We tested the hypothesis that neutrophil sequestration is required for the development of tumor necrosis factor- (TNF) induced neutrophil- (PMN) dependent pulmonary edema. TNF (3.2 X 10(5) U/kg ip) was injected into guinea pigs 18 h before lung isolation. After isolation, the lung was perfused with a phosphate-buffered Ringer solution. Dextran sulfate (mol wt 500,000) prevented the changes in pulmonary capillary pressure (Ppc; 8.5 +/- 0.8 vs. 12.8 +/- 0.8 cmH2O), lung weight gain (dW; +0.240 +/- 0.135 vs. +1.951 +/- 0.311 g), and pulmonary edema formation or wet-to-dry wt ratio [(W - D)/D; 6.6 +/- 0.2 vs. 8.3 +/- 0.5] at 60 min induced by PMN infusion into a TNF-pretreated lung. The unsulfated form of dextran had no protective effect [Ppc, dW, and (W - D)/D at 60 min: 11.9 +/- 0.9 cmH2O, +1.650 +/- 0.255 g, and 7.3 +/- 0.2, respectively], whereas the use of another anionic compound, heparin, inhibited the TNF + PMN response [Ppc, dW, and (W - D)/D at 60 min: 5.6 +/- 0.4 cmH2O, +0.168 +/- 0.0.052 g, and 6.4 +/- 0.2, respectively]. Isolated lungs showed increased PMN myeloperoxidase (MPO) activity compared with control in TNF-treated lungs at baseline and 60 min after PMN infusion. Dextran sulfate, dextran, and heparin inhibited the increase in MPO activity. The data indicate that inhibition of PMN sequestration alone is not sufficient for the inhibition of PMN-mediated TNF-induced hydrostatic pulmonary edema and that a charge-dependent mechanism mediates the protective effect of dextran sulfate.