Title: Co-Regulation of Transcellular and Paracellular Leak Across Microvascular Endothelium by Dynamin and Rac
Abstract: Increased permeability of the microvascular endothelium to fluids and proteins is the hallmark of inflammatory conditions such as sepsis. Leakage can occur between (paracellular) or through (transcytosis) endothelial cells, yet little is known about whether these pathways are linked. Understanding the regulation of microvascular permeability is essential for the identification of novel therapies to combat inflammation. We investigated whether transcytosis and paracellular leakage are co-regulated. Using molecular and pharmacologic approaches, we inhibited transcytosis of albumin in primary human microvascular endothelium and measured paracellular permeability. Blockade of transcytosis induced a rapid increase in paracellular leakage that was not explained by decreases in caveolin-1 or increases in activity of nitric oxide synthase. The effect required caveolin-1 but was observed in cells depleted of clathrin, indicating that it was not due to the general inhibition of endocytosis. Inhibiting transcytosis by dynamin blockade increased paracellular leakage concomitantly with the loss of cortical actin from the plasma membrane and the displacement of active Rac from the plasmalemma. Importantly, inhibition of paracellular leakage by sphingosine-1-phosphate, which activates Rac and induces cortical actin, caused a significant increase in transcytosis of albumin in vitro and in an ex vivo whole-lung model. In addition, dominant-negative Rac significantly diminished albumin uptake by endothelia. Our findings indicate that transcytosis and paracellular permeability are co-regulated through a signaling pathway linking dynamin, Rac, and actin. Increased permeability of the microvascular endothelium to fluids and proteins is the hallmark of inflammatory conditions such as sepsis. Leakage can occur between (paracellular) or through (transcytosis) endothelial cells, yet little is known about whether these pathways are linked. Understanding the regulation of microvascular permeability is essential for the identification of novel therapies to combat inflammation. We investigated whether transcytosis and paracellular leakage are co-regulated. Using molecular and pharmacologic approaches, we inhibited transcytosis of albumin in primary human microvascular endothelium and measured paracellular permeability. Blockade of transcytosis induced a rapid increase in paracellular leakage that was not explained by decreases in caveolin-1 or increases in activity of nitric oxide synthase. The effect required caveolin-1 but was observed in cells depleted of clathrin, indicating that it was not due to the general inhibition of endocytosis. Inhibiting transcytosis by dynamin blockade increased paracellular leakage concomitantly with the loss of cortical actin from the plasma membrane and the displacement of active Rac from the plasmalemma. Importantly, inhibition of paracellular leakage by sphingosine-1-phosphate, which activates Rac and induces cortical actin, caused a significant increase in transcytosis of albumin in vitro and in an ex vivo whole-lung model. In addition, dominant-negative Rac significantly diminished albumin uptake by endothelia. Our findings indicate that transcytosis and paracellular permeability are co-regulated through a signaling pathway linking dynamin, Rac, and actin. One of the critical functions of the endothelium is the maintenance of an intact barrier between the vascular lumen and the interstitium. Numerous inflammatory diseases, including acute lung injury, sepsis, and anaphylaxis, are characterized by prolonged and excessive microvascular leak.1Camerer E. Regard J.B. Cornelissen I. Srinivasan Y. Duong D.N. Palmer D. Pham T.H. Wong J.S. Pappu R. Coughlin S.R. Sphingosine-1-phosphate in the plasma compartment regulates basal and inflammation-induced vascular leak in mice.J Clin Invest. 2009; 119: 1871-1879PubMed Google Scholar, 2Kerschen E.J. Fernandez J.A. Cooley B.C. Yang X.V. Sood R. Mosnier L.O. Castellino F.J. Mackman N. Griffin J.H. Weiler H. Endotoxemia and sepsis mortality reduction by non-anticoagulant activated protein C.J Exp Med. 2007; 204: 2439-2448Crossref PubMed Scopus (255) Google Scholar, 3Lee W.L. Slutsky A.S. Sepsis and endothelial permeability.N Engl J Med. 2010; 363: 689-691Crossref PubMed Scopus (362) Google Scholar In such settings, the exodus of proteins and fluids has deleterious consequences,4Boyd J.H. Forbes J. Nakada T.A. Walley K.R. Russell J.A. Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure are associated with increased mortality.Crit Care Med. 2011; 39: 259-265Crossref PubMed Scopus (1018) Google Scholar, 5Ye X. Ding J. Zhou X. Chen G. Liu S.F. Divergent roles of endothelial NF-kappaB in multiple organ injury and bacterial clearance in mouse models of sepsis.J Exp Med. 2008; 205: 1303-1315Crossref PubMed Scopus (165) Google Scholar including the loss of serum albumin and oncotic pressure6Fleck A. Raines G. Hawker F. Trotter J. Wallace P.I. Ledingham I.M. Calman K.C. Increased vascular permeability: a major cause of hypoalbuminaemia in disease and injury.Lancet. 1985; 1: 781-784Abstract PubMed Scopus (645) Google Scholar and culminating in the development of hypotension and tissue edema. There is great interest, therefore, in developing novel agents that decrease microvascular leak. However, for this to be successful, understanding the mechanisms and regulation of microvascular permeability is essential.3Lee W.L. Slutsky A.S. Sepsis and endothelial permeability.N Engl J Med. 2010; 363: 689-691Crossref PubMed Scopus (362) Google Scholar Inflammatory mediators can induce vascular leak by opening gaps in endothelial cells, a process known as paracellular leak.7Majno G. Palade G.E. Studies on inflammation, 1: the effect of histamine and serotonin on vascular permeability: an electron microscopic study.J Biophys Biochem Cytol. 1961; 11: 571-605Crossref PubMed Google Scholar, 8Wu N.Z. Baldwin A.L. Transient venular permeability increase and endothelial gap formation induced by histamine.Am J Physiol. 1992; 262: H1238-H1247PubMed Google Scholar, 9Mehta D. Malik A.B. Signaling mechanisms regulating endothelial permeability.Physiol Rev. 2006; 86: 279-367Crossref PubMed Scopus (1338) Google Scholar Although the mechanisms of gap formation depend in part on the inflammatory stimulus, they can be broadly divided into three overlapping categories: i) injury and/or apoptosis of cells in the endothelial monolayer,10Menzies B.E. Kourteva I. Staphylococcus aureus alpha-toxin induces apoptosis in endothelial cells.FEMS Immunol Med Microbiol. 2000; 29: 39-45PubMed Google Scholar, 11Xu J. Zhang X. Pelayo R. Monestier M. Ammollo C.T. Semeraro F. Taylor F.B. Esmon N.L. Lupu F. Esmon C.T. Extracellular histones are major mediators of death in sepsis.Nat Med. 2009; 15: 1318-1321Crossref PubMed Scopus (1060) Google Scholar ii) internalization or disassembly of intercellular junctions,12Andriopoulou P. Navarro P. Zanetti A. Lampugnani M.G. Dejana E. Histamine induces tyrosine phosphorylation of endothelial cell-to-cell adherens junctions.Arterioscler Thromb Vasc Biol. 1999; 19: 2286-2297Crossref PubMed Scopus (202) Google Scholar, 13Gavard J. Gutkind J.S. VEGF controls endothelial-cell permeability by promoting the beta-arrestin-dependent endocytosis of VE-cadherin.Nat Cell Biol. 2006; 8: 1223-1234Crossref PubMed Scopus (784) Google Scholar and iii) remodeling of the actin cytoskeleton, leading to a change in cell shape. Cytoskeletal remodeling is regulated by members of the Rho GTPase family, which recruit downstream effectors when bound to GTP. For instance, the formation of the cortical actin fibers that normally buttress the plasmalemma is dependent on Rac,14Maruo N. Morita I. Shirao M. Murota S. IL-6 increases endothelial permeability in vitro.Endocrinology. 1992; 131: 710-714Crossref PubMed Google Scholar, 15Whitehead K.J. Chan A.C. Navankasattusas S. Koh W. London N.R. Ling J. Mayo A.H. Drakos S.G. Jones C.A. Zhu W. Marchuk D.A. Davis G.E. Li D.Y. The cerebral cavernous malformation signaling pathway promotes vascular integrity via Rho GTPases.Nat Med. 2009; 15: 177-184Crossref PubMed Scopus (283) Google Scholar whereas the generation and contraction of cytoplasmic actinomyosin stress fibers require RhoA.16Goldblum S.E. Ding X. Brann T.W. Campbell-Washington J. Bacterial lipopolysaccharide induces actin reorganization, intercellular gap formation, and endothelial barrier dysfunction in pulmonary vascular endothelial cells: concurrent F-actin depolymerization and new actin synthesis.J Cell Physiol. 1993; 157: 13-23Crossref PubMed Scopus (93) Google Scholar, 17Birukova A.A. Adyshev D. Gorshkov B. Bokoch G.M. Birukov K.G. Verin A.D. GEF-H1 is involved in agonist-induced human pulmonary endothelial barrier dysfunction.Am J Physiol Lung Cell Mol Physiol. 2006; 290: L540-L548Crossref PubMed Scopus (141) Google Scholar Although paracellular leak is more widely studied, leakage of plasma and proteins can also occur through the cytoplasm of endothelial cells by transcytosis.18Tuma P.L. Hubbard A.L. Transcytosis: crossing cellular barriers.Physiol Rev. 2003; 83: 871-932Crossref PubMed Scopus (507) Google Scholar, 19Predescu S.A. Predescu D.N. Malik A.B. Molecular determinants of endothelial transcytosis and their role in endothelial permeability.Am J Physiol Lung Cell Mol Physiol. 2007; 293: L823-L842Crossref PubMed Scopus (153) Google Scholar In fact, transcytosis is a physiologically important process that accounts for much of albumin transit across the resting endothelium.20Schubert W. Frank P.G. Razani B. Park D.S. Chow C.W. Lisanti M.P. Caveolae-deficient endothelial cells show defects in the uptake and transport of albumin in vivo.J Biol Chem. 2001; 276: 48619-48622Crossref PubMed Scopus (272) Google Scholar This involves the traffic of caveolae, small intracellular vesicles that are often docked to the plasmalemma, and begins with binding of albumin by multiple receptors on the apical endothelial surface.19Predescu S.A. Predescu D.N. Malik A.B. Molecular determinants of endothelial transcytosis and their role in endothelial permeability.Am J Physiol Lung Cell Mol Physiol. 2007; 293: L823-L842Crossref PubMed Scopus (153) Google Scholar, 21Tiruppathi C. Song W. Bergenfeldt M. Sass P. Malik A.B. Gp60 Activation mediates albumin transcytosis in endothelial cells by tyrosine kinase-dependent pathway.J Biol Chem. 1997; 272: 25968-25975Crossref PubMed Scopus (299) Google Scholar, 22Schnitzer J.E. Carley W.W. Palade G.E. Specific albumin binding to microvascular endothelium in culture.Am J Physiol Heart Circ Physiol. 1988; 254: H425-H437Crossref PubMed Google Scholar This leads to the invagination of caveolae from the plasma membrane. The GTPase dynamin-223Nabi I.R. Le P.U. Caveolae/raft-dependent endocytosis.J Cell Biol. 2003; 161: 673-677Crossref PubMed Scopus (599) Google Scholar forms a ring around the neck of the invaginating vesicles and mediates their scission from the plasmalemma.24Oh P. McIntosh D.P. Schnitzer J.E. Dynamin at the neck of caveolae mediates their budding to form transport vesicles by GTP-driven fission from the plasma membrane of endothelium.J Cell Biol. 1998; 141: 101-114Crossref PubMed Scopus (555) Google Scholar The vesicles then move into the cell, fuse with the basolateral membrane, and release albumin to the interstitial space. Accordingly, caveolae contain SNARE proteins and the hexameric ATPase N-ethylmaleimide (NEM)–sensitive factor, both of which are required for the fusion of vesicles to the plasmalemma.19Predescu S.A. Predescu D.N. Malik A.B. Molecular determinants of endothelial transcytosis and their role in endothelial permeability.Am J Physiol Lung Cell Mol Physiol. 2007; 293: L823-L842Crossref PubMed Scopus (153) Google Scholar In contrast to paracellular leak, nothing is known about the relationship between Rac and RhoA and transcytosis. In addition to albumin, transcytosis occurs for immunoglobulins, transferrin, aminopeptidase P,25Oh P. Borgstrom P. Witkiewicz H. Li Y. Borgstrom B.J. Chrastina A. Iwata K. Zinn K.R. Baldwin R. Testa J.E. Schnitzer J.E. Live dynamic imaging of caveolae pumping targeted antibody rapidly and specifically across endothelium in the lung.Nature Biotechnol. 2007; 25: 327-337Crossref PubMed Scopus (259) Google Scholar and numerous other molecules.18Tuma P.L. Hubbard A.L. Transcytosis: crossing cellular barriers.Physiol Rev. 2003; 83: 871-932Crossref PubMed Scopus (507) Google Scholar Although both paracellular and transcellular leak were described decades ago, remarkably little is known about the relationship between them. Endothelial cells from mice deficient in caveolin-1 lack caveolae and do not perform transcytosis.26Schubert W. Frank P.G. Woodman S.E. Hyogo H. Cohen D.E. Chow C.W. Lisanti M.P. Microvascular hyperpermeability in caveolin-1 (-/-) knock-out mice: treatment with a specific nitric-oxide synthase inhibitor, l-NAME, restores normal microvascular permeability in Cav-1 null mice.J Biol Chem. 2002; 277: 40091-40098Crossref PubMed Scopus (276) Google Scholar The vasculature from these animals demonstrates an increase in permeability, although the mechanism of this change is controversial. Although some have described a compensatory induction of paracellular leak,26Schubert W. Frank P.G. Woodman S.E. Hyogo H. Cohen D.E. Chow C.W. Lisanti M.P. Microvascular hyperpermeability in caveolin-1 (-/-) knock-out mice: treatment with a specific nitric-oxide synthase inhibitor, l-NAME, restores normal microvascular permeability in Cav-1 null mice.J Biol Chem. 2002; 277: 40091-40098Crossref PubMed Scopus (276) Google Scholar others attribute the effect to a change in capillary pressure or the endothelial glycocalyx.27Rosengren B.-I. Rippe A. Rippe C. Venturoli D. Sward K. Rippe B. Transvascular protein transport in mice lacking endothelial caveolae.Am J Physiol Heart Circ Physiol. 2006; 291: H1371-H1377Crossref PubMed Scopus (53) Google Scholar Furthermore, caveolin-1 has numerous functions in cell signaling and growth,28Quest A.F. Gutierrez-Pajares J.L. Torres V.A. Caveolin-1: an ambiguous partner in cell signalling and cancer.J Cell Mol Med. 2008; 12: 1130-1150Crossref PubMed Scopus (139) Google Scholar making it impossible to know whether the increase in permeability is due to the loss of transcytosis per se or to the loss of caveolin-1. Thus, how paracellular and transcellular leak are related remains largely unexplored. We hypothesized that, independent of caveolin-1, inhibition of transcytosis would lead to a compensatory increase in paracellular leak. To address this issue, we first devised methods of measuring transcytosis of albumin by primary human microvascular endothelium in culture. We then inhibited transcytosis by multiple distinct molecular and pharmacologic approaches and determined the effect on paracellular leak. Reciprocally, we inhibited paracellular leak and monitored for changes in transcytosis both in vitro and using an ex vivo lung model. Lastly, we determined the molecular mechanism for this relationship. Primary human microvascular endothelial cells of dermal origin were obtained from discarded human foreskin as previously described29Gillrie M.R. Krishnegowda G. Lee K. Buret A.G. Robbins S.M. Looareesuwan S. Gowda D.C. Ho M. Src-family kinase dependent disruption of endothelial barrier function by Plasmodium falciparum merozoite proteins.Blood. 2007; 110: 3426-3435Crossref PubMed Scopus (54) Google Scholar and were used throughout the study. Primary human microvascular endothelial cells of pulmonary origin were obtained from Lonza (Basel, Switzerland) and were used to confirm the results of key transwell experiments, as indicated below. Cells were cultured in EBM-2 media with the recommended supplements (Lonza) and were used in passages 3 to 8. Endothelial cells were seeded on gelatin-coated 0.4-μm-pore polyester transwells (Costar, Corning, NY) and grown until confluency 1 to 3 days later. Only monolayers that appeared healthy by phase-contrast microscopy and had a transendothelial electrical resistance (TEER; measured using Endohm-12 from World Precision Instruments, Sarasota, FL) of >19 Ωcm2 were used for permeability experiments. Replacement of the TEER machine led to higher TEER values in some experiments, but permeability to dextran remained unchanged. To measure transcytosis, endothelial monolayers were allowed to bind but not internalize fluorescein isothiocyanate (FITC)–conjugated albumin (Invitrogen, Eugene, OR; 0.4 mg/mL) for 10 minutes at 4°C. Unbound albumin was removed by washing the cells with cold media twice. Transcytosis was allowed to proceed by incubating the cells at 37°C for 2 hours. As controls, 80 μmol/L Dynasore (Sigma-Aldrich, St Louis, MO) or 4 mmol/L histamine (Sigma-Aldrich) were added for the incubations at 37°C. Aliquots of the lower and upper chambers were then taken and measured for fluorescence in a Spectramax Gemini EM (Molecular Devices, Sunnyvale, CA) using the appropriate filter sets and after subtraction of appropriate blanks. To measure transcytosis by a second method, serum-free media containing 0.3 mg/mL of biotin-tagged albumin (biotin-BSA; Thermo Scientific, Rockford, IL) and 50 μg/mL of FITC-conjugated dextran (Invitrogen; molecular weight, 70 kDa) were added to the upper chamber of the transwell, whereas serum-free media alone was added to the lower chamber. After 4 hours of incubation, aliquots from the lower chamber were taken for measurement of biotin-BSA by enzyme-linked immunosorbent assay (ELISA) (see below). As a control for paracellular leak, aliquots were also taken for measurement of FITC-dextran fluorescence intensity in a Spectramax Gemini EM (Molecular Devices) using the appropriate filter sets and after subtraction of appropriate blanks. For the ELISA, 100 μL of media from the lower chamber and 50 μL of 0.5 μg/mL of horseradish peroxidase–conjugated sheep anti-BSA (Immunology Consultants Laboratory, Inc., Newberg, OR) were incubated in 96-well plates coated with streptavidin (Thermo Scientific) at room temperature for 2 hours with shaking. After washing with buffer (25 mmol/L Tris, pH 7.2, 150 mmol/L NaCl, 0.05% Tween-20, 0.1% gelatin), 50 μL of 3,3′,5,5′-tetramentylbenzidine (Thermo Scientific) was added to the well and incubated at room temperature for 15 minutes before adding 50 μL of 2 mol/L sulfuric acid. Quantification of BSA was determined by measuring absorbance at 450 nm with a ThermoMax Microplate Reader (Molecular Devices) after verification with a standard curve using known amounts of biotin-BSA. We measured paracellular permeability using two methods and in both dermal and pulmonary microvascular endothelium. After measurement of the baseline TEER, cells were exposed to 0.1 mmol/L NEM (Sigma) for 2 minutes followed by repeat measurement. Permeability of NEM-treated monolayers was also measured by the addition of FITC-dextran (50 μg/mL; molecular weight, 70 kDa) to the upper chamber of the transwell with removal of aliquots of media from both the upper and lower chambers 40 minutes later. Using appropriate filter sets, we measured the fluorescence in aliquots of the lower and upper chambers after subtraction of the appropriate blanks. Permeability was calculated as P = [(ΔCb/Δt) × Vb]/(Ca × A), where ΔCb/Δt is the change in the basolateral concentration during the duration of the assay, Vb is the volume of the basolateral chamber, Ca is the apical concentration, and A is the area of the transwell.30DeMaio L. Tarbell J.M. Scaduto Jr., R.C. Gardner T.W. Antonetti D.A. A transmural pressure gradient induces mechanical and biological adaptive responses in endothelial cells.Am J Physiol Heart Circ Physiol. 2004; 286: H731-H741Crossref PubMed Scopus (47) Google Scholar Changes in Vb and Ca were negligible during the experiment, and Cb was much lower than Ca. In other experiments, monolayers were incubated with 80 μmol/L Dynasore [or dimethyl sulfoxide (DMSO) as a solvent control] for 40 minutes followed by repeat measurement of the TEER. Other cells were incubated with Dynasore or DMSO, and 50 μg/mL of FITC-dextran was added simultaneously to the upper chamber of the transwell. Forty minutes later, flux of dextran into the lower chamber was calculated as explained above. To determine the effect of cooling to <15°C, cells were incubated at 14°C or 37°C for 40 minutes followed by repeat measurement of TEER. Alternatively, FITC-dextran was added to the upper chamber immediately before incubation at 14°C or 37°C and flux was calculated 40 minutes later, as described above. In some experiments, cells were pretreated with the nitric oxide synthase inhibitor nitroglycerine-nitro-L-arginine methyl ester (l-NAME) (Sigma; 100 μmol/L) for 30 minutes before inhibition of transcytosis. Permeability to dextran was also tested under 10 cm H2O of pressure. A column was added to the upper chamber of the transwells and was filled with 10 cm of media. Transwells were acclimatized to the new pressure during 5 hours,30DeMaio L. Tarbell J.M. Scaduto Jr., R.C. Gardner T.W. Antonetti D.A. A transmural pressure gradient induces mechanical and biological adaptive responses in endothelial cells.Am J Physiol Heart Circ Physiol. 2004; 286: H731-H741Crossref PubMed Scopus (47) Google Scholar then the media in the column was replaced with media (also 10 cm in height) containing dextran and Dynasore or DMSO, and dextran flux was measured as described above with 80 μmol/L Dynasore or DMSO control. Transcytosis experiments were performed as described above, using gold-tagged albumin (6 nm, Electron Microscopy Sciences, Hatfield, PA). Fifteen minutes after incubation at 37°C, cells on their transwells were fixed in 2% glutaraldehyde, after which the cells were washed four times with 0.1 mol/L sodium cacodylate buffer. After the wash, the membranes were cut out of the transwell inserts. The cells were then fixed for 45 minutes in 1% osmium tetroxide and 1% potassium ferrocyanide in cacodylate buffer. The samples were washed three times in buffer followed by twice in distilled water. The cells were stained for 30 minutes in 4% aqueous uranyl acetate then washed twice in distilled water. They were dehydrated through a graded series of ethanol and infiltrated and embedded in Epon (JEMBED 812). Sections were cut at 80 nm, picked up on copper grids, and stained with uranyl acetate and lead citrate. Sections were examined using an FEI Tecnai 20 transmission electron microscope. Images were recorded with an AMT 16000 digital camera. Lysates were analyzed by SDS-PAGE using 6% to 12% polyacrylamide gels. Proteins were transferred to nitrocellulose membranes, blocked for 1 hour, and probed overnight with the primary antibody at 4°C. After washing in PBS with Tween 20, blots were incubated with horseradish peroxidase–conjugated secondary antibodies for 1 hour, washed, and then visualized by enhanced chemiluminescence (Amersham). For quantitation, blots were scanned using Image J (National Institutes of Health), and the integrated intensity was normalized to the amount of protein loaded after correction for background. Analysis of the distribution and density of caveolin-1, F-actin, vascular endothelial cadherin (VE-cadherin), claudin-5, heparan sulfate, and internalized albumin was performed using spinning disk confocal microscopy (Zeiss Axiovert 200M microscope with 25× or 63× objectives; Zeiss, Jena, Germany). The unit is equipped with diode-pumped, solid-state laser lines (440, 491, 561, 638, and 655 nm; Spectral Applied Research, Richmond Hill, ON), motorized XY stage (ASI), and a Zeiss focus drive. Images were acquired using back-thinned, electron-multiplied or conventional cooled, charge-couple device cameras (Hamamatsu, Shizuoka, Japan), driven by Volocity 4.1.1 software (Improvision, Coventry, England). In all cases, microscope settings were kept constant between conditions. All images were randomly chosen and were acquired as Z-stack projections (Z-interval, 0.2 to 0.5 μm). Total internal reflectance fluorescence (TIRF) images were acquired on an Olympus cell∧TIRF Motorized Multicolor TIRF module mounted on an Olympus IX81 microscope (Olympus, Hamburg, Germany). Samples were imaged using a 63×/1.49 objective with 491-nm excitation lasers and MetaMorph acquisition software. Retention of the yellow fluorescent protein–tagged p21-binding domain (PBD-YFP) was determined by the ratio of lamellipodia or filopodia that still exhibited the PBD-YFP at 30 minutes compared with time 0. To immunostain for F-actin, VE-cadherin, and heparan sulfate, cells were rinsed in PBS, fixed for 30 to 60 minutes in 4% paraformaldehyde in PBS, incubated in 0.15% glycine for 10 minutes, permeabilized in 0.1% Triton X-100 for 20 minutes, and blocked for 1 hour. Cells were stained with Alexa Fluor 488 phalloidin (Molecular Probes), anti–VE-cadherin (Santa Cruz Biotechnology, Santa Cruz, CA), or anti–heparan sulfate (US Biological, Marblehead, MA) for 1 hour. For caveolin-1 and claudin-5, cells were fixed in methanol then put in blocking buffer for 1 hour at room temperature. The primary antibodies for caveolin-1 (Santa Cruz Biotechnology) and claudin-5 (Abcam, Cambridge, MA) were added for 1 hour at room temperature. Heparinase III treatment (0.2 U/mL for 1 hour) was used as a negative control for heparan sulfate. In some experiments, the internalization of membrane-bound albumin was compared under different conditions. Briefly, fluorophore-tagged albumin was added to cells in cold media and spun down at 16 × g for 10 seconds; then the cells were incubated at 4°C for 10 minutes. Afterward, cells were rinsed with cold media and incubated at 37°C for 15 minutes. Cells were then fixed and processed as above. As a control for endocytosis, a lipophilic styryl dye, N-(3-triethylammoniumpropyl)−4-(p-diethylaminophenylhexatrienyl) pyridinium dibromide (FM 4-64, Invitrogen),31Vida T.A. Emr S.D. A new vital stain for visualizing vacuolar membrane dynamics and endocytosis in yeast.J Cell Biol. 1995; 128: 779-792Crossref PubMed Scopus (1136) Google Scholar was added to endothelial cells at a concentration of 5 μg/mL for 1 hour. Cells were then rinsed, fixed, and processed as above. Subcellular fractionation was performed as previously described.32London N.R. Zhu W. Bozza F.A. Smith M.C. Greif D.M. Sorensen L.K. Chen L. Kaminoh Y. Chan A.C. Passi S.F. Day C.W. Barnard D.L. Zimmerman G.A. Krasnow M.A. Li D.Y. Targeting Robo4-dependent Slit signaling to survive the cytokine storm in sepsis and influenza.Sci Transl Med. 2010; 2: 23ra19Crossref PubMed Scopus (258) Google Scholar Briefly, dermal endothelial cells were grown to confluency in 10 cm dishes. Cells were rinsed with PBS then scraped into PBS and centrifuged at 400 × g for 5 minutes to pellet the cells. The pellet was resuspended in hypotonic lysis buffer buffer containing 10 mmol/L Tris-HCl, 5 mmol/L KCl, 1 mmol/L MgCl2, protease inhibitor, and 1 mmol/L DTT. Cells were further lysed using a Dounce Homogenizer (Wheaton, Millville, NJ) with 20 strokes. The mixture was centrifuged at 400 × g for 5 minutes to pellet cell debris, and the supernatant was collected and subjected to repeat centrifugation at 16,000 × g for 30 minutes. The pellet was resuspended in radioimmunoprecipitation assay buffer (50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1% NP-40, 0.5% deoxycholic acid, and 0.1% SDS) and incubated on ice for 30 minutes before analysis with Western blotting. Preparations were probed for the transferrin receptor as a positive control. Dermal endothelial cells were grown to confluency in 6-cm dishes. After washing with PBS, cells were scraped and precipitated by centrifugation. Cells were lysed in 140 μL of lysis buffer containing 25 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, 10% sucrose, 1% Triton X-100, 1 mmol/L dithiothreitol, and protease inhibitor cocktail (R) for 30 minutes on ice. The lysate was then added to 0.66 mL of 60% OptiPrep (Axis-Shield, Oslo, Norway). To ensure complete lysis, cells were homogenized in an ice-cold Dounce Homogenizer (Wheaton) with 10 strokes. Cell lysates were then placed in an ultracentrifuge and overlaid with 0.8 mL of 40%, 30%, and 20% OptiPrep. The gradients were centrifuged at 28,000 × g at 4°C for 15 hours using an SW 55 Ti Rotor. Fractions were collected and analyzed by Western blot. Mouse anti–transferrin receptor antibody was from Invitrogen. Data were quantitated using ImageJ. After subtraction of background, the amount of caveolin-1 in the 20% fraction was divided by the total amount of caveolin-1.33Singleton P.A. Salgia R. Moreno-Vinasco L. Moitra J. Sammani S. Mirzapoiazova T. Garcia J.G. CD44 regulates hepatocyte growth factor-mediated vascular integrity: role of c-Met, Tiam1/Rac1, dynamin 2, and cortactin.J Biol Chem. 2007; 282: 30643-30657Crossref PubMed Scopus (104) Google Scholar Endothelial cell death was detected using trypan blue by enumerating the number of stained cells in 10 random fields (×20 magnification). Apoptosis of endothelial cells was determined by staining with Annexin V-FITC (EBioscience, San Diego, CA) as per the manufacturer's instructions and analyzed by flow cytometry using a BD FACS Calibur cytometer (Becton Dickinson) and then imported into FlowJo 8.8.6 (TreeStar Inc., Ashland, OR). For most experiments, primary human endothelial cells grown on coverslips or transwells were transfected with plasmids using Fugene HD (Roche) in accordance with the manufacturer's instructions and were used for experiments 24 hours later. For transfection of Rac constructs for cells on transwells, we used the Electro Square Porator according to the manufacturer's protocol (protocol 0394). The plasmids containing wild-type (WT) Rac, dominant-negative (DN) (T17N) Rac, dynamin-2 WT, K44A, and PBD-YFP were gifts from Sergio Grinstein and have been previously described.34Jin J.B. B