Cultured human mesangial cells produce both type 1 and type 2 plasminogen activator inhibitors.

Cultured human mesangial cells (HMC) derived from normal kidneys have been shown to synthesize tissue-type plasminogen activator (t-PA) and excess amounts of PA inhibitor type 1 (PAI-1). Conflicting results have been obtained concerning the production of urokinase-type PA (u-PA) and efforts to show PA inhibitor 2 (PAI-2) met with failure. We evaluated the fibrinolytic profile of cultured HMC lines obtained from 12 patients with renal carcinoma and one cadaveric kidney donor. Subconfluent cells (third passage) were incubated overnight in serum-free medium. t-PA, u-PA, PAI-1 and PAI-2 antigens were assayed by ELISA methods and PA and PAI activities by amidolytic methods both in conditioned medium (CM) and cell extracts (CE). Besides PAI-1, PAI-2 antigen was detected in all but one HMC lines. At variance with the former, which was largely released in the culture medium, PAI-2 was mainly cell-associated. t-PA antigen was found in all but two cell lines while u-PA antigen was detected in relatively high concentrations in 8 cell lines. PA activity, identified as u-PA by functional and immunological criteria, was measured in CM of six of the eight u-PA producing cell lines, whereas PAI activity was undetectable or very low in CM of all cell lines, suggesting that PAI-1 was largely inactive. Functional assays of cell extracts demonstrated the presence of PA activity, again identified as u-PA, only in samples (five lines) containing u-PA antigen in excess over PAI-2. PAI activity was found instead in the extracts in which the inhibitor was higher than the activator (six lines) and was identified as PAI-2, as it inhibited u-PA but not single-chain t-PA and was neutralized by a polyclonal anti-PAI-2 antibody. The heterogeneous fibrinolytic pattern of HMC lines was confirmed by mRNA analysis of three representative lines. Results were similar when HMC lines at passage five were used, except that the u-PA content was significantly reduced both in CM and CE. These findings indicate that the fibrinolytic profile of cultured HMC is more complex than previously reported. The production of large amounts of PAI-2 may represent an additional control mechanism of proteinase activity.

Monocytes, but not endothelial cells, downregulate the anticoagulant activity of activated protein C.

Activated protein C (APC) is a natural anticoagulant and inhibits thrombin generation by degrading factors Va and VIIIa. We evaluated the ability of APC to inhibit blood coagulation triggered by lipopolysaccharide (LPS)-stimulated [tissue factor (TF)-expressing] human mononuclear cells (MNCs) or umbilical vein endothelial cells (HUVECs). Using a plasma recalcification assay, we found that APC (up to 53.3 nmol/l final concentration) had a poor anticoagulant effect in the presence of LPS-stimulated MNCs, whereas it caused a marked prolongation of clotting time in the presence of LPS-stimulated HUVECs. A poor response to APC was also observed when platelet-free MNCs, monocyte-enriched preparations or the monocytoid cell line U937 were tested. Using a TF-independent (FXa-induced) thrombin generation assay, we demonstrated that both LPS-stimulated and unstimulated MNCs negated the inhibitory activity of APC. Direct determination of FVa activity indicated that MNCs were less efficient than HUVECs in promoting FVa inactivation by APC. Together, our results suggest that MNCs, at variance with HUVECs, protect factor Va from inactivation by APC, probably through the expression of a membrane component not present on endothelial cells. These strengthen the importance of monocytes in fibrin deposition associated with pathological conditions characterized by monocyte recruitment and activation.