Simultaneous Activation of p38 MAPK and p42/44 MAPK by ATP Stimulates the K+ Current ITREK in Cardiomyocytes*

Living cells exhibit multiple K+ channel proteins; among these is the recently reported atypical two-pore domain K+ channel protein TREK-1. Most K+ currents are modulated by neurohormones and under various pathological conditions. Here, in rat ventricular cardiomyocytes using the whole-cell patch-clamp technique, we characterize for the first time a native TREK-1-like current (ITREK) that is activated by ATP, a purine agonist applied at a micromolar range. This current is sensitive to arachidonic acid, intracellular acidosis, and various K+ current inhibitors. Reverse transcription-polymerase chain reaction reveals the presence of a TREK-1-like mRNA in rat cardiomyocytes that shows 93% identity with mouse TREK-1. ATP effects are greatly attenuated in the presence of arachidonic acid or HCO− 3-induced intracellular acidosis. Using a series of inhibitors, we further demonstrate that the ATP-induced stimulation of ITREKimplies the activation of cytosolic phospholipase A2 and the release of arachidonic acid. These events require the simultaneous involvement of p38 MAPK and p42/44 MAPK, respectively, via a cAMP-dependent protein kinase and a tyrosine kinase pathway, whereas the two MAPKs conjugate to activate a mitogen- and stress-activated protein kinase (MSK-1). Our results thus demonstrate the occurrence of a TREK-1-like current in cardiac cells whose activation by purine agonists implies a dual-MAPK cytosolic pathway.

The exceptional diversity of K ϩ currents has physiological significance in the heart, where the various currents underlie distinct phases of action potential repolarization. Recent cloning efforts have identified a large number of pore-forming subunits for K ϩ channels. They include the voltage-activated, outward rectifying K ϩ channels (Kv families) and the inward rectifying K ϩ channels (Kir families) that show a single poreforming region and six or two transmembrane domains, respectively. More recently, a novel class of K ϩ channel subunits (TWIK) that possess two pore-forming regions and four transmembrane domains has been cloned and expressed (1)(2)(3)(4). All expressed TWIK-related K ϩ channels produce instantaneous and noninactivating currents that do not display a voltage-dependent activation threshold. Some show properties of a background K ϩ current, whereas others are activated by free fatty acids and stretch (2,5) and by intracellular acidosis (6).
TREK-1 and TBAK-1 or TASK-1 are expressed in the whole heart (1,7,8). Whether these proteins carry a physiological ionic current is not yet known in any tissue.
Of the purinergic agonists released during ischemia and other pathological conditions, adenosine has been the most extensively studied. However, under similar conditions, ATP is also released into the extracellular space (9,10). Extracellular ATP modulates the inward rectifying K ϩ currents, I K1 and I K(Ach) , and a delayed outward rectifying current in atrial cells (11,12). Recently, it was postulated that tyrosine phosphorylation is involved in the enhancement of the delayed rectifier K ϩ current by ATP in guinea pig ventricular cells (13). Besides activating ionotropic P2X receptors, purinergic stimulation by ATP might involve several of the multiple metabotropic P2Y receptors found in cardiac cells (14). P2Y receptors have been shown to be linked to activation of PKA 1 (15), protein kinase C and MAPK (16,17), and TK (18). ATP also activates phospholipase A 2 (PLA 2 ) in several tissues (19 -21). In a variety of cell types, PLA 2 activation occurs as a result of phosphorylation by MAPKs (22,23). In neutrophils, it was recently reported that inhibitors of p38 MAPK and p42/44 MAPK individually reduced cytosolic PLA 2 (cPLA 2 ) activity, whereas their combined blockade caused a total inhibition of cPLA 2 (24). Arachidonic acid (AA), which is released by cPLA 2 during metabolic inhibition or ischemia, activates a K ϩ current, I K(AA) , in rat neonatal atrial and adult ventricular cells (25,26). The channel subunit protein underlying I K(AA) is unknown.
We now report a TREK-like K ϩ current in isolated rat ventricular cells. This current is activated by extracellular ATP via the release of arachidonic acid after cPLA 2 activation. The purinergic-dependent activation of cPLA 2 requires the simultaneous activation of both p38 MAPK and p42/44 MAPK by a cAMP-dependent protein kinase and a tyrosine kinase-dependent pathway, respectively.

EXPERIMENTAL PROCEDURES
Myocyte Isolation-Ventricular myocytes were isolated from the heart of urethane-anesthetized (2 g/kg, i.p.) Wistar rats as described previously (27). The heart was first perfused for 5 min at 35°C with a nominally Ca-free HEPES-buffered solution containing 117 mM NaCl, 5.7 mM KCl, 4.4 mM NaHCO 3 , 1.5 mM KH 2 PO 4 , 1.7 mM MgCl 2 , 21 mM HEPES, 11 mM glucose, 20 mM taurine and then perfused for 40 min with the same solution plus 30 M Ca 2ϩ and 1.2 mg/ml collagenase (CLS4; Worthington). The heart was then gently dissociated through the bore of a large-tip pipette, followed by two decantations to separate * 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.
‡ A recipient of a grant from the Fondation pour la Recherche Médicale.
Voltage-Clamp Recording-Recording of K ϩ currents was performed using the whole-cell patch-clamp technique at room temperature (22°C Ϯ 2°C). Electrode resistance was between 0.9 -1.1 megaohms. The series resistance (R s ), membrane capacitance (C m ), and time constant of membrane capacitance ( c ) were determined on voltageclamped cells according to the equations: in which I o is the maximum membrane current, I ϱ is the current at the end of the 10-ms pulse, and E m is the amplitude of the voltage step (2 mV from a holding potential of Ϫ70 mV). K ϩ currents were recorded during 400-ms-long pulses applied between Ϫ130 mV and ϩ50 mV every 6 s in 10-mV increments from a holding potential of Ϫ50 mV or from a holding potential of Ϫ80 mV in some cases, as specified. Recordings were digitized at 200 s with a 12-bit analog-to-digital converter acquired and analyzed using pClamp 6 software.
Solutions and Drugs-For experiments, a cell aliquot was put in a Petri dish containing the control solution (117 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl 2 , 1.7 mM MgCl 2 , 10 mM glucose, and 10 mM HEPES; pH was adjusted to 7.4 with NaOH). After achieving whole-cell patchclamp configuration, the cell was exposed to different extracellular solutions by positioning it at the extremity of one of six capillaries (250-m inner diameter). Such a system allowed rapid changes of solution (Ͻ1 s). The control solution was supplemented with 50 M tetrodotoxin and 2 mM CoCl 2 to block Na ϩ and Ca 2ϩ currents, respectively. The internal solution contained 120 mM KCl, 6.8 mM MgCl 2 , 5 mM Na 2 ATP, 5 mM sodium creatine phosphate, 0.4 mM Na 2 GTP, 11 mM EGTA, 4.7 mM CaCl 2 (120 nM free Ca 2ϩ ), and 20 mM HEPES; pH was adjusted with KOH to 7.2; total K ϩ ϭ 145 mM.
Intracellular cAMP Assay-Cells were preincubated for 10 min in the presence of 0.1 mM isobutylmethylxanthine. Batches of 250,000 cardiac cells were incubated for 6 min at 37°C with 30 M ATP␥S, 40 M arachidonic acid, or 1 M isoproterenol with or without preincubation with specific inhibitors. Incubation was stopped by addition of perchloric acid. After centrifugation, the pellet was dissolved in NaOH and used to estimate protein content (28). The supernatant was neutralized by K 2 CO 3 , and the perchlorate precipitate was spun down. An aliquot of the supernatant was used for the cAMP assay. A high specific binding protein assay kit was used to determine cAMP content.
[ 3 H]AA Release in Intact Ventricular Cardiomyocytes-Freshly isolated myocytes were prelabeled with [ 3 H]AA (0.5 Ci/ml) for 20 h at 37°C. Cells were then washed three times with Dulbecco's modified Eagle's medium to remove free [ 3 H]AA and incubated for 20 min at 37°C. The different inhibitors were then added for 20 min before applying the agonist (30 M ATP␥S). Radioactivity in the supernatant was determined by scintillation counting (22).
Western Blot Analysis-After stimulation, cells were lysed in an ice-cold lysis buffer containing 10 mM Tris, 10 mM Na 4 P 4 O 7 , 1 mM EDTA, 1 mM MgCl 2 , and 10 mM NaF, pH 8.0, supplemented with 0.1 mM phenylmethylsulfonyl fluoride. Lysed cells were then centrifuged at 14,000 rpm for 20 min to separate the membrane and contractile proteins from the cytosolic fraction. The pellet was then resuspended with NET buffer (150 mM NaCl, 5 mM EDTA, and 50 mM Tris-HCl, pH 8.0) supplemented with 0.2 M leupeptin, 2 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 0.6 mM NaF, 1 mM sodium orthovanadate, and 1% Nonidet P-40. Phosphorylation of MAPK was determined using the cytosolic fraction, whereas cPLA 2 activity was measured using the membrane fraction. Briefly, proteins from each fraction were loaded (100 g/lane) on 7.5% acrylamide gels and run overnight. After electrophoretic transfer, the nitrocellulose membranes were incubated overnight at 4°C with the different antibodies. Antibodies against cPLA 2 (Santa Cruz Biotechnology) or phosphorylated p38 MAPK and p42/44 MAPK (New England Biolabs) were used at a dilution of 1:1,000. Bound primary antibody was revealed using secondary peroxidase-conjugated anti-mouse IgG antibody (1:6,000) for cPLA 2 or anti-rabbit IgG antibody (1:10,000) for MAPK and enhanced chemiluminescence detection according to the manufacturer's instructions. The blots were quantified by a digital imaging system. If necessary, the blots were stripped for 20 min at 50°C in a Tris buffer containing 62.5 mM Tris, 2% SDS, and 10 mM mercaptoethanol adjusted to pH 6.7.
RT-PCR of TREK-1-Total RNA was prepared from isolated ventricular cells purified on Percoll gradient. Briefly, cells were lysed in gua-nidinium thiocyanate-containing buffer, and RNA was extracted by a modified phenol chloroform method (29,30). First-strand cDNA was synthesized from 1 g of total RNA using 50 units of SUPERSCRIPT II reverse transcriptase (Life Technologies, Inc.), 3 M oligo(dT) 16 , and 5 mM dNTP. PCR was carried out using a set of gene-specific primers designed to anneal to the cDNA of mouse TREK-1 (GenBank TM accession number U73488) (forward primer, 5Ј-TTTGGCTTTCTACTGGCT-GGGG-3Ј; reverse primer, 5Ј-TCGTCTTCTTAGAGATCACCCG-3Ј corresponding to sequences 991-1012 and 1372-1393 in the open reading frame of mouse cDNA). Amplification was performed using a touch down protocol comprising 20 cycles at 94°C for 45 s, 65°C to 55°C (Ϫ0.5°C/cycle) for 45 s, and 72°C for 90 s, followed by 20 cycles at a constant annealing temperature of 55°C and a final extension step of 8 min at 72°C. TREK cDNA was also amplified from 300 ng of a mouse 11-day-old embryo MATCHMAKER cDNA library (CLONTECH). The RT-PCR products were analyzed by 1.2% agarose gel electrophoresis, and the ethidium bromide-stained bands were visualized by scanning photographs of gels using a charge-coupled device camera and Image J 1.01 software (National Institutes of Health, Bethesda, MD). The nature of the PCR products was checked by size and confirmed by two sequencing runs using the same primers used for PCR and a second set of primers (forward primer, 5Ј-CTCCCTGCGGTCATATTCAAGC-3Ј; reverse primer, 5Ј-TTCTTAGAGATCACCCGTAGCC-3Ј) annealing to the two extremities of the fragment amplified by PCR).
Statistical Analysis-Comparisons of data were made using Student's t test, and p Ͻ 0.05 was considered statistically significant.

External Application of ATP Increases K ϩ Currents in Ventricular
Cardiomyocytes-In rat ventricular myocytes, three distinct K ϩ currents are usually recorded. Hyperpolarizing pulses activate the inward rectifier K ϩ current (I K1 ) sensitive to low (Ϸ100 M) Ba 2ϩ concentrations, whereas depolarizing pulses induce activation of the transient outward K ϩ current (I to ) sensitive to 4-aminopyridine (Ϸ3 mM) and of the delayed rectifier K ϩ (I K ) current sensitive to tetraethylamonium (Ϸ20 mM) (31). The external application of 30 M ATP induced an increase in both the outward and inward currents that reached steady state after 6 min (Fig. 1A). The ATP-induced outward current exhibited a fast activation with no inactivation during the imposed voltage steps. In parallel experiments performed in the presence of 3 mM 4-aminopyridine, the time to peak was 7.1 Ϯ 1.3 ms at ϩ50 mV (n ϭ 10). During hyperpolarizing voltage steps to Ϫ130 mV, the ATP-induced inward current was large, peaked at 5.5 Ϯ 0.3 ms (n ϭ 10), and demonstrated fast inactivation before reaching a steady value after about 100 ms. Mean current density/voltage relationships established during the first 20 ms of the applied voltage step at the peak of the inward and outward current demonstrated a significant increase in the inward current only (Fig. 1B). The outward peak current was hardly enhanced as a result of a weak inhibition of I to . The ATP-induced mean current density/voltage relationship established at the end of the 400-ms voltage pulses exhibited a weak outward-going rectification above Ϫ60 mV and showed inward rectification below this voltage (Fig. 1C). Similar effects were obtained by applying ATP␥S (30 M), or ATP in the presence of theophylline (100 M), an inhibitor of adenosine-sensitive P1-purinoceptors, whereas the ATP effects were inhibited by suramin, a P2-purinoceptor antagonist (data not shown). Thus, current activation results from ATP activation of P2-purinergic receptors.
ATP activated a muscarinic-like K ϩ current in rat ventricular cells (data not shown), as it does in atrial cells (11). In the continued presence of acetylcholine, which itself induced a large transient inward current, the further application of ATP was able to induce sustained outward and inward currents, but not the large transient inward current (Fig. 1D). Under these conditions, the ATP-induced current showed a fast activation phase with no inactivation for both depolarizing and hyperpolarizing pulses. Similarly, on cells that have been pretreated with PTX to inactivate the muscarinic cascade, ATP induced sustained inward and outward currents (Fig. 1E). Current/ voltage relationships of the ATP-induced sustained current in the presence of Ach or after PTX pretreatment show weak inward and outward rectification with similar reversal potential and conductance (Fig. 1F). Consequently, ATP activates both a muscarinic-like and a specific P2-purinergic-induced current in isolated ventricular myocytes of the rat. In the following experiments, the properties of the latter current were analyzed.
Pharmacological Characterization of the ATP-induced Outward K ϩ Current-Several compounds known to affect cardiac K ϩ currents were used to characterize the specific ATP-induced sustained K ϩ current. Their effects are quantified here on the sustained outward current elicited at ϩ50 mV (Table I). The ATP-induced outward K ϩ current was enhanced in K ϩ -rich solution, insensitive to glibenclamide, a specific inhibitor of the K (ATP) current, and significantly inhibited by tetracain, Gd 3ϩ , and by 4-aminopyridine, tetraethylamonium, and Ba 2ϩ only at high concentrations (Table I). AA is known to trigger an increase in sustained outward K ϩ currents in cardiomyocytes (25,26,32) and to activate several two-pore domain K ϩ channels (2). In the presence of AA, ATP had no further significant effect ( Fig. 2; Table I). The effects of both ATP and AA were similar in the presence of indomethacin and NDGA, respec-tively, cyclooxygenase and lipoxygenase inhibitors (data not shown). The application of 90 mM HCO Ϫ 3 , which is known to induce intracellular acidosis and activate TREK-1 (6), mimicked the effect of ATP by activating the sustained K ϩ current ( Fig. 2; Table I). The subsequent application of ATP was without further effect. Furthermore, the ATP-induced sustained K ϩ current was only slightly inhibited (39%; n ϭ 12) by extracellular acidosis from pH 7.4 to pH 6.8. This excludes the current to be carried by TASK-1 (1). Thus, besides a muscarinic-like inward current, ATP activates a sustained K ϩ current whose characteristics point to the recently described acid-sensitive two-pore domain TREK-1 or to a TREK-1-like channel, which we will refer to as I TREK .
A TREK-1-like Product Is Expressed in Rat Ventricular Cardiomyocytes-We used RT-PCR to check for the presence of TREK-1 mRNA in rat ventricular cardiomyocytes using genespecific primers designed to anneal the cDNA of the mouse TREK-1 K ϩ channel. Fig. 3 shows RT-PCR products generated from the total RNA of these cells, as well as a comparison with the RT-PCR products from mouse embryo hearts. The rat ventricular cardiomyocyte products showed 93% identity with TREK-1 (7).
ATP Activates cPLA 2 -To investigate whether I TREK activation by ATP would also require AA production after PLA 2 stimulation, ventricular myocytes were pretreated with AA-COCF 3 , a specific inhibitor of the cytosolic, Ca 2ϩ -sensitive cPLA 2 . This treatment prevented ATP activation of the sustained outward I TREK . Such an inhibition was not seen with HELSS, a specific inhibitor of another intracellular PLA 2 , iPLA 2 . Also, in experiments performed with a pipette solution containing a very low intracellular Ca 2ϩ concentration with 5 mM EGTA and no Ca 2ϩ added, ATP induced a significantly smaller increase in sustained outward K ϩ current (1.0 Ϯ 02 pA/picofarad). Moreover, in parallel experiments, ATP␥S induced translocation of cPLA 2 to the membrane as well as a significant increase in the release of AA by isolated rat ventricular myocytes (Fig. 4). These observations indicate that AA increases I TREK as a result of cPLA 2 activation.
I TREK Activation and cPLA 2 Activation Require Both p38 MAPK and p42/44 MAPK-Incubation of cells with either SB202190 or U0126 (or PD98059) (specific inhibitors of p38 MAPK or p42/44 MAPK, respectively) fully prevented ATPinduced cPLA 2 translocation, AA release, and I TREK activation (Fig. 4). Similarly, the effects of ATP on cPLA 2 , AA, and I TREK were prevented by the application of TK inhibitors such as herbimycin A (or tyrphostin and ST 638). Adding 2Јd3Ј-AMP (or SQ22536) and RpcAMPs in the pipette solution, (inhibitor of adenylyl cyclase and cAMP-dependent protein kinase PKA, respectively) antagonized the activation of I TREK by ATP, whereas incubating the cells with the two compounds prevented AA release and cPLA 2 translocation (Fig. 4). Indeed, extracellular ATP is known to increase myocardial cAMP content by activating the specific adenylyl cyclase isoform, ACV, in neonatal rat cardiomyocytes (15). cAMP content was increased by 32.5 Ϯ 8.7% (n ϭ 6) in freshly isolated rat ventricular myocytes after purinergic stimulation, about 10-fold less than that by maximal isoproterenol stimulation (see also Ref. 15). No effect of AA (40 M) was seen on the ATP-induced increase in cAMP, nor was this increase inhibited by preincubation with blockers of cPLA 2 and MAPK (data not shown). Involvement of PKA in the ATP signal transduction pathway was also suggested when the PKA inhibitor peptide PKI 6 -22 prevented the increase in I TREK (data not shown). After pretreatment of the cells with cholera toxin, ATP still significantly increased I TREK . Finally, Ro318220, a specific inhibitor of the mitogen-and stress-activated protein kinase (MSK)-1, prevented the activation of I TREK by ATP as well as the translocation of cPLA 2 .
ATP Activates p38 MAPK and p42/44 MAPK-Western blot analysis of p38 MAPK and p42/44 MAPK revealed that after 6 min, a time chosen to match the electrophysiological experiments, ATP induced activation of both MAPKs (Fig. 5). p38 MAPK activation by ATP was not blocked by herbimycin A (or tyrphostin and ST 638; data not shown) but was significantly inhibited by RpcAMPs (or 2Јd3ЈAMP) (Fig. 5). On the contrary, the former, but not the latter compound significantly inhibited p42/44 MAPK activation by ATP. This indicates that PKA and TK activate p38 MAPK and p42/44 MAPK, respectively. Ro318220 had no effect on p38 MAPK and p42/44 MAPK activation (Fig. 5).

FIG. 3. A TREK-like mRNA is present in isolated cardiomyocytes.
Total RNA (2 g) extracted from rat isolated ventricular myocytes was reverse-transcribed and subjected to 40 cycles of PCR amplification. cDNA from an embryonic library (mouse embryo; embryonic day 11; 300 ng) was used as a positive control. These results are representative of three similar experiments.

DISCUSSION
The present results demonstrate for the first time the occurrence of a native TREK-1-like current. Furthermore, purinergic stimulation of cardiomyocytes induces this K ϩ current after the activation of cPLA 2 by a dual-MAPK cascade. Like the expressed TREK-1 current activated by AA or acidosis, the ATP-induced current in cardiomyocytes is sustained, demonstrates fast activation and weak outward rectification, and shows specific sensitivity to various K ϩ channel antagonists. In support, RT-PCR products with high identity to the original mouse TREK-1 are generated in the rat ventricular cardiomyocytes. AA activates several TWIK-like currents in in vitro expression systems, including TREK-1, which is also specifically activated by intracellular acidosis (6). Our experiments cannot distinguish whether the AA-induced activation is direct or secondary to the lowering of the intracellular pH (33). However, a direct effect of AA is favored. A secondary effect is unlikely under our patch-clamp conditions, where the fast superfusion of the cell should wash away released compounds, and the pipette solution pH is buffered by 20 mM HEPES. Thus, besides inducing a muscarinic-like inward current (11), ATP activates the recently described acid-and stretch-sensitive two-pore domain K ϩ channel TREK-1 (2, 5, 6) or a channel with similar characteristics in mammalian cardiac myocytes. This current shares many similarities including activation by AA and intracellular acidosis with I K(AA) initially reported in rat neonatal atrial cardiomyocytes (25,26). It was also noticed that I TREK was reduced in the presence of most inhibitors of one or the other of the MAPK cascades. This is in line with the reduction in sustained outward current found on long-term (15-20-min) ATP application. This late current decrease could result from a secondary ATP-induced inhibitory effect that was emphasized in the presence of GTP␥S added to the pipette solution rather than from current run-down (data not shown). It could be due to a partial reversion of I TREK opening by cAMP/PKA activation as reported previously (7) or attributed to a slow inhibition of Kv1.5 channel activity by AA (34).
The MAPK pathway is found ubiquitously in eukaryotic organisms, where it regulates cell proliferation, differentiation, and many other biological functions, some of which follow phosphorylation of cytosolic proteins, whereas, most often, MAPKs are translocated to the nucleus. Our results demonstrate that two MAPKs are required to induce significant I TREK activation that occurs as a consequence of AA release after cPLA 2 translocation. p38 MAPK activation, as shown by Western blot, is a compulsory step in the ATP stimulation cascade because the specific p38 MAPK inhibitor, SB202190, prevents cPLA 2 translocation, AA release, and I TREK activation. p 38 MAPK activation is dependent on the primary activation of the PKA pathway and is prevented by inhibition of adenylyl cyclase and PKA. p38 MAPK activation is not altered by TK inhibitors. The regulatory role of PKA in the p38 MAPK signaling cascade is presently unclear. It might involve activation of intermediate molecules such as protein tyrosine phosphatases with a 16-amino acid-long KIM region containing a PKA consensus sequence that shares a common function as negative regulators of the p42/44 MAPK and p38 MAPK pathways (35). This is strengthened by our observation that the related tyrosine phosphatase PTP-SL, phosphorylated by PKA (36), is present in rat ventric- ular cells (data not shown). The ATP-induced I TREK activation, cPLA 2 translocation, and AA release also require activation of p42/44 MAPK as shown by Western blot and the inhibitory effects of PD98059 and U0126. The PKA pathway inhibitors did not significantly alter p42/44 MAPK activation. These observations might result from the fact that PKA has a dual activator and inhibitory effect on B-Raf and c-Raf-1, respectively, such that the resulting effect on p42/44 MAPK activation could be very moderate (37). Activation of the p42/44 MAPKpathwaybyGprotein-coupledreceptorsistyrosinekinasedependent because tyrosine kinase inhibitors prevent I TREK activation, cPLA 2 translocation, and AA release. Src-family tyrosine kinases, Fyn and Src, have been implicated in G protein-MAPK activation by angiotensin II in cardiac myocytes (38). Furthermore, we reported previously that purinergic stimulation activates Fyn and FAK to phosphorylate the Cl Ϫ / HCO Ϫ 3 exchanger in the rat heart and trigger cell acidosis (18). At odds with the significant inhibition of the ATP-induced cPLA 2 activation by PTX in the cPLA 2 -overexpressing Chinese hamster ovary cell line (39), in cardiomyocytes, PTX does not prevent the ATP effects on I TREK but it does antagonize the effects of acetylcholine on the fast inward current (Fig. 1D). This indicates that a G i/o protein is not involved in this pathway, whose more initial steps remain to be elucidated.
Considerable efforts have been made in recent years to study the mechanism of activation of cPLA 2 and the subsequent release of AA. cPLA 2 translocation from the cytosol to the membrane is a critical step for access to phospholipids that require Ca 2ϩ ions. cPLA 2 is known to be active at Ca 2ϩ concentrations found in the cytosol of cardiac myocytes (22,23). Thus, in vivo and during our biochemical experiments, some part of cPLA 2 activation might result from an increase in intracellular Ca 2ϩ induced by extracellular ATP stimulation (40,41). However, our electrophysiological experiments performed in the presence of a high EGTA-buffered Ca 2ϩ -containing pipette solution demonstrate a reproducible and large increase in I TREK (much greater than that seen in the absence of Ca 2ϩ ) and thus imply other regulatory mechanisms. cPLA 2 is a physiological target for both p42/44 MAPK and p38 MAPK (22,24,42,43). Under ATP stimulation, both p42/44 MAPK and p38 MAPK are activated (Fig. 5). Each specific MAPK inhibitor did not affect the activation level of the other MAPK but did prevent significant activation of the whole cascade. These observations in cardiomyocytes are in agreement with previous reports that cPLA 2 is the substrate for several MAPKs (44,45). We therefore suggest that in cardiomyocytes, p42/44 MAPK and p38 MAPK act independently and simultaneously to activate cPLA 2 , AA release, and I TREK . These effects might be due to each MAPK phosphorylating a different site among the several consensus sites on cPLA 2 . However, it should rather be considered that dual activation by both p42/44 MAPK and p38 MAPK has been reported for MSK-1, as well as for MAPKAPK5 (46 -48). MSK-1 is a widely expressed enzyme that is found in the heart, although with a 12-30-fold lower density in the cytosol than in the nucleus. During our experiments, Ro318220, a specific MSK-1 inhibitor, markedly reduces cPLA 2 translocation and I TREK activation, as did inhibition of each of p42/44 MAPK and p38 MAPK. One might thus consider that MSK-1 integrates phosphorylation by both MAPKs and then activates cPLA 2 (Fig. 6).
In conclusion, we first demonstrate the occurrence of a native TREK-1-like current in rat cardiomyocytes. We further elucidate multiple steps leading to TREK-1 activation during ATP application on cardiac cells. The cascade involved dual-MAPK activation, cPLA 2 translocation, and AA release. These purinergic-induced effects might result from the stimulation of different purinoceptors or from that of only the P2Y 11 subtype that is positively coupled to both cAMP and inositol trisphosphate production pathway (49). In vivo, these ATP-induced effects will be complemented by the internal acidosis that follows the ATP-induced activation of the Cl Ϫ /HCO Ϫ 3 exchanger (18). Both pathways are of importance under various pathological conditions including ischemia, a period during which ATP is released. Activation of I TREK is one of the numerous alterations induced by ATP that would contribute to electrophysiological disturbances in the ventricular wall and possibly in the nervous system, in which most cells expressed TREK-1 mRNA.