Effect of bacteria and antibiotic treatment on whole blood-mediated endothelial permeability in vitro

Abstract

Abstract Background During Gram-negative sepsis bacterial endotoxin or lipopolysaccharide (LPS) may activate and damage endothelial cells resulting in systemic vascular leakage. Antibiotic treatment of Gram-negative bacteria results in significant release of LPS, inducing production of cytokines by leucocytes in vitro. The biological relevance of this finding is not well understood. A whole blood assay and cultured endothelial monolayers have been used to evaluate the effect of antibiotic treatment of bacteria on endothelial permeability. Methods Heat-inactivated (HI) Escherichia coli (ATCC 25922) was treated with either cefuroxime (CXM), imipenem–cilastin (IPM) or polymyxin B (poly B). The cultures were incubated with heparinized whole blood from healthy donors in a final concentration of 105 colony-forming units ml−1 and incubated at 37°C for 24 h. Plasma was then collected and added to human umbilical vein endothelial cell monolayers cultured on semipermeable Transwell COL membranes. The permeability of the monolayers was determined using fluorescein isothiocyanate-labelled (FITC) albumin and rhodamine-β-isothiocyanate-labelled (RITC) dextran. The production of interleukin (IL) 1β, IL-6, IL-1ra and tumour necrosis factor (TNF) α in whole blood was measured using radioimmunoassay. Results In whole blood, HI bacteria alone induced the production of IL-1β, IL-6, IL-1ra and TNF-α. CXM treatment of HI E. coli before incubation in blood induced significantly more IL-1β, TNF-α and IL-6 than the HI control, while poly B treatment induced less cytokine production. IPM did not significantly alter the production of cytokines. In the permeability assay, whole blood treated with HI bacteria alone induced a permeability increase of 2·3 times control monolayer permeability (P < 0·05). While CXM or IPM treatment resulted in a significant further enhancement of the permeability, there was no significant difference between these antibiotic treatments (permeability increase 2·9 and 2·8 times the control respectively). In contrast, treatment with poly B almost completely abrogated the permeability effect of HI bacteria (1·1 times control; P > 0·05). Conclusion Since neutralization of LPS by poly B reduces the E. coli-induced production of cytokines in whole blood and prevents increased permeability, free LPS may be the driving force of cytokine-mediated endothelial permeability. Treatment of bacteria with CXM or IPM induces a significantly different cytokine production in whole blood. However, in both cases the resulting plasma leads to a similar increase in permeability. Thus, the fact that individual antibiotics can release different amounts of LPS may be irrelevant to endothelial permeability.