Glutamate transporters located in the plasma membrane of cerebral astrocytes take up excitatory neurotransmitters from the synaptic cleft. In diseases characterized by oxidative stress, the extracellular glutamate concentration increases and contributes to neuronal death. The authors wanted to determine whether propofol defends brain cells against oxidant-induced changes in their transport of glutamate.
Primary cultures of rat cerebral astrocytes were exposed to tert-butyl hydroperoxide (1 mM) to serve as an in vitro model of oxidative stress. Astrocytes were incubated with propofol for 2 h and tert-butyl hydroperoxide was added for the final hour. Alternatively, astrocytes were incubated with tert-butyl hydroperoxide for 30 min and then with propofol for another 30 min. Control cells received drug vehicle rather than propofol. The rate of uptake of glutamate, the efflux of the nonmetabolizable analog D-aspartate, and the intracellular concentration of the endogenous antioxidant glutathione were measured.
Tert-butyl hydroperoxide decreased the glutathione concentration and inhibited glutamate uptake but had a negligible effect on D-aspartate efflux. At clinically relevant concentrations, propofol did not affect the glutathione concentration but did prevent the effect of tert-butyl hydroperoxide on glutamate transport. Furthermore, the addition of propofol after tert-butyl hydroperoxide reversed the inhibition of glutamate uptake.
Propofol prevents and reverses the inhibition of excitatory amino acid uptake in astrocytes exposed to tert-butyl hydroperoxide. The ability of propofol to defend against peroxide-induced inhibition of glutamate clearance may prevent the pathologic increase in extracellular glutamate at synapses, and thus delay or prevent the onset of excitotoxic neuronal death.
GLUTAMATE is the principal excitatory amino acid transmitter in the central nervous system. It is released from nerve endings and activates postsynaptic receptors. The synaptic action of glutamate is terminated by its uptake into cells. [1–4] High-affinity glutamate transporters are localized in the plasma membranes of both neurons (containing EAAT3/EAAC1 and EAAT4 transporters) and astrocytes (containing EEAT1/GLAST and EEAT2/GLT-1 transporters). Astrocytic transporters are positioned appropriately to take up glutamate from the synaptic cleft.  Furthermore, there is evidence that they are essential for normal glutamate-mediated transmission. Knockout experiments have shown that eliminating these transporters leads to glutamate-induced excitotoxicity. [2–4] Continuous intraventricular administration of antisense oligonucleotides for GLAST and GLT-1 increases the extracellular glutamate concentration and neurodegeneration.  The brains of mice homozygous for the GLAST null mutation exhibit uptake of glutamate in vitro and increased sensitivity to trauma in situ.  Mutant mice deficient in GLT-1 show lethal spontaneous seizures and increased susceptibility to acute brain injury. 
Extracellular glutamate concentrations increase after trauma, stroke, and other conditions characterized by oxidative stress, which is defined as an excess production of free radicals relative to the rate at which they can be eliminated.  One probable reason that extracellular glutamate concentrations increase to destructive levels during trauma and stroke is that glutamate uptake is inhibited. Glutamate transporters are sensitive to redox agents. [6–9] For example, the oxidant hydrogen peroxide inhibits glutamate uptake by cultured astrocytes, and the thiol-specific reductant dithiothreitol reverses this inhibition. 
Often, anesthetics are administered to patients at risk of or after injuries that decrease cerebral antioxidant defenses. Propofol (2,6-diisopropylphenol) contains a phenolic hydroxyl group that donates electrons to free radicals and therefore may complement endogenous antioxidants. [10,11] When administered before or immediately after cerebral ischemia, it is neuroprotective in animal models. [12–16] Propofol also may be protective during intraoperative cerebral ischemia in anesthetized patients.  The beneficial effects of propofol are not caused by a direct action on excitotoxic receptors. On the contrary, propofol exacerbates the damage to neurons caused by activation of the neuronal N-methyl-D-aspartate type of glutamate receptors.  We hypothesize that this anesthetic combats free radical attack on neurotransmitter transporters. The oxidant tert-butyl hydroperoxide (t-BOOH) has been used in pharmacologic models of brain trauma and reperfusion injury. [19,20] The purpose of the current study was to determine whether propofol could prevent or reverse the inhibition of glutamate transport caused by t-BOOH.
Materials and Methods
L-[(3) H]glutamic acid ([(3) H]Glu, 38–46 Ci/mmol) and D-[(3) H]aspartic acid (D-[(3) H]Asp, 20 Ci/mmol) were purchased from Amersham Canada (Oakville, Ontario, Canada). L-Ascorbic acid; 4,4′-diisothiocyanatostilbene-2,2′-disulfonic acid; dithiothreitol; L-Glu; reduced glutathione; and t-BOOH were obtained from Sigma Chemical Company (St. Louis, MO). Horse serum was purchased from Gibco Laboratories (Burlington, Ontario, Canada). Propofol was purchased from Aldrich Chemical Company (Oakville, Ontario, Canada), and Intralipid was purchased from Clintec Nutrition Company (Mississauga, Ontario, Canada). Intralipid combines fatty acid with glycerol and lecithin to make micelles for carrying propofol. Propofol was dissolved in Intralipid, and those cell cultures used for vehicle controls were exposed to the same concentration of Intralipid as were the cultures incubated with propofol.
Primary cultures of astrocytes were prepared from the neopallium of 1-day-old Wistar rats and grown in horse serum-supplemented minimum essential medium according to our published procedure.  After 2 weeks, the cultures reached confluence and each 60-mm dish contained approximately 3 million cells. Astrocytes were used for transport experiments after 17–26 days in culture.
To determine the time course of glutamate uptake, astrocytes were incubated for 30 min in transport medium (containing 134 mM NaCl, 5.2 mM KCl, 1.8 mM CaCl2, 0.8 mM MgSO4, 10 mM D-glucose, and 20 mM HEPES; pH 7.3) without radiotracer and then for 30 s-2 min in transport medium containing [(3) H]Glu (100 [micro sign]M, 10 mCi/mmol, 37 [degree sign]C). Na+-independentuptake was measured in nominally Na+-freemedium in which N-methyl-D-glucamine+was substituted isoosmotically for Na+. Some cultures received the anion transport inhibitor 4,4′-diisothiocyanatostilbene-2,2′-disulfonic acid (1 mM), which was added at the same time that [(3) H]Glu was. Uptake was terminated by washing the cultures with ice-cold Tris-sucrose solution (pH 7.3). Cells were harvested by osmotic lysis (1 ml water/dish) and mechanical scraping. An aliquot of the cell harvest was used for protein measurement,  and the rest was combined with a scintillation cocktail. The radioactive contents of the buffer and cells were measured by liquid scintillation counting. Uptake rates were computed based on the specific activity of radiolabeled glutamate in the media and expressed as micromolar glutamate per gram of cell protein per minute.
To determine the prophylactic effect of propofol, astrocytes were first incubated in serum-free minimum essential medium for 3 h, and propofol or Intralipid was added to the medium for the final 1 h. The cells were incubated for an additional 1 h in transport medium containing propofol (or Intralipid) with or without 1 mM of the oxidant t-BOOH. Finally, the cells were washed with transport medium and glutamate uptake was measured in 1-min transport assays, as described before. In parallel experiments, instead of the glutamate transport assay, the astrocytes were harvested and the intracellular concentration of reduced glutathione was measured according to the spectrophotometric procedure described by Siushansian et al. 
To determine the therapeutic effect of propofol on glutamate uptake, astrocytes were incubated in serum-free minimum essential medium for 3 h. The cells were incubated for 30 min in transport medium with or without 1 mM t-BOOH. Subsequently, the cells were washed and incubated for another 30 min in transport medium containing propofol or Intralipid. Alternatively, dithiothreitol (0.1 mM, 3 min) was used to reverse the effect of t-BOOH. Finally, the cells were washed with transport medium and glutamate uptake was measured in 1-min transport assays.
The efflux of D-Asp, a nonmetabolizable analog of glutamate, was measured to determine the effects of oxidative stress and propofol on the ability of astrocytes to retain excitatory amino acid transmitters. Astrocytes were incubated overnight in serum-supplemented minimum essential medium containing D-[(3) H]Asp (10 [micro sign]M, 40 mCi/mmol) to load the cells with radiotracer. Then the cells were incubated for 30 min in transport medium with or without 1 mM t-BOOH. Finally, the cultures were transferred to transport medium that lacked t-BOOH but contained Intralipid, 0.02%(vol/vol), or 40 [micro sign]M propofol. Aliquots of media and cell harvests were analyzed by liquid scintillation counting. Experiments were also performed to quantify radiolabeling of proteins by D-[(3) H]Asp. After harvesting cells in 1 ml water, a 100-[micro sign]l aliquot was counted immediately in scintillation cocktail and the remaining 900 [micro sign]l of the cell harvest was frozen at -70 [degree sign]C overnight, thawed at 4 [degree sign]C, homogenized with a Tekmar electric homogenizer (Tekmar Co., Cincinnati, OH) for 15 s, and mixed with an equal volume of 20% trichloroacetic acid. The mixture was centrifuged and the pellet was washed three times with 2 ml ice-cold diethyl ether and dissolved in 1 ml of the scintillation cocktail. This procedure showed that only 0.2 +/- 0.03% of cell radioactivity was bound to acid-precipitable protein (n = 6 cultures).
Data are expressed as the mean +/- SEM values for n number of experiments. Bartlett's test was used to identify normal data distribution, and analysis of variance and the Tukey-Kramer test were used subsequently to identify differences between means. P < 0.05 was considered significant.
Uptake of [(3) H]Glu by astrocytes was inhibited by excess unlabeled glutamate (1 mM;Figure 1). 4,4′-Diisothiocyanatostilbene-2,2′-disulfonic acid (1 mM), which is an impermeant inhibitor of anion transport, also blocked glutamate uptake (data not shown), consistent with mediation of uptake by transporters located in the plasma membrane. In addition, substitution of N-methyl-D-glucamine+for Na+in the transport medium abolished glutamate uptake (Figure 1), consistent with the findings of previous studies that showed that astrocytic glutamate uptake is mediated by Na+-dependenttransporters. [2,3,9,24] Na+-dependentglutamate uptake proceeded linearly with time for at least 1 min (Figure 1), so subsequent experiments used 1 min transport assays to estimate the initial rate of uptake. The uptake rate was not affected by the age of the cultures during the period studied (from days 17–26 in culture; data not shown). Propofol, even at high concentration (40 [micro sign]M), did not affect the glutamate uptake rate in control astrocytes (Figure 2).
Exposure to 1 mM t-BOOH for 1 h inhibited the glutamate uptake rate by approximately 60%(Figure 2). Propofol prevented the decrease in glutamate transport activity caused by this severe oxidative stress. Most of the effect of the 1-h exposure to t-BOOH was prevented with 8 [micro sign]M propofol, and no significantly greater improvement occurred with 40 [micro sign]M propofol (Figure 2). Total cell protein was not affected by t-BOOH (Table 1). However, t-BOOH decreased the cell glutathione content to 21% of control values (Figure 3). This decrease was not prevented by propofol (40 [micro sign]M;Figure 3).
When astrocytes were exposed to t-BOOH for 30 min and the oxidant was removed, no spontaneous recovery of glutamate transport occurred during the following 30-min period (Figure 4). However, dithiothreitol (Table 2) and propofol (Figure 4) each reversed the inhibitory effect of t-BOOH on glutamate uptake. When exposure to aqueous vehicle or t-BOOH (1 mM, 30 min) was followed by incubation for 30 min with clinically relevant concentrations of Intralipid and propofol, the glutamate uptake rates were aqueous vehicle + Intralipid 39 +/- 3 [micro sign]mol/g cell protein/min; t-BOOH + Intralipid 19 +/- 6 [micro sign]mol/g cell protein/min; t-BOOH + 1 [micro sign]M propofol 31 +/- 4 [micro sign]mol/g cell protein/min; t-BOOH + 8 [micro sign]M propofol 37 +/- 8 [micro sign]mol/g cell protein/min; t-BOOH + 40 [micro sign]M propofol 39 +/- 3 [micro sign]mol/g cell protein/min (n = 3 experiments). Statistical analysis of a larger number of experiments confirmed that the effect of the 30-min exposure to t-BOOH was reversed completely by subsequent incubation with 8 - 40 [micro sign]M propofol (Figure 5).
Our investigation determined whether the effects of t-BOOH (1 mM) and propofol (40 [micro sign]M) observed in Figure 5were the result of the efflux of excitatory amino acids. D-[(3) H]Asp was lost slowly from preloaded astrocytes during conditions similar to those used for the transport experiments shown in Figure 5. Astrocytes were loaded with D-[(3) H]Asp, incubated for 30 min with t-BOOH or aqueous vehicle, and finally incubated for 30 min with Intralipid or propofol. The cell contents of D-[(3) H]Asp at the end of the efflux periods were aqueous vehicle + Intralipid 3.2 +/- 0.7 [micro sign]mol/g protein; aqueous vehicle + propofol 3.1 +/- 0.6 [micro sign]mol/g protein; t-BOOH + Intralipid 3.2 +/- 0.6 [micro sign]mol/g protein; t-BOOH + propofol 3.1 +/- 0.5 [micro sign]mol/g protein (n = 3 experiments). Thus, t-BOOH and propofol had no marked effect on the efflux of excitatory amino acids.
Anesthetics and analgesics alter neurotransmission. [25,26] They act presynaptically to slow glutamate release and postsynaptically to modulate receptor-mediated signaling. Only limited information is available about another postsynaptic action of anesthetics, specifically modulation of glutamate uptake by nervous system cells (see Griffiths and Norman  for a review of the older literature). Halothane, enflurane, isoflurane, and sevoflurane enhance glutamate uptake by cultured astrocytes.  Isoflurane also increases the maximal rate of uptake of glutamate through the high-affinity glutamate transport system in synaptosomes prepared from rat cerebral cortex.  The increase in the glutamate uptake rate in the presence of volatile anesthetics potentially attenuates excitatory synaptic transmission. Conversely, the intravenous anesthetics thiopentone, ketamine, and propofol fail to alter Na+-dependentglutamate uptake by rat brain synaptosomes during basal (control) conditions.  Similarly, we observed no effect of propofol on glutamate uptake by astrocytes during control conditions. Propofol inhibits the exocytosis of glutamate evoked by 4-aminopyridine and veratridine in rat brain synaptosomes.  In the current study, we observed no effect of propofol on the efflux of D-Asp, a nonmetabolizable analog of glutamate, from astrocytes during control conditions.
Oxidative injury occurs in many disorders of the central nervous system, including radiation injury, sepsis, neurodegenerative diseases, trauma, and stroke. [5,31] Reactive oxygen species influence the structure and function of high-affinity Na+-dependentglutamate transporters responsible for neuronal and astrocytic uptake of glutamate. [6–9] Inhibition of these transporters contributes to the increase in the extracellular glutamate concentration to excitotoxic levels. [1–5] The sensitivity of glutamate transport activity to oxidants may be explained by the requirement of keeping sulfhydryl groups on the transporters in their reduced forms. [6–9]
Here we used an in vitro model of oxidative injury in which t-BOOH inhibits glutamate uptake by astrocytes. t-BOOH did not affect markedly total cellular protein during the treatment period. A further indication that this oxidative stress did not kill the astrocytes within the treatment period is that it did not diminish the cellular retention of excitatory amino acids, as judged from the negligible effect on the release of D-Asp. We found that propofol defends glutamate uptake from the oxidative stress triggered by t-BOOH. Furthermore, propofol restores glutamate transport activity after astrocytes have been stressed. This anesthetic has chemical properties that retard the oxidation of lipids and protein thiols in the plasma membrane and subcellular organelles. The following discussion considers the mechanisms by which the antioxidant properties of propofol may restore excitatory amino acid transport after peroxide-induced oxidative stress.
Propofol slow lipid peroxidation in many test systems, including aqueous suspensions of arachidonic or linoleic acids, plasma, isolated mitochondria, liver microsomes, brain synaptosomes, and erythrocytes. [10,32–37] However, clinically relevant concentrations of propofol achieve only incomplete inhibition of lipid peroxidation in in vitro tests. As examples of the low potency of propofol, Ansley et al.  observed that 100 [micro sign]M propofol inhibits lipid peroxidation by approximately 50%, and Green et al.  found that 280 [micro sign]M propofol inhibits lipid peroxidation by only 40%.
Experiments with isolated mitochondria have shown that t-BOOH causes glutathione oxidation, protein thiol oxidation, and protein cross-linking, followed by the mitochondrial membrane permeability transition.  Thiols are more susceptible to t-BOOH in mitochondria from brain than in those from liver.  The mitochondrial permeability transition causes collapse of the mitochondrial membrane potential, mitochondrial swelling, calcium release, excessive production of oxygen free radicals, and apoptosis. Propofol has a direct inhibitory action on the permeability transition in cardiac mitochondria and thus protects these organelles from t-BOOH-induced swelling.  However, this effect requires propofol concentrations an order of magnitude larger than those used in the current study of astrocytic glutamate transport. The experimental findings reviewed here indicate that propofol's inhibition of lipid peroxidation and its direct effect on mitochondrial permeability are not potent enough to account entirely for the anesthetic's rescue of astrocytic glutamate transport after t-BOOH exposure.
Ischemia-reperfusion injury to the brain is marked by oxidation of protein and nonprotein thiols.  A pharmacologic model of reperfusion injury is the intracerebroventricular injection of t-BOOH.  This alkyl peroxide oxidizes thiols. [19,42] Oxidation of protein sulfhydryl groups precedes lipid peroxidation when microsomes are incubated with t-BOOH.  In contrast, propofol protects protein thiols. The latter is a potent effect because reduction of protein disulfides can be detected at propofol concentrations as low as 5 [micro sign]M.  We observed that both propofol and dithiothreitol, which also reduces disulfides, reverse the inhibitory effect of t-BOOH on astrocytic glutamate transport. It is possible that propofol rescues glutamate transporters in astrocytes exposed to t-BOOH by keeping sulfhydryl groups on the transporters in the reduced form necessary for normal function.
t-BOOH oxidizes glutathione to glutathione disulfide and glutathione sulfoxide in the brain.  It also oxidizes glutathione in cultured astrocytes.  Propofol does not prevent glutathione depletion by t-BOOH in erythrocytes  or astrocytes (in the current experiments), indicating that the anesthetic's action is targeted to the plasma membrane and not to the cytosol, where most glutathione is located. General anesthetics, because of their hydrophobicity, preferentially accumulate at the lipid-water interface on the surfaces of cells. Thus, antioxidant anesthetics may be especially well suited to influence redox reactions on these surfaces. Furthermore, it is tempting to speculate that propofol's most potent antioxidant actions on proteins occur in hydrophobic pockets that provide anesthetic-binding sites. 
The anesthetic concentrations of propofol are sufficient for neuroprotection in stroke. [12–17] The concentrations of propofol that maintain glutamate transport in cell cultures exposed to t-BOOH, are lower than plasma or tissue concentrations of total (free and protein-bound) propofol during general anesthesia. For instance, 30 min after anesthesia is established in the rat by infusion of propofol (60 mg [middle dot] kg-1[middle dot] h-1intravenously), the concentrations of propofol in plasma and the brain are 5 [micro sign]g/ml (28 [micro sign]M) and 39 [micro sign]g/g tissue (220 [micro sign]mol/g tissue), respectively.  As another example, the arterial blood concentrations of propofol that prevent movement after skin incision in healthy women are 14–21 [micro sign]g/ml (79–118 [micro sign]M) for 50–95% of the patients. 
In conclusion, propofol can defend against inhibition by reactive oxygen species of plasma membrane glutamate transport when astrocytes are exposed to t-BOOH. This contribution to antioxidant defense may model a generally applicable mechanism through which antioxidants with similar solubility characteristics could protect brain neurons. Rescue of glutamate transport activity from oxidative injury may slow the increase in glutamate concentration at postsynaptic receptors and thus delay or prevent excitotoxic injury.
The authors thank Raphael Yan for assistance with these experiments.