Caveolae are small, flask-like invaginations of the plasma membrane. Caveolins are structural proteins found in caveolae that have scaffolding properties to allow organization of signaling. The authors tested the hypothesis that delayed cardiac protection induced by volatile anesthetics is caveolae or caveolin dependent.
An in vivo mouse model of ischemia-reperfusion injury with delayed anesthetic preconditioning (APC) was tested in wild-type, caveolin-1 knockout, and caveolin-3 knockout mice. Mice were exposed to 30 min of oxygen or isoflurane and allowed to recover for 24 h. After 24 h recovery, mice underwent 30-min coronary artery occlusion followed by 2 h of reperfusion at which time infarct size was determined. Biochemical assays were also performed in excised hearts.
Infarct size as a percent of the area at risk was reduced by isoflurane in wild-type (24.0 +/- 8.8% vs. 45.1 +/- 10.1%) and caveolin-1 knockout mice (27.2 +/- 12.5%). Caveolin-3 knockout mice did not show delayed APC (41.5 +/- 5.0%). Microscopically distinct caveolae were observed in wild-type and caveolin-1 knockout mice but not in caveolin-3 knockout mice. Delayed APC increased the amount of caveolin-3 protein but not caveolin-1 protein in discontinuous sucrose-gradient buoyant fractions. In addition, glucose transporter-4 was increased in buoyant fractions, and caveolin-3/glucose transporter-4 colocalization was observed in wild-type and caveolin-1 knockout mice after APC.
These results show that delayed APC involves translocation of caveolin-3 and glucose transporter-4 to caveolae, resulting in delayed protection in the myocardium.
What We Already Know about This Topic
Anesthetic exposure can precondition the heart to protect it from subsequent ischemia
The role of specialized invaginations in the cell membrane, calveolae, and the structural proteins found in them, calveolins, on delayed anesthetic preconditioning is unknown
What This Article Tells Us That Is New
In mice, myocardial infarct size after coronary artery occlusion was decreased by delayed preconditioning with isoflurane
We show that delayed preconditioning was dependent on calveolin-3 and glucose transporter-4 using genetically altered mice
EXPOSURE to volatile anesthetics before a lethal ischemic insult can protect the heart.1–4This protection, termed anesthetic preconditioning (APC), has been described to be a biphasic event. Immediately after exposure to the volatile anesthetic, an early cardiac protection is observed (acute APC), which is transient and subsides after a few hours.1,2,4Protection resumes 12–24 h after the initial stimulus (delayed APC).3Acute APC involves the translocation and phosphorylation of preexisting proteins, whereas delayed APC is dependent on de novo protein synthesis.3,5–8Acute and delayed APC involve complex signal transduction cascades.9
Signal transduction molecules are organized by scaffolding molecules into molecular complexes.10Caveolae are small membrane invaginations on the plasma membrane that are enriched in glycosphingolipids, cholesterol, and caveolins.11,12Three isoforms of caveolin, Cav-1, -2, and -3, are involved in the formation of caveolae and interact with signaling molecules via a scaffolding domain.13–17All three caveolin isoforms are found in cardiac myocytes (CMs), with Cav-3 being the predominant isoform.18We have shown that both Cav-118and Cav-319are essential for acute APC-induced cardiac protection. We have also shown that acute ischemic preconditioning increases the formation of caveolae and that transgenic mice with CM-specific overexpression of Cav-3 are resistant to ischemia–reperfusion injury independent of a preconditioning stimulus.20Thus, there seems to be a clear role for caveolins or caveolae in the regulation of acute cardiac protection from ischemia–reperfusion injury.
The initial triggering events associated with acute and delayed cardiac protection are similar. Delayed protection induces gene and protein expression changes that ultimately lead to the induction of various mediators (e.g. , inducible nitric oxide synthase, 12-lipoxygenase, cyclooxygenase-2, and glucose transporter-4 [GLUT-4]).9,21It is unclear how organization of mediators of delayed cardiac protection is regulated and whether caveolins or caveolae play a role. Because caveolae serve as a nexus for regulating acute protective signaling, we hypothesized that caveolin or caveolae would be essential for delayed APC. In this article, we show that delayed APC is dependent on a specific caveolin isoform, Cav-3, and that the coordination of signaling is dependent on colocalization of Cav-3 and GLUT-4.
Materials and Methods
Antibodies
Antibodies used in this study are listed with their sources as follows: polyclonal antibody to Cav-1 (Abcam, Cambridge, MA and Cell Signaling, Danvers, MA), monoclonal and polyclonal antibody to Cav-3 (BD Biosciences, San Jose, CA; Abcam; and Santa Cruz Biotechnology, Santa Cruz, CA); monoclonal antibody to GLUT-4 (Abcam); and monoclonal antibody to glutaraldehyde phosphate dehydrogenase (Imgenex, San Diego, CA).
Animals
All animals were treated in compliance with the Guide for the Care and Use of Laboratory Animals (National Academy of Science, Washington, D.C.). Animal use protocols were approved by the Veterans Administration San Diego Healthcare System Institutional Animal Care and Use Committee (San Diego, California). C57BL/6 male and Cav-1 knockout mice were purchased from Jackson Laboratories (Bar Harbor, ME), and Cav-3 knockout mice were a kind gift from Yoshi Ishikawa, M.D., Ph.D. (Professor, Cardiovascular Research Institute, Yokohama City University School of Medicine, Yokohama, Japan), and Yasuko Hagiwara, Ph.D. (Professor, National Institute of Neuroscience, Kodaira, Tokyo, Japan), created as previously reported (8- to 10-week old, 21–26 g body weight, male).22Animals were daily randomly assigned into treatment groups by an independent observer. The animals were kept on a 12-h light–dark cycle in a temperature-controlled room. Mice were placed postoperatively in an animal care unit under daily supervision.
Immunoblot Analysis
Protein was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis 10% polyacrylamide precast gels (Invitrogen, Carlsbad, CA) and transferred to a polyvinylidene difluoride membrane by electroelution. Membranes were blocked in Tris-buffered saline and 0.1% Tween containing 2.0% nonfat dry milk and incubated with primary antibody overnight at 4°C. Bound primary antibodies were visualized using secondary antibodies conjugated with horseradish peroxidase from Santa Cruz Biotechnology and Lumigen TMA-6 chemiluminescent reagent from GE Healthcare (Piscataway, NJ). All displayed bands migrated at the appropriate size, as determined by comparison with molecular weight standards (Santa Cruz Biotechnology).
Electron Microscopy
Whole hearts were fixed with 2.5% glutaraldehyde in 0.1 m cacodylate buffer for 2 h, postfixed in 1% OsO4in 0.1 m cacodylate buffer (1 h), and embedded as monolayers in LX-112 (Ladd Research, Williston, VT). Sections were stained in uranyl acetate and lead citrate and observed with an electron microscope (Philips CM-10, Philips Electronic, New York, NY). Random sections were taken by an electron microscopy technician blinded to the treatments.
Experimental Preparation
Under light anesthesia (pentobarbital sodium, 40 mg/kg, intraperitoneal), mice were randomly divided into groups and received 30 min of 100% oxygen in the control group or 1.4% isoflurane vol./vol. in O2(1.0 minimum alveolar concentration for mice)23by using a pressure-controlled ventilator (TOPO Ventilator, Kent Scientific, Torrington, CT; peak inspiratory pressure: 15 cm H2O, respiratory rate: 100 breaths/min) followed by a 24-h recovery period. Core temperature was maintained with a heating pad and lamp, and electrocardiogram leads were placed to record heart rate.
Ischemia–Reperfusion Protocol
After a 24-h recovery period, mice were anesthetized with pentobarbital sodium (80 mg/kg intraperitoneal) and mechanically ventilated. Hemodynamics were measured through the right carotid artery with a 1.4F micro-tip pressure transducer (Model SPR-671; Millar Instruments, Inc., Houston, TX) as described earlier.24After thoracotomy, baseline was established, and mice were assigned to one of six experimental protocols as described in figure 1. Mice underwent 30-min index ischemia followed by 2 h of reperfusion. After reperfusion, mice were heparinized, and the coronary artery was again occluded. The heart was immediately excised and cut into 1.0-mm slices. Each slice of left ventricle was then counterstained with 2,3,5,-triphenyltetrazolium chloride (Sigma Chemical, St. Louis, MO). After overnight storage in 10% formaldehyde, slices were weighed and visualized under a microscope equipped with a charge-coupled device camera. The images were analyzed (Image-Pro Plus version 4.5; Media Cybernetics, Silver Spring, MD), and infarct size was determined by planimetry as previously described.25
Sucrose Density Membrane Fractionation
Mice were exposed to oxygen or isoflurane as described earlier and allowed to recover for 24 h. Mice were anesthetized with pentobarbital sodium (80 mg/kg intraperitoneal), and hearts were excised. We used whole left ventricle to prepare sucrose density membrane fractions as reported previously.18We defined fractions 4–6 as buoyant membrane fractions enriched in caveolae and proteins associated with caveolae. Fractions 9–12 were defined as nonbuoyant fractions.
Immunofluorescence
Ventricular tissue was mounted on a cryostat (−23°C), and 10-μm sections were cut in the long axis. Samples were fixed with paraformaldehyde, incubated with 100 mm glycine, permeabilized in 0.1% buffered Triton X-100, and blocked with 1% bovine serum albumin, phosphate-buffered saline, and 0.05% Tween. Samples were then incubated with primary antibody (1:100) in 1% bovine serum albumin, phosphate-buffered saline, and 0.05% Tween for 24 h. Excess antibody was removed, and samples were incubated with fluorescein Alexa-conjugated secondary antibodies (1:250) for 1 h. To remove excess secondary antibody, samples were washed with phosphate-buffered saline or 0.1% Tween and incubated for 20 min with the nuclear stain 4′,6-diamidino-2-phenylindole (1:5000) diluted in phosphate-buffered saline. Samples were mounted in gelvatol for microscopy imaging, and images were captured with DeltaVision deconvolution microscope system (Applied Precision, Inc., Issaquah, WA). The system includes a photometrics charge-coupled device mounted on a Nikon TE-200 inverted epifluorescence microscope (Nikon USA, Melville, NY). Three optical sections spaced 0.22 μm were taken. Exposure times were set such that the camera response was in the linear range for each fluorophore. Images were taken at 400 × magnification and were deconvolved and analyzed using SoftWorx software (Applied Precision, Inc.) on a Silicon Graphics Octane workstation (SGI, Fremont, CA). Colocalization of pixels was assessed quantitatively by CoLocalizer Pro 1.0 software (CoLocalization Research Software, Kochi, Japan). All images were normalized to a background threshold value of 50.
Adult Rat Cardiac Myocyte Isolation and Treatment
CMs were isolated via retrograde-perfused Langendorff enzymatic digestion as previously described.26Cells were plated on 12-well plates. CMs of rat were exposed to 1.4% isoflurane for 30 min in a temperature-controlled metabolic chamber. Chamber inflow was attached to the outflow of an isoflurane vaporizer. Chamber outflow was monitored with a Datex Capnomac capnograph (Datex, Helsinki, Finland). Isoflurane was infused with oxygen at 2 l/min flow. We have previously confirmed that 1.4% vol./vol. isoflurane produces 0.165 ± 0.003 mm isoflurane in media in our chamber.18After a 24-h recovery period, the CMs were exposed to ischemic stress. Ischemia was simulated by replacing the air content with a 95% N2and 5% CO2gas mixture at 2 l/min in a metabolic chamber and replacing the media with glucose-free Dulbecco's Modified Eagle's Medium (pH 6.2) for 60 min. This was then followed by 60 min of reperfusion by replacing the media with normal maintenance media and by incubating the cells with 21% O2and 5% CO2. Cell death was quantified by counting trypan blue-stained cells, with results expressed as a percentage of total cells counted. To determine the effect of intact caveolae on delayed APC-induced cardiac protection, we used methyl-β-cyclodextrin (MβCD, 1 mm, 1 h), which depletes membrane cholesterol, resulting in disruption of caveolae.
Statistical Analysis
Differences in hemodynamic data between groups were compared using a repeated-measures two-way ANOVA with post hoc Bonferroni analysis (GraphPad Software, San Diego, CA). All other statistical analyses were performed by one-way ANOVA followed by Bonferroni post hoc test or unpaired Student t test (two tailed). All data are expressed as mean ± SD. Statistical significance was defined as P < 0.05.
Results
Caveolin and Caveolae in Cardiac Myocytes
We investigated the expression of caveolin-1 and caveolin-3 proteins in heart tissue. Immunoblots revealed expression of both caveolin-1 and caveolin-3 in the wild-type (WT) mouse hearts and the absence of caveolin-1 or caveolin-3 proteins in caveolin-1 knockout or caveolin-3 knockout mice, respectively (fig. 2A). Electron microscopy revealed caveolae formation in WT and caveolin-1 knockout mice; however, no caveolae were observed in caveolin-3 knockout mice (fig. 2B).
Myocardial Area at Risk and Infarct Size
Hemodynamics (heart rate and mean arterial pressure) of mouse after carotid artery cannulation are shown in table 1. No significant differences in heart rate or mean arterial pressure were found between groups at the preocclusion time point.
The area at risk, as a percent of the left ventricle, was similar among all groups (fig. 3A). Twenty-four hours after isoflurane (1.0 minimum alveolar concentration) exposure, a reduction in myocardial infarction was observed compared with WT control. In caveolin-3 knockout mice, the protection produced by isoflurane was eliminated (fig. 3B), but APC-induced cardiac protection was maintained in caveolin-1 knockout mice (fig. 3B).
Isoflurane Modulates Caveolin Localization
We assessed the effect of delayed APC on cardiac caveolin localization in WT mice. Hearts from control and APC-treated animals (24-h post-APC or oxygen) were fractionated on a discontinuous sucrose gradient and analyzed for distribution of caveolin. APC increased the amount of caveolin-3 protein but not caveolin-1 into buoyant fractions (fig. 4, A and B).
Colocalization between Glucose Transporter-4 and Caveolins
Hearts subjected to sucrose density fractionation were probed for localization of various mediators in caveolae after APC. The whole heart expression of inducible nitric oxide synthase, 12-lipoxygenase, and cyclooxygenase-2 has previously been reported to increase with delayed protective stimuli7,27–29; the localization of these proteins was not increased in caveolar fractions of any group (data not shown). However, higher expression of GLUT-4 was observed in buoyant fractions of WT APC and caveolin-1 knockout APC but not in caveolin-3 knockout APC hearts (fig. 5A). To investigate the interaction of caveolins and GLUT-4, immunofluorescence microscopy was performed. At 24-h postisoflurane exposure (1.0 minimum alveolar concentration, 30 min), there was increased colocalization of GLUT-4 or caveolin-3 in left ventricle tissues (yellow pixels, fig. 5B) in WT APC and caveolin-1 knockout APC mice. The colocalization was confirmed by immunoprecipitation with and without isoflurane treatment. Lysates were immunoprecipitated with caveolin-3 and then probed with GLUT-4 antibody. We observed increased association of caveolin-3 and GLUT-4 after isoflurane treatment (fig. 5C). The reciprocal immunoprecipitation immunoblot was not performed because we were unable to obtain a GLUT-4 antibody suitable for immunoprecipitation.
In Vitro Assessment of Role of Caveolae in Delayed Anesthetic Preconditioning
Isolated adult rat CMs were subjected to simulated ischemia–reperfusion with and without disruption of caveolae using MβCD at different times as outlined in figure 6A. Cells treated with isoflurane and allowed to recover 24 h showed significant protection when subjected to simulated ischemia–reperfusion (fig. 6B). If caveolae were disrupted with MβCD before isoflurane treatment, protection was still observed after 24-h recovery; however, if caveolae were disrupted with MβCD after 24-h recovery but just before simulated ischemia–reperfusion, the protection afforded by isoflurane was lost (fig. 6B). To further confirm disruption of caveolae, sucrose density fractionation was used to evaluate buoyant caveolin-3, an indicator of caveolae. As in the in vivo data presented in figure 4B, isoflurane treatment resulted in increased caveolin-3 localization in buoyant fractions. Treatment of CM with MβCD before isoflurane or air had no effect on buoyant caveolin-3; however, treatment of CM with MβCD after 24-h recovery resulted in a decline in buoyant caveolin-3 indicative of loss in caveolae (fig. 6C). These data suggest that caveolae are not necessary to trigger protection but are necessary to organize downstream signaling associated with delayed APC.
Discussion
In this study, we observed that delayed APC induced by isoflurane cannot be elicited in caveolin-3-deficient mice in vivo , indicating that the presence of caveolae (dependent on caveolin-3 expression) is a prerequisite for delayed protection in the myocardium. We further clarified the role of caveolae in in vitro studies that suggest that caveolae are not necessary for the triggering of delayed APC but may be critical to organize downstream mediators. This is the first study to investigate the role of caveolins or caveolae in delayed APC. This study also showed that delayed APC involves translocation of caveolin-3 but not caveolin-1 into caveolae (which was also confirmed in an in vitro model of delayed APC), and this seems to be associated with specific upregulation and colocalization of GLUT-4 with caveolin-3.
Volatile anesthetics have a long history in the clinical management of anesthesia; however, recent evidence suggests a role in cardiac protection. Many studies have shown that volatile anesthetics exert biphasic cardiac protection1–3and that characteristics of this protection are similar to those observed during classic ischemic preconditioning.30Volatile anesthetics are short-chain halogenated alkanes and ethers that interact with cell membrane lipids and directly or indirectly interact with membrane-bound proteins to produce a number of cellular effects. Because caveolae are highly enriched in lipids and signaling molecules, they might serve as an ideal “docking” site for volatile agents to modulate cellular physiology. To this end, we have previously shown that acute treatment of isolated CMs with isoflurane results in increased caveolar invaginations of the sarcolemmal membrane.18This article extends this finding to delayed protection by showing that delayed APC causes translocation of caveolin-3 and GLUT-4 to buoyant fractions and that delayed APC is dependent on caveolae or caveolin-3. Previous reports have shown that lung tissue treated with sevoflurane did not show enhanced phosphorylation of caveolin-1, and no data were presented regarding modification of caveolae.31Therefore, it is likely that different anesthetics have varied effects on caveolins and caveolae. Unpublished preliminary data from our group suggest that sevoflurane may have a similar effect to isoflurane with respect to caveolae formation, but further work is necessary to validate cross-class effects of volatile anesthetics on caveolins and caveolae and downstream signaling. Volatile anesthetics may preferentially interact with and modulate lipid microdomains, such as caveolae, and the proteins localized within these microdomains to regulate function. This concept may represent a novel hypothesis for a mechanism of anesthetic action that may apply to multiple organs.
We have previously shown that isoflurane-induced acute cardiac protection is abolished in caveolin-1 knockout mice despite the fact that caveolae are present in CMs from caveolin-1 knockout mice.18Our laboratory has also shown that caveolin-3 colocalizes with opioid receptors in vitro , which can contribute to cardiac protection from ischemia.26More recently, we have reported that caveolin-3 knockout mice lose the ability to undergo isoflurane-induced acute cardiac protection from ischemia–reperfusion injury in both in vivo and in vitro models, although caveolin-1 is present at normal levels.19Collectively, these data implicate caveolae and caveolins as essential to the temporal and spatial organization of acute cardiac protective signaling molecules. No previous studies have described the role of caveolae and caveolins in cardiac protection induced by delayed APC. Interestingly, we show with delayed protection that only caveolin-3 is necessary because delayed APC was absent in caveolin-3 knockout mice but robust in caveolin-1 knockout mice. These data suggest that both caveolin-3 and caveolin-1 may be important in the regulation of signaling events involved in acute cardiac protection (e.g. , Src, C-terminal Src kinase, Akt, and glycogen synthase kinase 3β),18,20,32whereas caveolin-3 and localization of components to caveolae may be critical to induction and organization of mediators involved in delayed protection.
Caveolae are cholesterol and sphingolipid-enriched invaginations of the plasma membrane 50–100 nm in size33and play a role in important physiologic functions such as signal transduction,13–17endocytosis,34calcium homeostasis,35and intracellular cholesterol transport.36Proteomic studies have suggested that caveolae contain as many as 150 distinct signaling molecules37–40; however, it has also been suggested that not all caveolae are created equal with distinct subpopulations of caveolae containing specific molecules to control specific signaling pathways.41,42It is possible that in response to a protective stimulus, certain mediators are activated and organized into caveolae to form a molecular complex that facilitates delayed protection. Certain proteins (i.e. , inducible nitric oxide synthase, cyclooxygenase-2, and 12-lipoxygenase) have been shown by our group and others to be critical in mediating delayed cardiac protection in response to various stimuli in which expression and activity of these proteins are up-regulated in response to a stimulus and inhibition with specific pharmacologic agents abrogates protection.7,27–29We did not observe any changes in the translocation of inducible nitric oxide synthase, cyclooxygenase-2, or 12-lipoxygenase to buoyant caveolar fractions. Such data suggest that caveolae may compartment a different subset of mediators involved in delayed protection.
We found that the caveolin-3-dependent delayed cardiac protective effect was accompanied by GLUT-4 translocation to caveolae after 24-h isoflurane exposure. Others have shown that delayed ischemic preconditioning up-regulates GLUT-4 expression by activation of adenosine monophosphate-activated protein kinase in a protein kinase C-dependent manner21and that acute ischemic preconditioning-induced GLUT-4 translocation to the plasma membrane involves caveolin-3, Akt, and endothelial nitric oxide synthase.43Therefore, GLUT-4 seems to be an important downstream mediator of delayed APC that is dependent on caveolin or caveolae for its localization.
Our data put GLUT-4 temporally and spatially downstream of other mediators that regulate delayed protection. If this is the case, one would expect GLUT-4 to be regulated by classic mediators of delayed protection. In this regard, nitric oxide donors have been shown to induce messenger RNA for GLUT-4 and translocation of GLUT-4 to the plasma membrane.44,45Treatment of adipocytes with arachidonic acid (a precursor to cyclooxygenase-2 and 12-lipoxygenase metabolites) causes translocation of both GLUT-1 and GLUT-4 to the plasma membrane.46Specifically, it has been shown that inhibition of cyclooxygenase can reduce the expression of GLUT-4 messenger RNA.47Treatment of ventricular CMs with 12-lipoxygenase inhibitors leads to altered GLUT-4 translocation from intracellular stores to plasma membrane, and this was linked to the disassembly of the actin cytoskeleton.48It has previously been shown that caveolae and subsequent cardiac protection can be altered by cytoskeletal disruption agents.19,49,50Such findings suggest that GLUT-4 may be a more distal mediator of delayed protection than inducible nitric oxide synthase, cyclooxygenase-2, and 12-lipoxygenase, and translocation of GLUT-4 to caveolae may be a critical element of delayed APC. Our in vitro data further indicate that caveolae disruption does not alter triggering of delayed APC; however, disruption of caveolae just before simulated ischemia–reperfusion is sufficient to attenuate delayed APC. This suggests that caveolae are potentially important organizers of mediators induced by delayed APC.
Our findings should be interpreted within the constraints of potential limitations. We measured only heart rate and mean arterial pressure as hemodynamic parameters after isoflurane exposure. However, we showed previously that isoflurane exposure in a similar mouse protocol does not significantly alter hemodynamics or blood gases.18,25In addition, 24 h after isoflurane exposure, our hemodynamic data showed no differences between groups at the preocclusion time point. Therefore, the reductions in cardiac infarct size produced by APC most likely were not a result of changes in hemodynamic determinants. Caveolin-3 knockout mice have a variety of deleterious phenotypes (i.e. , muscle degeneration,22insulin resistance,51,52and progressive cardiomyopathy with age53) that may affect outcome after ischemia–reperfusion injury. A recent study shows that the hearts of caveolin-3 knockout mice have normal substrate metabolism and glucose uptake.54These contrasting data suggest that caveolin-3 knockout mice have a complex phenotype that may be organ specific. Our in vitro data suggests that caveolae are important to mediating delayed APC but not the triggering. The use of WT cells with pharmacologic disruption of caveolae get around the limitations of using caveolin-3 knockout mice, and the complementary data suggest that caveolae are regulating delayed APC.
In summary, these results show that isoflurane-induced delayed preconditioning involves translocation of caveolin-3 and GLUT-4 to caveolae and that the presence of microscopically distinct caveolae (dependent on caveolin-3 expression) is a requisite for delayed protection in the myocardium and specifically caveolae are necessary to mediate but not necessarily trigger delayed APC. Modulation of caveolin or caveolae may be a novel target for therapies to protect the heart from ischemia–reperfusion injury.