Previous electrophysiologic studies have implicated voltage-dependent Na+ channels as a molecular site of action for propofol. This study considered the effects of propofol on Na+ channel-mediated Na+ influx and neurotransmitter release in rat brain synaptosomes (isolated presynaptic nerve terminals).
Purified cerebrocortical synaptosomes from adult rats were used to determine the effects of propofol on Na+ influx through voltage-dependent Na+ channels (measured using 22Na+) and intracellular [Na+] (measured by ion-specific spectrofluorimetry). For comparison, the effects of propofol on synaptosomal glutamate release evoked by 4-aminopyridine (Na+ channel dependent), veratridine (Na+ channel dependent), KCi (Na+ channel independent) were studied using enzyme-coupled fluorimetry.
Propofol inhibited veratridine-evoked 22Na+ influx (inhibitory concentration of 50% [IC50] = 46 microM; 8.9 microM free) and changes in intracellular [Na+] (IC50 = 13 microM; 6.3 microM free) in synaptosomes in a dose-dependent manner. Propofol also inhibited 4-aminopyridine-evoked (IC50 = 39 microM; 19 microM free) and veratridine (20 microM)-evoked (IC50 = 30 microM; 14 microM free), but not KCi-evoked (up to 100 microM) glutamate release from synaptosomes.
Inhibition of Na+ channel-mediated Na+ influx, increased in intracellular [Na+], and glutamate release occurred in synaptosomes at concentrations of propofol achieved clinically. These results support a role for neuronal voltage-dependent Na+ channels as a molecular target for presynaptic general anesthetic effects.
The molecular mechanisms by which general anesthetics act are not completely understood. Several neuronal ion channels and neurotransmitter receptors have been implicated as targets for the depressant effects of anesthetics on synaptic transmission. Propofol (2,6-diisopropylphenol) is widely used as an intravenous agent to induce and maintain anesthesia. It has been shown to enhance gamma-aminobutyric acid-mediated synaptic inhibition, [2,3]to block Ca (2+) entry into aortic smooth muscle cells and Ca2+currents in ventricular myocytes, [4,5]and to block isolated voltage-dependent Na (+) channels (Na+channels). Presynaptic mechanisms of propofol action, such as effects on synaptic neurotransmitter release, have not been characterized.
Na+channels are integral membrane proteins that are involved in the generation and propagation of action potentials in excitable cells. Small fractional changes in Na+channel activation can have substantial effects on cellular functions. Synaptosomes, a subcellular fraction consisting of pinched-off nerve terminals, provide a useful system for analyzing the biochemical pharmacology of neurotransmitter release. They are largely devoid of functional glial and nerve cell body elements so that intercellular interactions are avoided. Na+channels have been demonstrated in synaptosomes by membrane-potential measurements, ion flux studies, and neurotoxin binding. [9–12]In the present study, we used synaptosomes isolated from rat cerebral cortex to study the effects of propofol on Na+uptake and intracellular Na+concentrations in the presence of Na+channel activators and blockers. The neurophysiologic significance of these effects was analyzed by measuring the effects of propofol on Na+channel-dependent glutamate release from synaptosomes. Glutamate, the major excitatory neurotransmitter in the central nervous system, is stored in and released from small synaptic vesicles present in cerebrocortical synaptosomes. Our results provide biochemical evidence for an action of propofol on neuronal Na+channels in isolated presynaptic nerve endings and support a role for the Na+channel as a molecular target for inhibitory presynaptic effects of propofol on excitatory synaptic transmission.
Materials and Methods
Isolation of Synaptosomes from Rat Cerebral Cortex
Experiments were done in accord with the National Institutes of Health guidelines for the care and use of laboratory animals as approved by the Cornell University Medical College Institutional Animal Care and Use Committee. Synaptosomes were prepared using a modification of the procedure of Dunkley et al. Adult male Sprague-Dawley rats (weighing 150–175 g) were anesthetized with 80% carbon dioxide/20% oxygen and killed by decapitation. Brains were immediately removed and rinsed in ice-cold 0.32 M sucrose. The cortical gray matter was homogenized in 0.32 M sucrose (10 ml/g tissue) using a motor-driven polytetrafluorethylene-glass (Potter-Elvehjem) homogenizer (Thomas Scientific, Swedesboro, NJ) at 900 rpm for ten up-and-down strokes. The homogenate was centrifuged at 1,000g for 2 min. The supernatant fraction was collected and centrifuged at 15,000g for 12 min. The resulting pellet was resuspended in 0.32 M sucrose (8 ml/brain). Aliquots (2.5 ml) of this fraction were loaded onto discontinuous gradients consisting of three 2.5-ml layers of filtered Percoll density gradient medium (23%, 10%, and 3%) in 0.32 M sucrose plus 0.25 mM dithiothreitol and 1 mM ethylenediaminetetraacetic acid, pH 7.4. The gradients were centrifuged at 25,000g for 6.5 min. The synaptosomal fraction was collected from the 23%/10% Percoll interface and diluted about five times in medium (either high [120 mM] or low [5 mM] Na+containing buffer) that was equilibrated with 95% oxygen/5% carbon dioxide. The synaptosomes were centrifuged at 23,000g for 10 min and resuspended in the appropriate medium to remove Percoll. The protein concentration of the synaptosomal preparation was determined by the method of Bradford using bovine serum albumin as a standard.
Measurement of Na+Uptake
Synaptosomal Na+uptake was measured by a modification of the method of Tamkun and Catterall. Synaptosomes were suspended in low Na+medium (130 mM choline chloride, 5.4 mM KCl, 5 mM NaCl, 0.8 mM MgSO4, 5.5 mM D-glucose, 50 mM HEPES-Tris, pH 7.4), and aliquots of the synaptosomal fraction (approximately 0.6–0.7 mg protein in 150 micro liter) were preincubated at 37 degrees C for 5 min, with or without the indicated concentrations of propofol. Immediately after preincubation with propofol, veratridine (60 micro Meter) was added and the samples were incubated for an additional 10 min at 37 degrees C, after which 1.3 micro Ci carrier-free22NaCl in 50 micro liter medium was added. Uptake was terminated after 5 s by adding 3 ml ice-cold washing buffer (163 mM choline chloride, 0.8 mM MgSO4, 1.8 mM CaCl2, mM HEPES-Tris, pH 7.4), and the reaction mixture was rapidly filtered under vacuum through Whatman GF/C glass microfiber filters (Brandel, Gaithersburg, MD). The filters were washed twice with 3 ml washing buffer. Filter radioactivity was determined by liquid scintillation spectrometry in Bio-Safe NA (Research Products International, Mount Prospect, IL) scintillation cocktail. Nonspecific 22Na+uptake was determined in the presence of 1 micro Meter tetrodotoxin, and this value was subtracted from total uptake to yield specific Na+uptake. The uptake of Na+by synaptosomes was linear up to 10 s. Based on this linearity, a 5-s uptake period was used for the remaining experiments.
Intrasynaptosomal Na+Concentration Measurement ([Na (+)]i)
Na+concentration was determined by ion-specific spectrofluorimetry using a spectrofluorimeter (Perkin Elmer luminescence spectrometer LS-50B; Beaconsfield, UK) with continuous computer-assisted data acquisition. The cell permeant acetoxymethyl ester precursor form of Na+-binding benzofuran isophthalate (SBFI-AM) was used as the fluorescent indicator. When entering the cell, the ester is hydrolyzed by cytosolic esterases, which releases the active free acid of the indicator, SBFI. Because the fluorescence of SBFI increases at lower excitation wavelengths (340 nm) without changes at a higher excitation wavelength (380 nm) after Na+binding, the dual-wavelength fluorescence ratio method is suitable for measuring changes in [Na+](i).
Synaptosomes (5 mg protein) were suspended in 1 ml Na+-free medium (120 mM choline chloride, 5 mM KCl, 5.0 mM D-glucose, 0.8 mM MgSO4, 50 mM HEPES-Tris, pH 7.4) containing 10 micro Meter SBFI-AM and 4 micro liter of 25%(w/v) Pluronic F-127 (Molecular Probes Company, Eugene, OR)(a nonionic detergent that facilitates dye uptake ) and incubated for 2 h at room temperature. At the end of the loading period, synaptosomes were centrifuged at 5,000g, resuspended in dye-free medium, and centrifuged again at 5,000g to remove excess dye. The synaptosomes were suspended in Na+-free medium and incubated for an additional 30 min to allow dye hydrolysis. After incubation, aliquots of synaptosomes (0.5 mg protein) were centrifuged and the pellets were stored on ice until use. For free [Na+]idetermination, synaptosome pellets were resuspended in 1.5 ml of 120 mM Na+medium (same as Na (+)-free medium except NaCl replaces choline chloride) and incubated in a stirred quartz cuvette at 37 degrees C in the absence or presence of propofol for 5 min followed by the addition of 60 micro Meter veratridine to activate Na+channels. Synaptosomal [Na+]iwas calculated from the ratio of the fluorescence intensity of SBFI at an emission wavelength of 510 nm, with excitation wavelengths of 340 and 380 nm (switched every 2 s). The signal ratio was converted into free [Na+]ibased on the method of Grynkiewicz et al. Calibration of the 340:380 nm excitation ratio in terms of free [Na+]iwas performed for each synaptosome preparation. For calibration, SBFI-loaded synaptosomes were added to solutions of known extracellular [Na+] made by appropriate mixtures of high [Na+] and high [K+] solutions in the presence of 40 micro Meter monensin, 2 micro Meter gramicidin, and 100 micro Meter ouabain. The high [Na+] solution contained 120 mM NaCl, 2 mM EGTA, and 10 mM HEPES, pH 7.4. The high [K+] solution was identical except that K+replaced Na+. In control experiments, there was no quenching of SBFI fluorescence by either veratridine or propofol observed in the presence of monensin, a Na+ionophore (data not shown).
Measurement of Glutamate Release from Synaptosomes
Endogenous glutamate release was measured by the method of Nicholls et al., as described by Schlame and Hemmings. Briefly, synaptosome pellets (0.5 mg protein) were resuspended in 1.5 ml incubation medium (120 mM NaCl, 5 mM KCl, 1.2 mM Na2HPO4, 5 mM NaHCO3, 1 mM MgSO4, 20 mM N-tris (hydroxymethyl)-methyl-2-aminoethanesulfonic acid (pH 7.4 with NaOH), 10 mM D-glucose, 16 micro Meter bovine serum albumin (essentially free of fatty acid), 1 mM NADP+, 100 U L-glutamate dehydrogenase, +/- 1.3 mM CaCl2). Samples were placed in a cuvette in a spectrofluorimeter equipped with a magnetic stirrer and a temperature-regulated (37 degrees Celsius) cuvette holder. The stirred samples were equilibrated at 37 degrees C for 4 min and then data acquisition was started with an excitation wavelength at 340 nm and an emission wavelength at 510 nm. To determine Ca2+-independent glutamate release, synaptosomes were preincubated with 1.3 mM EGTA in Ca2+-free medium. After recording basal glutamate release, veratridine (20 micro Meter or 60 micro Meter), 30 mM KCl, or 1 mM 4-aminopyridine (4-AP) was added and the rate of glutamate release from 0–60 s, which was essentially linear with respect to time, was measured. The fluorescence signal was calibrated by adding 5 nmol Dglutamate to the cuvette at the end of each experiment.
Na+, K+-Adenosine Triphosphatase Assay
Na+, K+-adenosine triphosphatase (ATPase) was assayed as described previously. The assay mixture for total ATPase (1 ml volume) contained 30 mM Tris-HCl buffer (pH 7.4), 3 mM Tris-ATP, 3 mM MgCl2, 100 mM NaCl, 20 mM KCl, +/- 1 mM ouabain. In the assay mixture for Mn2+-ATPase, NaCl and KCl were omitted. The reaction mixtures containing an appropriate dilution of synaptosomes were preincubated at 37 degrees C for 10 min. The reactions were initiated by adding Tris-ATP and terminated after 15 min by adding 1 ml of 10%(w/v) trichloroacetic acid. Inorganic phosphate (Pi) was determined in a 0.5-ml portion of the supernatant after centrifugation at 5,000g for 10 min according to the method of Bonting et al. The activity of Na+, K+-ATPase was calculated as the difference between total activity and Mn2+ATPase activity, expressed as micro mol Piliberated [center dot] mg protein-1[center dot] h-1.
Analysis of the Ratio of Free to Bound Propofol
The nominal propofol concentrations indicated in the figures include both bound and free (total) propofol. However, the pharmacologically relevant drug concentration that determines drug efficacy is the free or unbound fraction. For each experimental condition, the total:free ratio of propofol was determined by equilibrium dialysis using a Spectrum equilibrium dialyzer equipped with Spectra/Por 2 dialysis membrane (molecular weight cutoff = 12,000)(Spectrum Medical Industries, Los Angeles, CA). Samples (identical to assay conditions) were dialyzed against corresponding buffers at pH 7.4 for 4 h at 37 degrees C. Concentrations of propofol on both sides of the membrane were analyzed by high-pressure liquid chromatography according to the method of Pavan et al. The total:free ratios of propofol (25–100 micro Meter) were 5.2:1 in the assays of22Na+flux (0.65 mg synaptosomal protein in a 0.2-ml reaction volume) and 2.1:1 in the assays of [Na+]iand glutamate release (0.5 mg synaptosomal protein and 15 micro gram bovine serum albumin in a 1.5-ml reaction volume). These ratios were used to calculate the inhibitory concentration of 50%(IC50) values for free propofol.
Oxidized cicotinamide adenine dinucleotide phosphate, L-glutamate, L-glutamate dehydrogenase (Proteus sp.), bovine serum albumin (essentially fatty acid free), tetrodotoxin, veratridine, dimethylsulfoxide, monensin, gramicidin D, ouabain, 4-AP, and ATP were obtained from Sigma Chemical Company (St. Louis, MO). Percoll density gradient medium was obtained from Pharmacia (Uppsala, Sweden). [(22) NaCl](1,000 micro Ci/ml) was obtained from DuPont-New England Nuclear (Boston, MA). Benzofuran isophthalate and Pluronic F-127 were purchased from Molecular Probes Company (Eugene, OR). Propofol was obtained from Aldrich Chemicals (Milwaukee, WI) or was a gift from Zeneca Pharmaceuticals (Macclesfield, Cheshire, UK). All other chemicals were obtained from commercial sources and were of analytical grade. Dimethylsulfoxide at a final concentration of 0.05%(v/v) was used as a vehicle for propofol and veratridine. Control experiments showed that the vehicle alone had no effects on the activities measured.
Statistical differences between control and experimental values were determined by analysis of variance with the Fisher post hoc test. Concentration-effect data were analyzed using a graded dose-response program that performs linear regression analysis on data between 20% and 80% of the maximal response (Pharm/PCS pharmacologic calculation system, version 4.2; Springer-Verlag, New York, NY). The confidence limits (in parentheses) follow the derived IC50values in the text. Data for all curves were fit with simple polynomial or linear functions to obtain the best fit of the available data using commercially available software (Origin; Microcal Software Inc., Northhampton, MA).
Effect of Propofol on Synaptosomal Na+Uptake
Veratridine, an alkaloid neurotoxin that activates Na+channels by binding to receptor site 2 (see Catterall ) was used to induce22Na+uptake in synaptosomes.22Na+uptake increased with veratridine concentrations up to 60 micro Meter (Figure 1), where it reached a plateau. Basal Na+uptake (i.e., in the absence of Na+channel activators) was 26 +/- 5 nmol [center dot] min-1[center dot] mg-1. Veratridine (60 micro Meter) induced a fourfold increase in synaptosomal Na+uptake (98 +/- 1 nmol [center dot] min-1[center dot] mg-1). Veratridine-stimulated Na+uptake was completely blocked by 1 micro Meter tetrodotoxin, a Na+channel blocker that binds to receptor site 1 (see Catterall ), which indicates that the veratridine-stimulated22Na+uptake was mediated by Na+channels.
Preincubation of synaptosomes with propofol in vitro caused a dose-dependent inhibition of veratridine-stimulated22Na+uptake (Figure 2) with a nominal IC50value of 46 micro Meter (34–62 micro Meter), which corresponds to a free propofol concentration of 8.9 micro Meter. The data shown in Figure 2are reported as specific uptake corrected by subtracting nonspecific uptake in the presence of 1 micro Meter tetrodotoxin. The veratridine-independent (basal or resting) uptake of22Na+was unaffected by preincubation with 1 micro Meter tetrodotoxin or propofol (data not shown).
Effect of Propofol on Free [Na+]i
The resting free [Na+]iin synaptosomes increased 2.7 times in going from Na+-free buffer (4 +/- O.1 mM) to 120 mM Na+-containing buffer (11 +/- 0.6 mM). Adding veratridine (60 micro Meter) increased free [Na+]ito 56 +/- 6 mM in 120 mM Na+-containing buffer. This veratridine-induced increase in synaptosomal [Na (+)]iwas completely blocked by 1 micro Meter tetrodotoxin: In the presence of tetrodotoxin and veratridine, the [Na+]iwas 12 +/- 2 mM, as was resting [Na+]i. Preincubation of synaptosomes with propofol inhibited the veratridine-induced increase in free [Na+]iin a concentration-dependent manner (Figure 3), with a nominal IC50value of 13 micro Meter (range, 9–18 micro Meter), which corresponds to a free propofol concentration of 6.3 micro Meter. Propofol did not significantly affect basal [Na+]i(Figure 3) except at 5 and 10 micro Meter concentrations, which resulted in a small but significant increase in [Na+]i. Similar results were obtained in experiments conducted at 28 degrees C (data not shown). Propofol applied after veratridine was added also significantly antagonized the evoked increase in [Na+]irapidly (within 60 s) and to a similar extent as in the preincubation studies (Table 1). Depolarization of synaptosomes by either 4-AP (1 mM) or KCl (30 mM) did not significantly increase [Na+]ialone (12 +/- 1.3 mM and 12 +/- 0.7 micro Meter, respectively) or in the presence of propofol (50 mM)(14 +/- 1.4 mM and 12 +/- 0.9 mM, respectively).
Effect of Propofol on Glutamate Release Evoked by 4-Aminopyridine and KCl
Glutamate release evoked by 4-AP and KCl in synaptosomes was measured to assess the involvement of Na+channels as a target site for propofol inhibition. 4-Aminopyridine, a K+channel blocker, destabilizes the membrane potential and causes repetitive spontaneous Na (+) channel-dependent depolarizations that closely approximate in vivo depolarizations of the synaptic terminal and lead to activation of voltage-dependent Ca2+channels and neurotransmitter release. Elevated extracellular KCl concentrations depolarize the plasma membrane by shifting the K+equilibrium potential above the threshold potential for activation of voltage-dependent Ca2+channels (Ca2+channels), which leads to Ca2+entry and neurotransmitter release, while Na+channels are inactivated. Propofol inhibited 4-AP (1 mM)-evoked glutamate release in a dose-dependent manner (Figure 4) with a nominal IC50value of 39 micro Meter (18–87 micro Meter); in contrast, KCl (30 mM)-evoked glutamate release was not significantly ‘affected by concentrations of propofol as high as 100 micro Meter (Figure 4). The IC50of free propofol to inhibit 4-AP-evoked release was 19 micro Meter. These results indicate greater sensitivity to the action of propofol of Na+channels compared with Ca2+channels in the release of glutamate, because KCl-evoked glutamate release involves Ca2+channels exclusively and 4-AP-evoked release involves both Na+and Ca2+channel activation.
Effect of Propofol on Veratridine-evoked Glutamate Release from Synaptosomes
The role of Na+channels as a target for the inhibitory effect of propofol on glutamate release was analyzed further by testing the effects of propofol on veratridine-evoked glutamate release. Veratridine activates Na+channels directly and thereby leads to plasma membrane depolarization, Ca+entry, and neurotransmitter release. We studied the effects of propofol on glutamate release evoked by 20 micro Meter or 60 micro meter veratridine, because submaximal concentrations of veratridine (with respect to stimulation of Na+-influx) result in substantial increases in the release of neurotransmitters. Veratridine (20 micro Meter or 60 micro Meter) significantly elevated glutamate release in either the absence or presence of added Ca2+(1.3 mM). The amount of glutamate released by 20 micro Meter veratridine in the presence of Ca2+(3.6 +/- 0.3 nmol [center dot] min-1[center dot] mg-1) was six times higher than that released in the absence of Ca2+(0.61 +/- 0.04 nmol [center dot] min-1[center dot] mg-1). Similarly, the amount of glutamate released by 60 micro Meter veratridine in the presence of Ca2+(4.3 +/- 0.4 nmol [center dot] min (-1)[center dot] mg -1) was four times higher than the release in the absence of Ca2+(1.0 +/- 0.2 nmol [center dot] min -1 [center dot] mg-1). The major component of veratridine-evoked glutamate release was Ca (2+) dependent, although 60 micro Meter veratridine resulted in more Ca (2+)-independent release. Propofol inhibited both Ca2+-dependent and Ca2+-independent glutamate release evoked by veratridine in a dose-dependent manner (Figure 5, Figure 6, Figure 7). Figure 5shows typical glutamate-release curves obtained with veratridine (20 micro Meter) in the presence of Ca2+and increasing concentrations of propofol. The nominal IC50value for propofol inhibition of Ca2+-dependent glutamate release was 30 micro Meter (range, 8–106 micro Meter) for 20 micro Meter veratridine and was 50 micro Meter (range, 29–92 micro Meter) for 60 micro Meter veratridine (Figure 6). These values correspond to free propofol IC50values of 14 micro Meter and 24 micro Meter, respectively. Tetrodotoxin inhibited glutamate release evoked by veratridine in the presence and in the absence Ca2+(data not shown).
The correlations between the percentage inhibition by various concentrations of propofol of Ca2+-dependent glutamate release evoked by 20 micro Meter or 60 micro Meter veratridine and the percentage inhibition of22Na+uptake or of changes in [Na+]iwere essentially linear with comparable slopes. The statistical significance of the correlation was greater between changes in22Na (+) uptake or [Na+]iand changes in glutamate release evoked by 20 micro Meter veratridine (r = 0.932/P < 0.001 and r = 0.944/P <0.001, respectively) than by 60 micro Meter (r = 0.72/P < 0.001 and r = 0.67/P < 0.001, respectively).
Effect of Propofol Na+, K+-ATPase
(Table 2) shows the effect of propofol on Na+, K+-ATPase activity in rat cortical synaptosomes. No significant changes in activity were observed in the presence of 25 or 100 micro Meter total propofol.
Inhibition of glutamate release from synaptosomes by clinically relevant concentrations of volatile anesthetics was reported recently. [20,26]The greater sensitivity to volatile anesthetics of glutamate release evoked by veratridine or 4-aminopyridine compared with KCl implicated a site of anesthetic action proximal to Ca2+influx, probably Na+channels. This hypothesis is supported by independent studies that showed effects of various local, intravenous, and volatile anesthetics on Na+channel function. [6,27–30]In the present study, propofol significantly inhibited voltage-dependent Na+channels in rat cerebrocortical synaptosomes by blocking veratridine-evoked increases in22Na+influx; this action is supported by its effects on [Na+]iand glutamate release. These data provide functional support obtained in a nerve terminal preparation for electrophysiologic studies of single Na+channels from human cerebral cortex in which propofol depressed Na+channel function by reducing fractional open time. 
Our value for veratridine-induced22Na + influx into synaptosomes corresponds closely with published values of 65 (see Tamkun and Catterall ) and 55 nmol [center dot] min-1[center dot] mg-1(see Mullin and Hunt ). The Na+flux experiments were done in the presence of 5 mM Na+to allow measurable uptake rates. Because Na (+) uptake measured under these conditions is approximately proportional to Na+channel permeability and to the number of open channels, the reduction of Na+uptake by propofol reflects reduced Na+channel opening, Na+channel permeability to Na+, or both.
The inhibitory effect of propofol on Na+influx was confirmed by measuring its effect on changes in synaptosomal [Na+](i) levels using the fluorescence dye technique, as developed by Minta and Tsien and adapted to synaptosomes by Cousin et al. The resting [Na+]iin synaptosomes increased from 4 mM (in Na+free medium) to 11 mM (in 120 mM Na+medium), indicating that resting [Na+]iwas dependent on external [Na+]. This value for resting [Na+]i(11 + 0.6 mM) corresponds with values determined using the same method in guinea pig cortical synaptosomes (10.9 +/- 1.8 mM)and in rat cortical synaptosomes (11.3 +/- 0.9 mM). Inhibition of veratridine-evoked increases in synaptosomal [Na+]i, either by preapplication or postapplication of propofol, supports a specific action on Na+channels. The lack of effect of KCl on [Na+]iis consistent with earlier reports [15,33]and is probably the result of rapid Na+channel inactivation due to the fixed membrane depolarization produced by high K+. Similarly, 4-AP did not significantly affect [Na+]i. The repetitive nature of the stimulation by 4-AP results in a fraction of Na+channels that open and rapidly inactivate. This transient depolarization, which occurs in only a fraction of the total synaptosomes at any given time, may be insufficient to induce a measurable change in SBFI fluorescence.
Na+, K+-ATPase plays a major role in maintaining the transmembrane Na+electrochemical gradient. The activity of this enzyme was determined in the presence of propofol to rule out a possible role in the observed effects on [Na+]i. Propofol had no effect on synaptosomal Na+, K+-ATPase activity (Table 2), which further supports a direct action of propofol on Na+channels.
The functional significance of the inhibition by propofol of Na (+) influx was studied by comparing the effects of propofol on the release of the excitatory neurotransmitter glutamate from synaptosomes. Propofol did not inhibit glutamate release evoked by direct depolarization with KCl, whereas it significantly inhibited glutamate release evoked by 4-AP. Because 4-AP-induced depolarization involves the action of both Na (+) channels and Ca2+channels, whereas KCl-induced depolarization involves only Ca2+channels (i.e., is tetrodotoxin insensitive), Na (+) channels are apparently more sensitive to propofol inhibition than are the particular Ca2+channels coupled to glutamate release. This observation is consistent with the results of Bickler et al., in which propofol (1–100 micro Meter) did not affect KCl-evoked glutamate release from rat brain sections. Furthermore, selective inhibition of Na (+) currents and not of Ca2+currents by propofol (10–20 micro Meter) has been demonstrated in rat olfactory cortical sections. 
To further assess the hypothesis that Na+channels are a target for propofol in the presynaptic nerve terminal, we studied the effect of propofol on veratridine-evoked glutamate release. Activation of Na+channels by veratridine induces both Ca2+-dependent vesicular glutamate release and Ca2+-independent release of cytoplasmic glutamate due to reversal of Na+-dependent glutamate reuptake as a result of a collapse of the transmembrane [Na+] gradient. The effects of propofol on Ca2+-independent and Ca (2+)-dependent glutamate release from synaptosomes were examined separately using 20 micro Meter or 60 micro Meter veratridine. Propofol inhibited both Ca2+-dependent and Ca2+-independent veratridine-evoked glutamate release in a concentration-dependent manner. Because of its important contribution, inhibition of Ca2+-dependent glutamate release played the major role in the action of propofol. Both Ca (2+)-dependent and Ca2+-independent release of glutamate in the presence of veratridine were blocked completely by 1 micro Meter tetrodotoxin, which indicates that Na+channels are involved in both components of glutamate release. Propofol (100 micro Meter total) inhibited Ca2+-dependent glutamate release evoked by 60 micro Meter veratridine by 61%, and by 20 micro Meter veratridine by 84%(Figure 6). Reduced inhibition with increased stimulus intensity has been observed previously for inhibition by benzotropine and atropine of veratridine-evoked gamma-aminobutyric acid and glutamate release, and for inhibition by omega-agatoxin-IVA and omega-conotoxin GVIA of neurotransmitter release. Higher concentrations of veratridine may indirectly stimulate other mechanisms of Ca2+entry, which are less sensitive to inhibition by propofol. For example, the IC50for inhibition of Ca2+uptake in synaptosomes by phenytoin, an anticonvulsant, increased with increasing concentrations of veratridine due to an allosteric competitive interaction between phenytoin and veratridine. A direct competitive interaction between propofol and veratridine is unlikely given the observed inhibition by propofol of 4-AP-evoked glutamate release (which does not involve toxin-modified Na (+) channels) but cannot be ruled out completely based on the available data.
Several mechanisms for Ca2+entry coupled to glutamate release have been demonstrated in rat cortical synaptosomes : through Na+channels, through Ca2+channels, or by the electrogenic Na+-Ca2+exchange system. The increase in [Na+]iinduced by veratridine can lead to Ca2+influx by reversing the Na+-Ca2+exchanger, which is sensitive to tetrodotoxin. Ca (2+) entry through this exchange system contributes to dopamine release from synaptosomes. We hypothesize that propofol inhibits Ca2+-dependent veratridine-evoked glutamate release by inhibiting Na+channels, which prevents Na+channel-dependent membrane depolarization and consequent Ca2+channel activation or Ca2+entry through Na+channels, and indirectly inhibits Na+-Ca2+exchange. Ca2+-independent release is inhibited by an action at the Na+channel to reduce Na+influx and consequent reversal of the Na+/glutamatetransporter, or by a direct inhibition of the transporter acting in reverse. The fact that basal levels of [Na (+)]iand the spontaneous release of glutamate were unaffected by propofol in this study suggests that it has no direct effect on the Na+-glutamate transporter. Stimulation of glutamate reuptake by propofol is also unlikely as an explanation for our data because propofol has been shown to inhibit [(3) H]glutamate uptake, but at high concentrations (IC (50) > 100 micro Meter). 
Results of studies of the role of intracellular Ca2+stores in veratridine-evoked release of neurotransmitters in Ca2+-free medium are conflicting. Veratridine-evoked endogenous glutamate release from cerebellar sections and veratrine-evoked [(3) H]gamma-aminobutyric acid release from mouse cortical synaptosomes were not affected by ruthenium red (a mitochondrial Ca2+uptake inhibitor). In contrast, ruthenium red (10 micro Meter) potentiated veratrine-evoked Ca2+-independent [(3) H]-noradrenaline release from rat neocortical sections. The effects of propofol on mobilization of intracellular Ca2+stores are also equivocal. Propofol inhibited Ca2+release from mitochondria isolated from rat heart or liver, but did not affect thapsigargin-induced release of intracellular Ca2+in rat aortic smooth muscle cells. Further studies are needed to clarify the role of modulation of intracellular Ca2+stores in effects of anesthetics on neurotransmitter release.
Activation of Na+channels induces Na+influx, membrane depolarization, and Ca2+influx, which is a prerequisite for vesicular glutamate release. The Ca2+coupled to glutamate release can enter either through Ca2+channels, Na+channels, or reversal of the Na+-Ca2+exchange. The results described herein suggest that propofol inhibits the release of the excitatory neurotransmitter glutamate (IC50= 19 micro Meter in the presence of 1 mM 4-AP; IC50= 14 micro Meter in the presence of 20 micro Meter veratridine) by blocking Na+influx through Na+channels (IC50= 8.9 micro Meter at 5 mM [Na+]e; IC50= 6.3 micro Meter at 120 mM [Na+]e;Table 3). To our knowledge, this is the first report to show that propofol inhibits Na+influx through endogenous Na+channels in a semi-intact preparation of nerve terminals. Propofol reduced the open time of batrachotoxin-modified Na+channels in planar lipid bilayers with an EC50of 20 micro Meter. This effect is comparable in potency to the inhibition of veratridine-modified Na+channel function that we observed in this study.
Amorim et al. found that propofol (20 micro gram/ml; 111 micro Meter) inhibited the increase in Na+content in hippocampal brain sections associated with anoxia. A direct action of propofol on Na (+) entry was proposed, although the mechanism was not determined. Our results support this interpretation and suggest that propofol may have potential for use as a neuroprotective agent to limit Na+influx and glutamate release in cerebral ischemia. [51,52]
Correlation of the in vitro effects of propofol with its clinically relevant concentrations is an important, but not straightforward, consideration. Estimating appropriate concentrations of many intravenous anesthetics is complicated by their lipid solubility, redistribution, metabolism, and protein binding, and by the technically difficult feat of measuring their effect-site concentrations during anesthesia. Previous determinations of the propofol EC50for general anesthesia in humans were based on studies that used coadministration of opioids, variable anesthetic end points, or bolus injections, which complicate interpretation of these data. The whole-blood EC50and EC95values of propofol alone to suppress the response to skin incision (which corresponds to the minimum alveolar concentration of volatile anesthetics) in humans during continuous infusion were 85 micro Meter and 153 micro Meter, respectively. A recent study suggested that the actual brain concentration of propofol as it approaches equilibrium after 30 min of intravenous infusion is considerably higher, with a brain: plasma concentration ratio of 7.8–8.5. Estimation of the corresponding free propofol concentrations, accounting for approximate values for plasma protein binding of 98% and for the plasma:whole blood concentration ratio of 0.78 (see Servin et al. ), yields EC50, and EC95values of 1.3 micro Meter and 2.4 micro Meter, respectively. In tadpoles, an EC50value of 2.2 micro Meter was reported for loss of righting reflex, an end point that usually occurs at a lower concentration than surgical anesthesia. It is difficult to compare anesthetic EC50values, which are derived from quantal concentration-response relations, with IC50values obtained in vitro that examine a continuous response (variable). We observed significant reductions in22Na+flux (39%), [Na+](i)(34%), and veratridine-evoked glutamate release (20%) at free propofol concentrations of 4.8 micro Meter, 2.4 micro Meter, and 4.8 micro Meter, respectively. Because small reductions in Na+conductance can result in substantial changes in Na+channel-dependent events, these concentrations of propofol, which are relevant to those achieved clinically, may have physiologically significant effects on neuronal function and synaptic transmission.
Considerable data suggest that general anesthesia results from agent- and site-selective mechanisms and is not the result of a unitary mechanism of action. Inhibition of the release of glutamate, the major excitatory neurotransmitter in the mammalian central nervous system, may be involved in the generalized neuronal depression and antinociception characteristic of general anesthesia. The pharmacologic effects of propofol apparently result from various combinations of glutamate release inhibition (data from our study), gamma-aminobutyric acidAreceptor potentiation, [2,3]and/or Na+channel inhibition (and data from our study). Evidence implicating Na+channels as targets for general anesthetic effects is supported by the recent observation that central nervous system Na+channels are blocked by clinical concentrations of volatile anesthetics at physiologic resting membrane potential. Our data provide further support for a role of Na+channels as a molecular target of general anesthetics [60–62]and suggest that some of the effects of propofol, and probably volatile anesthetics, may result from the presynaptic inhibition of Na+channel-dependent glutamate release.
The authors thank Anna Adamo for technical assistance.