Inhibition of microvascular endothelial apoptosis in tissue explants by serum albumin.

Plasma factors appear to inhibit endothelial cell (EC) apoptosis in vivo so that flow influences microvascular form. The identity of these factors has not, however, been established. Earlier, we reported that apoptosis in isolated, serum-deprived human EC is inhibited by albumin (Alb). Here, we demonstrate likely biological relevance of this to vascular remodelling in experiments with tissue explants. Rat skin explants were incubated in medium M199 with or without serum, bovine Alb, or human Alb. EC in paraffin sections of explants were labelled by lectin histochemistry and the relative proportion of apoptotic was EC determined. Apoptosis was confirmed by transmission electron microscopy and terminal deoxynucleotidyl transferase labelling. Serum-free culture induced EC apoptosis (P < 0.02) and this was strongly inhibited by Alb at physiological concentrations (P < 0.01). This was not a nonspecific protein effect, as mercaptoethanol denaturation destroyed the activity and ovalbumin was not protective. Also, protection was not due to serum contaminants, as recombinant human Alb had activity identical to that of native material. The dose response was identical for all Alb preparations tested, with maximal activity at physiological concentrations. Protection was not limited to rat tissue as similar results were obtained with human gingival explants. These data support a role for Alb as a plasma antiapoptotic factor for EC in tissues.

Fibrinolytic proteins in apoptotic human umbilical vein endothelial cells.

Endothelial cells express fibrinolytic proteins including: urokinase (u-PA) and tissue type (t-PA) plasminogen activators, type-1 (PAI-1) and 2 (PAI-2) plasminogen activator inhibitors, and u-PA receptor (u-PAR). Apoptotic endothelial cells detach, potentially forming both local and circulating microthrombi in vivo. In this article, apoptotic human umbilical vein endothelium was obtained by serum starvation and compared with nonapoptotic cells rescued from death with fresh medium containing serum. Antigen levels for t-PA, PAI-1, PAI-2, and u-PAR were reduced greatly in apoptosis (p< 0.05). In contrast, u-PA levels were similar in apoptotic as compared with rescued cells (p<0.05). Radioactive amino acids were used to determine absolute levels of protein synthesis and degradation in these cells. Reduced antigen levels likely were due to proteolysis as there was 98% total protein degradation and very little protein synthesis in apoptotic endothelial cells. Also, u-PA levels in apoptotic endothelial cells were not affected by the protein synthesis inhibitor cycloheximide. Endothelial cells in inflammatory sites are exposed to cytokines, which increase both apoptosis and u-PA levels. Data from this article support the idea that maintained u-PA levels in apoptotic endothelium may protect from micro-thrombosis in inflammatory sites as well as in the circulation.

Serum albumin is a specific inhibitor of apoptosis in human endothelial cells.

Excess blood vessels are removed by apoptosis of endothelial cells, however, the signals responsible for this have not been defined. Apoptosis of cultured human umbilical vein endothelial cells is induced by deprivation of serum or adhesion. In this paper, apoptosis in human umbilical vein and microvascular endothelium was induced by deprivation of serum and or adhesion. Apoptosis was confirmed on the basis of morphology, ultrastructure and internucleosomal cleavage of DNA. Loss of endothelial adhesion was found to be an early event in cultured endothelial cell apoptosis and was exploited to quantitate apoptosis. The effect of: bovine serum albumin; human serum albumin; recombinant human albumin; dithiothreitol reduced human and bovine albumin; CNBr treated human and bovine albumin as well as ovalbumin upon endothelial apoptosis was determined. Native bovine and human albumin as well as recombinant human material inhibited apoptosis at physiological concentrations with identical dose response curves in both umbilical vein and microvascular cells. Dithiothreitol treatment destroyed all protective activity while bovine but not human albumin was partially inactivated by CNBr treatment. The unrelated protein ovalbumin was not protective. Albumin did not inhibit apoptosis if cells were also deprived of adhesion. The data suggest that albumin is a specific inhibitor of human endothelial apoptosis but does not protect cells also deprived of adhesion. Reduced supply of albumin to endothelium in poorly perfused blood vessels may provide a mechanism for the removal of excess blood vessels in remodelling tissues. Also, the failure of albumin to protect endothelial cells deprived of adhesion from apoptosis may reflect the need to remove potentially micro-embolic cells detached due to trauma.

Binding of urinary protein C inhibitor to cultured human epithelial kidney tumor cells (TCL-598). The role of glycosaminoglycans present on the luminal cell surface.

Binding of urinary protein C inhibitor (PCI) to cultured human epithelial kidney tumor cells (TCL-598) was studied. Binding was dose-dependent, time-dependent, and saturable. Heparin interfered in a dose-dependent way with PCI binding to TCL-598 as did heparan sulfate and to a lesser degree also dermatan sulfate. Pretreatment of TCL-598 with protamine sulfate inhibited subsequent binding of PCI in a dose-dependent manner and > 100 micrograms/ml protamine sulfate reduced binding of PCI to < 10% of the control. Binding of 125I-PCI was specific, and bound 125I-PCI was recovered from the cells by heparin treatment or detached together with intact cells by EDTA treatment, migrated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis with the same mobility (M(r) = 57,000) as unbound 125I-PCI. Furthermore, cell-bound PCI was functionally active as judged from its ability to inhibit the amidolytic activity of urokinase, and its inhibitory activity was stimulated approximately 3-4-fold as compared to fluid-phase PCI. Immunogold electron microscopy revealed that PCI-antigen presented to the cells from the luminal side bound exclusively to that surface in native as well as in prefixed cells. This binding of PCI was abolished in the presence of heparin (50 micrograms/ml) and after pretreatment of the cells either with protamine sulfate (400 micrograms/ml) or with heparinase III (0.5 unit/ml). A slight decrease in PCI binding was seen after pretreatment of the cells with chondroitinase ABC and chondroitinase AC. In contrast, binding of PCI to extracellular matrices of TCL-598 was decreased to approximately 70% after chondroitinase ABC treatment of the extracellular matrices, whereas both heparinase III or chondroitinase AC treatment only reduced matrix-bound PCI to approximately 95%. These data suggest that heparan sulfate-containing proteoglycans are predominantly involved in binding of PCI to the luminal side of TCL-598, while dermatan sulfate-containing proteoglycans, the overall predominant PCI-binding proteoglycans in TCL extracts, are responsible for PCI binding to the extracellular matrix. Heparan sulfate, however, exposed to an environment containing PCI under physiological conditions, might localize PCI and modulate its target enzyme specificity in vivo.