Lipoprotein Promotes Caveolin-1 and Ras Translocation to Caveolae
Role of Cholesterol in Endothelial Signaling
Abstract—To explore the role of LDL in caveolin-Ras regulation in human endothelial cells (ECs), we incubated confluent human umbilical vein endothelial cells (HUVECs) with LDL. This resulted in a high steady-state caveolin-1 (Cav-1) expression at both the mRNA and protein levels. LDL exposure appeared not to regulate the abundance of Cav-1. Immunofluorescence staining showed that Cav-1 protein migrated from the cytoplasm to the cell membrane after LDL exposure. Cav-1 protein and cholesterol partitioned mainly into the caveola fractions, and LDL increased both Cav-1 and cholesterol in these fractions. Ras protein in caveola fractions was also increased by LDL. Increased Ras was detected in Cav-1 immunoprecipitated samples, and conversely, increased Cav-1 was found in Ras-immunoprecipitated samples. We also demonstrated LDL-increased Ras activity in HUVECs by measuring the GTP/GTP+GDP ratio of Ras with [32P]orthophosphate labeling in the cells. Finally, we determined the binding of [3H]-labeled free cholesterol and recombinant H-Ras to Cav-1 fusion proteins in vitro. Both cholesterol and Ras bound to full-length GST–Cav-1, scaffolding domain (61–101), and C-terminal (135–178) Cav-1 fusion peptides. Addition of cholesterol enhanced Ras binding to the full-length and scaffolding domain of Cav-1 but not to the C-terminal Cav-1. These findings strongly suggest a role for Cav-1 in cholesterol trafficking and cholesterol-mediated intracellular signaling, which may mediate EC activation by LDL.
- Received March 27, 2000.
- Accepted July 12, 2000.
In peripheral cells, exposure to elevated plasma cholesterol results in the selective uptake of free cholesterol from native LDL and in the stimulation of cholesterol efflux.1 Regulation of cellular free cholesterol occurs as LDL levels are elevated over the physiological range, and this regulation has been linked to free cholesterol efflux in caveolae.2 Caveolae are vesicular organelles representing a subdivision of the plasma membrane. They are abundant in endothelial cells (ECs) both in vivo3 and in vitro,4 where the rate of receptor-mediated endocytosis of lipoproteins such as LDL is low.2 Numerous individual signal transduction molecules are located in caveolae and indicate that caveolae also serve to compartmentalize, modulate, and integrate signaling events at the cell surface.5 Cholesterol can move bidirectionally between the endoplasmic reticulum and the plasma membrane. Furthermore, cholesterol is one of the major components of the caveola lipid core and has been shown to play a key role in regulating the activity of membrane transport proteins, receptors, and enzymes.2 5
Caveolin-1 (Cav-1) is a 21- to 24-kDa cholesterol-binding membrane protein that is a major structural component of caveolae.6 Cav-1 appears to provide a direct means for resident caveola proteins to be sequestered within caveola microdomains.7 Cav-1 may organize the formation of caveola microdomains and regulate caveola-related signaling events as it interacts with protein kinase C, Src-like kinases, Ras, and endothelial NO synthase.7 8 Ras, a small GTPase molecule, plays a key role in promoting cell responses to mitogens, cytokines, UV irradiation, and other environmental stresses. After synthesis and posttranslational modification, Ras moves to the plasma membrane. Cav-1 can bind Ras in vivo and in vitro, but whether Cav-1 transports Ras to caveolae or retains Ras within the caveolae is not clear.9 Ras can trigger at least 3 diverging mitogen-activated protein kinase (MAPK) cascades, including the one operating through MAPK/extracellular signal-regulated kinase kinase-1 and c-Jun NH2-terminal kinase kinase to activate c-Jun NH2-terminal kinases (JNKs). Recently, we determined that LDL can activate nuclear c-Jun/AP-1 and JNK via Ras activation.10 11 Thus, we hypothesize that this activation occurs by a cholesterol–Cav-1 protein interaction. We postulate that LDL can increase cholesterol–Cav-1 binding and perturb the structure of caveolae in ECs. Because many receptor and signaling proteins are localized in caveolae, perturbation of caveolae by cholesterol from LDL probably initiates intracellular signaling, leading to cell activation.
Cell Culture and LDL Isolation
Human umbilical vein endothelial cells (HUVECs) were isolated and maintained in EC medium as described previously.10 All experiments were performed with cells from passages 2 and 3 cultured to confluence before LDL treatment. The human hepatoma cell line HepG2 was purchased from ATCC and cultured in MEM supplemented with 5% FBS. LDL was isolated from nonfrozen human plasma as described.10 For all studies, LDL was used at a final cholesterol culture concentration of 240 mg/dL (6.24 mmol/L).
Purification of Cav-1–Enriched Membrane Fractions
Cav-1–enriched membrane fractions were isolated from two 150-mm dishes of HUVECs by use of a modified detergent-free extraction procedure as previously described.9 For analysis of the resulting gradient, fractions were collected from the top of the gradient, and 40 μL of the fractions was subjected to SDS-PAGE. Subsequent Western blotting analyses with anti–Cav-1 antibodies were applied to establish Cav-1 abundance. Total cholesterol concentration of the different fractions was determined by an enzymatic assay.12
Immunoelectron microscopy was used to visualize Cav-1 in HUVECs.13 Indirect immunostaining was performed with rabbit anti–Cav-1 IgG and a horseradish peroxidase–conjugated secondary antibody. After postfixing and peroxidase reaction, cells were embedded in epoxy resin, and thin sections were visualized with an electron microscope (Philips CM300).
In Situ Immunofluorescence Staining
HUVECs grown on chamber slides were exposed to LDL (240 mg/dL) for 2 or 18 hours. After exposure, the cells were fixed with 2% paraformaldehyde and stained with indirect immunofluorescence using rabbit anti–Cav-1 IgG and an FITC-conjugated secondary antibody against rabbit IgG. The results were observed under a fluorescence microscope.14
Immunoprecipitation and Western Blotting Analysis
Protein isolation and coimmunoprecipitation (co-IP) were performed as previously described.15 16 Briefly, cell lysates were harvested in a buffer containing 50 mmol/L HEPES (pH 7.5), 125 mmol/L NaCl, 0.5% CHAPS, 1 mmol/L EDTA, 2 mmol/L DTT, and protease inhibitors. The membrane and cytosolic proteins from whole-cell lysate or from the ultracentrifugation-derived fractions were isolated.16 After the sample had been precleared with protein A/G plus agarose, aliquots of the supernatant were incubated with anti–Cav-1 or anti-Ras IgGs for 1 hour at 4°C. Then, protein A/G plus agarose was added for an additional hour of incubation at 4°C. Bound immune complexes were washed 3 times with the lysis buffer and then once with 50 mmol/L Tris-HCl (pH 7.4) and 150 mmol/L NaCl. Western analyses with antibodies against Cav-1 and H-Ras were performed as previously described.11
Ras Activation Assay
Confluent HUVECs were labeled with 0.2 mCi of [32P]orthophosphate/mL for 4 hours in a phosphate-free EC medium. After labeling, the cells were exposed to phosphate-free LDL (240 mg/dL) for various time periods. Cell lysates were harvested, and Ras proteins were immunoprecipitated. Then, bound guanine nucleotides were eluted from the precipitated protein complexes and analyzed by thin-layer chromatography.18 Ras GDP and GTP contents were assessed by autoradiography, and the ratio of GTP to GTP+GDP was determined by densitometry.
GST–Cav-1 Fusion Proteins and Cholesterol or H-Ras Binding Assay
The constructs of GST–Cav-1 fusion proteins encoding full-length Cav-1 and peptides corresponding to N-terminal residues (1 to 21, 1 to 81, and 61 to 101) and C-terminal residues (135 to 178) were provided by M. Lisanti (Albert Einstein College of Medicine, Bronx, NY). The expression and purification of the fusion proteins were conducted as previously described.9
Purified GST–Cav-1 fusion proteins (0.2 nmol) bound to GSH-agarose beads were prewashed with a binding buffer containing proteinase inhibitors.9 19 Equimolar diluted [3H]cholesterol or 0.5 μg of recombinant wild-type H-Ras (Oxford Biomedical Research, Inc) was added to the protein-bound beads in 0.5 mL of buffer. After rotation in 4°C for 16 hours, the bound complexes were washed 6 times with cold PBS. To measure the cholesterol binding on Cav-1, the [3H]cholesterol bound to GST–Cav-1 was eluted with a buffer containing 10 mmol/L reduced glutathione. The abundance of cholesterol on Cav-1 fusion proteins was determined by liquid scintillation counting. To measure H-Ras binding on Cav-1, the H-Ras bound to GST–Cav-1 was eluted and a Western analysis with antibody against Ras was performed.
LDL Increases Cellular Cholesterol Content in ECs
We previously reported that incubation of ECs with LDL increased cellular cholesterol content and cholesterol phospholipid molar ratios of the EC membrane.12 To determine the time course of this effect, we treated HUVECs with LDL for 30 minutes to 2 days. LDL exposure increased cellular cholesterol content within 30 minutes, and the cholesterol level remained elevated for at least 2 days. As seen in Figure 1⇓, the total cholesterol level in LDL-exposed cells increased 38% in 30 minutes and 49% in 3 hours compared with control (P<0.05). Furthermore, after 6 to 48 hours of LDL exposure, the ratio of cholesterol in the LDL-exposed samples over control ranged from 153% to 174%.
LDL Increased Abundance of Cholesterol and Cav-1 Protein in Caveolae
To determine how LDL affected the levels and distribution of cholesterol and Cav-1 in caveolae, cell fractions were isolated from ECs after LDL exposure.9 Cholesterol was present mainly in fractions 4 to 6, which contain most of the caveola membrane. After a 2-hour exposure of HUVECs to LDL, the cholesterol content in those fractions increased substantially (Figure 2A⇓). Western blot analysis showed that the Cav-1 protein was also located mainly in fractions 4 to 6, whereas the majority of total cell proteins were in fractions 9 to 13, determined by Coomassie blue staining of the SDS-PAGE. The Cav-1 protein in fractions 4 to 6 was increased after 2 hours of LDL exposure (Figure 2B⇓). Similar results were observed when HUVECs were exposed to LDL for 24 hours (data not shown). To ascertain that the change of cholesterol distribution was directly associated with Cav-1 in HUVECs, a hepatoma cell line (HepG2) was used as a negative control, because we did not detect Cav-1 in either mRNA or protein levels in this cell line (data not shown). Correspondingly, we did not detect any Cav-1 protein in any of the fractions, and the pattern of cholesterol distribution after fractionation in HepG2 cells was different from that in HUVECs (Figure 2⇓). Furthermore, to quantitatively compare the abundance of Cav-1 in the fractions in HUVECs, we pooled fractions 1 to 3, 4 to 6, and 7 to 9 and then concentrated the proteins from the fractions. The Cav-1 protein in equal amounts of protein from different fractions was measured by Western blot analysis. As shown in Figure 2C⇓, LDL exposure increased the Cav-1 protein abundance in caveola fractions compared with control by ≈47%.
LDL Increases Cav-1 Membrane Translocation but Does Not Affect Cav-1 Protein Synthesis
The accumulation of Cav-1 in caveolae after LDL exposure could be the result of either altered protein trafficking or increased de novo protein synthesis. To investigate whether LDL regulates Cav-1 at the mRNA or protein synthesis level, we incubated confluent HUVECs with LDL from 0.5 to 24 hours. In untreated cells, both Northern and Western analyses showed a high expression level of Cav-1, and immunoelectron microscopy with an anti–Cav-1 antibody also showed an abundance of caveola microstructures, which represent plasma membrane-bound vesicles 60 to 80 nm in diameter (data not shown). LDL exposure up to 24 hours did not appear to regulate the level of Cav-1 mRNA or protein in HUVECs. However, the results of in situ immunofluorescence staining with an anti–Cav-1 antibody showed that Cav-1 protein migrated from the cytoplasm to the cell membrane after 2 to 18 hours of LDL exposure (Figure 3⇓). Thus, LDL-regulated Cav-1 trafficking in human ECs is not due to increased de novo protein synthesis but rather to movement of preexisting Cav-1 protein from the cytoplasm to the membrane caveola microdomains.
Ras Is Coupled With Cav-1 and Can Be Enriched in Caveolae by LDL
Ras plays a key role in promoting cell responses to environmental stresses. After synthesis on free cytoplasmic ribosomes, Ras moves to the plasma membrane, where it is activated. To explore whether LDL affects Ras within ECs, we exposed HUVECs to LDL for different times and then measured Ras in whole-cell lysate and in the cell membrane. We did not detect any change in the amount of Ras protein in the whole-cell lysate up to 6 hours. However, Ras abundance in the cell membrane increased at 0.5 hour, reached a peak in 1 to 2 hours, and returned to basal level in 6 hours. Under the same conditions, Cav-1 translocation to the cytoplasmic membrane occurred in a sustained pattern after LDL exposure. The abundance of Cav-1 in the membrane increased after 0.5 hour and remained at a high level at 6 hours (data not shown). In addition, using the detergent-free sucrose gradient method to isolate cell fractions, we detected Ras in the caveola-enriched fractions (5 to 6) as well as in the bottom fractions (11 to 12). LDL exposure significantly increased Ras in the caveola-enriched fractions (Figure 4A⇓). To ascertain whether Ras and Cav-1 are directly associated with and enriched by LDL in ECs, we performed a cross-IP of Cav-1 and Ras with the caveola fractions. The direct association of Cav-1 and Ras in caveolae was confirmed in human ECs, because Ras could be detected in the Cav-1 IP samples and Cav-1 in Ras IP samples. The results also showed that LDL enriched the abundance of Ras in Cav-1 IP samples and vice versa (Figure 4B⇓). These data not only demonstrated that Ras bound directly to Cav-1 in human ECs but also showed that the abundance of Ras in caveolae was increased by LDL exposure.
LDL Activates Ras in Human ECs
To study whether LDL can activate Ras in ECs, we measured activation of Ras in LDL-exposed HUVECs by analysis of the change in GTP to GTP+GDP ratio in Ras protein. After labeling with [32P]Pi in a phosphate-free medium, the cells were exposed to LDL or phorbol 12-myristate 13-acetate (PMA) for different times. Cell proteins were extracted, and Ras protein was immunoprecipitated. The bound guanine nucleotides were eluted from the IP protein complexes and separated by thin-layer chromatography. As shown in Figure 5⇓, the ratio of GTP/GTP+GDP in the LDL-exposed sample was increased at 30 and 60 minutes, whereas PMA increased the ratio in 5 minutes and returned to basal level within 30 minutes. Taken together, the data from Figures 4⇑ and 5⇓ demonstrate that LDL promotes Ras activation and membrane translocation in human ECs.
Cholesterol and Ras Associate to Same Domains of Cav-1 Protein, and Cholesterol Increases Ras–Cav-1 Association
To evaluate the effect of cholesterol on Cav-1–Ras interaction, we used the recombinantly expressed full-length Cav-1 and portions of Cav-1 as GST fusion proteins produced in Escherichia coli. These GST–Cav-1 fusion proteins were constructed and characterized previously9 (also see Figure 6A⇓, top) and were incubated with purified wild-type H-Ras in vitro. H-Ras interacted specifically with both full-length Cav-1 and the scaffolding domain (61 to 101) as reported.9 Surprisingly, H-Ras also interacted with C-terminal (135 to 178) Cav-1 fusion peptides, as shown in Figure 6A⇓ (bottom). This effect was not due to nonspecific binding, because H-Ras did not bind to the other 2 N-terminal Cav-1 fusion peptides (1 to 21, 1 to 81). To directly confirm cholesterol association with Cav-1, the binding of [3H]-labeled free cholesterol to recombinant Cav-1 fusion proteins was determined in vitro. As shown in Figure 6B⇓, cholesterol interacted with full-length GST–Cav-1 and scaffolding domain (61 to 101) and C-terminal (135 to 178) Cav-1 fusion peptides but not to the other 2 N-terminal Cav-1 fusion peptides (1 to 21, 1 to 81). When H-Ras and the Cav-1 fusion proteins were incubated with increasing amounts of free cholesterol, cholesterol enhanced H-Ras interaction with full-length GST–Cav-1 and the scaffolding domain (61 to 101) but had little effect on H-Ras interaction with C-terminal (135 to 178) Cav-1 fusion peptides (Figure 6C⇓). The cholesterol-enhanced H-Ras–Cav-1 interaction in the Cav-1 scaffolding domain suggests a role for cholesterol interaction with Cav-1 interacting proteins, including Ras, through binding to Cav-1 within the same domain.
Numerous articles, including our own, have demonstrated that LDL, a risk factor for atherosclerosis, is an EC activator.10 11 12 20 In this report, evidence is presented showing a mechanism whereby LDL can initiate an intracellular signaling pathway leading to cellular phenotype modulation. Thus, we hypothesize that cholesterol is a biologically active molecule in LDL that initiates cell activation. Among the implications of this hypothesis is the possibility not only that LDL and serum cholesterol are risk factors for development of atherosclerosis but also that cholesterol can act directly as an atherogenic causative factor. In testing this hypothesis, we observed that LDL increases cholesterol in the cell membrane, mainly in the caveola fractions. LDL promotes the migration of Cav-1 protein from the cytosol to the cell membrane caveolae. As an important signaling mediator, Ras was activated by LDL and was also translocated to the caveola fractions associated with Cav-1 in LDL-exposed ECs. This is the first report showing the interaction of caveolin, cholesterol, and Ras in a model simulating a pathophysiological circumstance. Our findings strongly suggest a role for cholesterol–Cav-1 interaction in localizing Ras to the plasma membrane and enabling its signaling activity in ECs.
Prolonged incubation of ECs with LDL increased cellular cholesterol content and cholesterol/phospholipid molar ratios in the EC membrane, resulting in a reduction in relative EC plasma membrane fluidity.12 In the present report, LDL can increase cellular cholesterol in ECs as early as 30 minutes. The increase in cholesterol is located mainly in membrane caveolae. Cav-1 protein in caveolae also appears to be increased by LDL. Thus, LDL promotes Cav-1 protein translocation to membrane caveolae in human ECs. Pharmacological lowering of intracellular cholesterol levels decreases lipid-anchored membrane proteins clustering properly in caveolae, disassembles invaginated caveolae in the cell membrane, and decreases caveola-mediated intracellular and transcellular transport of macromolecules.21 22 These effects are consistent with the findings in this report. However, this report is the first demonstration that raising cellular cholesterol increases the membrane translocation of caveolin in human vascular ECs.
LDL/cholesterol loading upregulated Cav-1 mRNA and protein levels in fibroblasts and in bovine aortic ECs.23 24 However, we did not observe this pattern in human ECs with LDL exposure up to 24 hours using medium containing 20% FBS. These disparities are most likely due to cell type and culture conditions, which may, in turn, affect the abundance of Cav-1 and its response to extracellular stimuli. In our experiments, we used ECs in an early passage (passage 2 or 3). In these cells, high basal levels of Cav-1 were detected by both Northern and Western blots. However, serum deprivation for 24 hours lowered both Cav-1 mRNA and protein levels. This decrease was reversed by addition of LDL or medium containing 20% FBS (unpublished observation). Our experiments demonstrate that LDL-dependent regulation of Cav-1 can be due to enhanced Cav-1 translocation to membrane caveolae without de novo synthesis.
Cav-1 is postulated to organize formation of caveola microdomains and to regulate caveola-related signaling events, such as protein kinase C, Src-like kinases, Ras, and endothelial NO synthase.5 7 8 Because the structure and function of caveolae are sensitive to the amount of cholesterol in the membrane, there may be a direct link between the concentration of membrane cholesterol and the activation of the caveolin-interacting proteins, such as Ras. Depletion of cellular cholesterol lowered the cholesterol level of the caveola fractions. It also caused a reduction in amount of several key protein components of the MAPK complex, including Ras.25 Here, LDL increased the amount of cholesterol, Cav-1, and Ras in caveolae. This Ras translocation is probably associated with Cav-1 movement in ECs. These results support the notion that Cav-1 can function as a plasma membrane platform to localize caveolin-interacting signaling molecules, such as Ras, within caveola membranes.9 Recently, Roy et al26 reported that in hamster kidney cells, dominant-negative caveolin-3 completely blocked Raf activation mediated by H-Ras, and the inhibitory effect was reversed by replenishing cell membranes with cholesterol. Their results provide a link between the roles of caveolin in cholesterol trafficking and in Ras signal transduction. Consistent with their study, we found that elevating cellular cholesterol levels in human ECs could promote the translocation of Cav-1 and its binding signaling molecules, such as Ras, into caveolae, where cell signaling is triggered. Our results impart a pathophysiological significance to this link. Serum LDL has a direct correlation with the development of atherosclerosis, and overaccumulation of cellular cholesterol is common in cells within atherosclerotic lesions. Thus, cellular cholesterol levels and subsequent effects on caveolin trafficking may participate in the regulation of Ras and potentially other signaling molecules known to bind to caveolin.
We used a series of purified Cav-1 deletion mutants and [3H]-labeled free cholesterol to examine the cholesterol binding domains of Cav-1. The full-length Cav-1 showed the highest cholesterol-binding activity. Cholesterol also bound to the 2 Cav-1 mutants: a C-terminal domain peptide (135 to 178) and a scaffolding domain peptide (61 to 101). This result provides the first evidence showing that cholesterol binds to both the C-terminal lipid-binding sites and the scaffolding domain. These 2 domains also contribute to the membrane attachment of Cav-1.27 Thus, cholesterol may mediate Cav-1 membrane attachment via both domains. Furthermore, we observed that the C-terminal domain of Cav-1 also interacted with Ras. Addition of cholesterol increased the interaction of Ras with the scaffolding domain but not with the C-terminal component of Cav-1. It is known that the scaffolding domain of Cav-1 associates with inactivated Ras,9 but the function of the Cav-1 C-terminal domain for Ras needs further study. Nevertheless, because the scaffolding domain of Cav-1 binds signaling molecules, including Ras, and cholesterol can enhance Ras interaction with that part of Cav-1, interaction between these proteins and cholesterol, with the consequent signaling effects, will be an important issue in cholesterol-induced cell activation.
In view of our results, we postulate a hypothesis: In ECs, Cav-1 may function as a plasma membrane platform to localize caveolin-interacting signaling molecules, such as Ras, within caveola membranes. High concentrations of LDL elevate cellular cholesterol levels and promote the translocation of Cav-1 and its associated signaling molecules into caveolae, where cell signaling mediated by Ras is initiated. Ras activation, in turn, activates the JNK–c-Jun pathway.10 11
This study was supported in part by NIH grant HL-43023 (to M.B.S.) and American Heart Association, Western States Affiliate grant 98–252 (to Y.Z.). We thank M.P. Lisanti for providing the GST–Cav-1 plasmids. We also thank Dr J.A. Thompson for providing recombinant human fibroblast growth factor.
Fielding CJ, Fielding PE. Intracellular cholesterol transport. J Lipid Res.. 1997;38:1503–1521.
Schnitzer JE, Oh P, Jacobson BS, Dvorak AM. Caveolae from luminal plasmalemma of rat lung endothelium: microdomains enriched in caveolin, Ca(2+)-ATPase, and inositol trisphosphate receptor. Proc Natl Acad Sci U S A.. 1995;92:1759–1763.
Stolz DB, Jacobson BS. Macro- and microvascular endothelial cells in vitro: maintenance of biochemical heterogeneity despite loss of ultrastructural characteristics. In Vitro Cell Dev Biol. 1991;27A:169–182.
Murata M, Peranen J, Schreiner R, Wieland F, Kurzchalia TV, Simons K. VIP21/caveolin is a cholesterol-binding protein. Proc Natl Acad Sci U S A.. 1995;92:10339–10343.
Okamoto T, Schlegel A, Scherer PE, Lisanti MP. Caveolins, a family of scaffolding proteins for organizing “preassembled signaling complexes” at the plasma membrane. J Biol Chem.. 1998;273:5419–5422.
Garcia-Cardena G, Fan R, Stern DF, Liu J, Sessa WC. Endothelial nitric oxide synthase is regulated by tyrosine phosphorylation and interacts with caveolin-1. J Biol Chem.. 1996;271:27237–27240.
Song SK, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP. Co-purification and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains: detergent-free purification of caveolae microdomains. J Biol Chem.. 1996;271:9690–9697.
Zhu Y, Lin JH, Liao HL, Friedli OJ, Verna L, Marten NW, Straus DS, Stemerman MB. LDL induces transcription factor activator protein-1 in human endothelial cells. Arterioscler Thromb Vasc Biol. 1998;18:473–480.
Pritchard KA Jr, Schwarz SM, Medow MS, Stemerman MB. Effect of low-density lipoprotein on endothelial cell membrane fluidity and mononuclear cell attachment. Am J Physiol.. 1991;260:C43–C49.
Vasile E, Qu H, Dvorak HF, Dvorak AM. Caveolae and vesiculo-vacuolar organelles in bovine capillary endothelial cells cultured with VPF/VEGF on floating Matrigel-collagen gels. J Histochem Cytochem.. 1999;47:159–167.
Couet J, Sargiacomo M, Lisanti MP. Interaction of a receptor tyrosine kinase, EGF-R, with caveolins: caveolin binding negatively regulates tyrosine and serine/threonine kinase activities. J Biol Chem.. 1997;272:30429–30438.
Laufs U, Liao JK. Post-transcriptional regulation of endothelial nitric oxide synthase mRNA stability by Rho GTPase. J Biol Chem.. 1998;273:24266–24271.
Li YS, Shyy JY, Li S, Lee J, Su B, Karin M, Chien S. The Ras-JNK pathway is involved in shear-induced gene expression. Mol Cell Biol.. 1996;16:5947–5954.
Garcia-Cardena G, Martasek P, Masters BS, Skidd PM, Couet J, Li S, Lisanti MP, Sessa WC. Dissecting the interaction between nitric oxide synthase (NOS) and caveolin: functional significance of the NOS caveolin binding domain in vivo. J Biol Chem.. 1997;272:25437–25440.
Pritchard KA Jr, Tota RR, Lin JH, Danishefsky KJ, Kurilla BA, Holland JA, Stemerman MB. Native low density lipoprotein: endothelial cell recruitment of mononuclear cells. Arterioscler Thromb.. 1991;11:1175–1181.
Hailstones D, Sleer LS, Parton RG, Stanley KK. Regulation of caveolin and caveolae by cholesterol in MDCK cells. J Lipid Res.. 1998;39:369–379.
Ilangumaran S, Hoessli DC. Effects of cholesterol depletion by cyclodextrin on the sphingolipid microdomains of the plasma membrane. Biochem J.. 1998;335:433–440.
Fielding CJ, Bist A, Fielding PE. Caveolin mRNA levels are up-regulated by free cholesterol and down-regulated by oxysterols in fibroblast monolayers. Proc Natl Acad Sci U S A.. 1997;94:3753–3758.
Furuchi T, Anderson RG. Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-related kinase (ERK). J Biol Chem.. 1998;273:21099–21104.
Schlegel A, Schwab RB, Scherer PE, Lisanti MP. A role for the caveolin scaffolding domain in mediating the membrane attachment of caveolin-1: the caveolin scaffolding domain is both necessary and sufficient for membrane binding in vitro. J Biol Chem.. 1999;274:22660–22667.