Droperidol (C22H22FN3O2) is a butyrophenone related drug which is used clinically combined with a potent opioid to provide general (neurolept-) anaesthesia, as an efficient antiemetic drug and/or for the treatment of psychiatric disorders. The exact molecular mechanism(s) and site(s) of action of droperidol remain to be identified.
Existing studies demonstrated a significant droperidol mediated inhibition of dopamine DA-2, α1-adrenergic and serotoninergic (5-HT1-3) receptors within a nanomolar concentration range: receptorbinding studies on central dopamine DA-2 receptors showed an effective reduction of the dopamine uptake [1-3]. Furthermore interactions between droperidol and α1-adrenergic, serotoninergic (5-HT1-3) and histamine receptors were demonstrated [3,4]. A recent patch-clamp study showed significant droperidol interactions (concentration range: 1 μM) with voltage activated sodium channels from the peripheral nerve. Peak serum concentrations during neuroleptanaesthesia average around 2 μM (proteinunbound: 0.3 μM), corresponding to 0.7 μg mL−1[6,7]. Nevertheless, all these studies were performed on various animal but not on human preparations, and so far no information about droperidol's molecular mechanisms in a human in vivo or in vitro preparation is available.
As part of our systematic investigation of molecular effects of intravenous (i.v.) anaesthetically active substances, single sodium channels from human brain cortex, key components for neuronal signal integration and transduction , were used in this study as human CNS model protein to quantify the molecular effects of droperidol. Recently, we were able to show that hypnotics (e.g. propofol, pentobarbitone) depressed the sodium channel conductance in a dose range close to clinically relevant concentrations [9-11]. These effects occurred at concentrations predicted by the Meyer-Overton (anaesthetic potency vs. lipophilicity) correlation and indicated a lipophilic site of action for these substances. By contrast, more specific drugs used in anaesthesia (e.g. benzodiazepines, opioids) did not affect sodium channels at clinically relevant concentrations].
It is of interest, therefore, to establish the EC50 value and the concentration range at which droperidol - a specific inhibitor of central receptors (see above) - affects the'human' CNS sodium channel. The EC50 value will be compared with clinically relevant concentrations of droperidol and it will be tested whether the EC50 value follows the Meyer-Overton correlation. This study has been presented in abstract form .
Electrophysiological bilayer experiments were carried out as previously described [9,13,14]. A short summary is given here. Before the study was initiated approval of the institutional ethic committees was given (Committee on human rights in research, Cornell University Medical College, New York, NY, USA and Ethic's Committee of the Medical Faculty, University of Bonn, Germany).
All experiments were performed in symmetrical, 500 mM NaCl buffered at pH 7.4 with 10 mM HEPES, at room temperature. Planar lipid bilayers were formed from neutral phospholipid solutions, containing synthetic (4:1) phosphatidylethanolamine and phosphatidylcholine in decane. Samples of human brain cortex were obtained from neurosurgical waste, material was never removed for the purpose of this study alone. Synaptosomal fractions were prepared from the cortex tissue as described before [9,15]: biological membranes can be broken up in such a way that the remaining fragments reseal to form small membrane vesicles containing sodium channels. These membrane vesicles were added to the vicinity of the lipid bilayers with micropipets in such a way that through a spontaneous process of membrane fusion the sodium channels were transferred from the vesicle into the planar lipid bilayer.
After channel incorporation membrane potentials were applied across the bilayer and the resulting ioncurrent was measured using an experimental set-up described in Fig. 1. Identification and steady-state measurements of sodium channels were carried out in the presence of the sodium channel maker and activator batrachotoxin . The alkaloid-toxin batrachotoxin is commonly used in bilayer experiments removing channel inactivation and increasing channel opening time. Control measurements of the incorporated channels included measurements of the fractional open time, the single channel amplitude and conductance as well as the steady-state activation behaviour. For the measurement of the fractional open times, current traces were recorded by the computer, time-averaged and, after subtraction of the membrane capacitive transient, converted into conductances. The fractional open time, fo, was obtained by subtracting the background conductance through the lipid bilayer from the averaged conductance and divided by the number of channels and the single channel conductance . The steady-state activation behaviour was obtained by plotting a series of fractional open times from potential sequences as a function of the applied membrane potential. Subsequently, these fractional open times were fitted to a two-level Boltzmann distribution by a least squares fitting procedure. The fractional open time, plotted as a continuous function of membrane potential, is referred to as the steady-state activation curve.
After completion of the control measurements, increasing concentrations of droperidol (Janssen Pharmaceuticals, Neuss, Germany: clinical formulation; 0.05-0.8 mM) were added to the intra- or extracellular compartment and the same measurements repeated.
The literature provides contradictory information about the maximal solubility of pure droperidol in aqueous solutions (water, 25°C) ranging from 0.01 mg mL−1 or 0.026 mM  to 0.1 mg mL−1 or 0.26 mM . In order to reach higher concentrations, the clinical formulation of droperidol (containing the solubilizer: tartatric acid and manitol) was used in our experiments. To ensure a full solubilization of droperidol in our study, special experiments measuring the rate of sedimentation/precipitation were performed (ZETAPlus, Brockhaven Instruments, Holtsville USA). These measurements revealed that up to a final droperidol concentration of 1.65 mM no significant increase in the number of complex particles and no precipitation could be observed within the first 150 min. This is in line with results obtained by Janssen laboratories: no precipitation of 2.64 mM droperidol (0.9% NaCl solution, pH 7.4, 25°C) within 12 h .
A total of 12 membranes containing 30 channels (range: 1-5, average 2.5 ± 1.7 channels per membrane) were studied using our standard voltage-clamp protocol (see methods). The investigated channels were identified as sodium channels by tetrodotoxin block, sodium selectivity and reproducible single channel characteristics. The channels were monitored for 31.6 ± 16.0 min (mean ± SD) under control conditions before various concentrations of droperidol (0.05-0.8 mM) were added. Control measurements revealed a reproducible channel behaviour in line with previous studies using batrachotoxin modified sodium channels from human brain cortex [9,13].
The fractional open time (fo) of the channels under control conditions was 0.97± 0.4 (SD), with an average single channel slope conductance of 23.6 ± 3.7 pS (SD) within a membrane potential range between ± 45 mV. The channels showed increasing closing events with hyperpolarizing membrane potentials resulting in a steady-state activation midpoint potential (Va) of −97 ± 12 mV (mean ± SEM, n=4 membranes, 9 channels) and an effective gating charge (za) of 2.7 ± 2.5 (SD).
After addition of various concentrations of droperidol the number and duration of channel closing events significantly increased (Fig. 2). This was the main effect and accounted for about 90% of the overall sodium conductance block. A minor effect resulted in a decrease in the channel amplitude. This current suppression was quantified by time averaging the current traces. Time-averaged conductances were plotted vs. the applied membrane potential and proved to be independent between ± 45 mV (Fig. 3).
The overall blocking effect of sodium channel conductance was plotted vs. increasing droperidol concentrations and then fitted by a weighted computer fit. The resulting concentration-response curve (Fig. 4a) could be approximated by a rectangular hyperbola, yielding a maximal conductance block of 77%, a half-maximal blocking concentration (EC50) of 0.68 mM and an EC10 of 0.08 mM.
As the time-averaged channel conductance block consisted of two distinct mechanisms, the fractional opening time block and the reduction in the channel amplitude were analysed separately. Each effect showed a dependence on droperidol concentration and could also be computer fitted yielding two different concentration-response curves with a maximum block of the fractional open time of 69% (EC50: 0.71 mM, Fig. 4b) and 20% for the channel amplitude block (EC50: 1.28 mm, Fig. 4c).
Furthermore droperidol reduced the time during which the channel could be observed: 0.05 mM droperidol 24.5 ± 5.0 min (mean ± SD); 0.4 mM droperidol 4.5 ± 2.5 min. The steady-state-activation behaviour of the investigated sodium channels did not change significantly with droperidol exposure. The average closing midpoint potential under control conditions was measured to be −98 ± 10 mV (mean ± SEM, 3 membranes per 3 channels). In the presence of 0.1 mM droperidol the average midpoint potential was −104 ± 5 mV for the same (control) channels, not significantly different from the control values.
This initial study of the molecular effects of droperidol on a 'human' CNS ion channel shows two significant and substance-specific interactions with the ion conducting pathway of the channel protein.
Droperidol reduced the time averaged single sodium channel conductance in a concentration-dependent and voltage independent manner: the maximal conductance block was calculated to be 77% with a half-maximal blocking concentration (EC50) of 0.68 mM. This effect was mediated by at least two distinct mech-anisms: (1) major effect (≈90%): reduction of the time averaged fractional channel open time (EC50: 0.71 mM, Blockmax: 69%) and (2) minor effect: reduction of the single channel amplitude (EC50: 1.28 mM, Blockmax: 20%).
These distinct effects of channel block by droperidol represent a spectrum of actions different from other i.v. anaesthetic compounds investigated in the same model system. Pentobarbitone and ketamine merely induced fractional open-time block and had no effects on channel amplitude [9,11], while the hypnotic etomidate and the alcohol ethanol primarily blocked channel conductance via a reduction in the single-channel amplitude[19,20].
Furthermore, high droperidol concentrations (>0.2 mM) reduced the time that sodium channels could be observed in the bilayer as compared with low concentration conditions in a concentration-dependent manner (0.05 mM droperidol: mean 24.5 min; 0.4 mM: mean 4.5 min). This may indicate a droperidol-induced interaction at/with the batrachotoxin binding site of the channel or a drug-induced change in the channel quaternary structure.
These effects occurred at concentrations 100-fold beyond relevant serum concentrations (max. 2μM, protein unbound 0.3 μM) during general anaesthesia . Thus, the human brain sodium channel is unlikely to be a major relevant molecular target for pharmacological actions of droperidol in humans.
Interestingly, the human CNS sodium channels in this study proved to be far more insensitive to droperidol (EC50: 0.68 mM) than peripheral nerve sodium channels from a frog(node of Ranvier) preparation (1.4 μM) . Thus, anaesthetic potency may be species or tissue dependent, which makes the use of human protein preparations mandatory whenever comparative pharmacological studies in humans are proposed.
It has to be taken into consideration that planar bilayer studies of sodium channels still rely on the presence of alkaloid toxins. The extent to which anaesthetic actions are altered in the presence of batrachotoxin  has been discussed before . Nevertheless, batrachotoxin is a unique marker for labelling an individual sodium channel, allowing long observations of channel behaviour of the same, identified single channel.
The established correlation between lipophilicity and anaesthetic potency for several substances used for clinical anaesthesia (Meyer-Overton correlation) suggests a lipophilic mode of action for these substances [11,21]. Assuming a partition coefficient of 3100 for the pure substance  and postulating mainly lipophilic interactions between droperidol and the human brain sodium channel as a site or mechanism of molecular action, an effect concentration of about 0.1 mM should be expected . Droperidol reduced sodium channel conductance with an EC50 of 0.68 mM and an EC10 of 0.08 mM. Thus, the EC50 value is clearly beyond the predicted (Meyer-Overton) concentrations. The partition coefficient of droperidol in the clinical formulation will have to be smaller than for the pure substance. Because its actual value is unknown it is not possible to decide whether droperidol with a corrected value of the partition coefficient would still be outside the Meyer-Overton correlation.
One critical factor in the discussion of these results is the absolute solubility of droperidol in this system. Contradicting information for the solubility of pure droperidol(water, 25°C) range from 0.01 mg mL−1 or 0.026 mM  to 0.1 mg mL−1 or 0.26 mM , but at these measurements (see material and methods) for the rate sedimentation of the clinical formulation of droperidol (with solvents: tartatric acid and manitol) no precipitation could be observed. Thus, effective droperidol concentrations of up to 0.8 mM were actually achieved in our system.
Thus, we conclude: (1) droperidol depresses the human CNS sodium channel with distinct effects but only at high concentrations;(2) at least partly a lipophilic mode of action in this model protein is possible and (3) human CNS sodium channels do not serve as a relevant target protein for droperidol clinical effects, which are more likely mediated by specific interactions with central(dopamine, adrenergic and serotonin) receptors.
We thank Dr J. Daly (NIH, Bethesta, MD, USA) for a generous supply of BTX and Professor Süverkrüp (Institut für Pharmazeutische Technologie, Universität Bonn, Germany) for critical comments and performing the solubility experiments. Zita Dorner is acknowledged for her expert technical and experimental assistance.
1 Creese I, Burt DR, Snyder SH. Dopamine receptor binding predicts clinical and pharmacological potencies of antischizophrenic drugs. Science
2 Peroutka SJ, Snyder SH. Antiemetics: Neurotransmitter receptor binding predicts therapeutic action. Lancet
3 Leysen JE, Gommeren W. Receptor-drug dissociation rate of unlabeled neuroleptics: relation to binding affinity and lipophilicity. Br J Pharmacol
4 Testa R, Abiatti G, Ceserani R. Profile of in vitro affinities of neuroleptics at different rat brain receptors: cluster analysis comparison with pharmacological and clinical profiles. Pharmacol Res
5 Bräu M, Cronenberg R, Vogel W, Hempelmann G. The local anaesthetic effect of droperidol on single ion channels of peripheral nerves. Anaesthesist
6 Fischler M, Bonnet F, Trang H et al.
The pharmacokinetics of droperidol in anesthetized patients. Anesthesiology
7 Lehman KA, Van Peer A, Ikonomakis M, Gasparini R, Heykants J. Pharmacokinetics of droperidol in surgical patients under different conditions of anaesthesia. Br J Anaesth
8 Catterall WA. Structure and function of voltage-sensitive ion channels. Science
9 Frenkel C, Duch DS, Urban BW. Molecular actions of pentobarbital isomers on sodium channels from human brain cortex. Anesthesiology
10 Frenkel C, Urban BW. Human brain sodium channels as one of the molecular target sites for the new intravenous anaesthetic propofol (2,6-diisopropylphenol). Eur J Pharmacol
11 Frenkel C, Duch DS, Urban BW. Effects of i.v. anaesthetics on human brain sodium channels. Br J Anaesth
12 Frenkel C, Radke PW, Urban BW. Molecular interactions of droperidol with single human CNS sodium channels (Abstr.). Eur J Anaesthsiol
13 Duch DS, Recio-Pinto E, Frenkel C, Urban BW. Human brain sodium channels in bilayers. Mol Brain Res
14 Urban BW, Frenkel C, Duch DS, Kauff AB. Molecular models of anesthetic action on sodium channels, including those from human brain. Ann NY Acad Sci
15 Cohen RS, Blomberg F, Berzins K, Siekevitz P. The structure of postsynaptic densities isolated from dog cerebral cortex. I. Overall morphology and protein composition. J Cell Biol
16 Brown GB. Batrachotoxin. A window on the allosteric nature of the voltage-sensitive sodium channel. Int Rev Neurobiol
17 Janssen Pharmaceuticals. Internal research report: Droperdol-physico-chemical properties and solubility in various media. 1995.
18 Merck Index. Merck Co. Inc., Rahway, NY, USA. 1989.
19 Frenkel C, Weckbecker K, Wartenberg HC, Duch DS, Urban BW. Distinct blocking effects of the hypnotic etomidate on human CNS sodium channels (Abstr.). Anesthesiology
20 Frenkel C, Wartenberg HC, Rehberg BW. Interactions of ethanol with single human brain sodium channels. Neurosci Res Commun
21Rehberg B, Xiao YH, Duch DS. Central nervous system sodium channels are significantly suppressed at clinical concentrations of volatile anesthetics. Anesthesiology
22 Urban BW, Duch DS, Frenkel C, Rehberg B, Wartenberg HC. Do general anaesthetics act on specific receptors (in German)? Anästhesiol Intensivmed Notfallmed Schmerzther