Volatile and Intravenous Anesthetics Decrease Glutamate Release from Cortical Brain Slices during Anoxia

Methods: Glutamate released from cortical brain slices was measured fluorometrically with the glutamate dehydrogenase catalyzed formation of the reduced form of nicotinamide adenine dinucleotide phosphate from nicotinamide adenine dinucleotide phosphate. Glutamate release was measured in oxygenated (P^O2= 400 mmHg), hypoxic ((P^O2= 20 mmHg), and anoxic ((P^O2= 20 mmHg plus 100 micro Meter NaCN) solutions and with clinical concentrations of anesthetics (halothane 325 micro Meter, enflurane 680 micro Meter, propofol 200 micro Meter, sodium thiopental 50 micro Meter). The source of glutamate released during these stresses was defined with toxins inhibiting N and P type voltage-gated calcium channels, and with calcium-free medium.

Results: Glutamate released during hypoxia or anoxia was 1.5 and 5.3 times greater, respectively, than that evoked by depolarization with 30 mM KCl. Hypoxia/anoxia-induced glutamate release was not mediated by synaptic voltage-gated calcium channels, but probably by the reversal of normal uptake mechanisms. Halothane, enflurane, and sodium thiopental, but not propofol, decreased hypoxia-evoked glutamate release by 50–70%(P < 0.05). None of the anesthetics alter basal glutamate release.

Conclusions: The authors conclude that halothane, enflurane, and sodium thiopental but not propofol, at clinical concentrations, decrease extrasynaptic release of L-glutamate during hypoxic stress. (Key words: Anesthetics, intravenous: propofol; sodium thiopental. Anesthetics, volatile: enflurane; halothane. Animals: rat. Brain: excitatory neurotransmitters; ischemia. Experimental techniques: brain slices; neurotransmitter release.)

L-GLUTAMATE is the main excitatory neurotransmitter in the vertebrate central nervous system. During cerebral anoxia or ischemia, glutamate accumulates in the brain at levels that hyperexcite glutamate receptors and cause uncontrolled calcium influx through the N-methyl-D-aspartate subtype glutamate receptor. Calcium-mediated damage and cell death ensue. The key role of glutamate accumulation in the pathophysiology of cerebral anoxia/ischemia is well established (for review see references 1 and 2). The reversal or failure of glutamate reuptake transporters, rather than synaptic glutamate release, is the chief source of the glutamate accumulation in the extracellular space during ischemia. The pool of glutamate potentially released by anoxia may be as great as 5–20 mM. The glutamate reuptake systems fail during ischemia because of the collapse of the electrical and ionic gradients required by the glutamate transporters. Because synaptic glutamate release requires adenosine triphosphate energy, synaptic release is inhibited during energy-limited hypoxic or ischemic conditions ().

Based on much in vitro evidence, it is reasonable to assume that prevention of excessive glutamate accumulation may be protective during cerebral ischemia. Indeed, mild hypothermia, probably the most potent neuroprotectant known, greatly reduces infarct volume and ischemic tissue glutamate concentrations. Several reports suggest that anesthetics reduce infarct volume from focal cerebral ischemia (for barbiturates see ; for volatile agents, ), and it could be hypothesized that this is due to decreased glutamate accumulation in the periphery of infarcts. Propofol and volatile anesthetics do not inhibit glutamate accumulation in the infarct core, but it is not known if these agents reduce glutamate release under merely hypoxic, but not ischemic, conditions.

In this study an we developed an in vitro assay for measuring glutamate release from brain slices, and tested whether several different types of volatile anesthetic compounds (enflurane and halothane) or intravenous anesthetics (sodium thiopental and propofol) reduce glutamate release during hypoxic or anoxic insults.

The following methods were approved by the University of California San Francisco Committee on Animal Research and conform to relevant National Institutes of Health guidelines.

Cortical brain slices were prepared from 8–20-day old Sprague-Dawley rats anesthetized with 1.5–2% halothane in oxygen. After decapitation, brain hemispheres were rapidly dissected, glued with cyanoacrylate to a holder, and immersed in 1–3 degrees Celsius artificial brain extracellular fluid (ABECF; Earle's balanced salts, composition in mM: NaCl 116, NaHCO sub 3 25, KCl 5.4, CaCl²1.8, MgCl²0.9, NaH²PO⁴0.9, glucose 10, pH 7.40 bubbled with 5% CO²-95% Oxygen²). Slices (300–350-micro meter thick) were then prepared with a vibrating tissue slicer. Slices of the cerebral cortex were transferred to vials of gassed ABECF. To permit maximal slice recovery from slicing trauma, slices were not studied until 1–2 h had elapsed from slicing. During this period, slices were maintained at room temperature (approximately 25 degrees Celsius) in oxygenated ABECF.

Glutamate released from brain slices under study was detected with a fluorescence assay in a Hitachi (Tokyo, Japan) F-2000 fluorometer. Each slice was gently fixed to a mesh holder and placed in a fluorometer cuvette containing 1.6 ml of 37 degrees Celsius ABECF, 1 mM nicotinamide adenine dinucleotide phosphate, and 5 IU/ml glutamate dehydrogenase (Figure 1). The formation of the reduced form of nicotinamide adenine dinucleotide phosphate from nicotinamide adenine dinucleotide phosphate by glutamate dehydrogenase was measured fluorometrically (excitation light 340 nm, emission intensity 460 nm) in the solution above the slice. A stir bar ensured rapid detection of released neurotransmitter. The temperature of the cuvette fluid was maintained at 37 degrees Celsius throughout the study. The assay was calibrated by injecting known quantities of L-glutamate into the cuvette. The assay permitted detection of about 1 nmol (approximately 0.01 nmol/sec) of glutamate. Studies were performed to ensure that nicotinamide adenine dinucleotide phosphate (or other fluorochromes) and glutamate dehydrogenase were not released from slices during hypoxia. This was done by excluding the compound in question and measuring whether KCl- evoked glutamate release produced any change in fluorescence signal. The anesthetics, solvents (dimethyl sulfoxide), and reagents were tested to determine if they interfered with the glutamate assay. This involved measuring the rate and maximum extent of fluorescence change produced by addition of 10–100 nmol glutamate to the assay solution.

The glutamate release assay measures the net translocation or efflux of glutamate from the slice into the medium, and therefore reflects the balance between cellular release of glutamate and cellular uptake among the cells in the slice.

Glutamate efflux from brain slices was measured during three experimental stresses:(1) KCl-evoked depolarization. Synaptic glutamate release was induced with KCl (final cuvette concentration 30 mM);(2) Hypoxia. Slices were placed in ABECF equilibrated with 95% Nitrogen²-5% CO². For these slices, 95% Nitrogen²-5% CO²was used to fill the head space of the cuvette and the fluorometer analysis chamber. The final PO²of this solution was between 20 and 35 mmHg (Corning Blood Gas Analyzer, Pleasanton, CA), as small amounts of oxygen unavoidably entered the cuvette from the previously oxygenated slice;(3) Anoxia. To produce conditions essentially equivalent to anoxia, 100 micro Meter NaCN was added to the hypoxic solutions. Slices from a given animal were randomly assigned to different treatment groups. Slices were used only in one treatment then discarded. For all stresses, the appearance of glutamate in the cuvette was followed for a 5–10-min period, until a stable rate of formation was seen or a plateau was reached. Slices were then transferred to 1.0 ml tris(hydroxymethyl)aminomethane buffer at 100 degrees Celsius, sonicated, and stored at -80 degrees Celsius for later assay of total protein (Enhanced Protein Assay, Pierce, Rockford, IL).

To determine the cause of glutamate released during the different experimental stresses, we used the N and P type voltage-gated calcium channel toxins from Conus geographicus (omega-conotoxin GVIA, Research Biochemicals Inc., Natick, MA) and omega-agatoxin IVa from Agelenopsis aperta (gift from Pfizer, Inc., Groton, CT), respectively. These agents are highly specific for the calcium channels that trigger synaptic glutamate release in brain cortex. To define which calcium channels are responsible for depolarization-evoked glutamate release, slices were incubated in 1 micro Meter conotoxin and/or 0.5 micro Meter agatoxin before challenge with 30 mM KCl. In addition, we tested whether glutamate release during hypoxia was sensitive to inhibition with calcium-free medium with 10 mM ethylene glycol-bis tetraacetic acid, because calcium entry at the presynaptic nerve terminal is required for transmitter vesicle release.

Before studing effects of anesthetics on glutamate release, slices were incubated in a beaker of oxygenated ABECF containing the test anesthetic, at test concentration, for 10 min. The volatile anesthetics were also added to the study cuvette immediately before the slice, to minimize loss.

Propofol, thiopental, enflurane, and halothane were studied at concentrations equivalent to approximately 1 minimum alveolar concentration or 1 Cp⁵⁰, as tabulated by Franks and Lieb. Saturated solutions of volatile anesthetics prepared in ABECF (2.5 ml commercially available anesthetic per 15 ml ABECF) and were diluted into the study cuvette to produce the desired final concentrations. We corrected minimum alveolar concentration values for the effects of age on halothane potency in young rats and assumed that age affects enflurane and halothane minimum alveolar concentration similarly. Propofol was prepared in 53 mM stock solution in dimethyl sulfoxide, and diluted into the cuvette and incubation solutions to produce a final concentration of 1–200 micro Meter. In preliminary experiments, 1–100 micro Meter propofol had no effect on KCl-evoked glutamate release, so a concentration of 200 micro Meter propofol (dimethyl sulfoxide concentration 0.38%) was used in the final studies. This concentration represents a value at the uppermost range of that expected in the plasma in humans during anesthesia. No correction was made for plasma protein binding.

For the volatile anesthetics, saturated anesthetic solution was syringe injected into cuvettes 10 s before introduction of the brain slice. The partial pressure of anesthetic in the cuvette was determined by gas chromatographic analysis of ABECF samples from cuvettes at 37 degrees Celsius. Anesthetics were extracted into air in a constant temperature water bath. The air was then injected into a Gow Mac Gas Chromatograph (Bridgewater, NJ) calibrated with appropriate standards. Halothane and enflurane loss was 15 plus/minus 3%(n = 6) for both agents during mock experiments. These loss estimates were used to adjust the amount of anesthetic added to cuvettes to produce the desired partial pressure. The concentration of propofol in the test ABECF was measured with ultraviolet spectroscopy, using the 200-nm absorption peak. The molar extinction coefficient of propofol was determined in ethyl alcohol, a solvent in which propofol is highly soluble. No loss was detected.

Control glutamate release during anoxia, hypoxia, and depolarization (KCl) was compared by one way analysis of variance, with the Tukey-Kramer test to compare all pairs. The effect of anesthetic agent on glutamate release was analyzed by one way analysis of variance with Dunnett's test to compare to control values. The effect of agatoxin and conotoxin on glutamate release was determined by Student's t test. Results are reported as mean plus/minus standard error (SEM) unless otherwise noted. A P value of < 0.05 was considered statistically significant. Analyses were performed using JMP 3.1 statistical software (SAS Institute Inc., Cary, NC). n values refer to the number of slices studied; the number of animals is also indicated.

Basal glutamate release from brain slices incubated in oxygenated ABECF was less than 0.005 nmol glutamate *symbol* mg protein sup -1 *symbol* s sup -1. Within 30–60 s of transfer to hypoxic ABECF ((P^Osub 2 20–35 mmHg), glutamate release increased to 0.029 plus/minus 0.004 nmol glutamate *symbol* mg protein sup -1 *symbol* s sup -1, and remained steady for 8–10 min (examples in Figure 2and summarized in Figure 3). Glutamate release during anoxia (0.10 plus/minus 0.002 nmol glutamate *symbol* mg protein sup -1 *symbol* s sup -1) was more than 300% of that produced by hypoxia. KCl-evoked glutamate release, which only releases synaptic glutamate stores, increased glutamate release to only 0.019 plus/minus 0.002 nmol glutamate *symbol* mg protein sup -1 *symbol* s sup -1 (F3-14), an amount significantly smaller than that during hypoxia or anoxia.

KCl-evoked synaptic glutamate release was reduced 80–90% by either agatoxin and conotoxin, and was virtually eliminated by combining agatoxin and conotoxin together (F3-14). Calcium-free solution also abolished KCl-evoked glutamate release. These data indicate that activation of N and P type voltage-gated calcium channels are required for the synaptic glutamate release process triggered by KCl. However, neither agatoxin nor conotoxin reduced glutamate release during hypoxia or anoxia (F3-14), indicating that extrasynaptic glutamate release (nonvesicular release) is the only measurable source of glutamate efflux from hypoxic or anoxic cortical brain slices.

None of the four anesthetics changed basal glutamate release from the brain slices (data not shown).

(Figure 4) shows the effects of the different anesthetics on anoxia- and hypoxia-evoked glutamate release; these data are also summarized in Table 1. At 680 micro Meter (equivalent to 1.0 minimum alveolar concentration), enflurane decreased the mean rate of glutamate release during hypoxia by 65%, from 0.029 plus/minus 0.004 to 0.010 plus/minus 0.002 nmol glutamate *symbol* mg protein sup -1 *symbol* s sup -1 (P < 0.05;Figure 5). Enflurane also significantly decreased glutamate release during anoxia.

Halothane also reduced glutamate release from hypoxic brain slices. At 325 micro Meter (approximately 1 minimum alveolar concentration), halothane decreased hypoxia-evoked glutamate release by 61% and anoxia-evoked glutamate release by 58%(P < 0.05). To further investigate the effects of halothane on hypoxia-evoked glutamate release, a separate dose-response study was done (Figure 6). Increasing concentrations of halothane produced greater inhibition of hypoxia-evoked glutamate release from the brain slices.

Sodium thiopental (50 micro Meter) decreased the release of glutamate during hypoxia by 65%, from 0.029 plus/minus 0.004 to 0.010 plus/minus 0.003 nmol glutamate *symbol* mg protein sup -1 *symbol* s sup -1 (P < 0.05). During anoxia, the reduction was not significant (F4-14).

Propofol (200 micro Meter) did not alter the glutamate release induced by hypoxia. Propofol remained in solution during the study conditions, as determined by ultraviolet spectroscopy. Because propofol did not decrease hypoxia-evoked glutamate release, the effect of propofol on glutamate release during anoxia was not studied.

Halothane, enflurane, and sodium thiopental, at clinically relevant concentrations, substantially reduce glutamate release during both hypoxic and anoxic conditions. These properties indicate a mechanism by which anesthetics may protect neurons in the intact brain from hypoxic or ischemic injury.

In brain slices, the release of glutamate during hypoxia is predominately extrasynaptic, because antagonism of the calcium channels required for depolarization-evoked glutamate release have no measurable effect on the rate of glutamate release. The failure of hypoxia to trigger calcium-channel-dependent glutamate exocytosis may be a reflection of the finding that synaptic function is rapidly lost during oxygen deprivation, due perhaps to a protective hypoxic presynaptic hyperpolarization, or due to failure of synaptic vesicle transport and fusion, which are ATP-dependent processes. Adenosine triphosphate concentrations in our brain slices decline substantially after less than 5 min of anoxia. The anoxia-induced failure of synaptic glutamate release has been described previously. The difference in the magnitude of the depolarization-evoked glutamate release by KCl and the extrasynaptic hypoxia-evoked glutamate release means that even if anoxic depolarization led to synaptic glutamate release, it would contribute only a small amount to the total pool of released glutamate. The depression of hypoxia-evoked glutamate release by the anesthetics must therefore predominately involve extrasynaptic glutamate transport processes.

Other studies have reported that anesthetics reduce neurotransmitter release from brain synaptosomes and brain slices during electrical stimulation or during KCl-evoked depolarization. To our knowledge, this is the first demonstration that anesthetics reduce glutamate release during hypoxia. The mechanisms responsible for our findings must be related to reduced extrasynaptic glutamate release. Specific mechanisms may include:

1. Anesthetic-induced reduction in ATP consumption during hypoxia, an effect that would delay ATP depletion and preserve the conditions necessary for carrier-mediated glutamate reuptake.

2. Preservation of normal sodium, potassium, and electrical gradients, the collapse of which are thought to produce a reversal of glutamate reuptake transporters during anoxia. The preservation of these gradients may be due to anesthetic inhibition of sodium or potassium channels, secondary to preserving the ATP necessary for ionic homeostasis, or possibly on other processes such as ion pumps (e.g., the Sodium sup +-Potassium sup + ATPase).

3. Reduction in depolarization-induced vesicular (synaptic) release from anesthetic inhibition of synaptic calcium channels, or other presynaptic processes. As discussed earlier, this must be a small contribution.

4. Because the net glutamate release from a brain slice is a balance between release and reuptake processes, it is possible, although unlikely, that anesthetics stimulate the reuptake mechanisms, or the metabolism of glutamate.

The question of whether anesthetics have cerebro-protective qualities in intact animals remains controversial; protective effects are seen in some studies (e.g., ) but not in others (reviewed in ). There is little doubt that several types of anesthetics, under the right conditions, reduce focal ischemic cerebral injury, although the mechanisms are uncertain. Barbiturates, for example, reduce necrosis from focal ischemia. In humans undergoing cardiac surgery, barbiturates decrease postoperative neuropsychiatric disturbances. Although cerebral protection with volatile anesthetics has been considered to be relatively weak, other studies have recently shown protective effects. Still, concern has been raised that in some of these studies anesthetic-induced hypothermia may have contributed to the protective outcomes.

The effects of anesthetics on ischemic glutamate release and tissue necrosis in intact animals are unclear. One study, involving in vivo microdialysis, failed to show that volatile anesthetics (isoflurane) or intravenous anesthetics (propofol and pentobarbital) reduce extracellular accumulation of glutamate during global cerebral ischemia. That study, however, measured glutamate in densely ischemic cerebral tissues, where the insult may have been more severe than in the current study with brain slices. Recently, Patel et al. showed that isoflurane reduced glutamate accumulation in an in vivo rat model of forebrain ischemia. However, in the same model isoflurane does not decrease infarct volume. Moreover, other studies have noted reduction in dopamine release in the corpus striatum by volatile agents, again without clear reduction in infarct volume in the model. It is possible that anesthetics may reduce glutamate release only in partially ischemic or moderately hypoxic regions (i.e., the ischemic penumbra), which is where they are protective. This proposal compliments recent observations in this same brain slice preparation that anesthetics reduce glutamate receptor-mediated calcium accumulation and ATP loss in penumbralike conditions, but not in ischemia-like conditions. .

Propofol did not reduce hypoxia-induced glutamate release, in contrast to the other anesthetics. The causes for this cannot be determined from this study. However, we have excluded the possibilities that propofol stimulates basal glutamate release and thus depletes the releasable glutamate pool, or that propofol was not present free in solution at the concentrations studied. In microdialysis studies, propofol failed to inhibit accumulation of glutamate in the ischemic core, although glycine release (glycine acts to increase the activity of the N-methyl-D-aspartate receptor) was blunted. Because propofol decreases brain oxygen consumption at least as much as other anesthetics, a reduction in ATP consumption is unlikely to explain the reduction in glutamate release produced by the volatile anesthetics and thiopental.

We conclude that several anesthetics (enflurane, halothane, sodium thiopental) reduce hypoxia-induced glutamate release from rat cortical brain slices at clinically relevant concentrations. If this effect occurs in the brains of intact animals during hypoxia, it may reduce the potential for glutamate neurotoxicity and hypoxic brain damage.

The authors thank Pfizer Pharmaceuticals for donating the agatoxin and Zeneca Pharmaceuticals for the propofol. They also thank Pompi Ionescu, M.D., and E. I. Eger II, M.D., for performing the volatile anesthetic concentration measurements.