this study who had documented normal left ventricular and renal function. Patients with a previous history of coronary artery bypass graft, valvular heart disease, or recent myocardial infarction/acute coronary syndrome were excluded as were PCI procedures in patients with chronic total occlusions.

Spectrophotometry analysis of long-chain base fraction.

Procedures for the quantitative analysis of LCSB have been based on the determination of the long-chain base of sphingosine and the analysis of other sphingolipid components. A direct analysis for sphingosine is the preferred analytical approach to sphingolipid analysis since all known sphingolipids contain one molecule of sphingosine per molecule of sphingolipid. As all glycosphingolipids and sphingomyelin contain one molecule of long-chain base, the total amount of these lipids can be appreciated by the amount of fatty bases present in the hydrolyzed extract. Several procedures may be used but the procedure reported here were based on the determination of the long-chain base of SPH as described by Siakotos and colleagues (13, 44).

Isolation and culture of rat ventricular cardiomyocytes.

Neonatal rat ventricular cardiomyocytes were prepared and cultured from 2- to 3-day-old rats, as described previously (34). Experiments were performed after further 24-h cultivation.

Simulated ischemia model.

Cells were washed with PBS before addition of 1 ml ischemia buffer (in mM: 130 NaCl, 10 KCl, 0 glucose, 0.6 MgCl₂, 1.8 CaCl₂, 1 NaHCO₃, 0.6 NaH₂PO₄, and 10 HEPES, pH 6.6, gassed with 100% nitrogen for >30 min before the experiment was started). Subsequent experiments were performed using 24-h exposure to 100 μM cobalt chloride (CoCl₂). Cobalt has been widely used as an ischemia injury mimic in cell culture (49).

Langendorff-perfused hearts ischemic preconditioning study.

Langendorff-perfused hearts with ischemic preconditioning (IPC) protocol in cardiomyocyte-specific Pak1 knockout mice (Pak1^cko) and their wild-type control mice (Pak1^f/f). Generation of Pak1^f/f and Pak1^cko mice were described elsewhere by Liu et al. (27), hearts were equilibrated for 20 min and then subjected to three short cycles of I/R, each consisting of 2 min of global ischemia and 2 min of reperfusion, followed immediately by 30-min global ischemia and 30-min global reperfusion.

Measurement of lactate dehydrogenase activity.

The cells were subjected to global ischemia or CoCl₂ treatment for 20 min or 24 h, respectively. After the completion of simulation, the buffer was gently collected for lactate dehydrogenase (LDH) determination using a spectrophotometric LDH enzyme assay kit (Cayman Chemical, Cambridge, UK) as described previously. (28)

Determination of intracellular nitric oxide bioavailability.

Intracellular nitric oxide (NO) bioavailability was measured as described previously (14). Fluorescence was detected with a live cell imaging Leica AS MDW inverted fluorescence microscope (Leica Microsystems) in which the cells were kept at 37°C.

Western blotting analysis.

Western blot analysis was conducted as described previously (14) using specific antibodies described below.

Antibodies.

Rabbit anti-Pak1, rabbit anti-phospho-Pak1 Thr423, rabbit anti-Akt, and anti-phospho-Akt Thr308 residue, were all purchased from Cell Signaling Technology (Hitchin, UK).

Statistical analysis.

Data are expressed as means ± SE. For spectrophotometry data, repeated-measures one-way ANOVA was used to compare values of measurements between groups. When ANOVA revealed the existence of a significant difference among values, Tukey's test was applied to determine the significance of a difference between selected group means.

RESULTS

Spectrophotometry analysis of PCI-induced changes to plasma LCSB levels.

To determine if sphingolipids are involved in pathophysiological processes associated with acute cardiac ischemia, we measured LCSB concentrations in transient cardiac ischemia following temporary coronary artery occlusion during PCI in humans.

Plasma levels of LCSB in patients at baseline (preballoon inflation) and at different time points after balloon inflation were analyzed by spectrophotometry. Significant alterations were found in plasma levels of LCSB sampled from the CS in 7 patients, and peripheral veins in 24 patients, following induction of transient myocardial ischemia by balloon occlusion of target lesion. At baseline, LCSB levels from CS and peripheral blood samples averaged 1.7 ± 0.9 (n = 7) and 0.03 ± 0.005 μM (n = 31), respectively. There was a significant increase in LCSB levels at 1 and 5 min, compared with baseline levels, both in CS blood (Fig. 1A) and peripheral blood (Fig. 1B).

Fig. 1.

Changes in long-chain sphingoid base (LCSB) concentrations in coronary sinus (CS) blood (A) and comparative concentrations in peripheral blood (B) at different time-course points following balloon inflation. Values are means ± SE. *P < 0.001.

LCSB levels in CS and peripheral blood at different time points are shown in Fig. 1. At 1 min following balloon inflation in the CS, levels of LCSB increased by 1,072% compared with baseline levels (n = 7; all P < 0.001), whereas in peripheral blood levels of LCSB increased by 579% compared with baseline levels (n = 24; all P < 0.001). Peripheral sphingolipid levels were consistently very much lower than CS levels. At 5 min after balloon inflation in the CS blood, levels of LCSB increased by 941% compared with baseline levels (n = 7; all P < 0.001), while in peripheral blood, levels of LCSB increased by 617% compared with baseline levels (n = 24; all P < 0.001). At 12 h following the PCI procedure, in peripheral blood, levels of LCSB increased by 436% compared with baseline levels (n = 24; all P < 0.001; 95% confidence interval). These results implicate an important role of sphingolipids in pathophysiological processes that occur during early cardiac ischemia. It is known that I/R injury can be significantly minimized by a cardiac self-protective mechanism called ischemic preconditioning, a phenomenon describing a brief period of myocardium I/R that significantly reduces injury resulting from subsequent long-term I/R. Therefore, the release of LCSB might be involved in such a self defense system.

Effects of S1P, SPH, and FTY720 on cell viability in in vitro hypoxic and ischemic cell models.

The effects of S1P, SPH, and FTY720 on the viability of myocytes subjected to hypoxia and ischemia were first examined using in vitro cell models. Viability was gauged by LDH activity, a stable enzyme normally found in the cytosol of all cells, but rapidly released upon damage to the plasma membrane. The increase of LDH in the culture supernatant fraction provided a measurement of the number of lysed/damaged cells.

As shown in Fig. 2A, after 24-h exposure with 100 μM CoCl₂ to induce hypoxia, LDH release increased 10-fold from 2.55 ± 0.16 mU/ml in control conditions to 25.17 ± 0.20 mU/ml (n = 5, P < 0.01). In contrast, when the cells were subjected to a 24 h CoCl₂ treatment in the presence of 25 nM of S1P, FTY720, or SPH, respectively (Fig. 2A), there was a significant reduction of LDH released compared with cells treated with CoCl₂ alone. LDH decreased by 45.0% in the S1P-treated group, 44.7% in the FTY720-treated group, and 91% in the SPH-treated group. This indicates a protective effect against CoCl₂-induced hypoxia by these molecules.

Fig. 2.

Effects of sphingosine-1-phosphate (S1P), FTY720, and sphingosine (SPH) on CoCl₂-induced hypoxia (A) and on ischemic solution-induced ischemia (B). Protective effect of pretreatment with 100 μM CoCl₂ or with ischemic solution was studied in the presence or absence of S1P (25 nM), FTY720 (25 nM), and SPH (25 nM; 10 μM). *P < 0.01 vs. control. §P < 0.01 vs. ischemia or CoCl2.

As shown in Fig. 2B, a 20-min exposure to ischemic solution alone increased LDH release from 1.56 ± 0.11 at baseline to 2.55 ± 0.16 mU/ml (n = 5, P < 0.01). When the cells were subjected to a 20-min ischemic solution treatment in the presence of 25 nM S1P, 25 nM FTY720, or 25 nM SPH, respectively (Fig. 2B), LDH release was reduced by 40.0, 47.5, and 45.1% compared with the cells treated with ischemic solution alone, which indicates that these molecules display protective effects against ischemic induced cell injury.

FTY720 induces Pak1 and Akt activation.

Different cardioprotective signaling pathways converge to Pak1- and phosphatidylinositol 3-kinase (PI3K)-mediated Akt activation. Thus we focused on Akt and Pak1, a Ser/Thr kinase downstream of small G proteins, to clarify the key signaling mechanism(s) for the observed sphingolipid cardioprotection effect. Pak1 and Akt activation was assessed by quantifying the levels of (Thr423)-phosphorylated Pak1 and (Thr308)-phosphorylated Akt in FTY720-treated cardiomyocytes by Western blotting. Compared with control nontreated myocytes, cells treated with 25 nM FTY720 for 15 min showed an increase in phospho-Pak1 and phospho-Akt levels by 56 and 65.6%, respectively (P < 0.05 vs. control; Fig. 3, A, B, E, and F). There was no significant difference in the levels of total Pak1 and total Akt expression levels between nontreated and treated myocytes with FTY720 (Fig. 3, A, B, E, and F).

Fig. 3.

FTY720 induces Pak1 and Akt activation. A and B: representative blots for activation of phospho-Akt and phospho-Pak1 in isolated neonatal rat myocytes treated with FTY720. Myocytes were exposed for 15 min to 25 nM FTY720. C and D: representative blots for Akt and phospho-Akt proteins in control or 25 nM FTY720 plus 100 ng/ml PTX and for Pak1 and phospho-Pak1 proteins in control or FTY720 plus PTX. Myocytes were treated with FTY720/PTX for 15 min and then assayed for phosphorylation of Akt and Pak1 and by Western blotting. E and F: Akt and phospho-Akt proteins and Pak1 and phospho-Pak1 (Thr 423) were quantified by Western blotting from 4 experiments. Values are means ± SE (n =4 for each group). *P < 0.05 vs. vehicle. G and H: phosphorylation was quantified by densitometry and normalized to vehicle for Akt and phospho-Akt proteins (G) and Pak1 and phospho-Pak1 proteins (H) from 4 experiments. Values are means ± SE (n =4 for each group). *P < 0.05 vs. vehicle.

FTY720 stimulates NO production via a pertussis toxin-sensitive PI3K/Akt/endothelial NO synthase cascade.

To determine whether FTY720-mediated Pak1 and Akt activation, and NO release, were through G_i, myocytes were treated with 100 ng/ml pertussis toxin (PTX) overnight and then stimulated with 25 nM FTY720 for 15 min. As shown in Fig. 3, C, D, G, and H, FTY720 induced a 1.6-fold increase in Pak1 phosphorylation and a 1.45-fold increase in Akt phosphorylation relative to vehicle. After PTX treatment, FTY720-mediated activation of Pak1 was reduced by 94%, and activation of Akt was reduced by 55%. FTY720-induced NO release was abolished (data not shown).

Different cardioprotective signaling pathways also converge to release NO (29, 30). We thus determined how S1P/FTY720 induced NO production that results in cardioprotection. We detected NO release using DAF-FM, an NO-sensitive dye. As shown in Fig. 4, A and B (red arrows), myocytes exposed to 25 nM FTY720 and 25 nM SEW2871 (a specific S1P1 receptor agonist) displayed NO vesicles localized in specific areas near the cell membrane which disappear over time. However, NO vesicles were not observed when cells were preincubated for 1 h with 1 mM nitro-l-arginine methyl ester [l-NAME; a potent inhibitor of endothelial nitric oxide synthase (eNOS), inducible nitric oxide synthase (iNOS), and neuronal nitric oxide synthase (nNOS)] or with 1 μM N⁵-(1-iminoethyl)-l-ornithine (NIO; a potent inhibitor of eNOS) or for 30 min with 1 μM of A6730 (a Akt1/2 kinase inhibitor; Fig. 4, C, F, and G) or with 50 μM of LY294002 (a PI3K inhibitor; Fig. 4H), respectively. When cells were pretreated with 10 μM N⁶-(1-iminoethyl)-l-lysine (NIL), a potent and selective inhibitor of iNOS, and 1 μM SMLT, a potent inhibitor of nNOS, for 60 min, the appearance of FTY720-induced NO vesicles was still observed, although these were fewer in number (Fig. 4, D and E), suggesting that the effect of FTY720 is mainly through PI3K/Akt/eNOS signaling pathway. This ultimate finding is indirect evidence that FTY720 might trigger the production and the release of NO by cells.

Fig. 4.

Intracellular nitric oxide (NO) bioavailability. Intracellular NO levels were assessed in neonatal rat cardiomyocytes by staining with NO-sensitive dye (DAF-FM). A and B: effect of 25 nM FTY720 on NO production (A) and effect of 25 nM SEW2871 (a specific S1P1 receptor agonist) on NO production (B). C: effect of 25 nM FTY720 on NO production in the presence of nitro-l-arginine methyl ester [l-NAME; potent inhibitor of endothelial nitric oxide synthase (eNOS), inducible nitric oxide synthase (iNOS), and neuronal nitric oxide synthase (nNOS)]. D: effect of 25 nM FTY720 on NO production in the presence of S-methyl-l-thiocitrulline (SMLT; nNOS inhibitor). E: effect of 25 nM FTY720 on NO production in the presence of N⁶-(1-iminoethyl)-l-lysine (NIL; iNOS inhibitor). F: effect of 25 nM FTY720 on NO production in the presence of N⁵-iminoethyl-l-ornithine (NIO; potent inhibitor of eNOS). G: effect of 25 nM FTY720 on NO production in the presence of LY294002 [potent phosphatidylinositol 3-kinase (PI3K) inhibitor]. H: effect of 25 nM FTY720 on A6730 (Akt1/2 kinase inhibitor). n = 64 cells from 3 independent cell preparations in each group.

Disruption of Pak1 sensitizes the myocardium to I/R-induced ventricular arrhythmias.

IPC is a short period of I/R that rescues hearts from subsequent long term I/R injury. Reported triggers are adenosine, bradykinin, opioids, and S1P (4, 14, 16). Evidence suggests a definitive role of sphingolipid metabolites, including S1P, as important members of the IPC-mediated intracellular signaling process (14). In our previous study, we provided evidence that FTY720 prevents I/R injury associated arrhythmias and the cardioprotective effect of FTY720 is likely to involve activation of signaling through the Pak1. However, there is no direct evidence for the involvement of Pak1 in IPC signaling. Thus we took advantage of our recently developed mouse model of cardiac-specific deletion of Pak1 (27) to determine the role of Pak1 in IPC. Both WT control and Pak1^cko hearts were subjected to IPC protocol as described in the methods section. Figure 5 shows representative ECG recordings from hearts before and after subjected to IPC protocol. As summarized in Table 1, ventricular arrhythmic events including nonsustained and sustained episodes of ventricular tachycardia and ventricular fibrillation were more frequently observed in Pak1^cko hearts (5 out of 6 mice studied) than control WT hearts three out eight mice studied. Thus a disruption of Pak1 reduced the cardiac protection of IPC, in other words, a disruption of Pak1 sensitizes the myocardium to I/R-induced ventricular arrhythmias.

Fig. 5.

Representative ECG recordings from hearts before and after subjected to IPC protocol. Ventricular arrhythmic events including nonsustained and sustained episodes of ventricular tachycardia and ventricular fibrillation were more frequently observed in Pak1^cko hearts than control wild-type hearts. A–C: ECG recordings before ischemic preconditioning (IPC). C and D: ECG recordings during reperfusion.

Table 1.

VT/VF occurrence during reperfusion after IPC

	During Reperfusion
Genotype	Transient VT/VF	Sustain VT/VF	Total VT/VF Occurrence
WT	3/8 (37.5%)	0/8 (0%)	3/8 (37.5%)
PAK1CKO	2/6 (33.3%)	3/6 (50%)	5/6 (83.3%)

n = no. of mice studied. WT, wild type; VT/VF, ventricular tachycardia and ventricular fibrillation; IPC, ischemic preconditioning.

DISCUSSION

Transient coronary ischemia induced changes of plasma LCSB levels in human subjects.

Our study, for the first time, demonstrates a dramatic increase in LCSB levels following transient cardiac ischemia in humans. The levels of LCSB in both CS and peripheral vein markedly increased within 1 min of transient ischemia mediated by short periods of coronary vessel occlusion. The levels remained elevated after 5 min and started to decline after 12 h peripherally.

Sphingolipids are now emerging as important signaling molecules produced by cardiac tissue during ischemic stress or as a consequence of inflammation. Evidence indicating that cardiac tissue may be a putative source of serum sphingolipids comes from studies showing that cardiac cells produce sphingolipids in response to hypoxia and reoxygenation (1) as well as the finding that the ischemia caused by coronary microembolization in a canine model results in significant elevation in myocardial sphingolipids levels (47). Additionally, Deutschman et al. (5) reported that sphingolipid levels are elevated in patients with coronary artery disease where S1P had a greater predictive value in detecting coronary artery disease than traditional risk factors. The high concentration of LCSB levels seen in the CS compared with the periphery suggests that this increase in circulating LCSB are derived from the myocardium, rather than being released elsewhere as a consequence of myocardial ischemia.

Evidence is rapidly accumulating that suggests a definitive role of sphingolipid metabolites, including S1P, as important members of the IPC-mediated intracellular signaling process (14). It has also been shown that the protective effect of IPC could be mimicked by exogenous S1P, suggesting the possibility that ceramide and sphingosine formed during preconditioning must be converted to S1P to protect against I/R injury (30).

Cardioprotective effects of S1P, SPH, and FTY720.

Recent studies (14, 39, 50, 51) have provided evidence for a role of S1P signaling in protection against the stress of cardiac I/R injury. S1P, product of sphingosine kinase (SphK) activation, is recognized as a vital lipid mediator activating a family of five G-protein-coupled receptors (S1P1–5). These receptors regulate diverse cellular events, including cell survival, growth, motility, differentiation, cytoskeletal reorganization, and calcium mobilization (39). S1P is a phosphorylated derivative of sphingosine, the structural backbone of all sphingolipids, which was initially described as an intermediate in the degradation of LCSBs (21). FTY720 is a sphingolipid drug candidate displaying structural similarity to S1P and has demonstrated a protective effect in the prevention of liver I/R injury in an animal model (20). Importantly, FTY720 is currently under evaluation in a long-term, phase III clinical trial as an immunosuppressant agent for the treatment of autoimmune diseases and in organ transplantation (3). Our recent study on rat ex vivo heart model also demonstrated the important role of FTY720 in antagonizing both brady- and tachy-arrhythmias induced by I/R injury. The study also determined that FTY720 acts through Pak1/Akt signaling, which identifies this cascade as an important element in this cardiac protection effect. Our data (18, 39) also significantly extend earlier reports providing evidence for a role of S1P signaling in protection against I/R injury and its potential role in pre- and postconditioning.

In the present study, in cultured cardiac myocytes, S1P, SPH, and FTY720 all showed protective effects against hypoxic or ischemic cell injury by reducing LDH activity, which is consistent with previous findings (7, 11, 50, 51). Several studies (7, 11, 50, 51) have provided evidence for a role of S1P signaling in protection against the stress of I/R injury, in particular its critical role in pre- and postconditioning mechanisms of rescue of hearts from I/R injury. It is highly relevant that FTY720 is currently under evaluation in a long-term, Phase III clinical trial for the treatment of autoimmune disease and in organ transplantation (3, 41). FTY720 was found to attenuate liver graft injury by activation of cell-survival Akt signaling (54). We (7) recently also demonstrated the cardioprotective effect of FTY720 in an ex vivo rat heart I/R injury model.

Mechanism(s) underlying FTY720 cardioprotection in ischemia/hypoxia model.

In the present study, our data showed that FTY720 is able to stimulate, via the inhibitory G proteins, Pak1 and Akt autophosphorylation and activities and trigger NO release through eNOS. The NO release via S1P receptors appears to be mediated by PI3K and Akt activation. NO release induced by FTY720 was sensitive to G_i inhibition. We believe this is a strong evidence to demonstrate the mechanistic link between FTY720 and Pak1 and Akt activations. We (7) recently established cause-effect relationship for the observed phosphorylation of Pak-1 and Akt to the cardioprotection. In that study, to clarify whether the observed prevention of arrhythmias by FTY720 in I/R model was attributed to the activation of Pak1/Akt, activation of Pak1/Akt during I/R model was determined in I/R Langendorff hearts. The same hearts used for the functional-arrhythmias studies (made ischemic in the presence or absence of FTY720) were used to check the phosphorylation states of Pak1/Akt. This provided a potential link between the functional effects of FTY720 and intracellular signaling.

Thus the data presented here provide a new insight into the signaling pathways underlying an FTY720 cardioprotective mechanism. As illustrated in Fig. 6, signaling through activation of S1P receptors by S1P and its analog, FTY720, to Pak1/Akt/eNOS may serve as a key mechanism underlying the S1P/FTY720 cardioprotective effect.

Fig. 6.

Hypothesized mechanisms for sphingolipids in cardioprotective signaling in response to ischemia or hypoxia. First, ischemia/reperfusion activates sphingosine kinase 1 (SK1) through a PKCε-dependent mechanism and induces the enzymatic processing or “shedding” of membrane-bound TNFα (pro-TNF) by metallomatrix proteases (also known as TACE, or TNFα-converting enzyme) in the extracellular matrix of the cardiomyocyte. TNFα is released, which acts in a paracrine fashion on cardiomyocyte TNFα receptors (TNFRI) that, in turn, activate the sphingomyelin signal transduction system. The principal signaling molecule produced by the TNFα trigger is SPH, the majority of which crosses the sarcolemma membrane and is released into the extracellular fluid compartment. It is also known that blood platelets and other blood components that possess SK1 convert SPH to S1P. S1P is cardioprotective through both intracellular and putative “inside-out” pathways, the latter involving S1P1 and S1P3 receptors. SPH is also cardioprotective through a PKG/PKA-dependent pathway. Therefore, in contrast to S1P, the cardioprotective effects of sphingosine may not be mediated through S1P-specific G-protein-coupled receptor. PCI, percutaneous coronary intervention.

We report here that FTY720 is able to stimulate Pak1 and Akt autophosphorylation and activities in cardiomyocytes. Akt is a well-established regulator of myocardial growth and survival, contractile function, and coronary angiogenesis (43). Studies (9) showed that Pak1 is able to activate Akt whereas Akt can phosphorylate Pak1. Mao et al. (31) recently showed that both Pak1 and Akt can be activated by multiple hypertrophic stimuli and growth factors in a PI3K-dependent manner, suggesting that Pak1 and Akt may lie in the same signaling pathway in cardiomyocytes. Using both gain- and loss-of-function approaches in vitro and in vivo, the authors demonstrated that Pak1 is sufficient to activate Akt and is essential for growth factor-induced Akt activity in cardiomyocytes (31). The functional significance of Pak1-Akt signaling is underscored by the observation that the prosurvival effect of Pak1 is diminished by Akt inhibition (31). The Pak1-conferred protection was blocked by the Akt inhibitor X, suggesting that the protective effect of Pak1 is mediated, at least in part, by Akt signaling (31). These findings demonstrate an important role for Pak1-Akt signaling in cardiomyocyte survival. As show in Fig. 5 and Table 1, our further experiments on a mouse model of cardiac-specific deletion of Pak1 demonstrate a crucial role of Pak1 in cardiac protection against I/R injury. The recognition of the functional significance of Pak1 in preventing arrhythmogenesis associated with I/R injury may lead to the development of better therapies for treating I/R injury-induced ventricular arrhythmias.

Akt is a well-established mediator of cardioprotection in I/R injury, as demonstrated by transfection, gene delivery, and transgenic approaches (8, 32). However, mechanisms of Akt-mediated cardioprotection are still under intense investigation. Akt has been shown to increase eNOS phosphorylation, resulting in activation of eNOS, and experiments with eNOS knockout mice indicate a role for NO in protection against ischemic damage (17, 33). Described mechanisms for NO-mediated cytoprotection include antiapoptotic (22), antioxidant (45), anti-inflammatory (25), and cGMP-mediated effects (6). In endothelial cells, S1P activates eNOS via an Akt-mediated pathway, and this occurs via the S1P3 receptor (38). Kwon et al. (24) showed S1P increases eNOS activity via intracellular Ca²⁺ mobilization and resulting increases in NO protect endothelial cells from apoptosis by suppression of apoptotic signaling cascades. Importantly, that study revealed that S1P increased NO production from human umbilical vein endothelial cells by enhancing Ca²⁺-sensitive eNOS activity without significant increase in the eNOS protein. Two studies have addressed the cardioprotective role of NO during cardiac surgery by supplementation of blood cardioplegia with l-arginine (42) and SPM-5185, a cysteine-containing NO donor (35). These agents, respectively, produced both reduced infarct size and improved postischemic contractile performance.

Cellular subfractionation studies (48) have demonstrated that the cytoplasm of NOS containing neurons is homogeneously stained followed either immunohistochemistry or NADPH diaphorase histochemistry. In contrast, NOS-containing vascular endothelial cells exhibit a characteristic pattern of staining in which the reaction product is restricted to a small number of punctate regions within each endothelial cell (36, 37). Other studies (29, 40) have shown that this reaction product is present throughout the cytoplasm and in association with the membranes of vesicles, mitochondria, and endoplasmic reticulum. Our study has shown that neonatal cardiomyocytes show an identical pattern of staining following fluorescence using the NO sensitive dye DAF-FM, and thus the punctate staining might represent patches of NOS activity. Assuming that these small vesicles localized by NO fluorescent dye represent NOS activity, our results suggest that active NOS enzyme might be largely associated with the Golgi apparatus and with cytoplasmic vesicles, which thus accounts for the punctate appearance of neonatal cardiomyocytes following fluorescence using the NO sensitive dye DAF-FM. Neonatal cardiomyocyte NOS might probably be the result of a mature molecule in the Golgi apparatus and is incorporated into vesicles derived from the Golgi. Hecker et al. (10) have shown that the concentration of eNOS is highest in the cell fractions associated with the endoplasmic reticulum and the plasma membranes of endothelial cells. Whether neonatal cardiomyocytes patches represent a reserve pool of NOS, which may be required when the tissue is under a local stress, remains unclear. However, the subcellular punctate patches of NOS activity, as seen in our study to be localized to specific area toward cell periphery, may be the most opportune site for a readily available “pool” of NO.

Regulation of NOS activities by protein phosphorylation is complex. They are regulated by phosphorylation at Ser/Thr as well as tyrosine residues. Phosphorylation is both stimulatory and inhibitory depending on the site of phosphorylation (23). For example, phosphorylation at Thr495 and Ser1177 of eNOS may have opposite effects on the enzymatic activities (23). Phosphorylation of nNOS at threonine 1296 inhibits NO production (46), and activation of PKB activity by cAMP induces eNOS activation (53). On the other hand, phosphorylation of eNOS by Akt at Ser1177 increases the NO production (33). Regulation of NOS by other protein phosphatase, such as PP2A is not quite clear. Therefore, the exact signaling pathways from Pak1 to eNOS require further investigation.

In conclusion, the present study provides the first direct evidence of the behavior of plasma sphingolipids following transient cardiac ischemia with dramatic and early increases in LCSB in humans. We also demonstrated that S1P, SPH, and FTY720 have protective effects against hypoxic/ischemic cell injury, likely a Pak1/Akt1 signaling cascade and NO release. A further study on a mouse model of cardiac-specific deletion of Pak1 demonstrates a crucial role of Pak1 in cardiac protection against I/R injury.

GRANTS

The project was supported by the Wellcome Trust (to M. Lei); British Heart Foundation (to M. Lei, E. J. Cartwright, and S. E. Stringer); National Heart, Lung, and Blood Institute Grants RO1-HL-64035 and PO1-HL-62426 (Project 1; to R. J. Solaro); and Medical Research Council (to M. Lei).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).