Neutrophil cathepsin G increases transendothelial albumin flux

Neutrophils are involved in the development of inflammatory edema in some animal models, but their mechanisms of action are not completely understood. Among injurious agents available for neutrophil-mediated injury are cationic proteins such as cathepsin G. Previous articles have reported that cationic agents decrease the barrier properties of the endothelium both in vivo and in vitro. Using an ex vitro isolated, perfused rabbit lung and a cultured porcine pulmonary artery endothelial cell monolayer (ENDO) as models, we asked whether neutrophil cathepsin G could decrease the barrier properties of the intact vasculature and cultured endothelial monolayers. After the addition of cathepsin G (10 micrograms/ml) to the perfusate of the isolated, perfused lung, there was a fourfold increase in net weight gain after a venous pressure challenge (p less than 0.01). The addition of cathepsin G to a cultured ENDO directly increased the movement of albumin across the monolayers in a dose-dependent fashion (10 micrograms/ml led to a 59% +/- 5% increase, 15 micrograms/ml to a 135% +/- 55% increase, 20 micrograms/ml to a 247% +/- 78% increase, and 30 micrograms/ml to a 381% +/- 89% increase, p less than 0.05 at all concentrations). Heat-inactivation of the enzyme only partially protected the ENDO, but exposure in the presence of either the polyanion heparin or the polyanion dextran sulfate completely protected the ENDO. Normal human serum also protected the ENDO, but serum from two patients with alpha 1-proteinase inhibitor (alpha 1-PI) deficiency was only partially protective. The protection afforded by human serum was time dependent, because addition of the serum 5 or 15 minutes after the addition of cathepsin G offered no protection. Oxidation of alpha 1-PI, as may occur at sites of inflammation, also destroyed its protective effect. Cytotoxic injury of the porcine pulmonary artery endothelial cell monolayer by cathepsin G, which was also prevented by the polyanions, only partially explained these results, because cathepsin G increased albumin transfer across the ENDO at concentrations of 10 to 20 micrograms/ml, which were minimally cytolytic to the ENDO. Additionally, cathepsin G caused cell retraction, with the development of intercellular gaps visible on light microscopy of the ENDO. This effect was also prevented by the polyanions. These results demonstrate that cathepsin G has direct effects on a cultured ENDO that may be caused by the charge or charged site(s) on cathepsin G.