ABSTRACT: Endothelial cell dysfunction could explain many of the pathophysiological changes observed in preeclampsia. Markers of endothelial cell activation including increased circulating Von Willebrand factor (VWF) and cellular fibronectin (cFN) antedate clinically evident disease. We have therefore proposed that alteration of endothelial cell function by circulating agent(s) produced by the placenta initiates the clinical syndrome. This hypothesis predicts that there are a factor(s) in the blood of women with preeclampsia that are capable of altering endothelial cell function. We and others have examined in vitro interactions of maternal serum and endothelial cells to test this hypothesis. Our initial report indicating increased release of [51Cr]chromate from human umbilical vein endothelial cells (HUVE) suggested a lethal, lytic effect of serum from preeclamptic women. However, more specific indicators of endothelial cell structure and function do not support such a nonspecific effect. The morphology of HUVE was minimally altered after exposure to sera of preeclamptic women, and the entry of propidium iodide entry into cells, a sensitive indicator of membrane integrity, also was not increased. These findings, in combination with the increased expression of mRNA for platelet-derived growth factor (PDGF), suggest endothelial cell activation rather than cell death in response to sera from preeclamptic women. Comparison of the effects of endotoxin and sera from preeclamptic women also supports the specificity of this response. Whereas endotoxin strikingly increased VWF on the surface of HUVE and tissue factor activity in conditioned media while minimally increasing cFN, preeclamptic sera increased cFN but had no demonstrable effect on VWF or tissue factor activity. Thus, under the conditions of our in vitro assays, sera from preeclamptic women are not diffusely toxic but rather act selectively on specific activation pathways.