| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Cardiovascular Research Center, Department of Anatomy and Cell Biology, Temple University School of Medicine, Philadelphia, Pennsylvania, USA; and
Center for Cardiovascular Research, Albany Medical College, Albany, New York, USA
1Correspondence: Cardiovascular Research Center and Department of Anatomy and Cell Biology, Temple University School of Medicine, 3420 North Broad St., MRB826, Philadelphia, PA 19140, USA. E-mail: rizzov{at}temple.edu
ABSTRACT
Reactive oxygen species (ROS) generated during pathological events, such as inflammation and ischemia-reperfusion, activates both proapoptotic and antiapoptotic signaling programs in endothelial cells. Because cholesterol-rich, plasma membrane rafts serve as platforms for organizing and integrating signaling transduction processes, we asked whether these membrane regions play a mechanistic role in H2O2-induced responses. Bovine aortic endothelial cell cultures exposed to a 500-µM bolus of H2O2 showed progressive activation of caspase 3 and an increase in the number of TUNEL-positive cells. Pretreatment with either wortmannin or PD 098059 heightened these apoptotic responses, demonstrating that both PI3 kinase/Akt and ERK1/2 serve as signaling mediators to alleviate H2O2 cytotoxic effects. To investigate the role of lipid rafts in these signaling processes, endothelial cells were pretreated with methyl-ß-cyclodextrin (CD) or filipin to ablate raft structures. H2O2-induced phosphorylation of Akt and ERK 1/2 was attenuated, while caspase 3 and the number of TUNEL positive cells was enhanced in CD-pretreated cells exposed to H2O2. Reconstitution of raft domains restored H2O2-induced Akt and ERK1/2 phosphorylation, which was concomitant with reduction of caspase 3 activation and DNA fragmentation. Taken together, our findings suggest that plasma membrane compartments rich in cholesterol participate in signal transduction pathways activated by oxidative stress.—Yang, B., Oo, T. N., and Rizzo, V. Lipid rafts mediate H2O2 prosurvival effects in cultured endothelial cells.
Key Words: caveolae • apoptosis • reactive oxygen species • endothelium
PATHOLOGICAL EVENTS such as inflammation, ischemia/reperfusion injury, and atherosclerosis are accompanied by production of high quantities of reactive oxygen species (ROS) within the vasculature. Depending on the source of ROS generation and the concentration and duration of exposure to oxidants, endothelial cell responses to ROS vary from proliferation and survival to dysfunction and cell death. The molecularsignaling pathways that govern these responses are just beginning to emerge and includes activation of Akt (1
, 2)
, eNOS (2)
, Src-like kinases (3
4
5)
, growth factor receptors (4
, 6)
, and MAP kinases (1
, 4
, 7
8
9)
.
The concept that signal transduction occurs in spatially discrete subcompartments within the cell is becoming firmly established in the literature. Cholesterol- and sphingolipid-rich regions within the plasma membrane, termed lipid rafts, comprise one such signaling domain. The presence of caveolin proteins in these regions induces formation of discrete invaginated structures termed caveolae. Several laboratories have reported a variety of signaling molecules concentrated within these structures (10)
. Many of these signaling molecules have been implicated in endothelial cell responses to oxidative stress. For instance, Src-kinases, eNOS, and Ras are activated in response to ROS. Each of these molecules is enriched with rafts and caveolae and copurifies with caveolin-1 (10)
. In addition, ROS have been shown to cause the transactivation of epidermal growth factor receptor (EGFr) (4
, 6)
, which is partially localized to rafts/caveolae (11)
. The activation of Ras/Raf/ERK1/2 pathways in response to H2O2 has been shown to involve activation of Fyn and JAK2 (3)
. Both of these proximal signaling molecules are reported to be enriched in caveolin-containing raft microdomains (12
, 13)
.
Collectively, these studies implicate a role for cholesterol-rich rafts in oxidative stress-induced signaling. Surprisingly, few studies have been conducted to directly evaluate the functional significance of this cell compartment in ROS signaling transduction in endothelial cells. Here, we provide evidence that plasma membrane compartments rich in cholesterol participate in signal transduction pathways activated by oxidative stress.
MATERIALS AND METHODS
Antibodies
The following primary antibodies were purchased from Cell Signaling: anti-Akt (polyclonal antibody (pAb)), antiphospho-Akt (pAb; Ser-473, anti-ERK1/2 (pAb), anti-ERK2 (pAb), antiphospho-p38 (pAb), anti-p38 (pAb), anti-phospho-JNK1/2 (pAb), anti-JNK1 (pAb), anticleaved caspase 3 (pAb). Anticaveolin-1 (pAb) was from BD Transduction Labs (Franklin Lakes, NJ). Horseradish peroxidase-conjugated anti-rabbit secondary antibodies were from Amersham-Pharmacia Biotech (Piscataway, NJ).
Endothelial cell cultures
Bovine aortic endothelial cells (BAEC) were purchased from Cell Applications (San Diego, CA.). Cells were grown in MCDB-131 culture medium (Sigma) supplemented with 10% FBS (Atlanta Biologicals, Norcross, GA) and 0.04 mg/ml gentamicin sulfate (Cambrex Bio Science Walkersville, MD), and maintained at 37°C, 97% humidity, and 5% CO2.
Hydrogen peroxide (H2O2) treatments
All cultures were incubated in serum-free medium for 2 h before H2O2 challenge. BAEC were washed twice with fresh serum-free medium. H2O2 (Sigma, St. Louis, MO) was dissolved in fresh serum-free medium at a final concentration ranging from 100 µM to 1 mM and immediately applied to the cell cultures for indicated time at 37°C. Cell viability was determined by trypan blue exclusion assay. To verify the specificity of H2O2 effects, cell cultures were incubated in serum-free medium containing either catalase (1000 U/ml) alone or a mixture of catalase/H2O2 prepared just before addition to the cells. After treatments, cells were rinsed twice in PBS (pH 7.4) and processed as a whole cell lysate.
Preparation of lipid rafts
Caveolin-1-enriched lipid rafts were subfractionated from bulk plasma membranes, as described previously (14)
. Briefly, 6 x 106 BAEC were gently scraped into Tricine buffer (20 mM Tricine, 1 mM EDTA, 250 mM sucrose, pH 7.4), and centrifuged at 1400 g, 5 min, 4°C. Cell pellets were resuspended in 1 ml Tricine buffer and homogenized with a dounce homogenizer. Homogenates were then centrifuged at 1400 g for 10 min at 4°C. The supernatant was collected, mixed with 30% Percoll (Sigma), and centrifuged in a Beckman MLS50 rotor (OPTIMA MAX, Beckman Coulter, Fullerton, CA) at 77,000 g for 25 min. Plasma membranes were collected, sonicated (3x30 s bursts on ice), and mixed with 1.2 ml 60% sucrose. This mixture was placed in the bottom of 5-ml tube, layered with 1.3 ml 35% and 5% sucrose and finally centrifuged at 87,400 g, overnight at 4°C in a Beckman MLS50 rotor. Plasma membrane fractions were collected in 0.4 ml fractions from top to bottom of the tube and added to 0.25 vol TCA-DOC [100% (w/v) TCA in ddH2O, 0.1%(w/v) Deoxycholic acid] to precipitate proteins.
Lipid raft disassembly
In order to evaluate the role of cholesterol-rich microdomains in ROS signaling, the effects of two different cholesterol-binding agents were tested independently. Methyl-ß-cyclodextrin (CD) extracts cholesterol from the plasma membrane, and filipin, a statin that binds cholesterol, perturbs the clustering of cholesterol molecules within the membrane. Both compounds are known disruptors of lipid raft structure. Endothelial cells were incubated in serum-free media containing 10 mM CD for 30 min at 37°C before exposure to H2O2. In a separate set of experiments, cholesterol was added back to cholesterol-depleted cell cultures in order to reconstitute disassembled rafts. Briefly, repletion with cholesterol was accomplished by incubating cells in the presence of a cholesterol/methyl-ß-cyclodextrin mixture for 1 h at at 37°C. A stock solution of 0.4 mg/ml cholesterol and 10% CD was prepared by vortexing at 40°C in 10 ml of 10% CD with 200 µl of cholesterol (20 mg/ml in EtOH solution) (15)
.
Filipin is a macrolide polyene antibiotic, a class of drugs that binds sterols, such as cholesterol and can cause reversible disassembly of lipid rafts (16
, 17)
. As a complementary set of experiments to the CD studies, cells were pretreated with 5 µg/ml filipin for 5 min at 37°C before H2O2 exposure.
AnnexinV staining
To detect early stages of apoptosis, H2O2 challenged endothelial cell cultures were probed with Annexin V-FITC (BD Pharmingen; kit #556547). Briefly, BAEC were grown to a confluent monolayer on 0.2% gelatin-coated coverslips. Following H2O2 treatment, cells were washed with 1X Annexin V Binding Buffer [10mM HEPES/NaOH (pH 7.0), 140 mM NaCl, 2.5 mM CaCl2], incubated with Annexin V-FITC (5 µl) for 15 min at room temperature, rinsed with Annexin V Binding Buffer and fixed in 2% formaldehyde (Electron Microscopy Sciences, Hatfield, PA). Cells were subsequently viewed under fluorescence microscopy (Nikon Eclipse TE300). Images were captured with a Nikon digital camera (DXM1200) using ACT-1 software. Relative fluorescence intensity was quantified using Image J-software.
TUNEL assay
Peroxide-induced apoptosis was detected by performing terminal deoxynucleotidedyl transferase-mediated dUTP nick end-labeling (TUNEL) using the DeadEndTM Fluorometric TUNEL System kit (Promega). Briefly, BAEC were grown on gelatin-coated microscope slides until confluent. Following pretreatments and exposure to H2O2, cells were fixed, permeabilized, and labeled for DNA strand breaks. Cells were immediately analyzed under x40 with a fluorescence microscope (Nikon Eclipse TE 300). For each experiment, images from at least four fields were captured with a Nikon digital camera (DXM1200) using ACT-1 software. The number of labeled cells per field was quantified, and data from 3 independent experiments were pooled to determine a mean number of apoptotic cells.
Western analysis
Cell lysates were analyzed for protein content using bicinchoninic acid methodology. Equivalent amounts of protein (50 µg) from cell lysates were prepared in sample buffer (170 mM Tris, pH 6.8, 3% [w/v] SDS, 1.2% [v/v] ß-mercaptoethanol, 2 M urea, and 3 mM EDTA). For protein precipitated from sucrose gradient fractions, the entire protein yield was reconstituted in the same sample buffer. Samples were separated by SDS-PAGE (10% gel) followed by electrotransfer to nitrocellulose filters for immunoblotting. Filters were incubated with indicated primary antibody (Ab) followed by anti-rabbit horseradish peroxidase conjugated secondary Ab. Proteins were detected using enhanced chemiluminescence (ECL) substrate (ECL, Amersham). Densitometry quantification of immunoblots with Image J software allowed direct comparisons between experimental sets.
Statistical analysis
Data gathered from at least 3 independent experiments were pooled according to group. Mean and SD were calculated, and differences between groups were analyzed with an unpaired two-tailed Student’s t test using STATGRAPHICS 4.0 software (Statistical Graphics Corp., Herndon, VA). Differences between control and experimental groups were considered to be significant at P < 0.05.
RESULTS
Enhancement of H2O2 induced apoptosis in cholesterol-depleted endothelial cell monolayers
Previous studies demonstrate that prolonged exposure to micromolar concentrations of H2O2 induce endothelial apoptosis in vitro (9
, 18)
and in vivo (19)
. In our experimental system, exposure of BAEC cultures to a mid range dose of H2O2 (500 µM) induced significant loss of endothelial cell viability following 4 h of exposure (Fig 1
A). Preceding peroxide-induced cell death, we found that Annexin V showed initial binding to the endothelial cell plasma membrane following 30 min peroxide exposure (Fig 1B
). By 60 min, Annexin V binding increased markedly (Fig. 1B
). Additionally, TUNEL staining of monolayers following 4 h of H2O2 exposure showed a significant number of cells containing fragmented DNA, confirming H2O2-induced apoptosis (Fig. 1C
). Furthermore, an upstream apoptotic effector, caspase 3, was progressively activated in response to H2O2 treatment, as evidenced by detection of its 17-kDa active fragment (Fig. 2
A).
|
|
Because raft domains are enriched in cholesterol, compounds that interact with and sequester cholesterol provide useful tools for investigating the structural and functional significance of these plasma membrane compartments. To determine whether cholesterol-rich rafts play a role in ROS-mediated events, BAEC were pretreated 10 mM CD for 30 min before H2O2 exposure. Compared to H2O2 treatment alone, the percentage of TUNEL-positive cells nearly doubled in cholesterol-depleted BAEC monolayers after 4 h treatment with H2O2 (Fig. 1C
). Although statistically insignificant from control, CD appeared to directly induce a small degree of apoptosis (avg. of 1.3±0.9 cells per field). Therefore, TUNEL-positive cells in CD and H2O2 groups were added (avg. of 8.3± 1.1 cells per field) and compared to the cells pretreated with CD and subjected to H2O2 (avg. of 15.9±1.8 cells per field). Statistical analysis of the average number of TUNEL-positive cells between these groups revealed that enhancement of apoptosis observed in CD-pretreated cells subsequently exposed to H2O2 was significantly greater than those measured by the sum effect of CD and H2O2 treatments alone. Furthermore, the activated fragment of caspase3 could not detected in CD-pretreated cells after 4 h of replacement of CD media with CD-free media demonstrating that direct CD effects were minimal under the experimental conditions used in this study. In contrast, CD-pretreated cells exposed to H2O2 showed a somewhat earlier (30 min) and significantly greater level of cleaved caspase 3 signal at 60, 120, and 240 min compared to H2O2 treatment alone (Fig. 2A
). Moreover, restoration of plasma membrane cholesterol content halted the accelerated activation of caspase 3 observed in cholesterol-depleted cells treated with H2O2 (Fig. 2B
). These data suggest that depletion of plasma membrane cholesterol can enhance endothelial cell susceptibility to lethal effects of oxidative stress.
Akt is activated within lipid rafts/caveolae and contributes to prosurvival signaling
Both proapoptotic and antiapoptotic signaling pathways become activated within the endothelium in response to H2O2 (20)
. Protein kinase B/Akt has previously been shown to play an important prosurvival role in endothelial cells exposed to ROS (6
, 18)
. Consistent with these reports, pharmacological inhibition of the PI3kinase/Akt pathway with wortmannin showed substantially greater levels of H2O2-induced caspase 3 activation and DNA fragmentation (Figs. 3
A and B).
|
H2O2 induced transient phosphorylation of Akt on Ser-473, which peaked at 15 min post-treatment (Fig. 3C
). Disruption of lipid rafts by pretreatment with CD significantly attenuated this response (Fig. 3C
). The specificity of the cholesterol-depleting effect on Akt activation by H2O2 was verified by replenishing plasma membrane cholesterol content to reconstitute raft domains in CD-pretreated BAEC. Figure 3D
illustrates that cholesterol repletion served to fully restore H2O2-induced activation of Akt. In a complementary set of experiments, endothelial cell cultures were pretreated with an alternative raft-disrupting agent, filipin. Similar to CD, filipin pretreatment significantly reduced peroxide-induced Akt phosphorylation in these cells (Fig. 3E
).
To determine whether H2O2-induced Akt signaling responses were occurring specifically in raft membrane compartments, plasma membranes derived from H2O2-treated and nontreated BAEC monolayers were subfractionated by nondetergent sucrose gradient centrifugation methods. In both nontreated cells and BAEC exposed to H2O2 for 15 min, Akt was present in lipid rafts, as determined by location in light buoyant density plasma membrane fractions containing the raft marker caveolin-1 (Fig. 3F
). However, a significant pool of Akt was also found in the nonraft portions of the plasma membrane (Fig. 3F
). Probing these membrane fractions with phospho-specific antibodies showed that the pool of Akt phosphorylated by H2O2 was localized predominantly in raft compartment of the plasma membrane (Fig. 3F
).
To further explore the significance to Akt compartmentation to rafts, BAEC were pretreated with CD before H2O2 treatment. Figure 3F
(right) shows that depletion of plasma membrane cholesterol effectively caused lipid raft disruption, as evidenced by a significant shift in caveolin out of light buoyant density membrane fractions into heavier membrane fractions. In addition, raft ablation resulted in a similar redistribution of Akt. More importantly, the phosphorylation of Akt observed after H2O2 treatment remained attenuated, suggesting the raft localization plays an important role in Akt’s responsiveness to oxidative stress.
Erk1/2 prosurvival effects in response to H2O2 are dependent on plasma membrane cholesterol content
H2O2-induced activation of ERK1/2 peaked within 15 min of exposure and returned near baseline after 60 min. Pretreatment with PD098059 to pharmacologically inhibit ERK 1/2 activation significantly accelerated the detection of cleaved caspase 3 (Fig. 4
A) and enhanced the number of TUNEL-positive cells after a 4-h exposure to H2O2 (Fig. 4B
). Similar to the effects of PD098059 (data not shown), CD prevented H2O2-induced activation of ERK 1/2 (Fig. 4C
), which could be recapitulated on repletion of plasma membrane cholesterol (Fig. 4D
). Similar to our observations with Akt, raft disruption with filipin blocked ERK 1/2 phosphorylation by H2O2 (Fig. 4E
). Together, these data indicate that antiapoptotic effects of ERK 1/2 are dependent on the integrity of plasma membrane rafts in endothelial cells subjected to oxidative stress.
|
Effects of cholesterol depletion on H2O2-induced stress-activated MAP kinase activation
Because p38 and c-Jun NH2-terminal kinase (JNK)/SAPK are purported to relay proapoptotic signaling in endothelial cells subjected to oxidative stress, we sought to determine whether rafts/caveolae influence H2O2-induced activation of these signaling molecules. Similar to its effect on ERK 1/2 MAP kinase, H2O2 rapidly activated both p38 and JNK (Fig. 5
). However, CD pretreatment had little effect on p38 phosphorylation at early time points but appeared to sustain a significant concentration of p38 phosphorylation at 60 min post-treatment. The phosphorylation pattern for JNK was also altered in CD-pretreated cells, in which H2O2 induced JNK phosphorylation only after a 15 min of exposure. However, by 60 min, phosphorylation of JNK appeared to reach a plateau in both control and CD-pretreated cells in response to H2O2 (Fig. 5)
.
|
DISCUSSION
Experimental evidence converging from several distinct areas of cell biological research supports the concept that signal transduction events are localized in spatially discrete subcompartments within the cell. One compartment of particular interest is composed of proteins and lipids that are organized in a cohesive structure on the plasma membrane and termed lipid rafts. The lipid composition consists of cholesterol and sphingolipids that together provide the unique physical characteristics of these membrane regions. Caveolae comprise a subset of lipid rafts in that they are also enriched in cholesterol and sphingolipid relative to the membrane proper. However, caveolae contain a family of structural proteins, caveolins, which are responsible, together with cholesterol, for the unique invaginated morphology of caveolae. By mechanisms that are not completely understood, a wide variety of receptors and signaling molecules concentrate within both rafts and caveolae (10
, 21)
. Thus, lipid rafts and caveolae have been hypothesized to function as platforms for efficient signal coordination.
The results from our study provide additional support for this concept through the observation of a functional relationship between cholesterol-rich plasma membrane domains and prosurvival signaling pathways stimulated in response to oxidative stress. It should be noted however, that because cholesterol depletion and/or sequestration disrupts both lipid rafts and caveolae, our study does not distinguish between rafts and caveolae in this signaling process. Thus, further studies will be necessary to determine the precise role that caveolae play in endothelial cell signaling response to exogenous ROS.
Our present findings are consistent with previous studies (9
, 18)
demonstrating that 500 µM H2O2 rapidly induced apoptosis in endothelial cell cultures. We detected rapid and progressive Annexin V binding to the plasma membrane (Fig. 1B
) and DNA fragmentation events via TUNEL analysis (Fig. 1C
). Concomitant with these observations, caspase 3 activation occurred within 1 h of H2O2 exposure and progressively increased during the 4 h experimental period (Fig. 2A
). To address whether lipid rafts play a role in H2O2-induced apoptosis, BAEC monolayers were pretreated with methyl-ß-cyclodextrin (CD) to deplete plasma membrane cholesterol, resulting in disruption of raft structural integrity. In response 500 µM H2O2, CD-pretreated cells showed heightened sensitivity to the cytotoxic effects of peroxide. The number of cells demonstrating apoptotic nuclei in response to H2O2 was significantly greater in monolayers pretreated with CD (Fig. 1C
). Additionally, compared to H2O2 alone, CD-pretreated cells showed activation of caspase 3, as soon as 30 min and was significantly greater at all other time points measured after H2O2 treatment (Fig. 2A
).
These data are in agreement with a previous study in which staurosporin-induced apoptosis was enhanced following CD pretreatment of cardiac endothelial cells (22)
. In a separate study, cholesterol depletion was shown to mitigate, not enhance, TNF-
induced apoptosis (23)
. Considering that the TNF- receptor and associated factors can localize to cholesterol-enriched plasma membrane domains in endothelial cells (23
24
25)
, differing results may be explained by disruption of TNF- signaling complex on cholesterol depletion and loss of ability to propagate death signaling initiated by TNF-. Indeed, our preliminary studies strongly support these findings (our unpublished observations). In cultured endothelial cells, we found that the superoxide-generating enzyme system, NADPH oxidase, is localized and functional in plasma membrane rafts and caveolae. TNF- enhanced ROS production within these membrane compartments concomitant with recruitment of p47phox regulatory subunit to these domains. In addition, TNF- induced activation and phosphorylation of eNOS present in plasma membrane raft/caveolae compartments. The dual activation of superoxide- and NO-generating systems within the same membrane domains provided a spatially favorable environment for formation of peroxynitrite, a known cytotoxic radical. Taken together with our present study, lipid rafts not only appear to be sensitive to external sources of ROS, which trigger prosurvival signaling events, but also play a role in regulating NADPH oxidase and subsequent ROS generation that may mediate a death response in endothelial cells. Therefore, the context in which endothelial cells are exposed to ROS appears to determine the type of signaling responses elicited within lipid rafts and caveolae.
Akt is a crucial prosurvival signaling factor in several cell types exposed to a variety of proapoptotic stimuli (20)
. In endothelial cells exposed to apoptotic stressors, Akt is rapidly and transiently activated, an event that occurs predominantly in a PI3 kinase-dependent manner (1
, 2)
. Consistent with these reports, we found that H2O2 induced phosphorylation of Akt. We also found that wortmannin effectively blocked H2O2-induced phosphorylation of Akt (data not shown) and heightened peroxide-induced apoptotic events (Fig. 3)
, thus verifying Akt’s role as an antiapoptotic mediator. These data are in close agreement with prosurvival signaling functions reported for Akt in endothelial cells undergoing serum withdrawal-induced apoptosis where another PI3 kinase inhibitor, LY294002, attenuated Akt activation and enhanced apoptosis (26)
.
Here, we tested whether lipid raft domains play a role in propagation of H2O2-induced signaling to Akt. We found that both CD and filipin pretreatment significantly attenuated H2O2-induced Akt phosphorylation (Fig. 3)
. Akt was distributed in both light buoyant density membrane fractions containing caveolin-1, as well as in heavier membrane fractions. Interestingly, H2O2 appeared to induce phosphorylation of the Akt molecules specifically associated with raft membranes (Fig. 3)
. Moreover, raft disruption caused a redistribution of Akt into nonraft membrane fractions where the phosphorylation response to H2O2 was blunted (Fig. 3)
. Collectively, these data suggest that H2O2 signaling to Akt occurs primarily within discrete plasma membrane compartments. Our observations that wortmannin attenuates H2O2-induced Akt phosphorylation supports previous observations that PI3 kinase is localized to rafts and caveolae in endothelial cells (27)
. Consistent with our findings (Fig. 3F
, right), raft ablation served to disrupt the organization of signaling molecules (i.e., PI3 kinase) in rafts, which is necessary for efficient propagation of H2O2 signals.
Previous studies have shown that activation of ERK1/2 can confer protection from detrimental effects of ROS (18
, 28)
. Here, we observed that depleting or sequestering cholesterol from plasma membrane raft compartments inhibited H2O2 activation of ERK1/2 (Fig. 4)
. Furthermore, H2O2 -induced ERK1/2 phosphorylation was fully restored on repletion of plasma membrane cholesterol (Fig. 4)
. Although these observations correlated with enhanced apoptosis, we pretreated cells with the MEK inhibitor PD98059 in order to more fully explore the link between ERK1/2 phosphorylation and H2O2-induced apoptosis in our system. Pretreatment with PD 98059 effectively blocked ERK1/2 phosphorylation (data not shown) and served to enhance both H2O2 activation of caspase 3 activity and the number of TUNEL-positive cells (Fig. 4)
, substantiating the prosurvival role of ERK1/2 in H2O2-induced cytotoxicity reported previously (18
, 28)
. In addition, we found that wortmannin and PD98059 did not alter H2O2-induced ERK1/2 or Akt phosphorylation levels, respectively (data not shown). These data suggest that both Akt and ERK1/2 pathways function in parallel to mediate prosurvival signaling pathways in endothelial cells subjected to oxidative stress. More importantly, our data collectively suggest that plasma membrane rafts constitute a common, proximal point of signaling transduction that relays prosurvival messages through Akt and ERK1/2.
The localization of both PI3 kinase (27)
and Akt (Fig. 3)
, and in some cases enrichment, in rafts suggest that raft domains serve as a site for H2O2-induced signaling. Additionally, upstream mediators of PI3 kinase, such as Src-like kinases are also localized in raft/caveolae membranes (27)
. H2O2 can rapidly inactivate phosphatases, resulting in loss of opposition to actions of Src-like kinases. Hence, H2O2 inactivation of phosphatases would result in Src induction of PI3 kinase activity and Akt phosphorylation. Given these observations, it could be predicted then that disruption of raft structures would alter effective propagation of these pathways. Indeed, our data showing that raft ablation through depletion of plasma membrane cholesterol attenuates activation of Akt by H2O2 support this mechanism of H2O2-induced signaling.
Our data also show that rafts mediate H2O2-induced activation of ERK1/2 (Fig. 5)
. A potential mechanism whereby rafts can effect ERK1/2 activation involves transactivation of epidermal growth factor receptor (EGFr) by H2O2 (27)
. Since EGFr can localize to caveolin-containing domains (11)
, it would be expected that alteration of raft integrity would induce loss of receptor compartmentation or disorganization of the signaling machinery necessary to propagate epidermal growth factor (EGF) receptor responses to ERK1/2 following oxidant challenge. To substantiate a mechanistic relationship between rafts and EGF receptors, experiments that directly examine localization and/or transactivation of EGFr in cholesterol-depleted endothelial cell cultures subjected to H2O2 would be required.
Data from several recent studies (9
, 18)
and this report indicate that blockade of either ERK1/2 or Akt heighten the sensitivity of cells to H2O2-induced apoptosis, thus demonstrating the importance of these signaling molecules in compensating for ROS toxic effects. Although our discussions above speculate on likely mechanisms for rafts/caveolae regulation of H2O2-induced activation of both these prosurvival signaling molecules, alternative mechanisms also related to rafts/caveolae may be activated in antiapoptotic responses to oxidative stress. In cardiac endothelial cells, staurosporin-induced apoptosis occurred through the activation of caspases that were specifically localized to lipid rafts (22)
. It was posited that this compartmentation served to concentrate caspase proteolytic activity on local raft-residing substrates to propagate and direct apoptotic signals. By altering plasma membrane localization of caspases with CD, caspases became hyperactivated and showed substrate promiscuity resulting in an enhanced apoptotic response to staurosporin. Whether pro- and active-caspases are localized in vascular endothelial cell rafts and participate in H2O2 induced apoptotic signaling events requires further investigation.
In addition to the proposed mechanisms offered above to explain H2O2 activation of signaling events within raft domains, several additional alternatives are worth consideration. First, ROS can react with cellular lipids to create products that contain functional groups capable of modifying proteins (29)
. For example, electrophilic lipids, formed as the reaction products of oxidation reactions, can directly or indirectly react with nucleophilic cysteine residues found in many signaling molecules. Such modification has been shown to alter the activity of signaling molecules, such as those involved in the MAP kinase pathway (30)
. Second, signaling molecules such as Ras (31
32
33
34)
can be redox activated through ROS reaction with thiol targets present in the signaling molecule. Considering the unique lipid milieu and localization of signaling molecules, such as Ras in rafts and caveolae, these potential mechanisms remain consistent with the experimental evidence provided in this study, demonstrating that lipid rafts serve as signaling platforms for propagation of compensatory survival pathways activated in response to damaging levels of H2O2.
In summary, endothelial cell cultures respond to oxidative stress by activating several signaling pathways. Our observations extend the general concept that compartmentalization of lipid and protein signaling mediators within the plasma membrane provides an efficient means to respond to external stimuli, including oxidants. Moreover, regulation of lipid raft or caveolae expression may be an important mechanism used by endothelial cells to adjust their sensitivity to ROS.
ACKNOWLEDGMENTS
This work was supported by grants from the National Institutes of Health (HL66301-VR) and the American Heart Association (0030300T-VR).
Received for publication November 7, 2005. Accepted for publication February 27, 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  W. Han, H. Li, V. A. M. Villar, A. M. Pascua, M. I. Dajani, X. Wang, A. Natarajan, M. T. Quinn, R. A. Felder, P. A. Jose, et al. Lipid Rafts Keep NADPH Oxidase in the Inactive State in Human Renal Proximal Tubule Cells Hypertension, February 1, 2008; 51(2): 481 - 487. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Soundararajan, A. D. Wishart, H. P. V. Rupasinghe, M. Arcellana-Panlilio, C. M. Nelson, M. Mayne, and G. S. Robertson Quercetin 3-Glucoside Protects Neuroblastoma (SH-SY5Y) Cells in Vitro against Oxidative Damage by Inducing Sterol Regulatory Element-binding Protein-2-mediated Cholesterol Biosynthesis J. Biol. Chem., January 25, 2008; 283(4): 2231 - 2245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kuehnle, A. Crameri, R. E. Kalin, P. Luciani, S. Benvenuti, A. Peri, F. Ratti, M. Rodolfo, L. Kulic, F. L. Heppner, et al. Prosurvival Effect of DHCR24/Seladin-1 in Acute and Chronic Responses to Oxidative Stress Mol. Cell. Biol., January 15, 2008; 28(2): 539 - 550. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Carlile-Klusacek and V. Rizzo Endothelial cytoskeletal reorganization in response to PAR1 stimulation is mediated by membrane rafts but not caveolae Am J Physiol Heart Circ Physiol, July 1, 2007; 293(1): H366 - H375. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
