Histamine increases microvascular permeability by creating small transitory (100–400 nm) gaps between adjacent endothelial cells at sites of vascular endothelial (VE)-cadherin-based adhesion. We examined the effects of histamine on the proteins within the VE-cadherin-based adherens junction in primary human umbilical vein endothelial cells. VE-cadherin is linked not only by β- and α-catenin to cortical actin but also by γ-catenin to the intermediate filament vimentin. In mature human umbilical vein cultures, the VE-cadherin immunoprecipitate contained equivalent amounts of α- and β-catenin, 130% as much β- as γ-catenin, and 50% as much actin as vimentin. Within 60 s, histamine decreased the fraction of VE-cadherin in the insoluble portion of the cell lysate by 35 ± 1.5%. At the same time, histamine decreased the amount of vimentin that immunoprecipitated with VE-cadherin by 50 ± 6%. Histamine did not affect the amount of actin or the amount of α-, β-, or γ-catenin that immunoprecipitated with VE-cadherin. Within 60 s, histamine simulated a doubling in the phosphorylation of VE-cadherin and β- and γ-catenin. The VE-cadherin immunoprecipitate contained kinase activity that phosphorylated VE-cadherin and γ-catenin in vitro.
- vascular endothelial cadherin
histamine increases microvascular permeability in postcapillary venules (37). Early studies from Majno and Palade (21) indicated that one of the mechanisms by which histamine increased venular permeability was the transient development of gaps between adjacent endothelial cells. McDonald and colleagues (22) quantitated these changes and measured gaps of 100–400 nm, which have a half-life of ∼1.9 min. Restriction of the edemagenic effects of histamine to venules may reflect the high concentration of histamine receptors in venules relative to other vascular segments (15).
Activation of the histamine H1 receptor caused similar small gaps to develop between adjacent human umbilical vein endothelial (HUVE) cells (5, 6, 24). These decreases in cell-cell adhesion occurred in the absence of any detectable increases in tension within the cells (23). The absence of an increase in tension suggested that histamine might cause gaps to form by altering cell adhesion, unbalancing the forces resisting the resting tension within the cells. Histamine could affect adhesion at sites of apposition between adjacent cells or at sites of apposition of the ventral surface of the cells to their matrix. Histamine altered adhesion at sites between adjacent cells (24).
In endothelial cells, cadherin-5, or vascular endothelial (VE)-cadherin, is an important constituent of cell-cell adhesion (7). We hypothesized that histamine acted at sites of VE-cadherin-based adhesion. ECV304 cells respond to histamine with an increase in cell calcium, similar to endothelial cells (35). ECV304 cells do not express VE-cadherin, and histamine did not change the adhesion of control ECV304 cells. However, histamine decreased cell-cell adhesion of ECV304 cells transfected with VE-cadherin by 60%. The decrease in adhesion followed a time course similar to that observed in HUVE cells, which is similar to the kinetics of gap formation and resolution in situ (22, 35). Hence, histamine alters VE-cadherin based cell-cell adhesion.
Changes in adherens junction adhesion are frequently associated with changes in phosphorylation of elements of the adherens junction complex (8). Assessing phosphorylation of the VE-cadherin complex is complicated by the fact that tyrosine phosphorylation of elements of the complex decreases with maturation of the cells (19). Several authors have reported that agonists that alter endothelial adhesion initiate increases in tyrosine phosphorylation of VE-cadherin, p120, and β- and γ-catenin in endothelial cells (1, 9,32). However, none of these authors could detect increases in phosphorylation of VE-cadherin or the catenins after exposure to the agonists without pretreating the cells with inhibitors of tyrosine phosphatases. More recent reports demonstrate that the phosphatase inhibitors, themselves, increase phosphorylation and alter adhesion. Hence, the physiological inference of the observations made in the presence of the inhibitors is uncertain (12, 36).
Andriopoulou et al. (1) reported that histamine decreased the amount of VE-cadherin in the detergent-insoluble fraction of the cell lysate, implying a decrease in VE-cadherin association with the cytoskeleton. However, no one has detected a change in the amount of any of the proteins that are associated with VE-cadherin when it is immunoprecipitated from cells exposed to thrombin, histamine, or vascular endothelial growth factor (VEGF) (1, 9, 32). Hence, although there is general agreement that VE-cadherin-based adhesion is altered, there are not currently any data to explain how it is altered.
The classic type I cadherins, E-, N-, and P-cadherin, are linked to cortical actin by β- and α-catenin (2). VE-cadherin is a type II cadherin, and it is linked not only by β- and α-catenin to cortical actin but also by γ-catenin and desmoplakin to the intermediate filament protein vimentin (18, 33). The importance of the linkage to vimentin is highlighted by the fact that desmoplakin-deficient mice do not survive embryogenesis. They have a defective microvasculature and resemble VE-cadherin-deficient mice (10, 11). When endothelial cell cultures are wounded, VE-cadherin complexes with γ-catenin respond to the wounding and facilitate cell movement much more than β-catenin-containing VE-cadherin complexes (20). This raises the possibility that γ-catenin-based VE-cadherin complexes respond more to agonists that affect cadherin-based adhesion.
In the experiments reported in this manuscript we used 32P labeling to assess changes in phosphorylation of elements of the VE-cadherin complex after exposing HUVE cells to histamine, and we did not pretreat the cells with phosphatase inhibitors. We used [35S]methionine labeling to determine changes in protein binding within the VE-cadherin complex. We used only mature cultures (≥4 days postconfluence), and we examined these events 60 s after initiating exposure to histamine, a physiologically relevant time when adhesion has reached its nadir (23, 24). We found that histamine stimulates an increase in phosphorylation of VE-cadherin, β-catenin, γ-catenin, and vimentin within 60 s. Consistent with the observations of Andriopoulou et al. (1), we also found that histamine decreased the amount of VE-cadherin that was in the detergent-insoluble, cytoskeleton-associated fraction of the cell lysate. We did not detect a change in the amount of any of the catenins or actin present within the VE-cadherin immunoprecipitate, but we did detect a 50% decrease in the amount of vimentin that was present. These observations suggest that histamine primarily affects the interaction of VE-cadherin with the vimentin cytoskeleton.
Fibronectin was from Collaborative Research (Bedford, MA). Tissue culture media and serum were from the Tissue Culture Core, University of Iowa. Antibody to VE-cadherin (mouse IgG, monoclonal, clone 55-7H1) was from PharMingen. Antibody to E-cadherin was clone rr1 from the Hybridoma Facility. Antibodies to β-catenin and γ-catenin were rabbit polyclonals directed to the epitopes described by Hinck et al. (16). Antiphosphotyrosine was from Upstate Biotechnology (Lake Placid, NY). Secondary antibodies were from Amersham. [32P]ATP and [γ-32P]ATP were from New England Nuclear.
Cultured HUVE cells were prepared by collagenase treatment of freshly obtained human umbilical veins and cultured. Cultures were identified as endothelial by their characteristic uniform morphology, uptake of acetylated low-density lipoprotein, and by indirect immunofluorescent staining for factor VIII (35).
Immunoprecipitation and blotting.
For immunoprecipitations, cells were grown on Transwell filters for 5–7 days. In some experiments cells were placed in methionine-depleted Selectamine (DuPont NEN) media with 1 mCi per Transwell of [35S]methionine (DuPont NEN) 36 h before the experiment. In other experiments, cells were placed in phosphate-free media, and 0.8 mCi of 32P as inorganic phosphate was added to the media 3 h before the experiment. Cells were washed and then exposed to histamine for the indicated times. After exposure, cells were lysed and immunoprecipitated as described by Hinck et al. (16). Cells were immediately lysed in 4°C CSK buffer [50 mM NaCl, 10 mM PIPES, pH 6.8, 3 mM MgCl2, 1% Triton X-100, 300 mM sucrose, 5 mg/ml phenylmethylsulfonyl fluoride (PMSF), 5 μg/ml aprotinin, 1 mg/ml leupeptin, 5 μg/ml pepstatin A, 4 mM Na orthovanadate, 10 mM NaF, and 10 mM NaP2O7] and kept on ice for 20 min. The cells were scraped from the filter and added to the CSK buffer, which was centrifuged at 12,400 g for 10 min. Soluble material was transferred to a separate tube for immunoprecipitation.
The soluble fraction was precleared with 30 μl of Sepharose (Sigma CL 4B200) and 5 μl of gamma bind (Pharmacia 17-0886-01) preincubated with 250 μg of mouse IgG in a total volume of 100 μl of Tris-buffered saline [TBS (10 mM Tris, pH 7.5, 150 mM NaCl, 2 mM MgCl2, 2 mM CaCl2, 1 mM PMSF, 2 μg/ml aprotinin, 15 μg/ml leupeptin, and 360 μM 1.10-phenanthroline)] for 2 h at 4°C. The beads were pelleted, and the precleared solutions were then incubated with 30 μl of Sepharose and 5 μl of gamma bind that had been preincubated in 500 μl of anti-cadherin-5 (VE-cadherin) media for 2 h at 27°C and then rinsed in 500 μl of TBS. Samples were rotated overnight with anti-cadherin-5-Sepharose-gamma bind at 4°C. Beads were pelleted at 12,400 g for 2 min at 4°C. Beads were then resuspended in 1 ml of high-stringency buffer (15 mM Tris, pH 7.5, 5 mM EDTA, 2.5 mM EGTA, 1% Triton X-100, 1% Na-deoxycholate, 0.1% SDS, 120 mM NaCl, and 25 mM KCl) underlain with 100 μl of 10% sucrose and centrifuged at 12,400 g for 2 min at 4°C (16). The pellet was then washed in high-salt buffer (15 mM Tris, 5 mM EDTA, 2.5 mM EGTA, 1% Triton X-100, 1% Na-deoxycholate, 0.1% SDS, and 1 M NaCl) followed by low-salt buffer (15 mM Tris, pH 7.5, and 5 mM EDTA).
For a selected set of experiments, the cells were preexposed to peroxide and pervanadate for 7 min before being exposed to histamine as described by Andriopoulou et al. (1). These experiments were used to determine the effects of peroxide and pervanadate, alone, on phosphorylation, and they confirmed the observations of Garcia et al. (12). The data from these experiments are not presented in the text.
Immunoprecipitated proteins were separated on 8% polyacrylamide gels. The proteins in the gels were transferred to polyvinylidene difluoride (PVDF) membranes. Autoradiograms were generated from the [35S]methionine-labeled transfers and quantitated by densitometry. Proteins were identified by immunoblotting, the same transfers with the indicated primary and secondary antibodies. Blots were quantitated by densitometry. In assessing the effects of histamine on the 35S mass of proteins immunoprecipitated with VE-cadherin, we normalized the mass of individual proteins (as assessed by the 35S signal) to the number of methionines and then divided by the mass of VE-cadherin immunoprecipitated in the same sample, so that each sample was normalized to itself and to methionine content.
In assessing the changes in phosphorylation, we quantitated the32P activity associated with individual proteins on the transfers by densitometry of an autoradiogram developed from the transfer. Again, proteins were identified by blotting the transfers. The mass of VE-cadherin in each transfer was quantitated using phosphorimaging (Molecular Dynamics) of a Western blot for VE-cadherin. The 32P activity for each protein in the sample was normalized to the mass of VE-cadherin in the sample estimated from this blot. This ensured that unequal loading did not contribute to estimating changes in phosphorylation.
Briefly, for the soluble kinase assay, immunoprecipitates were prepared as above but washed 2× in kinase buffer (20 mM MgCl2, 25 mM HEPES, 20 mM β-glycerophosphate, 20 mM p-nitrophenyl phosphate, 0.1 mM NaVO4, 2 mM dithiothreitol, 1.25 mM EGTA, and 1.15 mM CaCl2) after the low-salt wash. Individual samples were then resuspended in 20 μl of kinase buffer with 20 μM ATP, 5 μCi [γ-32P]ATP, and 10 μg myelin basic protein (MBP). Samples were rotated for 30 min at 27°C, and the reaction was terminated with 50 μl of 2× SDS sample buffer. Samples were heated to 95°C for 5 min and centrifuged to pellet the beads, and proteins in the supernate were separated on polyacrylamide gels. Gels were transferred to PVDF membranes, which were then used for autoradiography and Western blotting.
For the in-gel assay, immunoprecipitates were prepared as described above. After the low-salt wash, beads were placed in SDS sample buffer and boiled for 5 min. The proteins released from the beads were separated in a thin 6% polyacrylamide gel with 0.33 mg/ml MBP incorporated into the gel. The gel was washed and renatured overnight and then incubated with kinase buffer (25 mM HEPES, 10 mM MgCl2, 90 μM Na vanadate, and 5 mM β-mercaptoethanol), 26 μM ATP, and [γ-32P]ATP for 1 h. The reaction was stopped with 5% TCA in 1% disodium pyrophosphate. An autoradiogram was generated from the gel.
Differences among groups were compared by analysis of variance, and individual group comparisons were done using a Tukey honestly significant difference test for post hoc comparisons of means. Differences were considered significant at the P < 0.05 level.
Proteins in the VE-cadherin immunoprecipitate.
The VE-cadherin immunoprecipitate from HUVE cells more than 4 days postconfluence contains α-, β-, γ-catenin, and consistent amounts of actin and vimentin (Fig. 1). We quantitated autoradiograms of the VE-cadherin immunoprecipitate from [35S]methionine-labeled cells for each of these proteins with densitometry and normalized the values for the amount of VE-cadherin in the immunoprecipitate and for the number of methionines in the respective protein. There were equivalent amounts of α- and β-catenin in the VE-cadherin immunoprecipitate (Fig.2). There was 68 ± 0.02% as much γ- as β-catenin, and there was 52 ± 0.05% as much actin as vimentin (Fig. 2).
Histamine decreased the cytoskeletal association of VE-cadherin and the amount of vimentin in the VE-cadherin immunoprecipitate.
Histamine decreases endothelial cell-cell adhesion at sites of VE-cadherin-based adhesion. To begin to understand how histamine alters VE-cadherin-based adhesion, we examined the effects of histamine on the association of VE-cadherin with the cytoskeleton and on the binding of proteins within the VE-cadherin immunoprecipitate.
Histamine decreased the fraction of VE-cadherin that was present in the detergent-insoluble (cytoskeletal associated) portion of the cell lysate by 35 ± 1.5% (Fig. 4). Histamine did not change the amount of α-, β-, or γ-catenin that immunoprecipitated with VE-cadherin (Fig.5, n = 33 control and 33 histamine-stimulated monolayers, six separate experiments). Histamine also did not change the amount of actin that immunoprecipitated with VE-cadherin [−16 ± 8%, P = not significant], but histamine decreased the amount of vimentin that immunoprecipitated with VE-cadherin by 50 ± 6% (Fig.6, n = 15 control and 15 histamine-stimulated monolayers, three separate experiments). The decrease in vimentin was evident in both the autoradiograms from35S-labeled cells and when the immunoprecipitate was blotted for vimentin (Fig. 6).
Histamine increased phosphorylation of VE-cadherin, β-catenin, γ-catenin, and vimentin.
Histamine increased phosphorylation of VE-cadherin, β- and γ-catenin, and vimentin that immunoprecipitated with VE-cadherin. After 60 s, the 32P activity of VE-cadherin, β-catenin, γ-catenin, and vimentin that immunoprecipitated with the VE-cadherin increased by two to five times above control levels (Fig.7, n = 16 separate control and 16 histamine-stimulated monolayers). It was not possible to detect phosphorylation of the trace amount of VE-cadherin-associated p120, consistent with other reports (19). These increases in phosphorylation occurred without pretreatment of the cells with phosphatase inhibitors, indicating that they are physiological changes within the first 60 s of the cells' exposure to histamine. For each sample, the 32P activity in each of the proteins was normalized for the mass of VE-cadherin in the same sample. The increase in vimentin phosphorylation was evident even though there was ∼50% less vimentin present.
Blots of immunoprecipitate from control and histamine-stimulated cells not pretreated with peroxide and vanadate did not demonstrate a difference in tyrosine phosphorylation of the immunoprecipitated proteins, consistent with reports of others (data not shown) (1,12). When HUVE cells were pretreated with peroxide and vanadate, there was a significant increase in tyrosine phosphorylation of VE-cadherin in control cells, also consistent with reports from others (data not shown) (12).
Kinase activity in the cadherin-5 immunoprecipitate.
Because histamine increased phosphorylation of proteins in the VE-cadherin immunoprecipitate, we examined the VE-cadherin immunoprecipitate for kinase activity. An autoradiogram of a transfer of proteins separated on a PAGE gel after a soluble kinase assay demonstrated phosphorylation of MBP (Fig.8).
The identity of some of the phosphorylated bands other than MBP was investigated by blotting the transfers used for the autoradiograms for components of the VE-cadherin complex. One of the additional bands migrated at a molecular mass (MM) of ∼130 kDa, and this band labeled with antibody to VE-cadherin (Fig. 8). The larger of the other two bands migrated with an MM of 80–90 kDa, and it reacted with antibody to γ-catenin (Fig. 9). The identity of the bands that migrate just above MBP and at ∼60 kDa are unknown (see Fig. 8).
To begin to characterize the kinase activity, we examined the ability of proteins from the immunoprecipitate, separated on a 6% PAGE gel, to phosphorylate MBP that was incorporated into a gel. We found a dominant band of activity that migrated between the 46- and 66-kDa markers, a fainter band that migrated below the 97-kDa marker, and a very faint band below the 46-kDa marker (Fig. 10).
Adhesive interactions between cells are essential for morphogenesis and homeostasis. For both purposes the adhesive interactions must be dynamic (14). Cadherins are calcium-dependent intercellular adhesion molecules with a single transmembrane domain and five extracellular repeats that mediate calcium-dependent homophilic binding to adjacent cells. The cadherin cytoplasmic domain binds catenins and thereby mediates attachment to the cortical cytoskeleton (3, 8, 14, 16, 25, 27, 30, 34,38).
The response of endothelium to histamine is a rapid, dynamic loss and restoration of cadherin-based adhesion (35). In this study, we rigorously demonstrate that histamine stimulates phosphorylation of elements of the VE-cadherin complex. More importantly, we observed that histamine alters the link of VE-cadherin to the cytoskeleton within the linkage to the intermediate filament protein vimentin. At the same time, we detected no evidence that histamine alters the linkage of VE-cadherin to actin. An effect of histamine on VE-cadherin linkage to vimentin, but not actin, has significant implications for the mechanisms by which histamine alters VE-cadherin-based adhesion.
VE-cadherin is a type II cadherin, retaining some of the properties of classical type I cadherins but also demonstrating significant differences (2). As indicated in the Introduction, VE-cadherin is linked not only to cortical actin, but also to the intermediate filament vimentin (2, 4, 18, 33). Our data shed some light on the magnitude of these linkages.
The estimated molar ratio of α- and β-catenin in all (before and after histamine) the immunoprecipitates of VE-cadherin from HUVE cells was 1:1. This suggests that either very little α-catenin is linked to γ-catenin or β-catenin is linked to another protein in addition to α-catenin. If all of the β-catenin is linked to α-catenin, and none of the γ-catenin is linked to α-catenin, then almost all of the γ-catenin could be linked to vimentin, the linkage altered by histamine.
The VE-cadherin immunoprecipitates from mature HUVE cell cultures contained ∼70% as much γ-catenin as β-catenin. Lampugnani and colleagues (19) previously reported that immature HUVE cultures were rich in β-catenin, whereas mature cultures were enriched in γ-catenin. They estimated the relative amounts of the two catenins by immunoblotting rather than metabolic labeling, and their blot suggests even less β-catenin in mature cultures than do our metabolic labeling data. These differences may reflect differences between primary (ours) and passaged HUVE cells (theirs). When we examined passaged human dermal microvessel cells, they had 130% as much γ-catenin as β-catenin, indicating that the relative amounts of the proteins can vary among cell types (data not shown). Regardless, in both cells there is abundant potential for linking VE-cadherin to the intermediate filament vimentin.
The VE-cadherin immunoprecipitates also contained two lower MM bands, one at 58 kDa and one at 41 kDa. These bands blotted with antibodies to vimentin and actin, respectively. Because both vimentin and actin exist as multimeric proteins, it was not possible to make a good estimate of how many actin or vimentin molecules bound to a single VE-cadherin molecule. However, after normalizing for methionine and the amount of VE-cadherin in the same immunoprecipitate, we found about two times as much vimentin as actin, indicating that the linkage to the intermediate filament vimentin is very insignificant.
In mature endothelial cultures, Lampugnani et al. (19) found that very little p120 was bound to VE-cadherin, and it was never tyrosine phosphorylated. We also found only trace amounts of p120 that immunoprecipitated with VE-cadherin. The limited amount of p120 bound to VE-cadherin may relate to the recent observation that the new armadillo protein p0071 appears to compete with p120 for binding to the same juxtamembrane domain on VE-cadherin. Interestingly, p0071 also may link VE-cadherin to desmoplakin and vimentin (4).
We found that histamine decreased the amount of VE-cadherin that was present in the detergent-insoluble phase of the cell lysate by ∼40%. This is similar to the earlier observations of Andriopoulou et al. (1). We did not detect any changes in the amounts of α-, β-, or γ-catenin that immunoprecipitated with VE-cadherin. However, while examining blots from 35S-labeled cells exposed to histamine for 60 s, we consistently detected a large (∼50%) decrease in the band at 58 kDa, which blots with antibody to vimentin. Another band at 41 kDa, which comigrates with a band that blots for actin, did not change after exposing the cells to histamine.
Gallicano and colleagues (10, 11) have established an essential role for desmoplakin and the intermediate filament cytoskeleton in the developing microvasculature. Kowalczyk et al. (4, 18) has demonstrated important roles for γ-catenin (plakoglobin) and now p0071 in recruiting desmoplakin to the VE-cadherin complex. We had taken special care to preclear all the lysates with nonspecific mouse IgG bound to the same type of beads that were subsequently used for the VE-cadherin immunoprecipitation. Despite this, we initially had some concerns about the specificity of the vimentin in the VE-cadherin precipitates. However, the dramatic change in vimentin binding, but not in actin binding, after histamine could not be explained on the basis of nonspecific binding within the immunoprecipitation, and it provided new information indicating that the effects of histamine target the linkage of VE-cadherin to vimentin.
The observation that histamine primarily affects the link to the vimentin cytoskeleton is also consistent with an earlier observation. When mature endothelial cell cultures were wounded, γ-catenin (plakoglobin)-containing complexes near to, but not within, the wound were altered, but β-catenin-containing complexes were not (20). This would be consistent with cytokines preferentially acting on γ-catenin-based complexes.
We do not yet know how histamine alters the link between VE-cadherin and vimentin. Gaudry et al. (13) and Stappenbeck (29) have reported that tyrosine phosphorylation of γ-catenin (plakoglobin) reduces its ability to bind desmoplakin and that serine phosphorylation of desmoplakin reduces its ability to bind vimentin. In addition, the structure and polymerization of vimentin, itself, is affected by phosphorylation (17, 31). The new observations of Calkins et al. (4) that p0071 links VE-cadherin and desmoplakin offers additional potential mechanisms that might alter the link of VE-cadherin to vimentin.
Andriopoulou et al. (1) reported that histamine increased tyrosine phosphorylation of elements of the VE-cadherin complex in endothelial cells. Other authors reported similar effects of VEGF and thrombin (1, 9, 32). However, each of these studies found that it was not possible to detect changes in tyrosine phosphorylation unless the endothelial (20) cells were pretreated with pervanadate. More recently, Garcia et al. (12) found that similar treatment with pervanadate altered endothelial cell adhesion and increased phosphorylation of the same proteins. The results Garcia et al. raised questions about the physiological inference of the prior observations. Did the increase in tyrosine phosphorylation caused by the agonists detected in the presence of pervanadate represent a physiological response, or did it reflect results that would occur only with aberrant regulation? The observations by Wong et al. (36) that VEGF and histamine decreased, and not increased, phosphorylation of p120 in HUVE cells heightens this concern.
In our experiments, histamine increased phosphorylation of VE-cadherin and β- and γ-catenin in the VE-cadherin complex in HUVE cells. We measured phosphorylation as the incorporation of 32P into the proteins. We did not pretreat the cells with phosphatase inhibitors, and therefore these increases in phosphorylation represent a physiological response to histamine that occurs with intact phosphatase activity. As others have reported, we could not detect changes in phosphorylation of tyrosines in these proteins in cells not pretreated with tyrosine phosphatase inhibitors. The fact that we detected increased phosphorylation with 32P and not with antiphosphotyrosine suggests either that the 32P is more sensitive or that the phosphorylation is not on tyrosines.
We identified kinase activity in the VE-cadherin immunoprecipitate that phosphorylates VE-cadherin and γ-catenin. We do not yet know if this activity contributes to the effects of histamine on VE-cadherin- based adhesion.
In summary, histamine stimulates phosphorylation of elements of the VE-cadherin complex, and this can be detected without inhibiting phosphatases in the cells. We have confirmed that histamine causes VE-cadherin to move from the detergent-insoluble to the detergent-soluble fraction of the cell lysate. We have made the new observation that histamine affects the link between VE-cadherin and the vimentin cytoskeleton without causing a detectable change in the link between VE-cadherin and the actin cytoskeleton. These observations will help direct future investigations into how histamine alters VE-cadherin-based adhesion.
This work was supported by National Heart, Lung, and Blood Institute Grants HL-335450 and HL-63100.
Address for reprint requests and other correspondence: D. M. Shasby, Univ. of Iowa College of Medicine, Iowa City, IA 52242 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published January 11, 2002;10.1152/ajplung.00329.2001
- Copyright © 2002 the American Physiological Society