Plasma obtained during human endotoxemia increases endothelial albumin permeability in vitro.

To gain insight in the pathogenesis of increased vascular permeability during sepsis, we studied the effect of plasma obtained during human experimental endotoxemia on the permeability of cultured endothelial monolayers. Eight healthy subjects received an i.v. dose of 2 ng/kg Escherichia coli O:113 lipopolysaccharide (LPS). The concentration of various plasma mediators that supposedly induce vascular permeability was measured over time. Plasmas that were obtained before, and 2 and 4 h after the administration of LPS were added to human umbilical venular endothelial cells that were cultured on semipermeable membranes.The permeability of the endothelial monolayers to fluorescein isothiocyanate-labeled bovine serum albumin was determined and expressed as the relative concentration of fluorescein isothiocyanate-bovine serum albumin when compared with that measured across empty Transwell-COL (Corning Life Sciences B.V., Schiphol-Rijk, The Netherlands) membranes (i.e., without endothelial monolayers). The permeability levels were correlated with the concentrations of various mediators.Experimental endotoxemia resulted in elevated levels of tumor necrosis factor alpha, interleukin (IL) 1beta, IL-6, IL-8, IL-10, and vascular endothelial growth factor and a moderate increase of IL-12 and IFN-gamma (all P values < 0.01). Incubation of human umbilical venular endothelial cells with plasma obtained 2 and 4 h after the administration of LPS increased the relative permeability from a baseline level (median) of 17% (range, 14% - 31%) to 23% (range, 12% - 39%; P = not significant) and 28% (range, 11% - 40%; P < 0.05), respectively. Plasma levels of vascular endothelial growth factor and IL-10, but not TNF-alpha or any other mediators, significantly correlated with the increase in endothelial permeability (r = 0.47, P = 0.038; r = 0.43, P = 0.038, respectively). The data presented here demonstrate that plasmas obtained from experimental human endotoxemia increase endothelial albumin permeability in vitro. Thus, cultured human endothelial monolayers provide a model to study sepsis-associated vascular changes.

Whole blood-mediated endothelial permeability and adhesion molecule expression: a model study into the effects of bacteria and antibiotics.

AIM: To investigate whether the inflammatory response of cultured endothelial cells, as induced by conditioned plasma, depends on the bacterial species or type of antibiotic used for incubation with whole blood. MATERIALS AND METHODS: Blood from healthy volunteers was stimulated ex vivo with different microorganisms, and with bacteria killed with different antibiotics. The resultant plasmas were incubated on monolayers of cultured human endothelial cells, followed by measurement of their permeability to albumin and expression of E-selectin and intercellular adhesion molecule-1. RESULTS: Incubation of Escherichia coli in blood yielded plasmas that induced a marked increase in endothelial permeability and E-selectin expression. The response to Bacteroides fragilis or Enterococcus faecalis was generally weaker. Similar effects were observed after incubation of whole blood with lipopolysaccharide (LPS). Much of the permeability and adhesion molecule response to E. coli remained after removal of intact microorganisms from the culture. Whereas antibiotic treatment of E. coli with imipenem or cefuroxime resulted in a divergent production of tumour necrosis factor-alpha (TNF-alpha) in blood, no significant differences between these treatments were observed with respect to the plasma-induced endothelial response. CONCLUSION: Bacteria differ in their capacity to generate a whole blood-mediated increase of endothelial permeability and adhesion molecule expression; this response depends, at least in part, on the presence of soluble bacterial components, such as LPS. Whereas treatment with various antibiotics may generate varying amounts of TNF-alpha, these differences are not translated into differences in endothelial permeability or adhesion molecule expression.

Tumor necrosis factor-alpha and interleukin-1beta mediate endothelial permeability induced by lipopolysaccharide-stimulated whole blood.

OBJECTIVE: To investigate the role of endotoxin-induced inflammatory mediators in blood on the permeability of endothelial monolayers. DESIGN: Whole blood of healthy volunteers was treated with bacterial lipopolysaccharide (Escherichia coli, B55:05), and the resultant plasma was added to human umbilical venular endothelial cells (HUVEC) cultured on semipermeable membrane inserts (Transwells). SETTING: University hospital laboratory. SUBJECTS: Whole blood of healthy volunteers. INTERVENTIONS: Donor plasma was treated with excess antibodies against either tumor necrosis factor-alpha, interleukin-1beta, or both, before the incubation on HUVEC. MEASUREMENTS AND MAIN RESULTS: The permeability of HUVEC monolayers to fluorescent-labeled albumin and dextran was measured over a 6-hr period, after removal of the stimulus. The production of tumor necrosis factor-alpha and interleukin-1beta in lipopolysaccharide-treated whole blood was determined by radioimmunoassay. Individually, lipopolysaccharide (10 microg/mL), tumor necrosis factor-alpha (10 ng/mL), and interleukin-1beta (50 ng/mL) all increased endothelial permeability by about 2.5-fold. A much larger increase could be achieved by preincubation of lipopolysaccharide (10 microg/mL) in whole blood: the resultant plasma induced a ten-fold increase of the permeability. The permeability response after preincubation of lipopolysaccharide in whole blood was time- and dose-dependent. Moreover, this treatment increased the sensitivity of endothelial monolayers to lipopolysaccharide by a factor of several thousand-fold: Whereas high doses of lipopolysaccharide were required for direct stimulation of the permeability, picomolar amounts of lipopolysaccharide in whole blood induced a similar increase. Significant amounts of tumor necrosis factor-alpha and interleukin-1beta were produced in blood at similar doses of lipopolysaccharide. The addition of antibodies against tumor necrosis factor-alpha or interleukin-1beta to plasma partially but significantly abrogated the permeability increase. However, a complete inhibition could be achieved by the simultaneous addition of anti-tumor necrosis factor-alpha and anti-interleukin-1beta to plasma. CONCLUSIONS: Although lipopolysaccharide is capable of directly inducing endothelial permeability, blood-borne tumor necrosis factor-alpha and interleukin-1beta mediate lipopolysaccharide-induced endothelial permeability at low endotoxin concentrations. These findings support the idea that multifactorial inhibition of inflammatory mediators may improve survival in septic patients.