Secondary Logo

Journal Logo

Ketamine and propofol differentially inhibit human neuronal K+ channels

Friederich, P.1; Benzenberg, D.1; Urban, B. W.1 2

European Journal of Anaesthesiology: March 2001 - Volume 18 - Issue 3 - p 177-183
Original papers

Background and objective Interaction of intravenous anaesthetic agents with voltage-dependent potassium channels significantly correlates with clinical concentrations. If potassium channels were to play an important part in anaesthesia, one might expect different effects at the molecular level of those anaesthetics that show different clinical effects. Our aim was to analyse the interaction of general anaesthetics with voltage-dependent K channels.

Methods Whole cell patch-clamp experiments were analysed in detail in order to compare the effects of two clinically diverse intravenous hypnotics, ketamine and propofol, on voltage-dependent potassium channels in human neuroblastoma SH-SY5Y cells.

Results Both anaesthetics inhibited the potassium conductance in a concentration-dependent and reversible manner with IC50-values of 300 µM and 45 µM for ketamine and propofol respectively. Whereas ketamine shifted the midpoint of current activation by maximally 14 mV to more hyperpolarized potentials, propofol had the opposite effect on the activation midpoint. Current inhibition by ketamine increased with voltage but decreased with propofol at higher membrane potentials. Propofol but not ketamine induced concentration-dependent but voltage-independent decline, akin to inactivation, of the voltage-dependent potassium channels.

Conclusions The anaesthetics differed not only in their clinical profiles but they also showed differential actions on voltage-dependent potassium channels in several ways. This provides additional evidence for the hypothesis that voltage-dependent potassium channels play an important role in anaesthesia.

1Department of Anaesthesiology and Intensive Care Medicine, University of Bonn, Germany and

2Departments of Anesthesiology and Physiology, Weill Medical College of Cornell University, New York, New York, USA

Correspondence: Dr P. Friederich, Klinik für Anästhesiologie, Martinistr. 52, 20251 Hamburg, Germany.

Accepted October, 2000.

Back to Top | Article Outline


Recent unexpected evidence demonstrates that inhibition of voltage-dependent K currents by intravenous (i.v) anaesthetics significantly correlates with clinical concentrations [1]. Furthermore, voltage-dependent K channels are suppressed at clinically relevant concentrations [1]. Investigating the inhibition of these ion channels may not only help to establish molecular determinants of anaesthetic potency [1] but half-maximal inhibition of these K channels also seem to predict the clinical potency of anaesthetic agents [1]. The interaction of general anaesthetics with this class of ion channels, thus, warrants further analysis. In view of the excellent correlation between clinical anaesthesia and in vitro action on human K channels [1] it would be important to establish if inhibition of K channels may also allow discrimination between clinically diverse anaesthetic agents. If, as suggested by the correlation with clinical potency, K channels were to play an important part in anaesthesia one would expect different effects at the molecular level of those anaesthetics that show different clinical effects.

Ketamine produces a unique anaesthetic state, described as ‘dissociative anaesthesia’ [2]. It is manifest clinically as a spontaneously breathing cataleptic patient with open eyes, intact light corneal reflexes, spontaneous movements unrelated to surgical stimuli and a satisfactory state of hypnosis and analgesia [3]. Propofol is widely used both for induction and maintenance of general anaesthesia [4] without exerting ‘dissociative’ or analgesic actions. The effect of ketamine is attributed to its influence on N-methyl-D-aspartate (NMDA)-receptors [5] and that of propofol to enhancement of the GABAA response [6]. However, both anaesthetics have also been demonstrated to inhibit voltage-dependent K channels [1,7,8]. It is presently unknown if ketamine and propofol exhibit different molecular effects on voltage-dependent K channels.

In view of the unique clinical properties of ketamine, we intended to compare the action of ketamine on these targets with that of propofol. If investigating the suppression of voltage-dependent K channels may help to reveal molecular determinants of anaesthetic action clinically divers agents such as ketamine and propofol are expected to act differently on these molecular targets. It was, therefore, the aim of our study to establish whether the divers clinical profile of both agents may be paralleled by different effects already at the level of human neuronal K channels.

Back to Top | Article Outline


Cell Culture

SH-SY5Y cells were grown in non-confluent monolayer using RPMI medium (Biochrom, Berlin, Germany) at 37°C with 95% air and 5% CO2. Growth medium contained 10% fetal calf serum, glutamine (2 mM), penicillin (100 U Ml−1) and streptomycin (100 µg mL−1). Neuronal differentiation was induced by exposure to retinoic acid (10 µM) for 3–7 days. Treatment of these cells with retinoic acid results in a reduction in cell division and neurite extension [9].

Back to Top | Article Outline

Electrophysiological recordings

Voltage-sensitive outward currents were recorded at room temperature (22–25°C) with an EPC-7 amplifier (List Electronic, Darmstadt, Germany) and pclamp Software Version 5.71 (Axon Instruments, Foster City, CA, USA) using the whole cell patch-clamp technique [10]. Pipettes with an input resistance between 1.8 and 3 MΩ were pulled from borosilicate glass capillary tubes (World Precision Instruments, Saratoga, Fl, USA) and were filled with the following internal solution to record K currents (mM): KCl 115, MgCl2 1, MgATP 3, HEPES 10, EGTA 10, and pH was adjusted to 7.2 with KOH. The extracellular solution contained (mM): NaCl 135, KCl 4, MgCl2 1, CaCl2 2, glucose 10, HEPES 10, and pH was adjusted to 7.4 with NaOH. The holding potential in all experiments was −80 mV, the test potentials were rectangle pulses with a duration of 84 ms increasing from −50 mV to +90 mV in 10 mV steps. Ketamine-HCl (Parke-Davis, Freiburg, Germany) and propofol (Disoprivan®, AstraZeneca, Planckstadt, Germany) were perfused on the cells using a hydrostatic perfusion system. The carrier of propofol (Intralipid 10%) had no effect on the K currents. The concentration of propofol in the cell bath was measured by high-performance liquid chromatography (HPLC) (RF-10 AXL, Shimatzu, Duisburg, Germany). The bath level was kept constant with a level control driven pump (MPCU-3, Lorenz, Lindau, Germany). The recorded signal was filtered at 3–10 kHz, digitized using an analogue-to-digital converter (Digidata 1200, Axon Instruments, Foster City, CA, USA) and stored on a 3/86 IBM-compatible personal computer with a sampling rate of 5 kHz for later analysis. The results of this study are based on the electrophysiological recordings presented by Friederich and Urban [1].

Back to Top | Article Outline

Data analysis

For the analysis of our data current-voltage curves were converted by the algorithm presented below to conductance-voltage curves. Ionic currents result from changes of the ion channel protein and from changes in the driving force for the ion that permeates the ion channel. Changes of the ion channel conductance reflect direct changes imposed on the ion channel protein. The maximal conductance describes the moment when all ion channels are in the open state (gKmax). The membrane potential at which half of the channels are in the open state is called the activation midpoint (Vmid). Changes of both parameters caused by the anaesthetics were analysed. For this purpose, the maximal amplitudes of the K current were measured by fitting mono-exponential functions to the current traces using pclamp 6.0 Software (Axon Instruments, Foster City, CA, USA). The amplitudes of these functions represented the current maximum at each test potential. The size of these amplitudes was then measured with the same software and used for further analysis. Activation curves were obtained by fitting K conductances corresponding to the maximal current amplitudes (gK = IK/[V-VK], where VK = − 84 mV) with the Boltzmann-function (gK = gKmax/{1 + exp[-zaF(V-Vmid)/RT]}), where gKmax is the maximal K conductance, za is the effective charge or slope factor, V is the membrane potential, Vmid is the midpoint of activation, T is the temperature, and F and R are the Faraday and gas constant. Inhibition of K conductances was measured as 1–gKmax in the presence of the drug in relation to the mean of gKmax under control conditions and after wash out of the drug effect. Inhibition of steady-state current after 50–60 ms by propofol is also given. This was necessary as inhibition of K conductance and steady-state current differed akin to inactivation induced by propofol but not by ketamine. Concentration-dependent shifts of Vmid were measured as the difference between Vmid in the presence of the drug and the mean of Vmid under control conditions and after wash out of the drug effect. The concentration-response curves were established using non-linear regression [11] and Sigma-Plot 4.0 Software (Sigma-Plot, Jandel Scientific, Erkrath, Germany). The measured inhibition of gKmax and the measured shift of Vmid where fitted to the Hill equation (b/Bmax = cy/[IC50y + cy]). Here b = block of gKmax or shift of Vmid, Bmax = maximal block of gKmax or maximal shift of Vmid, c = concentration, y = Hill coefficient and IC50 = concentration of half-maximal effect. Time-constants of anaesthetic-induced K current decline were established by fitting monoexponential functions to the K currents after they had reached their maximum. Statistical comparison was performed with paired or unpaired Student's t-test as appropriate and considered significant for P ≤ 0.01. Data are always shown as mean ± SEM.

Back to Top | Article Outline


Original recordings of the K currents under control condition, under propofol and ketamine, and after the wash out of the drugs are shown in Figure 1. Propofol and ketamine differentially affected the time course of the currents during the test pulse (Figure 1). Propofol caused a pronounced current decline or inactivation-like behaviour that was not observed under control condition. This decline was observed at all membrane potentials between +10 mV and +70 mV. The time constants of the anaesthetic-induced decline were dependent on drug concentration but not on voltage. Time-constants ranged between 2 and 5 ms at membrane potentials of maximal current activation (+40 to +70 mV) and they were more than 100 times faster compared with time constants of K current inactivation without propofol (n = 31).

Figure 1.

Figure 1.

The induction of inactivation-like behaviour by propofol allowed discrimination between inhibition of peak current at the beginning of the test pulse and inhibition of the K current after 50–60 ms of the test pulse. In order to quantify this time-dependent effect the ratio of maximal current inhibition to current inhibition after 50–60 ms of the test pulse (I50-60) was calculated for both anaesthetics at concentrations that caused nearly 50% inhibition of the K currents. Whereas ketamine (250 µM) suppressed current maxima and I50-60 equally as well (47 ± 2% vs. 45 ± 1%, mean ± SEM, n = 5, P ≥ 0.01), inhibition of I50-60 by propofol (43 µM) was 1.6 times higher (P ≤ 0.01) than inhibition of the current maxima (44.9 ± 0.02% vs. 28.9 ± 0.02%, mean ± SEM, n = 9, P ≤ 0.01).

For the comparison of the action of both drugs, current-voltage curves were generated for inhibition of the maximal outward currents and subsequently converted to conductance-voltage curves (see Methods). This allowed comparison of the action of both drugs independent of time. Additionally, a separate analysis of concentration-dependent effects of ketamine and propofol on K conductance and current activation was possible.

The maximal K conductance (GKmax) and the midpoints of current activation (Vmid) of the control currents did not differ significantly between ketamine and propofol experiments. GKmax for the control currents of ketamine and propofol experiments was 3.75 ± 0.16 and 4.31 ± 0.17 (mean ± SEM, n = 29, 34, P > 0.01). Vmid was 7.54 ± 1.31 and 7.00 ± 0.69 for the respective experiments (mean ± SEM, n = 25, 34, P > 0.01). Both anaesthetics inhibited GKmax in a concentration-dependent and reversible manner (Figure 2, white symbols). As estimated by their respective IC50-value, ketamine inhibited GKmax 5–6 times more potently than propofol. The concentration-response curve for inhibition of I50-60 by propofol is also shown in order to illustrate the time-dependent drug effect of propofol (Figure 2, black symbols, data taken from [1]).

Figure 2.

Figure 2.

Propofol, as well as ketamine, shifted the midpoint of current activation along the voltage axis in a concentration-dependent and reversible manner (Figure 3). Vmid was, however, shifted in opposing directions by both drugs. Whereas propofol caused a depolarizing shift of Vmid by maximally 15 mV, ketamine maximally shifted Vmid by the same amount towards more hyperpolarized membrane potentials (Figure 3). The voltage dependence of conductance block as analysed at concentrations of propofol and ketamine close to their respective IC50-values differed between both anaesthetics, as well. Inhibition of the K conductance by propofol decreased with increasing test potentials. This decrease could be described by regression analysis. The slope of the regression line was −1.3 and the regression coefficient r2 was 0.9 (n = 7). In contrast to propofol, inhibition by ketamine increased with voltage. This increase could best be described with a function of exponential rise [inhibition = a(1-e–bx) with values for a and b of 4.9 e−1 and 1.4 e−1, respectively;r2 = 0.9, n = 5].

Figure 3.

Figure 3.

Back to Top | Article Outline


It had previously been shown that i.v. anaesthetics, including propofol and ketamine, reversibly inhibited human neuronal K currents in SH-SY5Y cells. This suppression correlated with their clinical concentrations [1]. The detailed analysis presented in this paper provides evidence that propofol and ketamine differ in their actions on potassium currents in several ways. The pattern of divers pharmacological action contradicts the view of non-specific anaesthetic interaction of propofol and ketamine with this family of human K channels. Both drugs differ in their effects on current activation (Figure 3) that would lead to opposite effects on the threshold of action potentials. As a result of the inactivation-like behaviour, propofol will have a greater impact on the refractory time than ketamine when inhibition is normalized to peak current inhibition (Figures 1 and 2). Furthermore, as the membrane potential changes during an action potential the different voltage-dependence of inhibition would lead to different effects of propofol and ketamine on the shape of the action potential.

Systemic consequences of molecular effects are difficult to predict from in vitro experiments. However, the results show that diverse clinical profiles are reflected already at the molecular level of human K channels. SH-SY5Y cells express voltage-dependent K channels of the Kv3.1 subtype [12]. Kv 3.1 channels are, for example, expressed in GABAergic interneurones of the hippocampus [13]. As these ion channels are critical for the integration of temporal information in central neurones [14], differential pharmacological effects on these K channels may also cause different responses within these neuronal networks.

At first, the small effects at clinically relevant concentrations [1] may suggest that inhibition of this type of voltage dependent K channel is not involved in pharmacological action during clinical anaesthesia. However, it is increasingly recognized that small effects on the molecular level may be modulated on the way up the integrative structure of the entire central nervous system [15,16]. Ion channel proteins are integrated with other membrane proteins within a single cell. Single cells form neuronal networks with other cells, and these neuronal networks in turn are integrated into higher functional units. A concentration-response curve at the bottom of this hierarchical structure does not need to be identical to the concentration-response curve at the top [15,16].

Anaesthetic agents exert their effects at each level of the central nervous system (Figure 4). At the molecular level, the anaesthetic effects are described by the right most concentration-response curve. Because the neuronal output of the lower level is the input of the next higher level (Figure 4), the concentration-response curve of the higher level would be multiplied with the concentration-response curve of the preceding level [15,16]. Therefore, depending on the level of integration, the composite concentration-response curve will constitute the result of an increasing number of multiplications by small effect concentration-response curves. As a consequence the composite concentration-response curve will show larger and larger effects by shifting to the left on the concentration axis and becoming increasingly steeper (Figure 4, left concentration-response curve).

Figure 4.

Figure 4.

How much alteration of a molecular target is needed to account for changes at the highest level of neuronal integration is unclear. However, examples of small molecular effects that add up to severe systemic effects have been given. Modification of a very small number (1–2%) of voltage-dependent ion channels can have severe clinical consequences such as myotonia and familial periodic paralysis [17]. On the other hand, because of the wide margin of safety of neuromuscular transmission, more than 70% of the acetylcholine receptor must be occupied by relaxants before any mechanical weakness results [18]. Thus, a small pharmacological effect at the molecular level can not be dismissed as being relevant for clinical anaesthesia [15,16].

We find several differences between propofol and ketamine on the molecular level. To interpret their meaning clinically we would have to know the neuronal networks involved in anaesthetic drug effects. Until now, these networks have not been identified. K channel alterations, for several reasons, reflect clinical anaesthetic action. Therefore, establishing anaesthetic drug effects on these molecular targets remains an important strategy in identifying molecular determinants of general anaesthesia.

In summary, inhibition of human neuronal K channels by propofol and ketamine resulted from qualitative distinct effects. Ketamine and propofol differ not only in their clinical profiles but they also show differential actions at the molecular level. Our results provide further support for the hypothesis that Kv3.1 channels constitute a molecular target relevant for clinical anaesthesia.

Back to Top | Article Outline


1 Friederich P, Urban BW. Interaction of intravenous anesthetics with human neuronal potassium currents in relation to clinical concentrations. Anesthesiology 1999; 91: 1853–1860.
2 Corssen G, Domino EF. Dissociative anesthesia: further pharmacologic studies and first clinical experience with the phencyclidine derivative CI-581. Anesth Analg 1966; 45: 29–40.
3 White PF, Way WL, Trevor AJ. Ketamine-its pharmacology and therapeutic use. Anesthesiology 1982; 56: 119–136.
4 Smith I, White PF, Nathanson M, Gouldson R. Propofol. An update on its clinical use. Anesthesiology 1994; 81: 1005–1043.
5 Yamamura T, Harada K, Okamura A, Kemmotsu O. Is the site of action of ketamine anesthesia the N-methyl-D-aspartate receptor? Anesthesiology 1990; 72: 704–710.
6 Hara M, Kai Y, Ikemoto Y. Enhancement by propofol of the gamma-aminobutyric acid. A response in dissociated hippocampal pyramidal neurons of the rat. Anesthesiology 1994; 81: 988–994.
7 Bartschat DK, Blaustein MP. Phencyclidine in low doses selectively blocks a presynaptic voltage-regulated potassium channel in rat brain. Proc Natl Acad Sci USA 1986; 83: 325–330.
8 Baum VC. Distinctive effects of three intravenous anesthetics on the inward rectifier (IK1) and the delayed rectifier (IK) potassium currents in myocardium: implications for the mechanism of action. Anesth Analg 1993; 76: 18–23.
9 Pahlman S, Ruusala AI, Abrahamsson L et al. Retinoic acid-induced differentiation of cultured human neuroblastoma cells: a comparison with phorbol ester induced differentiation. Cell Differ 1984; 12: 165–170.
10 Hamill OP, Marty A, Neher E et al. Improved patch-clamp techniques for high resolution current recording from cells and cell-free membrane patches. Pflügers Arch 1981; 391: 85–100.
11 Seber GAF, Wild CJ. Nonlinear Regression. New York: John Wiley. 1989.
12 Friederich P, Dilger JP, Pongs O, Urban BW. Kv3.1 expression in human neuroblastoma SH-SY5Y cells. Pflügers Arch 2000; 439: R427.
13 Brew HM, Forsythe ID. Two voltage-dependent K+ conductances with complementary functions in postsynaptic integration at a central auditory synapse. J Neurosci 1995; 15: 8011–8022.
14 Rudy B, Chow A, Lau D et al. Contributions of Kv3 channels to neuronal excitability. Ann NY Acad Sci 1999; 868: 304–343.
15 Urban BW, Friederich P. Anesthetic mechanisms in-vitro and in general anaesthesia. Toxicol Lett 1998; 100–101: 9–16.
16 Eckenhoff RG, Johansson JS. On the relevance of ‘clinically relevant concentrations’ of inhaled anesthetics in in vitro experiments. Anesthesiology 1999; 91: 856–860.
17 Cannon SC, Brown RH, Corey DP. Theoretical reconstruction of myotonia and paralysis by incomplete inactivation of sodium channels. Biophys J 1993; 65: 270–288.
18 Paton WD, Waud DR. The margin of safety of neuromuscular transmission. J Physiol (Lond) 1967; 191: 59–90.


© 2001 European Academy of Anaesthesiology