A novel two-pore domain K+ channel, TRESK, is localized in the spinal cord.

To find a novel human ion channel gene we have executed an extensive search by using a human genome draft sequencing data base. Here we report a novel two-pore domain K+ channel, TRESK (TWIK-related spinal cord K+ channel). TRESK is coded by 385 amino acids and shows low homology (19%) with previously characterized two-pore domain K+ channels. However, the most similar channel is TREK-2 (two-pore domain K+ channel), and TRESK also has two pore-forming domains and four transmembrane domains that are evolutionarily conserved in the two-pore domain K+ channel family. Moreover, we confirmed that TRESK is expressed in the spinal cord. Electrophysiological analysis demonstrated that TRESK induced outward rectification and functioned as a background K+ channel. Pharmacological analysis showed TRESK to be inhibited by previously reported K+ channel inhibitors Ba2+, propafenone, glyburide, lidocaine, quinine, quinidine, and triethanolamine. Functional analysis demonstrated TRESK to be inhibited by unsaturated free fatty acids such as arachidonic acid and docosahexaenoic acid. TRESK is also sensitive to extreme changes in extracellular and intracellular pH. These results indicate that TRESK is a novel two-pore domain K+ channel that may set the resting membrane potential of cells in the spinal cord.

To find a novel human ion channel gene we have executed an extensive search by using a human genome draft sequencing data base. Here we report a novel twopore domain K ؉ channel, TRESK (TWIK-related spinal cord K ؉ channel). TRESK is coded by 385 amino acids and shows low homology (19%) with previously characterized two-pore domain K ؉ channels. However, the most similar channel is TREK-2 (two-pore domain K ؉ channel), and TRESK also has two pore-forming domains and four transmembrane domains that are evolutionarily conserved in the two-pore domain K ؉ channel family. Moreover, we confirmed that TRESK is expressed in the spinal cord. Electrophysiological analysis demonstrated that TRESK induced outward rectification and functioned as a background K ؉ channel. Pharmacological analysis showed TRESK to be inhibited by previously reported K ؉ channel inhibitors Ba 2؉ , propafenone, glyburide, lidocaine, quinine, quinidine, and triethanolamine. Functional analysis demonstrated TRESK to be inhibited by unsaturated free fatty acids such as arachidonic acid and docosahexaenoic acid. TRESK is also sensitive to extreme changes in extracellular and intracellular pH. These results indicate that TRESK is a novel two-pore domain K ؉ channel that may set the resting membrane potential of cells in the spinal cord.
The potassium (K ϩ ) channel was first identified and characterized as Drosophila Shaker. It is activated by a change in membrane potential and induces selective permeation of K ϩ (1-3). To date, more than 70 K ϩ channel genes have been identified in mammals and are divided into several subfamilies on the basis of their sequence similarities and channel properties (4). The K ϩ channels are characterized by having two, four, or six transmembrane domains and one or two pore-forming domains that are essential for selective permeation of K ϩ . It was reported that the K ϩ channel is present in all living cells and organs and is involved in a variety of cellular functions, including regulation of the neuronal firing rate, heart rate, muscle contraction, and hormone secretion (5,6).
The most recently identified K ϩ channels were two-pore domain K ϩ channels, which have four transmembrane do-mains (TM1-4), two pore-forming domains (P1, P2), cytoplasmic N-and C-terminal regions, and an extended extracellular loop between TM1 and P1, which are evolutionarily conserved (7). Especially, the pore-forming domain has a consensus amino acid motif of TXG(Y/F)G. In mammals the first gene identified as coding for the two-pore domain K ϩ channel was named TWIK-1, and to date 14 genes have been isolated as members of the subfamily (8 -25). In addition to mammals, two-pore domain K ϩ channels have been isolated and characterized in Drosophila, Caenorhabditis elegans, and plants (26 -28). Most clones of these genes form functional K ϩ channels when expressed in Xenopus oocytes or in mammalian cell lines and exhibit the properties of a background K ϩ current. Functional characterizations show that these channels have the following sensitivities: (i) mechanical stress, TREK-1, TREK2, and TRAAK; (ii) changes in pH, TASK-1, TASK-2, TASK-3, THIK-1, TREK-2, TALK-1, and TALK-2; (iii) heat stress, TWIK-2, TREK-1; (iv) fatty acids, TASK-1, TASK-3, TREK-1, TREK-2, TRAAK, and THIK-1 (8 -24). However, it is not clear whether these sensitivities are involved in cellular function.
The human genome project was completed in 2001, and the genome sequence is now open to the public (29,30). As a result, many genes whose existence had not previously been reported have now been isolated by using the human draft sequencing data base. Many channel genes have been isolated and analyzed for their cellular mechanisms. This has been especially true for two-pore domain K ϩ channel genes. In this paper we describe the cloning of a novel two-pore domain K ϩ channel, TRESK, from the human draft sequencing data base and describe its tissue distribution and electrophysiological and pharmacological properties.

EXPERIMENTAL PROCEDURES
Molecular Cloning and Isolation of TRESK-The human TWIK-1 amino acid sequence was used for a TBLASTN search against the human draft genome sequencing data base at the National Center for Biotechnology Information (NCBI). The genomic clone (GenBank TM accession number AL731557) was identified, and some fragments of AL731557 showed the highest similarity to human TREK2. To identify the 5Ј-and 3Ј-ends of the cDNA corresponding to this fragment, 5Ј and 3Ј rapid amplification of cDNA ends was performed with human spinal cord poly(A) ϩ RNA using a Marathon cDNA amplification kit (Clontech) and primers derived from the genomic fragment. Amplified fragments were directly cloned into the plasmid pCR-TOPO (Invitrogen) and sequenced on both strands (Applied Biosystems, model 3700). Sequencing of the rapid amplification of cDNA end products revealed that this genomic fragment was a part of a cDNA. The determined sequences were assembled into one contiguous sequence with an open-reading frame encoding 385 amino acids. This is called human TRESK.
Finally, to verify that human TRESK cDNA including the entire open-reading frame could be cloned from an independent source, human spinal cord poly(A) ϩ RNA (Clontech) was used for the reverse * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AB087138.
RT-PCR Analysis-Multiple human poly(A) ϩ RNA (Clontech) was reverse-transcribed to generate first-strand cDNA with an Advantage RT-for-PCR kit (Clontech). PCR was carried out with this first-strand cDNA, corresponding to 5 ng of poly(A) ϩ RNA. PCR primers were designed from the human TRESK cDNA sequence, 5Ј-TGCTTTCTGGT-GACCTACGC-3Ј and 5Ј-GCTGCACAGGAAATGTAGGC-3Ј. PCR was performed for 45 cycles at 98°C for 15 s and 70°C for 1 min using Taq polymerase. The PCR products were run on 1.5% polyacrylamide gels. Ethidium bromide-stained gels were examined under UV illumination. Amplified fragments were subcloned into the pCR-TOPO vector and sequenced on both strands to confirm the PCR product as human TRESK. As a control, glyceraldehyde-3-phosphate dehydrogenase was amplified using specific primers (Clontech).
Transfection of TRESK-The previously amplified cDNA of TRESK was cloned into the pcDNA3.1 vector (Invitrogen). L929 cells were transfected with TRESK cDNA for electrophysiological studies as described elsewhere (31)(32)(33). Cells were co-transfected with the TRESK expression vector described previously and the green fluorescent protein expression vector phGFP S65T (Clontech) using FuGENE 6 (Roche Applied Science). The cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Transfected cells were identified by observing green fluorescent protein fluorescence with an epifluorescence microscope. Electrophysiological studies were carried out 1-3 days after transfection.
Electrophysiological Study-Electrophysiological recordings of TRESK transfectants were performed using a voltage-clamp technique (34). Recordings were made with an Axopatch 1D amplifier (Axon Instruments) using patch electrodes with a resistance of 3-5 megaohms. The internal pipette solution contained 149 mM KCl, 1.8 mM MgCl 2 , 4.5 mM EGTA, and 9 mM HEPES-K (pH 7.3). The external solution contained 149 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , and 10 mM HEPES-Na (pH 7.3). For the inside-out patch experiments, both the pipette and bath solutions contained 150 mM KCl, 2 mM MgCl 2 , and 10 mM HEPES-K (pH 7.3). All recordings were performed at room temperature (22°C). Analysis was carried out on a personal computer using pCLAMP6 software (Axon Instruments). All values are expressed as means Ϯ S.E. Statistical significance was tested with Student's t test.

Molecular Cloning of Human TRESK-
The two-pore domain K ϩ channel-related sequence, accession number AL731557, was identified in the human draft genome sequencing data base of GenBank TM using a TBLASTN search for the amino acid sequence of human TWIK-1. Genomic clone AL731557 has a region encoding an amino acid sequence having the highest similarity to human TREK-2. As described under "Experimental Procedures," a series of rapid amplification of cDNA end studies using primers designed from this region resulted in identification of a full-length coding sequence (CDS) (1158 bases) with an open-reading frame encoding 385 amino acids (Fig. 1A). Because this new clone is included in human genomic clone AL731557, consists of 142,068 bases, and has been assigned to chromosome 10 of the human genome, we performed sequence alignment with the CDS and AL731557. The alignment revealed that the CDS consisted of 3 exons composed of 222, 129, and 804 bases and two introns spanning 12,811 bases. 1 The abbreviations used are: RT, reverse transcription; TM, transmembrane domain. Hydrophobicity analysis of the amino acid sequence showed that the new clone belongs to the K ϩ channel family since it has two pore-forming domains (P) and four transmembrane domains (TM) (Fig. 1, A and B). Two N-glycosylation sites are present in the TM1-P1 linker region (Fig. 1, A and C). Because the N-terminal region is located on the intracellular side, similar to other two-pore domain K ϩ channels, the putative K ϩ channel has a short N-terminal region, extended intracellular loop between TM2 and TM3, and a short C-terminal region (Fig. 1C).
Searching the GenBank TM data base with the BLAST sequence alignment program revealed that the full amino acid sequence of the new clone shows the greatest similarity to human TREK2, but even so, it still shares only 34% identity. Alignment of the amino acid sequence of the new clone with other two-pore domain K ϩ channels shows low homology, 13.5-19.7%. Also, phylogenetic analysis suggested that this clone has a common ancestor with previously reported two-pore domain K ϩ channels but represents an additional branch in this family of the channels (Fig. 2). These results indicate that this new clone is a novel two-pore domain K ϩ channel that is included in a novel subfamily. Thus, we named this new clone the TWIK-related spinal cord K ϩ channel, TRESK.
Tissue Distribution of TRESK-To determine the tissue distribution of TRESK in humans, RT-PCR analysis was carried out using a specific primer set, one primer from the first exon and the other from the third exon of TRESK, with human multiple tissue poly(A) ϩ RNA. The expected 800-bp PCR product of TRESK, confirmed by sequencing, was amplified only in the spinal cord (Fig. 3). To be sure of this result, we repeated the same experiment with another specific primer set, one primer from the first exon and the other from the third exon of TRESK, and another set of human multiple tissue poly(A) ϩ RNA. Again, the expected PCR product was amplified only in the spinal cord (data not shown). Northern blot analysis was performed using human multiple tissue blots (MTN Blots, Clontech), but no specific signal was detected in the 23 tissues (brain, heart, skeletal muscle, colon, thymus, spleen, kidney, liver, small intestine, placenta, lung, peripheral blood leukocytes, adrenal gland, bladder, bone marrow, lymph node, mam-mary gland, prostate, spinal cord, stomach, thyroid, trachea, and uterus) (data not shown). The reason is thought to be that the TRESK transcript level is very low and/or expressed in a specific region of the spinal cord.
Basic Electrophysiological Properties of TRESK-To characterize the electrophysiological properties of TRESK using the whole-cell voltage clamp technique, L929 cells were transiently transfected with the TRESK expression vector. The cells were clamped at a holding potential of Ϫ80 mV and were depolarized to voltages between Ϫ120 and ϩ100 mV. In TRESK-transfected cells, large currents were recorded (Fig. 4A) that were not present in the control cells (data not shown). The activation kinetics of TRESK current was rapid. Depolarization steps induced a two-step current composed of large instantaneous and delayed components (Fig. 4A). The ion selectivity of TRESK was studied by changing the concentration of K ϩ in the external solution from 5 to 150 mM while maintaining the pipette [K ϩ ] constant at 150 mM. The current-voltage (I-V) relationship is outwardly rectifying, and almost no inward currents were recorded in an external medium containing 5 mM K ϩ (Fig. 4B). The plots of the reversal potential as a function of [K ϩ ] showed that the slope was 49.0 Ϯ 0.8 mV/decade (n ϭ 3), which is close to the calculated Nernst value of 58 mV, as expected for a highly selective K ϩ channel (Fig. 4C). As previously reported for other two-pore domain K ϩ channels, the membrane potential of TRESK-expressed L929 cells with a high current density was strongly polarized (Fig. 4D). In contrast, the membrane potential of TRESK-expressed cells that had a low current density was not polarized (Fig. 4D).
Pharmacological Properties of TRESK-The effect of various pharmacological agents on currents elicited by voltage pulse to 0 mV have been studied in TRESK-transfected cells using whole-cell recording. The classical K ϩ channel blocker, Ba 2ϩ , applied extracellularly at 3 mM inhibited TRESK current. This inhibition was gradually released by depolarization to voltages from Ϫ20 to ϩ80 mV (Fig. 5A). To clarify the inhibition by Ba 2ϩ , we tested the effect of Ba 2ϩ at the start and end of pulses at different voltages. The inhibition by Ba 2ϩ at the end of the pulse was significantly decreased at depolarized potentials. On the other hand, the inhibition at the start of the pulse was not affected by the depolarization (Fig. 5B). These results suggest that Ba 2ϩ acts as an open channel blocker. TRESK was insen- sitive to 4-aminopyridine (non-selective K ϩ channel blocker), paxilline, penitrem A (Ca 2ϩ -activated K ϩ channel blocker), tolazamide, and glipizide (ATP-sensitive K ϩ channel blocker) when applied extracellularly at 100 M. Propafenone, glyburide (ATP-sensitive K ϩ channel blocker), lidocaine (Na ϩ channel blocker), quinine, quinidine, and triethanolamine (non-selective K ϩ channel blocker) inhibited TRESK currents by 95.3 Ϯ 1.8% (n ϭ 3), 76.8 Ϯ 8.6% (n ϭ 4), 75.7 Ϯ 5.4% (n ϭ 5), 75.2 Ϯ 7.1% (n ϭ 4), 90.6 Ϯ 1.5% (n ϭ 4), and 34.4 Ϯ 1.0% (n ϭ 3), respectively (Fig. 5C). Because these compounds were reported to inhibit K ϩ channels, the results suggest that TRESK has the property of K ϩ channels.
Inhibition of TRESK by Free Fatty Acids-Outside-out patches were used to test whether TRESK is sensitive to free fatty acids. Arachidonic acid, which is a polyunsaturated free fatty acid, inhibited TRESK current when applied to the extracellular side of outside-out patches (Fig. 6A). We also tested whether arachidonic acid could inhibit TRESK from the cytoplasmic side of the membrane. In inside-out patches, intracellular application of arachidonic acid inhibits TRESK current (Fig. 6B). This inhibition was expressed in a concentration-dependent manner (Fig. 6C). The onset of inhibition was generally rapid (within 1 min), and steady-state inhibition was observed within 1 min after application. The relative channel current was plotted as a function of the arachidonic acid concentration in Fig. 6D. With the reasonable assumption that 50 M arachidonic acid produces maximal inhibition of TRESK, the averaged data were fit to a Hill equation of the form y ϭ 1/ (1 ϩ (K 1/2 /[arachidonic acid])n), where K 1/2 is the apparent concentration of arachidonic acid that produces half-maximal inhibition, and n is the Hill coefficient (apparent K 1/2 ϭ 6.6 M, Hill coefficient ϭ 1.3, n ϭ 3). We tested whether other free fatty acids, i.e. saturated and unsaturated free fatty acids, were also able to inhibit TRESK current. When applied to the intracellular side, unsaturated free fatty acids such as docosahexaenoic acid, linoleic acid, and oleatic acid significantly inhibited TRESK current at 20 M (Fig. 6, E and F). Saturated free fatty acids such as palmitic acid and stearic acid had almost no effect on TRESK current at 20 M (Fig. 6, E and G, n ϭ 3 and 4,  respectively). These results show that unsaturated free fatty acids inhibit TRESK current.  5. Pharmacological properties of TRESK. A, the addition of 3 mM Ba 2ϩ to the bath solution reversibly reduced the TRESK current. Cells were held at Ϫ80 mV and depolarized to a voltage between Ϫ120 and ϩ100 mV. B, inhibition of TRESK currents by 3 mM Ba 2ϩ at the start and end of pulses at different voltages. Cells were held at Ϫ80 mV and depolarized to a voltage between Ϫ20 and ϩ80 mV. C, inhibition by various K ϩ channel inhibitors, 50 M propafenone, 50 M glyburide, 1 mM lidocaine, 0.1 mM quinine, 0.1 mM quinidine, and 2 mM triethanolamine (TEA), of TRESK currents elicited by 0 mV pulse.

Regulation of TRESK Current by pH-We tested whether
TRESK is sensitive to changes in the extracellular and intracellular pH. We first used whole-cell recording to test whether TRESK is sensitive to changes in the extracellular pH. The cells were clamped at a holding potential of Ϫ40 mV and then depolarized to voltages between Ϫ80 and ϩ80 mV. TRESK currents were measured at different extracellular pH values (Fig. 7A). Changes in the extracellular pH from 7.3 to 5.6 and 8.9 resulted in inhibition (80 Ϯ 7%, n ϭ 3) and enhancement (120 Ϯ 5%, n ϭ 3) of TRESK currents, respectively, from the basal channel activity observed at pH 7.3 (Fig. 7B). Thus, the TRESK channel is sensitive to changes in the extracellular pH, acid, and alkali, but those sensitivities are not remarkable compared with other K ϩ channels. To test the effect of changes in the intracellular pH, we used inside-out patches. Fig. 7C shows that the TRESK current was remarkably inhibited by a change in the intracellular pH from 7.3 to 5.6. Changes in the intracellular pH from 7.3 to 5.6 and 8.9 resulted in inhibition (39 Ϯ 4%, n ϭ 3) and enhancement (140 Ϯ 12%, n ϭ 3) of TRESK currents, respectively, from the basal channel activity observed at pH 7.3 (Fig. 7D). These effects, especially at pH 5.6, were greater than those of the extracellular pH. These results show that TRESK is much more sensitive to changes in intracellular pH than extracellular pH.
We also tested whether TRESK is sensitive to changes in the temperature. Several two-pore K ϩ channels were also reported to be sensitive to changes in the temperature (16,19). We recorded the TRESK current in whole-cell recording at 15, 22, 39, and 43°C. Lowering the temperature from 22 to 16°C had no effect on the TRESK current at Ϫ40 mV (102.7 Ϯ 11.5%, n ϭ 3). Raising the temperature from 22 to 39 or 43°C also had no effect on the TRESK current (106.3 Ϯ 5.6%, n ϭ 3; 118.8 Ϯ 6.4%, n ϭ 3, respectively). These results show that TRESK is insensitive to changes in temperature.

Molecular
Cloning of a Novel Two-pore Domain K ϩ Channel-We isolated TRESK as a novel member of the emerging family of two-pore domain K ϩ channels. Its cloning extends to 15 the number of these channels identified to date in mammals. Despite general conservation of the structure, TRESK does not share more than 13.5-19.7% amino acid identity with the previously reported channels and does not seem to be more related to any one of them from a phylogenetic point of view. Although the major structure of this family is a long C-terminal cytoplasmic region, TRESK has a short C-terminal region. These data indicate that this channel represents a new subfamily of twopore domain K ϩ channels. Our study shows that expression of TRESK mRNA is very low in humans because there was no positive signal in Northern blot analysis. However, RT-PCR analysis revealed that this mRNA was localized only in the spinal cord, indicating that the TRESK transcript level is very low and/or expressed in a specific region of the spinal cord. Thus, we decided to name this new two-pore domain K ϩ chan-  7. pH-dependent changes in TRESK current. A, the current-voltage (I-V) relationships deduced from whole-cell currents of TRESK elicited by voltage ramp ranging from Ϫ80 to ϩ80 mV, measured at three different extracellular pH levels (5.6, 7.3, and 8.9). I m , current density. B, effects of changes in extracellular pH on TRESK current at 0 mV. The current at pH out 7.3 was taken as 100%. C, inhibition of TRESK current by change in intracellular pH in inside-out patches held at ϩ40 mV. D, effects of changes in intracellular pH on TRESK current at ϩ40 mV. The current at pH in 7.3 was taken as 100%.
The asterisks indicate a significant difference from the value at pH in 7.3, as judged by Student's t test (p Ͻ 0.01).
nel the TWIK-related spinal cord K ϩ channel, TRESK, because this differentiates it from other two-pore domain K ϩ channels.
Electrophysiological Behavior of TRESK-Electrophysiological studies demonstrated many profiles of the TRESK channel. The reversal potential of the current induced by expression of TRESK-cDNA in mammalian cell lines followed the calculated K ϩ equilibrium potential when the external [K ϩ ] was changed. This suggests that the TRESK channel selectively permeates K ϩ . The whole-cell current-voltage I-V relationship in a symmetric K ϩ solution showed that TRESK induced outward rectification. When TRESK is transiently expressed in mammalian cell lines, TRESK currents are observed at all membrane potentials, including the resting membrane potential. This channel activity is also seen in other two-pore domain K ϩ channels. The basal current that is present at the resting membrane potential is thought to drive the membrane potential toward the K ϩ equilibrium potential. TRESK current is activated instantaneously by a voltage step and shows no voltage threshold for activation. Moreover, the membrane potential of TRESK-expressing cells is strongly polarized. These properties suggest that TRESK may function as a background K ϩ channel that sets the resting membrane potential.
The pharmacological profiles of the TRESK channel are similar to those of previously reported two-pore domain K ϩ channels, which are inhibited by K ϩ channel inhibitors. We analyzed the detailed mechanism of the inhibition by Ba 2ϩ . The inhibition by Ba 2ϩ at the end of the pulse was significantly decreased at depolarized potentials, but the inhibition at the start of the pulse was not affected. These properties suggest that Ba 2ϩ inhibited the open state of TRESK, and this blockage was released by depolarized stimulus.
Free Fatty Acid and pH Sensitivity of TRESK-To test for functional similarities with other members of the two-pore domain K ϩ channel family, we studied modulation of the TRESK channel by various experimental interventions such as application of free fatty acids and changes in the pH.
Like TASK-1 and TASK-3, TRESK was inhibited by intracellular and extracellular application of arachidonic acid. This inhibition was shown by intracellular application of not only arachidonic acid but also other unsaturated free fatty acids, such as docosahexaenoic acid, linoleic acid, and oleatic acid. However, our studies also showed that saturated free fatty acids such as palmitic acid and stearic acid did not significantly affect the TRESK current. Thus, the TRESK channel is sensitive only to unsaturated free fatty acids, and it seems that there may be an unsaturated free fatty acid-selective site in the TRESK protein.
Furthermore, we demonstrated here that TRESK currents were inhibited and enhanced by extreme changes in the pH values from 7.3 to 5.6 and 8.9, respectively. Regulation of the channel activity by the pH is an important functional property of two-pore domain K ϩ channels. TRESK exhibited greater sensitivity to acidic conditions than alkaline conditions. Our studies also show that TRESK is more sensitive to changes in the intracellular pH than changes in the extracellular pH. This property has also been reported for TASK-3 and TREK-2 (18,20).
Predicted Physiological Role of TRESK in the Spinal Cord-Until now, some two-pore domain K ϩ channels were reported to be expressed in the spinal cord, but they were not specific to the spinal cord (11). On the other hand, the expression of TRESK mRNA is specifically localized in the spinal cord, indicating that TRESK may play a unique role in the spinal cord that is different from other two-pore domain K ϩ channels. Actually, we revealed that the TRESK channel is regulated by free fatty acids, which have been reported to participate in the function of the spinal cord. Spinal cord trauma is known to cause a marked release of free fatty acids, especially arachidonic acid, from the cell membrane. These released free fatty acids are suggested to lead to the secondary injury to spinal cord neurons that is observed in spinal cord trauma (35). However, the specific effects of released arachidonic acid in relation to the mechanisms of spinal cord injury remain to be established. One would suspect that Ca 2ϩ influx, which is known to be one of the most common cellular signaling factors, is involved in spinal cord injury. Therefore, if injured neurons show TRESK currents, released free fatty acids may cause Ca 2ϩ influx via a voltage-gated Ca 2ϩ channel by depolarizing the membrane potential through changing the TRESK current.
In summary, we have isolated TRESK, an outward rectifying K ϩ channel that belongs to the two-pore domain K ϩ channel family. The functional properties of TRESK show that it is sensitive to unsaturated free fatty acids and extreme changes in extracellular and intracellular pH. Furthermore, its tissue distribution and basic electrophysiological properties suggest that TRESK may be involved in regulation of the resting membrane potential in spinal cord-specific cells.