Anesthetics Affect the Uptake but Not the Depolarization-evoked Release of GABA in Rat Striatal Synaptosomes 

Handling procedures, as written in the Guide for the Care and Use of Laboratory Animals, were followed throughout. Experiments were performed on male Sprague-Dawley rats (Iffa-Credo, France) weighing 200–225 g and housed on a 12:12 light/dark cycle with food and water ad libitum.

The effects of the following anesthetic and/or pharmacologic agents were studied on³H-GABA uptake and release: nipecotic acid (a specific inhibitor of GABA uptake, 10⁶to 10³M; Sigma, La Verpilliere, France), thiopental (10⁵to 10²M: Nesdonal, Specia, France), ketamine (10⁵to 10³M; Sigma), propofol (10 sup 5 to 10³M), etomidate (10⁵to 10³M), clonidine (Sigma; 10⁹to 10⁶M), halothane (Fluothane, Zeneca Pharmaceuticals, Cergy, France), enflurane (Ethrane), and isoflurane (10 sup 5 to 10²M; Forane, Abbott, Rungis, France). Volatile anesthetics were first prepared as a 10¹-M stock solution in dimethyl sulfoxide (DMSO, Merck Sharp and Dohme, Darmstadt, Germany). Dilutions of the volatile anesthetics were obtained by using glass syringes (Hamilton, Reno, NV) of appropriate volumes of the 10¹-M solution into 5 ml CSF. The final aqueous concentrations of halogenated agents in the tubes (5 ml capacity) after dilution and equivalent periods were determined in a parallel experimental set up by chromatography. [18] Precautions were taken to minimize evaporation of volatile anesthetics leaving the tubes covered with Teflon foils during the serial dilutions and with Parafilm (Greenwich, CT) for the final tubes placed into the superfusion device. The vehicles tested were DMSO (used at a 0.1–10% concentration range for preparation of halogenated agents, propofol, and etomidate) and 0.01% thymol (Merck, Sharp and Dohme, used for stabilizing halothane).

Preparation of synaptosomes has been reported in detail elsewhere. [8] Synaptosomes were diluted up to 0.16 mg/ml in an ice-cold artificial cerebrospinal fluid (CSF; in mM: NaCl 126.5, NaHCO sub 3 27.5, KCl 2.4, KH²PO⁴0.5, CaCl²1.1, MgCl²0.83, Na²SO⁴0.5, glucose 11.8, aminooxyacetic acid (Sigma, an inhibitor of GABA catabolism) 0.1, and beta-alanine (an inhibitor of the GABA carrier in glial cells; Calbiochem, San Diego, CA) 1, adjusted to pH 7.3 with 95%/5%(v/v) oxygen/carbon dioxide mixture. In Calcium sup 2+-free experiments, CaCl²was omitted from the artificial CSF.

Uptake of [2.3]³H-GABA (Amersham, UK; 60 Ci/mmol) was performed by incubating 1-ml aliquots of synaptosomes (5 min, 37 degrees Celsius) with 20 nM³H-GABA in either the absence (control) or presence of anesthetic and/or pharmacologic agents. This duration of incubation was determined according to preliminary experiments indicating that³H-GABA uptake increased linearly and reproducibly with time between 3 and 10 min of incubation at 37 degrees Celsius in this preparation. The reaction was stopped by vacuum-filtration through Whatman GF/F filters (0.70 micro meter retention capacity) and washing with ice-cold CSF (30 ml), and the radioactivity retained on the filters (counts per minute) measured by liquid scintillation spectrometry using Aquasol-2 (New England Nuclear, Boston, MA). Specific³H-GABA uptake was considered the difference between the radioactivity measured in both the presence and absence of 10³M nipecotic acid.

To determine whether any observed affects of anesthetics were not due to irreversible synaptosomal damage,³H-GABA uptake was examined after removal of the anesthetic from the buffer/synaptosomal mixture by vacuum and stirring rapidly for 60 min, as previously reported in dopamine uptake studies. [7] Aliquots were removed at 30 and 60 min from an uptake mixture initially containing a concentration of anesthetic corresponding to that one inhibiting 50% of the specific high affinity GABA uptake (IC⁵⁰), and³H-GABA uptake was determined as a function of the time that the mixture had been exposed to vacuum and stirring. Control experiments also were run in parallel to allow comparison with anesthetic-free mixtures exposed to the same conditions of vacuum and stirring.

To distinguish between competitive and noncompetitive mechanisms of any observed inhibition of³H-GABA uptake by anesthetics, kinetics of anesthetic-induced inhibition were investigated by incubating synaptosomes (37 degrees Celsius, 5 min) with increasing concentrations of³H-GABA (20–200 nM). Reaction was stopped by the addition of ice-cold CSF followed by vacuum filtration, and the radioactivity retained on the filters was measured. The kinetic parameters of the reaction (V^max, Km) were determined from double reciprocal plots of experiments performed in the absence (control) and presence of a fixed concentration of anesthetic corresponding to the IC sub 50 value.

Release experiments were performed as previously reported by Galli et al. [19] The synaptosome suspension was first incubated with sup 3 H-GABA (10 min, 37 degrees Celsius). Aliquots (1 ml) were pipetted into each of the 12 superfusion chambers and embedded in Whatman GF/F glass filters by light suction, then superfused at a 0.5 ml/min flow rate with artificial CSF using a superfusion device equipped with an automatic fraction collector (Brandel, Gaithersburg, MD). After a 30-min washing step (37 degrees Celsius), serial 0.5 ml fractions were collected during three consecutive periods. The mean value of³H-GABA radioactivity measured over the 8 first min was considered the basal (spontaneous)³H-GABA release. During the next 5 min of superfusion, either no treatment (time-dependent control) or various pharmacologic and/or anesthetic agents (evoked release) were delivered to the synaptosome preparation, and radioactivity was again estimated in each 0.5-ml fraction. The potassium chloride concentrations used were low (9 and 15 mM) to mimic physiologic depolarizations. For the last 5 min, radioactivity was determined from 1-min fractions in the absence of any treatment. The influence of a drug on³H-GABA release was assessed by calculating the difference between the maximal peak of radioactivity observed in both the presence and the absence (time-dependent control) of the pharmacologic or anesthetic agents used. To allow comparison with previous studies, the effect of an anesthetic agent on spontaneous release was expressed as a percentage of either increase or decrease from the time-dependent control basal release. The action of an anesthetic on depolarization-evoked release was displayed as a fraction of the control³H-GABA peak elicited by KCl. This presentation of data facilitated comparison with results obtained from previous studies. [19–21] Nipecotic acid was used in the release experiments as well to eliminate the possibility of reuptake occurring during superfusion of the synaptosomes. Therefore, the values taken into account were those obtained by subtracting results done in the absence nipecotic acid from those in the presence of nipecotic acid (1 mM) only.

In uptake experiments, concentration-response curves were generated directly from computer and IC⁵⁰values for inhibition of sup 3 H-GABA uptake were calculated using the GraphPAD Software (San Diego, CA). The functions used to fit the curves to the data was the following four-parameter logistic equation:Equation 1where x represents the logarithm of anesthetic concentration, A and B, respectively, the minimum (bottom) and maximum (top) of x values. C is the logarithm of the IC⁵⁰value, and D the Hill coefficient or slope factor that is positive for curves in which the Y value increases with increasing x and is negative for curves in which the Y value decreases with increasing x. Results were considered reliable only if they had been reproduced in four independent experiments at least (each of them run in triplicate). Uptake and release data were analyzed by ANOVA followed by Student's t test corrected for the number of comparisons. The measured concentrations of volatile anesthetics were compared to the theoretical ones using the Student's t test. A P value less than 0.05 was considered the threshold for significance. Unless specified, data are expressed as mean plus/minus SD.

The concentrations of volatile anesthetics were measured after dilution and equivalent periods of time in three independent experiments (each of them run in triplicate). The actual concentrations measured in the final samples and corresponding to the anticipated 100, 10, 2, 1, 0.5, and 0.1 mM, respectively, were 93 plus/minus 11, 9.4 plus/minus 0.4, 1.7 plus/minus 0.3, 0.96 plus/minus 0.06, 0.43 plus/minus 0.05, and 0.07 plus/minus 0.01 mM, respectively, for halothane; 95 plus/minus 10, 9.2 plus/minus 0.5, 1.8 plus/minus 0.2, 0.91 plus/minus 0.07, 0.41 plus/minus 0.06, and 0.07 plus/minus 0.01 mM, respectively, for isoflurane; and 91 plus/minus 9, 9.6 plus/minus 0.4, 1.8 plus/minus 0.3, 0.93 plus/minus 0.06, 0.44 plus/minus 0.05, and 0.08 plus/minus 0.01 mM, respectively, for enflurane. As was observed in a previous study in which a similar methodology was used for dilution of halogenated agents from a concentrated DMSO solution, [22] these values were nonsignificantly different from those anticipated theoretically.

The³Hydrogen-GABA concentration estimated in the incubation medium during the uptake phase was 125 plus/minus 30 pM. Synaptosomal³Hydrogen-GABA uptake after the 5-min incubation averaged 0.87 plus/minus 0.08 pM/mg protein. The specific uptake inhibitor nipecotic acid produced a dose-related inhibition of³Hydrogen-GABA transport into synaptosomes (IC⁵⁰= 3.5 plus/minus 0.3 x 10⁶M, Figure 1). A 10³-M concentration of nipecotic acid was found to inhibit more than 95% of³Hydrogen-GABA uptake. Of the intravenous anesthetics tested, propofol, etomidate, thiopental, and ketamine induced an inhibitory effect on³Hydrogen-GABA transport (IC⁵⁰values (mean plus/minus SEM)= 4.6 plus/minus 0.3 x 10 sup -5 M, 5.8 plus/minus 0.3 x 10 sup -5 M, 2.1 plus/minus 0.4 x 10 sup -3 M, and 4.9 plus/minus 0.5 x 10⁴M for propofol, etomidate, thiopental, and ketamine, respectively (Figure 1). In contrast, neither clonidine nor volatile anesthetics were found to significantly affect³Hydrogen-GABA uptake by the synaptosomal preparation (Figure 1). In agreement with previous studies, DMSO and thymol, applied at the vehicle concentrations used in the current study, failed to affect GABA uptake. [23].

Reversibility of the effects of propofol, etomidate, thiopental, and ketamine on the uptake of³Hydrogen-GABA was obtained in all cases. As shown on Figure 2, uptake activities of the anesthetic-exposed solutions were not significantly different from time-dependent control uptake activities measured at either 30 or 60 min after washout.

The kinetic parameters of³Hydrogen-GABA uptake were determined in either the absence (control) or presence of the concentrations of intravenous anesthetics inhibiting 50% of the transport of GABA into synaptosomes (IC⁵⁰) reported above. Changes in Km values due to anesthetic exposure were nonsignificant in comparison with control. In contrast, all anesthetic agents tested produced a significant reduction in V^max(Figure 3and Table 1). These data suggest that inhibition of³Hydrogen-GABA uptake by these anesthetics is mediated by noncompetitive mechanisms.

Results obtained from the release assay are illustrated in Figure 3. No significant variation was observed in the time course of spontaneous³Hydrogen-GABA release. No significant modification in baseline occurred when Calcium²+ was omitted from the superfusion medium. The effect of nipecotic acid (10 sup -3 M) on spontaneous release failed to achieve statistical significance. In contrast, KCl (9 and 15 mM) application induced a significant and dose-related increase from baseline. KCl, 9 and 15 mM, respectively, elicited a 34 plus/minus 10%(P < 0.01) and 61 plus/minus 13%(P < 0.001), respectively, increase from basal release (Figure 4). This³Hydrogen-GABA peak occurred rapidly after KCl application and was followed by a rapid return of radioactivity to baseline, which was observed before the stimulus was discontinued. The KCl-evoked³Hydrogen-GABA release was completely blocked when external Calcium²+ was omitted from the medium (Figure 4). All anesthetic agents tested failed to affect significantly both the spontaneous and KCl-evoked release of³Hydrogen-GABA (Figure 5and Table 2).

The current study demonstrates that some intravenous but not volatile anesthetics exert significant actions on the presynaptic GABAergic nerve terminals located in the rat striatum in vitro. Propofol, etomidate, thiopental, and ketamine were found to inhibit the specific high-affinity GABA uptake in this brain area. In contrast, the release of³Hydrogen-GABA evoked by KCl depolarization was not affected by the anesthetics studied.

In the current study, the effects of anesthetics on the uptake and on the depolarization-evoked release of³Hydrogen-GABA were examined on synaptosomes prepared from the rat striatum. Advantages and drawbacks of working on this preparation have been addressed in a study originating from our laboratory. [9] Synaptosomes provide a reliable and reproducible model for analyzing, in vitro, the direct presynaptic effects of anesthetics on neurotransmitter release or uptake. [4–9,19] However, unlike striatal slices, they preclude looking at indirect, tetrodotoxin-sensitive mechanisms or to address the role played by diffusible mediators involved in neurotransmitter release, such as nitric oxide. Therefore, the results obtained here apply only to uptake and release in synaptosomes, whereas nerve terminals in situ subjected to impulse activity and a transmitter-enriched environment may exhibit a different behavior.

Basically, the striatal synaptosomes consist of different subpopulations of presynaptic nerve endings originating from a restricted but functionally important brain area. The specific GABA carrier is present on synaptosome membranes of the GABA terminals located in this structure. [19] GABA released from striatal nerve endings is likely to originate from either striatal GABA interneurons or from recurrent collaterals arising from striatopallidal or striatonigral output neurons. [21,25] Interestingly, convincing evidence has been provided that the striatum is anatomically and functionally heterogeneous. Subcompartmentalization of this brain area into matrix and striosomes correlates with major biochemical and pharmacologic differences, such as the control of presynaptic dopamine release in the cat. [24,26] To date, whether these differences also apply to GABA release in the rat striatum remains unknown. However, it cannot be excluded that differences might exist in the actions of anesthetics on GABA uptake or release between matrix- and striosome-enriched areas of the rat striatum.

The originality of the current study is that we separately investigated the uptake and release processes. The specific GABA uptake inhibitor nipecotic acid was found to potently block the incorporation of³H-GABA into synaptosomes. This indicates that the uptake process studied here was the specific one for GABA. [27] Of the different anesthetics examined, only propofol, etomidate, ketamine, and thiopental were found to significantly inhibit³H-GABA uptake. The effects of propofol and etomidate were observed at concentrations that were clinically relevant, because these corresponded to the range of plasma concentrations that might be observed in the anesthetized patient. Reversibility of anesthetic effects on GABA uptake further suggests that these actions might occur in vivo, because they were not associated with irreversible damage of synaptosome membranes. However, the IC⁵⁰values reported for thiopental and ketamine suggest that the effects of these compounds on GABA uptake were obtained for anesthetic concentrations surpassing clinical relevance.

Kinetics of³H-GABA uptake were in agreement with the characteristics of the high-affinity GABA carrier located on the neuronal membrane. [28] Also consistent with previous studies of membrane transporters, the kinetics of GABA uptake inhibition by intravenous anesthetics were characterized by a noncompetitive profile. [7,8] Our data do not allow clarification of the subtle molecular mechanisms involved in these effects. Whether GABA carrier inhibition proceeds through anesthetic/protein interactions or via indirect effects on lipidic membranes of the transporter environment remains to be delineated. The widely varying structures of the lipid and nonlipid soluble substances, such as propofol, etomidate, thiopental, or ketamine, could be supportive of a nonspecific mechanism involving lipid interactions and synaptosomal membranes.

On the other hand, clonidine and the volatile anesthetics failed to interfere with GABA transport. It might have been argued that the lack of effect of halogenated agents on GABA uptake was explained by much lower final concentrations than those expected theoretically with respect to the dilution procedure used here (serial dilutions from a concentrated DMSO stock solution). However, data obtained from gas chromatographic analysis of the final volatile anesthetic concentrations rule out this possibility. Interestingly, clinical concentrations of halothane have been reported to inhibit dopamine transport into synaptosomes. [7] Although these apparently differential effects of volatile anesthetics on dopamine versus GABA uptake could be explained in part by differences in the experimental conditions achieved in the two studies, it can be suggested that the specific GABA and dopamine carriers exhibit distinct sensitivities to inhalational anesthetics in the rat brain.

We found that anesthetic agents had no effect on either spontaneous or depolarization-evoked³H-GABA release. Also, clonidine had no effect on unstimulated³H-GABA release from the striatal nerve endings. This may appear surprising, because clonidine has been shown to stimulate³H-GABA release on synaptosomes originating from the rat hippocampus. [29] Furthermore, elegant studies performed on the rat hippocampal slice preparation have shown that alpha-adrenergic agonists increase GABAergic inhibition by presynaptically enhancing endogenous GABA release. [30] These differences likely can be attributed to the lack of noradrenergic innervation of the striatum, while the hippocampus receives massive noradrenergic inputs from the locus ceruleus. [24,31].

It can be speculated that the effects of anesthetics on GABA release evoked by depolarization may be more relevant to the in vivo situation than those observed on basal, unstimulated release, because depolarization is the normal way synapses are activated. The release of GABA from synaptosomes is classically mediated by both exocytotic mechanisms, which depend on depolarization and calcium entry, and reversal of the GABA carrier, which is independent from the presence of external Calcium²+ and is driven by the sodium gradient. [32–34] KCl depolarization is able to trigger both mechanisms on the synaptosome preparation. However, it has been shown that the exocytotic release of sup 3 H-GABA is more pronounced than the carrier-mediated process in synaptosomes from the adult rat brain. [35] In agreement with these findings, we found that the KCl-evoked release of³H-GABA was completely blocked in the absence of external calcium. Second, the magnitude of the KCl-evoked³H-GABA peaks reported here are in good agreement with previous studies. [19] Third, anesthetics failed to affect depolarization-evoked GABA release, regardless whether the specific high-affinity uptake process was blocked by nipecotic acid. These results strongly support that the release process investigated there was primarily the exocytotic one. [19] Therefore, anesthetics seem not to interfere with exocytotic release of GABA from preloaded striatal synaptosomes in vitro. Halothane has been shown to reduce both norepinephrine and dopamine release elicited by KCl stimulation from rat brain synaptosomes in vitro. [6,9] In contrast, this agent was ineffective in modulating acetylcholine release from the same preparation. [6] It can be speculated that differences are present between the sensitivity to anesthetics of voltage-operated Calcium²+ channels involved in the release of neurotransmitters from distinct synaptosome subpopulations. [36] Alternatively, it has been shown that the release of amino acids from isolated nerve terminals requires higher elevations in the Calcium²+ concentration in the bulk cytoplasm than that of catecholamines. [37] This might contribute to the differences observed in the effects of halothane, enflurane, and isoflurane on GABA versus dopamine release in the rat striatum. More generally, it can be suggested that volatile anesthetics will interfere selectively with stimulus-secretion coupling in specific neuronal phenotypes.

In summary, we have shown that intravenous but not volatile agents interfere with the specific high-affinity GABA uptake process in a synaptosome preparation in vitro. However, the similarities between the concentrations required to modulate³H-GABA uptake in vitro and those associated with propofol and etomidate anesthesia in vivo suggest that synaptic GABA accumulation may occur during anesthesia induced by both agents. Whether these findings are of potential clinical relevance cannot be extrapolated from our experimental data. It has been demonstrated that only events that are elicited by large stimuli or corresponding to paroxysmal activity are enhanced by blocking GABA uptake. [38,39] Whether these patterns of activities occur during anesthesia remains speculative. However, it can be suggested that a major consequence of inhibiting GABA uptake is to increase the background levels of GABA-activating extrasynaptic GABA^Aor even GABA^B-mediated synaptic transmission. [38] Thus, some actions at selective presynaptic CNS target sites might contribute to the mechanisms by which propofol or etomidate anesthesia is produced.

The authors thank Dr. Denis Herve, for his assistance in the computerized analysis of uptake data.