In vitro reendothelialization. Microfilament bundle reorganization in migrating porcine endothelial cells

Exosomes from Adipose Mesenchymal Stem Cells Overexpressing Stanniocalcin-1 Promote Reendothelialization After Carotid Endarterium Mechanical Injury

OBJECTIVE

Stanniocalcin-1 (STC-1) is a secreted glycoprotein that participates in the regulation of inflammation, apoptosis, and necrosis. We investigated the reendothelialization effect of exosomes from adipose stem cells (ADSC) overexpressing STC-1 on injured carotid endarterium.

METHODS

ADSCs were transfected with lentivirus vectors containing pre-STC-1. PHK-26 as molecular probe was used to track the exosomes engulfed by mice arterial endothelial cells (MAEC). The role of STC-1-ADSC-Exosome (S-ADSC-Exo) in MAECs was verified through scratch test and tube forming. Expressions of STC-1 and NLRP3 inflammasome were detected by western blot and quantitative reverse transcription polymerase chain reaction. Reendothelialization effect was inhibited by the antagonist of siRNA targeting STC-1. Carotid endarterium mechanical injury was induced by insertion with a guidewire into the common carotid artery lumen. Carotid arteries were harvested for histological examination, immunofluorescence staining, and Evan's blue staining.

RESULTS

Transfection of STC-1 significantly enhanced STC-1 levels in ADSCs, their exosomes, and MAECs. Compared with the control group and the ADSC-Exo group, STC-1 enriched exosomes markedly inhibited the expressions of NLRP3, Caspase-1, and IL-1β in MAECs, exhibited good lateral migration capacity, and promoted angiogenesis. Administration of siRNA targeting STC-1 completely abolished down-regulation of NLRP3, Caspase-1, and IL-1β by STC-1 and inhibited effects of S-ADSC-Exo on lateral migration and angiogenesis. In vivo administration of S-ADSC-Exo had reendothelialization effect on post-injury carotid endarterium as evidenced by thinner arterial wall, low-expressed NLRP3 inflammasome, and more living endothelial cells.

CONCLUSIONS

The reendothelialization effect of exosomes from ADSCs on post-injury carotid endarterium could be enhanced by genetic modification of the exosomes to contain elevated STC-1, possibly through suppression of NLRP3 inflammasome-mediated inflammation.

The Endothelial Cytoskeleton: Organization in Normal and Regenerating Endothelium*

The components of the endothelial cell cytoskeleton that have been shown to be important in maintaining endothelial structural integrity and in regulating endothelial repair include F-actin microfilament bundles, including stress fibers, and microtubules, and centrosomes. Endothelial cells contain peripheral and central actin microfilaments. The dense peripheral band (DPB) consists of peripheral actin microfilament bundles which are associated with vinculin adhesion plaques and are most prominent in low or no hemodynamic shear stress conditions. The central microfilaments are very prominent in areas of elevated hemodynamic shear stress. There is a redistribution of actin microfilaments characterized by a decrease of peripheral actin and an increase in central microfilaments under a variety of conditions, including exposure to thrombin, phorbol-esters, and hemodynamic shear stress. During reendothelialization, there is a sequential series of cytoskeletal changes. The DPB remains intact during the rapid lamellipodia mediated repair of very small wounds except at the base of the lamellipodia where it is splayed. The DPB is reduced or absent when cell locomotion occurs to repair a wound. In addition, when cell locomotion is required, the centrosome, in the presence of intact microtubules, redistributes to the front of the cell to establish cell polarity and acts as a modulator of the directionality of migration. This occurs prior to the loss of the DPB but does not occur in very small wounds that close without migration. Thus, the cytoskeleton is a dynamic intracellular system which regulates endothelial integrity and repair and is modulated by external stimuli that are present at the vessel wall-blood interface.

An agent-based model approach to multi-phase life-cycle for contact inhibited, anchorage dependent cells

Cellular agent-based models are a technique that can be easily adapted to describe nuances of a particular cell type. Within we have concentrated on the cellular particularities of the human Endothelial Cell, explicitly the effects both of anchorage dependency and of heightened scaffold binding on the total confluence time of a system. By expansion of a discrete, homogeneous, asynchronous cellular model to account for several states per cell (phases within a cell's life); we accommodate and track dependencies of confluence time and population dynamics on these factors. Increasing the total motility time, analogous to weakening the binding between lattice and cell, affects the system in unique ways from increasing the average cellular velocity; each degree of freedom allows for control over the time length the system achieves logistic growth and confluence. These additional factors may allow for greater control over behaviors of the system. Examinations of system's dependence on both seed state velocity and binding are also enclosed.

Quantification and regulation of cell migration

Cell migration is essential for many physiological and pathological processes that include embryonic development, the immune response, wound healing, angiogenesis, and cancer metastasis. It is also important for emerging tissue engineering applications such as tissue reconstitution and the colonization of biomedical implants. By summarizing results from recent experimental and theoretical studies, this review outlines the role played by growth factors or substrate-adhesion molecules in modulating cell motility and shows that cell motility can be an important factor in determining the rates of tissue formation. The application of cell motility assays and the use of theoretical models for analyzing cell migration and proliferation are also discussed.

A 3D hybrid model for tissue growth: the interplay between cell population and mass transport dynamics

To provide theoretical guidance for the design and in vitro cultivation of bioartificial tissues, we have developed a multiscale computational model that can describe the complex interplay between cell population and mass transport dynamics that governs the growth of tissues in three-dimensional scaffolds. The model has three components: a transient partial differential equation for the simultaneous diffusion and consumption of a limiting nutrient; a cellular automaton describing cell migration, proliferation, and collision; and equations that quantify how the varying nutrient concentration modulates cell division and migration. The hybrid discrete-continuous model was parallelized and solved on a distributed-memory multicomputer to study how transport limitations affect tissue regeneration rates under conditions encountered in typical bioreactors. Simulation results show that the severity of transport limitations can be estimated by the magnitude of two dimensionless groups: the Thiele modulus and the Biot number. Key parameters including the initial seeding mode, cell migration speed, and the hydrodynamic conditions in the bioreactor are shown to affect not only the overall rate, but also the pattern of tissue growth. This study lays the groundwork for more comprehensive models that can handle mixed cell cultures, multiple nutrients and growth factors, and other cellular processes, such as cell death.

Functional and ultrastructural analysis of endothelial-like cells derived from bone marrow stromal cells

BACKGROUND

Recent studies have suggested that bone marrow stromal cells (BMSC) have the potential to differentiate into endothelial cells. However, the physiologic functions of the endothelial-like cells derived from BMSC have not been well studied.

METHODS

Human BMSC were induced to differentiate into endothelial-like cells with a combination of cytokines. Morphologic, phenotypic, ultrastructural and functional characterizations of the endothelial-like cells were made.

RESULTS

Human BMSC were successfully differentiated into cells with endothelial-like morphology and phenotype in vitro. These cells expressed various endothelial cell functions in vitro, such as release of von Willebrand factor (vWF) mediated by histamine, acetylated low-density lipoprotein (acLDL) uptake, binding of Ulex europaeus agglutinin-1 (UEA-1) and in vitro capillary formation. The cells also acquired important ultrastructural and physiologic properties of endothelial cells as they contained Weibel-Palade bodies, abundant mitochondria with a homogeneous mitochondrial matrix, diluted rough endoplasmic reticula, enlarged Golgi complexes, a regular arrangement of microfilaments and many surface cytoplasmic processes and plasmalemmal vesicles, as well as intercellular tight junctions and desmosome-like structures. Subcutaneous implantation of the endothelial-like cells in Matrigel plugs in immunodeficient mice resulted in the formation of functional blood vessels that contained erythrocytes. Moreover, these cells contributed to in vivo neovascularization during wound healing in rabbit ischemic hindlimb models.

DISCUSSION

Physiologic features of the endothelial-like cells derived from BMSC suggest the potential use of these cells as a functional cell source for therapeutic applications.

Cell population dynamics modulate the rates of tissue growth processes

The development and testing of a discrete model describing the dynamic process of tissue growth in three-dimensional scaffolds is presented. The model considers populations of cells that execute persistent random walks on the computational grid, collide, and proliferate until they reach confluence. To isolate the effect of population dynamics on tissue growth, the model assumes that nutrient and growth factor concentrations remain constant in space and time. Simulations start either by distributing the seed cells uniformly and randomly throughout the scaffold, or from an initial condition designed to simulate the migration and cell proliferation phase of wound healing. Simulations with uniform seeding show that cell migration enhances tissue growth by counterbalancing the adverse effects of contact inhibition. This beneficial effect, however, diminishes and disappears completely for large migration speeds. By contrast, simulations with the "wound" seeding mode show a continual enhancement of tissue regeneration rates with increasing cell migration speeds. We conclude that cell locomotory parameters and the spatial distribution of seed cells can have profound effects on the dynamics of the process and, consequently, on the pattern and rates of tissue growth. These results can guide the design of experiments for testing the effectiveness of biomimetic modifications for stimulating tissue growth.

Regulation of the actin cycle in vivo by actin filament severing

Cycling of actin subunits between monomeric and filamentous phases is essential for cell crawling behavior. We investigated actin filament turnover rates, length, number, barbed end exposure, and binding of cofilin in bovine arterial endothelial cells moving at different speeds depending on their position in a confluent monolayer. Fast-translocating cells near the wound edge have short filament lifetimes compared with turnover values that proportionately increase in slower moving cells situated at increasing distances from the wound border. Contrasted with slow cells exhibiting slow actin filament turnover speeds, fast cells have less polymerized actin, shorter actin filaments, more free barbed ends, and less actin-associated cofilin. Cultured primary fibroblasts manifest identical relationships between speed and actin turnover as the endothelial cells, and fast fibroblasts expressing gelsolin have higher actin turnover rates than slow fibroblasts that lack this actin-severing protein. These results implicate actin filament severing as an important control mechanism for actin cycling in cells.

Reduced in vitro repair in endothelial cells harvested from the intercostal ostia of porcine thoracic aorta

The ability of large-vessel endothelium to repair itself rapidly after injury is important in the maintenance of its barrier function and in limiting the development and progression of atherosclerosis. Because dysfunctional repair may be involved in the pathogenesis of some atherosclerotic plaques, including those at the ostia of aortic branches, linear mechanical denuding wounds were made in confluent monolayers of endothelial cells harvested by scraping from the flow divider, the upstream wall of the intercostal branch and unbranched regions in the thoracic aorta. The extent of wound closure was significantly lower in cells derived from either side of the intercostal branches, compared with cells from unbranched areas. The wound edge of cells harvested from the flow divider and its opposite wall closed by 22+/-0.084 microm and 22+/-1.3 microm, respectively, versus control, unbranched endothelial cells (30+/-2.2 microm) at 24 hours and by 48 hours, 48+/-3.4 microm and 47+/-3.6 microm compared with control (61+/-3.4 microm). Extent of wound closure in cells harvested by scraping from unbranched regions was comparable with collagenase-harvested endothelial cells at 24 and 48 hours. Distribution of F-actin microfilaments, tubulin and centrosomes have been shown to be disrupted at the wound edge in poorly migrating cells. In our study, however, no differences were observed in cytoskeletal distribution between cells from branched, unbranched and control areas. Thus, aortic endothelial cells from the intercostal branch region show a reduced ability to repair wounds compared with cells harvested from unbranched aorta. The mechanism for this difference is currently unknown.

Measuring actin dynamics in endothelial cells

Cytoplasmic actin distributes between monomeric and filamentous phases in cells. As cells crawl, actin polymerizes near the plasma membrane of expanding peripheral cytoplasm and depolymerizes elsewhere. Thus, the finite actin filament lifetime, the diffusivity of actin monomer, and the distribution of actin between the polymer and monomer phases are key parameters in cell motility. The dynamics of cellular actin can be determined by following the evolution of fluorescence in the techniques of photoactivated fluorescence (PAF) or fluorescence recovery after photobleaching (FRAP) of microinjected actin derivatives. A mathematical model is discussed that measures monomer diffusion coefficients, filament turnover rates, and the fraction of actin polymerized from measurements of the evolution of fluorescence from a photoactivated band [Tardy et al. (1995) Biophys. J., 69:1674-1682; McGrath et al. (1998) Biophys. J., in press]. Applying this model to subconfluent endothelial cells shows that approximately 40% of the actin is polymer and that these filaments turn over on average every 6 minutes. This report discusses how PAF and FRAP can be combined with more traditional biochemistry to probe actin cytoskeleton remodeling in endothelial cells.

Effects of disturbed flow on endothelial cells

Atherosclerotic lesions tend to localize at curvatures and branches of the arterial system, where the local flow is often disturbed and irregular (e.g., flow separation, recirculation, complex flow patterns, and nonuniform shear stress distributions). The effects of such flow conditions on cultured human umbilical vein endothelial cells (HUVECs) were studied in vitro by using a vertical-step flow channel (VSF). Detailed shear stress distributions and flow structures have been computed by using the finite volume method in a general curvilinear coordinate system. HUVECs in the reattachment areas with low shear stresses were generally rounded in shape. In contrast, the cells under higher shear stresses were significantly elongated and aligned with the flow direction, even for those in the area with reversed flow. When HUVECs were subjected to shearing in VSF, their actin stress fibers reorganized in association with the morphological changes. The rate of DNA synthesis in the vicinity of the flow reattachment area was higher than that in the laminar flow area. These in vitro experiments have provided data for the understanding of the in vivo responses of endothelial cells under complex flow environments found in regions of prevalence of atherosclerotic lesions.

Human endothelial cell damage by neutrophil-derived cathepsin G. Role of cytoskeleton rearrangement and matrix-bound plasminogen activator inhibitor-1

Cathepsin G, a major protease released by activated neutrophils, induces functional and morphological damage to human endothelial cells. We studied the mechanisms involved and ways to reverse this damage. Cathepsin G induced a concentration- and time-dependent injury to human umbilical vein endothelial cell (HUVEC) morphology simultaneous with cytoskeleton rearrangement. Preincubation of the endothelial monolayer with phallacidin completely prevented damage to cell morphology by cathepsin g, whereas preincubation with cytochalasin b potentiated its activity. Damage to cell shape and F-actin cytoskeleton were prevented by eglin C, and inhibitor of the active site of cathepsin G. Furthermore, cathepsin G increased transcellular permeability to albumin and induced a time-dependent detachment of PAI-1 from the extracellular matrix of a cell-free system. The inhibition of matrix-bound PAI-1 activity by specific antibodies induced matrix-bound PAI-1 activity by specific antibodies induced changes in HUVEC monolayers similar to those observed after cathepsin G. However, although stabilization of F-actin microfilaments by phallacidin prevented changes in cell shape, it did not prevent the ability of cathepsin G to increase cell permeability and release matrix PAI-1. The damage of cathepsin G to cell morphology and cytoskeleton arrangement was reversed within 12 hours if the deendothelialization area was < 50% to 55% and the subendothelial matrix was still able to bind the newly synthesized PAI-1. Thrombin, whose role in the thrombotic process is well known, also induced changes in cell morphology and cytoskeleton arrangement of HUVEC. Cathepsin G reaches the subendothelial matrix through an increase in cell permeability and injures endothelial cell morphology by detaching matrix-bound PAI-1. These events expose a highly thrombogenic surface to which platelets can adhere, become activated, attract further neutrophils, and trigger thrombus formation.

A cellular automaton model for the proliferation of migrating contact-inhibited cells

A cellular automaton is used to develop a model describing the proliferation dynamics of populations of migrating, contact-inhibited cells. Simulations are carried out on two-dimensional networks of computational sites that are finite-state automata. The discrete model incorporates all the essential features of the cell locomotion and division processes, including the complicated dynamic phenomena occurring when cells collide. In addition, model parameters can be evaluated by using data from long-term tracking and analysis of cell locomotion. Simulation results are analyzed to determine how the competing processes of contact inhibition and cell migration affect the proliferation rates. The relation between cell density and contact inhibition is probed by following the temporal evolution of the population-average speed of locomotion. Our results show that the seeding cell density, the population-average speed of locomotion, and the spatial distribution of the seed cells are crucial parameters in determining the temporal evolution of cell proliferation rates. The model successfully predicts the effect of cell motility on the growth of isolated megacolonies of keratinocytes, and simulation results agree very well with experimental data. Model predictions also agree well with experimentally measured proliferation rates of bovine pulmonary artery endothelial cells (BPAE) cultured in the presence of a growth factor (bFGF) that up-regulates cell motility.

Insulin does not disrupt actin microfilaments, microtubules, and in vitro aortic endothelial wound repair

In the face of small denuding injuries, the endothelium undergoes a process of rapid repair involving actin microfilaments, microtubules, and centrosomes to reestablish an intact monolayer. Failure to maintain an intact endothelial monolayer is an important factor in the pathogenesis of the atherosclerotic plaque. It was hypothesized that increased susceptibility to atherosclerosis in diabetes mellitus may be, in part, due to delayed reendothelialization following endothelial injury. To test this, the effects of high insulin concentrations on the reendothelialization of small wounds were examined using an in vitro porcine aortic endothelial cell wound model. Elevated concentrations of insulin did not disrupt the confluent endothelial monolayer or alter endothelial cell shape. Insulin also did not induce detectable alterations in the distribution of microtubules and microfilaments in the confluent monolayer. High insulin did not reduce the extent of reendothelialization of a linear wound made in the confluent monolayer. Centrosomal reorientation was similar to that of control wounded cultures as was the reorganization of the microfilaments and microtubules. The data suggest that the atherogenic effects of hyperinsulinemia are not due to disruption of endothelial repair.

Vascular smooth muscle cell migration mediated by thrombin and urokinase receptor

To determine whether thrombin directly modifies mobility of vascular smooth muscle cells (SMC), in Transwell systems (modified Boyden chambers), we exposed SMC to alpha-thrombin. In concentrations as low as 1 NIH U/ml, thrombin induced migration as well as proliferation of SMC. Inhibition of protein synthesis by cycloheximide (2 micrograms/ml) obviated thrombin's chemotactic effect. Neither gamma-thrombin nor D-phenylalanyl-L-prolyl-L-arginine chloromethyl ketone (PPACK)-inactivated alpha-thrombin (both used as controls) exerted a chemotactic effect. Concomitant hirudin or antithrombin III plus heparin inhibited chemotaxis by thrombin when added up to 2 h after addition of thrombin. alpha-Thrombin increased SMC synthesis of urokinase receptor (uPAR) and its cell surface expression as shown by metabolic labeling and immunoprecipitation as well as by flow cytometry. Thus alpha-thrombin, in concentrations thought to be present in vivo at sites of vascular injury, can stimulate not only proliferation but also migration of vascular SMC though a mechanism(s) possibly involving synthesis of uPAR, which is known to influence migration in diverse types of cells. Accordingly, both proliferation and migration dependent on thrombin may accelerate atherosclerosis, restenosis, or both after interventions such as angioplasty.

Three patterns of distribution characterize the organization of endothelial microfilaments at aortic flow dividers

Since actin microfilaments are essential in the maintenance of endothelial integrity and in the repair of injured endothelium, we have carried out a detailed study of the distribution of microfilaments in the immediate vicinity of aortic branches. Branches are of major interest because there is a predilection for atherosclerotic lesions near branch ostia. We made an extensive, systematic examination of branches of the aorta and iliac arteries using in situ staining of perfusion-fixed arteries. Microfilaments were localized using rhodamine phalloidin. Three patterns of staining were observed. Some endothelial cells showed prominent central stress fibers. Others had few central stress fibers but prominent peripheral fibers. Still others showed an intermediate pattern with some central and some peripheral fibers present. At small branch sites, the lip of the divider was more blunt, and there were more cells with peripheral actin. At large branches, cells with peripheral actin were confined mainly to the lip, while there were many more cells with prominent central fibers. We also found that major differences can occur over very small distances, so adjacent cells may have strikingly different patterns of microfilament distribution. These patterns appear to reflect the geometry of the flow divider and local variations in hemodynamic shear stress. The differences in microfilament distribution may reflect differences in endothelial functions which are essential in maintaining endothelial integrity.

Role of Ca2+ and protein kinase C in shear stress-induced actin depolymerization and endothelin 1 gene expression

Vascular endothelial cells adapt to changes in blood flow by altering the cell architecture and by producing various substances. We have previously reported that low shear stress induces endothelin 1 (ET-1) expression in endothelial cells and that this induction is mediated by depolymerization of actin fiber. In the present study, we examined the role of Ca2+ and protein kinase C (PKC) in shear stress-induced actin depolymerization and subsequent ET-1 gene expression. Exposure of cultured porcine aortic endothelial cells to low shear stress (5 dyne/cm2) for 3 hours increased the ratio of G-actin to total actin from 54 +/- 0.8% to 80 +/- 1.0%. This shear stress-induced actin depolymerization was completely blocked by chelation of extracellular Ca2+ with EGTA and partially inhibited by intracellular Ca2+ chelation with the tetraacetoxymethyl ester of BAPTA (BAPTA/AM). Pretreatment with staurosporine, a PKC inhibitor, or desensitization of PKC by treatment with 12-O-tetradecanoylphorbol 13-acetate (TPA) for 24 hours also resulted in partial inhibition of shear stress-induced actin depolymerization. Although PKC activation by TPA mildly increased G-actin content, the effect of TPA and shear stress on actin depolymerization was not additive. Moreover, shear stress-induced ET-1 gene expression was inhibited by EGTA, BAPTA/AM, and staurosporine to a degree similar to the inhibition of actin depolymerization. In contrast, ET-1 gene expression induced by cytochalasin B, an actin-disrupting agent, was not affected by staurosporine.(ABSTRACT TRUNCATED AT 250 WORDS)

Experimental in vitro endothelialization of cardiac valve leaflets

This study reports our results with vitro endothelialization of fresh nonpreserved homograft valve leaflets compared with mild alternatively preserved valves and valves treated by preservation procedures commonly used for commercially available tissue valves. In vitro lining of biological heart valves with cultured autologous endothelial cells might help prevent the detrimental effects of degeneration on valve durability. To investigate the growth characteristics of endothelial cells on valve bioprostheses, three different methods of storage and preservation were compared. After precoating with fibronectin and seeding of 4.4 x 10(4) endothelial cells/cm2 onto the different leaflet surfaces, primary adherence, growth kinetics, morphology, and maintenance of monolayer integrity were studied over a period of 10 days. On valve leaflet surfaces of group 1 (fresh nonpreserved homograft valve leaflets) and group 2 (mild alternatively preserved valves), endothelial cells grew to persistent monolayers between days 6 and 10. In contrast, endothelial cell proliferation with monolayer growth could not be achieved on the group 3 leaflets (preserved like commercially available biological valve prostheses). In that group, no viable endothelial cells could be found on the valve surfaces 2 days after seeding. These results demonstrate the theoretical feasibility of endothelializing biological heart valve leaflets in vitro if they are not preserved and stored according to commonly used procedures. Provided such an endothelium can withstand the mechanical forces after implantation in vivo, in vitro endothelialization might contribute either to the development of new biological heart valves for modern cardiac surgery or to the improvement of clinical results with homograft valve transplants.

Disruption of endothelial actin microfilaments by protein kinase C inhibitors

In this study, we report that the isoquinolinesulfonamide inhibitors of protein kinase C (PKC), H-7 [1-(5-isoquinolinesulfonyl)-2-methylpiperazine] and its related derivatives H-8 and HA-1004, in addition to staurosporine cause depletion and reorganization of microfilament bundles of porcine aortic endothelial cells in both low-density and confluent monolayer cultures. Concomitantly, significant loss of cell adhesion was noted following treatment with H-7. The effects of these compounds were found to be reversible upon wash-out, with restoration of the microfilament network. In addition, longer term incubation with phorbol myristate acetate (PMA) carried out to deplete PKC results in depletion of microfilaments as well. After 24 hr of PMA incubation, however, addition of H-7 or staurosporine is associated with further loss of the remaining microfilaments, suggesting that these agents act, at least in part, through a PKC-independent mechanism.

Ultrastructure appearance of atherosclerosis in human and experimentally-induced animal models

We describe here the basic structure of the aorta, the changes with aging and ultrastructural appearance of atherosclerosis of human and animal models. The architecture of the aortic wall is highly organized, for adaptation to changes of blood pressure. The main cells composing the vessel are endothelial cells and smooth muscle cells. They maintain the integrity and homeostasis of the aorta along with the extracellular matrix of collagen fibrils, elastic fibers and glycosaminoglycans. The structural changes with aging and atherogenesis are a compensative or degenerative phenomenon caused by many factors. Three major cells are the endothelial cell, smooth muscle cell and monocyte-derived macrophages (as well as platelets) all of which are involved in atherogenesis. Foam cells in atheromatous lesions are derived from macrophages and smooth muscle cells. Recently, the molecular biological nature and function of these cells and their derived-factors have been thoroughly investigated in cell culture and in experimental animal models caused by a mechanical injury of the endothelium or by a dietary induced hypercholesterolemia. However, the mechanism of the endothelial injury in vivo as well as formation of atheromatous cores of human atherosclerosis is not exactly understood. Some structural and functional changes inherent to the arterial wall during aging may play an important role in initiation or progression of human atherosclerosis.

Cytoskeleton in TFG-beta- and bFGF-modulated endothelial monolayer repair

Endothelial cell proliferation and migration in vitro is depressed by transforming growth factor beta (TFG-beta) and enhanced by basic fibroblast growth factor (bFGF) treatment. This study examines interactions between cytoskeletal changes and cell proliferation in regenerating endothelial monolayers treated with bFGF, TFG-beta, and both factors. As previously described by others, monolayer regeneration is enhanced by bFGF and reduced by TFG-beta. Endothelial cell morphology is altered by TFG-beta treatment. Cells lose their cobblestone appearance and assume a pleomorphic shape. Actin microfilament staining is modified in both intact and regenerating TFG-beta-treated monolayers as well. There is a loss of dense peripheral band staining and an enhancement in staining intensity of cytoplasmic stress fibers. No such alterations are seen in bFGF-treated cultures. Cell proliferation at the wound edge, as indicated by bromodeoxyuridine incorporation, is inhibited by TGF-beta. Although monolayer repair is modulated by growth factor treatment, centrosome reorientation and microtubule staining patterns are not altered by either factor. Thus these factors appear to have effects on a mechanism(s) other than centrosome reorientation which may be involved in repair of denuded endothelial monolayers.

Immunocytochemistry of intermediate filaments in cultured arterial smooth muscle cells: differences in desmin and vimentin expression related to cell of origin and/or plating time

The objective of this study was to determine whether intermediate filament expression, including desmin and vimentin, in cultured smooth muscle cells (SMCs) is related to cyto-differentiation or proliferation. Using antibodies to desmin and vimentin, we studied by immunoperoxidase technique the distribution of these proteins in subcultured SMCs derived from porcine aorta and coronary artery. In addition, the proliferative potentiality of the cells was estimated by the incorporation of [3H]thymidine into DNA. The frequency of desmin-positive cells in coronary arterial SMCs of 3 and 6 population doubling levels was significantly higher as compared to findings with the aortic SMCs and depended on the plating time. No difference was evident at the 12 population doubling level. In contrast, vimentin was present in the majority of both aortic and coronary arterial SMCs. With regard to the localization of vimentin, two cell types were observed, one had reaction products to vimentin in both perinuclear and cell-peripheral areas (type-I cell), the other only in the cell-peripheral region (type-II cell). The relative proportion of the type-I and -II cells varied with the period of culture. Most of the SMCs showed the type-I cell on the first day and the number of type-II cells was increased on the sixth day. Quiescent SMCs in serum-free media had the same percentage of desmin-positive cells and frequency distribution of type-I and -II cells as did the proliferating SMCs incubated in media containing 5% serum. These results suggest that intermediate filament expression, including desmin and vimentin in cultured SMCs, is related to cell origin and/or plating time, but not to the proliferating activity, per se.

Role of the cytoskeleton during injury-induced cell migration in corneal endothelium

The role of microfilaments and microtubules during injury-induced cell migration of corneal endothelial cells in situ along their natural basement membrane has been investigated using organ culture. In the noninjured tissue, actin is localized at or near the plasma membrane, whereas tubulin is observed as a delicate lattice pattern throughout the cytoplasm. Twenty-four hours after a circular freeze injury, cells surrounding the wound area extend processes into this region. Fluorescent microscopy using phallotoxins and anti-tubulin antibodies demonstrated the presence of stress fibers and microtubule reorganization within these cells. Between 24 and 48 h post-injury endothelial cells move into the wound region, and by 48 h, the injury zone is repopulated and the monolayer is becoming reestablished. When injured corneas are placed in media containing 5 x 10(-7) M cytochalasin B, endothelial cell migration occurs; but it is slow, and wound closure is not complete even by 72 h. In contrast, when tissues are cultured in the presence of 10(-8) M colchicine, cell movement is greatly reduced, complete wound closure does not occur, and endothelial cells at the wound edge fail to display extensions typical of migrating cells. Furthermore, when injured endothelia are exposed to 0.05 micrograms/ml of actinomycin D for 15 min within the first hour after injury and transferred back into culture media lacking the drug for the duration of the experiment, migration does not occur and the wound persists. These actinomycin D treated cells remain viable as shown by their ability to incorporate 3H-uridine and 3H-thymidine. Fluorescence microscopy of actinomycin D treated tissues revealed the presence of stress filaments but disorganized microtubule patterns. Interestingly, 24 h after injury, if the tissue is exposed to actinomycin D, even for periods of up to 1 h, migration is not inhibited. Our results indicate that injury-induced endothelial cell movement appears to be more dependent on microtubule than microfilament reorganization and may require a critical timing of macromolecular synthesis.

Endothelial monolayer integrity. Perturbation of F-actin filaments and the dense peripheral band-vinculin network

The role of the actin microfilaments in maintaining the integrity of the monolayer and activating endothelial repair processes is not well understood. This study was designed to characterize the prominent changes in F-actin distribution in endothelial cells that are associated with shape changes in the cells after perturbation of a confluent monolayer. F-actin was localized by using rhodamine phalloidin and fluorescence microscopy. The dense peripheral band (DPB) and vinculin cell-cell junctions were co-localized by using double fluorescence and immunofluorescence microscopy. Thrombin and 12-o-tetradecanoyl-myristyl-13-acetate (TPA) caused loss of the DPB and an increase in the central microfilament bundles, while agents that caused rounding of the cells (including plasmin, trypsin, and chymotrypsin) did not cause loss of the DPB although large gaps were formed between cells. The thrombin and TPA effects were rapid and reversible and were associated with an accompanying loss of vinculin cell-cell plaques. The mechanisms of the effects were not studied. It was postulated that thrombin and TPA were activating endothelial repair processes.

Effect of platelet factors on migration of cultured bovine aortic endothelial and smooth muscle cells

Endothelial cell (EC) injury and the response of EC and smooth muscle cells (SMCs) to injury contribute to the pathophysiology in patients with vascular disease and atherosclerosis. Since platelets have been suggested to play an important role in modulating vascular injury, the present study was undertaken to examine the influence and mechanism of action of individual platelet factors on bovine aortic EC and SMC migration using an in vitro wound assay system. Serotonin decreased EC proliferation and reduced EC migration 21 +/- 1% (p less than 0.005), which was attenuated by imipramine. Transforming growth factor-beta reduced EC proliferation and decreased EC migration 52 +/- 3% (p less than 0.005). Norepinephrine increased EC proliferation but decreased EC migration 26 +/- 2% (p less than 0.005), which was abolished by phenoxybenzamine. Histamine increased EC proliferation but reduced EC migration 29 +/- 2% (p less than 0.005), which was attenuated by diphenhydramine. Platelet-derived growth factor decreased EC proliferation and decreased EC migration 40 +/- 2% (p less than 0.005). In contrast, serotonin increased SMC proliferation and increased SMC migration 31 +/- 2% (p less than 0.005), which was abolished by ketanserin. Transforming growth factor-beta increased SMC migration 35 +/- 5% (p less than 0.005). Norepinephrine increased SMC proliferation and increased SMC migration 43 +/- 4% (p less than 0.005), which was abolished by propranolol. Histamine increased SMC proliferation and increased SMC migration 38 +/- 3% (p less than 0.005), which was abolished by cimetidine. Platelet-derived growth factor increased SMC proliferation and increased SMC migration 40 +/- 3% (p less than 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)

Enhanced responsiveness of endothelium in the growing/motile state to tumor necrosis factor/cachectin

Some in vivo observations have suggested that growing or perturbed endothelium, such as that which occurs during angiogenesis, is more sensitive to the action of cytokines (TNF/cachectin, TNF, or IL-1) than normal quiescent endothelial cells. This led us to examine the responsiveness of endothelium to TNF as a function of the growth/motile state of the cell. TNF-induced modulation of endothelial cell surface coagulant function was half-maximal at a concentration of approximately 0.1 nM in subconfluent cultures, whereas 1-2 nM was required for the same effect in postconfluent cultures. Perturbation of endothelial cell shape/cytoskeleton was similarly more sensitive to TNF in subconfluent cultures. Consistent with these results, radioligand binding studies demonstrated high affinity TNF binding sites, Kd approximately 0.1 nM on subconfluent cultures, whereas only lower affinity sites (Kd approximately 1.8 nM) were detected on postconfluent cultures. The mechanisms underlying this change in the affinity of endothelium for TNF were studied in four settings. Crosslinking experiments with 125I-TNF and endothelium showed additional bands corresponding to Mr approximately 66,000 and approximately 84,000 with subconfluent cultures that were not observed with postconfluent cultures. Experiments with X-irradiated endothelium, whose growth but not motility was blocked, indicated that proliferation was not required for induction of high affinity TNF sites. Postconfluent endothelium, triggered to enter the proliferative cycle by microbutuble poisons, expressed high affinity TNF binding sites together with changes in cell shape/cytoskeleton well before their entry into S phase. Using wounded postconfluent monolayers, cells that migrated into the wound and those close to the wound edge displayed enhanced TNF binding and modulation of coagulant properties. These results suggest a model for targetting TNF action within the vasculature; regulation of high affinity endothelial cell binding sites can direct TNF to activated cells in particular parts of the vascular tree.

Stress fibers in endothelial cells overlying atherosclerotic lesions in rabbit aorta

Endothelial injury or dysfunction has long been postulated to promote atherogenesis, but structural alterations of endothelium in atherosclerosis have remained obscure. We report the common occurrence of actin-containing stress fibers, stainable by rhodamine-phalloidin, in endothelium overlying atherosclerotic lesions in cholesterol-fed rabbits. Nonlesioned areas in the same aortas showed normal endothelium with minimal development of stress fibers, which was no different from the appearance of endothelium in chow-fed rabbits. Microtubule organization revealed by immunofluorescence appeared normal in all areas. The development of stress fibers may be related to an altered substratum for endothelial attachment. This study provided no evidence to relate stress fiber formation with lesion initiation, but an association with well-developed foam cell lesions was evident.

Patterns of endothelial microfilament distribution in the rabbit aorta in situ

The available data on F-actin microfilament distribution in vascular endothelial cells in vivo is limited. In this study, the appearance and distribution of endothelial cell microfilaments in the rabbit thoracic aorta, the abdominal aorta and its major arterial branch points, and the aortic bifurcation were examined. Perfusion fixed rabbit aortas were stained in situ for F-actin by infusing rhodamine phalloidin via a peristaltic pump into the aortas at a slow flow rate. This new technique resulted in excellent visualization of branch points and allowed for a precise description of the actin microfilament bundles in endothelial cells along flow dividers. In the thoracic and abdominal aorta, away from branch ostia, actin microfilaments were localized in two regions of the endothelial cells, as a prominent band that completely outlined the cell periphery, and also as short central stress fibers. The central stress fibers were more frequent and prominent in cells of the abdominal aorta. At branch sites and at the aortic bifurcation, long, thick microfilament bundles were present in endothelial cells extending from the tip of the flow divider to a few millimeters along the branch arteries, the aorta, and the iliac arteries. Peripheral actin, however, no longer completely surrounded the cells. The thick bundles were not prominent in endothelial cells located adjacent to the proximal lip of branches or at the iliac arteries opposite the flow divider. This study shows that endothelial cell F-actin microfilament distribution in vivo is well defined along the aortic-arterial system. The prominent central microfilament bundles and the reduced peripheral microfilaments seen at localized regions may reflect an adaptive response to elevated shear stress at these sites.

The reorganization of microfilaments, centrosomes, and microtubules during in vitro small wound reendothelialization

The repair of small endothelial wounds is an important process by which endothelial cells maintain endothelial integrity. An in vitro wound model system was used in which precise wounds were made in a confluent endothelial monolayer. The repair process was observed by time-lapse cinemicrophotography. Using fluorescence and immunofluorescence microscopy, the cellular morphological events were correlated with the localization and distribution of actin microfilament bundles and vinculin plaques, and centrosomes and their associated microtubules. Single to four-cell wounds underwent closure by cell spreading while wounds seven to nine cells in size closed by initially spreading which was then followed at approximately 1 h after wounding by cell migration. These two processes showed different cytoskeletal patterns. Cell spreading occurred independent of centrosome location. However, centrosome redistribution to the front of the cell occurred as the cells began to elongate and migrate. While the peripheral actin microfilament bundles (i.e., the dense peripheral band) remained intact during cell spreading, they broke down during migration and were associated with a reduction in peripheral vinculin plaque staining. Thus, the major events characterizing the closure of endothelial wounds were precise in nature, followed a specific sequence, and were associated with specific cytoskeletal patterns which most likely were important in maintaining directionality of migration and reducing the adhesion of the cells to their neighbors within the monolayer.

Migration-related changes in the cytoskeleton of cultured neural crest cells visualized by the monoclonal antibody I-5G9

An epitope recognized by the monoclonal antibody I-5G9 was expressed by all neural crest cells shortly after explantation into culture. At this time all neural crest cells actively migrated away from the neural tube. Immunoreactivity was localized intracellularly and organized into stress fiber-like filaments. Often, immunofluorescence was particularly high in short fibers in the lamellipodia of the leading edge of migrating cells. Two-week-old cultures had a diameter of 8-10 mm. At that stage a ring of immunoreactive cells was present at the periphery of each culture, an area where cells were still migratory. An inner concentric circle had reduced and more granular staining. In this area cells had ceased to migrate. In the center of the culture cells were multilayered, nonmigratory, and did not bind I-5G9. After creating a lesion in the nonreactive central region, some cells resumed migration into the lesioned area and reexpressed the epitope. I-5G9 staining and phalloidin fluorescence colocalized partially in some cells and completely in others. It is concluded that the epitope recognized by I-5G9 is expressed in a migration-dependent manner. The partial colocalization of I-5G9 and phalloidin fluorescence supports the notion that the epitope recognized by I-5G9 is specifically expressed in stress fibers of migratory cells, possibly in one of the actin-associated proteins or an F actin-associated protein complex.

Urokinase-type plasminogen activator is induced in migrating capillary endothelial cells

Cellular migration is an essential component of invasive biological processes, many of which have been correlated with an increase in plasminogen activator production. Endothelial cell migration occurs in vivo during repair of vascular lesions and angiogenesis, and can be induced in vitro by wounding a confluent monolayer of cells. By combining the wounded monolayer model with a substrate overlay technique, we show that cells migrating from the edges of an experimental wound display an increase in urokinase-type plasminogen activator (uPA) activity, and that this activity reverts to background levels upon cessation of movement, when the wound has closed. Our results demonstrate a direct temporal relationship between endothelial cell migration and uPA activity, and suggest that induction of uPA activity is a component of the migratory process.

In situ localization of F-actin microfilaments in the vasculature of the porcine retina

The organization of F-actin microfilaments in the vascular endothelium of the porcine retina was studied in situ using rhodamine phalloidin labelling and fluorescence microscopy. A comparison was made between arterial and venous endothelial-cell microfilament distribution. The arterial cells in straight segments, bifurcations and branch points were elongated with their long axis in the direction of flow. Venous endothelial cells, on the other hand, were ellipsoid to rhomboid in shape throughout. F-actin was localized at the periphery of both arterial and venous endothelial cells. Prominent central microfilament bundles, similar to in vitro stress fibres, were oriented parallel to the long axis of arterial cells but were rarely present in venous cells. Only occasional venous endothelial cells contained short central actin filaments which were mainly in the venules. Central microfilaments were not identified in pre-capillary, capillary, or post-capillary endothelial cells. Incubation of the retinal organ cultures for 24 hr resulted in loss of the central microfilaments while peripheral staining persisted. Short-term incubation of the retinas in organ culture with low-dose cytochalasin B resulted in disruption of the central microfilaments while the peripheral actin microfilaments remained intact. The central microfilament bundles may reflect an adaptive response to arterial blood flow and may indeed be a sensitive dynamic system reflecting the influence of environmental factors on endothelial cells.

Stimulation of smooth muscle cell glycosaminoglycan synthesis by cultured endothelial cells is dependent on endothelial cell density

Smooth muscle cells (SMC) cultured in the presence of endothelial cells (EC), or in EC-conditioned medium, show increased synthesis of glycosaminoglycans (GAG). We have found that both the amount and type of GAG produced by the SMC are dependent on the density of the EC. EC (porcine) at a low density (0.1-0.5 X 10(6) cells/25 cm2), or their conditioned media, where the most active per cell in stimulating GAG. All GAG were stimulated but the increase was due mostly to hyaluronic acid (HA). At intermediate densities (1.0 X 10(6)/25 cm2) stimulation was markedly reduced, but still present, and both HA and sulphated GAG were similarly increased. At high densities (1.5-3 X 10(6)/25 cm2) where EC were confluent there was very little stimulation of HA but continued stimulation of sulphated GAG synthesis. The shift in stimulation from HA to sulphated GAG with increasing density was most clearly demonstrated by the decrease in the HA to the chondroitin sulphate ratio. These findings provide support for the general concept that SMC metabolism may be affected by changes in the state of the endothelium.

Mechanisms of cytoskeletal regulation: modulation of aortic endothelial cell protein band 4.1 by the extracellular matrix

The bovine aortic endothelial cell (BAEC) cytoskeleton is a complex structure modulated by many stimuli including release from contact inhibition and various components of the extracellular matrix (ECM). Transduction of information from the ECM to the cell nucleus proceeds via several complex pathways including the cytoskeleton. We have demonstrated the presence of an immunoreactive isoform of the human erythrocyte cytoskeletal protein band 4.1 (4.1) in BAEC. BAEC 4.1 is similar in molecular weight to the erythroid protein by immunoblot analyses and produces a similar pattern of cysteine specific cleavage products consistent with a cluster of cysteine residues previously described in the erythroid molecule. We have also examined the effects of defined ECM proteins on the distributions of cultured BAEC 4.1 and actin filaments (AF) at confluency and following release from contact inhibition. The distribution of 4.1 in BAEC on a plasma fibronectin substrate is complex, having partial codistribution with cytoplasmic AF and a unique perinuclear staining. In contrast, on a collagen type I/III substrate, 4.1 is localized, in part, to peripheral areas of cell-cell contact distinct from the dense peripheral band staining of AF. During migration on this substrate, 4.1 had a filamentous distribution having partial codistribution with AF. Indirect immunofluorescence staining of cross-sections of bovine calf aortae revealed a cortical staining pattern in the aortic endothelial cells with staining noted on the luminal and basolateral aspects of the cells. These data suggest that, in endothelial cells, protein 4.1 is a cortical membrane protein which may function to link actin filaments to other skeletal proteins such as spectrin. These findings also suggest an active role for protein 4.1 in cytoskeletal reorganization events which can occur in response to external stimuli, such as the extracellular matrix or contact with other cells.

Endothelial cell monolayer integrity. I. Characterization of dense peripheral band of microfilaments

Although the endothelial cell (EC) cytoskeleton has been studied both in vitro and in vivo, little is known about its role in endothelial integrity. We have previously suggested that specific EC microfilament (MF) structure, which we have termed the "dense peripheral band (DPB)," may play a major role in this process. We have extended our studies to characterize this structure in pig aortic ECs in vitro. During the growth of EC cultures, the DPB appears only when the cultures have attained confluency. Using double fluorescent labeling, we found that alpha-actinin, myosin, and tropomyosin colocalized with the F-actin making up the DPB. Occasional microtubules were present in this region, although there was no preferred association between microtubules and the DPB. Colocalization studies revealed vinculin plaques at the cell-cell interface. Thin MFs extended from the DPB into the cytoplasmic side of these plaques. The DPB was completely disrupted by low dose cytochalasin B within 30 minutes, whereas many central MF bundles were still present at 24 hours. The results of this study suggest that the DPB is a distinct structure in the confluent EC monolayer and is closely associated with the ability of ECs to form and maintain the EC monolayer. The disruption of the DPB as an important initial event in the pathogenesis of vascular diseases such as atherosclerosis is discussed.

In situ localization of F-actin in the normal and injured guinea-pig tympanic membrane

Although cell migration is an important function of the epithelial cells of the tympanic membrane (TM), little is known about the distribution of the F-actin cytoskeleton, a contractile protein important in cell motility. The purpose of this experiment was to study the in situ localization of F-actin in the epithelial cells of the TM. F-actin, localized using Rhodamine-phalloidin, was present as a thin cortical band at the margin of both the mucosal cells on the inner side of the drum, and the suprabasal cells of the epidermis. The basal cells showed diffuse circumferential F-actin staining sometimes appearing as short microfilaments. Following a full thickness injury, changes in the distribution of F-actin could be observed with in situ localization. While the diffuse F-actin staining of the basal cells was reduced, both long F-actin microfilament bundles extending parallel to the long axis of the cell and focal aggregates of F-actin were prominent. The suprabasal cells became elongated, and while the F-actin remained localized to the cell margin, faint central F-actin microfilaments were observed. The staining of the mucosal cells remained unchanged. This study showed that the guinea pig TM is a useful model to study the distribution of epithelial F-actin in situ under normal and repair conditions, and that the basal cell layer may be important in regulating migration in the epidermis.

The distribution of centrosomes in endothelial cells of non-wounded and wounded aortic organ cultures

The distribution of centrosomes in porcine vascular endothelial cells of the thoracic aorta maintained in organ culture was determined in en face preparations using immunofluorescence. Rectangular pieces of aorta that had the distal half (with respect to the heart) of their endothelial surface gently denuded with a scalpel blade and pieces with intact endothelium were cultured for up to 96 h. At time 0, centrosomes were found to be preferentially oriented toward the heart, both in the cells of intact monolayers and in cells at the wound edge. This distribution was maintained in the intact monolayers for at least 24 h, but by 72 h the number of centrosomes in the center of the cells exceeded the number oriented toward the heart as the cells changed from a fusiform to a polygonal shape. The centrosomes of most endothelial cells at the wound edge began to redistribute themselves within the first 24 h in culture, moving from a position toward the heart to a position either in the center of the cell or away from the heart. By 72 h, the majority of centrosomes in endothelial cells at the wound edge were oriented away from the heart toward the denuded region. It is concluded that the centrosomes in the endothelial cells maintained in organ culture respond to injury in a manner similar to those grown in monolayer cell culture except that the reorientation of centrosomes occurs more slowly.

Vasoactive amines modulate actin cables (stress fibers) and surface area in cultured bovine endothelium

Cultured bovine aortic endothelial cells were fixed and stained with NBD-phallicidin and quantitated with a digital image analyzer for changes in actin cables and surface area. Serotonin (5-HT), norepinephrine (NE), dopamine and histamine (all at 10(-4)M concentrations) were tested for their ability to induce cytoskeletal changes. Only 5-HT and NE increased actin cables significantly (p less than 0.01), 80.7% and 97.9%, respectively. Dopamine and histamine treated cells showed a 67.4% and 80.8% decrease in actin cables respectively (p less than 0.01). Stimulated increases of actin cables by 5-HT were inhibited by Ketanserin, and propranolol inhibited NE stimulation of actin cables. Treatment of cells with these blockers alone also decreased actin cables below control values (p less than 0.01). Pretreatment of cells with diphenhydramine, but not cimetidine, inhibited histamine-induced decreases in actin cables. Stimulation of surface area by 5-HT and NE was also observed, with 40.8% and 80.7% increases respectively, when compared with controls (p less than 0.01). The increases in actin cables were associated with a lack of ruffled edges that are indicative of motile cells. In contrast, induced decreases in actin cables resulted in cells with ruffled edges. Exogenous 5-HT and NE have been shown to prevent the increased permeability visible as extravasation of red blood cells from postcapillary venules in thrombocytopenic animals. The present data suggest that 5-HT and NE may be involved in maintaining the endothelial barrier function by a receptor-mediated stimulation of actin cables. Also, histamine-induced decreases in actin cables may be correlated with the amine's action in vivo as a mediator of increased inflammatory permeability.

Actomyosin organization in stationary and migrating sheets of cultured human endothelial cells

We used immunofluorescence microscopy to study the organization of actin, myosin and vinculin in confluent endothelial cells and in cells migrating into an experimental wound and interference reflection microscopy to assess the cell-substratum adhesion pattern in these cells. In confluent stationary endothelial cell monolayers actin showed a distinct cell-to-cell organization. Myosin, on the other hand, was diffusely distributed and was clearly absent from cell peripheries. Vinculin was confined as linear arrays to cell-cell contact areas. Interference reflection microscopy revealed areas of close and distant adhesion but no focal adhesion sites in these cultures. Twelve hours after experimental wounding a distinct zone of advancing cells was seen at the wound edge. These cells showed a spreadout morphology and, in contrast to stationary cells, had a stress fibre-type organization of both actin and myosin. Vinculin was in the migrating cells seen as plaques at the ventral cell surface. In interference reflection microscopy numerous focal adhesions were seen. The results indicate that the actomyosin system forms the structural basis for monolayer organization of endothelial cells and responds by reorganization upon cell migration.