10th International Symposium on Pharmacology of Cerebral Ischemia
July 25-28, 2004, Marburg, Germany

Summary
The 10th International Symposium on Pharmacology of Cerebral Ischemia was a gathering of more than 250 scientists from all over the world at the Philipps-University of Marburg, Germany. Organized by Josef Krieglstein with the support of an international advisory board the meeting was again a great success combining international contributions of outstanding scientists and clinicians in the field of ischemia and neuroscience research. During the conference basic research on mechanisms of neuron death and therapeutic strategies explored in experimental models of cerebral ischemia, as well as clinical trials of known drugs and novel compounds under development were presented in 90 poster presentations and 56 oral presentations divided in 7 sessions and a plenary lecture. Prominent areas discussed were: newly discovered mechanisms involved in ischemic brain damage such as the activation of acid-sensing ion channels (ASIC), matrix-metalloproteinases and different (apoptotic) intracellular death signalling pathways; neuroprotective drugs including activated protein C (APC), novel docosahexaenoic acid derivatives and a carbamylated erythropoietin-derivative; new strategies to enhance endogenous repair mechanisms involving angiogenesis and neurogenesis. In addition, promising results from clinical stroke trials on edaravone and albumin therapy were reported. On the basis of known mechanisms and novel insights into the pathology of ischemic brain damage new cerebroprotective strategies were presented that will - individually or in combination - lead to hopeful approaches in stroke therapy.

Vascular damage and remodeling after stroke
Cerebrovascular dysregulation is a key feature in stroke pathology, and, in particular, in hemorrhagic transformation after ischemia. Under physiological conditions the balance between energy demands due to neural activity and substrate delivery through blood flow is tightly regulated by a functional unit of neurons, astrocytes and vascular cells as reviewed by Gregory del Zoppo (La Jolla, USA). After ischemic injury, the activation of matrix metalloproteinases (MMP) contributes to the breakdown of extracellular matrix constituents which leads to a destruction of the microvasculature in the ischemic brain tissue. As also emphasized by Eng Lo (Charlestown, USA) and Carolina Maier (Stanford, USA) the activation of MMP-9 is a major cause for endothelial apoptosis, vascular dysfunction and hemorrhagic transformation after stroke. MMP levels are therefore suggested as a biomarker that indicate an increased risk of hemorrhagic conversion after an ischemic insult. Inhibitors of MMPs could be useful in stroke therapy to rescue the extracellular matrix thereby reducing endothelial cell apoptosis and the risk of hemorrhagic transformation.

Berislav Zlokovic (Rochester, N.Y., USA) introduced the therapy with activated protein C (APC) as another promising strategy to protect the neurovascular unit after stroke. APC is a systemic anti-coagulant and anti-inflammatory factor that protects the brain against ischemic injury by acting directly on brain cells. In endothelial cells the antiapoptotic effect of APC correlated with a reduced activation of caspases 3 and 8, and a reduced release of cytochrome c from mitochondria. In addition, APC also blocked p53 activity and normalized the pro-apoptotic Bax/Bcl-2 ratio in these cells. Since APC also prevented apoptosis in neurons and enhanced cerebral blood flow after ischemia/reperfusion in mice, the protein appears attractive for stroke therapy.

In addition to the rescue of the neurovascular unit against ischemic injury the recovery of cerebral blood flow may be achieved by angiogenesis. Samuel Valable (Caen, France) demonstrated a role for the angiogenic factors vascular endothelial growth factor (VEGF) and angiopoietin-1 (Ang-1) in neuroprotection and vascular remodelling after cerebral ischemia. In the first hours after stroke, these factors prevented further damage of brain tissue by activation of the phosphoinositide-3-phosphate kinase (PI3K)/protein kinase B (Akt) survival pathway, whereas at later phases after the insult the support of repair mechanisms involving angiogenesis may contribute to an improved recovery of neurological functions. VEGF alone, however, may support hemorrhagic transformation of an ischemic infarct by induction of MMP-9. Interestingly, Ang-1 counteracts VEGF-mediated upregulation of MMP-9 thereby preserving the integrity of the blood-brain-barrier.

Konstantin-Alexander Hossmann (Cologne, Germany) demonstrated that granulocyte macrophage colony stimulating factor (GM-CSF) stimulates arteriogenesis in hypoperfused brain areas, which appears to be a useful strategy for remodeling of supply vessels that improve brain hemodynamic parameters and prevent tissue damage during a following ischemic episode.
Overall, neurovascular protection, angiogenesis or arteriogenesis emerged as a hopeful strategy to preserve the delicate balance of blood supply and energy demand in the brain's microenvironment and, therefore, to reduce neurological deficits after an ischemic insult.

Intracellular signalling in ischemic neuron death
After cerebral ischemia, glutamate-mediated excitotoxicity and the extensive production of reactive oxygen species mediate, in large part, the neurotoxic effect of an ischemic insult in the brain. However, the intracellular signalling cascades involved in ischemic neuron death have not been fully unravelled. While most neurons in the core of an ischemic infarct are damaged in a passive manner and die by necrosis due to the lack of energy and oxygen supply, neurons in the penumbra area often expose hallmarks of an active biochemical cell death cascade called apoptosis. This cascade is triggered by the disruption of cellular calcium homeostasis and oxidative stress, and involves mitochondrial dysfunction, cytochrom c release and activation of caspases. The cytotoxic accumulation of intracellular calcium has been well established as a key step in ischemic neuronal cell death. So far, it has been widely accepted that cytotoxic intracellular calcium overload after ischemia is mainly mediated through stimulation of glutamate receptors, namely through N-methyl-D-aspartate (NMDA) receptors, metabotropic glutamate receptors and voltage-dependent Ca2+-channels. Roger Simon (Portland, Oregon) now added a new class of ion channel-coupled receptors to this list, that could be highly relevant for ischemia-induced toxic calcium influx into neurons. He showed that acid-sensing ion channels (ASIC) were expressed in neurons and mediated calcium influx and cell death after lowering of the extracellular pH. Since pH also rapidly drops in ischemic brain tissue, proton-activated ASIC could likely contribute to ischemia-induced calcium influx in neurons. Blocking ASIC by amilorid or Psalmotoxin-1 (from tarantula venom), or siRNA-mediated ASIC knockdown prevented proton-induced neuronal death exposing ASIC as a new target for therapeutic approaches in stroke. Indeed, evidence from animal models of stroke supported the important role of ASIC in ischemic brain damage, because ASIC1-knockout or pharmacological inhibition of ASIC significantly reduced the infarct volume and did so more potently than glutamate antagonism.

In the plenary lecture Pierluigi Nicotera (Leicester, UK) reviewed the current knowledge on molecular mechanisms of ischemic neuron death that may result from various signalling pathways, i.e. different subroutines of cell death ranging from classic apoptosis to primary or secondary necrosis. In particular, the excitotoxic increase in intracellular calcium levels has been established as an integrating trigger for ischemic cell death that can activate different cytotoxic cascades. Calcium overload can set off cell demise by activating proteases, lipases and DNases, change the balance of neuronal death from apoptosis to necrosis by depleting energy stores, or amplify other subroutines of the apoptotic death program. According to new data presented by Nicotera, Ca2+ may be also the crucial link between the different postischemic apoptotic and necrotic processes. He demonstrated that ion pumps that under physiological conditions could rapidly pump out Ca2+ to preserve a steady state cytosolic calcium were cleaved in apoptotic cells resulting in a fatal accumulation of Ca2+ in the cell. In neurons, isoforms of the plasma membrane Ca2+ ATPase (PMCA) have been identified to play an essential role in rectifying changes in intracellular Ca2+ in the long term. In addition to the cleavage of PMCA2 and PMCA4, Nicotera now presented new evidence for the cleavage of the Na+-Ca2+ exchanger (NCX) in apoptotic neurons. The latter Ca2+ ion pump effectively contributes to remove large amounts of calcium accumulated in the cytosol. During apoptosis, activated caspases eventually cleave these ion pumps which results in the disruption of calcium homeostasis that can finally switch apoptotic signalling to necrosis.

Caspases are proteases that become activated during apoptotic processes and then further promote key steps in the final execution of the apoptotic death program. Therefore, inhibition of caspases has been considered as an effective strategy to prevent apoptosis as well as secondary necrosis in ischemic neurons. Indeed, previous studies employing peptide caspase inhibitors or caspase knockout mice demonstrated a reduction of ischemic brain damage due to inactivation of caspases. However, caspase inhibition only partly prevented ischemic brain damage, and a considerable large number of neurons exposing apoptotic features were still detectable in the cortical penumbra region after transient focal ischemia in caspase-3 knockout mice. As reported by Nicotera, caspase inhibitors were also unable to prevent the degeneration of neuronal dendrites and axons which are essential for neuronal plasticity and, thus, for complex brain functions. Therefore, additional key factors may be involved in ischemic neuronal degeneration and death that need to be identified to develop effective new strategies for the treatment of stroke.

Apoptotic factors as novel target molecules in stroke therapy
Mitochondrial damage has been considered as the 'point of no return' in the cell death cascade triggered in neurons after an ischemic insult. Therefore, mechanisms upstream of mitochondrial dysfunction that are triggered early after an ischemic insult seem to be of particular interest for the development of neuroprotective stroke therapies. In particular, the regulation of Bcl-2 protein family members may be crucial for the maintenance of mitochondrial integrity and function thereby deciding a cell's fate after an apoptotic stress. While earlier reports established a proapoptotic role of Bax and (truncated) Bid in ischemic neuron death, Pak Chan (Stanford, USA) now presented studies that elucidated the activation of Bad in ischemic brain tissue. After ischemia, dephosphorylation of Bad appears to be the crucial step for it's activation. While under physiological conditions phosphorylated Bad remains bound and inactive in a complex with the cytosolic 14-3-3 protein, dephosphorylated Bad is released from this complex to interact with the anti-apoptotic Bcl-xL. Enhanced Bad phosphorylation by protein kinase A (PKA) or the PI3K/Akt pathway stabilizes the Bad/14-3-3 complex and prevents neuronal cell death. In contrast, dephosphorylation of Bad by protein phosphatases such as calcineurin (protein phosphatase 2B, PP2B) or protein phosphatase 2C (PP2C) may be a key step in the initiation of the apoptotic death program in ischemic neurons. Strikingly, new results from Josef Krieglstein, Juergen Schaefer (Marburg, Germany) and Susanne Klumpp (Muenster, Germany) proposed PP2C-induced dephosphorylation of Bad as an underlying apoptotic mechanism also in endothelial cells. In their study endothelial cell apoptosis was induced by fatty acids that are part of low density lipoproteins (LDL) and selectively enhance PP2C activity. Here, they provided evidence for enhanced PP2C activity and Bad dephosphorylation and, furthermore, demonstrated colocalization of PP2C and Bad in the apoptotic endothelial cells. Overall, these results point at a pivotal role for Bad dephosphorylation for ischemic brain damage as well as in (fatty acid-induced) apoptosis in endothelial cells which may be relevant to the pathology of atherosclerosis. Therefore, PP2C is proposed as a novel target molecule for therapeutic approaches aiming at neuroprotection in the acute phase after stroke. Further studies are required to clarify whether PP2C could also serve as a target for the development of effective therapeutic strategies to prevent atherosclerosis thereby reducing the risk of cardiovascular diseases and stroke.

After an ischemic insult, mitochondrial dysfunction is a major cause for ATP depletion and further disruption of the intracellular calcium homeostasis. In addition, damaged mitochondria release proapoptotic proteins such as cytochrome c, Smac/DIABLO or HtrA2/OMI which activate caspase-dependent apoptotic pathways. Other released mitochondrial proteins include apoptosis-inducing factor (AIF) and endonuclease G, both of which contribute to apoptotic nuclear DNA damage in a caspase-independent way. Klas Blomgren (Gothenborg, Sweden) and Nikolaus Plesnila (Munich, Germany) now provided evidence for a substantial role of AIF in ischemic brain damage. They demonstrated that AIF was released from mitochondria and translocated to the nucleus within the very first hours after an ischemic insult. This nuclear translocation of AIF colocalized with DNA damage and apoptotic nuclear condensation in the ischemic penumbra. Of note, mitochondrial release of AIF occurred several hours before cytochrom c release, and AIF-mediated cell death appeared to be caspase-independent, suggesting that AIF is in the first line of cell death signalling after ischemia. Indeed, the reduction of AIF protein levels in siRNA-treated cultured neurons or in harlequin (Hq) mutant mice resulted in a significant reduction of neuronal cell death in the respective experimental models of ischemia by approximately 50%. These results implicate AIF as a promising target for neuroprotective strategies in stroke therapy and will therefore stimulate further research on mechanisms of AIF-mediated ischemic cell damage after stroke. While it is still unclear how AIF exerts its apoptogenic function data presented by Valina Dawson (Baltimore, USA) elucidated the regulation of AIF release from mitochondria after an apoptotic insult. Using pharmacological inhibitors and knock out animals she demonstrated that activation of poly (ADP-ribose) polymerase-1 (PARP-1) was keys step in AIF-mediated caspase-independent apoptosis. Her data now also imply that the AIF fraction that is released from mitochondria and translocated to the nucleus during apoptosis is probably not located in the mitochondrial intermembrane space but rather bound to the outer mitochondrial membrane. This new finding may also explain the observation that AIF was released very rapidly and independently of other proapoptotic proteins that are also located in the mitochondrial intermembrane space, such as cytochrome c or Smac/DIABLO.

These and further novel insights into pathological mechanisms of ischemic cell death as presented on the Marburg symposium are the basis for promising therapeutic strategies directed against the multiple subroutines of cell death to significantly reduce brain damage after stroke.

Activation of endogenous protection mechanisms
Yet another emerging strategy towards neuroprotection is the activation of endogenous mechanisms of protection as, for example, the activation of intracellular survival signalling pathways by growth factors. Activation of such intracellular survival cascades often depend on the phosphorylation of key signalling proteins and require the fine-tuned regulation of protein kinases and phosphatases. It has been well established, that activation of such intracellular survival pathways can be achieved through stimulation of membrane-located receptors, and this activation depends on the extracellular concentration of the respective ligand. New results from the laboratories of Susanne Klumpp (Münster, Germany) and Josef Krieglstein (Marburg, Germany) now suggest that extracellular growth factors also require phosphorylation to activate intracellular survival pathways. In particular, they demonstrated that the neuroprotective activity of basic fibroblast growth factor (bFGF) depends on its phosphorylation in the extracellular space whereas dephosphorylated bFGF was inactive. This result implies the activation of protein kinases or inhibition of protein phosphatases by CNS penetrating small molecules as a new strategy to enhance growth factor-induced survival signalling.

A major obstacle for the use of growth factors as therapeutics in stroke treatment is their poor ability to cross the blood brain barrier. Therefore, induction of endogenous growth factor synthesis in the brain by small lipophilic molecules appears to be a useful strategy to overcome such application problems. As established by the group of Josef Krieglstein (Marburg, Germany) the lipophilic b2-adrenoceptor agonist clenbuterol mediated protection against ischemic brain damage through the induction of neuroprotective growth factors such as NGF, bFGF and TGF-b1. However, clenbuterol had to be applied hours before the insult to achieve neuroprotection. According to new data from combination therapy studies (Carsten Culmsee, Munich and Josef Krieglstein, Marburg, Germany) it is possible to extend the therapeutic window of clenbuterol to several hours after the onset of ischemia when combined with memantine, an NMDA receptor antagonist that is used in the treatment of Parkinson's disease and Alzheimer's disease. Moreover, the combination of both substances was always more efficient than treatment with either substance alone.

Frank Sharp (Cincinnati, USA) exposed heat shock protein 70 (HSP70) as potential factor of endogenous survival signalling. HSP70 blocks apoptotic factors such as AIF, apoptotic protease activating factor-1 (Apaf-1), and Jun-kinase (JNK). However, HSP70 also inhibits NF-kB activity which is important for neuronal survival. Therefore, a mutant HSP70 protein (HSP70c) was generated that retained the anti-apoptotic function but did not interfere with NF-kB signalling. The HSP70c mutant protein exposed pronounced neuroprotective effects and may therefore be useful for stroke therapy.

Erythropoietin (EPO) is another compound with a pronounced neuroprotective effect as established in experimental studies in vitro and in vivo, and, most importantly, in very promising initial clinical trials. According to new findings the cerebroprotective effect of EPO may not only result from the interference with pathological mechanisms of ischemic brain damage but, in addition, from enhanced endogenous repair mechanisms. According to Michael Chopp (Detroit, USA) EPO induced synthesis of growth factor, e.g. BDNF and VEGF in the brain tissue, thereby stimulating angiogenesis and improving the blood supply in the ischemic tissue. In addition, Chopp demonstrated that EPO stimulated neurogenesis which may also support the recovery of brain function after stroke. Most impressingly, such enhanced recovery of neurological functions was still observed when EPO was applied 24 h after stroke. Similar neuroprotective effects were achieved with a new carbamylated derivative of EPO (CEPO) as presented by Marcel Leist (Valby, Denmark). In various models of stroke and brain trauma CEPO exposed also an extensive therapeutic window of 24 h regarding the improvement of neurological functions. Notably, in contrast to EPO the carbamylated derivate CEPO did not affect hematopoiesis. It has been proposed that EPO-induced hematopoiesis enhances the risk of thrombosis, in particular because of the increase in the level of thrombocytes.

Further, regeneration of ischemic brain tissue was achieved by compounds of two other substance classes that are established in other therapeutic categories: Statins (cholesterol synthesis inhibitors), e.g. atorvastatin and of phosphodiesterase-5 (PDE-5) inhibitors, e.g. sildenafil enhanced angiogenesis and neurogenesis after stroke. Similar to erythropoietin, an improvement of neurological functions was detectable even when the compounds were applied 24 h after the onset of ischemia.

Improve and combine treatment strategies - towards an effective stroke therapy in the clinic
Although experimental studies generated a high number of compounds with a pronounced neuroprotective or regenerative potential, an effective drug therapy for the treatment of stroke patients is still not available. Many successful experimental strategies may have failed in stroke patients either because the animal models may have insufficiently reflected the human pathology after stroke or because the neuroprotective compounds were applied too late or at insufficient low doses in the clinical trials. Recombinant tissue plasminogen activator (rtPA) is the only approved therapeutic compound for the treatment of thromboembolic stroke. However, side effects such as haemorrhagic transformation and neurotoxic effects of rtPA may counteract the beneficial effect of thrombolysis in stroke patients, in particular if rtPA was applied later than 3 h after the insult. Because of these severe side effects the application of rtPA is rather restricted, and only about 4-5% of the stroke patients actually meet the criteria for rtPA therapy. Eng Lo (Charlestown, USA) presented new data on rtPA toxicity that was apparently mediated through activation of matrix metalloproteinases (MMP). Activation of MMP causes the destruction of the extracellular matrix and eventually endothelial cell apoptosis. The following disruption of the microvascular structure leads to hemorrhagic transformation after rtPA therapy. Lo suggested that combination therapy with rtPA and MMP inhibitors may be an appropriate strategy to avoid the fatal hemorrhagic transformation and, furthermore, could significantly expand the therapeutic window as well as the current tight treatment criteria. With such combination therapy a larger number of stroke patients could then profit from rtPA treatment.

Kyuya Kogure (Saitama, Japan) presented stimulating data from the application of the free radical scavenger edaravone (RadicutTM) in 3000 stroke patients. Edaravone has been shown to inhibit lipid peroxidation and vascular endothelial cell injury in vitro, and reduced brain edema, tissue injury, delayed neuronal death and neurological deficits in animal models of stroke. The results of various clinical studies demonstrated a significant improvement in neurological deficits without serious safety problems suggesting edaravone as a promising neuroprotective agent in the treatment of acute ischemic stroke. Moreover, edaravone may offer future advantages in combination therapy with fibrinolytic agents and antithrombotics by scavenging the free radicals associated with reperfusion injury.

Albumine is another neuroprotective compound under clinical investigation. Albumin represents 55-62% of plasmaproteins and albumin solutions are currently used as plasma expander in the treatment of edema. Studies by Myron Ginsberg and colleagues (Miami, USA) demonstrated pronounced neuroprotective effects in various stroke models. Disturbances of lung function were the only major side effect of albumin treatment observed in clinical trials. Meanwhile, such side effects were successfully reduced by using low dose albumin solutions in combination therapy with docosanoid derivates, and clinical phase II and III trials are in preparation. Docosanoids are derivatives of the docosahexaenoic acid, that have been identified as major products of the oxidative degradation of membrane lipids after cerebral ischemia. Nicolas Bazan (New Orleans, USA) reported on the pronounced neuroprotective effects of such docosanoids. In particular, he presented data on the new docosahexaenoic acid derivate 10,17S-docosatriene that prevented the infiltration of leucocytes and inhibited an increase in inflammatory markers such as COX-2 and NF-kB in the ischemic brain tissue, and reduced brain damage by approximately 50%.

Farewell to Marburg
Overall, new data from all areas of stroke research including cell culture and animal models of stroke as well as clinical studies were presented at the10th International Symposium on Pharmacology of Cerebral Ischemia in Marburg. In addition to neuroprotective strategies directed against key factors of cell death the stimulation of endogenous repair mechanisms, including angiogenesis and neurogenesis emerged as a promising attempt to improve neurological functions after stroke. Notably, a lot of drugs that are already in clinical use in other therapeutic categories were employed in these studies, for example statins, PDE-5-inhibitors, albumin, EPO, memantine and clenbuterol. With these established drugs and new compounds developed on the basis of the latest insights into the pathology of ischemic brain damage, and combinations thereof, an effective stroke therapy will likely be established in the near future.

The 10th symposium was at the same time the last one organized by Prof. Krieglstein in Marburg. All speakers and attendants acknowledged the achievements of Prof. Josef Krieglstein, who succeeded to organize an outstanding series of international stroke meetings over the last 18 years that were well recognized for their continuous high quality of the scientific program and the charming atmosphere in Marburg where renowned scientists and young researchers from all areas of stroke research came together as an international scientific family. The members of the international advisory board especially thanked Prof. Krieglstein by presenting a plate of honor reading: "The International Advisory Board recognizes Prof. Josef Krieglstein for his outstanding contributions to the neurosciences, for organizing the International Symposium on Pharmacology of Cerebral Ischemia during the last two decades and for academic leadership."

PD Dr. Carsten Culmsee1, Prof. Dr. Dr. Josef Krieglstein2
1Pharmazeutische Biologie-Biotechnologie, Department Pharmazie, Ludwig-Maximilians-Universität München;
2Institut für Pharmakologie und Toxikologie, Fachbereich Pharmazie, Philipps-Universität Marburg