Endothelial cell associated platelet-activating factor (PAF), a costimulatory intermediate in TNF-alpha-induced H2O2 release by adherent neutrophil leukocytes.

TNF is a strong secretagogue for surface-contacting neutrophils. During inflammation, endothelium offers the first substrate for neutrophil adherence and for modulation of the toxic response of neutrophils to soluble agonists such as TNF. In this in vitro study, evidence is presented that endothelium participates actively in TNF-induced neutrophil respiratory burst activity by expressing platelet-activating factor (PAF) in response to initial neutrophil H2O2 release. Three findings are shown that favor such a mechanism. First, PAF receptor antagonists reduced H2O2 release by TNF-activated neutrophils placed on endothelium approximately by 50%, whereas H2O2 responses by neutrophils placed on serum-coated polystyrene remained intact. Second, preincubation of HUVEC with known PAF-inducing agents PMA, H2O2, and thrombin, followed by fixation, enhanced neutrophil H2O2 release in response to TNF. H2O2 release by these neutrophils was sensitive to the presence of PAF receptor antagonists, whereas H2O2-release from neutrophils placed on fixed nonactivated endothelial cells was not. Finally, replacing endothelium by monolayers of human renal cortical epithelial cells and human fibroblasts, cells that are known to produce less PAF than endothelial cells, reduced the effect of PAF receptor antagonists. P-selectin expression and IL-8 release, two other ways by which endothelial cells might influence H2O2-release by TNF preincubated neutrophils, were examined in parallel, and were found not to influence TNF-induced neutrophil H2O2-release. We conclude that during neutrophil-endothelial interaction in inflammation, endothelium modulates the toxic response of neutrophils to TNF. Endothelial cell-associated PAF, but not endothelial cell IL-8 release and P-selectin expression, is likely to participate in TNF-induced neutrophil respiratory burst activity.

IL-6, IL-8 and TNF production by cytokine and lipopolysaccharide-stimulated human renal cortical epithelial cells in vitro.

The capacity of renal epithelial cells to produce IL-6, IL-8 and TNF was investigated. Cultures of explanted human renal cortical epithelial cells (RCEC) were established, and cytokine-release and mRNA expression by these cells were measured. IL-6, IL-8 and TNF release were measured after stimulation with IL-1 beta TNF-alpha, LPS and the phorbol esther PMA. All these agents were found to induce increased release of the three cytokines. Whilst no spontaneous TNF-release occurred, IL-6 and IL-8 were continuously released by non-stimulated RCEC cultures. IL-1 beta was the most potent trigger, enhancing both RCEC cytokine release and expression of IL-6, IL-8 and TNF mRNA. Indomethacin, budesonide, cyclosporin and FK 506 were tested for their influence on RCEC cytokine release. Only the steroid budesonide appeared to reduce both spontaneous and IL-1 beta induced cytokine release. Our data demonstrate stimulus specific release of IL-6, IL-8 and TNF by RCEC, and suggest that cytokine cell-to-cell communication may be important in regulating inflammatory processes in the kidney.

Antagonistic effects of lipopolysaccharide binding protein and bactericidal/permeability-increasing protein on lipopolysaccharide-induced cytokine release by mononuclear phagocytes. Competition for binding to lipopolysaccharide.

Serum proteins play an important role in LPS-induced cell activation. The LPS binding protein (LBP) enhances cellular responses to LPS, whereas the polymorphonuclear leukocyte product bactericidal/permeability-increasing protein (BPI) inhibits LPS-induced cell activation. In this study the influences of LBP and BPI, two proteins with opposite effects, but with considerable sequence homology, on LPS-induced mononuclear phagocytic cell cytokine release was studied. LBP was shown to enhance LPS-induced TNF-alpha, IL-6, and IL-8 release by mononuclear phagocytic cells, whereas BPI inhibited the release of these cytokines. Furthermore, the effects of LBP and BPI on LPS-induced cytokine release by mononuclear phagocytic cells were shown to be counteractive. BPI interfered with the enhancing effect of LBP on the LPS-induced cytokine release. At high LBP to BPI ratios, BPI could no longer inhibit LBP-induced enhancement. In accordance, increasing concentrations of BPI abrogated the LBP effect. Next, it was shown that LBP and BPI compete for binding to LPS by using an assay system that detects binding of free BPI to an anti-BPI mAb. LPS prevented binding of BPI to anti-BPI mAb, whereas preincubation of LPS with LBP prevented the LPS-induced inhibition. Also, it was observed that both BPI and LBP inhibited LPS activity in the chromogenic LAL assay. We conclude from this study that LBP and BPI have counteractive effects on LPS-induced mononuclear phagocytic cell cytokine release by competing for binding to LPS.

Anti-CD14 antibodies reduce responses of cultured human endothelial cells to endotoxin.

Lipopolysaccharide (LPS) activates both myeloid and endothelial cells. Whereas CD14 has been shown to be involved in LPS recognition by myeloid cells, the mechanism responsible for the strong response of endothelial cells to LPS remains to be elucidated. The role of CD14 in this process was studied using CD14-specific antibodies (Ab). Anti-CD14 Ab inhibited LPS-induced interleukin-6 (IL-6) release and E-selectin expression by cultured human umbilical vein endothelial cells (HUVEC). Messenger RNA encoding IL-6 and E-selectin was reduced in parallel. The inhibitory effect of anti-CD14 Ab was epitope dependent, maximal at low LPS concentrations and dropping with increasing LPS doses. Anti-CD14 Ab did not affect endothelial cell activation induced by IL-1 beta, tumour necrosis factor-alpha (TNF-alpha) and phorbol 12-myristate 13-acetate (PMA). IL-6 release and E-selectin expression of HUVEC were strongly reduced when LPS activation was performed in the absence of serum, indicating involvement of serum components in LPS activation of HUVEC. Nevertheless, anti-CD14 Ab also blocked LPS-induced HUVEC activation in the absence of serum. Although the role of serum components in LPS activation remains to be elucidated, CD14 seems to be a key mediator in LPS-induced activation of endothelial cells.

Involvement of CD14 in lipopolysaccharide-induced tumor necrosis factor-alpha, IL-6 and IL-8 release by human monocytes and alveolar macrophages.

CD14 has been reported to function as a receptor for bacterial LPS complexed with serum proteins, transducing an activation signal for TNF-alpha production. We found that the anti-CD14 mAb MEM-18 inhibited not only LPS-induced release of TNF-alpha, but also LPS-induced, TNF-alpha independent release of IL-6 and IL-8 by human monocytes and alveolar macrophages. Inhibitory effect of MEM-18 was detected both in the presence of human or bovine calf serum and under serum-free conditions. In contrast, MEM-18 did not block release of these cytokines induced by IL-1 beta, TNF-alpha, PMA, and zymosan. We conclude that CD14 is involved in LPS-induced release of TNF-alpha, IL-6, and IL-8 by monocytes and alveolar macrophages and that this receptor appears to be able to recognize LPS directly in the absence of serum.

Evidence for endocytosis of E-selectin in human endothelial cells.

E-selectin is an inducible adhesion molecule on endothelial cells. The internalization of this glycoprotein was investigated on tumor necrosis factor (TNF)-activated cultured human umbilical vein endothelial cells (HUVEC). Kinetics of intercellular adhesion molecule-1 (ICAM-1) were studied in parallel experiments. Internalization studies were performed with radioiodinated antibodies in an acid elution endocytosis assay, and by immunohistology; both approaches gave equivalent results. [125I]ENA1, a monoclonal antibody (mAb) specific for E-selectin, was internalized at a rate of approximately 1.7% of the membrane-bound [125I]mAb per minute. In contrast, less than 0.1% of membrane-bound [125I]RR1/1, an mAb specific for ICAM-1, was internalized per minute. TNF-activated HUVEC were immunostained and examined by light microscopy (LM) and electron microscopy (EM). LM revealed the presence of ENA1, but not RR1/1, after 30 minutes of incubation with these mAb in cytoplasmic vesicles, which were characterized as multivesicular bodies by EM. Without previous mAb exposure of the endothelial cells, both high amounts of E-selectin and bovine serum albumin complexed to colloidal gold, used as a marker for fluid-phase internalization, were detected in the same organelles, thus arguing against mAb interaction-induced E-selectin internalization. Furthermore, the amount of E-selectin surface expression was not influenced by ongoing mAb presence, also arguing against mAb interference with normal E-selectin kinetics. Taken together, these results indicate that TNF-activated HUVEC constitutively internalize E-selectin. Physiological significance of E-selectin internalization in the regulation of E-selectin membrane expression, and in clearing E-selectin ligands from the circulation, needs further investigation.

A comparison of substrates for human umbilical vein endothelial cell culture.

We have studied culture conditions which facilitate the growth of stable, non-proliferating, human umbilical vein endothelial cell (HUVEC) monolayers. Gelatin and fibronectin coatings, with or without glutaraldehyde cross-linking, on both plastic and glass were investigated for initial attachment of HUVEC and growth characteristics. The presence during culture of intercellular (IC) junctions demonstrated by silver staining, expression of platelet endothelial cell adhesion molecule-1 (PECAM-1) and maintenance of a cobblestone appearance of HUVEC monolayers were assessed over time. Glutaraldehyde cross-linked fibronectin and gelatin coatings on glass and glutaraldehyde cross-linked gelatin or untreated fibronectin coatings on plastic served as good substrates for short term culture. Long term (20 days) cultures of HUVEC which maintained silver and PECAM-1 staining of IC junctions and a cobblestone appearance could be achieved if glutaraldehyde cross-linked gelatin coatings on glass were used as substrates.

Involvement of the CD11b/CD18 integrin, but not of the endothelial cell adhesion molecules ELAM-1 and ICAM-1 in tumor necrosis factor-alpha-induced neutrophil toxicity.

TNF-alpha can incite neutrophil-mediated endothelial cell damage and neutrophil H2O2 release. Both effects require adherent neutrophils. Using specific mAb, we showed in this in vitro study that the CD18 beta 2-chain and the CD11b alpha M-chain of the CD11/CD18 integrin heterodimer have a major role in both TNF-alpha-induced neutrophil-mediated detachment of human umbilical vein endothelial cells and H2O2 release by TNF-alpha-activated human neutrophils. In contrast to anti-CD18 mAb, which consistently prevented neutrophil activation, anti-CD11a mAb and two of three anti-CD11b mAb did not reduce endothelial cell detachment and neutrophil H2O2 release, although they decreased neutrophil adhesion to human umbilical vein endothelial cells. mAb 904, directed against the bacterial LPS binding region of CD11b, reduced endothelial cell detachment for about 40% and neutrophil H2O2 release for more than 50%, demonstrating that CD11b/CD18 is engaged in TNF-induced neutrophil activation. Dependence on CD11b/CD18 could not be overcome by CD18-independent anchoring of neutrophils via PHA. Additionally, neither induction of increased expression of the endothelial cell adhesion molecules ICAM-1 and ELAM-1, nor subsequent addition of specific mAb, influenced endothelial cell injury or H2O2 release by TNF-activated neutrophils. Interaction with ICAM-1 and ELAM-1 therefore appears not to induce additional activation of TNF-stimulated neutrophils. These studies suggest that a specific, CD11b/CD18-mediated signal, instead of adherence only, triggers toxicity of TNF-activated neutrophils.

Tumour necrosis factor-alpha induces neutrophil-mediated injury of cultured human endothelial cells.

We investigated the ability of TNF-alpha to mediate damage of endothelial cells in the presence of neutrophils, by measuring detachment of cultured human umbilical vein endothelial cells (HUVEC). Endothelial cell detachment was increased from 5% to about 75% by the presence of 1-10 ng/ml TNF-alpha during incubation with neutrophils, whereas negligible endothelial cell lysis was observed as measured by 51Cr release. TNF-alpha was compared with the cytokines IL-1 alpha and IFN-gamma and with PMA and LPS. Both TNF-alpha and PMA appeared to be strong triggers for neutrophil-induced endothelial cell detachment, whilst reduced injury was seen after addition of IL-1 alpha and LPS. IFN-gamma did not induce endothelial cell detachment, but potentiated the effect of both TNF-alpha and IL-1 alpha. TNF-alpha-induced endothelial cell detachment was neutrophil dependent, since pre-incubation of neutrophils, but not pre-incubation of endothelial cells with TNF-alpha, caused endothelial cell detachment. Thus, TNF-alpha-induced increase in neutrophil-adhesiveness of HUVEC was found not to be essential for endothelial damage. Pre-incubation of neutrophils in suspension with TNF-alpha induced rapid activation, followed by nearly complete deactivation of neutrophils, as measured by their capacity to induce detachment of endothelial cells after removal of TNF-alpha. These results indicate that local presence of TNF-alpha might be critical in tissue or organ damage during early, neutrophil-mediated inflammatory processes, independent of enhanced adhesiveness of endothelium for neutrophils.

LPS and cytokine-induced endothelial cell IL-6 release and ELAM-1 expression; involvement of serum.

In this in vitro study, the influence of serum-concentration, heat inactivation of the serum and the origin of the serum on the responsiveness of cultured human umbilical vein endothelial cells (HUVEC) to immunological challenges was investigated. Addition of human serum during stimulation with 1 microgram/ml bacterial lipopolysaccharide (LPS) increased endothelial cell ELAM-1 expression and interleukin (IL)-6 release five to ten-fold. Full endothelial cell responsiveness to LPS required 10 to 50% human serum and was largely abrogated after heating the serum for 30 minutes at 56 degrees C. Addition of newborn or fetal bovine serum instead of human serum, induced even higher IL-6 release and ELAM-1 expression in response to LPS, whilst heat-inactivation of these serum-batches only moderately decreased endothelial cell responses. Endothelial cell IL-6 release and ELAM-1 expression after stimulation with IL-1 beta and tumor necrosis factor-alpha (TNF-alpha) were less influenced by heat inactivation of the serum and by omission of serum, whilst responses to PMA remained completely unaffected by such modifications in assay media. Finally, we demonstrated that endothelial cell IL-8 release also and ICAM-1 expression in response to LPS and cytokines were increased by addition of human serum, indicating that the use of serum-free assay media, or the use of media enriched with heat-inactivated (HI) human serum interferes with physiological endothelial cell responsiveness.(ABSTRACT TRUNCATED AT 250 WORDS)

The interrelation between TNF, IL-6, and PAF secretion induced by LPS in an in vivo and in vitro murine model.

The interrelation between TNF, IL-6, and PAF secretion in an in vivo and in vitro murine model was studied. Mice were injected with LPS, giving rise to considerable TNF and IL-6 serum levels. To determine the influence of TNF production on the IL-6 secretion, one group of mice was treated with an anti-TNF mAb before LPS administration. The LPS-induced IL-6 secretion was reduced to 48% in the anti-TNF-pretreated group. Parallel to the in vivo experiments, cultures of murine peritoneal macrophages were stimulated with LPS. IL-6 secretion was diminished for 25% in presence of anti-TNF mAb. Further, the role of PAF released in response to a challenge with LPS in the regulation of TNF and IL-6 production was investigated. TNF secretion was strongly reduced when cultures of peritoneal macrophages were stimulated with LPS in the presence of a PAF antagonist, whereas IL-6 secretion was not altered by the PAF antagonist. However, pretreatment of mice with a PAF antagonist did not influence serum TNF, nor serum IL-6 levels induced by LPS injection. These data show that TNF was an intermediate in the induction of IL-6 production in vivo and in vitro. PAF played a central role in the TNF release in the in vitro experiments. This could, however, not be established in vivo. LPS-induced PAF secretion was not involved in the regulation of IL-6 secretion.

Tumour necrosis factor alpha (TNF-alpha) and interleukin 6 in a zymosan-induced shock model.

TNF plays a central role in septic shock induced by endotoxin or Gram-negative bacteria. Zymosan can elicit a septic shock-like syndrome in rodents in the absence of endotoxin. TNF and IL-6 release in mice treated with zymosan was investigated. One hour after intraperitoneal zymosan injection, maximal TNF levels were measured in serum, followed by IL-6 peak levels 1 h later. Treatment with a monoclonal antibody against TNF lowered zymosan-induced mortality from 63 to 11.6%, while maximal IL-6 levels were lowered by about 40%. Mechanisms triggering zymosan-induced cytokine release in murine macrophages were analysed in vitro. Cytokine release was only slightly triggered by uncoated zymosan particles. Thirty-nine per cent of TNF release by macrophages appeared to be triggered by zymosan-bound activated complement. Maximal TNF release also required the presence of natural antibodies against zymosan and zymosan-activated serum. In contrast, maximal IL-6 release was reached upon stimulation with zymosan-activated serum only, while the presence of zymosan particles lowered this response. We conclude that TNF is a crucial mediator in zymosan-induced shock. TNF release can be induced by different immunological pathways, without the need for the direct presence of endotoxins. Although IL-6 release during septic shock is partly dependent on TNF, in vitro trigger mechanisms for IL-6 and TNF differ remarkably.

IFN-gamma regulates the expression of the adhesion molecule ELAM-1 and IL-6 production by human endothelial cells in vitro.

In this study two new in vitro effects of IFN-gamma on human umbilical vein endothelial (HUVE) cells were described. First, it was shown that the expression of the adhesion molecule ELAM-1 on activated HUVE cells can be modulated by IFN-gamma. ELAM-1 is normally not expressed by HUVE cells, but its expression can rapidly be induced by TNF, IL-1, or LPS. Maximal expression is reached after 4 to 6 h of activation, and after 24 h the expression disappeared. Whereas IFN-gamma per se did not induce expression of ELAM-1, it enhanced and prolonged the expression of ELAM-1. This enhancement occurred when IFN-gamma was added before activation as well as when added simultaneously with activation. When IFN-gamma was added 6 or 9 h after the activation, the normally ongoing reduction of expression was not only retarded, but the expression increased for at least 3 h. Moreover, IFN-gamma abrogated the refractory period for restimulation. Neither IFN-beta nor IL-6 had any effect on the expression of ELAM-1. The second effect of IFN-gamma on HUVE cells is the capacity to enhance the IL-6 production by these cells. Prestimulation as well as coincubation of IFN-gamma with TNF, IL-1, or LPS resulted in a strongly augmented production of IL-6. The effects of IFN-gamma may in vivo play a role in the regulation of an inflammatory reaction, because ELAM-1 is an adhesion molecule for neutrophils, and IL-6 has an enhancing effect on the cytotoxicity of neutrophils.