Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Ethanol at high concentrations acts as a general anesthetic agent, preventing movement in response to a noxious stimulus. Prevention of nocifensive movement is the most common endpoint for comparing potencies among volatile anesthetic agents (Fang et al., 1997a,b). For such agents the anesthetic concentration at this endpoint is MAC (minimum alveolar anesthetic concentration), which prevents movement in response to a noxious stimulus (Eger et al., 1965). MAC is determined by anesthetic actions in the spinal cord (Rampil et al., 1993; Antognini and Schwartz, 1993; Rampil, 1994; Antognini, 1997). Thus, anesthetic actions on spinal cord are directly relevant to general anesthesia.

We have shown previously that ethanol depresses synaptic transmission to motor neurons in intact spinal cord in vitro (Wong et al., 1997). However, the previous studies could not discriminate between postsynaptic depression of responses to transmitter and presynaptic depression of transmitter release. Moreover, postsynaptic actions might be mediated via enhancement of inhibition rather than depression of response to excitatory transmitter. Both volatile anesthetic agents and ethanol enhance currents at both glycine and -aminobutyric acid_A (GABA_A) receptors (Jones et al., 1992; Lin et al., 1992; Mihic et al., 1994, 1997). Enhancement of GABA_A inhibition is considered to be an important common factor in general anesthesia produced by a variety of agents (Franks and Lieb, 1993). The present studies were designed to test the following hypotheses: 1) that ethanol acts postsynaptically on motor neurons to depress synaptic transmission; 2) that ethanol directly depresses glutamate-evoked responses independent of actions on inhibitory chloride channels; and 3) that both -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) glutamate currents are sensitive to ethanol.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Experiments were carried out in spinal cord slices from Sprague-Dawley rats 14 to 23 days old; most were in the range 14 to 20 days. In a protocol approved by Stanford's panel on laboratory animal care and use, the animals were anesthetized with halothane, decapitated, and spinal cords quickly removed and placed in a protective chilled oxygenated calcium-free low sodium artificial cerebrospinal fluid (ACSF). The ACSF was composed as follows (mM): KCl 5, MgSO₄ 2, NaHCO₃ 26, NaH₂PO₄ 1.25, d-glucose 10, and sucrose 252. Slices 400 µm thick were prepared as described previously (Wang and Dun, 1990). Briefly, slices were sectioned from the lumbar region on a Vibratome (Technical Products International, St. Louis, MO), and removed to an oxygenated ACSF of the following composition (mM): NaCl 123, KCl 4, NaH₂PO₄ 1.2, MgSO₄ 1.3, NaHCO₃ 26, d-glucose 10, CaCl₂ 2. In this solution the slices were allowed to recover at room temperature for 0.5 to 1 h. Individual slices were transferred to a chamber constantly superfused with oxygenated ACSF of the same composition as the recovery solution. All experiments were carried out at room temperature.

Cell bodies beneath the cut surface of the slice were viewed on a closed circuit TV monitor using infrared illumination and a 40X water immersion objective. In preliminary studies the large cell bodies in the ventral horn, most commonly seen in the ventrolateral area, were identified as motor neurons by fluorescent labeling with Evans blue dye injected into the hind limb of the rat the day before sacrifice. Once this identity was established, fluorescent labeling was not used in the pharmacological studies because of concerns that the dye might alter glutamate receptor properties (Price and Raymond, 1996).

Patch pipettes were pulled on a Flaming-Brown pipette puller (Sutter Instrument Co., Novato, CA) and filled with a solution of the following composition (mM): NaCl 15, K gluconate 110, HEPES 10, MgCl₂ 2, EGTA 11, CaCl₂ 1, MgATP 2, pH adjusted with KOH to 7.3. Pipettes typically had a tip resistance of 3 to 5 M. The patch pipette was directed toward a motor neuron cell body under visual control. After establishment of a Gigohm seal the patch was ruptured by brief negative pressure and subsequent measurements made in the whole cell ruptured patch configuration in either current-clamp or voltage-clamp mode using an Axopatch 1D patch clamp amplifier (Axon Instruments, Inc., Burlingame, CA). Motor neuron responses were evoked by electrical stimulation of the dorsal root fragment or the dorsal root entry zone via a concentric bipolar platinum electrode with tip diameter 0.025 mm (Frederick Haer & Co., Brunswick, ME) using square wave stimuli 0.1 ms in duration, 1 to 20 V nominal intensity, frequency 0.03 to 0.1 s¹ (Wang and Dun, 1990). Excitatory postsynaptic potentials (EPSPs) or currents (EPSCs) were averaged in groups of 5 to 10. In addition to synaptic currents evoked by dorsal root stimulation, responses were evoked by direct pressure application of glutamate from a pipette positioned at the surface of the slice near the cell (Picospritzer, General Valve Corporation, Fairfield, NJ). Pressure pulses were 200 kPa, 10 to 400 ms in duration; glutamate concentration in the pipette was 100 mM. In each experiment duration of pressure to the glutamate-containing pipette was adjusted to give a reproducible inward current of good amplitude. Glutamate applications were separated by a 3-min interval to minimize receptor desensitization and were not averaged. In voltage-clamp studies, holding potential was usually 70 mV. Pharmacologic agents [tetrodotoxin (TTX), bicuculline methiodide, strychnine hydrochloride, 6-cyano-7-nitroquinoxaline-2,3-dione disodium (CNQX), D,L-2-amino-5-phosponopentanoic acid (AP-5), and ethanol] were made up as stock solutions, dissolved in ACSF at the desired concentration, and applied in the superfusate. Ethanol effects were measured 15 to 20 min after application.

Ethanol was obtained from commercial sources (Gold Seal Chemical Co., Hayward, CA) as the 95% compound and diluted into the ACSF to the desired concentration. Ethanol from another source (Grain Processing Corporation, Muscatine, IA) was used in a few experiments and gave similar results. Ethanol was applied by superfusion in the bath by gravity feed from a container similar to that containing the control ACSF. Superfusion rates were maintained constant between control and alcohol-containing solutions, and sham experiments in which both containers held control solutions showed that switching from one to the other container had no effect.

A commercially available software package (pClamp, Axon Instruments) was used to acquire data, which were digitally stored and analyzed off-line. Experiments were carried out on a single cell in each slice. Ethanol effects were expressed as mean ± S.E.M. and statistical significance was tested with Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Ethanol Effects on Synaptically Evoked Potentials and Currents. To examine the effects of ethanol on synaptically evoked potentials and currents a stimulus intensity was chosen that evoked a stable response of amplitude sufficient to measure but below the threshold for impulse initiation. Stimuli to the dorsal root entry area were given at a constant frequency of 0.03 s¹ during the control period, ethanol application, and washout. EPSPs were recorded from eight cells exposed to concentrations of ethanol between 12.5 and 200 mM. Two cells were exposed to a single ethanol concentration (130 mM) for 15 min. EPSCs were also recorded from these cells. The remaining six cells were exposed to cumulatively increasing ethanol concentrations with 15-min exposure at each concentration. Examples of ethanol depression of EPSPs and EPSCs at 130 mM are shown in Fig. 1, A and B. In the two cells exposed to a single concentration of 130 mM, peak response amplitudes were decreased to 23 and 55% of the control values, and the area under the curve of the response was depressed to 16 and 84% of control. The cumulative dose-response curve for area of the EPSP in six cells is shown in Fig. 1C. Ethanol significantly (P < .05) depressed both EPSP amplitude and area at concentrations 25 mM and higher.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Effects of ethanol on potentials and currents evoked by dorsal root stimulation. Sample records (A, B) are from the same cell in a slice taken from a 15-day-old animal. Resting potential in current clamp (A) was 63 mV, holding potential in voltage clamp (B) was 70 mV. A stimulus of a constant intensity was given to the dorsal root at the times indicated by the arrows. A, ethanol (130 mM) reversibly depressed the EPSP at 20 min after application; the effect was fully reversible on washing for 30 min. B, ethanol (130 mM) reversibly depressed the inward currents evoked by dorsal root stimulation after 20-min application, with recovery after 35-min washing. Records are averages of five responses. C, cumulative dose-response curve for ethanol actions on EPSP area. Ethanol actions were measured 15 min after application at each concentration. Data points are means of six cells, each from a different slice; error bars are S.E.M. Ethanol significantly depressed EPSP area at concentrations 25 mM and higher; depression at 12.5 mM was not quite significant.

Evoked Responses: Glutamate Application. Pulses of pressure applied to the glutamate-containing pipette produced inward currents in the motor neurons whose size increased with pulse duration and thus with amount of glutamate ejected (Fig. 2A). In the presence of 0.3 µM TTX glutamate-evoked currents were reduced and wave form shortened and simplified, indicating that some glutamate-induced inward current in the absence of TTX is due to impulse activity in neurons presynaptic to the motor neuron (Fig. 2A). Ethanol reversibly depressed glutamate-induced inward currents in the presence of TTX at all pulse durations (Figs. 2 and 3A). Cumulative dose-response curves to ethanol application were carried out in four cells exposed to ethanol for 15 to 20 min at each concentration without TTX (Fig. 3B) and at a single 100-mM ethanol concentration in another four cells in the presence of TTX for comparison (Fig. 3B). At 15 to 20 min after application, ethanol significantly depressed the response to glutamate at concentrations from 25 to 200 mM in the absence of TTX (Fig. 3B). The cumulative dose-response curve was not linear; the peak response was less depressed at 100 mM (three cells) or 50 mM (one cell) than at the preceding lower ethanol concentration. At 100 mM ethanol significantly (P < .05) depressed glutamate-evoked currents either in the presence or the absence of TTX (Fig. 3B), indicating a direct postsynaptic effect of ethanol on the motor neurons. TTX did not significantly attenuate the effect of ethanol.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Ethanol actions on currents evoked by different durations of glutamate pulses given at the time indicated by the arrows. A, under control conditions increasing the duration of the pressure pulse to the glutamate-containing pipette increases the duration of the response; the late components may be secondary to glutamate-induced activity in neurons presynaptic to the motor neuron. Downward sharp deflections are brief hyperpolarizing voltage steps to monitor input resistance. TTX reduces the amplitude and duration of the response. Ethanol (100 mM) reduces the amplitude of glutamate-evoked responses at all durations of glutamate application and affects both early and late components of the response. B, peak currents measured from the cell shown in (A), current amplitude plotted against duration of glutamate application in control, TTX, and ethanol plus TTX.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Ethanol effects on glutamate-evoked currents in motor neurons in a slice from a 17-day-old rat. A, inward currents were evoked by glutamate application at the time indicated by the arrow. Hyperpolarizing voltage steps as indicated in the top protocol trace were imposed to measure changes in membrane conductance. TTX (0.3 µM) was used to block action potentials in this cell. Holding potential was 70 mV. Ethanol, 100 mM applied for 15 min, reversibly depressed the inward current evoked by glutamate. There was a slight decrease in membrane conductance and a slight decrease in holding current in this cell after ethanol application. B, () show a cumulative dose-response curve to ethanol in four motor neurons in slices not treated with TTX. Error bars are S.E.M. Measurements were peak amplitude of the glutamate-evoked current (IGlu) normalized to control, at 15- to 20-min ethanol exposure at each concentration. For comparison, a single 100-mM ethanol application () was made to four cells in the presence of TTX. TTX did not significantly attenuate the effects of ethanol, confirming a postsynaptic effect directly on motor neurons.

Block of Inhibitory Chloride Channels. In the presence of TTX, bicuculline (50 µM) and strychnine (5 µM) were used to block GABA_A and glycine receptors, respectively. Ethanol depressed glutamate-evoked responses when either inhibitory receptor was blocked (Fig. 4, A and B) or when both were blocked together (Fig. 4C). In four cells treated with a combination of bicuculline and strychnine 100 mM ethanol significantly depressed peak glutamate-evoked current to 66% of control ± 8.6 (mean ± S.E.M., P < .05). This is not significantly different from the effect of ethanol in untreated preparations (Fig. 3).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Ethanol actions on glutamate-induced inward currents are not blocked by inhibitory chloride channel receptor antagonists. Experiments were done in the presence of TTX; Control and Wash are with the antagonist(s) present in the ACSF. Glutamate pulses were given at the time indicated by the arrows. Sharp deflections are brief hyperpolarizing voltage steps to monitor input resistance. A and B, ethanol 100 mM reversibly reduces the glutamate-evoked response in the presence of either the GABAA receptor antagonist bicuculline (50 µM) or the glycine receptor antagonist strychnine (5 µM). C, pronounced depressant effect of ethanol in the presence of both antagonists. Records in (A), (B), and (C) are each from a different cell.

Glutamate Receptor Subtypes. Selective antagonists for AMPA/kainate receptors (CNQX) and for NMDA receptors (AP-5) were used to determine which receptors mediated inward currents in motor neurons. The results are shown in Fig. 5, A and B. Both CNQX (10 µM) and AP-5 (30-40 µM) inhibited the response to dorsal root stimulation and to glutamate application, suggesting that both NMDA and non-NMDA (AMPA/kainate) receptors contribute to the inward current evoked either synaptically or by direct application of glutamate. In the presence of both antagonists together the response was nearly abolished, although a small residual inward current remained, indicating complete or nearly complete block of each receptor subtype at these concentrations of antagonist (Fig. 5).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Pharmacology of motor neuron responses to dorsal root stimulation and glutamate application. A, inward currents evoked by dorsal root stimulation are slightly sensitive to the NMDA receptor antagonist AP-5; sensitivity is most pronounced in the later components of the response. The glutamate non-NMDA receptor antagonist CNQX when applied with AP-5 almost completely abolishes the response. B, inward currents evoked by pressure application of a pulse of glutamate at the time indicated by the arrows. Downward square deflections are hyperpolarizing voltage steps to monitor resistance. In many cases the inward currents elicited spikes that were abolished by the sodium channel-blocking agent TTX. TTX also reduced the amplitude and shortened the duration of the inward currents, indicating that in the absence of TTX some of the glutamate-induced current was secondary to activity in neurons presynaptic to the motor neurons. AP-5 slightly reduced the inward current; the combination of AP-5 and CNQX reduced it further but a small amount of current remained.

Ethanol Depresses Currents at Both AMPA and NMDA Receptors. Ethanol depressed glutamate-evoked currents in the presence of CNQX (10 µM), indicating an action on NMDA receptors (Fig. 6A). Ethanol also depressed inward currents in the presence of 40 µM AP-5, even when inhibitory channels were blocked by bicuculline and strychnine (Fig. 6B). Six cells were studied in the presence of each of the glutamate antagonists. When NMDA receptors were blocked with AP-5, 100 mM ethanol significantly depressed the residual glutamate current to a mean of 59% of control ± 8.7 (mean ± S.E.M., P < .01). When CNQX was used to block AMPA/kainate receptors, 100 mM ethanol significantly depressed the remaining NMDA current to a mean of 51% of control ± 22.5 (P < .01). These results are not different from the effects of ethanol in untreated preparations. Ethanol thus acts on both major subtypes of glutamate receptors in motor neurons, NMDA and AMPA/kainate.

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Ethanol affects both NMDA and non-NMDA receptor-mediated inward currents evoked by glutamate application. A, in the presence of the non-NMDA receptor antagonist CNQX glutamate applied at the time indicated by arrows evokes inward currents presumably through NMDA receptors. Ethanol reversibly depresses these currents. B, ethanol depresses inward currents when NMDA receptors are blocked by AP-5 and when both GABAA and glycine inhibitory chloride channels are blocked by their respective antagonists bicuculline and strychnine.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results show that ethanol, at and below general anesthetic concentrations, depresses inward currents evoked by glutamate when impulse generation is blocked by TTX. The depressant effects of ethanol are still present when inhibitory chloride channels are blocked by bicuculline and strychnine. Currents are also sensitive to ethanol when either NMDA or AMPA/kainate receptors are blocked. Previous studies have shown no depressant action of a kainate-specific antagonist in spinal cord (D. L. Tauck and J. J. Kendig, unpublished data), leading to the conclusion that the residual current when NMDA receptors are blocked is carried by AMPA receptors. These results show that ethanol acts directly on motor neurons to depress currents mediated by both AMPA and NMDA receptors, in addition to any indirect effects due to presynaptic actions or enhancement of GABA_A or glycine inhibition.

A Direct Action on Motor Neurons. Experiments on synaptically evoked potentials in either the intact cord or the spinal cord slice do not allow depression of the monosynaptic reflex and its underlying EPSP to be partitioned into presynaptic depression of transmitter release versus postsynaptic effects on the motor neurons. For volatile agents, there is evidence based on studies of the F-wave that motor neuron excitability is decreased at MAC (King and Rampil, 1994); however, depression of this reflex also may include changes in input from tonically active neurons presynaptic to the motor neurons. Direct glutamate application as in the present study is the classic way to resolve this issue. In similar studies with volatile agents, all depress synaptically evoked glutamatergic transmission at various central nervous system sites, but some do not alter responses to glutamate application, a result suggesting that their actions are pre- rather than postsynaptic (Richards, 1973, 1983; Richards and White, 1975; Richards and Smaje, 1976; Perouansky et al., 1994, 1995). For ethanol we have observed inhibition of sodium currents in rat dorsal root ganglion cells at 100 to 200 mM, suggesting the possibility of presynaptic inhibitory actions upstream from the calcium channels that mediate transmitter release (Wu and Kendig, 1998). There are reports that presynaptic sodium channels are also sensitive to volatile anesthetic agents (Ratnakumari and Hemmings, 1998), suggesting that this mechanism may also contribute to depression of synaptic transmission. The present study shows that ethanol depresses both synaptically evoked responses and responses to glutamate application, suggesting that its actions on synaptic transmission to motor neurons in the spinal cord are at least in part postsynaptic rather than on primary afferent terminals or interneurons presynaptic to the motor neurons.

Effectiveness of Receptor Blockade. The concentrations of strychnine and bicuculline were sufficient to block all the spinal cord glycine and GABA_A receptors respectively (Wang and Dun, 1990; Jonas et al., 1998). They were added to TTX not only for the concerns about presynaptic transmitter release of GABA and glycine from inhibitory neurons as outlined above, but also because ethanol has been reported to increase the frequency of spontaneous miniature inhibitory postsynaptic currents in spinal cord motor neurons (Cheng et al., 1996). If all inhibitory amino acid-gated chloride channels were blocked by this strategy, then the only remaining way in which ethanol effects could be mediated via these channels would be if ethanol directly gated the channels in addition to enhancing the effects of the inhibitory transmitter, and did so at a site not blocked by the antagonists. Direct gating of GABA_A receptors is known for some i.v. anesthetics and there is one report of direct gating with volatile agents (Yang et al., 1992). There have been reports that ethanol may directly stimulate chloride flux (Mehta and Ticku, 1994). However, other groups do not find increases in baseline chloride flux independent of GABA, nor any evidence of ethanol-induced GABA currents in either oocytes or hippocampal slices (R. A. Harris, personal communication).

Ethanol Effects on Excitatory Amino Acid Receptors. The present study shows that ethanol directly depresses both AMPA and NMDA receptor-mediated responses. There is a broad consensus that NMDA receptors are sensitive to ethanol at both intoxicating and anesthetic concentrations (Lovinger et al., 1989, 1990; Peoples and Weight, 1995; Dildy-Mayfield et al., 1996). There has been debate about the sensitivity of glutamate non-NMDA receptors (Crews et al., 1996; Lovinger, 1997). The results of the present study as well as others (Morrisett and Swartzwelder, 1993; Crews et al., 1996; Dildy-Mayfield et al., 1996) suggest that in the spinal cord currents mediated by both AMPA and NMDA glutamate receptors are sensitive to ethanol, certainly at the concentrations associated with general anesthesia. At other sites ethanol also acts on both NMDA and non-NMDA glutamate receptors. In locus ceruleus ethanol (100 mM) equally inhibits NMDA- and AMPA-induced inward currents (Nieber et al., 1998). In nucleus accumbens ethanol reduces NMDA- and kainate-induced currents but not AMPA; NMDA currents are more sensitive to ethanol (Nie et al., 1994).

The Role of GABA_A and Glycine Inhibition. It is a pervasive theory in the field of anesthesia that actions on GABA_A receptors are the dominant factor in producing the anesthetic state (Franks and Lieb, 1994). It has been established that the spinal cord is the anatomic site responsible for the most common anesthetic endpoint used to compare potencies among agents (Rampil et al., 1993; Antognini, 1997). The results of the present study exclude both GABA_A and glycine receptors as essential to anesthetic depression of glutamate-evoked currents in spinal cord. As outlined in the introduction, ethanol and volatile general anesthetics enhance and prolong GABA_A and glycine currents (Mihic et al., 1994). In addition, both may increase tonic inhibition by increasing spontaneous inhibitory transmitter release (Mody et al., 1991; Cheng et al., 1996). Enhancement of GABA_A inhibition has been proposed as a common mechanism of general anesthesia (Tanelian et al., 1993; Franks and Lieb, 1993). However, with respect to abolition of nocifensive movement as the definition of anesthesia, a case may be made that at least some types of GABA_A receptors are not an important target. Benzodiazepines, which have prominent effects on receptors of certain subunit compositions, require very high concentrations to abolish movement (J. W. Mandema, personal communication). A recent study shows that an agent that potentiates activity at benzodiazepine-sensitive receptors does not alter the potency (MAC) of the inhalation agent desflurane (Yost et al., 1998).

Summary. The results of the present study show a direct depressant effect of ethanol on glutamate responses in spinal cord motor neurons. Depression persists when inhibitory receptors are blocked. These results are consistent with the hypothesis that immobility as an anesthetic endpoint is due to actions on motor neurons, and that direct depression of glutamate excitatory responses plays a role independent of enhancement of GABA_A or glycine inhibition.