Direct effects of endotoxin on the endothelium: barrier function and injury.

D. Bannerman,S. Goldblum
Abstract:LPS directly disrupts EC barrier function in vitro and in vivo. This barrier dysfunction has been reported to occur in EC derived from both the macro- and microvasculature of varying species, including humans. Unlike other EC responses, LPS-induced loss of endothelial barrier function is protein-synthesis independent. In fact, protein synthesis inhibition enhances the LPS effect. The lipid A moiety is responsible for LPS-induced activation of the non-CD14-bearing EC, and agents that bind to and neutralize this highly conserved portion of the LPS molecule can crossprotect against EC barrier dysfunction elicited by LPS derived from diverse species of Gram-negative bacteria. Although the presentation of LPS to CD14-bearing cells such as macrophages and monocytes has been well characterized, far less is known about the interactions of LPS with the non-CD14-bearing EC. An EC receptor involved in LPS binding and cellular activation has yet to be identified. The presence of the accessory molecules, LBP and sCD14, are prerequisite to LPS-induced activation of EC at clinically relevant LPS concentrations. As with monocytes and macrophages, the CD14 dependence of LPS-induced endothelial barrier dysfunction can be overcome with high concentrations of LPS. In the absence of LBP and sCD14, a 200,000-fold increase in LPS concentration is required to elicit the same increments in EC monolayer permeability relative to when these accessory molecules are present. Within 30 minutes after LPS exposure, PTK activation is observed. PTK inhibition blocks LPS-induced EC actin depolymerization and endothelial barrier dysfunction which are seen only after a > or = 2-hour stimulus-to-response lag time. Furthermore this LPS-induced actin depolymerization is a prerequisite to opening up the paracellular pathway and loss of monolayer integrity. Interestingly LPS-induced increments in transendothelial 14C-BSA flux and EC detachment parallel caspase-mediated cleavage of ZA and FA proteins that participate in cell-cell and cell-matrix adhesion. The cleavage of the ZA components, beta- and gamma-catenin, does not affect their ability to bind the transmembrane protein, cadherin, or the actin-binding protein, alpha-catenin, suggesting that the linkage of the ZA to the actin cytoskeleton remains intact. LPS-induced cleavage of the FA protein, FAK, leads to dissociation of its catalytic domain from paxillin substrate and decreased paxillin phosphotyrosine content. Caspase inhibition protects against LPS-provoked apoptosis, cleavage of adherens junction proteins, paxillin dephosphorylation, cell-shape changes, and EC detachment. In contrast it fails to block LPS-induced increments in transendothelial 14C-BSA flux. PTK inhibition, which does protect against increased transendothelial 14C-BSA flux, does not block LPS-induced proteolytic cleavage events and only partially inhibits EC detachment. These findings suggest that the EC detachment and endothelial barrier dysfunction elicited by LPS are mediated through distinct pathways (Fig. 6). Much of the work to date has focused on LPS interactions with mCD14-bearing cells, such as monocytes and macrophages, which are central to the inflammatory response elicited by endotoxin. EC, which line the vasculature, are one of the first host tissue barriers to encounter circulating LPS. Because damage to the endothelium is known to contribute to the development of multiorgan failure, including ARDS, understanding LPS-induced EC dysfunction in the setting of Gram-negative septicemia has clear pathophysiologic implications. (ABSTRACT TRUNCATED)
What problem does this paper attempt to address?