Studies on the role of protein kinases in the TNF-mediated enhancement of murine tumor cell-endothelial cell interactions.

We have previously demonstrated that the exposure of mouse microvascular endothelium (MME) to tumor necrosis factor-alpha (TNF) led to the increased binding of mouse mastocytoma cells (P815) to endothelial monolayers (Bereta et al., in press). In the current study we examined the possible involvement of protein kinases in TNF signal transduction in the endothelial cells. PKA does not appear to play a role in the potentiation of binding by TNF. We found that the TNF-generated signal is inhibited by H-7 and sangivamycin, but not by staurosporine. TNF did not cause translocation of PKC to the cell membrane and its effect could not be completely mimicked by PMA nor by PMA in the presence of calcium-raising agents. Thus, we concluded that the "classical" PKC pathway is not completely responsible for TNF signalling in this system. We also found that staurosporine itself strongly enhanced adhesion of tumor cells to endothelium, utilizing a mechanism distinct from that of TNF. Although the data provide evidence for the role of kinases in the effect of TNF on binding of tumor cells to MME, this role appears to be a complex one.     

Protein kinase C-mediated endothelial barrier regulation is caveolin-1-dependent

Protein kinase C (PKC) is activated in response to various inflammatory mediators and contributes significantly to the endothelial barrier breakdown. However, the mechanisms underlying PKC-mediated permeability regulation are not well understood. We prepared microvascular myocardial endothelial cells from both wild-type (WT) and caveolin-1-deficient mice. Activation of PKC by phorbol myristate acetate (PMA) (100 nM) for 30 min induced intercellular gap formation and fragmentation of VE-cadherin immunoreactivity in WT but not in caveolin-1-deficient monolayers. To test the effect of PKC activation on VE-cadherin-mediated adhesion, we allowed VE-cadherin-coated microbeads to bind to the endothelial cell surface and probed their adhesion by laser tweezers. PMA significantly reduced bead binding to 78±6% of controls in WT endothelial cells without any effect in caveolin-1-deficient cells. In WT cells, PMA caused an 86±18% increase in FITC-dextran permeability whereas no increase in permeability was observed in caveolin-1-deficient monolayers. Inhibition of PKC by staurosporine (50 nM, 30 min) did not affect barrier functions in both WT and caveolin-1-deficient MyEnd cells. Theses data indicate that PKC activation reduces endothelial barrier functions at least in part by the reduction of VE-cadherin-mediated adhesion and demonstrate that PKC-mediated permeability regulation depends on caveolin-1.     

Human umbilical vein and dermal microvascular endothelial cells show heterogeneity in response to PKC activation

Changes in endothelial cell (EC) phenotype are central to thefunction of endothelium in inflammation. Although these events mainlyoccur in the microvasculature, previous studies have predominantly usedlarge-vessel EC. Using enzyme-linked immunosorbent and flow cytometricassays, we compared the responses of human umbilical vein endothelialcells (HUVEC) and dermal microvascular endothelial cells (DMEC) to theactivation of protein kinase C (PKC). Stimulation with phorbol12,13-dibutyrate and more selective PKC agonists, including12-deoxyphorbol 13-phenylacetate 20-acetate (dPPA), inducedmorphological changes and proliferation in both EC types. PKCactivation induced a marked increase in Thy-1 expression on DMEC andonly a moderate rise on HUVEC. Furthermore, heterogeneity in theinduction of the adhesion molecules intercellular adhesion molecule 1, vascular cell adhesion molecule 1 (VCAM-1), and E-selectin between thetwo EC types following activation of PKC was demonstrated. Inparticular, E-selectin and VCAM-1 were significantly upregulated onHUVEC but not DMEC. The data indicate that the PKC pathway is unlikelyto be important for E-selectin and VCAM-1 expression in themicrovasculature but are consistent with a role for PKC inangiogenesis. This diversity in signaling in response to PKC activationmay depend on differential utilization of PKC isozymes and mayfacilitate specialized endothelial responses.

     

IL-4 regulates endothelial cell activation by IL-1, tumor necrosis factor, or IFN-gamma.

Alteration in the surface membrane of endothelial cells (EC) is a feature of endothelial activation both at sites of inflammation in vivo and after stimulation with cytokines in vitro. The effects of stimulating EC with IL-1 or TNF include enhanced adhesiveness for polymorphonuclear leukocytes (PMN) and T cells, the induction of EC leukocyte adhesion molecule-1 (ELAM-1) expression, and the increased expression of intercellular adhesion molecule-1 (ICAM-1) and the 1.4C3 Ag. In contrast, IFN-gamma stimulation increases EC binding of T cells but not PMN and enhances ICAM-1 expression but not ELAM-1 or 1.4C3 Ag expression. Recently we have reported that the T cell-derived cytokine IL-4 also increases EC adhesiveness for T cells but not PMN. In this study we have examined the effect of IL-4 on the expression of several cytokine-inducible EC activation Ag, by using a previously described ELISA technique. IL-4 modulation of activation Ag expression was concentration dependent, optimal at around 100 U/ml, and exhibited a unique pattern compared to that seen with the other cytokines. Although, IL-4 stimulation increased 1.4C3 Ag expression (p less than 0.001), it significantly inhibited constitutive ICAM-1 expression (p less than 0.01) and did not induce ELAM-1. Furthermore, IL-4 exhibited significant synergy with IL-1 or TNF in inducing 1.4C3 Ag expression (p less than 0.001) but inhibited the increased expression of ICAM-1 produced by IL-1, TNF, or IFN-gamma (p less than 0.01) and inhibited the induction of ELAM-1 by IL-1 and TNF (p less than 0.001). In contrast, IL-4 had no effect on the expression of EC HLA-class I, -DR, -DP, or -DQ and neither enhanced nor inhibited the effect of IFN-gamma on the expression of these molecules. Finally, although IL-4 alone caused little if any shape change in EC monolayers, it strongly synergized with TNF or IFN-gamma in causing a change in shape to a more fibroblastic morphology. These observations indicate that IL-4 increases EC adhesiveness for T cells by the induction of a different adhesion molecule to ICAM-1. Furthermore, the ability of IL-4 to both enhance and inhibit the expression of activation Ag on EC already activated by IL-1, TNF, or IFN-gamma suggests that it may be important in altering the quality of inflammatory responses such as may occur during the development and maintenance of chronic or immune-mediated inflammation.     

Role of LFA-1 and VLA-4 in the adhesion of cloned normal and LFA-1 (CD11/CD18)-deficient T cells to cultured endothelial cells. Indication for a new adhesion pathway.

Patients with the leukocyte adhesion deficiency (LAD) syndrome have a genetic defect in the common beta 2-chain (CD18) of the leukocyte integrins. This defect can result in the absence of cell surface expression of all three members of the leukocyte integrins. We investigated the capacity of T cell clones obtained from the blood of an LAD patient and of normal T cell clones to adhere to human umbilical vein endothelial cells (EC). Adhesion of the number of LAD T cells to unstimulated EC was approximately half of that of leukocyte function-associated antigen (LFA)-1+ T cells. Stimulation of EC with human rTNF-alpha resulted in an average 2- and 2.5-fold increase in adhesion of LFA-1+ and LFA-1- cells, respectively. This effect was maximal after 24 h and lasted for 48 to 72 h. The involvement of surface structures known to participate in cell adhesion (integrins, CD44) was tested by blocking studies with mAb directed against these structures. Adhesion of LFA-1+ T cells to unstimulated EC was inhibited (average inhibition of 58%) with mAb to CD11a or CD18. Considerably less inhibition of adhesion occurred with mAb to CD11a or CD18 (average inhibition, 20%) when LFA-1+ T cells were incubated with rTNF-alpha-stimulated EC. The adhesion of LFA-1- T cells to EC stimulated with rTNF-alpha, but not to unstimulated EC, was inhibited (average inhibition, 56%) by incubation with a mAb directed to very late antigen (VLA)-4 (CDw49d). In contrast to LAD T cell clones and the LFA-1+ T cell line Jurkat, mAb to VLA-4 did not inhibit adhesion of normal LFA-1+ T cell clones to EC, whether or not the EC had been stimulated with rTNF-alpha. We conclude that the adhesion molecule pair LFA-1/intercellular adhesion molecule (ICAM)-1 plays a major role in the adhesion of LFA-1+ T cell clones derived from normal individuals to unstimulated EC. Adhesion of LFA-1-T cells to TNF-alpha-stimulated EC is mediated by VLA-4/vascular cell adhesion molecule (VCAM)-1 interactions. Since we were unable to reduce significantly the adhesion of cultured normal LFA-1+ T cells to 24 h with TNF-alpha-stimulated endothelium with antibodies that block LFA-1/ICAM-1 or VLA-4/VCAM-1 interactions, and lectin adhesion molecule-1 and endothelial leukocyte adhesion molecule-1 appeared not to be implicated, other as yet undefined cell surface structures are likely to participate in T cell/EC interactions.     

Vascular endothelial growth factor expression of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1), and E-selectin through nuclear factor-kappa B activation in endothelial cells

Vascular endothelial growth factor (VEGF) induces adhesion molecules on endothelial cells during inflammation. Here we examined the mechanisms underlying VEGF-stimulated expression of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1), and E-selectin in human umbilical vein endothelial cells. VEGF (20 ng/ml) increased expression of ICAM-1, VCAM-1, and E-selectin mRNAs in a time-dependent manner. These effects were significantly suppressed by Flk-1/kinase-insert domain containing receptor (KDR) antagonist and by inhibitors of phospholipase C, nuclear factor (NF)-kappaB, sphingosine kinase, and protein kinase C, but they were not affected by inhibitors of mitogen-activated protein/extracellular signal-regulated kinase kinase (MEK) 1/2 or nitric-oxide synthase. Unexpectedly, the phosphatidylinositol (PI) 3'-kinase inhibitor wortmannin enhanced both basal and VEGF-stimulated adhesion molecule expression, whereas insulin, a PI 3'-kinase activator, suppressed both basal and VEGF-stimulated expression. Gel shift analysis revealed that VEGF stimulated NF-kappaB activity. This effect was inhibited by phospholipase C, NF-kappaB, or protein kinase C inhibitor. VEGF increased VCAM-1 and ICAM-1 protein levels and increased leukocyte adhesiveness in a NF-kappaB-dependent manner. These results suggest that VEGF-stimulated expression of ICAM-1, VCAM-1, and E-selectin mRNAs was mainly through NF-kappaB activation with PI 3'-kinase-mediated suppression, but was independent of nitric oxide and MEK. Thus, VEGF simultaneously activates two signal transduction pathways that have opposite functions in the induction of adhesion molecule expression. The existence of parallel inverse signaling implies that the induction of adhesion molecule expression by VEGF is very finely regulated.     

Two pathways of CD11b/CD18-mediated neutrophil aggregation with different involvement of protein kinase C-dependent phosphorylation

The characteristics of homotypic neutrophil aggregation, mediated by the adhesion molecule CD11b/CD18, differ according to whether activation takes place via intracellular protein kinase C(PKC) inducers or chemoattractants. In response to diacylglycerol (DAG) analogues such as PMA and 1,2-dioctanoyl-sn-glycerol, a prolonged cellular aggregation occurs that is associated with intense phosphorylation of the CD18 beta-chain. In response to the chemoattractant FMLP, a more transient aggregation event results that is associated with minimal beta-chain phosphorylation. By using the PKC inhibitor staurosporine, we now show that these differences are likely to reflect two different pathways of activation. Both aggregation and phosphorylation induced by DAG analogues are completely abolished by staurosporine in a parallel dose-dependent manner. Conversely, FMLP-induced aggregation is enhanced and prolonged by staurosporine whereas the associated minimal phosphorylation event is further diminished by staurosporine. Accordingly, activation of neutrophil aggregation by DAG analogues is associated with and presumably due to phosphorylation of the CD18 beta-chain. This intense phosphorylation occurs via a staurosporine-sensitive kinase such as PKC. FMLP, on the other hand, appears to activate CD11b/CD18 by a distinct mechanism. This latter mechanism does not seem to be dependent on what may be a minor PKC-induced phosphorylation of the beta-chain, and indeed is enhanced by inhibition of PKC. Of note, staurosporine was also found to cause selective release of specific granules with concomitant increase in surface display of CD11b/CD18. These data further support previous observations that up-regulation of this adhesive molecule is not the primary event in the induction of cellular adhesiveness.     

Endothelial activation during interleukin 2 immunotherapy. A possible mechanism for the vascular leak syndrome

A major sequela of immunotherapy with interleukin 2 (IL-2) is development of a vascular leak syndrome. The pathogenesis of this toxic effect is not known. We have examined pre- and post-treatment skin biopsies from 14 patients undergoing systemic administration of IL-2 for evidence of endothelial cell activation. Specifically, we have used the immunoperoxidase technique to detect the expression of three different activation antigens: endothelial-leukocyte adhesion molecule 1, detected with monoclonal antibody H4/18; intercellular adhesion molecule 1, detected with antibody RR1/1; and histocompatibility leukocyte antigen-DQ, detected with antibody Leu 10. Each of these antigens may be induced on cultured endothelial cells by various cytokines (although not by IL-2) and is expressed during endothelial cell activation in vivo at sites of delayed hypersensitivity and other immune responses. Pretreatment biopsies from each patient showed no endothelial expression of endothelial-leukocyte adhesion molecule 1 and only weak to moderate expression of intercellular adhesion molecule 1 and histocompatibility leukocyte antigen-DQ (except for one specimen unreactive with Leu 10). After 5 days of treatment, every patient showed marked endothelial expression of all three antigens (except for the same patient who remained unreactive with Leu 10). Endothelial-leukocyte adhesion molecule-1 expression was confined to postcapillary venular endothelium whereas intercellular adhesion molecule-1 and Leu 10 also were expressed on stromal cells and mononuclear cells. Thus, we conclude that i.v. administration of IL-2 leads to endothelial cell activation. Because IL-2 fails to induce the same antigens on cultured endothelial cells, we infer that IL-2 acts in vivo by inducing the production of other cytokines (e.g., interleukin 1, tumor necrosis factor, lymphotoxin, and interferon-gamma). Finally, since endothelial cell activation at sites of cell-mediated immune responses is well known to result in vascular leakiness to macromolecules, we propose that the vascular leak syndrome accompanying IL-2 therapy may arise from widespread inappropriate endothelial cell activation.     

Activation of dense human tonsilar B cells. Induction of c-myc gene expression via two distinct signal transduction pathways.

Antibodies to surface Ig or to the B cell marker CD20 trigger resting human B cells in similar yet distinct ways. Either antibody induces five-fold increases in the expression of the protooncogene, c-myc, as detected with semi-quantitative Northern blot assays. The induction of c-myc mRNA by anti-IgM or anti-CD20 is blocked by inhibitors of protein kinase C (PKC) such as staurosporine and by pretreatment of B cells with phorbol esters to reduce cellular PKC levels. This suggests that PKC is involved in the pathways stimulated by both anti-IgM and anti-CD20. However, anti-CD20, unlike anti-IgM, does not activate significant increases in inositol triphosphate or intracellular-free calcium. Further, anti-CD20-triggered elevation of c-myc mRNA is inhibited by pertussis and cholera toxins, whereas the pathway initiated by anti-IgM if anything is stimulated by pertussis toxin and unchanged by cholera toxin. Further differences in the nature of these two signals were seen when the expression of adhesion/recognition molecules were examined. Anti-IgM consistently induces increased expression of the adhesion molecules CD54 (I-CAM-1) and B7/BB-1 on B cells, but anti-CD20 does not. Yet both anti-CD20 and anti-IgM increase class II MHC, CD18 (LFA-1 beta-chain) and LFA-3 levels. These data suggest that the way in which B cells are activated may influence their surface phenotype and possibly subsequent migration or cell-cell interactions.     

Thrombin causes increased monocytic-cell adhesion to endothelial cells through a protein kinase C-dependent pathway.

The coagulation protein thrombin has been shown to stimulate multiple endothelial-cell (EC) functions, including production of platelet-derived growth factor and of platelet-activating factor (PAF), and neutrophil adhesion. We have found that thrombin causes increased binding of monocytic cells (U937 cells and normal human monocytes) to cultured EC of various species. Maximum adhesion of monocytes to pig aortic EC occurred 6 h after thrombin treatment and remained elevated through 24 h. Stimulation of adherence by bovine alpha-thrombin was half-maximal at 15 units/ml, and reached a plateau at 50 units/ml. Catalytically inactive thrombin (phenylmethanesulphonyl fluoride-treated) had no effect on monocyte adhesion to EC. Heparin, but not the endotoxin antagonist polymyxin B, suppressed the stimulation of adhesion by thrombin without altering basal adhesion. Two lines of evidence suggested that protein kinase C (PKC) was involved in the intracellular signalling to increase monocyte adhesion to EC. First the PKC activator phorbol 12-myristate 13-acetate (PMA) stimulated monocytic-cell adhesion to EC at a dose consistent with stimulation of PKC (half-maximal response at 1-3 nM) and with a time course similar to that for thrombin stimulation (maximal by 4 h). Diacylglycerol, a physiological activator of PKC, also stimulated U937-cell adhesion to EC. Secondly, H7, a PKC inhibitor, completely blocked stimulation of monocyte adhesion to EC by thrombin or PMA. The structural analogue of H7, HA1004, which preferentially inhibits cyclic-AMP- and cyclic-GMP-dependent protein kinases, had no effect on stimulated monocyte adhesion. The PKC inhibitor also blocked the stimulation of monocyte adhesion to EC by interleukin-1 and endotoxin, but did not alter the basal level of monocyte binding to unstimulated EC. Thrombin stimulation of monocyte adhesion differed from the reported stimulation of neutrophil adhesion by thrombin in that the latter process reached a maximum in minutes rather than hours. In addition, neither PAF itself nor agents known to stimulate PAF production by EC, such as arachidonate and the Ca2+ ionophore A23187, had any effect on monocyte adhesion. These results demonstrate a PKC-dependent cytokine-like action of the coagulation protein thrombin in modulating monocytic-cell adhesion to EC, a phenomenon of potential importance in many pathological and physiological processes.     

Effects of tumor necrosis factor, lipopolysaccharide, and IL-4 on the expression of vascular cell adhesion molecule-1 in vivo. Correlation with CD3+ T cell infiltration.

We have injected human TNF, LPS, and IL-4 into the skin of baboons to examine regulation of endothelial leukocyte adhesion molecules (ELAM) in vivo and to determine which endothelial adhesion molecules correlate temporally and spatially with cytokine-induced T cell infiltration. The expression of adhesion molecules ELAM-1 (E-selectin), VCAM-1, and ICAM-1 (CD54) were quantified by immunocytochemical staining of frozen sections obtained from skin biopsies; T cell infiltration was measured by immunocytochemical staining of CD3+ T cells in serial sections. We found that injection of TNF causes late (24 to 48 h) T cell infiltration whereas injection of LPS, in doses that do not cause tissue necrosis, does not. The ability of TNF (but not LPS) to recruit T cells correlates with the ability of TNF to cause sustained endothelial cell adhesion molecule expression. Expression of VCAM-1 on post-capillary venules showed the highest degree of spatial localization with infiltrates. IL-4, although not proinflammatory by itself, can cause T cell infiltration in combination with an ineffective dose of TNF. The ability of IL-4 to augment TNF-induced inflammation best correlates with the ability of the combination of IL-4 and TNF to increase endothelial VCAM-1 expression. In contrast, IL-4 does not promote T cell infiltration or endothelial VCAM-1 expression in combination with LPS. In cytokine-injected tissues, VCAM-1 is also expressed on connective tissue cells other than endothelium, including smooth muscle and perineural cells, where it is induced by cytokines in parallel with endothelial VCAM-1. Overall, our data support the hypothesis that endothelial VCAM-1 expression contributes to T cell extravasation at sites of inflammation. Furthermore, we find that IL-4, a product a Ag-activated T cells, can interact with TNF to selectively promote VCAM-1 expression and the development of T cell-rich infiltrates, characteristic of Ag-induced inflammatory reactions.     

Modulation of endothelial autacoid release by protein kinase C: Feedback inhition or non-specific attenuation of receptor-dependent cell activation?

Receptor-mediated elevations of intracellular Ca²⁺ in endothelial cells may be controlled by a negative feedback mechanism through activation of protein kinase C (PKC). To test this hypothesis, we studied the effects of an activation or inhibition of PKC on the release of nitric oxide (NO) and prostacyclin (PGI₂) from cultured bovine and porcine aortic endothelial cells (EC). Preincubation with the PKC activators phorbol-12-myristate-13-acetate (PMA) (3-300 nM) or 1-oleyl-2-acetyl-glycerol (OAG) (30 μM) significantly attenuated the release of NO and PGI₂ from EC stimulated with bradykinin (0.3–30 nM), whereas phorbol-12, 13-didecanoate (PDD) (30–300 nM), which does not activate PKC, had no effect. UCN-01 (10 nM), a specific PKC inhibitor, significantly augmented the bradykinin-stimulated release of NO from EC. These effects were correlated with a reduced (PMA) or enhanced (UCN-01) elevation of intracellular Ca²⁺ in response to bradykinin in both types of EC. Neither the PKC activators nor the inhibitor had any effect on resting intracellular Ca²⁺ or basal endothelial autacoid release. Several isoforms of PKC (namely PKC_α, PKC_δ, PKC_?, and PKC_ζ) were detected in bovine, human, and porcine EC by immunoblotting analysis with isotype-specific anti-PKC antibodies, which, except PKC_?, were predominantly located in the cytosol. Incubation of bovine EC with PMA elicited a significant increase in membrane-bound PKC_α immunoreactivity, whereas there was no translocation of PKC_α from the cytosolic to the membrane fraction with bradykinin. As determined by histone phosphorylation, PKC activity was similarly reduced in the cytosol, but increased in the membrane fraction of bovine EC exposed to PMA, whereas bradykinin had no significant effect. These findings indicate that endothelial autacoid release can be modulated by activators and inhibitors of PKC. However, stimulation of EC with bradykinin does not lead to a detectable activation of PKC, suggesting that PKC does not exert a negative feedback in the signal transduction pathway of this receptor-dependent agonist. © 1993 Wiley-Liss, Inc.     

Zymosan-stimulated tumor necrosis factor-alpha production by human monocytes. Down-modulation by phorbol ester.

In this study, we showed that human monocytes produced TNF-alpha in response to zymosan, a particulate agonist. Protein kinase C (PKC) seems to play a regulatory role in zymosan-induced TNF-alpha secretion. The pretreatment of monocytes with PMA induced a dose-dependent inhibition of zymosan-stimulated TNF production. This inhibition was likely due to an activation of PKC because it was prevented by inhibitors of PKC, sphingosine, and staurosporine. Moreover, PMA elicited a profound down-modulation of zymosan binding to monocytes. The inhibition of zymosan binding and TNF production displayed similar dose-dependence, suggesting that both events were closely related. In addition, PMA did not modify the expression of CD11b/CD18 receptor that is involved in zymosan recognition. In view of these findings, qualitative changes of CD11b/CD18 molecules might account for the inhibition of zymosan binding and TNF production. Thus, PMA specifically increased the association of CD11b/CD18 with the detergent-insoluble cytoskeleton. Cytochalasin B but not microtubule disrupters, nocodazole and colchicine, partially prevented the inhibition of zymosan binding. Hence, the inhibitory action of PMA on zymosan binding seems to be mediated by an increase in attachment of zymosan receptor to cytoskeleton and more likely to microfilaments. The regulatory activity of PKC might represent a first way of limiting cytokine over-production in response to pathogens which interact with monocytes via CD11/CD18 molecules.     

Role of ras-dependent ERK activation in phorbol ester-induced endothelial cell barrier dysfunction

The treatment of endothelial cell monolayers with phorbol 12-myristate 13-acetate (PMA), a direct protein kinase C (PKC) activator, leads to disruption of endothelial cell monolayer integrity and intercellular gap formation. Selective inhibition of PKC (with bisindolylmaleimide) and extracellular signal-regulated kinases (ERKs; with PD-98059, olomoucine, or ERK antisense oligonucleotides) significantly attenuated PMA-induced reductions in transmonolayer electrical resistance consistent with PKC- and ERK-mediated endothelial cell barrier regulation. An inhibitor of the dual-specificity ERK kinase (MEK), PD-98059, completely abolished PMA-induced ERK activation. PMA also produced significant time-dependent increases in the activity of Raf-1, a Ser/Thr kinase known to activate MEK ( approximately 6-fold increase over basal level). Similarly, PMA increased the activity of Ras, which binds and activates Raf-1 ( approximately 80% increase over basal level). The Ras inhibitor farnesyltransferase inhibitor III (100 microM for 3 h) completely abolished PMA-induced Raf-1 activation. Taken together, these data suggest that the sequential activation of Ras, Raf-1, and MEK are involved in PKC-dependent endothelial cell barrier regulation.