To evaluate the contribution of cellular dysfunction and neuronal loss to brain N-acetylaspartate (NAA) depletion, NAA was measured in brain tissue by HPLC and UV detection in rats subjected to cerebral injury, associated or not with cell death. When lesion was induced by intracarotid injection of microspheres, the fall in NAA was related to the degree of embolization and to the severity of brain oedema. When striatal lesion was induced by local injection of malonate, the larger the lesion volume, the higher the NAA depletion. However, reduction of brain oedema and striatal lesion by treatment with the lipophilic iron chelator dipyridyl (20 mg/kg, 1 h before and every 8 h after embolization) and the inducible nitric oxide synthase inhibitor aminoguanidine (100 mg/kg given 1 h before malonate and then every 9 h), respectively, failed to ameliorate the fall in NAA. Moreover, after systemic administration of 3-nitropropionic acid, a marked reversible fall in NAA striatal content was observed despite the lack of tissue necrosis. Overall results show that cellular dysfunction can cause higher reductions in NAA level than neuronal loss, thus making of NAA quanti®cation a potential tool for visualizing the penumbra area in stroke patients.

The pathogenesis of ischemic stroke is a complex sequence of events including inflammatory reaction, for which the microglia appears to be a major cellular contributor. However, whether post-ischemic activation of microglial cells has beneficial or detrimental effects remains to be elucidated, in particular on long term brain plasticity events. The objective of our study was to determine, through modulation of post-stroke inflammatory response, to what extent microglial cells are involved in some specific events of neuronal plasticity, neurite outgrowth and synaptogenesis. Since microglia is a source of neurotrophic factors, the identification of the brain-derived neurophic factor (BDNF) as possible molecular actor involved in these events was also attempted. As a means of down-regulating the microglial response induced by ischemia, 3-aminobenzamide (3-AB, 90 mg/kg, i.p.) was used to inhibit the poly(ADP-ribose) polymerase-1 (PARP-1). Indeed, PARP-1 contributes to the activation of the transcription factor NF-kB, which is essential to the upregulation of proinflammatory genes, in particular responsible for microglial activation/proliferation. Experiments were conducted in rats subjected to photothrombotic ischemia which leads to a strong and early microglial cells activation/proliferation followed by an infiltration of macrophages within the cortical lesion, events evaluated at serial time points up to 1 month post-ictus by immunostaining for OX-42 and ED-1. Our most striking finding was that the decrease in acute microglial activation induced by 3-AB was associated with a long term down-regulation of two neuronal plasticity proteins expression, synaptophysin (marker of synaptogenesis) and GAP-43 (marker of neuritogenesis) as well as to a significant decrease in tissue BDNF production. Thus, our data argue in favour of a supportive role for microglia in brain neuroplasticity stimulation possibly through BDNF production, suggesting that a targeted protection of microglial cells could represent an innovative approach to potentiate post-stroke neuroregeneration.

Although brain-derived neurotrophic factor (BDNF) plays a central role in recovery after cerebral ischemia, little is known about cells involved in BDNF production after stroke.The present study testes the hypothesis that neurons are not the unique source of neosynthesized BDNF after stroke and that non neuronal-BDNF producing cells differ according to the delay after stroke induction. For this purpose, cellular localization of BDNF and BDNF content of each hemisphere were analysed in parallel before and after (4h, 24h and 8d) ischemic stroke in rats. Stroke of different severities was induced by embolization of the brain with variable number of calibrated microspheres allowing us to explore the association between BDNF production and neuronal death severity. The main results are that a) unilateral stroke increased BDNF production in both hemispheres with a more intense and long-lasting effect in the lesioned hemisphere, b) BDNF levels either of the lesioned or unlesioned hemispheres were not inversely correlated to neuronal death severity whatever the delay after stroke onset, c) in the unlesioned hemisphere, stroke resulted in increased BDNF staining in neurons and ependymal cells (at 4h and 24h), d) in the lesioned hemisphere, beside neurons and ependymal cells, microglial cells (at 24h), endothelial cells of cerebral arterioles (at 4h and 24h) and astrocytes (at 8d) exhibited a robust BDNF staining as well. Taken together, overall data suggest that non neuronal cells are able to produce substantial amount of BDNF after ischemic stroke and that more attention should be given to these cells in the design of strategies aimed at improving stroke recovery through BDNFrelated mechanisms.

Cardiovascular BDNF is mainly localized within endothelial cells in which its expression is dependent on endothelial function. These results open new perspectives on the role of endothelial BDNF in cardiovascular health.

Physical exercise constitutes an innovative strategy to treat deficits associated with stroke through the promotion of BDNF-dependent neuroplasticity. However, there is no consensus on the optimal intensity/duration of exercise. In addition, whether previous stroke changes the effect of exercise on the brain is not known. Therefore, the present study compared the effects of a clinically-relevant form of exercise on cerebral BDNF levels and localization in control versus stroke rats. For this purpose, treadmill exercise (0.3 m/s, 30 min/day, for 7 consecutive days) was started in rats with a cortical ischemic stroke after complete maturation of the lesion or in control rats. Sedentary rats were run in parallel. Mature and proBDNF levels were measured on the day following the last boot of exercise using Western blotting analysis. Total BDNF levels were simultaneously measured using ELISA tests. As compared to the striatum and the hippocampus, the cortex was the most responsive region to exercise. In this region, exercise resulted in a comparable increase in the production of mature BDNF in intact and stroke rats but increased proBDNF levels only in intact rats. Importantly, levels of mature BDNF and synaptophysin were strongly correlated. These changes in BDNF metabolism coincided with the appearance of intense BDNF labeling in the endothelium of cortical vessels. Notably, ELISA tests failed to detect changes in BDNF forms. Our results suggest that control beings can be used to find conditions of exercise that will result in increased mBDNF levels in stroke beings. They also suggest cerebral endothelium as a potential source of BDNF after exercise and highlight the importance to specifically measure the mature form of BDNF to assess BDNF-dependent plasticity in relation with exercise.

We examined the efficacy of the liposoluble iron chelator 2,2Ј-dipyridyl (DP) in reducing histological damage in rats submitted to cerebral ischemia and the mechanisms involved in the potential cytoprotection. For this purpose, DP (20 mg/kg, i.p.) was administered 15 min before and 1 h after induction of cortical photothrombotic vascular occlusion in rat. Histological studies were performed to assess infarct volume (at days 1 and 3 postischemia) and astromicroglial activation (at day 3 postischemia). Damage to endothelial and neuronal cells was evaluated at day 1 postischemia by quantitative measurements of Evans Blue extravasation and N-acetylaspartate levels, respectively. Cerebral blood flow was recorded in the ischemic core by laser-Doppler flowmetry within the 15 min to 2 h period after photothrombosis. At 4-h postischemia, radical oxygen species (ROS) production was evaluated by measuring brain glutathione concentrations. The cortical expression of the proteins heme oxygenase-1 (HO-1) and hypoxia-inducible factor-1␣ (HIF-1␣) was analyzed by Western blotting at day 1 postischemia. Infarct volume and ischemic damage to endothelial and neuronal cells were significantly reduced by DP treatment. This cytoprotection was associated with a reduction in ROS production, perfusion deficits, and astrocytic activation. DP treatment also resulted in significant changes in HO-1 (ϩ100%) and HIF-1␣ (Ϫ50%) protein expression at the level of the ischemic core. These results report the efficacy of the liposoluble iron chelator DP in reducing histological damage induced by permanent focal ischemia.

Because arginase and nitric oxide (NO) synthases (NOS) compete to degrade l-arginine, arginase plays a crucial role in the modulation of NO production. Moreover, the arginase 1 isoform is a marker of M2 phenotype macrophages that play a key role in tissue remodeling and resolution of inflammation. While NO has been extensively investigated in ischemic stroke, the effect of stroke on the arginase pathway is unknown. The present study focuses on arginase expression/activity and localization before and after (1, 8, 15 and 30 days) the photothrombotic ischemic stroke model. This model results in a cortical lesion that reaches maximal volume at day 1 post-stroke and then decreases as a result of astrocytic scar formation. Before stroke, arginase 1 and 2 expressions were restricted to neurons. Stroke resulted in up-regulation of arginase 1 and increased arginase activity in the region centered on the lesion where inflammatory cells are present. These changes were associated with an early and long-lasting arginase 1 up-regulation in activated macrophages and astrocytes and a delayed arginase 1 down-regulation in neurons at the vicinity of the lesion. A linear positive correlation was observed between expressions of arginase 1 and glial fibrillary acidic protein as a marker of activated astrocytes. Moreover, the pattern of arginase 1 and brain-derived neurotrophic factor (BDNF) expressions in activated astrocytes was similar. Unlike arginase 1, arginase 2 expression was not changed by stroke. In conclusion, increased arginase 1 expression is not restricted to macrophages in inflammation elicited by stroke but also occurs in activated astrocytes where it may contribute to neuroplasticity through the control of BDNF production.

BackgroundWhereas brain-derived neurotrophic factor (BDNF) levels are measured in the brain in animal models of stroke, neurotrophin levels in stroke patients are measured in plasma or serum samples. The present study was designed to investigate the meaning of circulating BDNF levels in stroke patients.Methods and ResultsUnilateral ischemic stroke was induced in rats by the injection of various numbers of microspheres into the carotid circulation in order to mimic the different degrees of stroke severity observed in stroke patients. Blood was serially collected from the jugular vein before and after (4 h, 24 h and 8 d) embolization and the whole brains were collected at 4, 24 h and 8 d post-embolization. Rats were then selected from their degree of embolization, so that the distribution of stroke severity in the rats at the different time points was large but similar. Using ELISA tests, BDNF levels were measured in plasma, serum and brain of selected rats. Whereas plasma and serum BDNF levels were not changed by stroke, stroke induced an increase in brain BDNF levels at 4 h and 24 h post-embolization, which was not correlated with stroke severity. Individual plasma BDNF levels did not correlate with brain levels at any time point after stroke but a positive correlation (r = 0.67) was observed between individual plasma BDNF levels and stroke severity at 4 h post-embolization.ConclusionCirculating BDNF levels do not mirror brain BDNF levels after stroke, and severe stroke is associated with high plasma BDNF in the very acute stage.