Effects of Thiopental and Its Optical Isomers on Nicotinic Acetylcholine Receptors

This study conforms to the United Kingdom Animals (Scientific Procedures) Act of 1986.

Adult female Xenopus laevis  frogs (Blades Biological, Cowden, Kent, United Kingdom) were maintained in fresh-water holding tanks with a 12-h light–dark cycle in a temperature-controlled room at 20–22°C. Animals were anesthetized by immersion in a 0.2% (wt/vol) solution of tricaine (3-aminobenzoic acid ethyl ester methanesulfonate). A surgical incision into the abdomen was made, and portions of the ovaries containing oocytes were removed and teased apart with forceps. The ovarian fragments were then briefly washed in calcium-free oocyte incubation buffer (Ca²⁺-free OIB; composition: 88 mm NaCl, 1 mm KCl, 2.4 mm NaHCO³, 0.8 mm MgSO⁴, 15 mm HEPES, titrated to pH 7.5 with NaOH). Individual oocytes were obtained by digestion of these fragments in Ca²⁺-free OIB containing 2 mg/ml of type 1A collagenase (Sigma Chemical Co., Poole, Dorset, United Kingdom), for 3 h at room temperature. The oocytes were then carefully washed in Ca²⁺-free OIB to remove all traces of collagenase, before being transferred into normal OIB (composition 88 mm NaCl, 1 mm KCl, 2.4 mm NaHCO³, 0.8 mm MgSO⁴, 0.4 mm CaCl², 0.3 mm Ca(NO³)², 15 mm HEPES, titrated to pH 7.5 with NaOH). Oocytes were then scrutinized under a dissection microscope, and healthy eggs at stages 5–6 of development were set aside for injection with either DNA (chick α7 subunit; rat α⁴and β2 subunits) or RNA (mouse α, β, γ, and δ subunits). Injections of cDNA were made with 10 nl of diethyl pyrocarbonate–treated water containing 0.1– 1.0 pg of the appropriate cDNA directly into the nucleus of the oocyte, using the “blind” method. In the case of the α4β2 combination, a premixed solution with an equimolar ratio of the α4 and β2 cDNAs was used. Injections of cRNA were made into the cytosol of the eggs with 50 nl of diethyl pyrocarbonate–treated water containing a total of 10–40 ng of the cRNAs premixed at an equimolar ratio of the α, β, γ, and δ subunits. All injections were conducted with use of calibrated micropipettes (10–16-μm tip diameter) in combination with a Picospritzer II valve (General Valve Corp., Fairfield, NJ), which provided short pressure pulses of nitrogen gas to expel known volumes of the nucleic acid solutions from the micropipettes. Injected oocytes were maintained in a cooled incubator at 19–20°C in normal OIB supplemented with antibiotics (penicillin 100 IU/ml, streptomycin 100 μg/ml; Life Technologies, Paisley, United Kingdom) for 2–7 days before use. All chemicals, unless otherwise stated, were obtained from Sigma Chemical Company (Poole, Dorset, United Kingdom).

The nAChR cDNAs used here were kindly supplied as follows. Rat neuronal nAChR cDNAs for the α4 and β2 subunits (in the pSM expression vector) were from Jim Patrick (Baylor College of Medicine, Houston, TX). Mouse muscle nAChR cDNAs encoding the α, β, γ, and δ subunits were from Jim Boulter (Salk Institute, San Diego, CA), contained in either the pSP64 (γ) or pSP65 (α, β, and δ) vectors. The chick α7 cDNA was supplied by Marc Ballivet (Université de Genève, Switzerland) in the “flip” expression vector. 16The pSM- and “flip”-derived plasmids contain SV40 promoter sequences, which allows direct use of these cDNAs by the nuclear injection method. However, in the case of the mouse muscle subunits, individual cRNA transcripts suitable for injection were prepared from the cDNAs supplied with use of a Riboprobe System cRNA production kit (Promega, Southampton, United Kingdom).

The extracellular saline used for electrophysiological recordings had the following composition: 110 mm NaCl, 2 mm KCl, 1 mm MgCl², 2 mm BaCl², and 10 mm HEPES (titrated to pH 7.5 with NaOH). Additionally, atropine (100 nm) was added to all solutions to suppress the activity of endogenous muscarinic receptors. At this concentration atropine did not affect the acetylcholine-evoked nicotinic currents. The R  (+) and S  (−) enantiomers of thiopental were prepared from racemic thiopental stock (94.3% sodium salt–sodium carbonate mixture) to a purity level of 99.0% or more, with use of high-performance liquid chromatography with a permethylated β-cyclodextrin column, as described elsewhere. 2Acetylcholine chloride and thiopental solutions were prepared freshly from stock salts on the day of the experiment.

Inward currents evoked by bath application of acetylcholine chloride (0.3 μm to 10.0 mm) were recorded at −70 mV from oocytes expressing the desired subunit combinations and situated in a bath (volume ≈ 50 μl), where they were continuously perfused at a rate of 2 ml/min with appropriate solutions. Currents were recorded with use of an Axoclamp 2A amplifier (Axon Instruments, Foster City, CA) in the two-electrode voltage clamp mode. Recording electrodes were made from thin-walled filamented borosilicate glass capillary tubes (GC150TF, Clark Electromedical Instruments, Reading, Berkshire, United Kingdom); the passive electrode was filled with 3 m KCl and the current-passing electrode with 3 m KCl supplemented with BAPTA [1,2-bis(2-aminophenoxy)ethane-N,N,N’,N’-tetraacetic acid, tetrapotassium salt, 100 mm]. These electrodes typically had resistances of 0.4–0.8 MΩ. Currents were filtered at 10 Hz (−3 dB, 8-pole Bessel filter; Model 902, Frequency Devices, Haverhill, MA) before being digitized (sampling rate 20 Hz) and stored on a computer for offline analysis. All experiments were performed at room temperature (21–23°C).

Current responses were evoked by continuous application of acetylcholine-containing solutions until a well-defined peak in the response was observed (usually at 20–30 s); this peak was then used as the measure of receptor activity. During data collection a standard acetylcholine concentration was chosen and applied alternately throughout the experiment, to allow the data to be normalized and to correct for any fluctuation in the control response. Two protocols were used concurrently to assess the effects of thiopental. In the first instance, a protocol was used where appropriate concentrations of acetylcholine were simply coapplied with the various concentrations of thiopental. In the second, or preapplied protocol, the oocyte was bathed in the appropriate concentration of thiopental for 1 min before its coapplication with acetylcholine. (The effect of preincubation was complete after approximately 30 s.) In both cases, application of acetylcholine was maintained until an unambiguous peak in the response was observed. Such applications were repeated to ensure consistency of the response. Thiopental by itself had no significant effect on the resting current. Inhibition by thiopental of the response to acetylcholine was shown to be fully reversible at the thiopental concentrations used here. Inhibition of the acetylcholine-evoked current by thiopental was calculated by averaging controls before and after application of anesthetic and comparing these mean control values with the current responses in the presence of thiopental.

Control acetylcholine concentration–response data were fitted to a standard activation Hill equation of the form where E is the peak acetylcholine-induced current expressed as a percentage of the maximal current, A is the concentration of acetylcholine, n ^H  is the Hill coefficient, and EC ⁵⁰  is the acetylcholine concentration that gives a half-maximal effect.

Similarly, data describing the inhibition of the acetylcholine-evoked response by thiopental were fitted to an inhibition Hill equation of the form where y is the percentage of the control acetylcholine current remaining in the presence of thiopental at a concentration I, n ^H  is the Hill coefficient, and IC ⁵⁰  is the concentration of thiopental required to inhibit the control response by 50%.

Numerical fits to these Hill equations were obtained using the method of unweighted least-squares (Marquardt-Levenberg algorithm, KaleidaGraph version 3.0.2, Abelbeck Software, Reading, PA) applied to individual (not mean) data points. Values throughout are given as mean ± SEM, and statistical significance was assessed with use of the Student t  test.

Acetylcholine concentration–response relationships were determined for Xenopus  oocytes expressing rat neuronal heteromeric α4β2, chick neuronal homomeric α7, and mouse muscle heteromeric αβγδ nAChRs. For each nAChR, the data were fitted to the activation Hill equation (equation 1) with the results given in figure 1and table 1. The insets to figure 1show typical inward current responses to acetylcholine; note the differing degrees of desensitization, depending on receptor type and acetylcholine concentration. The EC ⁵⁰  concentrations and Hill coefficients (n ^H ) for the relatively slowly desensitizing heteromeric α4β2 and αβγδ receptors are comparable to those reported previously from this laboratory. 9For the homomeric α7 receptor, our EC ⁵⁰  value is in reasonable agreement with previously published values, 16–18although these are all, almost certainly, substantial overestimates (perhaps by a factor of 10). This is an unavoidable consequence of combining fast α7 receptor desensitization with the slow-mixing rates characteristic of the Xenopus  oocyte perfusion system, and it is because of the peak of the acetlycholine receptor response occurring before the steady state acetylcholine concentration is attained. 19Experiments (not shown) comparing the time courses of receptor activation and solution exchange suggested that this was a problem for the homomeric α7 but not for the heteromeric α4β2 and αβγδ nAChRs.

For each nAChR, thiopental concentration-dependent inhibition of acetylcholine-activated inward currents was determined during four conditions: at a low and a high concentration of acetylcholine, both with and without preincubation with thiopental. The low and high ACh concentrations for each nAChR (table 2) differed by at least an order of magnitude, with the low concentration being selected for experimental convenience as one that activated conveniently large but essentially nondesensitizing currents (for the heteromeric α4β2 and αβγδ receptors) or minimally desensitizing but still measurable currents (for the homomeric α7 receptor). Taking into account a possible 10-fold overestimate of our EC ⁵⁰  for the α7 receptor (see above), the low acetylcholine concentrations for all three nAChRs turned out to be near to their actual EC ⁵⁰  concentrations.

Preincubation with racemic thiopental of oocytes expressing nAChRs increased their sensitivity to inhibition by this agent (fig. 2and table 2). This was true for all three nAChRs at both low and high concentrations of acetylcholine, with the effects of preincubation being greater at the high acetylcholine concentrations. These enhancements of inhibition could be surprisingly large. For example, consider the neuronal α4β2 nAChR. The IC ⁵⁰  for thiopental at the high (1,000 μm) concentration of acetylcholine decreased from 204 ± 38 to 18 ± 2 μm thiopental because of preincubation, whereas the corresponding decrease at the low (100 μm) concentration of acetylcholine was from 81 ± 5 to 26 ± 2 μm thiopental. (For comparison, EC ⁵⁰ = 25 μm thiopental for clinical general anesthesia, using as the anesthetic end point a lack of response of patients to a painful stimulus. 1)

Using the thiopental preapplication protocol (which is arguably more relevant pharmacologically; see Discussion), for all three nAChRs the racemic thiopental concentration–response curves were shifted to the left by the high concentrations of acetylcholine (fig. 2). This effect, which corresponds to an acetylcholine-induced increase in thiopental sensitivity, was very much more marked for the muscle (αβγδ) nAChR than for the neuronal (α4β2 and α7) nAChRs. All the nAChRs, however, could be sensitive to thiopental. For example, at the high concentrations of acetylcholine, the IC ⁵⁰  concentrations for inhibition of the three receptors (α4β2, α7, and αβγδ) were 18 ± 2, 34 ± 4, and 20 ± 2 μm thiopental, respectively (table 2). Comparing these IC ⁵⁰  values with the EC ⁵⁰  value for general anesthesia (25 μm thiopental), a case can be made for the substantial inhibition of all three nAChRs during maintenance of anesthesia with racemic thiopental.

In addition to the aforementioned experiments with racemic thiopental, we separated the two optical isomers (enantiomers) of thiopental, S  (−) and R  (+), and determined their individual effects on each of the three nAChRs at a fixed concentration (25 μm, the EC ⁵⁰  of the racemate for general anesthesia). For each nAChR we used eight conditions: a low and high concentration of ACh, with and without preapplication of each thiopental enantiomer.

The results of these experiments with the enantiomers are given in figure 3. Neither the neuronal α7 nor the muscle αβγδ nAChR differentiated between the R  (+) and S  (−) enantiomers of thiopental during any of the conditions used (P > 0.5). The α4β2 nAChR, conversely, was consistently more sensitive to the R  (+) enantiomer, and this difference was statistically significant (P < 0.01 or P < 0.05), except during the condition of high acetylcholine concentration and no preincubation with the anesthetic enantiomers. Overall, then, using the pharmacologically relevant preincubation protocol at both high and low concentrations of acetylcholine, thiopental was found to stereoselectively inhibit the neuronal α4β2 but not the neuronal α7 or the muscle αβγδ nAChR. Interestingly, the stereoselectively observed with the rat neuronal α4β2 nAChR is the opposite of that observed for general anesthesia in mice. 11–13 

Nicotinic acetylcholine receptors are members of an anesthetic-sensitive superfamily of structurally and genetically related fast neurotransmitter-gated receptor channels. 5–7This superfamily also includes GABA^A, glycine, and 5-hydroxytryptamine type 3 receptors (but not glutamate-gated receptors). Most members of this superfamily are very sensitive to minimum alveolar concentration levels of volatile general anesthetics but show varying susceptibilities to anesthetizing EC ⁵⁰  concentrations of intravenous general anesthetics such as propofol, etomidate, and thiopental. Low concentrations of propofol and etomidate, for example, substantially potentiate GABA^Areceptors, 20–22while having relatively modest effects on glycine receptors, 20,21,235-hydroxytryptamine type 3 receptors 24,25and neuronal nAChRs. 9,26In addition, thiopental and etomidate potentiate GABA^Areceptors in a stereoselective manner (propofol is achiral and cannot act stereoselectively), with the same rank orders of potency found for general anesthesia. 2,22These results are consistent with major roles for GABA^Areceptors in the production of general anesthesia with propofol, etomidate, or thiopental.

Thiopental, however, appears to be a more promiscuous intravenous agent than either propofol or etomidate, substantially affecting not only GABA^Areceptors but some other superfamily members as well. For example, although 5-hydroxytryptamine type 3 receptors in N1E-115 neuroblastoma cells are insensitive to thiopental, 25,27recombinant glycine receptors in the Xenopus  expression system are sensitive. 28Furthermore, a steady current (but not the peak current) through nicotine-activated ganglionic-type neuronal nAChRs in rat PC12 cells can be more than half-inhibited by surgical EC ⁵⁰  concentrations of thiopental. 29The situation with respect to thiopental inhibition of CNS-type neuronal nAChRs is unknown and is addressed in the current study.

Neuronal nAChRs in vertebrates have a wide range of compositions. 30–33Each functional receptor is a pentamer consisting of two to five subunits of the α type (α2, α3, α4, α⁵, α⁶, α7, α8, α9) and zero to three subunits of the β type (β2, β3, β4). (Muscle-type nAChRs are made up of other subunits: two α¹, one β¹, one δ, and either one γ or one ε.) The heteromeric α4β2 (two α4 and three β2 subunits) and homomeric α7 (five α7 subunits) nAChRs are representative of two important classes of neuronal CNS receptors that account for most of the high-affinity binding sites in the brain for nicotine and α-bungarotoxin, respectively. In addition, the α7 nAChR is especially permeable to the physiologically important cation Ca²⁺. Furthermore, the α4β2 receptor is supersensitive to inhibition by volatile anesthetics, 9whereas the α7 receptor is insensitive to these agents. 10For these reasons, and taking into account the availability or otherwise of specific subunit clones, we chose to study the rat neuronal heteromeric α4β2 nAChR, the chick neuronal homomeric α7 nAChR, and (for comparison) the embryonic mouse muscle heteromeric αβγδ nAChR. We expressed all three nAChRs in the same system (the Xenopus  oocyte) to make fair comparisons of their behaviors toward thiopental. However, the use of clones from different species, although necessary at the time this study began, may have introduced some variability into these comparisons.

The physiologic functions of neuronal nAChRs in the CNS are still largely unknown, 32and much of the interest in these nicotinic receptors stems from the agonist effects of tobacco-derived nicotine. There is little evidence for fast nicotinic cholinergic synaptic transmission in the brain, which is surprising in view of its important roles in peripheral ganglionic and neuromuscular transmission. 34This points to a presynaptic rather than a postsynaptic role for central nAChRs. It is partly for this reason that presynaptic neuronal nAChRs in the brain are currently receiving so much attention, 34,35the idea being that acetylcholine or nicotine binding to such receptors modulates the release of a number of different types of neurotransmitters (e.g. , dopamine, norepinephrine, glutamate, GABA, and acetylcholine) from presynaptic terminals. Although this is an attractive hypothesis that has received some experimental support, many questions remain to be answered. For example, although the source of nicotinic agonist poses no conceptual problems for cholinergic synapses or for experiments with exogenous nicotine, it is still not clear in most instances what are the physiologic sources of endogenous agonist (presumably acetylcholine) and what are the relevant concentrations at the presynaptic receptors. Whatever the answers, it seems likely that elucidation of the mechanisms and roles of neuronal nAChRs will produce more surprises.

Thiopental inhibited the peak current responses to acetylcholine in a protocol-dependent manner. We found that preincubation of oocytes with thiopental resulted in enhanced inhibition for all nAChRs (fig. 2and table 2). This effect, which could be very large, was presumably a result of preincubation favoring the attainment of equilibrium between receptor and anesthetic before the application of the acetylcholine agonist. (This is the simplest explanation, although we cannot rule out slow second messenger effects.) Pharmacologically, preapplication of anesthetic would appear to be the protocol most relevant to the steady state maintenance of general anesthesia. For induction, the situation is less clear but would still seem to favor preapplication, because the thiopental concentration is at a quasi–steady state for times on the order of seconds, whereas acetylcholine release, presumably from nerve terminals, occurs at times on the order of milliseconds. For interpreting mechanisms at the molecular level, anesthetic–acetylcholine coapplication without anesthetic preapplication is the easier option but should be resisted, we believe, because it can be misleading. In our experiments, for example, we would have erroneously concluded from the results using the coapplication protocol alone that both the neuronal α4β2 and α7 nAChRs are relatively insensitive to thiopental.

Our results using the pharmacologically more relevant thiopental preincubation protocol, conversely, show that both neuronal and muscle nAChRs can be very sensitive to inhibition by thiopental. Indeed, at high concentrations of acetylcholine agonist IC ⁵⁰ = 18 ± 2, 34 ± 4, and 20 ± 2 μm thiopental for the neuronal α4β2, neuronal α7 and muscle αβγδ nAChRs, respectively, all with Hill coefficients near unity (table 2). These IC ⁵⁰  concentrations for inhibiting the nAChRs can be compared with the EC ⁵⁰  of 25 μm thiopental for surgical general anesthesia 1; at this anesthetizing concentration of thiopental, the aforementioned IC ⁵⁰  concentrations translate into nAChR inhibitions of 58, 42, and 56%, respectively, for the neuronal α4β2, neuronal α7^,and muscle αβγδ nAChRs. At the low concentrations of acetylcholine, the neuronal nAChRs are slightly less sensitive to thiopental than at the high concentrations of acetylcholine (this is significant at the 5% level for the α4β2 but not the α7 nAChR), whereas the muscle nAChR is profoundly less sensitive to thiopental at low concentrations of acetylcholine. This result is consistent with substantial open-channel block for the muscle-type receptor 36,37but not for the neuronal α4β2 receptor, 9as has been found for volatile anesthetics.

Does thiopental inhibition of central neuronal nAChRs play an important role in producing general anesthesia? If it does, these receptor channels might be expected to be substantially inhibited at thiopental EC ⁵⁰  concentrations for surgical general anesthesia. Using as an anesthetic end point a failure of patients to respond purposefully to a painful stimulus equivalent to a surgical incision, we previously calculated 1that this EC ⁵⁰  is approximately 25 μm thiopental. It can be seen from figures 2A and 2Bthat, using the thiopental preincubation protocol with both high and low concentrations of acetylcholine, the α4β2 and α7 neuronal nAChRs were indeed substantially (35–60%) inhibited at this anesthetizing EC ⁵⁰  concentration of thiopental (arrows in fig. 2). Although this suggests that the activities of neuronal nAChRs in the CNS are severely compromised during thiopental general anesthesia, it does not logically follow that thiopental disruption of nAChR function actually causes general anesthesia.

Clearly, another criterion (in addition to sensitivity) is needed to judge relevance. A convenient yet powerful test 1,7exploits the fact that general anesthesia is stereoselective for many chiral (i.e. , optically active) anesthetics, including thiopental. 13If a given receptor is one of at most a few important targets for the anesthetizing actions of a particular chiral agent, then the rank order and ratio of effectiveness of its optical isomers should be similar for producing general anesthesia and for modulating the activity of the receptor being tested. Using this test with optical isomers of not only thiopental but also of pentobarbital, hexobarbital, etomidate, and anesthetic neurosteroids, previous studies from this 2,22and other 38laboratories have identified potentiation of GABA^Are-ceptor activity as a major factor underlying the production of general anesthesia by these intravenous agents. Conversely, using this test with ketamine enantiomers identifies the glutamate N -methyl-d-aspartate receptor as a major target for this agent. 1,39,40 

What happens when this test is applied to our results for the R  (+) and S  (−) optical isomers of thiopental acting on neuronal nAChRs? S  (−)-thiopental is more potent than R  (+)-thiopental at abolishing the mouse righting reflex, 11,12whereas it is less effective at inhibiting the neuronal heteromeric α4β2 nAChR and equally effective at inhibiting the neuronal homomeric α7 nAChR (fig. 3). (For a comparison, the muscle αβγδ nAChR, like the neuronal α7 nAChR, showed no discrimination between the thiopental enantiomers.) Thus both neuronal nAChRs fail to pass the stereoselectivity criterion for important roles in thiopental anesthesia.

A very recent study 41using depressant and convulsant enantiomers of the barbiturate MPPB (1-methyl-5-phenyl-5-propyl barbituric acid) examined their effects on ganglion-type nAChRs in PC12 cells. The two MPPB enantiomers had opposite (inhibitory and excitatory) effects on animals, whereas they were found to have uniformly inhibitory effects on the activities of the neuronal nAChRs. The investigators concluded that inhibition of these ganglion-type neuronal nAChRs does not contribute to the anesthetic action of barbiturates.

What are the implications of our results for understanding the molecular mechanisms of thiopental general anesthesia? First, we conclude from our studies with the pure optical isomers of thiopental that neither of the neuronal nAChRs tested plays a major role in the anesthetizing actions of thiopental. Instead, from the positive results of our previous GABA^Areceptor study 2with these same optical isomers, it would appear that potentiation of the activities of inhibitory GABA^Areceptors is the main factor. Second, although thiopental inhibition of nAChRs does not actually produce anesthesia, the fact that these receptors can be very sensitive to anesthetizing concentrations of this intravenous agent suggests that they may be involved in the generation of anesthetic side effects.

The authors thank the Medical Research Council for its support and Marc Ballivet, Jim Boulter, and Jim Patrick for their kind gifts of the individual nAChR subunit cDNAs.