Ischemia-induced depolarizations and associated hemodynamic responses in incomplete global forebrain ischemia in rats.

Spontaneous depolarizations around the core are a consistent feature of focal cerebral ischemia, but the associated regional hemodynamic changes are heterogeneous. We determined how the features of depolarizations relate to subsequent cerebral blood flow (CBF) changes in global forebrain ischemia. Forebrain ischemia was produced in halothane-anesthetized rats (n=13) by common carotid artery occlusion and hypovolemic hypotension. Mean arterial blood pressure (MABP) was monitored via a femoral catheter. Specific illuminations allowed the capture of image sequences through a cranial window to visualize: changes in membrane potential (voltage-sensitive dye method); CBF (laser speckle contrast imaging); cerebral blood volume (intrinsic optical signal, IOS at 540-550nm); and hemoglobin deoxygenation (IOS at 620-640nm). A depolarization occurred (n=9) when CBF fell below 43.4±5% of control (41±4mmHg MABP), and propagated with a distinct wave front at a rate of 2.8mm/min. Depolarizations were either persistent (n=4), intermediate (n=3) or short, transient depolarization (n=2). Persistent and intermediate depolarizations were associated with sustained hypoperfusion (-11.7±5.1%) and transient hypoperfusion (-17.4±5.2, relative to CBF before depolarization). Short, transient depolarizations did not generate clear CBF responses. Depolarizations during incomplete global ischemia occurred at the lower limit of CBF autoregulation, propagated similar to spreading depolarization (SD), and the hemodynamic responses indicated inverse neurovascular coupling. Similar to SDs associated with focal stroke, the propagating event can be persistent or transient.

A new model for the study of high-K(+)-induced preconditioning in cultured neurones: role of N-methyl-d-aspartate and alpha7-nicotinic acetylcholine receptors.

Spreading depression (SD), whether elicited by local application of high K(+) medium to the cortical surface or by other stimuli, can increase the brain's tolerance to a subsequent, severe ischaemic insult in vivo, a phenomenon termed preconditioning. Herein, we have developed and validated a robust in vitro protocol for high-K(+)-preconditioning of cultured neurones. This new model is especially appropriate to unravel the molecular mechanisms underlying neuronal preconditioning and subsequent ischaemic tolerance. With this new, optimised preparation, preconditioning was found to be dependent upon culture day in vitro, cell density, K(+) concentration and duration of treatment. Finally, preconditioning was shown to be dependent upon N-methyl-d-aspartate (NMDA), CAM-kinase II signalling and alpha7-nicotinic (alpha7 nACh) receptor function, which is analogous to in vivo preconditioning induced by various stimuli.

Effects of the nitric oxide donor, DEA/NO on cortical spreading depression.

Cortical spreading depression (CSD) is a transient disruption of local ionic homeostasis that may promote migraine attacks and the progression of stroke lesions. We reported previously that the local inhibition of nitric oxide (NO) synthesis with Nomega-nitro-L-arginine methyl ester (L-NAME) delayed markedly the initiation of the recovery of ionic homeostasis from CSD. Here we describe a novel method for selective, controlled generation of exogenous NO in a functioning brain region. It is based on microdialysis perfusion of the NO donor, 2-(N,N-diethylamino)-diazenolate-2-oxide (DEA/NO). As DEA/NO does not generate NO at alkaline pH, and as the brain has a strong acid-base buffering capacity, DEA/NO was perfused in a medium adjusted at alkaline (but unbuffered) pH. Without DEA/NO, such a microdialysis perfusion medium did not alter CSD. DEA/NO (1, 10 and 100 microM) had little effect on CSD by itself, but it reversed in a concentration-dependent manner the effects of NOS inhibition by 1 mM L-NAME. These data demonstrate that increased formation of endogenous NO associated with CSD is critical for subsequent, rapid recovery of cellular ionic homeostasis. In this case, the molecular targets for NO may be located either on brain cells to suppress mechanisms directly involved in CSD genesis, or on local blood vessels to couple flow to the increased energy demand associated with CSD.

Spreading depression-induced preconditioning in the mouse cortex: differential changes in the protein expression of ionotropic nicotinic acetylcholine and glutamate receptors.

Preconditioning of the cerebral cortex was induced in mice by repeated cortical spreading depression (CSD), and the major ionotropic glutamate (GluRs) and nicotinic acetylcholine receptor (nAChRs) subunits were compared by quantitative immunoblotting between sham- and preconditioned cortex, 24 h after treatment. A 30% reduction in alpha-amino-3-hydroxy-5-methyl-4-iso- xazolepropionate (AMPA) GluR1 and 2 subunit immunoreactivities was observed in the preconditioned cortex (p < 0.03), but there was no significant change in the NMDA receptor subunits, NR1, NR2A and NR2B. A 12-15-fold increase in alpha7 nAChR subunit expression following in vivo CSD (p < 0.001) was by far the most remarkable change associated with preconditioning. In contrast, the alpha4 nAChR subunit was not altered. These data point to the alpha7 nAChR as a potential new target for neuroprotection because preconditioning increases consistently the tolerance of the brain to acute insults such as ischaemia. These data complement recent studies implicating alpha7 nAChR overexpression in the amelioration of chronic neuropathologies, notably Alzheimer's disease (AD).

Asymmetric propagation of spreading depression along the anteroposterior axis of the cerebral cortex in mice.

The purpose of this study was to ascertain whether or not spreading depression (CSD) propagates symmetrically along the anteroposterior axis of the cortex of mice, and to determine where CSD should be elicited to achieve a uniform exposure of the cortex to this phenomenon. Experiments were performed in halothane-anesthetized mice, with three different locations aligned 1.5 mm from the midline used for either KCl elicitation of CSD or the recording of its propagation. Our results demonstrated that, at least in the mouse cortex, CSD propagated much more effectively from posterior to anterior regions than in the opposite direction. This feature was due to a different efficacy of propagation in the two opposite directions, and not to a reduced susceptibility of occipital regions to CSD elicitation. Heterogeneous CSD propagation constitutes a potential pitfall for neurochemical studies of post-CSD changes in mice, as brain tissue samples collected for this purpose should be uniformly exposed to CSD. Occipital sites for CSD induction are clearly optimal for this purpose. If CSD propagation is confirmed to be more effective from posterior to anterior regions in other species, this may be relevant to the pathophysiology of classical migraine because the most frequent aura symptoms (i.e., visual disturbances) originate in the occipital cortex.

Quinolinic acid accumulation during neuroinflammation. Does it imply excitotoxicity?

It is often proposed that quinolinic acid (QUIN) contributes to the pathophysiology of neuroinflammation because this kynurenine pathway metabolite is a selective agonist of N-methyl-D-aspartate (NMDA) receptors, and both its brain tissue and cerebrospinal fluid concentrations increase markedly with inflammation. However, whether or not the extracellular levels of QUIN reached during neuroinflammation are high enough to promote excitotoxicity, remains unclear because QUIN is a weak NMDA receptor agonist. We have addressed this issue by evaluating the extracellular concentrations of QUIN that must be reached to initiate potentially excitotoxic changes in the cerebral cortex of rats, under normal conditions, and when superimposed on another insult. We have also examined the increase in extracellular lactate associated with the perfusion of increasing concentrations of QUIN through a microdialysis probe. The extracellular EC50 for induction of local depolarisation was 228 microM with QUIN alone; that is, about 30 times the levels of QUIN measured previously in immune activated brain. Furthermore, at least 20 microM extracellular QUIN needed to be reached to reduce K+ induced spreading depression, an unexpected effect since spreading depression is inhibited by NMDA receptor antagonists. Our data suggest that, although synthesis of QUIN from activated microglia and invading macrophages can increase its extracellular concentration 10-100-fold, the levels that are reached in these conditions remain far below the concentrations of QUIN that are necessary for excessive NMDA receptor activation. However, the possibility that QUIN accumulation may be a deleterious feature of neuroinflammation cannot be ruled out at this stage.

Microdialysis coupled to online enzymatic assays.

In vivo sampling of interstitial fluid by using microdialysis fibers has become a standard and accepted procedure. This sampling method is generally coupled to offline analysis of consecutive dialysate samples by high-performance liquid chromatography or capillary electrophoresis, but this combination is not the best approach for some applications, especially those which require high temporal resolution and rapid data collection. The purpose of this review is to provide information on enzyme-based online assays, i.e., continuous analysis of the dialysate as it emerges from the outlet of the sampling device. We have focused on methods developed specifically for the analysis of solutions perfused at a very slow flow rate, i.e., a feature of microdialysis and ultrafiltration techniques. These methods include flow enzyme-fluorescence assays, flow enzyme-amperometric assays, and sequential enzyme-amperometric detection. Each type of assay is discussed in terms of principle, applications, advantages, and limitations. We also comment on implantable biosensors, an obvious next step forward for in vivo monitoring of molecules in neuroscience.

Kynurenine 3-hydroxylase inhibition in rats: effects on extracellular kynurenic acid concentration and N-methyl-D-aspartate-induced depolarisation in the striatum.

Inhibition of kynurenine 3-hydroxylase suppresses quinolinic acid synthesis and, therefore, shunts all kynurenine metabolism toward kynurenic acid (KYNA) formation. This may be a pertinent antiexcitotoxic strategy because quinolinic acid is an agonist of NMDA receptors, whereas kynurenic acid antagonises all ionotropic glutamate receptors with preferential affinity for the NMDA receptor glycine site. We have examined whether the kynurenine 3-hydroxylase inhibitor Ro 61-8048 increases extracellular (KYNA) sufficiently to control excessive NMDA receptor function. Microdialysis probes incorporating an electrode were implanted into the striatum of anaesthetised rats, repeated NMDA stimuli were applied through the probe, and the resulting depolarisation was recorded. Changes in extracellular KYNA were assessed by HPLC analysis of consecutive dialysate samples. Ro 61-8048 (42 or 100 mg/kg) markedly increased the dialysate levels of KYNA. The maximum increase (from 3.0 +/- 1.0 to 31.0 +/- 6.0 nM; means +/- SEM, n = 6) was observed 4 h after administration of 100 mg/kg Ro 61-8048, but the magnitude of the NMDA-induced depolarisations was not reduced. A separate study suggested that extracellular KYNA would need to be increased further by two orders of magnitude to become effective in this preparation. These results challenge the notion that kynurenine 3-hydroxylase inhibition may be neuroprotective, primarily through accumulation of KYNA and subsequent attenuation of NMDA receptor function.

In vivo assessment of kynurenate neuroprotective potency and quinolinate excitotoxicity.

Three complementary questions related to the kynurenine pathway and excitotoxicity were addressed in this study: (i) Which extracellular levels of quinolinic acid (QUIN) may be neurotoxic? (ii) Which extracellular levels of kynurenic acid (KYNA) may control excessive NMDA-receptor function? (iii) Can "anti-excitotoxic" levels of KYNA be reached by inhibition of kynurenine-3-hydroxylase (i.e. inhibition of QUIN synthesis and shunts of kynurenine metabolism toward KYNA)? Multifunctional microdialysis probes were used in halothane anaesthetised rats to apply NMDA or QUIN directly to the brain, with or without co-perfusion of KYNA, to record the resulting local depolarisations, and to monitor changes in dialysate KYNA after kynurenine-3-hydroxylase inhibition. QUIN produced concentration-dependent depolarisations with an estimated EC50 (i.e. concentration in the perfusion medium) of 1.22mM. The estimated ED50 for KYNA inhibition of NMDA-responses was 181microM. Kynurenine-3-hydroxylase inhibition (Ro-61-8048, 100mg/kg i.p.) increased dialysate KYNA 11 times (to 33.8nM) but without any reduction of NMDA-responses. These data challenge the notion that extracellular accumulation of endogenous QUIN may contribute to excessive NMDA-receptor activation in some neurological disorders, and the suitability of kynurenine-3-hydroxylase inhibition as an effective anti-excitotoxic strategy.

Pharmacological investigation into the involvement of nitric oxide in K+-induced cortical spreading depression.

Cortical spreading depression (CSD) is a transient, local disruption of cellular ionic homeostasis that propagates slowly across the cerebral cortex. As previous data have suggested a possible link between nitric oxide (NO) formation and CSD, we have examined whether CSD is suppressed by local inhibition of NO synthesis with 7-nitroindazole (7-NINA), a compound which may have a greater selectivity for the neuronal NO synthase isoform. Multifunctional microdialysis probes were implanted in the cortex of halothane-anaesthetised rats, and used for (1) elicitation of repetitive CSD by perfusion of 160 mM K+ through the probe, (2) recording of CSD as a negative shift of the extracellular direct current (DC) potential, and (3) perfusion of 7-NINA before and during CSD elicitation. Elicitation of CSD was moderately inhibited by 1 mM 7-NINA in the perfusion medium, as shown in one treated group (n=8) by a significant reduction of both number (from 5.1+/-0.4 to 3.6+/-0.4; P<0.05) and cumulative DC negativity (from 16.4+/-0.7 mV x min to 13.3+/-0.9 mV x min; P<0.01). However, effective concentrations of 7-NINA were at least 100-fold higher than its Ki for the target enzyme in vitro, the moderate inhibition of CSD by 7-NINA was not reversed by the NO precursor, L-arginine, and the amplitude of the K+-induced sustained DC potential negative shift was also reduced significantly by 7-NINA (from 27.9+/-0.9 mV to 23.9+/-1.2 mV; P<0.05). These data do not support the hypothesis that NO formation contributes to the elicitation of CSD by high extracellular K+. The finding that 7-NINA reduced the intensity of K+-induced depolarisation may be relevant to previous investigations that used this drug to examine the role of NO in the modulation of K+-induced neurotransmitter release.

Neuroprotective potency of kynurenic acid against excitotoxicity.

The aim of this study was to determine in vivo which extracellular levels of kynurenic acid (KYNA) are required to control excessive NMDA receptor activation in the rat cortex. As excitotoxicity is coupled to marked ion movements, local depolarisations induced by perfusion of NMDA or quinolinic acid (QUIN) through microdialysis probes were recorded at the site of excitotoxin application. Perfusion of KYNA through the dialysis fibre inhibited the excitotoxin responses with an IC50 of 32-66 microM (extracellular concentration corrected for microdialysis delivery), but > 10-fold lower levels of endogenous KYNA were reported to be neuroprotective. Accordingly, these results strengthen the notion that KYNA accumulation may protect the brain parenchyma by acting on different molecular target(s), besides the NMDA receptor glycine site.

Excitotoxicity in neurological disorders--the glutamate paradox.

Beneficial effects of glutamate-receptor antagonists in models of neurological disorders are often used to support the notion that endogenous excitotoxicity (i.e. resulting from extracellular accumulation of endogenous glutamate) is a major contributor to neuronal death associated with these conditions. However, this interpretation conflicts with a number of robust and important experimental evidence. Here, emphasis is placed on two key elements: (i) very high extracellular levels of glutamate must be reached to initiate neuronal death, far above those measured in models of neurological disorders; and (ii) changes in extracellular glutamate as measured by microdialysis are not related to changes in the synaptic cleft, i.e. the compartment where neurotransmitter glutamate interacts with its receptors. It has become clear that the diversity and complexity of glutamate-mediated processes allow for a wide range of potential abnormalities (e.g. loss of selectivity of glutamate-operated ion channels, abnormal modulation of glutamate receptors). In addition, as neuronal death subsequent to ischemia and other insults is likely to result from multifactorial processes that may be inter-related, inhibition of glutamate-mediated synaptic transmission may be neuroprotective by increasing the resistance of neurons to other deleterious mechanisms (e.g. inadequate energy supply) that are not directly related to glutamatergic transmission.

The relationship between the apparent diffusion coefficient measured by magnetic resonance imaging, anoxic depolarization, and glutamate efflux during experimental cerebral ischemia.

A reduction in the apparent diffusion coefficient (ADC) of water measured by magnetic resonance imaging (MRI) has been shown to occur early after cerebrovascular occlusion. This change may be a useful indicator of brain tissue adversely affected by inadequate blood supply. The objective of this study was to test the hypothesis that loss of membrane ion homeostasis and depolarization can occur simultaneously with the drop in ADC. Also investigated was whether elevation of extracellular glutamate ([GLU]e) would occur before ADC changes. High-speed MRI of the trace of the diffusion tensor (15-second time resolution) was combined with simultaneous recording of the extracellular direct current (DC) potential and on-line [GLU]e from the striatum of the anesthetized rat. After a control period, data were acquired during remote middle cerebral artery occlusion for 60 minutes, followed by 30 minutes of reperfusion, and cardiac arrest-induced global ischemia. After either focal or global ischemia, the ADC was reduced by 10 to 25% before anoxic depolarization occurred. After either insult, the time for half the maximum change in ADC was significantly shorter than the corresponding DC potential parameter (P < 0.05). The [GLU]e remained at low levels during the entire period of varying ADC and DC potential and did not peak until much later after either ischemic insult. This study demonstrates that ADC changes can occur before membrane depolarization and that high [GLU]e has no involvement in the early rapid ADC decrease.

Long-term potentiation in the dentate gyrus is not linked to increased extracellular glutamate concentration.

Long-term potentiation (LTP) of excitatory transmission is a likely candidate for the encoding and storage of information in the mammalian brain. There is a general agreement that LTP involves an increase in synaptic strength, but the mechanisms underlying this persistent change are unclear and controversial. Synaptic efficacy may be enhanced because more transmitter glutamate is released or because postsynaptic responsiveness increases or both. The purpose of this study was to examine whether increased extracellular glutamate concentration was associated with the robust and well-characterized LTP that can be induced in the rat dentate gyrus. To favor the detection of any putative change in extracellular glutamate associated with LTP, our experimental strategy included the following features. 1) Two separate series of experiments were carried out with animals under pentobarbital or urethan anesthesia; 2) changes in extracellular concentration of glutamate were monitored continuously by microdialysis coupled to enzyme amperometry; and 3) dialysate glutamate levels and changes in the slope of excitatory postsynaptic potential evoked by activation of the perforant path were recorded precisely at the same site. Tetanic stimulation of the perforant path increased persistently test-evoked responses in the dentate gyrus (by 19 and 14% in barbiturate and urethan group, respectively), but there was no glutamate change either during or after LTP induction and no indication of increased glutamate efflux when low-frequency stimulation was applied. The results do not rule out a possible contribution of enhanced glutamate exocytosis to LTP induction and/or maintenance because such a presynaptic change may not be detectable extracellularly. However, our findings and other data supporting the notion that neurotransmitter glutamate may hardly leak out of the synaptic cleft conflict with the hypothesis that LTP could also involve a broad synaptic spillover of glutamate.

High extracellular glutamate and neuronal death in neurological disorders. Cause, contribution or consequence?

In models of neurological disorders, increased extracellular glutamate and beneficial effects produced by glutamate-receptor antagonists are consistently taken as supporting evidence of excitotoxicity. This systematic interpretation is over-simplified and potentially misleading. High extracellular glutamate is not a reliable indicator of endogenous excitotoxicity, i.e., the intrinsic, potential neurotoxicity of endogenous glutamate whenever it accumulates extracellularly. Firstly, because the extracellular levels of glutamate necessary to produce depolarization and death in vivo, are far above those measured in models of neurological disorders. Secondly, because changes in the concentration of glutamate in the synaptic cleft (i.e., the relevant compartment for endogenous excitotoxicity) are not reflected extracellularly. Protection by glutamate-receptor antagonists does not necessarily imply inhibition of excitotoxic abnormalities. Indeed, neuronal death initiated by insults such as ischemia results from multifactorial processes that may be interrelated. Therefore, beneficial effects resulting from an interaction with glutamate-mediated transmission may actually render the cell more resistant to other deleterious mechanisms (e.g., mitochondrial injury, oxidative stress).

Neuroprotective strategies: voltage-gated Na+-channel down-modulation versus presynaptic glutamate release inhibition.

Insufficient ATP production relative to cellular requirements is the key factor detrimental to neurons in neurological disorders associated with deficient oxygen/glucose supply or mitochondrial dysfunction. As a large part of the energy consumed by brain cells is used to maintain the Na+ gradient across the cellular membrane, reduction of energy demand by down-modulation of voltage-gated Na+-channels is a rational strategy for neuroprotection against these conditions. Preservation of the inward Na+ gradient is likely to be also beneficial as it is an essential driving force for vital ion exchanges and transport mechanisms (e.g. Ca2+-homeostasis and cell volume regulation). From these elements, I propose that use-dependent Na+-channel blockers increase the resilience of nerve cells to the primary insult and/or subsequent deleterious events, and that reduced efflux of glutamate and other compounds is only a consequence of cellular stress attenuation. The widespread hypothesis that down-modulation of Na+-channels is neuroprotective primarily through reduction of presynaptic glutamate release conflicts with strong experimental evidence.

Hypoosmolarity induces an increase of extracellular N-acetylaspartate concentration in the rat striatum.

We previously showed that extracellular levels of N-acetylaspartate (NAA) increase when a medium with reduced NaCl concentration is perfused through a microdialysis probe, and proposed that NAA may be released during hypoosmotic swelling. Here, we demonstrate that this effect is due to hypoosmolarity of the perfusion medium, and not to low NaCl. NAA changes in the dialysate were compared with those of taurine as the osmoregulatory role of this amino acid is established. Reduction of the NaCl concentration in the perfusion medium increased the dialysate levels of NAA and taurine, but this effect was abolished when NaCl was replaced by sucrose to maintain isosmolarity. The NAA response to hypoosmolarity was smaller than that of taurine, but it may still be important to neurons as NAA is predominantly neuronal in the mammalian CNS.

Effects of pharmacological inhibition of glutamate-uptake on ischaemia-induced glutamate efflux and anoxic depolarization latency.

It has been proposed that deficient glutamate uptake, by increasing the extracellular concentration of this excitatory neurotransmitter, may contribute to the pathophysiology of cerebral ischaemia. This study aimed to examine whether pharmacological inhibition of glutamate uptake altered the kinetics of ischaemia-induced glutamate efflux, and precipitated anoxic depolarisation. Microdialysis was used for application of the glutamate-uptake inhibitor L-trans-pyrrolidine-2,4-dicarboxylate (L-trans-PDC), recording of the EEG and extracellular direct current (DC) potential with an electrode within the probe, and continuous monitoring of changes in extracellular glutamate. L-trans-PDC was applied locally from 8 min prior to cardiac arrest to the end of the recording period. L-trans-PDC (2.5 mM) barely altered the time course of postmortem glutamate efflux in the cortex. Only the maximum rate of efflux during the first exocytotic phase, and the concentration reached at the end of this phase, appeared slightly increased. L-trans-PDC (5 mM) reduced significantly the delay between EEG silence and anoxic depolarization in the cerebral cortex (59.2 +/- 9.2 s vs. 79.7 +/- 11.5 s; n = 6), but not in the striatum and hippocampus. These effects contrast with the marked increase in dialysate glutamate that L-trans-PDC produces in all these three brain regions. Together, these data do not support the hypothesis that inhibition of glutamate uptake plays a critical role, early in cerebral ischaemia. However, a contribution of reversed glutamate uptake to the secondary Ca2+-independent phase of ischaemia-induced glutamate efflux cannot be ruled out.