What is the prototype for the classification of drugs referred to as benzodiazepine?

Benzodiazepines cause sedation by positive allosteric modulation of γ-aminobutyric acid type A (GABAA) receptor function. Native GABAA receptors are inhibitory ligand-gated ion channels formed from 5 protein subunits. The number of known subunits approaches 20, leading to a large diversity of GABAA receptor subtypes.1 Many, but not all subtypes of the GABAA receptor harbor a binding site for benzodiazepines. The classical benzodiazepine binding site is located between an α1, α2, α3, or α5 and a γ subunit.2 However, in recent years, evidence emerged that further, “nonclassical” benzodiazepine binding sites are present on GABAA receptors. Walters et al.3 showed that a high- and a low-affinity binding site coexist on the α1β2γ2 GABAA receptor subtype. This GABAA receptor subtype is highly expressed in cortical neurons.2 Furthermore, Baur et al.4 proposed yet another binding site for benzodiazepines, which, curiously, prevents modulation via the classical benzodiazepine binding site, perhaps contributing to the safety of these drugs.

At present, it is unknown whether these postulated multiple binding sites and the resulting multiple molecular actions of benzodiazepines on GABAA receptors translate into altered activity patterns of neuronal networks. Studies on genetically modified mice suggest that diazepam causes sedation by acting on GABAA receptors located on glutamatergic cortical neurons.5 Benzodiazepines enhance inhibitory synaptic transmission, and therefore decrease neural activity and excitability in the cerebral cortex. This action is possibly a key mechanism in producing sedation and hypnosis.

In this study, we addressed the question as to what extent the actions of diazepam mediated via the classical benzodiazepine binding site contribute to the overall effect of this drug on action potential firing of cortical neurons. The major finding is that diazepam-induced depression of neuronal activity displays a biphasic concentration-response relationship. The high-affinity, but not the low-affinity component was sensitive to the benzodiazepine antagonist flumazenil, suggesting that only the former is mediated via the classical benzodiazepine binding site. Together with previous studies, these results suggest that the classical benzodiazepine binding site, which is involved in sedation,6 contributes only in part to the anesthetic properties of diazepam.

METHODS

Organotypic Slice Cultures

Wild-type C57black6 mice of both sexes were used for this study. All procedures were approved by the animal care committee (Eberhard Karls University, Tuebingen, Germany) and were in accordance with German law on animal experimentation. Neocortical slice cultures were prepared from 2- to 5-day-old mice as described by Gähwiler.7 All efforts were made to minimize both the suffering and number of animals used. In brief, animals were deeply anesthetized with isoflurane and decapitated. Cortical hemispheres were aseptically removed and 300-μm-thick coronal slices were cut. Slices were fixed on glass coverslips by a plasma clot, transferred into plastic tubes containing 750 μL of nutrition medium and incubated in a roller drum at 37°C. After 1 day in culture, antimitotics were added. The suspension was renewed twice a week. Cultures were used after 2 weeks in vitro.

Electrophysiology

Extracellular network recordings were performed in a recording chamber mounted on an inverted microscope. Slices were perfused with artificial cerebrospinal fluid (ACSF) consisting of (in mM) NaCl 120, KCl 3.3, NaH2PO4 1.13, NaHCO3 26, CaCl2 1.8, and glucose 11, bubbled with 95% oxygen and 5% carbon dioxide. The recording chamber consisted of a metal frame with a glass bottom. A heating wire was glued onto the metal frame, which was heated by passing an appropriate direct current through the heating wire. All experiments were conducted at 34°C. ACSF-filled glass electrodes with a resistance of approximately 3 to 5 MΩ were advanced into the tissue until extracellular single- or multiunit spike activity exceeding 100 μV in amplitude were visible. Data were acquired on a personal computer with the DigiData 1200 AD/DA interface and AxoScope 9 software (Axon Instruments, Foster City, CA).

Preparation and Application of Test Solutions

Diazepam (Braun, Melsungen, Germany) was diluted in ACSF and placed into gastight glass syringes. The drug containing ACSF was applied via bath perfusion using syringe pumps (ZAK, Marktheidenfeld, Germany), connected to the experimental chamber via Teflon tubing (Lee, Frankfurt, Germany). The flow rate was approximately 1 mL/min and the recording chamber had a volume of 1.5 mL. When switching from ACSF to drug-containing solutions, the medium in the experimental chamber was replaced by at least 95% within 2 minutes. To ensure steady-state conditions, recordings during diazepam treatment were conducted 10 to 12 minutes after commencing the change of the perfusate. This time interval has been proven to be sufficient for steady-state conditions,8 because diffusion times in slice cultures are considerably shorter compared with acute slice preparations.9,10 Bicuculline-methiodide and flumazenil were obtained from Sigma-Aldrich (Taufkirchen, Germany).

Data Analysis

Extracellular recorded signals were filtered and counted offline using self-written programs in MATLAB (MathWorks, Natick, MA). The activity pattern of neocortical slice cultures is characterized by bursts of spontaneous action potential firing separated by periods of neuronal silence (Fig. 1A). Action potentials were detected and the average firing rate was computed using an automated event detection algorithm with a threshold set approximately 2 times higher than the baseline noise (Fig. 1B). All parameters are shown as relative change compared with control condition. We used the Student t test for statistical testing; P values <0.05 were considered significant. All results are given as mean ± SEM.

Figure 1:

A, Original extracellular recording of action potential activity in a cultured neocortical slice under control condition. Six bursts of action potentials are separated by neuronal silence. B, A burst of action potentials under control conditions at a higher temporal resolution. All action potentials that cross the threshold are counted to detect the network activity. C, A burst of action potentials in the presence of 12.5 μM diazepam. The number of action potentials crossing the detection threshold is decreased in the presence of the drug.

RESULTS

Biphasic Depression of Spontaneous Action Potential Activity by Diazepam

Neocortical slice cultures displayed spontaneous action potential firing after 2 weeks in vitro. The mean multiunit firing rate was 18.0 ± 0.77 Hz (n = 237) and lies in the range of previously reported values.11 Diazepam, the prototype of all benzodiazepines, depressed spontaneous action potential firing in a biphasic manner. Over a broad concentration range from 50 nM to 6.25 μM, diazepam reduced neocortical firing rates nearly identically by approximately 20% (Fig. 2). A linear regression fit of this region of the concentration-response curve revealed a slope of 0.04 ± 0.07 (P > 0.1), indicating the lack of concentration dependence and the existence of a plateau.

Figure 2:

Diazepam induces a biphasic depression of spontaneous cortical network activity. Over a large concentration range (50 nM to 6.25 μM, 10 ≤ n ≤ 41), diazepam depresses network activity by approximately 20%. At even higher concentrations, a second, stronger, and concentration-dependent depression by diazepam is observed. A normalized firing rate of 1.0 would correspond to control values, a firing rate of 0 to a complete depression induced by diazepam. Data were fit according to Walters et al.3 The normalized firing rate in the presence of artificial cerebrospinal fluid (“sham-application,” gray dot) is included as zero value.

At concentrations >12.5 μM, a further, concentration-dependent depression of action potential firing was induced by diazepam, leading to an almost complete depression at a concentration of 100 μM. This biphasic depression induced by diazepam is shown in Figure 2.

Actions of the Benzodiazepine Antagonist Flumazenil

In clinical practice, flumazenil is used as an antagonist of the benzodiazepine binding site on the GABAA receptor and supposedly does not have a modulatory effect. However, it was shown in vitro that flumazenil acts as a weak benzodiazepine agonist in GABAA receptors containing γ1 subunits.12 In our cortical slice cultures, 250 nM flumazenil did not change the action potential firing rate. At a concentration of 25 μM, flumazenil showed a trend to depress spontaneous action potential firing; however, this did not reach statistical significance (Fig. 3).

Figure 3:

The benzodiazepine antagonist flumazenil does not alter neocortical network activity in vitro. Flumazenil altered action potential firing at neither the low concentration of 250 nM (−0.8% ± 5.5%, n = 16, P > 0.5) nor at the high concentration of 25 μM (−8.0% ± 10.1%, n = 15, P > 0.5). Mild depression of neocortical network activity induced by small concentrations of diazepam is antagonized by flumazenil. Diazepam at a concentration of 250 nM leads to a depression of neocortical activity by 15.93% ± 5.23% (n = 41, P < 0.01). In the presence of 250 nM flumazenil, this depression is nullified (change in firing rate + 6.80% ± 4.61%, n = 24, P > 0.1). Strong depression induced by a large concentration of diazepam is not antagonized by flumazenil. The high concentration of 25 μM diazepam leads to a depression of cortical firing by 32.81% ± 3.67% (n = 37, P < 0.001). This depression is unaffected by the additional presence of 25 μM flumazenil (depression by 34.48% ± 4.90%, n = 16, P < 0.001 compared with control condition). Flu = flumazenil; Dia = diazepam; n.s. = not significant.

Diazepam-Induced Depression of Neuronal Firing at Higher Concentrations Is Not Antagonized by Flumazenil

Walters et al.3 observed that the nanomolar effect of diazepam could be antagonized by flumazenil, whereas the micromolar could not, using 1 GABAA receptor subtype expressed in oocytes and exposed to concentrations of GABA that induced only 3% of the maximal current. Therefore, we were curious to perform analogous experiments in our cultured cortical slices, which express a natural variety of GABAA receptor subtypes13 and in which synaptic GABA concentrations are presumed to reach the millimolar range.14 The results of these experiments are summarized in Figure 3. The moderate depression induced by 250 nM diazepam was nullified in the presence of 250 nM flumazenil, but 25 μM flumazenil failed to antagonize the strong depression induced by 25 μM diazepam.

Diazepam-Induced Depression of Neuronal Firing at Higher Concentrations Is Partly Mediated by GABAA Receptors

To test whether the pronounced depression of neuronal activity observed at high concentrations of diazepam was mediated via GABAA receptors, a third set of experiments was performed. GABAA receptors were blocked by the competitive antagonist bicuculline, which was used at a low and a moderate concentration (1 and 10 μM). The effects of 50 and 100 μM diazepam were significantly reduced in the presence of bicuculline, although not completely abolished. In the presence of the GABAergic blocker bicuculline, action potential activity observed at 100 μM diazepam was nearly twice as high as compared with the absence of the blocker (Fig. 4).

Figure 4:

Effects of diazepam at high concentrations during blockade of γ-aminobutyric acid type A (GABAA) receptors by bicuculline. The network activity depressing actions of 50 and 100 μM diazepam under control conditions (white columns) and in the presence of bicuculline (1 μM, sparse patterning and 10 μM, dense patterning) are displayed. In the absence of the blocker, 50 μM diazepam leads to a remaining network spike activity of 40.2% ± 0.03% (n = 29), whereas in the presence of bicuculline, this value is significantly higher (1 μM bicuculline: 55.8% ± 0.05%, n = 36, P < 0.01; 10 μM bicuculline: 55.4% ± 0.04%, n = 41, P < 0.01). At 100 μM, the highest concentration of diazepam tested, the neuronal activity is 17.3% ± 0.02% of control activity (n = 12). This value is almost doubled in the presence of the GABAA receptor antagonist bicuculline: 30.7% ± 0.03% at 1 μM (n = 35, P < 0.05) and 31.0% ± 0.02% at 10 μM (n = 41, P < 0.001). Bicu = bicuculline.

DISCUSSION

Effects of Diazepam Mediated via the Classical Benzodiazepine Binding Site

This study demonstrates that nanomolar concentrations of diazepam produce a significant depression of spontaneous action potential firing in cultured cortical neurons. Because this nanomolar component is abolished by flumazenil, it is likely mediated by the classical benzodiazepine binding site. All tested concentrations between 50 nM and 12.5 μM caused a significant decrease in action potential firing of approximately 20%. Because the slope of a linear fit is not different from zero, we conclude that there is a lack of concentration dependence, i.e., the existence of a plateau. Therefore, it is tempting to speculate that the nanomolar component, which displays a 50% effective concentration lying between 10 and 50 nm, is saturated by concentrations of diazepam higher than 50 nm. The molecular mechanisms that are involved in forming this plateau are not known. Either the classical benzodiazepine binding site is completely occupied at these concentrations, or alternatively, a novel binding site as proposed by Baur et al.4 prevents further, concentration-dependent modulation via the classical benzodiazepine binding site.

Using knockin mice, Rudolph et al.6 demonstrated that the sedative properties of diazepam, which involve the classical benzodiazepine binding site, are mediated by GABAA receptors containing α1 subunits. In a subsequent study, the same group provided evidence that diazepam causes sedation via GABAA receptors localized on glutamatergic cortical neurons.5 In light of the present study, it is hence concluded that diazepam actions at nanomolar concentrations predominantly involve α1 subunit–containing GABAA receptors and that these actions are related to the sedative properties of this drug.

The Actions of Diazepam in Cortical Networks and in Expressed GABAA Receptors

In the present study, diazepam concentrations exceeding 12.5 μM decreased neuronal activity in a concentration-dependent manner. Walters et al.3 investigated the effects of diazepam on α1β2γ2 GABAA receptors, expressed in Xenopus oocytes. This GABAA receptor subtype is densely expressed in cortical neurons.2 When comparing the concentration-response relationship published by Walters et al. with Figure 2 of the present study, a striking similarity is evident. Akin to our findings in neocortical slice cultures, Walters et al. reported that only the high-affinity component was sensitive to flumazenil. Another analogy between the studies concerns the efficacy of the high- and low-affinity component. In both investigations, the low-affinity component was the more powerful one in modulating GABAA receptor function and neuronal activity, respectively. In summary, there is an astonishing similarity between the effects of diazepam on expressed α1β2γ2-GABAA receptors and on action potential activity of cortical neurons.

Estimating Concentrations of Diazepam Causing Hypnosis

Although diazepam has been introduced more than 50 years ago, the effective concentrations that cause hypnosis or immobility remain a matter of debate. First, different terms and definitions dealing with concentration and also end points of anesthesia have to be distinguished. Little and Bichard15 proposed a membrane concentration of 20 to 60 mM corresponding to ablation of movements. At first, this value seems quite high, but one has to keep in mind that this is a calculated number based on several assumptions and needs to be distinguished from the diazepam plasma concentration. Concerning the plasma concentration, Baird and Hailey16 reported that, after an IV injection of 20 mg diazepam, a plasma level of 1.1 μg/mL corresponding approximately to 3 μM is obtained. Furthermore, it was stated that the concentration of diazepam in the central nervous system is 3 times higher, compared with plasma levels.17 In addition, approximately half of the in vivo actions of diazepam are in fact caused by its metabolites,17 which are not present in ex vivo systems such as in the present study. When considering the aforementioned, brain concentrations of diazepam in the low double-digit micromolar range seem still quite reasonable. They might be achieved during surgical tolerance in rodents and during the induction of general anesthesia with benzodiazepines in humans. To approach the question of concentrations of diazepam causing hypnosis, we propose the following. We and others have studied the depressant effects of various anesthetics on action potential firing of cortical neurons and on the loss of righting reflex, a measure frequently used as a surrogate variable for hypnosis in rodents. An excellent correlation between the efficacy of anesthetic drugs in decreasing action potential firing of cortical neurons and in ablating righting reflexes in rodents was found (Fig. 5). Because all anesthetics included in Figure 5 predominantly act via GABAA receptors,8 a common binding site on cortical GABAA receptors for anesthetics is conceivable. Walters et al.3 observed that mutations in the transmembrane domain 2 of the GABAA receptor abolished the micromolar action of diazepam. These transmembrane regions are thought to be involved in the binding site for propofol and etomidate. Taken together, we speculate that this binding site on the GABAA receptor is also modulated by diazepam. This binding site is referred to as the “nonclassical” benzodiazepine binding site here.

Figure 5:

Comparison of the action potential depression potencies of various general anesthetics in cultured neocortical slices (50% effective concentration [EC50] inhibition of action potential firing) with in vivo potencies for the loss of righting reflex, which is frequently used as a surrogate variable for hypnosis in rodents. The values are taken from the literature.8 , 11 , 24 – 28 The slope of the regression line was 0.94 and correlation coefficient was 0.997. As can be seen from Figure 2, diazepam's EC50 of action potential depression in cultured cortical slices is approximately 40 μM. Assuming that the correlation is also true for diazepam, the proposed value for hypnosis (loss of righting reflex) should be below this value (gray arrow). AP = action potential.

From the correlation displayed in Figure 5, it is estimated that the concentration of diazepam that is required to cause loss of righting reflex (as a surrogate variable for hypnosis) should be approximately 40 μM, as opposed to 7 to 20 mM according to the estimation by Little and Bichard.15 However, this estimate only represents an uppermost limit because it is based on the assumption that diazepam exclusively acts via the nonclassical binding site, which is clearly not the case. Because diazepam decreases neuronal activity via both the high- and low-affinity binding site on GABAA receptors, the effective concentration that causes hypnosis should be smaller than 40 μM. This conclusion is further corroborated by the finding that loss of righting reflexes is antagonized by flumazenil in mice, pigeons, and squirrel monkeys.15,18 Therefore, it is very likely that the classical benzodiazepine binding site, to a yet unknown extent, contributes to the hypnotic properties of diazepam.

However, the immobilizing action of diazepam was not antagonized by flumazenil in mice,15 suggesting that the classical benzodiazepine binding site does not considerably contribute to this behavioral end point. We hypothesize that immobility is produced at concentrations of diazepam exceeding 30 μM, because at these concentrations flumazenil failed to antagonize the effect of diazepam on expressed GABAA receptors3 and on action potential firing of cortical neurons in the present study.

Actions of Diazepam in the Micromolar Range—Specific, Unspecific, or Both?

In general, the actions of benzodiazepines are mediated by GABAA receptors containing specific α subunits,2 but the effects of the high concentrations of diazepam used in the current study are not necessarily mediated exclusively by GABAA receptors. Alternatively, diazepam, at high concentrations, might also act via non-GABAergic receptor systems. The drug might impair the function of other ion channels or proteins. It also might increase membrane surface tension or change the membrane lateral pressure profile, resulting in alterations in the bilayer interaction with the channel proteins, altering their function, and thereby changing neuronal excitability. There is a large body of literature documenting such unspecific, i.e., not directly receptor-mediated actions of anesthetics, benzodiazepines, and other neuroactive drugs19,20 (for review see Ref. 21). To ascertain the extent to which the effects of high concentrations of diazepam are mediated via GABAA receptors, a series of experiments in the presence of the GABAA receptor blocker bicuculline was performed. Although bicuculline also blocks small conductance calcium-activated potassium channels at high concentrations,22 the 2 low concentrations used in the current study (1 and 10 μM) can be regarded as rather selective for antagonism of GABAA receptors.23 Against the background of GABAA receptor blockade, the remaining spontaneous cortical network activity in the presence of 100 μM diazepam is approximately twice as high, compared with control conditions in which diazepam can act via the GABAergic way. This implicates that, even at high concentrations, diazepam acts via GABAA receptors. However, diazepam induces a remarkable depression of cortical activity even in the presence of bicuculline. Therefore, we conclude that, besides the classical GABAergic pathway, diazepam at high concentrations in addition acts via non-GABAergic receptors or via unspecific membrane effects, which we cannot distinguish with the experiments performed in this study. In summary, the present study demonstrates that diazepam concentration-dependently induces cortical network depression by 3 different ways: via the classical benzodiazepine binding site, via a nonclassical binding site at the GABAA receptor, and via a non-GABAergic mechanism.

CONCLUSIONS

Our findings provide further evidence that the sedative actions of diazepam, which are mediated via the classical benzodiazepine binding site, involve GABAA receptors on cortical neurons.5,6 At nanomolar concentrations, diazepam decreases the activity of cortical neurons almost exclusively via the classical benzodiazepine binding site. Furthermore, our results suggest that the hypnotic action of diazepam is in part mediated by the classical benzodiazepine binding site and in part by a nonclassical benzodiazepine binding site on GABAA receptors. The effective concentration of diazepam causing hypnosis is estimated to be 10 to 30 μM. The immobilizing action of diazepam requires a 3- to 4-fold higher concentration and almost exclusively involves this nonclassical binding site.

REFERENCES

1. Olsen RW, Sieghart W. International Union of Pharmacology. LXX. Subtypes of gamma-aminobutyric acid(A) receptors: classification on the basis of subunit composition, pharmacology, and function—update. Pharmacol Rev 2008;60:243–60

2. Möhler H, Fritschy JM, Rudolph U. A new benzodiazepine pharmacology. J Pharmacol Exp Ther 2002;300:2–8

3. Walters RJ, Hadley SH, Morris KD, Amin J. Benzodiazepines act on GABAA receptors via two distinct and separable mechanisms. Nat Neurosci 2000;3:1274–81

4. Baur R, Tan KR, Luscher BP, Gonthier A, Goeldner M, Sigel E. Covalent modification of GABAA receptor isoforms by a diazepam analogue provides evidence for a novel benzodiazepine binding site that prevents modulation by these drugs. J Neurochem 2008;106:2353–63

5. Zeller A, Crestani F, Camenisch I, Iwasato T, Itohara S, Fritschy JM, Rudolph U. Cortical glutamatergic neurons mediate the motor sedative action of diazepam. Mol Pharmacol 2008; 73:282–91

6. Rudolph U, Crestani F, Benke D, Brünig I, Benson JA, Fritschy JM, Martin JR, Bluethmann H, Möhler H. Benzodiazepine actions mediated by specific γ-aminobutyric acidA receptor subtypes. Nature 1999;401:796–800

7. Gähwiler BH. Organotypic monolayer cultures of nervous tissue. J Neurosci Methods 1981;4:329–42

8. Antkowiak B. Different actions of general anaesthetics on the firing patterns of neocortical neurons mediated by the GABAA receptor. Anesthesiology 1999;91:500–11

9. Gredell JA, Turnquist PA, MacIver MB, Pearce RA. Determination of diffusion and partition coefficients of propofol in rat brain tissue: implications for studies of drug action in vitro. Br J Anaesth 2004;93:810–7

10. Benkwitz C, Liao M, Laster MJ, Sonner JM, Eger EI II, Pearce RA. Determination of the EC50 amnesic concentration of etomidate and its diffusion profile in brain tissue: implications for in vitro studies. Anesthesiology 2007;106:114–23

11. Drexler B, Jurd R, Rudolph U, Antkowiak B. Distinct actions of etomidate and propofol at beta3-containing gamma-aminobutyric acid type A receptors. Neuropharmacology 2009;57:446–55

12. Wafford KA, Bain CJ, Whiting PJ, Kemp JA. Functional comparison of the role of gamma subunits in recombinant human gamma-aminobutyric acidA/benzodiazepine receptors. Mol Pharmacol 1993;44:437–42

13. Brunig I, Scotti E, Sidler C, Fritschy JM. Intact sorting, targeting, and clustering of gamma-aminobutyric acid A receptor subtypes in hippocampal neurons in vitro. J Comp Neurol 2002;443:43–55

14. Mody I, Pearce RA. Diversity of inhibitory neurotransmission through GABA(A) receptors. Trends Neurosci 2004;27:569–75

15. Little HJ, Bichard AR. Differential effects of the benzodiazepine antagonist Ro 15–1788 after “general anaesthetic” doses of benzodiazepines in mice. Br J Anaesth 1984;56:1153–60

16. Baird ES, Hailey DM. Delayed recovery from a sedative: correlation of the plasma levels of diazepam with clinical effects after oral and intravenous administration. Br J Anaesth 1972;44:803–8

17. Klockowski PM, Levy G. Kinetics of drug action in disease states. XXIV. Pharmacodynamics of diazepam and its active metabolites in rats. J Pharmacol Exp Ther 1988;244:912–8

18. Witkin JM, Barrett JE, Cook JM, Larscheid P. Differential antagonism of diazepam-induced loss of the righting response. Pharmacol Biochem Behav 1986;24:963–5

19. Garcia DA, Perillo MA. Benzodiazepine localisation at the lipid-water interface: effect of membrane composition and drug chemical structure. Biochim Biophys Acta 1999;1418: 221–31

20. Song C, Nerdal W. Olanzapine interaction with dipalmitoyl phosphatidylcholine (DPPC) and 1-palmitoyl-2-oleoyl phosphatidylserine (POPS) bilayer: a (13)C and (31)P solid-state NMR study. Biophys Chem 2008;134:47–55

21. Urban BW, Bleckwenn M, Barann M. Interactions of anesthetics with their targets: non-specific, specific or both?Pharmacol Ther 2006;111:729–70

22. Debarbieux F, Brunton J, Charpak S. Effect of bicuculline on thalamic activity: a direct blockade of IAHP in reticularis neurons. J Neurophysiol 1998;79:2911–8

23. Seutin V, Johnson SW. Recent advances in the pharmacology of quaternary salts of bicuculline. Trends Pharmacol Sci 1999;20:268–70

24. Franks NP. General anaesthesia: from molecular targets to neuronal pathways of sleep and arousal. Nat Rev Neurosci 2008;9:370–86

25. Franks NP, Lieb WR. Molecular and cellular mechanisms of general anaesthesia. Nature 1994;367:607–14

26. Liang J, Suryanarayanan A, Abriam A, Snyder B, Olsen RW, Spigelman I. Mechanisms of reversible GABAA receptor plasticity after ethanol intoxication. J Neurosci 2007;27:12367–77

27. Antkowiak B, Helfrich-Förster C. Effects of small concentrations of volatile anesthetics on action potential firing of neocortical neurons in vitro. Anesthesiology 1998;88:1592–602

28. Grasshoff C, Drexler B, Hentschke H, Thiermann H, Antkowiak B. Cholinergic modulation of sevoflurane potency in cortical and spinal networks in vitro. Anesthesiology 2007; 106:1147–55