The molecular site of volatile anesthetic action remains an unsolved question despite >150 yr of investigation. The leading hypothesis is that ion channels within excitable membranes of the central nervous system (CNS) are modulated by these drugs to cause immobility in response to noxious surgical stimulation (minimum alveolar anesthetic concentration [MAC]). Over the past 20 yr, many ion channels found in the spinal cord have been studied for their potential role in MAC. Most of these channels, except for glycine receptors, N-methyl-d-aspartate receptors, and potassium (K+) channels have now been excluded as likely targets (1). The same ion channels may also mediate the suppression of higher CNS functions such as wakefulness, awareness, explicit and implicit memory produced by volatile anesthetics.
Inhibitory ion channels whose activities are enhanced by volatile anesthetics include γ-aminobutyric type A receptors, glycine receptors, and background K+ channels. Enhancement of background K+ channel activity remains a highly plausible mechanism by which volatile anesthetics depress the CNS (2). The tandem-pore domain family of K+ channels (K2P) is the principal mediator of background K+ currents. The first mammalian member of this family isolated is called TWIK-1 (tandem pore weak-inward rectifier K channel) and the family now comprises 15 members. The currents of 7 of these K2P channels are enhanced by volatile anesthetics: TREK-1 (TWIK-related K channel), TASK-1 (TWIK-related acid-sensitive K channel), TASK-2, TASK-3, TREK-2, TALK-2 (tandem pore domain alkaline-activated K channel), and TRESK (TWIK-related spinal cord K channel). TRESK is the most recently isolated family member (3), and was originally isolated from human spinal cord RNA but is also expressed in human brain (4). Human TRESK is more strongly enhanced by volatile anesthetics than any of the other K2P channels tested (4).
Therefore, we have investigated the pharmacology of TRESK cloned from other species in order to better understand species-specific differences in function and pharmacology. In this article, we report the cloning and expression of rat and mouse TRESK and compare their responses to volatile anesthetics, local anesthetics, pH, and zinc ions (Zn+2). In addition, we studied the effects that the enantiomers of isoflurane have on functional rodent and human TRESK channels because previous studies have shown different sensitivity of mice (5) and rats (6) to the isoflurane enantiomers.
All animal experiments were done according to protocols approved by the Institutional Animal Care and Use Committee of the University of California San Francisco.
Rat and mouse TRESK genes were identified by searching their respective genome databases with the human TRESK sequence using the basic local alignment search tool (BLAST) (7). A three-stage overlapping extension strategy using reverse transcription coupled to the polymerase chain reaction (RT-PCR) was devised for cloning rat and mouse TRESK. The first step involved reverse transcribing rat and mouse spinal cord poly(A)+ RNA (Clontech, Palo Alto, CA) as template RNA using the Superscript One-Step RT-PCR kit (Invitrogen, Carlsbad, CA). The RT reaction was primed with a DNA oligonucleotide complementary to the 3′ end of the predicted TRESK sequence. The first round of the PCR was performed using oligonucleotides complementary to the 5′ and 3′ ends of the predicted TRESK sequence. For mouse 5′-ATGGAGGCTGAGGAGCCAC-3′ was the 5′ primer, and 5′-TTACCAAGGTAGCGAAACTTCCCT-3′ was the 3′ primer. For rat: 5′-ATGGAGGCTGAGGAGCCA-3′ was the 5′ primer, and 5′-TTAGCAAGGTAGCGAAACCTCTC-3′ was the 3′ primer. The second round of PCR was performed with the aim of increasing the specificity of the PCR products by using two new sets of primers with the PCR products from the first reaction as templates. For mouse, this set of primers consisted of 5′-ATGGAGGCTGAGGAGCCA-3′ as 5′ primer and 5′-CTCCACATTGCAGCTCGG-3′ as 3′ primer. The other set of primers used 5′-CCGAGCTGCAATGTGGAG-3′ as 5′ primer and 5′-TTACCAAGGTAGCGAAACTTCCCT-3′ as 3′ primer. For rat, this set of primers consisted of 5′-ATGGAGGCTGAGGAGCCA-3′ as 5′ primer and 5′-TGGCTTCATCTGCGGGTTTG-3′ as 3′ primer. The other set of primers used 5′-TGACAGCAAACCCGCAGATG-3′ as 5′ primer and 5′-TTAGCAAGGTAGCGAAACCTCTC-3′ as 3′ primer. The PCR products generated from these two reactions produced overlapping fragments that spanned the whole TRESK coding sequence. These products were separated by agarose gel electrophoresis and the DNA bands were isolated from the gel by using the Qiagen fragment purification kit (QIAGEN Inc., Valencia, CA). The final round of PCR was performed to assemble the full-length TRESK gene, for mouse using 5′-CACCATGGAGGCTGAGGAGCCAC-3′ as 5′ primer, 5′-TTACCAAGGTAGCGAAACTTCCCT-3′ as 3′ primer, and for rat using 5′-CACCATGGAGGCTGAGGAGCCA-3′ as 5′ primer, and 5′-TTAGCAAGGTAGCGAAACCTCTC-3′ as 3′ primer with gel-purified PCR products as template. The PCR reactions were conducted under the following conditions: 94°C for 15 s, 52°C for 30 s, and 72°C for 120 s for 35 cycles. This overlapping extension technique produced a single band that was gel-purified and subcloned into pcDNA3.1-TOPO vector (Invitrogen). Identities of the final cloned sequences with the predicted mouse and rat TRESK sequences were confirmed by sequencing the generated full-length fragments.
Various point mutations were made in the coding sequences of human, mouse, and rat TRESK using PCR-based overlapping extension using a high-fidelity DNA polymerase (AccuPrime Pfx DNA Polymerase; Invitrogen). Mutants were confirmed by DNA sequencing.
Total RNA was isolated from surgically resected tissue samples by homogenization in Trizol reagent (Sigma, St. Louis, MO). Each RNA sample was treated with DNAse I to avoid DNA contamination. The cDNA was synthesized by using the Moloney murine leukemia virus reverse transcriptase enzyme. Template cDNA was added to Taqman Universal Master Mix (Applied Biosystems, Foster City, CA) in a 12.5-μL reaction with specific primers and probe for each gene. The primer and probe sets were designed using primer Express 2.0 Software (Applied Biosystems). Quantification of specific transcript levels was performed using the ABI Prism 7900HT Sequence Detection System (Applied Biosystems). Mouse TRESK primers and 5′-labeled fluorescent reporter dye (6FAM) probe were as follows: forward 5′-CTGTTCATTGCCTTCAAGCTGAT-3′, reverse 5′-TTGGCAAACAAACAGCATGAG-3′, probe 5′-CAGAACCGGCTCCTGCACACCTACAA-3′. The conditions for cycling were: 95°C for 10 min then 40 cycles of 95°C for 15 s followed by 60°C for 1 min. Relative gene expression quantification was calculated according to the comparative cycle threshold method using glyceraldehyde-3-phosphate dehydrogenase as an endogenous reference. For each sample analyzed, a reverse transcriptase minus control was run to assure the absence of genomic DNA contamination.
cRNA transcripts were synthesized from linearized cDNA template of TRESK using T3 RNA polymerases (mMessage mMachine; Ambion, Austin, TX). Defolliculated Xenopus laevis oocytes were injected with 1–15 ng of cRNA using standard methods for oocyte preparation and maintenance (8). One to four days after injection, two-electrode voltage clamp recordings were performed at room temperature (GeneClamp 500B; Axon Instruments, Foster City, CA). Most voltage pulse protocols were applied from a holding potential of −60 mV using 1-s voltage-pulse steps ranging from −120 to +60 mV in 20-mV increments, with 1.5-s interpulse intervals. In some experiments, longer voltage steps (up to 2 s) with higher sampling rates were used. Except where noted, all two-electrode voltage clamp experiments were performed using frog Ringer's solution (in mM: 115 NaCl, 2.5 KCl, 1.8 CaCl2, 10 HEPES, pH 7.4) as perfusate. Recordings were obtained in a 25-μL recording chamber at flow rates of 1–4 mL/min. Signals were filtered using a low-pass filter set at a 50–100 Hz cutoff before sampling at 100–1000 Hz. Water-injected oocytes were used as controls, undergoing the same treatment as transcript-injected oocytes.
Racemic isoflurane (Baxter Healthcare, New Providence, NJ) and halothane with 0.01% thymol (w/w) (Ayerst Laboratories Inc., Philadelphia, PA and Halocarbon Laboratories, Augusta, SC) were used for the experiments. The enantiomers of isoflurane were provided as a gift from Baxter Healthcare, and were 99.8% optically pure. The nonimmobilizer 1,2-dichlorohexafluorocyclobutane (“2N”; Lancaster Synthesis Inc., Windham, NH) was also studied (9). Methods for volatile anesthetic delivery and analysis have been previously described (10). Briefly, solutions and their dilutions to the experimental concentrations were prepared immediately before use. Saturated stock solutions of racemic, S(+) and R(−) isoflurane were prepared by adding approximately 10 mL of racemic volatile anesthetic or 250 μL of an enantiomer to an airtight glass bottle with 200 mL or 50 mL frog Ringer's solution, respectively. An equilibration period of 24 h at room temperature was allowed. Aliquots were taken from the bottle through a needle with its tip in a fixed position close to the phase of volatile anesthetic to assure a constant concentration. Volatile anesthetics were applied using an airtight perfusion system including glass syringes and Teflon tubing. Their concentrations in the recording chamber were determined by gas chromatography (GowMac Series 750; GowMac, Bethlehem, PA) from samples taken at the outlet of the perfusion system. Isoflurane enantiomers, applied at a clinical concentration (approximately 1.0 MAC, 270 μM) were also tested on oocytes expressing mouse, rat, and human TRESK in a paired exposure paradigm. Baseline responses were first determined then retested after being superfused in random order with either the S(+) or R(−) isomer. After washout, the effect of the other isomer was determined. Thus, each oocyte was tested with both drugs to identify differential sensitivity.
Lidocaine and racemic bupivacaine were purchased from Astra Pharmaceuticals (Söodertölje, Sweden). Stock solutions of local anesthetics were prepared in frog Ringer's solution and kept at 4°C for no more than 4 weeks. Stock solutions of Zn+2 were freshly made before each experiment.
Except where noted, data are reported from at least three oocytes. The background current response is defined as the current measured for the –60 to +60 mV pulse during the treatment condition relative to the control condition. These currents are defined as the current measured 430 ms after the voltage jump. Control currents are the average currents before and after treatment. Mean values are expressed ± se with n values indicating the number of oocytes studied. Concentration response curves were derived by nonlinear least square fit performed by the graphing and statistical package Prism 3.0a (Prism, San Diego, CA) providing 50% effective concentration and Hill coefficient (nHill) values. For comparison of enantiomeric differences, statistical significance was determined using a Student's paired t-test. Statistical significance is defined by P < 0.05.
A search of the mouse genome database for sequences related to the human TRESK sequence identified a single gene composed of 3 exons on mouse chromosome 19 whose coding sequence showed 65% identity and 71% overall amino acid similarity to the human gene. Using BLAST analysis of the draft-rat sequence available at the Baylor University website, we discovered a sequence nearly identical to the mouse gene (92% amino acid identity). Reverse BLAST query of the human genome with either mouse or rat protein or nucleotide sequence identified no gene other than the original TRESK sequence.
Full-length cDNAs were amplified from mouse and rat total RNA by RT-PCR using species-specific primers. The full-length cDNA sequences are deposited in GenBank under accession numbers AY542902 (mouse) and AY567970 (rat). The mouse sequence was identical to the previously reported mouse TRESK cDNA (11,12). Sequence analysis predicted a 394 amino-acid protein for the mouse clone and a 405 amino-acid protein for the rat clone with intracellular amino- and carboxy-termini, four predicted transmembrane (TM) sequences, and two K+ pore sequences between TM1/2 and TM3/4. The difference between the mouse and rat clones is an extra 11 amino-acid repeat sequence present in the amino-terminus of the rat clone. As in human TRESK, a long sequence (91 residues) in rodent TRESK between TM2 and TM3 is predicted to be an intracellular domain not found in other K2P family members.
We expressed the rodent TRESKs by injection into Xenopus oocytes of in vitro synthesized cRNA made from the cDNA templates. Both mouse and rat TRESK passed large (>2 μA at the +60 mV pulse) outwardly rectifying currents over the voltage range from –120 to +60 mV. Healthy uninjected and water-injected oocytes displayed minimal background currents (<100 nA at +60 mV). The volatile anesthetic isoflurane enhanced currents passed by both mouse and rat TRESK, qualitatively similar to what we have described for human TRESK (Fig. 1A). Figure 1B shows the concentration-response curves for isoflurane enhancement of mouse and rat TRESK. The concentration of isoflurane producing 50% of the maximal effect (EC50) was 226 μM for mouse TRESK and 346 μM for rat TRESK; the Hill coefficient of the fitted curve was 2.6 for mouse and 2.1 for rat. The nonanesthetic compound 2N, at a concentration more than six-fold larger than its predicted 1.0 MAC concentration (250 μM) had no effect on TRESK currents.
Mouse and rat TRESK have a conserved histidine residue immediately downstream of the selectivity filter sequence (GYG) of the first-pore domain (his132 in mouse and his143 in rat); human TRESK has a tyrosine residue at this location. This residue is critical for the extracellular pH sensitivity of the K2P channels TASK-1 and TASK-3 (13). Because human TRESK shows minimal extracellular pH sensitivity (3,4) and lacks this histidine residue, we hypothesized that the rodent TRESKs would be sensitive to changes in extracellular pH. As shown in Figure 2, mouse and rat TRESK currents were inhibited by acidic pH and enhanced by alkaline pH, whereas human TRESK currents showed little change over the same pH range. The pKa for mouse TRESK was 7.00 and for rat TRESK it was 7.16. We were unable to determine if replacement of the native tyrosine residue with histidine in human TRESK by site-specific mutagenesis conferred extracellular pH sensitivity because this change resulted in a nonfunctional mutant. Conversely, mutating the histidine residue in either mouse or rat TRESK to tyrosine resulted in nonfunctional subunits.
The pore vicinal histidine can also contribute to an inhibitory Zn+2 binding site. Clarke et al. (14) have recently shown that Zn+2 inhibits TASK-3 by interaction at the site of this histidine with a 50% inhibitory concentration (IC50) of 19.8 μM. Rodent TRESKs with this same amino acid residue show about two-fold greater sensitivity to Zn+2; mouse and rat TRESK were reversibly inhibited by Zn+2 with IC50s of 11.1 and 11.8 μM, respectively. In contrast, human TRESK is completely insensitive to Zn+2 inhibition up to 1 mM concentration (Fig. 2D).
In a previous study, we amplified human TRESK transcript with a qualitative PCR technique not only from the spinal cord, as originally described (3), but also from higher regions of the CNS. Accordingly, we used mouse TRESK-specific primers and Taqman quantitative PCR to more accurately determine TRESK expression levels in various regions of the mouse CNS; TRESK transcript was detected at approximately the same level at various divisions of the spinal cord including the dorsal root ganglion (Fig. 3). Significantly, almost four-fold more TRESK transcript was detected in both the cortex and cerebellum of the mouse brain than in the spinal cord but it was undetectable in heart, lung, liver, and kidney.
To further explore the role of mouse and rat TRESK in volatile anesthetic mechanisms, we studied the responses of the channels to the S(+) and R(−) enantiomers of isoflurane applied at concentrations of approximately 1.0 MAC (270 μM). As seen in Figure 4, both isoflurane enantiomers increased mouse and rat TRESK currents, consistent with the enhancement shown in Figure 1. In every oocyte, the S(+) enantiomer produced significantly greater enhancement of mouse (Fig. 4A) and rat (Fig. 4B) TRESK currents than the R(−) enantiomer. The average increases for mouse TRESK were: S(+) = 12.1% ± 1.1%, R(−) = 9.3% ± 1.0%, n = 6 (P < 0.001), R/S ratio = 0.77; and for rat TRESK: S(+) = 14.4% ± 1.9%, R(−) = 11.0% ± 1.5%, n = 7 (P < 0.002), R/S ratio = 0.76. In contrast, currents from oocytes expressing human TRESK were enhanced identically by exposure to each enantiomer: S(+) = 78.9% ± 17.5%, R(−) = 88.4% ± 17.1%, n = 9 (P = 0.56), R/S ratio = 1.12.
The rodent TRESKs, similar to most members of the K2P channel family, are inhibited by local anesthetics (15,16). Figure 5 shows the sensitivity of mouse and rat TRESK to the amide local anesthetics bupivacaine and lidocaine. Rodent TRESKs were inhibited by both drugs at 10-fold greater potency than human TRESK. The IC50 of bupivacaine for mouse TRESK was 3.5 μM and for rat TRESK it was 6.2 μM compared with 80 μM for human TRESK (4); the IC50 for lidocaine for mouse TRESK was 198 μM and for rat TRESK it was 432 μM compared with 3400 μM for human TRESK (4).
We studied the sensitivity of the mouse and rat TRESK to isoflurane and to local anesthetics. Similarities as well as differences to human TRESK were found. In comparison with human TRESK, isoflurane had lesser enhancing efficacy on mouse and rat TRESK. Human TRESK was enhanced up to 300% by clinically relevant volatile anesthetic concentrations, whereas mouse and rat TRESK showed a maximal enhancement of 40%–80% at somewhat larger anesthetic concentrations. The rodent forms but not the human orthologue were modulated by extracellular hydrogen and Zn+2 ions. In addition, both mouse and rat TRESK showed differential sensitivity to the enantiomers of isoflurane, whereas human TRESK did not. The rodent TRESKs were also significantly more potently inhibited by the amide local anesthetics bupivacaine and lidocaine. And just as expression of human TRESK can also be found in higher CNS regions, mouse TRESK transcript was found at four-fold higher levels in the cortex and cerebellum than in the spinal cord.
These results show that although there are some physiologic and pharmacologic differences between rodent and human TRESK, their basic responses are conserved. The difference in isoflurane efficacy on TRESK in these species might be predicted to produce a greater anesthetic effect by isoflurane in humans. The MAC of isoflurane is only 20% smaller in humans than in rats; we hypothesize that a higher level of expression of TRESK in critical regions of the rodent CNS explains the remaining difference in volatile anesthetic sensitivity between human and rodent TRESK.
A search of the human, mouse, and rat genomic sequences available through Genbank using the BLAST program failed to identify any other similar or closely related sequences to TRESK. Residue-by-residue comparison of our mouse sequence with the mouse TRESK-2 previously published detected no amino acid differences. Similarly, our human sequence, which is identical to that published initially by Sano et al. (3) is also identical to the human TRESK-2 sequence registered with Genbank. We conclude that there is a single TRESK gene in the human and rodent genomes that shares a high degree of amino-acid identity among the rodent subunits and a moderate amount between the rodent and human subunits.
The features of rodent TRESKs are similar to those displayed by some other K2P channels. TASK-1, TASK-2, and TASK-3 are also inhibited by extracellular H+ ions and activated by volatile anesthetics. The pH response curve described here for rat TRESK is most similar to the pH sensitivity of rat TASK-3 (17). TASK channels and TRESK also overlap in their brain and spinal cord expression patterns. These similar characteristics can make it difficult to unequivocally associate native currents to the expression of a particular K2P channel. Thus, the activity of TRESK may partially contribute to pH and volatile anesthetic-sensitive currents in cerebellar granule cells, hypoglossal motor neurons, and locus caeruleus neurons assigned previously to TASK-1 and TASK-3 channels (17–19).
At least part of the molecular mechanism underlying pH-sensing of K2P channels involves a histidine residue lying immediately distal to the first-pore domain. Mutation of this amino acid in TASK-1 and TASK-3 causes a large reduction of pH sensitivity (20). The rodent TRESK sequences also contain a histidine residue at this location (mouse histidine 132; rat histidine 143) but human TRESK does not. This histidine may be a significant contributor to extracellular pH sensing in TRESK. We found that the rodent TRESKs showed pH sensitivity comparable to TASK-3 whereas human TRESK showed little or none (3). Our mouse pH response results were similar to those reported by Kang et al. (12), which were obtained with outside-out patches expressing an identical mouse TRESK sequence (approximately 50% inhibition at pH 6.3). Based on inhibition studies with TASK-1, the pH sensitivity of K2P channels likely involves titration of amino acids but even with mutation of this residue the channel is still somewhat sensitive to the external pH (11). Evidently there are other, as yet not well understood, pH-sensing mechanisms in TASK-1 that probably also contribute to pH sensitivity of other K2P channels such as TASK-2 and rodent TRESK. This residue probably also contributes a coordinate for Zn+2 inhibition, as has been shown for TASK-3 (14), because rodent but not human TRESK were inhibited by micromolar concentrations of Zn+2.
Stereoselectivity of the responses to volatile anesthetics is one of four principal criteria proposed to discriminate a molecular site of anesthetic action (21). Other important criteria include physiologic plausibility, responses at clinical concentrations, and lack of effect by nonanesthetics/nonimmobilizers (i.e., compounds predicted to be anesthetics that do not exert an immobilizing effect in vivo). The differential potency of the enantiomers (R/S ratio) that we found for rat TRESK (0.76) is between the values reported by in vivo rat studies (0.65 and 0.85) (6,22). The lack of stereoselectivity shown for human TRESK suggests that isoflurane enantiomers would not produce stereoselective potency differences in humans; however, the stereoselectivity of volatile anesthetics has not yet been tested in humans.
The mechanism of volatile anesthetic activation of TRESK could result from changes in intracellular calcium. Czirjak et al. (11) discovered that mouse TRESK can be activated up to 15-fold directly by the application of the calcium ionophore ionomycin or indirectly by G protein-coupled receptors that increase cytoplasmic calcium. A previous study found that isoflurane can induce calcium release from intracellular stores in cultured neurons to increase cytosolic calcium (23). However, single channel studies with excised patches from Aplysia neurons in which the production of intracellular second messenger molecules was blocked found that volatile anesthetic activation of background K channels occurred through direct interaction of the volatile anesthetic with the background channel (24). Additional studies with this technique in transfected mammalian cells will be needed to determine whether isoflurane is acting directly to modulate the TRESK ion channel or indirectly via cytosolic calcium.
The volatile anesthetic sensitivity of the K2P channel TREK-1 has recently been shown in a mouse knockout model to contribute directly to the CNS depression produced by volatile anesthetics (25). We hypothesize that TRESK also contributes a significant component to volatile anesthetic action. TRESK expression in the cortex and cerebellum suggests that higher CNS functions such as consciousness, memory, and motor coordination could be modulated by volatile anesthetic-enhanced TRESK activity. TREK-1 and TRESK share only about 29% sequence similarity, mainly in the transmembrane and pore-forming domains. TRESK not only has short intracellular amino- and carboxy-terminal segments, but also quite a lengthy TM2-TM3 connector segment (approximately 91 amino acids). In contrast, TREK-1 has a long carboxy-terminal tail and a short connector segment (approximately 10 amino acids). Thus, there are no obvious consensus sequences that suggest a functional link between TRESK and TREK, and we believe that volatile anesthetic sensitivity arises from either a higher level (tertiary or quaternary) of protein folding or from interactions at the protein-lipid interface. The results presented here show that mammalian TRESK clones fulfill the basic criteria for a site of anesthetic action and support the hypothesis that activation of background K+ currents, mediated by members of the K2P channel family, contributes to the mechanism of action of volatile anesthetics.
The authors acknowledge the help of Drs. Michael Laster, James Sonner, and Edmond I Eger II for helpful discussions and for assisting in the analyses of anesthetic concentrations. We also acknowledge the gift of isoflurane enantiomers from Baxter Healthcare (New Providence, NJ).
1. Sonner JM, Antognini JF, Dutton RC, et al. Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth Analg
2. Franks NP, Lieb WR. Background K+ channels: an important target for volatile anesthetics? Nat Neurosci
3. Sano Y, Inamura K, Miyake A, et al. A novel two-pore domain K+ channel, TRESK, is localized in the spinal cord. J Biol Chem
4. Liu C, Au J, Zou H, et al. Potent activation of the 2P K channel TRESK by clinical concentrations of volatile anesthetics. Anesth Analg
5. Harris B, Moody E, Skolnick P. Isoflurane anesthesia is stereoselective. Eur J Pharmacol
6. Lysko GS, Robinson JL, Casto R, Ferrone RA. The stereospecific effects of isoflurane isomers in vivo. Eur J Pharmacol
7. Altschul SF, Gish W, Miller W, et al. Basic local alignment search tool. J Mol Biol
8. Quick MW, Lester HA. Methods for expression of excitability proteins in Xenopus
oocytes. In: Conn PM, ed. Methods in neuroscience
. San Diego: Academic Press, 1994:261–79.
9. Koblin DD, Chortkoff BS, Laster MJ, et al. Polyhalogenated and perfluorinated compounds that disobey the Meyer-Overton hypothesis. Anesth Analg
10. Gray AT, Leonoudakis DL, Winegar BD, et al. TOK-1 is an anesthetic activated K+ channel. Anesthesiology
11. Czirjak G, Toth ZE, Enyedi P. The two-pore-domain K+ channel, TRESK, is activated by the cytoplasmic calcium signal through calcineurin. J Biol Chem
12. Kang D, Mariash E, Kim D. Functional expression of TRESK-2, a new member of the tandem-pore K+ channel family. J Biol Chem
13. O'Connell AD, Morton MJ, Hunter M. Two-pore domain K+ channels-molecular sensors. Biochim Biophys Acta
14. Clarke CE, Veale EL, Green PJ, et al. Selective block of the human 2-P domain potassium channel, TASK-3, and the native leak potassium current, IKso, by zinc. J Physiol
15. Kindler CH, Yost CS, Gray AT. Local anesthetic inhibition of baseline potassium channels with two pore domains in tandem. Anesthesiology
16. Kindler CH, Paul M, Zou H, et al. Amide local anesthetics potently inhibit the human tandem pore domain background K+ channel TASK-2 (KCNK5). J Pharmacol Exp Ther
17. Berg AP, Talley EM, Manger JP, Bayliss DA. Motoneurons express heteromeric TWIK-related acid-sensitive K+ (TASK) channels containing TASK-1 (KCNK3) and TASK-3 (KCNK9) subunits. J Neurosci
18. Sirois JE, Lei Q, Talley EM, et al. The TASK-1 two-pore domain K+ channel is a molecular substrate for neuronal effects of inhalation anesthetics. J Neurosci
19. Kang D, Han J, Talley EM, et al. Functional expression of TASK-1/TASK-3 heteromers in cerebellar granule cells. J Physiol
20. Morton MJ, O'Connell, AD, Sivaprasadarao A, Hunter M. Determinants of pH sensing in the two-pore domain K(+) channels TASK-1 and -2. Pflugers Arch
21. Franks NP, Lieb WR. Molecular and cellular mechanisms of general anaesthesia. Nature
22. Eger EI 2nd, Koblin DD, Laster MJ, et al. Minimum alveolar anesthetic concentration values for the enantiomers of isoflurane differ minimally. Anesth Analg
23. Kindler CH, Eilers H, Donohoe P, et al. Volatile anesthetics increase intracellular calcium in cerebrocortical and hippocampal neurons. Anesthesiology
24. Winegar BD, Yost CS. Volatile anesthetics directly activate baseline S K+ channels in aplysia neurons. Brain Res
25. Heurteaux C, Guy N, Laigle C, et al. TREK-1, a K(+) channel involved in neuroprotection and general anesthesia. EMBO J