Bat plasminogen activator: Desmoteplase - From bat to bench to bedside of stroke victims

Abstract

Desmoteplase (DSPA) was identified in the salivary venom of Desmodus rotundus (a blood-feeding vampire bat common in Latin America) triggered by observations as early as 1964. The initial interest in DSPA as a therapeutic modality was raised by the success of recombinant tissue plasminogen activator (rt-PA) in acute myocardial infarction (AMI) and the evolving paradigm of fibrin-specificity as the key to safe and effective thrombolysis. The early research on DSPA confirmed an extremely high fibrin specificity and a potential for a lower bleeding propensity, demonstrated in a variety of preclinical studies. Obviously, the unique task in the Vampire Bat - curbing clot formation without disintegrating the other salivary proteins - has led to a protease which serves no other known function than activating plasminogen in the presence of fibrin. This fibrin and substrate specificity distinguishes DSPA from rt-PA, which, apart from clot lysis, has a number of additional physiological roles. The first clinical trial in AMI confirmed the absence of fibrinogen depletion even at doses of 750∈μg/kg (90∈μg/kg is effective in acute ischaemic stroke, AIS). DSPA was abandoned by Schering AG for strategic reasons and, in 2001, its further development was redirected by PAION to acute ischemic stroke (AIS), based on the assumption that a more fibrin-specific thrombolytic should pose a lower bleeding risk and allow a longer post-stroke treatment window. Also in 2001, the discovery of the ability of rt-PA to enhance NMDA-induced neurotoxicity (which models the glutamate neurotoxicity known in vivo) gave rise to the speculation that DSPA, by virtue of its high specialization, might be different. The subsequent multi-level research indeed confirmed that DSPA is devoid of any neurotoxic properties. The main reason seems to lie in a structural difference: The deleterious augmentation of NMDA neurotoxicity by rt-PA (and its mutants, shown in vitro) requires the kringle 2 (K2) domain, a moiety lacking in DSPA. This distinction may prove advantageous also in other areas, as, more generally, the kringle 2 seems to capacitate (r)t-PA for numerous mechanisms it is known to operate, sometimes to a harmful end. An important example is its K2-dependent ability to activate platelet-derived growth factor (PDGF), which is instrumental in (r)t-PA induced weakening of the blood brain barrier. DSPA is now in a global phase III program in AIS, sponsored by PAION's partner Lundbeck, treating patients up to 9 h after stroke onset, a wide extension of the current window of 3-4.5 h.