Apolipoprotein A-1 mimetic D-4F enhances isoflurane-induced eNOS signaling and cardioprotection during acute hyperglycemia

Abstract

Acute hyperglycemia (AHG) decreases the availability of nitric oxide (NO) and impairs anesthetic preconditioning (APC)-elicited protection against myocardial infarction. We investigated whether D-4F, an apolipoprotein A-1 mimetic, rescues the myocardium by promoting APC-induced endothelial NO signaling during AHG. Myocardial infarct size was measured in mice in the absence or presence of APC [isoflurane (1.4%)] with or without AHG [dextrose (2 g/kg ip)] and D-4F (0.12 or 0.6 mg/kg ip). NO production, superoxide generation, protein compartmentalization, and posttranslational endothelial NO synthase (eNOS) modifications were assessed in human coronary artery endothelial cells cultured in 5.5 or 20 mM glucose with or without isoflurane (0.5 mM) in the presence or absence of D-4F (0.5 μg/ml). Myocardial infarct size was significantly decreased by APC (36 ± 3% of risk area) compared with control (54 ± 3%) in the absence, but not presence, of AHG (49 ± 4%). D-4F restored the cardioprotective effect of APC during AHG (36 ± 3% and 30 ± 3%, 0.12 and 0.6 mg/kg, respectively), although D-4F alone had no effect on infarct size (53 ± 3%). Isoflurane promoted caveolin-1 and eNOS compartmentalization within endothelial cell caveolae and eNOS dimerization, concomitant with increased NO production (411 ± 28 vs. 68 ± 10 pmol/mg protein in control). These actions were attenuated by AHG (NO production: 264 ± 18 pmol/mg protein). D-4F reduced superoxide generation and enhanced caveolin-1 and eNOS caveolar compartmentalization and posttranslational eNOS modifications, thus restoring NO production during isoflurane and AHG (418 ± 36 pmol/mg protein). In conclusion, D-4F restored the cardioprotective effect of APC during AHG, possibly by decreasing superoxide generation, which promoted isoflurane-induced eNOS signaling and NO biosynthesis.

acute hyperglycemia (AHG) alone, in the presence or absence of overt diabetes, is a major predictor of cardiovascular morbidity and mortality (18, 21, 23, 30), but the mechanisms that contribute to increased risk are poorly understood. Recent evidence strongly implicates perioperative hyperglycemia, in the absence of diabetes, as an independent predictor of death after noncardiac surgery (15). AHG increases the production of ROS (7, 34), decreases the availability of nitric oxide (NO) (20), impairs endothelial function (31), and attenuates coronary microcirculatory responses to myocardial ischemia (25). Evidence from our laboratory also indicates that AHG markedly attenuates cardioprotective signal transduction produced by pharmacological agents such as volatile anesthetics (1, 24). The American College of Cardiology/American Heart Association 2007 guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery recommend the use of volatile anesthetic agents in patients at risk for myocardial ischemia due to their cardioprotective effects (13). As the beneficial effects of these agents may be impaired by AHG (1, 53), additional strategies to address the adverse consequences of perioperative AHG are required.

Apolipoprotein (Apo)A-1 is the major component of high-density lipoprotein, and ApoA-1 mimetics that scavenge oxidized lipids and modulate cholesterol transport to membrane microdomains (52) have been suggested to possess properties that may decrease cardiovascular risk during diabetes and AHG (19). Membrane (lipid) rafts containing cholesterol serve an important function to organize signaling molecules, such as endothelial NO synthase (eNOS), into temporally and spatially regulated caveolar microdomains that facilitate signal transduction during anesthetic cardioprotection (22, 29). We hypothesized that AHG adversely modulates eNOS function during cardioprotective signaling and that, conversely, the ApoA-I mimetic D-4F enhances isoflurane-induced intracellular compartmentalization and activation of eNOS within caveolae, thereby increasing NO and restoring the cardioprotective effects of anesthetic preconditioning (APC).

MATERIALS AND METHODS

D-4F.

D-4F (Ac-DWFKAFYDKVAEKFKEAF-NH₂) was synthesized using all d-amino acids as previously described (32, 36); it was dissolved in PBS before administration.

Animals.

Male C57BL/6J mice (weight: 26.9 ± 2.0 g, age: 9–12 wk) were purchased from Jackson Laboratory (Bar Harbor, ME). Animals were housed in an environmentally controlled room and maintained on a 12:12-h light-dark cycle. All experimental procedures used in this study were approved by the Animal Care and Use Committee of the Medical College of Wisconsin (Milwaukee, WI) and conformed with the “Guiding Principles in the Care and Use of Laboratory Animals” of the American Physiological Society and were in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.

Myocardial ischemia-reperfusion injury in vivo.

We have previously described the in vivo murine model of myocardial ischemia-reperfusion injury (16, 17). Briefly, mice were anesthetized by an intraperitoneal injection of pentobarbital sodium with an initial dose of 80 mg/kg and ventilated using a positive-pressure mouse ventilator (Minivent type 845, Hugo-Sachs Eletronik-Harvard Apparatus, March, Germany). Additional doses of pentobarbital (15 mg/kg) were given during the experiment (1 h after the initial dose) as needed to maintain anesthesia (movement to noxious stimulation). Body temperature was maintained between 36.8 and 37.5°C. Myocardial ischemia was produced by occluding the left coronary artery, and reperfusion was initiated by loosening the suture. At the conclusion of each experiment, mice were killed with intraperitonal pentobarbital (80 mg/kg) and intravenous 50 mM KCl (0.1 ml). The infarct area was delineated by perfusing the coronary arteries with 2,3,5-triphenyltetrazolium chloride (Sigma-Aldrich, St. Louis, MO) via the aortic root, and the area at risk (AAR) was determined by perfusing phthalo blue dye (Heath Acrylics, Effingham, IL) into the aortic root after ligation of the coronary artery at the site of previous occlusion. As a result of these procedures, the nonischemic portion of the left ventricle was stained dark blue. Viable myocardium within the AAR was stained bright red, and infarcted tissue was light yellow.

Experimental protocol.

The concentration-dependent effects of D-4F (0.12 or 0.6 mg/kg ip) or vehicle (control) to restore APC [1.0 minimum alveolar concentration of isoflurane (1.40%) (51) administered for 30 min followed by 15 min of washout] in the presence or absence of hyperglycemia [dextrose (2 g/kg ip)] were determined in C57BL/6J mice randomly assigned to six separate experimental groups. In two additional groups, the contribution of enhanced NO production to D-4F protection during hyperglycemia was evaluated in mice treated with the NO synthase inhibitor N^G-nitro-l-arginine methyl ester (l-NAME; 1 mg/kg ip). All mice were subjected to 30 min of coronary occlusion followed by 2 h of reperfusion (Fig. 1A). Blood glucose levels were monitored throughout the experiment, and heart rate was monitored by electrocardiogram. AAR for infarction and myocardial infarct size were determined after each experiment.

Fig. 1.

Experimental protocols used to determine myocardial infarct size in C57BL/6J mice (A); O₂^·− formation, endothelial nitric oxide (NO) synthase (eNOS) dimerization, and eNOS, phosphorylated (p-)eNOS, and caveolin (Cav)-1 compartmentalization in human coronary artery endothelial cells (HCAECs; B), and NO production in HCAECs (C). AHG, acute hyperglycemia; APC, anesthetic preconditioning; BFA, brefeldin A; CON, control; D-4F, Ac-DWFKAFYDKVAEKFKEAF-NH₂ (apolipoprotein A-1 mimetic); l-NAME, N^G-nitro-l-arginine methyl ester (NO synthase inhibitor); OCC, occlusion.

Human coronary artery endothelial cells.

Human coronary artery endothelial cells (HCAECs; Cell Applications, San Diego, CA) isolated from healthy coronary arteries were purchased and cultured in growth medium (Cells Applications) as previously described (1). To assess the effects isoflurane and high glucose concentrations on HCAECs, 24 h before experimentation, growth medium was replaced with normal 5.5 mM media (containing 5.5 mM d-glucose, 14.5 mM mannitol, 81 mM NaCl, 4.0 mM KCl, and 1.6 mM CaCl₂; pH 7.4) or high 20.0 mM glucose media (containing 20.0 mM d-glucose, 81 mM NaCl, 4.0 mM KCl, and 1.6 mM CaCl₂; pH 7.4) media. HCAECs were also pretreated with 0.5 μg/ml D-4F for 1 h before 1 minimum alveolar concentration isoflurane exposure for 30 min using continuous air flow as previously described (1) to determine if D4-F mitigated the adverse effects of high glucose concentrations on HCAECs. Additional experiments were completed using 4 μg/ml brefeldin A (B5936, Sigma-Aldrich) for 1 h before D-4F to inhibit protein trafficking. Brefeldin A has previously been shown to impair cholesterol and caveolin (Cav) transport to caveolae (45) (Fig. 1B).

Ozone chemiluminescence.

To determine the effects of isoflurane on NO production with and without high glucose, nitrite concentration, an index of NO production, was measured in cell culture media by ozone-based chemiluminescence as previously described (56). Samples (30 μl) were refluxed in reaction solution (50 mg KI in 1 ml of double-distilled water) mixed with glacial acetic acid (4 ml). Nitrite concentrations were calculated after the subtraction of background levels and normalized to protein concentration (Bradford method). As NO production in HCAECs has previously been shown to peak at 60 min after stimulation (8), NO measurements were performed after 60 min of exposure to isoflurane. Additional experiments were completed using 1 mM l-arginine (A8094, Sigma-Aldrich) for 1 h before isoflurane exposure to stimulate NO synthase activity that could be substrate limited, particularly during high glucose. Since gas flow can induce shear-stress-dependent NO release, control groups were exposed to air and CO₂ alone at the same flow rate (Fig. 1C).

O₂^·− analysis in living HCAECs.

To evaluate the effects of high glucose on ROS production during isoflurane, the formation of O₂^·− was assessed in living HCAECs with dihydroethidium (10 μM, Invitrogen, Carlsbad, CA), which, upon oxidation, yields red fluorescent ethidium. Dihydroethidium was added to cell media 5 min before isoflurane. Controls using antimycine A (10 μM, Sigma-Aldrich) were positive, and those using DMSO Hybri-Max (D2650, Sigma-Aldrich) alone were negative. Quantification of ethidium fluorescence intensity was performed at the end of the experiment using a laser-scanning confocal microscope (Nikon Eclipse TE2000-U, Tokyo, Japan) with a ×60/1.4 oil-immersion objective (Nikon) (46). Ethidium fluorescence (excitation wavelength/emission wavelength = 543/570–610) was acquired, and the data were analyzed using ImageJ software (NIH, Bethesda, MD).

Western blot analysis.

Cells were lysed with 250 μl buffer (20 mM MOPS, 5 mM EDTA, 2 mM EGTA, and 0.5% Nonidet P-40) and protease and phosphatase inhibitor cocktails as previously described (1). Protein concentration was determined using the Bradford method, and equivalent amounts of protein (25 μg) were resolved by SDS-PAGE. Proteins were transferred to a polyvinylidene difluoride (PVDF) membrane, blocked, and exposed overnight to primary antibodies against Cav-1 (1:1,000, ab17052, Abcam, Cambridge, MA), eNOS (1:5,000, sc-654, Santa Cruz Biotechnology, Santa Cruz, CA), and phosphorylated Ser¹¹⁷⁷ eNOS (p-eNOS; 1:1,000, Cell Signaling Technology, Danvers, MA). Blots were washed and incubated with secondary antibodies (Cav-1: 1:2,000, goat anti-mouse, Bio-Rad, Hercules, CA, and eNOS and p-eNOS: 1:10,000, donkey anti-rabbit, GE Healthcare). To investigate eNOS homodimer formation in endothelial cells as an index of coupled eNOS activity during isoflurane, low-temperature SDS-PAGE was performed with nonreducing Laemmli buffer. Equivalent amounts of protein (15 ug) were loaded, transferred to a PVDF membrane, blocked, and then probed with eNOS antibody (1:500, Santa Cruz). After being stripped, blots were reprobed with CD-31 antibody (1: 2,000, ab28364, Abcam). The secondary antibody was donkey anti-rabbit (eNOS: 1:2,000 and CD-31:1:8,000). Immunoreactive bands were detected by enhanced chemiluminescence followed by densitometric analysis using ImageJ software (NIH). eNOS dimer and monomer band densities were normalized to CD-31, and the change in the dimer-to-monomer ratio during isoflurane was expressed relative to control conditions in the absence of isoflurane.

Discontinuous sucrose gradient.

To investigate the actions of isoflurane with or without high glucose to modulate the intracellular localization of eNOS-related proteins, buoyant fractions, membrane fractions enriched in caveolae and lipid rafts, were isolated on a discontinuous sucrose gradient as previously described (35). HCAECs were washed, scraped into 2 ml of 500 mM Na₂CO₃, and homogenized on ice. Cell homogenates (600 μg protein/sample) were fractionated by sucrose density ultracentrifugation for 18 h. Fractions (1 ml) were collected, and an equal volume of each fraction was used for eNOS, p-eNOS, and Cav-1 Western blot analyses. All fractions were probed for the specific endoplasmic reticulum marker calreticulin (1:500, ab 4, Abcam).

Statistics.

Data are expressed as means ± SE. Comparison of two means was performed using Student's t-test. Comparison of several means was performed using one-way (one factor tested) or two-way (two factors tested) ANOVA, when appropriate, and the Newman-Keuls post hoc test. Hemodynamic data were analyzed with repeated-measures ANOVA. Changes within and between the groups were considered statistically significant at P < 0.05 (two-tailed). Statistical analysis was performed using NCSS 2007 software (Statistical Solutions, Cork, Ireland).

RESULTS

D-4F restored the cardioprotective effect of APC in mice with AHG.

Ninety-two mice were instrumented to obtain seventy successful experiments. Twenty-two mice died during the experiment and were excluded from analysis [control: 2 mice, APC: 2 mice, D-4F: 2 mice, AHG + APC: 2 mice, AHG: 5 mice, D-4F (0.12 mg/kg) + AHG + APC: 3 mice, D-4F (0.6 mg/kg) + AHG + APC: 2 mice, l-NAME: 2 mice, and D-4F + AHG + APC + l-NAME: 2 mice]. Heart rate at baseline and after occlusion and reperfusion was similar among groups (Table 1), and blood glucose levels were not affected by APC or D-4F throughout the experiment (Fig. 2). There were no differences in AARs for infarction among groups (Table 2). Myocardial infarct size (Fig. 3) was significantly (P < 0.05) decreased by APC (36 ± 3% of AAR) compared with control experiments (54 ± 3% of AAR), and this action was blocked by AHG (49 ± 4% of AAR). Pretreatment of mice with D-4F at 0.12 and 0.6 mg/kg restored APC during AHG (36 ± 3% and 30 ± 3% of AAR, respectively), although D-4F alone had no effect on infarct size (53 ± 3% of AAR). l-NAME alone had no effect on infarct size (63 ± 5%, n = 7) but partially attenuated the beneficial effects of D-4F during APC and AHG (44 ± 3% of AAR, n = 7) compared with D-4F + APC + AHG in the absence of NO synthase inhibition (30 ± 3% of AAR).

Table 1.

Heart rates in mice during in vivo experiments

			Reperfusion
Experimental Group	Baseline	Occlusion	30 min	60 min	120 min
Control	407 ± 15	420 ± 11	418 ± 10	416 ± 12	404 ± 11
AHG	401 ± 14	410 ± 12	411 ± 12	412 ± 12	419 ± 13
APC	413 ± 17	383 ± 27	424 ± 19	417 ± 14	431 ± 15
D-4F	402 ± 8	387 ± 11	419 ± 11	404 ± 10	402 ± 14
AHG + APC	434 ± 22	407 ± 10	415 ± 12	424 ± 13	438 ± 13
D-4F (0.12 mg/kg) + AHG + APC	420 ± 12	412 ± 15	405 ± 19	433 ± 10	417 ± 13
D-4F (0.6 mg/kg) + AHG + APC	429 ± 15	408 ± 18	398 ± 14	389 ± 11	399 ± 10
l-NAME	388 ± 12	386 ± 9	398 ± 13	394 ± 12	392 ± 12
D-4F (0.6 mg/kg) + l-NAME + AHG + APC	422 ± 16	420 ± 18	427 ± 18	414 ± 15	418 ± 16

Data are means ± SE (in beats/min). AHG, acute hyperglycemia; APC, anesthetic preconditioning; D-4F, Ac-DWFKAFYDKVAEKFKEAF-NH₂ (apolipoprotein A-I mimetic); l-NAME, N^G-nitro-l-arginine methyl ester (nitric oxide synthase inhibitor).

Fig. 2.

Blood glucose measurements in C57BL/6J mice with or without AHG [dextrose (2 g/kg ip)] in the presence or absence of the apolipoprotein A-1 mimetic D-4F (0.6 mg/kg ip) during APC. Data are means ± SE; n = 6 mice/group. *P < 0.05 compared with the CON group.

Table 2.

Areas at risk for infarction

Experimental Group	Area at Risk, % of the Left Ventricle
Control	49 ± 11
AHG	50 ± 12
APC	50 ± 9
D-4F	51 ± 12
AHG + APC	53 ± 12
D-4F (0.12 mg/kg) + AHG + APC	52 ± 11
D-4F (0.6 mg/kg) + AHG + APC	51 ± 10
l-NAME	53 ± 3
D-4F (0.6 mg/kg) + l-NAME + AHG + APC	52 ± 3

Data are means ± SE.

Fig. 3.

Myocardial infarct size shown as a percentage of the area at risk for infarction in C57BL/6J mice. Circles are individual infarct size measurements (with means ± SE). *P < 0.05 compared with the CON group; †P < 0.05 compared with the AHG + APC group.

D-4F enhanced isoflurane-stimulated NO production during high glucose.

Isoflurane significantly (P < 0.05) increased NO production in HCAECs (Fig. 4A) cultured under normal glucose conditions (411 ± 28 pmol/mg protein, n = 3). This was attenuated by high glucose (264 ± 18 pmol/mg protein, n = 3) and restored when HCAECs were pretreated with D-4F (418 ± 36 pmol/mg protein, n = 5). l-Arginine enhanced NO production in unstimulated HCAECs (n = 8; Fig. 4B) but not in those stimulated by isoflurane (n = 6). High glucose prevented l-arginine-induced increases in NO with (n = 7) or without (n = 9) isoflurane. Interestingly, during high glucose, combined treatment of cells with D-4F and isoflurane (n = 9) substantially enhanced the ability of eNOS to use the substrate and increased NO production, whereas this did not occur with D-4F alone (n = 6).

Fig. 4.

A: NO production in HCAECs cultured in normal (5.5 mM) media or high glucose (20 mM) media. Data are means ± SE; n = 3–5 experiments/group. *P < 0.05 compared with the respective CON group without isoflurane (ISO); †P < 0.05 compared with 20 mM glucose + ISO. B: NO production in HCAECs after the addition of l-arginine to the media. Data are expressed as mean percent changes from the respective CON group without l-arginine (means ± SE; n = 6–9 experiments/group). *P < 0.05 compared with the respective CON group.

D-4F attenuated O₂^·− formation during high glucose.

High glucose significantly (P < 0.05) increased O₂^·− formation (435 ± 36 arbitrary units, n = 6; Fig. 5) compared with normal glucose conditions (306 ± 25 arbitrary units, n = 7). Isoflurane had no effect on O₂^·− formation during high glucose (417 ± 38 arbitrary units, n = 6). However, D-4F decreased O₂^·− formation to levels observed during normal glucose in either the absence (295 ± 10 arbitrary units, n = 6) or presence (308 ± 26 arbitrary units, n = 6) of isoflurane.

Fig. 5.

O₂^·− formation measured with dihydroethidium, which, upon oxidation, yields red fluorescent ethidium in living HCAECs cultured in normal (5.5 mM) or high glucose (20 mM) media. Scale bar = 10 μm. Confocal images of ethidium fluorescence are shown for the following groups: 5.5 mM glucose CON (A), 20 mM glucose (B), 20 mM glucose + D-4F (C), 20 mM glucose + ISO (D), and 20 mM glucose + D-4F + ISO (E). F: quantification of ethidium fluorescence in HCAECs. Data are means ± SE; n = 6 experiments/group. *P < 0.05 compared with the 5.5 mM CON group; †P < 0.05 compared with the 20 mM glucose group.

D-4F enhanced isoflurane-stimulated eNOS homodimerization during high glucose.

Isoflurane significantly (61 ± 10%, n = 5, P < 0.05) increased eNOS dimerization (Fig. 6) during normal glucose conditions compared with unstimulated cells (n = 4), whereas high glucose blocked this effect (−38 ± 11%, n = 5). D-4F enhanced the ability of isoflurane to increase the eNOS protein dimer-to-monomer ratio (123 ± 27%, n = 5).

Fig. 6.

eNOS protein dimer-to-monomer ratio in HCAECs. A: representative Western blot showing eNOS dimers (280 kDa), eNOS monomers (140 kDa), and CD-31 (140 kDa) protein expression in HCAECs. B: histogram showing the eNOS protein dimer-to-monomer ratio normalized to CD-31 and expressed as the percent change from the respective CON group without ISO. Data are means ± SE; n = 5 experiments/group. *P < 0.05 compared with the respective CON group without ISO.

D-4F promoted isoflurane-induced Cav-1, eNOS, and p-eNOS caveolar compartmentalization.

During unstimulated, normal glucose conditions, Cav-1 was located in heavy/noncaveolar fractions, eNOS was equally distributed between caveolar and noncaveolar compartments, and Ser¹¹⁷⁷ phosphorylation occurred predominantly in buoyant/caveolar fractions (fractions 4 and 5; n = 3 per condition; Fig. 7). Isoflurane significantly (P < 0.05) promoted the redistribution of eNOS and Cav-1 from noncaveolar to caveolar fractions and increased Ser¹¹⁷⁷ phosphorylation of eNOS exclusively in caveolar fractions. In contrast, high glucose enhanced the retention of eNOS, p-eNOS, and Cav-1 in noncaveolar fractions. Isoflurane or D-4F alone had no effect on protein redistribution toward caveolae during high glucose. However, combined D-4F and isoflurane treatment significantly (P < 0.05) increased the redistribution of Cav-1, eNOS, and p-eNOS to buoyant fractions, with the latter being blocked by brefeldin A, a pharmacological inhibitor of protein trafficking at a dose that had no effect on cell viability measured with trypan blue (data not shown).

Fig. 7.

Subcellular localization of eNOS, p-eNOS, and Cav-1 in HCAECs evaluated by discontinuous sucrose gradient fractionation. Fractions 4 and 5 represent buoyant caveolar microdomains, whereas fractions 9–12 represent heavy membrane fractions. A: representative sucrose density gradient separations by Western blot analysis. B: quantification of the ratio of buoyant to total protein expression in HCAECs. Data are means ± SE; n = 3 experiments with 5 dishes/group. *P < 0.05 compared with the 5.5 mM glucose CON group; †P < 0.05 compared with the 20 mM glucose group; ‡P < 0.05, 20 mM glucose + D-4F + ISO + BFA group compared with 20 mM + D-4F + ISO group.

DISCUSSION

AHG and diabetes have been demonstrated to abrogate the cardioprotective effects of APC (1, 24, 44, 53); however, the responsible mechanisms have been incompletely elucidated. Hyperglycemia is well known to impair the bioavailability of NO, an important trigger and mediator of APC (2, 9). We (1) have previously shown that hyperglycemia adversely modulates NO signaling during pharmacological protection with volatile anesthetics, such as isoflurane, by attenuating heat shock protein 90 interactions with eNOS and by decreasing tetrahydrobiopterin concentrations. Recently, the contribution of lipid rafts to intracellular signaling has emerged as a critical component of pharmacological cardioprotection (38, 54); however, the role of high glucose to adversely impact NO production through interactions with lipid rafts has not been investigated. The results of this study indicate that volatile anesthetics modulate intracellular compartmentalization and posttranslational modifications of eNOS in lipid rafts, that this process is sensitive to glucose concentration, and that the adverse effects of AHG were ameliorated by an ApoA-1 mimetic.

Lipid/membrane rafts are sterol-, sphingolipid-, and cholesterol-enriched microdomains present in the plasma membrane, endoplasmic reticulum, and mitochondria that allow for the dynamic, temporal regulation of protein trafficking and control of intracellular signal transduction events (43). Caveolae are a subclass of membrane rafts distinguished by the presence of the scaffolding proteins Cav-1, Cav-2, and Cav-3 (10). Cav-1 and Cav-2 are differentially expressed in adipocytes, endothelial cells, and fibroblasts, whereas Cav-3 is expressed predominantly in skeletal, cardiac, and smooth muscle cells. The number of caveolae and functional signal transduction events that occur in caveolae are indirectly regulated by Cav expression. The fluidity of membrane rafts is modulated by Cavs through the binding of cholesterol, which, in turn, alters membrane composition and signaling effects (4). Although high glucose has been shown to impair the organization of lipid rafts in pancreatic β-cells (50), the effects of high glucose to disrupt the compartmentalization of proteins involved in APC signal transduction have not been investigated.

The present findings confirm and extend recent evidence showing that lipid rafts and Cavs play an important role in cardioprotection. Cav-1^−/− mice have been demonstrated to be resistant to the cardioprotective effects of isoflurane (38). Isoflurane increased the formation of caveolae in wild-type adult mouse cardiomyocytes and enhanced the localization of Cav-1 to buoyant membrane fractions in the myocardium in vivo (38). Similarly, isoflurane induced the compartmentalization of Cav-1 and eNOS within endothelial cell caveolae in the present investigation, with concurrent increased Ser¹¹⁷⁷ phosphorylation (2) and production of NO. Cav-1 has been shown to regulate eNOS activity through protein-protein interaction events that are localized in specific membrane domains (11). Intracellular compartmentalization of eNOS was thus correlated with eNOS activity, and eNOS activation was sevenfold greater in the plasma membrane fraction compared with the cytosolic fraction. In addition, NO synthase activity was ninefold greater in caveolar versus noncaveolar membranes (47). Paradoxically, basal activity of eNOS was negatively regulated through direct interactions with Cav-1. However, eNOS localization within the caveolae was essential for eNOS activation in response to physiological stimuli (11). During pathophysiological conditions, such as increased oxidative stress, oxidation of cholesterol in caveolae induced the translocation of Cav-1 away from the plasma membrane to the Golgi apparatus (49). Increased levels of oxidized low-density lipoprotein (LDL) that caused the depletion of caveolar cholesterol also promoted Cav-1 and eNOS redistribution to the cytosolic compartment with subsequent eNOS inactivation (4, 42). Diabetes has also been shown to downregulate membrane-associated Cav-1 expression in the renal cortex concomitantly with a decrease in p-eNOS (27). Our results similarly demonstrated that high glucose impaired isoflurane-induced Cav-1 compartmentalization within caveolae and Ser¹¹⁷⁷ phosphorylation of eNOS. Taken together, the previous and present findings implicate AHG as a source of oxidative stress that impairs eNOS signaling by disrupting normal caveolar compartmentalization of proteins.

D-4F is an 18-amino acid peptide composed of d-amino acids that contains four phenylalanine residues (4F) and has no sequence homology with ApoA-1 but shares the lipid-associating structural ApoA-1 motif (3). ApoA-1 mimetics have been shown to possess antiatherogenic and anti-inflammatory effects and to produce favorable vascular effects in animal models of type I diabetes (28, 40), type II diabetes and obesity (39, 41), atherosclerosis (32), and hypercholesterolemia (36). The 4F peptides are proposed to exert anti-inflammatory effects by binding and sequestering proinflammatory oxidized phospholipids and fatty acid hydroperoxides (3), thereby reducing LDL oxidation (55). ApoA-1 or its mimetic have also been demonstrated to modulate cellular cholesterol metabolism (12, 48, 55), thus rendering plausible the assumption that D-4F facilitates eNOS compartmentalization. A single oral dose of D-4F was safe, well tolerated, and improved an in vitro inflammatory index in patients with coronary heart disease (5). The efficacy of ApoA-1 mimetics to improve outcomes in patients with coronary artery disease, AHG, or diabetes, however, has never been tested.

The present results demonstrated that a dose of D-4F, which by itself was not cardioprotective, restored APC during AHG. D-4F normalized isoflurane-stimulated NO production during high glucose, which was accompanied by enhanced translocation of Cav-1 and eNOS to the caveolar membrane fraction. High glucose increased the formation of O₂^·− in endothelial cells, although the intracellular source of ROS was not identified. High glucose has been demonstrated to increase ROS generation from multiple enzymatic sources, including mitochondria (40). Nevertheless, under conditions of oxidative stress and tetrahydrobiopterin deficiency, eNOS fails to produce NO and can itself generate O₂^·−, a condition referred to as eNOS “uncoupling” (14). Uncoupling of eNOS has been linked to monomerization of the protein, a process exacerbated by high glucose and diabetes (6, 57). To our knowledge, the present results are the first to show that isoflurane increases eNOS coupling and that high glucose abrogates isoflurane-stimulated dimerization of eNOS and the production of NO even under conditions of excess substrate availability. Interestingly, D-4F ameliorated O₂^·− formation during high glucose with or without isoflurane and enhanced eNOS dimerization and NO production in the presence of isoflurane, consistent with an action of D-4F to promote coupled eNOS activity. This is consistent with previous evidence showing that the ApoA-1 mimetic L-4F enhanced NO and decreased O₂^·− generation in endothelial cells treated with LDL (37). Finally, the beneficial actions of D-4F were not entirely abolished by l-NAME in vivo; thus, D4-F may have additional effects on lipid oxidation and cholesterol metabolism that could contribute to cardioprotection. This hypothesis requires further investigation to confirm, however.

Our findings should be interpreted within the constraints of several potential limitations. Heart rates were similar during coronary artery occlusion in mice with or without prior exposure to isoflurane or D-4F. Thus, it is unlikely that differences in hemodynamics among the groups contributed to the observed results. However, myocardial O₂ consumption was not measured in the present investigation. In vitro experiments were performed to elucidate the detailed mechanisms contributing to isoflurane-stimulated NO regulation in endothelial cells during high glucose. To recapitulate the in vivo setting as precisely as possible, the time course of drug and glucose administration during these experiments was different. For example, the pharmacokinetics of D-4F (time to peak concentration: 1.5–2 h after application) (33) and the expected in vivo duration of the peak drug effect after intraperitoneal injection required a more prolonged application of D-4F in vivo than in vitro. AHG in mice was produced by an intraperitioneal injection of dextrose compared with the intravenous administration of dextrose previously performed in a large animal species (26). Canine, rabbit, and rodent species are acutely sensitive to short-term exposure to high glucose in vivo; however, preliminary experiments indicated that isoflurane-stimulated NO production in HCAECs was insensitive to short-term (<24 h) culture in high glucose media. The latter observation is consistent with other reports in the literature investigating the effects of high glucose on cells in culture. Our cellular model of NO-related signaling mechanisms during APC has consistently replicated findings in vivo using models of AHG and diabetes (1, 9).

In conclusion, these results demonstrate that AHG impairs isoflurane-stimulated eNOS activity and subsequent cardioprotection by interfering with intracellular protein compartmentalization, phosphorylation, and homodimerization and that this action is mitigated by the ApoA-1 mimetic D-4F. AHG portends a worse clinical outcome in the perioperative period in patients with or without diabetes. Thus, further studies are needed to define strategies for improving eNOS function during AHG as a potential means for decreasing cardiovascular morbidity and mortality.

GRANTS

This work was supported by National Institutes of Health Grants R01-HL-063705 (to J. R. Kersten) and P01-GM-066730 (to J. R. Kersten).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

Author contributions: I.B., Z.-D.G., F.S., D.C.W., and J.R.K. conception and design of research; I.B., Z.-D.G., F.S., A.C., and D.W. performed experiments; I.B., Z.-D.G., F.S., A.C., D.W., and J.R.K. analyzed data; I.B., Z.-D.G., F.S., D.W., and J.R.K. interpreted results of experiments; I.B. and J.R.K. drafted manuscript; I.B., Z.-D.G., F.S., A.C., D.W., D.C.W., and J.R.K. approved final version of manuscript; Z.-D.G., D.W., D.C.W., and J.R.K. edited and revised manuscript; F.S. prepared figures.