Anesthesiology 2005; 102:93–101 © 2004 American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins, Inc. Molecular Mechanisms Underlying Ketamine-mediated Inhibition of Sarcolemmal Adenosine Triphosphate- sensitive Potassium Channels Takashi Kawano, M.D.,* Shuzo Oshita, M.D.,† Akira Takahashi, M.D.,‡ Yasuo Tsutsumi, M.D.,* Katsuya Tanaka, M.D.,§ Yoshinobu Tomiyama, M.D.,࿣ Hiroshi Kitahata, M.D.,# Yutaka Nakaya, M.D.** Background: Ketamine inhibits adenosine triphosphate-sen- sensitive channel pore.3 These channels, as metabolic sitive potassium (KATP) channels, which results in the blocking sensors, are associated with such cellular functions as of ischemic preconditioning in the heart and inhibition of va- insulin secretion, cardiac preconditioning, vasodilata- sorelaxation induced by KATP channel openers. In the current 4–7 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/102/1/93/357819/0000542-200501000-00017.pdf by guest on 27 September 2021 study, the authors investigated the molecular mechanisms of tion, and neuroprotection. ketamine’s actions on sarcolemmal KATP channels that are re- In cardiac myocytes, intravenous general anesthetics, associated by expressed subunits, inwardly rectifying potas- such as ketamine racemate, propofol, and thiamylal, di- sium channels (Kir6.1 or Kir6.2) and sulfonylurea receptors 8–10 rectly inhibit native sarcolemmal KATP channels. Al- (SUR1, SUR2A, or SUR2B). though these observations suggest that intravenous an- Methods: The authors used inside-out patch clamp configura- tions to investigate the effects of ketamine on the activities of esthetics may impair the endogenous organ protective reassociated Kir6.0/SUR channels containing wild-type, mutant, mechanisms mediated by KATP channels, the possibility or chimeric SURs expressed in COS-7 cells. that KATP channel inhibition by intravenous anesthetics Results: Ketamine racemate inhibited the activities of the might have adverse consequences in clinical practice reassociated KATP channels in a SUR subtype-dependent man- ␮ remains controversial. Indeed, propofol and thiamylal ؍ ␮ ؍ ner: SUR2A/Kir6.2 (IC50 83 M), SUR2B/Kir6.1 (IC50 77 M), -␮ are known to possess cardioprotective and neuroprotec ؍ ␮ ؍ SUR2B/Kir6.2 (IC50 89 M), and SUR1/Kir6.2 (IC50 1487 M). S-(؉)-ketamine was significantly less potent than ketamine tive properties, respectively, with these being mediated racemate in blocking all types of reassociated KATP channels. by other well-established mechanisms that do not in- The ketamine racemate and S-(؉)-ketamine both inhibited 11,12 volve KATP channels. In recent in vitro studies, how- channel currents of the truncated isoform of Kir6.2 ever, ketamine racemate, but not the stereoisomer S-(ϩ)- (Kir6.2⌬C36) with very low affinity. Application of 100 ␮M mag- nesium adenosine diphosphate significantly enhanced the in- ketamine, was found to block early and late hibitory potency of ketamine racemate. The last transmem- preconditioning in rabbit hearts and inhibit vasorelax- brane domain of SUR2 was essential for the full inhibitory effect 13–16 ation induced by a KATP channel opener. It is there- of ketamine racemate. fore possible that the mechanisms underlying ketamine- Conclusions: These results suggest that ketamine-induced in- induced inhibition of KATP channel activity may differ hibition of sarcolemmal KATP channels is mediated by the SUR subunit. These inhibitory effects of ketamine exhibit specificity from those of propofol and thiamylal. A previous mu- for cardiovascular KATP channels, at least some degree of ste- tagenesis study demonstrated that the major effects of reoselectivity, and interaction with intracellular magnesium both propofol and thiamylal on KATP channel activity are adenosine diphosphate. mediated via the Kir6.2 subunit.17 However, organ spec- ificity and the molecular site of action of ketamine have ADENOSINE triphosphate-sensitive potassium (KATP) not been investigated in detail. In addition, although it is channels are inhibited by intracellular adenosine triphos- now well established that intracellular MgADP can mod- phate (ATP) and activated by magnesium adenosine ulate the sensitivity of KATP channel activators and inhib- diphosphate (MgADP) and thus provide a link between itors,18,19 there is no evidence that intracellular MgADP 1,2 the cellular metabolic state and excitability. KATP modulates intravenous anesthetics inhibitory effects on channels are composed of an ATP-binding cassette pro- sarcolemmal KATP channels. tein, sulfonylurea receptor (SUR), and an inwardly recti- In the current study, we used patch clamp techniques ϩ fying K channel (Kir) subunit, Kir6.0; SUR acts as a to examine the electrophysiological effects and molecu- regulatory subunit whereas Kir subunits form the ATP- lar mechanisms of racemic ketamine and S-(ϩ)-ketamine on different types of reassociated KATP channels contain- * Resident, † Professor and Chairman, ࿣ Assistant Professor, # Associate Pro- ing wild-type, mutant, or chimeric SURs expressed in fessor, § Instructor, Department of Anesthesiology, Tokushima University School COS-7 cells (African green monkey kidney cells). We also of Medicine, Tokushima, Japan; ‡ Associate Professor, ** Professor and Chairman, Department of Nutrition, Tokushima University School of Medicine, Tokushima, investigated the effects of intracellular MgADP on the Japan. inhibitory actions of ketamine racemate. Received from the Department of Anesthesiology, Tokushima University School of Medicine, Tokushima, Japan. Submitted for publication April 26, 2004. Accepted for publication August 30, 2004. Supported in part by a Grant-in Aid for Scientific Research (C, 15591636) from the Japan Society for the Promotion of Materials and Methods Science, Tokyo, Japan. Address correspondence to Dr. Kawano: Department of Anesthesiology, To- Molecular Biology kushima University School of Medicine, 3-18-15 Kuramoto, Tokushima 770-8503, Japan. Address electronic mail to: [email protected]. Individual article re- cDNAs (The human Kir6.2, rat Kir6.1, rat SUR1, rat prints may be purchased through the Journal Web site, www.anesthesiology.org. SUR2A, and rat SUR2B) and expression vector pCMV6C Anesthesiology, V 102, No 1, Jan 2005 93 94 KAWANO ET AL. were kindly provided by Susumu Seino, MD., Ph.D. (Pro- The sampling frequency of the single-channel data were fessor and Chairman, Department of Cellular and Molec- 5 KHz with a low-pass filter (1 KHz). ular Medicine, Chiba University, Chiba, Japan). Coex- pressing SUR1 and Kir6.2 (SUR1/Kir6.2) forms the Electrophysiological Data Analysis ␤ pancreatic cell KATP channel, SUR2A and Kir6.2 Channel currents were recorded with a patch clamp (SUR2A/Kir6.2) forms the cardiac KATP channel, SUR2B amplifier (CEZ 2200; Nihon Kohden, Tokyo, Japan) and and Kir6.2 (SUR2B/Kir6.2) forms the nonvascular stored in a personal computer (Aptiva; IBM, Armonk, smooth muscle KATP channel, and SUR2B and Kir6.1 NY) with an analog-to-digital converter (DigiData 1200; (SUR2B/Kir6.1) forms the vascular smooth muscle KATP Axon Instruments, Foster City, CA). pClamp version 7 3,20,21 channel. A truncated form of human Kir6.2 lacking software (Axon Instruments) was used for data acquisi- the last 36 amino acids at the C terminus was obtained tion and analysis. The open probability (Po) was deter- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/102/1/93/357819/0000542-200501000-00017.pdf by guest on 27 September 2021 by polymerase chain reaction amplification as previously mined from current amplitude histograms and was cal- 17 described. Chimeric cDNA constructs were produced culated as follows: by splicing, using the overlap extension polymerase chain reaction technique.22 The exact amino acid com- N ͑͸ ⅐ ͒ position of the SUR1-SUR2A chimeric constructs was: tj j ϭ chimera SUR1–2A ϭ (1–1035, SUR1)-(1013–1261, ϭ j 1 Po ⅐ (1) SUR2A)-(1297–1581, SUR1); SUR2A-1 ϭ (1–1013, Td N SUR2A)-(1035–1277, SUR1)-(1241–1545, SUR2A). All DNA products were sequenced using BigDye terminator cycle sequencing kit (Applied Biosystems, Foster City, where tj is the time spent at current levels corresponding ϭ CA), and an ABI PRISM 377 DNA sequencer (Applied to j 0, 1, 2, N channels in the open state, Td is the Biosystems) was used to confirm the sequence. duration of the recording, and N is the number of the channels active in the patch. Recordings of 2–3 min were analyzed to determine P . The channel activity was Cell Culture and Transfection o expressed as NPo. The NPo in the presence of drugs was K channel-deficient COS-7 cells were plated at a ATP normalized to the baseline NPo value obtained before density of 3 ϫ 105 per dish (35 mm diameter) and drug administration and presented as the relative chan- cultured in Dulbecco’s modified Eagle’s medium supple- nel activity. When the concentration-dependent effects mented with 10% fetal calf serum. A full-length Kir cDNA of drugs were studied, the superfusion was stopped for and a full-length SUR cDNA were subcloned into the approximately 1 min at each concentration, and these mammalian expression vector pCMV6c. For electrophys- drugs were injected into the cell bath using a glass iological recordings, mutated pCMV6c Kir alone (1 ␮g) syringe to five final concentrations in a cumulative man- or either wild-type or mutated pCMV6c Kir (1 ␮g) plus ner (total volume injected was approximately 10–20 ␮l). pCMV6c SUR (1 ␮g) were transfected into COS-7 cells Therefore, the superfusion was stopped for approxi- with green fluorescent protein