Bacterial interaction with the fibrinolytic system and extracellular matrix

Bacterial infection of host organisms is a state of complex interaction between the parasite and the host. Two critical steps in the infectious process are: Adhesion to the epithelial surfaces and penetration of tissue barriers to gain entrance to tissues and into the circulation. These two steps have been the subjects of this thesis.

A method to spectrophotometrically measure the rate of formation of plasmin was established. This method was used to quantitatively measure how bacterial binding of plasminogen affects t-PA and u-PA mediated activation to plasmin. A method to study bacterial penetration of basement membranes in vitro based on a double chamber system was also established and used to investigate how bacterial capture of plasminogen affect the bacterial ability to penetrate membranes.

63 strains representing 7 species with the capability to bind plasminogen were tested for their ability to enhance t-PA and u-PA mediated plasminogen activation. Most strains were able to enhance the t-PA mediated formation of plasmin considerably, in some cases up to 130-fold (Strepococcus equisimilis). Within a species there was a correlation between the degree of enhancement of activation and level of binding of plasminogen (r=0.74-0.92). Both binding and activation could be inhibited by addition of the lysine analogue 6-aminolhexanoic acid, indicating that lysine binding structures of the plasminogen molecule are essential for both these phenomena.

Three strains of Haemophilus influenzae were tested for interaction with the extracellular matrix. We were able to demonstrate adhesion to enzymatically prepared extracellular matrix as well as to a reconstituted basement membrane. When testing with purified matrix components it was found that fibronectin, collagen type I and collagen type III conferred adhesion. The unencapsulated strains also adhered to laminin and collagen type V. Preincubation of the bacteria with plasminogen improved the capability to penetrate basement membranes, reducing the time for penetration from over 20 hours to about 3 hours.

A following study of by Neisseria meningitidis demonstrated a binding to collagen type I, III, V and fibronectin also for this bacterium. Adhesion to fibronectin was detected only to surface associated but not solubilised fibronectin. This interaction was attributed to the central 75 kD fragment without involvement of the cell binding RGD sequence: This is the first report of such an interaction by a microorganism.

In a subsequent investigation 17 strains of Streptococcus pneumoniae were tested for interaction with ECM components and all adhered to fibronectin and collagen type Ill. A majority of strains also adhered to laminin and collagen type I and V. Furthermore, pneumococci could use captured plasminogen to penetrate reconstituted basement membranes as time for penetration was reduced from more than 21 hours to 4 hours.

Finally, we were able to detect vitronectin binding to Haemophilus influenzae. Only surface associated vitronectin would bind and the interaction was not mediated by fimbriae or a 28 kD invasion factor. The adhesion can be inhibited by heparin indicating that the heparin binding domain of the vitronectin molecule is involved in this interaction.