Inflammation-induced subcellular redistribution of VE-cadherin, actin, and gamma-catenin in cultured human lung microvessel endothelial cells.

The inflammation-induced subcellular redistribution of key cytoskeletal and junctional proteins in cultured human lung microvessel endothelial cells is investigated as part of a study on the posttranslational regulation of paracellular permeability. Inflammatory agonist-stimulated cells are detergent fractionated into three subcellular compartments followed by quantitative immunoblot analysis. Actin, gamma-catenin, and VE-cadherin increasingly associate with the cytoskeletal fraction upon thrombin stimulation. Concomitantly, actin is reduced in the cytosol fraction, whereas gamma-catenin and VE-cadherin are reduced in the membrane fraction. alpha- and beta-catenin show baseline distributions similar to those of VE-cadherin and gamma-catenin, but do not significantly redistribute. Additionally, vimentin is found exclusively in the cytoskeletal fraction and also does not significantly redistribute following thrombin treatment. The VE-cadherin response is independent of the presence of F-actin or actin redistribution. Immunofluorescence microscopy reveals that membrane and cytoskeletal VE-cadherin is present in alternating patches along the cell junctions. Furthermore, VE-cadherin is lost from zones of interendothelial cell pore formation. A model is formulated describing these membrane-associated VE-cadherin patches as predetermined zones of potential intercellular gap formation. During inflammation, VE-cadherin is lost from these zones and sequestered at the remaining cell-cell contact sites, anchored to the cytoskeleton in an actin-independent fashion.

Changes in the biomechanical properties of neutrophils and endothelial cells during adhesion.

This study examined changes in the biomechanical properties of cultured pulmonary microvascular endothelial cells (ECs) and neutrophils induced by adhesion of neutrophils to these ECs. The biomechanical properties of cells were evaluated using magnetic twisting cytometry, which measures the angular rotation of ferromagnetic beads bound to cells through antibody ligation on application of a specified magnetic torque. Adhesion of neutrophils to 24-hour tumor necrosis factor-alpha (TNF-alpha)-treated ECs, but not to untreated ECs, induced an increase in EC stiffness within 2 minutes, which was accompanied by an increase and a reorganization of F-actin in ECs. A cell-permeant, phosphoinositide-binding peptide attenuated the EC stiffening response, suggesting that intracellular phosphoinositides are required. The stiffening response was not inhibited by ML-7, a myosin light-chain kinase inhibitor, or BAPTA, an intracellular Ca2+ chelator. Moreover, the phosphorylation pattern of the regulatory myosin light chains was unaltered within 15 minutes of neutrophil adherence. These data suggested that the EC stiffening response appeared not to be mediated by myosin light-chain-dependent mechanisms. Concomitantly, neutrophil adhesion to 24-hour TNF-alpha-treated ECs also induced changes in the biomechanical properties of neutrophils compared to neutrophils bound to untreated ECs. Taken together, these results demonstrated that neutrophil adhesion to TNF-alpha-treated ECs induces changes in the biomechanical properties of both cell types through actin cytoskeletal remodeling. These changes may modulate neutrophil transmigration across the endothelium during inflammation.

NFkappaB translocation in human microvessel endothelial cells using a four-compartment subcellular protein redistribution assay.

Protein distribution profiles may be used to characterize both physiological and pathophysiological cellular changes, but rigorous biochemical assays for measuring such movements are lacking. This paper reports on a protein redistribution assay that combines reversible metal chelate-based total protein detection with a four-fraction subcellular detergent fractionation procedure. TNF-alpha stimulated cultured human omental microvessel endothelial cells are fractionated into cytosol, membrane/organelle, nuclear (envelope and associated), and cytoskeletal/DNA compartments. Protein fractions are separated electrophoretically and electroblotted or slot-blotted onto PVDF membranes without electrophoretic separation. A key feature is that total protein is measured and analyzed directly on the resultant PVDF membrane, using a Ferrozine/ferrous metal-chelate stain, without the added step of a prior solution-phase protein assay. As a result, factors that may adversely affect NFkappaB quantification, such as saturation of the solid-support membrane, are rigorously evaluated and controlled. Following removal of the Ferrozine/ferrous total protein stain, NFkappaB distribution is determined via standard immunodetection procedures. This assay reveals a new level of complexity regarding NFkappaB distribution and translocation. NFkappaB is shown to translocate from the cytosol to the membrane/organelle and cytoskeletal/DNA fractions, whereas trace levels of NFkappaB are observed in the nuclear (envelope and associated) fraction. Dose-curve analysis reveals that the response is initiated at 10 U/ml of TNF-alpha, plateaus at approximately 1000 U/ml, and remains essentially constant up to 2000 U/ml. Time-course analysis demonstrates a measurable response as early as 5 min and a peak response at approximately 30 min, after which the distribution begins to return to baseline. The assay should provide a valuable tool for rapid evaluation and mechanistic studies of NFkappaB redistribution.

Comparative effects of garlic and aspirin on diabetic cardiovascular complications.

By using streptozotocin-induced diabetic rats as a studied model, our previous experimental results have indicated that daily oral feeding of garlic extract (100 mg/kg BW) could increase the cardiovascular functions in streptozotocin (STZ) rats; the abnormality of lipid profile was prevented; and garlic extract could increase fibrinolitic activities with the decrease of platelet aggregation. Moreover, the plasma insulin level was increased concomitantly with the decrease of plasma glucose level. However, due to the high incidence of atherosclerosis in diabetes, the present study has been continued for further investigation of the effect of garlic extract on the coronary vascular ultrastructural changes. In addition, to identify the possible mechanism(s) of garlic's therapeutic effects, the cyclooxygenase inhibitor, aspirin, has been included in this present study. By using transmission electron microscopic studies, 16 weeks of daily oral feeding of garlic extract (100 mg/kg BW) caused as an antiatherosclerotic agent at the coronary arteriolar (15-30 microns) wall in STZ-rats. Interestingly, the thickening of coronary capillary (5-10 microns) basement membrane also was significantly attenuated within the group of STZ-rats treated with garlic extract. However, the possible direct action of garlic through the cyclooxygenase pathway has not been confirmed by the results of aspirin: The daily oral feeding of aspirin (10 mg/kg BW) in 16-week STZ-rats has not showed reduced arteriolar vascular wall abnormalities. The irregular patterns of fiber matrix, arranging the basement membrane at the arteriolar walls, were still recognized in the same manner as in STZ-rats. Interestingly, the thickening of the capillary basement membrane occurred in 16-week STZ-rats seems to be attenuated by the aspirin received. At present, garlic extract may open the new era in the medicinal use of garlic to prevent diabetic cardiovascular complications.

Soluble P-selectin moderates complement-dependent reperfusion injury of ischemic skeletal muscle.

P-selectin is an adhesion molecule expressed on activated endothelial and platelet surfaces. The function of the short consensus repeats (SCRs) of P-selectin, homologous with the SCRs of complement regulatory proteins is largely unknown. In a model of murine hindlimb ischemia where local reperfusion injury is partly mediated by IgM natural antibody and classical complement pathway activation, we hypothesized that human soluble P-selectin (sP-sel) would moderate the complement component of the inflammatory response. Infusion of sP-sel supernatant or purified (p) sP-sel prepared from activated human platelets, reduced ischemic muscle vascular permeability by 48% and 43%, respectively, following reperfusion. Hindlimb immunohistochemistry demonstrated negligible C3 staining colocalized with IgM in these groups compared with intense staining in the untreated injured mice. In vitro studies of mouse serum complement hemolytic activity showed that psP-sel inhibited the classical but not alternative complement pathway. Flow cytometry demonstrated that psP-sel inhibited C1q adherence to sensitized red blood cells. From these data we conclude that sP-sel moderates skeletal muscle reperfusion injury by inhibition of the classical complement pathway.

Myosin translocation in retinal pericytes during free-radical induced apoptosis.

Vascular pathologies induced by ischemia/reperfusion involve the production of reactive oxygen species (ROS) that in part cause tissue injury. The production of ROS that occurs upon reperfusion activates specific second messenger pathways. In diabetic retinopathy there is a characteristic loss of the microvascular pericyte. Pericytes are more sensitive than endothelial cells to low concentrations of ROS, such as hydrogen peroxide (H(2)O(2)) when tested in vitro. Whether the pericyte loss is due to toxic cell death triggered by the noxious H(2)O(2) or apoptosis, due to activation of specific second messenger pathways, is unknown. During apoptosis, a cell's nucleus and cytoplasm condense, the cell becomes fragmented, and ultimately forms apoptotic bodies. It is generally assumed that apoptosis depends on nuclear signaling, but cytoplasmic morphological processes are not well described. We find that exposing cultured retinal pericytes to 100 microM H(2)O(2) for 30 min leads to myosin heavy chain translocation from the cytosol to the cytoskeleton and a significant decrease in cell surface area. Pericyte death follows within 60-120 min. Exposing cells to 150 mJ/cm(2) ultraviolet radiation, an alternate free radical generating system, also causes pericyte myosin translocation and apoptosis. Proteolytic cleavage of actin is not observed in pericyte apoptosis. 3-aminobenzamide, a pharmacological inhibitor of the cleavage and activation of the DNA-repairing enzyme poly (ADP-ribose) polymerase (PARP) inhibits pericyte apoptosis, and prevents myosin translocation. Deferoxamine, an iron chelator known to interfere with free radical generation, also inhibits pericyte myosin translocation, contractility, and cell death. Myosin translocation to the cytoskeleton may be an early step in assembly of a competent contractile apparatus, which is involved in apoptotic cell condensation. These results suggest that pericyte loss associated with increased free radical production in diabetic retina may be by an apoptotic phenomenon.

Selective propagation of retinal pericytes in mixed microvascular cell cultures using L-leucine-methyl ester.

Endothelial cell (EC) propagation has been simplified by developing cell-specific selection criteria. Methods commonly used for selectively isolating EC include: (i) differential sieving of disaggregated tissue, (ii) differential plating of cells on extracellular matrices, (iii) lectin affinity isolation of cell populations and (iv) fluorescence-activated cell sorting of cells labeled with a carbocyanine dye of acetylated low-density lipoprotein (DiI-Ac-LDL). Few criteria for selectively propagating pericytes (PC) are currently available. Nonspecific esterases exhibit a high degree of multiplicity when compared with other mammalian isozymes and may be suitable for the identification and selective propagation of cells of the microvasculature. Evaluation of esterase isotype expression in PC and EC by zymography indicates PC contain alpha-naphthyl acetate and alpha-naphthyl butyrate hydrolyzing esterases as well as dipeptidyl peptidase I, while EC only contain alpha-naphthyl acetate esterase. The cytotoxic response of PC and EC to various amino acid esters is assessed by monitoring vital dye uptake and by light microscopy. Several amino acid esters are cytotoxic to both cell types, whereas 50 mM L-leucine methyl ester (L-Leu OMe) is toxic to EC but not to PC. This amino acid ester is also toxic to mesothelial and retinal pigmented epithelial cells, other common contaminants of PC cultures. Analysis of protein composition by two-dimensional gel electrophoresis indicates that L-Leu OMe does not stimulate expression of stress response proteins in PC. Thus, L-Leu OMe can be utilized to cultivate PC selectively from mixed cell populations.

Expression and subcellular distribution of filamin isotypes in endothelial cells and pericytes.

Two principal forms of the actin binding protein, filamin, are expressed in mammalian cells: nonmuscle and muscle isotypes (FLN-1 and FLN-2). A protein that copurifies with an alpha-naphthyl acetate hydrolyzing esterase from human omentum microvessel endothelial cells (EC) is isolated by nondenaturing electrophoresis, sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and electroblotting. The purified protein is subjected to in situ trypsin cleavage, reversed-phase high performance liquid chromatography (HPLC) and automated Edman degradation. Six peptide fragments from the protein are identified to have 60-66% identity with nonmuscle filamin (ABP-280). Two of these peptides are 100% identical to a previously sequenced human muscle filamin fragment. Polyclonal antibody is produced using a 16-residue synthetic peptide corresponding to a structural beta-sheet region of muscle filamin. Compared with a variety of vascular cells evaluated, retinal pericytes express an abundance of both muscle and non-muscle filamin isotypes. Pericytes contain at least 10 times more muscle filamin than human umbilical vein EC and at least three times the amount expressed in human omentum microvessel and bovine pulmonary artery EC. Differential detergent fractionation indicates that both filamin isotypes are primarily localized in the cytosol and membrane/organelle fractions of pericytes. Another actin crosslinking protein, alpha-actinin, is primarily found in the cytosol and cytoskeletal fractions. The dynamic regulation of actin microfilament organization in pericytes may be controlled in part by the two filamin isotypes, which in turn may contribute to pericyte contractility.

Metabolites of the phospholipase D pathway regulate H2O2-induced filamin redistribution in endothelial cells.

Hypoxia/reoxygenation injury to cultured endothelial cells results in cytoskeletal rearrangement and second messenger activation related to increased monolayer junctional permeability. Cytoskeletal rearrangement by reactive oxygen species may be related to specific activation of the phospholipase D (PLD) pathway. Human umbilical vein endothelial cell monolayers are exposed to H2O2 (100 microM) or metabolites of the PLD pathway for 1-60 min. Changes in cAMP levels, Ca2+ levels, PIP2 production, filamin distribution, and intercellular gap formation are then quantitated. H2O2-induced filamin translocation from the membrane to the cytosol occurs after 1-min H2O2 treatment, while intercellular gap formation significantly increases after 15 min. H2O2 and phosphatidic acid exposure rapidly decrease intracellular cAMP levels, while increasing PIP2 levels in a Ca2+-independent manner. H2O2-induced cAMP decreases are prevented by inhibiting phospholipase D. H2O2-induced cytoskeletal changes are prevented by inhibiting phospholipase D, phosphatidylinositol-4-phosphate kinase, phosphoinositide turnover, or by adding a synthetic peptide that binds PIP2. These data indicate that metabolites produced downstream of H2O2-induced PLD activation may mediate filamin redistribution and F-actin rearrangement.

Human omental microvascular endothelial and mesothelial cells: characterization of two distinct mesodermally derived epithelial cells.

Human omental microvascular endothelial (HOME) and mesothelial (MESO) cells share many phenotypic properties, but can be characterized from one another based upon a comprehensive panel of endothelial and mesothelial markers. Traditional cell markers such as von-Willebrand factor, DiI-Ac-LDL, and Ulex europaeus I lectin are not sufficient to distinguish between HOME and MESO cells. Furthermore, immunoreactivity to a panel of endothelial cell-specific monoclonal antibodies, including representatives from the known clusters of differentiation (CD), indicate that some of these antigens are coexpressed in HOME and MESO cells. In distinguishing between the two cell types, HOME and not MESO cells express E-selectin, E/P-selectin, P-selectin (CD62), Le-y, and VLA-6 (CDw49f*). Moreover, HOME cells and not MESO cells form tube-like structures when cultured on Matrigel. MESO cells differ from HOME cells based upon (1) the expression of cytokeratins; (2) their rapid proliferation in response to platelet-derived growth factor; and (3) a change from an epitheliod to fibroblast-like morphology in response to tumor necrosis factor and epidermal growth factor. Both HOME and MESO cells express tissue plasminogen activator and plasminogen activator inhibitor, but urokinase activity is only expressed by MESO cells. As there is no one universal endothelial or mesothelial cell marker that can specifically confirm the identity of these cells, it appears necessary to employ a comprehensive panel of cell markers to rule out the possibility of misidentifying a cell culture.

Two-stage isolation procedure for obtaining homogenous populations of microvascular endothelial and mesothelial cells from human omentum.

The human omentum is a highly vascularized tissue often advocated as a source of human microvascular endothelial (HOME) cells. The omentum also contains mesothelial (MESO) cells and isolation protocols published to date do not describe a separation of the two cell populations. Using a two-stage collagenase digestion procedure, homogenous populations of HOME and MESO cells are obtained from the same omental tissue sample. HOME and MESO cells are both simple squamous epithelial cells and consequently are often difficult to discriminate between based on morphology and reactivity with many of the conventional endothelial and mesothelial cell markers. Both HOME and MESO cells form typical cobblestone, contact-inhibited monolayers, metabolize DiI-Ac-LDL, and are immunoreactive to von Willebrand Factor and Ulex europeaus I lectin. However, MESO cells are distinguishable from HOME cells based upon their expression of cytokeratins. Moreover, HOME cells and not MESO cells form capillary-like structures when cultured on Matrigel. It appears that HOME and MESO cells share many phenotypic properties, but are distinguishable from one another based upon a comprehensive panel of endothelial and mesothelial markers. Both cell types should be useful for studying the biology and pathology of the human microvasculature in vitro.

H2O2-induced filamin redistribution in endothelial cells is modulated by the cyclic AMP-dependent protein kinase pathway.

Hypoxia/reoxygenation injury in vitro causes endothelial cell cytoskeletal rearrangement that is related to increased monolayer permeability. Nonmuscle filamin (ABP-280) promotes orthogonal branching of F-actin and links microfilaments to membrane glycoproteins. Human umbilical vein endothelial cell monolayers are exposed to H2O2 (100 microM) for 1-60 min, with or without modulators of cAMP-dependent second-messenger pathways, and evaluated for changes in filamin distribution, cAMP levels, and the formation of gaps at interendothelial junctions. Filamin translocates from the membrane-cytoskeletal interface to the cytosol within 1 min of exposure to H2O2. This is associated with a decrease in endothelial cell cAMP levels from 83 pmoles/mg protein to 15 pmoles/mg protein. Intercellular gaps form 15 min after H2O2 treatment and progressively increase in number and diameter through 60 min. Both filamin redistribution and actin redistribution are associated with decreased phosphorylation of filamin and are prevented by activation of the cAMP-dependent protein kinase pathway. A synthetic peptide corresponding to filamin's C-terminal, cAMP-dependent, protein kinase phosphorylation site effectively induces filamin translocation and intercellular gap formation, which suggests that decreased phosphorylation of filamin at this site causes filamin redistribution and destabilization of junctions. These data indicate that H2O2-induced filamin redistribution and interendothelial cell gap formation result from inhibition of the cAMP-dependent protein kinase pathway.

A novel anti-inflammatory peptide inhibits endothelial cell cytoskeletal rearrangement, nitric oxide synthase translocation, and paracellular permeability increases.

The endothelial cell (EC) membrane-cytoskeletal interface in part maintains plasma membrane integrity and promotes cell-cell apposition. Nonmuscle filamin (ABP-280), an actin crosslinking protein, promotes orthogonal branching of F-actin and is the major protein that links the peripheral actin network to the plasma membrane through its C-terminal glycoprotein binding site. In response to bradykinin, filamin translocates from the cell periphery to the cytosol within 1 min. A synthetic peptide, corresponding to filamin's C-terminal calcium/calmodulin-dependent protein kinase II phosphorylation site (CaM peptide), prevents calcium-activated filamin translocation in permeabilized bovine pulmonary artery EC. The myristoylated permeable form of this peptide inhibits bradykinin-induced filamin translocation and F-actin rearrangement in cultured intact ECs. In addition, bradykinin-induced paracellular gap formation is significantly attenuated by CaM peptide, which suggests that the presence of a filamin-based peripheral F-actin network is essential for maintaining EC barrier function. Moreover, CaM peptide reduces wound-induced EC migration rate by 40%, which indicates that F-actin rearrangement is required for efficient cell motility. The CaM peptide affects other bradykinin-induced inflammatory responses. EC nitric oxide synthase (eNOS) translocates from the cell membrane to the nuclear fraction within 1-2 min of bradykinin treatment. Pretreatment with CaM peptide inhibits eNOS translocation. However, the peptide has no effect on bradykinin-induced von Willebrand Factor release. In summary, the CaM peptide exhibits several anti-inflammatory properties that include maintaining EC junctional stability and inhibiting eNOS translocation.

A luminescent europium complex for the sensitive detection of proteins and nucleic acids immobilized on membrane supports.

Certain metal complexes selectively interact with proteins immobilized on solid-phase membrane supports to form brightly colored products. Detecting the absorbance of colorimetric stains is limited by the molar extinction coefficient of the product, however. Development of light-emitting complexes should improve detection sensitivity, but fluorescent labels described to date modify free amino, carboxyl, or sulfhydryl groups often rendering proteins unsuitable for further analysis. Bathophenanthroline disulfonate (BPSA) forms a luminescent europium (Eu) complex that reversibly binds to proteins and nucleic acids. Analysis of charge-fractionated carrier ampholytes and synthetic polymers of different L-amino acids indicates that protein binding is chiefly through protonated alpha- and epsilon-amino side chains. Proteins or nucleic acids immobilized to a nitrocellulose or polyvinyl difluoride membrane by electroblotting, dot-blotting, or vacuum slot-blotting are incubated with the lanthanide complex at acidic pH. Membranes are rinsed, illuminated with UV light and the phosphorescence of BPSA-Eu is measured at 590 to 615 nm using a CCD camera or spectrofluorimeter. The linear dynamic range of the stain is 476- and 48-fold for protein and DNA, respectively. A strong chelating agent such as ethylenediaminetetraacetic acid combined with a shift to basic pH (PH 8-10) elutes BPSA-Eu from the membrane. The reversible nature of the protein staining procedure allows for subsequent biochemical analyses, such as immunoblotting, lectin staining, and mass spectrometry.

Filamin redistribution in an endothelial cell reoxygenation injury model.

Ischemia-reperfusion injury increases vascular permeability in part by generating reactive oxygen species that disassemble the endothelial cell actin dense peripheral band. This is followed by an increase in the number and diameter of intercellular gaps. Millimolar concentrations of reactive oxygen metabolites lead to nonspecific endothelial cell injury, but micromolar concentrations activate inflammatory second messenger cascades which produce distributional changes in endothelial cell cytoskeletal proteins. H2O2 (100 microM) causes translocation of filamin, from the membrane to the cytosol within 1 min. Subsequently, gap formation occurs within 10-25 min, which is attributed to rearrangement of the dense peripheral band of F-actin. Plasma membrane blebbing occurs after 90 min and decreases in mitochondrial activity occur after 1-2 h. Deferoxamine (iron chelator) and TEMPO (nonspecific free radical scavenger) inhibit these changes. H2O2 (100-1000 microM) does not increase endothelial cell intracellular Ca2+ through 30 min and pretreating cells with a Ca2+-calmodulin kinase inhibitor or an intracellular Ca2+ chelator does not prevent filamin translocation. Filamin redistribution and actin rearrangement are early events in H2O2-mediated endothelial cell injury that appear to occur through Ca2+-independent pathways.

Antamanide Prevents Bradykinin-lnduced Filamin Translocation by Inhibiting Extracellular Calcium Influx.

Bradykinin-induced paracellular gap formation in cultured endothelial cells (ECs) is preceded by cytoskeletal rearrangement. Actin binding proteins, such as nonmuscle filamin, perform a pivotal role in modulating actin organization. In response to bradykinin treatment, EC filamin translocates from the cell periphery to the cytosol within 1 min and the dynamics of filamin translocation parallel intracellular Ca(2+) increases. Intracellular Ca(2+) increases are essential for bradykinin-induced filamin translocation. In this study, we examine the role of extracellular Ca(2+) influx in mediating bradykinin-induced filamin translocation. Several K(+) channel blockers, including antamanide, tetraethylam-monium chloride (TEA), and charybdotoxin (CTX), are evaluated. All of these agents inhibit extracellular Ca(2+) influx with minimal or partial inhibition of intracellular Ca(2+) release. Bradykinin-induced filamin translocation is prevented by pretreatment with these K(+) channel blockers. Moreover, incubation of ECs in high-K(+) saline inhibits bradykinin-induced extracellular Ca(2+) influx as well as filamin translocation. To examine the efficacy of antamanide as an anti-inflammatory drug that affects filamin translocation, bradykinin-induced paracellular gap formation is quantified and compared in the presence or absence of antamanide pretreatment. Antamanide does not completely block bradykinin-induced gap formation; however, a significant attenuation is observed. This suggests that extracellular Ca(2+) influx is required for bradykinin-induced filamin translocation, which in part regulates microvascular EC barrier function, and that antamanide may be a useful anti-inflammatory agent.