p190 RhoGTPase-Activating Protein Links the β1 Integrin/Caveolin-1 Mechanosignaling Complex to RhoA and Actin Remodeling
Objective—To determine whether the β1 integrin/caveolin-1 signaling complex plays a role in shear stress regulation of RhoA activity .
Methods and Results—Hemodynamic shear stress influences the phenotype of the endothelium. Integrins and RhoA are essential components in the process that allows endothelial cells to adapt to flow. However, the signaling mechanisms that relay from integrins to RhoA are not well defined. Bovine aortic endothelial cells were subjected to laminar shear stress (10 dyne/cm2) for up to 6 hours. β1 integrin blockade inhibited Src family kinases and p190RhoGAP tyrosine phosphorylation observed after the immediate onset of shear stress. Depletion of caveolin-1 blocked the decline in p190RhoGAP tyrosine phosphorylation observed at later points by sustaining Src family kinase activity. The manipulation of β1 integrin and caveolin-1 also altered shear regulation of RhoA activity. More importantly, cells depleted of p190RhoGAP showed faulty temporal regulation of RhoA activity. Each of these treatments attenuated actin reorganization induced by flow. Similarly, stress fibers failed to form in endothelial cells exposed to enhanced blood flow in caveolin-1 knockout mice.
Conclusion—Our studies demonstrate that p190RhoGAP links integrins and caveolin-1/caveolae to RhoA in a mechanotransduction cascade that participates in endothelial adaptation to flow.
The hemodynamic environment in which an endothelial cell resides strongly influences cell morphological features through regulation of cytoskeletal structures.1 Several, now classic, studies2–5 illustrate that in vitro and in vivo, actin stress fibers orient parallel to the direction of flow and are prominent in endothelial cells subjected to high shear velocities. These actin bundles reflect an adaptive response to shear stress that may aid endothelial cells in withstanding elevated hemodynamic stress, as in the case of hypertension.
Although the influence of flow on regulating endothelial cell phenotype is widely recognized, the fundamental process by which these cells detect and transduce fluid mechanical forces into biochemical signals is not completely clear. Previous studies have described individual cellular elements with potential mechanosignaling properties, including ion channels,6 integrins,7 the glycocalyx,8 cilia,9 platelet endothelial cell adhesion molecule-1 (PECAM-1),10 various receptors for humoral compounds,11 the cytoskeleton,12 and the plasma membrane, including lipid raft and caveolar subdomains.13,14 Interestingly, many of these link to similar sets of second messenger signaling molecules and regulate the same flow responses, such as endothelial NO synthase regulation of NO production and events downstream of extracellular signal–regulated kinase activation. These observations suggest that primary mechanosignaling elements likely engage in a substantial degree of interaction and cross talk. Indeed, recent findings show that integrin/vascular endothelial growth factor receptor 215 and PECAM-1/vascular endothelial–cadherin16 associations form important mechanosignaling complexes.
From our own work, we found that a signaling network composed of β1 integrin and caveolin-1 develops in response to shear stress. Acute shear stress applied to cultured endothelial cells resulted in integrin-dependent phosphorylation of caveolin-1 (pY14) via a Src family kinase (SFK).17 The phosphorylation of caveolin-1 served to recruit C-terminal Src-like kinase (Csk) to the integrin/caveolin-1 complex, resulting in additional regulation of SFK activity and induction of myosin light chain (MLC) phosphorylation. Key components of this pathway were recruited to caveolar microdomains and dependent on the presence of the caveolin-1 protein.18
The small GTPase, RhoA, is a key second messenger in the mechanotransduction pathway that allows endothelial cells to adapt to changes in hemodynamic shear forces. Shear stress applied to cultured endothelial cells temporally regulate RhoA activity, a process that depends on upstream integrin activation.19,20 Moreover, the overexpression of either dominant-negative or constitutively active RhoA prohibits induction of actin stress fibers and morphological restructuring,19 demonstrating that shear modulation of short-term signaling events, such as temporal regulation of RhoA activity, is crucially linked to long-term outcomes of exposing endothelial cells to flow. Although a general relationship between RhoA and caveolin-1–enriched membranes has been described,21 whether caveolae microdomains, either alone or in concert with integrins, influence the mechanotransducing properties of RhoA requires experimental evaluation.
In addition to integrins and caveolin-1/caveolae as proximal regulators of RhoA, guanine nucleotide exchange factors and GTPase-activating proteins (GAPs) directly modulate RhoA activity by stimulating GTP/GDP cycling (ie, on/off), respectively. During early stages of cell adhesion, integrin ligation to the extracellular matrix leads to a rapid decrease in basal RhoA activity. Several studies show that the suppression of RhoA activity after integrin engagement involves SFK-dependent phosphorylation and activation of a major GAP, p190RhoGAP.22 Interestingly, recent reports23,24 suggest that p190RhoGAP may regulate RhoA activity through its association with plasma membrane lipid raft domains. Based on these observations, we hypothesize that hemodynamic shear stress is translated via an integrin/caveolin-1 signaling complex that serves to temporally regulate a SFK/p190RhoGAP/RhoA axis and subsequent actin remodeling.
Antibodies and Reagents
All general buffers and reagents were obtained commercially, unless noted otherwise, as were SFK inhibitor PP2 and its negative control, PP3. The following primary antibodies were also obtained from commercial sources: caveolin-1 and p190RhoGAP polyclonal antibodies (pAbs), SFK and RhoA pAbs, β1 integrin mAb JB1A, MLC and β-actin monoclonal antibodies (mAbs), MLC diphosphorylated pAb, pY416 SFK pAb, anti–phosphotyrosine 4G10, and horseradish peroxidase–conjugated anti–rabbit and anti–mouse secondary antibodies.
Enhanced Flow In Vivo
All procedures were performed according to an Institutional Animal Care and Use Committee protocol approved by Temple University, Philadelphia, Pa. Briefly, wild-type and caveolin-1 (Cav1)−/− mice (aged 8 weeks; Jackson Laboratories, Bar Harbor, ME) were anesthetized with a mixture of ketamine and xylazine and body heat maintained at 37°C with a heating pad. The right common carotid artery was exposed and ligated by closure of a suture looped around the vessel. Because of blood shunting, this procedure increases blood flow in the contralateral left common carotid artery and enhances formation of stress fibers within endothelial cells lining the vessel wall.3,25,26 To verify flow changes, blood flow was measured at vessel midpoint using ultrasonography (VisualSonic Vevo 770 Doppler Ultrasound with a Scan Head [RMV 716]). One day after ligation, blood flow was enhanced by 50% in the left common carotid artery compared with sham-operated animals. Blood flow through the common carotid arteries of caveolin-1–deficient mice was similar to wild-type matched animals, and ligation induced the same increase in flow through the contralateral vessel. At the conclusion of the experimental point, the mice were euthanized with sodium pentobarbital and vessels were harvested, fixed, and immunolabeled.
Bovine aortic endothelial cells, purchased from Cell Applications, San Diego, CA, were grown in culture medium (MCDB-131) supplemented with 10% fetal bovine serum and 0.04-mg/mL gentamicin sulfate and maintained at 37°C, 95% humidity, and 5% CO2. All experiments were performed using cells below passage 8.
Cav1, p190RhoGAP, and Csk Small-Interfering RNA
Similar to previous reports,17,27 bovine aortic endothelial cells were transfected with caveolin-1, p190RhoGAP or Csk SMARTpool small-interfering RNA (siRNA), or siScrambled control. Briefly, BAEC's at 80% confluence were transfected with 100nM siRNA using DharmaFECT-1. Similar to previous results, cells were used 48 hours after transfection when expression levels of caveolin-1, p190RhoGAP, and Csk were reduced by >90% (supplemental Figure I; available online at http://atvb.ahajournals.org).
Cav1 pY14 Peptide
As described in detail previously,17 caveolin-1 peptides spanning the pY14 region were constructed with the Y14 site either phosphorylated (pY14) or unmodified (Y14). Peptides consisting of amino acids 1 to 27 (MSGGKYVDSEGHLY*TVPIREQCNIYKPNNC) of caveolin-1 with the tyrosine-14(*) either phosphorylated or unmodified were individually linked to biotinylated penetratin. Penetratin (300 nmol/L), coupled to either control or phosphopeptides, was incubated with endothelial cells for 1 hour before shear stress. To determine the efficiency of peptide uptake, cells were incubated with streptavidin–horseradish peroxidase, and biotinylated penetratin was detected via the oxidation of diaminobenzidine (supplemental Figure II).
In Vitro Shear Stress System
A parallel plate flow chamber (Streamer model) was used to subject endothelial cell monolayers to laminar shear stress at 10 dyne/cm2, as previously detailed.17 The parallel plate flow chamber was connected to a recirculating flow circuit composed of a variable-speed peristaltic pump, a fluid capacitor that dampens pulsation, and a reservoir with culture medium. A computer-operated pump (Masterflex L/S) was controlled with software (Streamsoft v1.0) and produced fluid flow. Shear stress was determined by the following equation: Shear Stress=6μQ/bh2, where Q is the flow rate in cm3/min, b is the channel width in cm, h is the channel height in cm, and μ is the dynamic viscosity of the fluid in dyne/cm2. Before placement in the parallel-plate apparatus, endothelial monolayers are acclimated for 4 hours in “flow media” consisting of culture medium (MCDB-131) containing 1% fetal bovine serum. The temperature was maintained at 37°C, and pH and oxygen levels were maintained in a 95% air/5% CO2 humidified incubation chamber.
Briefly, p190RhoGAP monoclonal antibody was conjugated to sheep–anti–mouse-coated paramagnetic dynabeads (Dynal). Endothelial cell lysates (250 μg) were incubated with the antibody/bead conjugates for 4 hours at 4°C. The bound fraction was separated from the unbound material and processed for Western blot analysis.
RhoA Activation Assay
After shear stress, cell lysates were added to beads (Rhotekin-RBD GST) for 1 hour at 4°C. GTP-RhoA associated with the beads was determined by Western blotting.
Endothelial cells were processed for Western blot analysis as previously described.17 Autorads were scanned and digitized, and band intensities were quantified using software (Image J).
Caveolae Immunoaffinity Isolation
Caveolar vesicles were purified as previously described.18 Briefly, static and shear-exposed bovine aortic endothelial cells were scraped into detergent-free Tricene buffer (250-mmol/L sucrose, 1-mmol/L EDTA, and 20-mmol/L Tricene, pH 7.4) and processed to enrich plasma membranes. Plasma membranes were subsequently sonicated and incubated with anticaveolin-1–conjugated goat anti–mouse IgG-coated magnetic beads for 1 hour at 4°C. Bound material, representative of caveolae vesicles, was separated magnetically from unbound noncaveolar membranes.
Microscopy and Quantification of Stress Fibers
Cell monolayers and carotid vessels were fixed (4% formaldehyde) and permeabilized with 100-μg/mL saponin in HEPES buffer (10-mmol/L HEPES, 100-mmol/L potassium chloride, and 5-mmol/L magnesium chloride) and labeled with phalloidin AF488 to visualize F-actin (stress fibers) and 4≪,6-diamidino-2-phenylindole (nucleus). Images were captured using a confocal microscope (Nikon Eclipse TE300). Fluorescence signal intensity was quantified using EZ C1 3.90 Free viewer. Briefly, the average fluorescence intensity along a line connecting the edges of the nucleus between 2 paired endothelial cells was recorded. For each experiment (n=3), 4 random fields were chosen and 15 paired cells per field were unbiasly selected for analysis.
For each study, data were gathered from at least 3 independent experiments and pooled according to group. Mean and SD were calculated, and differences between groups were analyzed with an unpaired 2-tailed Student t test or an ANOVA with a post hoc Tukey test using software (STATGRAPHICS 4.0). Differences between control and experimental groups were significant at P<0.05.
Induction of Actin Stress Fibers in Endothelial Cells Under Enhanced Flow In Vivo Requires Cav1
Past in vitro studies indicate that caveolin-1 and β1 integrin associate under enhanced flow conditions to form a signaling complex that regulates endothelial cytoskeletal elements. To evaluate key aspects of this mechanotransduction process in vivo, F-actin was labeled with a phalloidin-fluorophore conjugate in endothelial cells residing in carotid arteries of wild-type or caveolin-1–deficient mice after surgical enhancement of blood flow. In sham-operated wild-type animals, F-actin was organized in dense peripheral bands and as short stress fibers that were randomly oriented through the endothelial cell whereas F-actin appeared less developed in Cav1−/− mice (Figure 1). In response to flow, the F-actin signal increased 2-fold in wild-type vessels with the development of stress fibers that oriented parallel to the direction of blood flow. However, in carotid vessels of caveolin-1–deficient mice, stress fibers failed to form and F-actin content of the endothelium was similar to sham-operated animals.
Integrin/Caveolae Mechanosignaling Complex Regulates p190RhoGAP Phosphorylation
To more effectively evaluate the molecular signaling mechanism that relays from the β1 integrin/caveolin-1 complex to the actin cytoskeleton in response to flow, we exposed cultured endothelial cells to well-defined fluid shear stress using a parallel plate apparatus. Previous studies indicate that signaling components localized to an integrin- and caveolae-based mechanosignaling complex can spatially and temporally regulate shear-mediated SFK activity.17 A more complete evaluation of each element within this complex on shear-induced SFK activation revealed that β1 integrin and caveolin-1 serve to temporally regulate shear-induced SFK activity (supplemental Figure III).
An important function of SFK is to propagate and amplify mechanosignals that influence cell structure in response to shear stress.28 To address the molecular events that bridge integrin/caveolin regulation of SFKs to endothelial cell adaptation to flow, we focused on signaling molecules that are influenced by SFKs and regulate downstream pathways that govern the cells' cytoskeletal architecture (ie, p190RhoGAP and RhoA). Figure 2 illustrates that after the short-term onset of shear stress, p190RhoGAP tyrosine phosphorylation initially increases, followed by a decrease over time. Antibody blockade of β1 integrins prohibited these shear-induced phosphorylation events, whereas both caveolin-1 siRNA and a caveolin-1 phosphopeptide prevented the decline in p190RhoGAP phosphorylation (Figure 2A, 2C, and 2D, respectively). The data regarding both the caveolin-1 and caveolin-1 peptide experiments are consistent with their effects on enhancing SFK activity and thereby maintaining phosphorylation of p190RhoGAP. Last, inhibition of SFK with PP2 significantly decreased basal levels of p190RhoGAP phosphorylation and subsequent shear-induced tyrosine phosphorylation of p190RhoGAP (Figure 2B). Taken together, these findings are the first to describe p190RhoGAP as a shear-sensitive signaling molecule. Fold change in p190RhoGAP tyrosine phosphorylation is presented in Figure 2E.
Integrin/Caveolae Mechanosignaling Complex Temporally Regulates RhoA Activity
In a similar manner to the experiments previously described, we evaluated shear-induced activity status of the p190RhoGAP target, RhoA. Figure 3 illustrates that RhoA activity initially declines from baseline levels, followed by an increase at 30-minute exposure to shear stress. To determine whether integrins relay shear-induced signals to enhanced RhoA activity, cells were pretreated with JB1A. We found that JB1A significantly attenuated the activation of this small GTPase in response to shear stress (Figure 3A). We also found that depletion of caveolin-1 enhanced basal RhoA, which was sustained at all shear stress points (Figure 3B). In cells expressing the Cav1 phosphopeptide, in which both SFK and p190RhoGAP phosphorylation is sustained, a marked inhibition of shear-induced RhoA activity was observed (Figure 3C). Similar to caveolin-1 siRNA treatments, depletion of p190RhoGAP substantially elevated basal RhoA activity in endothelial cell monolayers (Figure 3D). Fold change in RhoA activity is depicted in Figure 3E.
Integrin/Caveolae Mechanosignaling Complex Regulates MLC Phosphorylation and Actin Stress Fiber Formation
The phosphorylation of MLC is a critical event in the formation of distinct stress fibers that accompany endothelial cell acclimation to shear stress. A recent report17 demonstrates that both integrin activation and caveolae/caveolin-1 are upstream of shear-induced MLC phosphorylation. Similar to our past findings and consistent with the effects of JB1A on SFK, p190RhoGAP, and RhoA, blockade of β1 integrin attenuated MLC phosphorylation (Figure 4A). Similar results were observed with both the 3S3 and 6S6 β1 integrin blocking antibodies that target alternative epitopes of the integrin molecule (data not shown). Although loss of caveolin-1 enhanced SFK and RhoA activity, these events did not translate to phosphorylation of MLC (Figure 4B). We also found that temporal regulation of SFK was important for proper phosphorylation of MLC because caveolin-1 phosphopeptide and Csk siRNA also attenuated shear-induced MLC phosphorylation (Figure 4C and 4D). Last, depletion of p190RhoGAP, which activates RhoA, significantly enhanced baseline phosphorylation of MLC, which increased modestly in response to shear stress (Figure 4E).
Our in vivo findings demonstrated that caveolae/caveolin-1 are necessary cellular components for actin stress fiber formation in response to enhanced flow (Figure 1). To evaluate the other components of the integrin/caveolae mechanosignaling pathway on actin reorganization, cultured endothelial cells were also examined for actin stress fiber formation after shear stress. Exposing endothelial monolayers to 6 hours of laminar shear stress significantly enhanced F-actin labeling (Figure 5) and formation of stress fibers (supplemental Figure IV). These events correlated with continued and progressive phosphorylation of MLC throughout the 6-hour shearing period (supplemental Figure V). Interestingly, RhoA activity was also sustained through a 3-hour observation point but decreased by 6 hours (supplemental Figure V), perhaps indicating that the endothelial monolayers are adapting to the flow environment. As in the case of short-term responses to shear stress, JB1A blocked long-term RhoA activity, MLC phosphorylation (supplemental Figure V), and stress fiber formation (Figure 5 and supplemental Figure IV). Depletion of caveolin-1 also inhibited shear-induced MLC phosphorylation and enhanced F-actin labeling and the development of stress fibers in endothelial cell monolayers (Figure 5 and supplemental Figure IV). Similar to the earlier observation points, decreasing caveolin-1 protein expression enhanced the basal activity of RhoA through 6 hours of surveillance (supplemental Figure V).
Caveolar Localization of p190RhoGAP and RhoA
Recently, it was reported that shear stress induces the transposition of β1 integrin into caveolae plasma membrane domains, where they participate in the formation of the integrin/SFK/phosphocaveolin-1/Csk signaling complex.18 To determine whether the downstream elements examined herein (ie, p190RhoGAP and RhoA) are present in caveolar vesicles, we immunoaffinity isolated caveolae from endothelial cells. Figure 6 shows that under static culture conditions, little p190RhoGAP is present in caveolae. Consistent with other reports,21,27,29 a pool of RhoA is constitutively targeted to caveolae domains. After the onset of shear stress, p190RhoGAP was detected in caveolae, with significant transposition observed by 5 minutes after shear, whereas the expression level of RhoA remained unchanged. The association of p190RhoGAP appeared to be transient because we observed a significant loss of p190RhoGAP in the caveolae fraction after cells were subjected to 30 minutes of shear stress. On the contrary, the RhoA content increased by more than 5-fold in caveolae domains at this later point.
Subjecting endothelial cells to short-term increases in shear stress, both in vivo and in vitro, sets off a signaling cascade that results in the alteration of cell morphological features that serves the endothelium in adapting to new flow environments. A major feature of this process is the reorganization of the cytoskeletal network and the development of actin stress fibers. Although the precise signaling mechanism that regulates this process is not fully understood, the temporal regulation of RhoA activity appears to be essential for proper cytoskeletal realignment and adjustment of endothelial cells to changes in shear.19,30,31 Several studies have also demonstrated that shear-induced RhoA activation is secondary to integrin activation; however, the mechanosignaling mechanisms that link integrins to RhoA and actin remodeling remain unclear. Past work from our laboratory17,18 indicates that the shear-induced phosphorylation of MLC, an event that is associated with actin stress fiber formation, is mediated by a β1 integrin/caveolin/Csk signaling complex and depends on the presence of caveolar organelles. In this study, we systematically evaluated whether this mechanosignaling complex constitutes a bridge from integrins to RhoA and stress fiber formation in response to shear stress.
Consistent with another report,19 we observed an initial reduction in basal RhoA activity, followed by an increase in activity through 3-hour exposure to shear stress before returning toward baseline. We found that β1 integrin participated in this temporal pattern of shear-induced RhoA activity because significant changes in activity were not observed in cells that were pretreated with a β1 integrin blocking antibody, JB1A (Figure 3A and supplemental Figure V). These findings are in agreement with events described for regulation of RhoA activity during the process of cell adhesion, in which β1 integrin ligation to extracellular matrix components activates signaling pathways that first suppress RhoA activity and then, during later stages of cell spreading, are enhanced.22,32 Thus, cell adhesion and mechanotransduction processes appear to share similar signaling properties.
Next, we examined the influence of SFKs on RhoA activity given that the formation of a β1 integrin/caveolin-1/Csk signaling complex was crucial for the temporal regulation of SFK activity and subsequent relay to important RhoA targets, such as MLC. As previously described,17 SFK activation in response to shear stress was rapid, yet transient (supplemental Figure III). In experiments in which SFK activity was sustained, either through depletion of caveolin-1 or sequestration of Csk with a phosphorylated N-terminal caveolin-1 peptide, the temporal regulation of RhoA activity by shear stress was significantly altered (Figure 3). These findings illustrate that SFKs play a role in regulation of RhoA activity by shear stress. Because SFK can influence RhoA activity through SFK-dependent phosphorylation of signaling intermediates in response to integrin engagement,22 we evaluated the pattern of p190RhoGAP phosphorylation following shear stress. After the onset of shear stress, p190RhoGAP phosphorylation increased, followed by dephosphorylation events that correlated with GDP/GTP loading of RhoA. More important, the reduction of p190RhoGAP protein expression resulted in the loss of temporal regulation of RhoA activity induced by shear stress (Figure 3D). Together, these findings suggest that p190RhoGAP function is sensitive to shear stress and that RhoA is a direct p190RhoGAP target in mechanotransduction.
Changes in the temporal sequence of RhoA activity observed in endothelial cells during the first hour of exposure to shear stress are critical for long-term adaptive responses, such as actin reorganization and cell alignment.19 Herein, we show that, in addition to serving as proximal signaling elements that govern RhoA activity, β1 integrin, caveolin-1, and p190RhoGAP are necessary for the formation of stress fibers induced by prolonged exposure to shear stress (Figures 1 and 5 and supplemental Figure IV). Collectively, the studies demonstrate that maintaining the shear-induced off/on activity status of RhoA is essential for endothelial cells to morphologically adapt to flow. Moreover, our findings advance this concept by describing a molecular signaling mechanism that relays from integrins to RhoA to the cytoskeleton with p190RhoGAP being implicated in the shear stress responses in endothelial cells for the first time.
Signaling transduction is often compartmentalized within particular cellular microdomains. Previously, it was reported that plasma membrane caveolae served as sites for the recruitment and formation of a β1 integrin signaling complex that communicated with downstream elements involved in endothelial morphological adaptation to flow.17 In the current study, we also discovered that both p190RhoGAP and RhoA associated with caveolae, although to varying degrees, after short-term exposure of endothelial cells to shear stress. Similar to previous reports,23,24 we found that a pool of RhoA was basally targeted to caveolae and RhoA agonists, including shear (present study), induced further recruitment of this GTPase to these domains. In contrast, we found little p190RhoGAP present in caveolae derived from control, nonsheared, endothelial cells. However, shear stress induced recruitment of p190RhoGAP to caveolae at an early point when RhoA activity was depressed. When RhoA became most active, p190RhoGAP appeared to be released from these domains. Although we did not measure RhoA activity status in caveolae, per se, the observed compartmentation of both p190RhoGAP and RhoA suggests that caveolae serve as platforms that allow for temporal regulation of RhoA activity in response to shear stress. This concept is supported by recent evidence23 showing p190RhoGAP/RhoA interactions in lipid raft domains of the plasma membrane. In those studies, RhoA is most active in nonadherent cells and p190RhoGAP was observed to reside mainly outside of the lipid raft, where its activity was downregulated. On adhesion, p190RhoGAP accumulated in raft domains, where it became active and able to suppress RhoA activity. The similarity in observations between these studies and the present study suggests that initiation of shear stress activates signaling events that may recapitulate those of cell adhesion.
Interestingly, depletion of caveolin-1 elevated the activity of RhoA and SFKs over baseline (supplemental Figure III) in our system. Although these observations are consistent with the function of caveolin-1 as a signaling molecule inhibitor, the increase in basal SFK and RhoA did not translate to larger changes in the cells' phenotype, such as remodeling of the actin cytoskeleton (Figure 5 and supplemental Figure IV). On the surface, this may seem surprising. However, when considering that each of these signaling molecules can be compartmentalized within caveolae microdomains (Figure 6), along with several of their downstream targets, then the loss of caveolin-1 and, more important, caveolae (as a consequence) would be expected to disrupt efficient signal propagation from these sites. In support of this concept and consistent with previous observations,29 we found that disruption of caveolae and lipid rafts with cholesterol-depleting agents, such as methyl-β-cyclodextrin, shift SFK and RhoA from these membrane microdomains to other cellular compartments (supplemental Figure VI) and provide additional evidence for the localization of these mechanosensitive signaling molecules to rafts and caveolae.
Although other integrin subtypes, GTPases, and second messengers have been implicated in mechanotransduction pathways that contribute to endothelial adaptation to flow,33 our collective studies in this area indicate that shear stress also induces the formation of a caveolae-based signaling complex in which mechanically sensitive integrins translocate to caveolae and associate with caveolin-1. This complex appears to regulate endothelial functions that participate in vascular remodeling events initiated by altered flow. In support of this concept, recent studies using endothelial cell–specific heterozygous β1 integrin gene depletion (which reduced β1 integrin protein expression in endothelial cells by 40%) show abnormal vascular remodeling in response to experimentally altered blood flow in vivo.34 Our experiments in endothelial cell cultures and animal models of flow enhancement demonstrate that flow-induced remodeling of the actin cytoskeleton requires caveolin-1/caveolae. Taken together, these studies demonstrate that both caveolin-1/caveolae and β1 integrin are linked in their functions as mechanotransduction elements within endothelial cells and may serve as appropriate targets to modulate endothelial sensitivity to hemodynamic forces.
Sources of Funding
This study was supported by grant HL086551 from the National Institutes of Health; and grant GIA-2280303 from the American Heart Association.
We thank John Wilkins, PhD, University of Manitoba, Winnipeg, for the 3S3 and 6S6 a1 integrin- blocking antibodies.
- Received October 16, 2009.
- Accepted October 26, 2010.
- © 2011 American Heart Association, Inc.
- Davies PF
- Kim DW,
- Langille BL,
- Wong MK,
- Gotlieb AI
- White GE,
- Fujiwara K
- Wong AJ,
- Pollard TD,
- Herman IM
- Shyy JY,
- Chien S
- Nauli SM,
- Kawanabe Y,
- Kaminski JJ,
- Pearce WJ,
- Ingber DE,
- Zhou J
- Osawa M,
- Masuda M,
- Kusano K,
- Fujiwara K
- Rizzo V,
- Sung A,
- Oh P,
- Schnitzer JE
- Wang Y,
- Miao H,
- Li S,
- Chen KD,
- Li YS,
- Yuan S,
- Shyy JY,
- Chien S
- Radel C,
- Rizzo V
- Tzima E,
- Angel del Pozo M,
- Shattil S,
- Shein S,
- Schwartz M
- Wojciak-Stothard B,
- Ridley AJ
- Mammoto A,
- Huang S,
- Ingber DE
- Wickstrom SA,
- Alitalo K,
- Keski-Oja J
- Korshunov VA,
- Berk BC
- Rudic RD,
- Bucci M,
- Fulton D,
- Segal SS,
- Sessa WC
- Carlile Klusacek M,
- Rizzo V
- Kawamura S,
- Miyamoto S,
- Brown JH
- Lin T,
- Zeng L,
- Liu Y,
- DeFea K,
- Schwartz MA,
- Chien S,
- Shyy JY
- Ren XD,
- Kiosses WB,
- Schwartz MA
- Katsumi A,
- Orr AW,
- Tzima E,
- Schwartz MA
- Lei L,
- Liu D,
- Huang Y,
- Jovin I,
- Shai SY,
- Kyriakides T,
- Ross RS,
- Giordano FJ