In vivo evaluatie van de simultane inhibitie van TAFI en PAI-1: op weg naar een effectieve en veilige trombolyse

Thrombin activatable fibrinolysis inhibitor (TAFI) and plasminogen activator inhibitor-1 (PAI-1) are important regulators of fibrinolysis. Bothproteins attenuate the tissue-type plasminogen activator (tPA)-mediatedconversion of plasminogen into fibrin-degrading plasmin, however with distinct modes of action. Activated TAFI (TAFIa) removes C-terminal lysines from the fibrin surface, rendering fibrin less susceptible to fibrinolysis. PAI-1 neutralises tPA by forming a covalent complex and is considered as the main inhibitor of tPA. Elevated levels of TAFI and PAI-1 arelinked to an increased risk for thrombosis. There is an unmet clinical need to improve thrombolytic treatment, especially in acute ischemic stroke (AIS). The only drugs available for thrombolysis comprise plasminogen activators. However, plasminogen activators remain underutilised owing to significant drawbacks, e.g. a low recanalising rate, high bleeding risk and possible neurotoxicity. Therefore, we embarked on a novel thrombolytic strategy: dual targeting of TAFI and PAI-1 to effectively enhance endogenous fibrinolysis. In a previous study, a first heterodimer diabody against human TAFI and human PAI-1 was generated that exhibited a strong profibrinolytic capacity in lysing experimental human thrombi. A heterodimer diabody is a recombinant antibody derivative with dualspecificity and was used in this study to circumvent complications associated with the administration of two different inhibitors. In the present study, we evaluated the application of dual TAFI/PAI-1 inhibition through a diabody in vivo. In chapter 2, a recombinant bispecific diabody was generated from MA-RT36A3F5 directed towards rat TAFI (with cross-reactivity to mouse, but not human TAFI) and MA-33H1F7 directed towards human PAI-1 (with cross-reactivity to mouse and rat PAI-1). MA-RT36A3F5 inhibits TAFI by destabilising TAFIa. MA-33H1F7 converts active PAI-1 to a substrate of tPA resulting in the inactivation of PAI-1. Although the diabody was successfully expressed in bacteria and it preserved the inhibitory properties of the parental antibodies, the profibrinolytic effect was transient because of its low plasma stability. Through antibody engineering (e.g. CDR-grafting and linker addition), we obtained a stabilised version of this diabody with comparable profibrinolytic properties as those of the parental antibodies. Inchapter 3, a second candidate diabody was generated from MA-TCK26D6, directed towards a stable variant of human TAFI, and MA-33H1F7, also used in the previous diabody, against PAI-1. MA-TCK26D6 is a first-in-class antibody to have cross-reactivity towards human and rodent TAFI. A great asset of the corresponding diabody, Db-TCK26D6x33H1F7, is that it can beevaluated in both preclinical as clinical settings. The diabody retained the functional properties of the parental antibodies and did not encounter stability issues. This resulted in a strong profibrinolytic effect in vitro which exceeded that of the combination of parental antibodies. Moreover, the diabody did not impair the beneficial anti-inflammatory properties of TAFIa. Therefore, Db-TCK26D6x33H1F7 was selected for furtherevaluation in vivo. In chapter 4, the profibrinolytic properties of the parental antibodies or corresponding diabody (Db-TCK26D6x33H1F7) were first assessed in a mouse model of thromboembolism evoked bysystemic administration of tissue factor. Whereas the antibodies, administered separately or combined, caused a half-maximal clearance of fibrin from the lungs, the diabody, administered at a corresponding dose, resulted in a maximal effect. In a model of cerebral ischemia/reperfusion, whereas the administration of a single parental antibody at a certain dose caused no effect, the administration of the combination of antibodiesor the diabody resulted in a complementary effect in mitigating ischemia/reperfusion injury (reduced lesion volumes and improved neurological and motor outcomes). The diabody was slightly more effective than the combination of antibodies in improving functional outcome. The superior profibrinolytic capacity of the diabody is most likely ascribed to its relatively small (58 kDa) and flexible structure. In chapter 5, the thrombolytic capacity of Db-TCK26D6x33H1F7 was assessed by administering the diabody post thrombus formation in two types of thrombosis-induced stroke models. The effect of the diabody was compared to thatof tPA, the current treatment of AIS. In the thrombin-induced model, fibrin-rich clots are produced which are susceptible to tPA-mediated fibrinolysis. 24 hours post stroke onset, diabody administration caused a substantial reduction in lesion sizes which exceeded the effect of tPA administration. In the FeCl(3)-induced model, platelet-rich clots are produced which are resistant to tPA-mediated fibrinolysis. 24 hours post stroke onset, whereas tPA was non-effective, the diabody recanalised the artery, improved brain perfusion and reduced lesion sizes. Moreover, in contrast to tPA, the diabody did not prolong bleeding times. Inconclusion, this study clearly demonstrates the strong profibrinolytic properties of a dual targeting strategy against two antifibrinolytic proteins, TAFI and PAI-1, in vivo. The two main findings of this study are that (i) dual targeting of TAFI and PAI-1 is more effective with a diabody than with the combined use of parental antibodies and (ii) the diabody is more effective and potentially safer than tPA in the treatment of experimental stroke.nrpages: 140status: publishe