Cerebral malaria (CM) is associated with a high mortality rate and long-term neurocognitive impairment in survivors. The murine model of experimental cerebral malaria (ECM) induced by Plasmodium berghei ANKA (PbA)-infection reproduces several of these features. We reported recently increased levels of IL-33 protein in brain undergoing ECM and the involvement of IL-33/ST2 pathway in ECM development. Here we show that PbA-infection induced early short term and spatial memory defects, prior to blood brain barrier (BBB) disruption, in wild-type mice, while ST2-deficient mice did not develop cognitive defects. PbA-induced neuroinflammation was reduced in ST2-deficient mice with low Ifng, Tnfa, Il1b, Il6, CXCL9, CXCL10 and Cd8a expression, associated with an absence of neurogenesis defects in hippocampus. PbA-infection triggered a dramatic increase of IL-33 expression by oligodendrocytes, through ST2 pathway. In vitro, IL-33/ST2 pathway induced microglia expression of IL-1β which in turn stimulated IL-33 expression by oligodendrocytes. These results highlight the IL-33/ST2 pathway ability to orchestrate microglia and oligodendrocytes responses at an early stage of PbA-infection, with an amplification loop between IL-1β and IL-33, responsible for an exacerbated neuroinflammation context and associated neurological and cognitive defects.

Glufosinate-ammonium (GLA), the active compound of a worldwide-used herbicide, acts by inhibiting the plant glutamine synthetase (GS) leading to a lethal accumulation of ammonia. GS plays a pivotal role in the mammalian brain where it allows neurotransmitter glutamate recycling within astroglia. Clinical studies report that an acute GLA ingestion induces convulsions and memory impairment in humans. Toxicological studies performed at doses used for herbicidal activity showed that GLA is probably harmless at short or medium range periods. However, effects of low doses of GLA on chronically exposed subjects are not known. In our study, C57BL/6J mice were treated during 10 weeks three times a week with 2.5, 5 and 10mg/kg of GLA. Effects of this chronic treatment were assessed at behavioral, structural and metabolic levels by using tests of spatial memory, locomotor activity and anxiety, hippocampal magnetic resonance imaging (MRI) texture analysis, and hippocampal GS activity assay, respectively. Chronic GLA treatments have effects neither on anxiety nor on locomotor activity of mice but at 5 and 10mg/kg induce (1) mild memory impairments, (2) a modification of hippocampal texture and (3) a significant increase in hippocampal GS activity. It is suggested that these modifications may be causally linked one to another. Since glutamate is the main neurotransmitter in hippocampus where it plays a crucial role in spatial memory, hippocampal MRI texture and spatial memory alterations might be the consequences of hippocampal glutamate homeostasis modification revealed by increased GS activity in hippocampus. The present study provides the first data that show cerebral alterations after chronic exposure to GLA.

Background and Purpose-The present study was designed to investigate whether neuronally derived nitric oxide (NO) plays a toxic role in the cascade of cellular events triggered by global cerebral ischemia in rats. Methods-7-Nitroindazole (7-NI) was used as a selective inhibitor of neuronal NO synthase. Global ischemia was induced for 20 minutes in anesthetized rats following the four-vessel occlusion model. Electroencephalogram and brain and body temperatures were continuously monitored. All rats were thermoregulated for the entire duration of anesthesia. 7-NI (25 mg/kg) or its vehicle was given intraperitoneally just after the carotid clamping and again 1 hour later. Rats were randomly divided into four groups: (1) vehicle (nϭ7); (2) 7-NI (nϭ7); (3) L-arginine (300 mg/kg IP)ϩ7-NI (nϭ7); and (4) 7-NI associated with warming to 37°C for 7 hours after disruption of anesthesia to compensate for the decrease in temperature induced by 7-NI (nϭ9). Seven days after ischemia, hippocampal CA1 damage was evaluated by classic histology. The lesion was scored with the use of a point scale, and the surviving neurons were counted. Results-Lesion scores were significantly lower and neuron counts higher in the two (warmed and unwarmed) groups of rats in which 7-NI was given alone than in vehicle-and L-arginineϩ7-NI-treated rats. Conclusions-The results indicate that 7-NI was neuroprotective in 20-minute global ischemia in rats and that the neuroprotective effect of 7-NI was mostly due to the blockade of NO synthesis, suggesting that NO released from neurons in ischemic conditions has a deleterious influence on hippocampal pyramidal neurons. (Stroke. 1998;29:1248-1254.)Key Words: cerebral ischemia Ⅲ hippocampus Ⅲ neuroprotection Ⅲ nitric oxide T ransient cerebral ischemia is associated with NO release, 1-5 but the question of whether NO is beneficial or toxic in this pathology remains unanswered. It is suggested that the effect of NO depends on the stage of evolution of the ischemic process and on the cellular source of NO. 6 The balance between the activation of two different (neuronal and endothelial) calcium-dependent isoforms of NOS in the acute stage of ischemia has been invoked to explain the contradictory results obtained with nonselective NOS inhibitors in rat focal ischemia 7,8 and in rat 9,10 and gerbil 11-13 global ischemia. On the basis of studies in knockout mice lacking specific NOS isoforms, it has been proposed that, in cerebral ischemia, the activation of type 3 (endothelial) NO synthase (eNOS) is beneficial, whereas the activation of type 1 (neuronal) NO synthase (nNOS) is detrimental.14 -16 NO produced by these enzymes is involved in the relaxation of cerebral blood vessels and the neurotoxicity of glutamate, respectively. However, no direct validation of this hypothesis has been demonstrated in rats submitted to severe global cerebral ischemia, which is the most widely used model. See Editorial Comment, page 1253The selective and delayed hippocampal damage induced in pyramidal neurons by transient global is...