13S contribute to regulation of epithelial permeability.56 We asked whether endothelial cell cytoskeletons could contribute to regulation of endothelial permeability and whether this might have any relationship to granulocyte-mediated increases in lung microvascular permeability. To answer these questions, we used 2 models to measure endothelial permeability. Porcine pulmonary artery endothelial cells were grown on micropore filters. On the filters, the cells form a monolayer impermeable to albumin while the filter itself is permeable. The filters with adherent monolayers are mounted in modified chemotaxis chambers and the movement of albumin from the luminal to the subluminal side is measured. To support the relevance of this in vitro system, we also used isolated rabbit lungs perfused with a balanced salt solution. To determine if microfflaments of the endothelial cell cytoskeleton were important in determining vascular permeability, we added the microfilament disrupting agents cytochalasin D (2 x 106M) or cytochalasin B (lx 10-5M) to monolayers of endothelial cells. We observed a 2-fold increase in endothelial permeability that was readily reversed by washing the cytochalasins off the monolayers. Parallel fluorescence and phase contrast microscopy showed disruption of endothelial cell microfilaments and gaps between adjacent cells, both defects reverting to normal within 4 hours. To support the relevance of these findings, we added lx 105M cytochalasin B to perfusates of isolated rabbit lungs and observed that the lungs became edematous within 1.5 hours (lung weight increase 20 x above control and lung lavage albumin concentration/perfusate albumin concentration was 20 x above control). We had recently shown that granulocytes and PMA caused a granulocyte-dependent increase in permeability of endothelial monolayers grown on micropore filters and of isolated perfused lungs. Phase contrast and fluorescence microscopy of endothelial monolayers exposed to granulocytes and PMA showed marked disruptions of the microfilaments with gaps between adjacent endothelial cells. These studies demonstrate that primary alterations in endothelial cell cytoskeletons can alter endothelial permeability and that at least 1 model of granulocyte-mediated edema is associated with marked disruption of endothelial cell cytoskeletons. The endothelial cell cytoskeleton may contribute to regulation of vascular permeability in vivo.