Fibrin formation by wounded bronchial epithelial cell layers in vitro is essential for normal epithelial repair and independent of plasma proteins.

BACKGROUND: The bronchial epithelium is in contact with, and continually damaged by, the environment. Animal models have indicated that normal epithelial repair is rapid and supported by the formation of a provisional fibrin matrix that is exclusively plasma-derived. OBJECTIVES: Our objectives were to demonstrate the ability of normal human bronchial epithelial (NHBE) cells to produce coagulation cascade proteins and form fibrin in response to damage, independently of plasma proteins, and to show that formation of a cross-linked fibrin matrix is essential for normal epithelial repair in vitro. METHODS: Primary NHBE cells and cells of the 16HBE 14o- bronchial epithelial cell line were grown and maintained in vitro prior to mechanical wounding of confluent monolayers in serum-free media. Tissue factor (TF) and factor XIII (FXIII) were visualized on 16HBE 14o- monolayers using immunohistochemistry. The time-dependent expression of TF, factor VII (FVII), factor X (FX), fibrinogen, soluble fibrin, FXIII subunit A (FXIIIA) and D-dimers following wounding of confluent 16HBE 14o- monolayers was investigated using immunoassays. TF and FVII expression at the mRNA level was investigated by RT-PCR. The role of coagulation cascade proteins in the repair response of NHBE and 16HBE 14o- monolayers was investigated using neutralizing antibodies. RESULTS: Active TF was constitutively expressed in 16HBE 14o- cells. Levels of FVII, FX, fibrinogen, soluble fibrin, FXIIIA and D-dimers in culture supernatants increased rapidly and were maximal 20 min after wounding the monolayers. Expression of TF and FVII mRNA was significantly increased 10 and 4 h, respectively, after wounding. Neutralizing antibodies to TF, fibrinogen and FXIIIA significantly inhibited repair of NHBE and 16HBE 14o- cell layers. CONCLUSIONS: The bronchial epithelium has the potential to respond rapidly to mechanical damage by forming a cross-linked fibrin matrix that is essential for normal epithelial repair, independently of plasma proteins.

Preferential antagonism of the interactions of the integrin alpha IIb beta 3 with immobilized glycoprotein ligands by snake-venom RGD (Arg-Gly-Asp) proteins. Evidence supporting a functional role for the amino acid residues flanking the tripeptide RGD in determining the inhibitory properties of snake-venom RGD proteins.

The inhibitory properties of a panel of snake-venom-derived RGD (Arg-Gly-Asp) proteins, including the disintegrins kistrin, elegantin and albolabrin, and the neurotoxin homologue dendroaspin, were investigated in a platelet-adhesion assay using three immobilized ligands of the glycoprotein IIb-IIIa complex (alpha IIb beta 3), namely fibrinogen, fibronectin and von Willebrand factor (vWF). The snake-venom proteins preferentially inhibited the adhesion of ADP-treated platelets to one or more of the immobilized ligands. Kistrin and dendroaspin exhibited similar inhibitory characteristics, abrogating platelet adhesion to fibrinogen and vWF at nanomolar concentrations, but poorly inhibiting adhesion to fibronectin. Kistrin and dendroaspin share little overall amino-acid-sequence identity, but a considerable level of sequence similarity exists around the RGD tripeptide. Synthetic cyclic peptides corresponding to these regions of kistrin and dendroaspin inhibited platelet adhesion to both fibrinogen and fibronectin with approximately equal potency, but were 100-fold weaker antagonists of the interactions of the alpha IIb beta 3 complex with fibrinogen than their parent proteins. The disintegrins elegantin and albolabrin, which share approx. 60% overall amino-acid-sequence similarity with kistrin but have different residues around the RGD tripeptide, exhibited different antagonistic preferences. Elegantin inhibited platelet adhesion to immobilized vWF and fibronectin, but was significantly less effective at disrupting adhesion to fibrinogen. Albolabrin selectively inhibited platelet adhesion to immobilized vWF and was less effective with fibrinogen and fibronectin as adhesive ligands. In contrast with the behaviour of these venom proteins, the adhesion of ADP-treated platelets to immobilized fibrinogen, fibronectin and vWF was inhibited non-selectively by a range of monoclonal antibodies with specificity for the alpha IIb beta 3 complex. These observations, therefore, define antagonistic preferences in this panel of venom proteins towards the interactions of the alpha IIb beta 3 complex with three immobilized glycoprotein ligands.

An independent haemostatic mechanism: shear induced platelet aggregation.

We have published (1) evidence indicating that high shearing forces alone applied to platelets expose and activate a unique domain on glycoprotein IIb/IIIa (GPIIb/IIIa) at the platelet surface. In the presence of von Willebrand's factor (vWf) and divalent cations the platelets will aggregate. This paper reviews the extensive literature on high shear effects. It describes a device in which high shear produced by forcing heparinised whole blood through a complex filter normally results in platelet activation; the platelets aggregate and then block the filter. This system is inhibited by antibodies to GPIIb/IIIa and to vWf: fibrinogen is apparently not involved. The same antibodies to GPIIb/IIIa and vWf prevent high shear induced thrombosis occurring in vivo in animal models. The filter blockage is not influenced by aspirin, heparin and ticlopidine and so involves a different mechanism from the aspirin sensitive mechanisms involved in clinical thrombosis prevention in vivo in man. While there are a number of unexplained phenomena in this global test nevertheless this filter model is a simple way of studying a recently recognised pathway which is almost certainly involved in thrombogenesis in man.

Shear stress activation of platelet glycoprotein IIb/IIIa plus von Willebrand factor causes aggregation: filter blockage and the long bleeding time in von Willebrand's disease.

The article explores the finding that high shear alone applied to normal, native blood results in platelet aggregation. A filter with tortuous capillary-sized channels permits a study of the effect of shearing forces at different pressures. Native, heparinized, citrated and EDTA blood and platelet-rich plasma (PRP) were forced through the filter. Normal and von Willebrand's blood were studied, as were the effects of antibodies to platelet glycoproteins (GPr) and to von Willebrand's factor (vWf) and of "membrane-active" drugs. Normally, the filter blocked at 40 mmHg but not at 5 mmHg. Transmission electronmicroscopy of the filter at 40 mmHg showed blockage by platelet aggregates. Initially, the mean transit time through the filter was 8 milliseconds. Platelet retention in the filter occurred in two phases. From 0 to 3 seconds, only high-shear, vWf, and GPrIIb/IIIa were required. From 10 to 20 seconds, retention presumably involved these three attributes, but divalent cations were also essential. Only this phase was inhibited by some membrane-active drugs. ADP- and thrombin-induced aggregation requires GPrIIb/IIIaand fibrinogen. Shear-induced blocking of the filter by blood with a normal concentration of fibrinogen requires GPrIIb/IIIa and vWf. This indicates a different type of exposure of GPrIIb/IIIa. The long bleeding time in vW disease highlights the absolute requirement for vWf and emphasizes the difference in exposure of GPrIIb/IIIa induced by shear stress. Evidently, a process similar to that occurring in the filter is required in normal capillary hemostasis.