Can RAP save your brain?

Stroke is still one of the major causes of death and disability worldwide. Despite decades of clinical research, the therapeutic value of neuroprotective agents that can limit damage to the postinfarct neuronal tissue is limited.¹  The only highly effective treatment is a fast reperfusion of the ischemic brain tissue through dissolution of the culprit blood clot using thrombolytic agents. Based on a landmark study from the National Institute of Neurological Disorders and Stroke,²  it is well established that treatment with tissue-type plasminogen activator (t-PA) within 3 hours after onset of symptoms results in a clear benefit for patients during follow-up, including a 30% higher chance of remaining disability-free after 3 months. Still, a major complication is an almost 10-fold increase of intracranial bleeding, with a prevalence as high as 6.4% in treated patients. Debate continues on the therapeutic window when benefits of t-PA treatment still outweigh its deleterious bleeding complications.³  Clinical studies currently focus on the use of imaging technology to select patients who might benefit most from this therapy.³ 

In the current issue of Blood, the paper by Suzuki et al takes a more optimistic approach in trying to prevent the complications of thrombolytic therapy without compromising on clot lysis.⁴  It builds on a series of basic research studies into the molecular mechanisms that control the breach of the brain vasculature by t-PA–provoked degradation of the endothelial basal membrane. This degradation has been linked to the activation of matrix metalloproteinases (MMP) by profibrinolytic plasmin, the active protease produced by t-PA from circulating plasminogen.⁵  Initial studies showed that broad inhibitors of MMPs indeed limit bleeding complications induced by t-PA administration in mouse models for stroke. Unfortunately, there are many noxious side effects, and prolonged treatment with such inhibitors actually impairs functional recovery, as MMPs are crucial for postischemic neurovascular repair.⁵  A crucial observation to bypass this caveat came from an earlier study of Suzuki et al. In that study, they identified MMP3 to be solely responsible for intracranial bleeding after t-PA treatment.⁶  Plasmin did not play a significant role in this process, however, making the link between t-PA and MMP3 rather obscure. Novel insights came from Dr Yepes' group, who showed that LRP, the LDL-receptor related protein, is the receptor for t-PA in the brain, and this binding is essential for opening of the blood brain barrier.⁷  This occurs via signaling through nuclear factor kappa B (NF-κB), indicating that a signaling event rather than plasmin drives the effects of t-PA on breakdown of the vascular basal membrane.⁸  This identified a crucial moonlighting function for a receptor that was cloned in 1988 by Dr J. Herz and was shown to play a crucial role in the lipoprotein metabolism in the liver, although already an exceptionally high expression in the brain was noted.

The current paper by Suzuki et al now integrates many of these findings and resolves several of the basal questions.⁴  Using both cultured brain endothelial cells and a mouse model for ischemic stroke, Suzuki et al show that t-PA binding to LRP indeed induces activation of NF-κB, which increases expression of MMP3. Interestingly, this was found in endothelial cells, which normally express LRP at modest levels compared with astrocytes,⁷  but boost LRP expression in response to ischemia, making it highly relevant for ischemic stroke.⁴  Indeed, the mouse model showed that induced expression of LRP and concomitantly of MMP3 after t-PA treatment is found specifically in the endothelial cells surrounding the infarcted brain area. The key finding is that the natural LRP-inhibitor, receptor-associated protein (RAP), was able to prevent these effects. Indeed, injection of RAP before administration of t-PA completely prevents the t-PA–induced intracranial bleeding, bringing it to levels almost identical to those in the control animals. This shows for the first time in a therapeutic setting rather than a genetically modified mouse model that it is possible to maintain t-PA's clot lysing abilities while at the same time preventing its deleterious MMP3-activating effects.

Two major issues remain. First, it is shown that inactive t-PA does not signal through LRP. As plaminogen has been excluded as target protease for the t-PA–induced breakdown of the blood brain barrier,^6,7  this leaves us with the question: what would be the target for t-PA's proteolytic activity? It certainly suggests that the signaling events to NF-κB are not directly involving LRP itself, but rather that LRP might merely serve to concentrate t-PA at the endothelial cell surface to attack it true substrate. Could it be the thrombin receptors, the PARs known to signal through NF-κB but which are normally not a substrate for t-PA? Resolving this issue might be crucial to developing a better agent to treat stroke patients. Second and more disturbing is the fact that ischemia induction of LRP in both cultured cells and mouse brain vasculature is a rather slow event. Inductions occur after about 6 to 24 hours of ischemia. Indeed, in the current study, t-PA was administered 6 hours after the onset of ischemia, concurrent with maximal LRP induction. Relevance for human patients treated within the first hour to 3 hours after onset of symptoms thus remains to be established. In effect, there are no significant differences in bleeding complications between groups of patients treated in either the first 3 hours or after 3 to 6 hours.³  Follow-up studies in different animal models for stroke, using different timing and administration regimens, are needed to better understand the exact implications of the current findings.

Still, the Suzuki paper establishes the possibility to devise conjunctive therapies to t-PA treatment in stroke patients that would prevent the strongly enhanced bleeding tendencies. Furthermore, this might lead to inclusion of “late-arriving” patients who, under the current regimen, would not benefit from thrombolytic treatment because the complications outweigh the benefits. Whether RAP itself would be a good adjunctive agent for thrombolytic therapy or better pharmaceuticals can be developed based on these insights does not limit the promise held by these results.