Stimulated endothelial cells withdraw from the junctions with neighboring cells and develop a mounded morphology. These cells have mixed procoagulant and anticoagulant function, in contrast to the dominant anticoagulant function of quiescent cells. Recent findings from our laboratory indicate that tumor necrosis factor-a (TNFα) treated endothelial cells support localized assembly of the prothrombinase complex on the margins of cells and on filopodia that span the inter-cellular space. The prothrombinase binding sites are characterized by exposed phosphatidylserine and high convex curvature. We asked whether this inter-cellular space is a focal procoagulant environment while the body of the endothelial cell maintains an overall anticoagulant function. We also asked whether focal procoagulant activity has a function in innate immunity.
Human umbilical vein endothelial cells (HUVECs) were grown to a confluent monolayer on coverslips coated with 1% gelatin or Sigmacote. Cells were treated with 5 nM TNFα or an equal volume of saline for 12 hr. Rinsed cells were then overlaid with re-calcified plasma, diluted 1:4 with buffered saline and placed on a shaker plate at 37 °C at 1000 rpm. Plasma was removed from cells at various times and the cells and fibrin were assessed by confocal microscopy. Fluorescein-labeled fibrinogen was added to plasma to enable imaging fibrin deposition. Alternatively, fibrin derived from native fibrinogen was imaged with fluorescein-labeled mAb 59D8, specific for the beta-chain of fibrin. Streptococci pyogenes strain 700294 was obtained from ATCC and grown and plated using standard techniques.
Quiescent endothelial cells responded to TNFα by retracting from the cell-cell junctions, as previously described. As we previously reported, the filopodia and localized regions on the cell margins stained with lactadherin, indicating phosphatidylserine exposure and convex curvature, while the cell bodies did not. The plasma-saline mixtures incubated over endothelial cells did not clot over a 2 hr time span, but plasma incubated over control gelatin-coated coverslips clotted within 10 min. Confocal microscopy indicated that fibrin strands were deposited between TNFa-treated endothelial cells but were largely absent from cover slips with untreated cells. Soluble, fluorescent fibrin bound to the margins of TNFα-treated endothelial cells but not to quiescent endothelial cells or to the Sigmacote-treated cover slips. Fibrin deposition was greatly diminished by reducing Ca++ to < 1 mM, by an anti-tissue factor mAb, and by hirudin. The quantity of fibrin was increased by an anti-tissue factor pathway inhibitor antibody but not by corn trypsin inhibitor, anti-activated protein C mAb, anti-thrombomodulin mAb, or anti-factor VIII mAb. Thus, focal procoagulant activity is mediated by the extrinsic coagulation pathway. To determine whether the fibrin deposition has a function in innate immunity, Streptococcus pyogenes, was added to plasma at a concentration of 106cfu/ml. The streptococci concentration was decreased to <40% in three hours in the presence of TNFa treated endothelial cells, but increased more than 3-fold in the presence of control cells. Addition of hirudin to the TNFα-treated cells enabled Streptococci to grow at the same rate as in citrated plasma alone or re-calcified plasma incubated over quiescent endothelial cells. Confocal microscopy indicated that streptococci decorated the fibrin strands in the inter-endothelial space but not the intercellular matrix or the endothelial cell bodies.
TNFα treated HUVECs support focal procoagulant activity via the extrinsic coagulation pathway while maintaining anticoagulant function toward bulk plasma. Focal endothelial procoagulant activity leads to cell-bound fibrin and anti-streptococcal activity, independent of leukocytes and platelets.
Focal support of procoagulant activity by endothelial cells may have a role in innate immunity without clotting of bulk plasma and independent of leukocytes or platelets. The conditions under which vascular beds react in this manner, the breadth of anti-bacterial activity, and the mechanisms of coagulation-dependent anti-bacterial activity are promising areas for further investigation.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.