Am J Phorol Cell Physiol 316: C312C824, 2018, iret pobied December 21,2018; do 10 1152aipest.00817 2018 RESEARCH ARTICLE Triple arginine residues in the proximal C-terminus of TREK K~ channels are critical for biphasic regulation by phosphatidylinositol 4,5-bisphosphate JooHan Woo,'* Young Keul Jeon,! Yin-Hua Zhang,'** Joo Hyun Nam, Dong Hoon Shin,$ and © Sung Joon Kim'** ‘Department of Physiology, Seoul National University College of Medicine, Seoul, South Korea; *Deparment of Biomedical Sciences, Seoul National University College of Medicine, Seoul, South Korea; "schemic/Hypoxic Disease Institute, Seoul National University College of Medicine, Seoul, South Korea; “Department of Physiology and Ion Channel Disease Research Center, Dongguk University College of Medicine, Seoul, South Korea: and "Department of Pharmacology, Yonsei University College of Medicine, Seoul, South Korea Sebmited 25 October 2018; accepted in final form 17 December 2018 ‘Woo J, Jeon YK, Zhang YH, Nam JH, Shin DH, Kim SJ. Triple arginine residues in the proximal C-terminus of TREK K" channels axe critical fr biphasic regulation by phosphatidylinostel4,5-bispho- sphate. Am J Physiol Cell Physiol 316: C312-C324, 2019. First published December 21, 2018; doi:0,1152/ajpoell 0417.2018,— ‘TWIK-related two-pore domain K* channels (TREKs) are activated by acidic intracellular pH (pH), membrane stretch, temperature, and arachidonic acid (AA). Phosphatiéylinositol 4,5-bisphosphate (PIP2) ‘exerts conceattation-Jepencent biphasic regulations, which have been ‘observed: inhibition by high PIP, activation by partial decrease of PPIP,, and inhibition by depletion of PIP,. Consistently, the stimulation cof voltage-sensitive PIP: phosphatase (Dr-VSP) induces initial acti- vation and subsequent inhibition of TREKS. Lys in the proximal (C-terminus (pC0) is responsible for te inhibition by high PIP, whick is generated by phosphatidylinositol kinases with ATP, its neuraiz- ing mutation [KA of human TREK-2 (XTREK-2)] induces tonic high activity, irespective of ATP. Hece we focus on tiple successive Arg in pCt (R3-pC1) as a candidate region forthe stimulatory reg lation by lower PIP;. Their neutralized mutant (R3A-pCt; RRRS!=A and RRR®™"A in BTREK-1 and -2, respectively) showed negligible ‘basal curtent and was not affected by ATP removal or by Dr-VSP activation. Phosphatidic acid, a phospholipid agonist of TREKS, did pot activate R3A-pCt In contrast, acidic pH, AA, and high temper- ature activated R3A-pCt normally, whereas activation by membrane stretch was attenuated. In NTREK-2, combined neutralizations of the inhibitory K® and R3-pCt (C!"ARRR™*A) did not recover the suppressed current. In contrast, combined aculralization of pHl-seas- ing Glu (E"A/R"°7A) induced tonic high current and no further activation by pH, Interestingly, when the Gly between K°/8 and R3-pCt was mutated (G°A), BTREK.2 was tonic activated with reversed responses to ATP and acidic pH. Therefore, we propose that the PIP;-dependent converse regulation of TREKs by Lys and R3-pCt ‘with Gly implies suctural lexiblity (C-terminal PIPs; TREK-1; TREK-2; two-pore domain K channel ‘equsts and ober conespondence: S.J. Kim Depts of Physiology and Biomedical Sciences, Ischenie/lyporie Disease Tania, ‘Seoul National University Collegeof Meine, 103 Dachangno, Jongno-Gi. ‘Seoul 03080, Kore (nal piysolksj@ gral com). cs 0365.6 119 Copyright ©2018 the American Physiological Society INTRODUCTION ‘Two-pore domain K* (K2P) channels assemble as dimers, as opposed to other K* channels with a tetrameric structure Each subunit las four transmembrane domains (TM1~TM4) two-pore domains (PI and P2), and intracellular N- and C-ter- mini. Up until now, 15 members of K2P channels have been identified in various mammalian cell types, and they are divided into six clades, ie, TWIK (andem of P domains in a weak inwardly rectifying K* chanel), THIK (tandem-pore-domain hhalothane-inhibited K* channel), TREK (TWiK-related K'~ channel gene), TRAAK (TWIK-related arachidonic-aci-stimu- lated K channel), TASK (TWix-related acid-senstive K* chan- nels), TALK (TWik-related alkaline-pH-activated K* channel), and TRESK (TWIK-related spinal-cord K* channels). Although K2P channels were initially regarded as “leak-type” background K* channels, subsequent studies revealed complex ways of reg- ulation by intracellular signals and by pharmacological agents. K2P channels also exhibit various responses to a wide variety of physicochemical stimuli to cells, suggesting their critical roles in the physiological sensory and effector func- tions (9, 20, 23, 24, 26) ‘Among the K2P channels, TREK-1 (KCNK2) and TREK-2 (KCNK10) are markedly activated by membrane stretch, poly- ‘unsaturated fatty acid, acidic intracellular pH (pH), anionic phospholipids e-g., phosphatidic acid (PA), temperature, and 2-aminoethoxy diphenyl borinate (2-APB) (2, 9, 16, 23. 24) The strong activation of TREK-1 and -2 (TREKS) indirectly, indicates that their basal activity is maintained at low levels. However, the intrinsic cellular conditions that limit the back- ground activity of TREKs are not well understood, and it has been suggested that phosphatidylinositol 4,5-bisphosphate (PIP,) could be the critical factor (4~6, 29). on channel regulation by PIP», the membrane phospholipids with a large anionic head group, is important in the cell physiology, such as phospholipase C (PLC)-coupled signaling for electrical activity. It is generally thought that cationic amino acids (aa; eg, Lys and Arg) in the cytoplasmic domains of ion channels are responsible forthe PIP,-dependent regulation via electrostatic interaction withthe inner leaflet the plasma membrane. However, the direction of responses, Inept alpeeorg Downloaded fom jounals. physiology: orjournalajpel (112.154.090.060) on fly 8, 2023,‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION C313 i.e., activation or inhibition of the ion channels by PIP, and the gets, such as poly-L-lysine (PLL), and their recovery by PIP extent of PIPs-dependent regulations are variable, depending applied to the cytoplasmic side (Fig. 1A). However, subsequent ‘on the types of ion channels (10-12, 17). Among the PIP.- study by the same group showed both stimulatory and inbibi- sensitive ion channels, the directions of TREKs regulation are tory effects of PIP, apparently depending on the concentration particularly complex and controversial (see below). ranges (6). By site-directed mutagenesis of the proximal cyto- In an early patch clamp study by Chemin et al. (5), PIP, was _ plasmic C-terminal region (pCt) of mouse TREK-1, they pro- suggested as a stimulatory factor for TREK-1 based on two key posed that a cluster of five cationic residues in early pC, ic. findings, i¢., complete inhibition by polycationic PIP, scaven- close to the fourth transmembrane domain (post-M) encom: = = ad SS Pips {| Pu Honore etal, 2007 ~e B ™ T™4 ™e T™ sy exe o rru rmonerataoe c Chemin et al., 2005 Cc ‘Chemin etal, 2007 [nTREK-]... WLRVISKKTKEEVGEFR AHAAEWTANVTAEFKE svel. sk 330, PIP, inhibitory, Woo et al, 2018. Rs-pcT ig I. Schemutie models of TWIK lated two-potedomaia K* chansel (TREK) regulation and candidate regulatory sts in proximal C-erminus (pC) A: ealy rmovel by Honore eal. (13), Schematic drawing of mouse TREKCt dimer with pat of TM4 domain conected 1 pC conaining cationic (ed ele) and anionic {ray ciel) amino avi resides. Top righ itation of Giu™ by acidic itracelllar pT (pt) ie indicated by removing the minis symbol in the gray’ ice Pontve cages isthe pC eectosticlly interact with phosphatidyinositel 4S-bisphoephae (PIP). Top lft epasion between Gla” and PUP, snteropt PiPy-dependent channel sctvasion Bottom: ckmination ofthe eleccsttic interaction between pC and plata membrane by poly-i-ssine (PLL) close the suanel.B: out model of TREK regulation by PIP. Top left: under pysologcal conditions, genctatio of PIP, by phosphatidylinsiel (PD) nase inkubts TTREK. Tep right without ATP, dephosplorslason of PIP, by intlaic pid phosphatase releases TREK trom toni inhibition by PIP Bottom: PIP seavenging ‘with PLL or fore! dephorphorylation a PP, by sstainedaevason of vollage-sestive phosphatase (D-VSP) inhibits TREK. C:requenes of pCt amino se ‘ip hunan TREK-1 QTREK-1) and BTREK-2. Yellow and gieen bores indicat the two clusters of tpleeaole residues investigate in tis study. Pak eucle Indicates pulalive hinge residae Gy?" in ATREK-2. Yellow and blue citles indiate PIPrdependentitbitory Lys” and pHrsensig Glu!” inthe pose region of HTREK2, PIP, stimulatory and inhibitory ranges from Chemin etl (5,6) ate marked (A an meied fom Honore eta. (13) and Woo etl (32), ‘espectvly. by pesmssion fom NetelSprngerPaleave] AdP-Celt Physiol oi t0.11523 172018 - wor ipl o Downloaded tom journals physiology or journalapell (112 154090.560) on uly 8, 2023,cia passing the pH, sensor Glu, i critical for the positive effects ‘of anionic phospholipids (5, 13) (Fig. 1, A and C, PIP: stimulatory), Optical probing of the interaction between TREK-1 pCt and phospholipids also suggested that the in- creased association between the post-M4 pCt and plasma membrane may induce higher activity (26). For inhibitory regulation, a cluster of aa., A334-E346 of mouse TREK-I, next to the post-Mé pCt (late pCt) has been suggested based on a deletion mutant study (4) (Fig. 1C, PIPs inhibitory) However, more recent reports, including those from our laboratory, have suggested inhibitory and nonstimulatory in- teraction of the post-M4 basic motif with PIP; (3, 30, 32) ‘Through fluorescence resonance energy transfer (FRET) anal- ysis method, Cabanos et al. G) suggested that PIP, inhibits TTREK-1 via interacting with the cationic residues in the post-MA pCi, and that the inhibitory influence from the phos- phorylated inositol ring of PIPa is relieved by anionic phos- pholipids with a smaller head group, such as PA. To explain, the positive effects of PIP, on TREK-1, it has been suggested that the conversion of phosphatidyleholine into PA by phos- pholipase D is stimulated, at least partly, by PIP: G, 7). For the effects of PIP, our laboratory has demonstrated biphasic regulations of TREKs by PIP. (21, 32, 33). In whole cell (W-C) and inside-out (J-0) patch-clamp studies, ATP-free condition alone significantly inereased the TREK currents in heterologous overexpression as well asin the native TREK-2 expressing cells (B lymphocytes and astrocytes) (32). Consis- tent with their relatively low basal activity, TREKs are sup- pressed with millimolar ATP in the cytoplasmic side, by the intrinsic level of PIP, generated by phosphatidylinositol (PI) kinases associated with the plasma membrane (21, 32, 33). The inhibitory effects of higher PIP: on TREKs could be consis tently demonstrated by the total inhibition by exogenous ap- plication of PIP: (32, 33). Despite the inhibition by higher PIP2, scavenging with PL or excessive hydrolysis by PLC activation also abolishes the ‘TREK activity. Moreover, after scavenging PIP, with PLL, stepwise increase in PIP> application to I-O patches demon- strated the PIP; concentration-dependent biphasic activation and inhibition of TREKs (32), Concentration-dependent regu- lation by PIP: can be dynamically demonstrated by the inital increase and subsequent decrease in TREK current in W-C patches during the stimulation of voltage-sensitive phosphatase (Dr-VSP) (32). Such results indicate that a certain level of PIP is essential for the activity of TREKs. Based on previous studies (13, 30, 32), the reversible inhibition of TREKs by ATP via memibranc-delimited generation and decay of PIP: by PI kinases and lipid phosphatase are schematically presented in ig. 1B. ‘Through site-directed mutagenesis study, we have recently suggested that a Lys residue (K°"? and of human TREK-1 (TREK-1) and -2 (ATREK-2), respectively; Fig. 10) in the post-M4 Ct is critical for the inhibitory regulation by ATP. ‘When substituted with Ala, both KA (bTREK-1) and K™°A. (TREK-2) showed tonic high activity with no inhibition by cytoplasmic ATP (30). However, the study did not reveal the PIPr-dependent stimulatory site(s), and the identification of positive regulatory site(s) requires further mutagenesis and patch-clamp studies. Previous studies on TREKs Ct deletion mutants have commonly demonstrated markedly decreased activity wien pCt is truncated (15, 19, 25). Thus we hypoth- AUP-Cell Physiol do:10.1152Ijpcelt 00417 2018 - wor aipecl on Downloaded fom journals physiology -orgJounalgpeat (112.134 096.080) ‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION esized that the positive regulatory site(s) for PIPs may be located in pCt next to the post-Mi region Inthe pCt residues of ATREK-1 and -2, two candidate sites bearing three consec- ttive cationic residues gained our attention (Fig. 1C, colored boxes), i¢., 1) RRR“? (hTREK-1) and RRR"? (hTREK-2), and 2) KRK***- (hTREK-1) and RRR*7"* (hTREK-2). ‘On these backgrounds, we explored whether the tuple cat- ionic residues are cxtcal for basal activity of TREKS in both the presence and tne absence of ATP. Through the stady, we newly found thatthe relative proximal tiple Arg in Ct (R3- CD is the eritial polyeationie group in both TREK-1 and -2 (Fig. 10). Next, we investigated whether R3-pCt is also nec- essary for the activation of TREKs by PA, arachidonic acid (AA), acidic pF, high temperature, and membrane stretch, Another intriguing question was the putative interactions be- tween R3-pCt and previously identified regulatory sites in the post-Mé region for PIPy-dependent inhibition and pl, sensi- tivity. Thus we investigated the properties of combined neutal- izing man in STREK 2, ie, Ala substation in R3-pCt with K* (K*°A/RRR™*"A) or with B™? (E™?A/RRR"*"A), Fi- nally, as a hypothetical structural flexibility for the opposite regulation of hTREK-2 by PIP, via post-M4 K*? and RRR**7, ‘we suspected that the only Gly residue between them (G**") may be eritcal, Thus GA of KTREK-2 was investigated. Part ofthe ‘key findings, ie, the positive regulatory role of RRR?5*7 in NTREK-2, has been presented in a conference, Experimental Biology 2014 GD, AND METHODS. Celt culture and preparation. Hunan embryonic kidney (HEK2937) cells were purchased from ATCC (Manassas, VA) and incubated in Dulbecco's modified Eagle's medium (DMEM, GIRCO, Grand Is- land, NY) supplemented with 10% (50/500 ml) fetal bovine serum (FBS; Hyclone, Logan, UT) and 1% penicillin-streptomycin (GIBCO) at 37°C in 20% O;-10% CO., Heterologous expression of KTREK:1 and ATREK:2. The burn complementary DNAs of TREK-1 (sTREK-1, RC214082) and TREK-2 (BTREK-2, $C222775) were purchased from ORIGENE Voltage-sensi- live phosphatase gene of Danio rerio (DE-VSP) was kindly donated by Dr. Byung-Chang Sua (DGIST). The hTREK-1 and BTREK-2 were twansfected into HEK2937 cells using TurboFeet transfection re- agent (ThermoFisher Scientific, Waltham, MA), The day before transfection, 1X 10° celle were seeded in 12-well dish. Twenty-four hours after wansfection, the cells were passaged at a higher dilution (50 cells35-mm culture dish) into fresh medium. In some cases, Dr-VSP and TREK were transiently cotransfected into the cells by a ratio of 1:10. All of the point mutations in BTREK-1 and BTREK-2 were generated using the QuikChenge II Site-Directed Mutagenesis Kit (Agilent Technologies). Sequences ofthe mutants were confirmed bby DNA sequencing Electrophysiology. Cells were transferred into a bath mounted on the stage of an inverted microscope (TE2000-S, Nikon). The bath (015 mi) was perused at 5 mlimin, and voltage-clamp experiments ‘were performed at room temperature (22-25°C). Patch pipetes with a liee-up resistance of ~3 5 and T MA were used for W-C and 10 pate clamp, respectively, The pipettes were connected tothe head stage of a patch-clamp amplifier (Axopateh-IC, Axon Instruments, Poster City, CA), pCLAMP software version 9.2 and Digidata-1322A (Axon Instruments) were used to acquire data and apply command pulses The recorded currents were sampled at 10 kHz and were low-pass filtered at 0.5 (W-C) of 2 Kilz (-0), The stored currents were analyzed using Clampfit version 9.2 and Origin version 8.0 (Microcal, ‘Norlhampton, MA) uly 8, 2023,‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION Solutions s. Normal Tyrode's bath solution for whole cell patch clamp contained (in mM) 145 NaCl, 3.6 KCL 1 MgCl, 1.3 CaCl, 5 glicose, and 10 HEPES [4-(2-hydroxyetbyl)-1- piperazine ethanesulfonic acid], with a pH of 7-4 (titated with [NaOW). The pipote solution for W-C patch clamp contained (in mM) 135 KCL, 6 NaCl, 10 HEPES, 3 MgATP. and 5 EGTA (ethyles glycol tetraacetic acid), with a pH of 72 (titrated with KOH). For WC recordings of ATP-lee concition, MgATP was omitted, and pH was titrated to 7.2 (Fig. 2 and Fig. 8). The pipette and bath solutions ‘of celkattached (C-A) and 1-0 patch clamp contained (in mM) 145 KCl, 0.1 EGTA, 10 HEPES, and 10 sucrose, with a pH of 74 (rated with KOH. For acidic (pH 5,0) solution applied to the cytoplasmic side of [-0 patch clamp, one-half of HEPES was substituted with 5 mM MES [2-(N-morpholinoJethanesulfonie acid ind chem Fig, 2, Inbibition of whol ‘ATP-sre pipet solution war not observed a RRR? bar graph of normalized WC ‘was no dtfrence bet wotage (-V) eurves obtained by a Downloaded from journals py sl (W-C) TWIK:telated two-pore domain K* shanael (TREK) cress byt ke depolarizing pulses (025 Vis) in W-C 2. fooman TREK-I (WTREKC1)) or in RRRO"A (MTREK- utente al “40 mi, according to the amplitude of AAinduced curent (J @ —40 mV. means + SB) in RTREK-1 (C) and re lower im RSAC (od bars) than thore of wil 1ype (WT ‘a WT std the mote dtl ple eatin mtats (teed bas ogy -rgjoumal spell (112134090, c3is For the investigation of temperature sensitivity of ATREK-2 (see 65), temperature contol was achieved by perfusion of prebeated or room temperate bathing solstion, Hot bathing solution was main- tained a the expected temperate with an SH-27B incline solution heater controlled by a TC-S24C temperature contzoller (Warner In- struments), The thermistor was placed right next to the pipette 0 ensure accurate monitoring of local emperstare ‘The chemicals and drugs used in this slady were purchased from Sigma-Aldrich (St. Louis, MO), except for 18:1 PA (catalog no 840875, Avant Polar Lipide, Alabaster, AL). To prepare the PA olution, PA was dissolved in chloroform (stock, 5 mid) and kept at 20°C until use, Before bath application, the PA stock was diluted in bath rnd was sonicated (=30 min). The final level of chloroform in tke bath application wat 0.1%, and we also confirmed sion of REC (proximal Cerin). and dings with ATP-re pipette solution After INI. nommal Tyrode's solton.C and D- open bars), Tere tee OL: 4p = 00 IS aEC316 ‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION thatthe chloroform alone did not affect the activity of TREKS in the used for comparison of data collected from two mutations, One-way LO patches, with or without ATP (data not shown), ANOVA wat used lo investigale satistical diferences among more Statistical analysis, Data age presented as means * SE, Results than two diferent conditions. When significant, Turkey correction ‘were compared using paired or unpaired two-sample test, which was was performed for post hoc testing, P values = 0.05 were considered Tig 3, No effet of vollage-ensitve phosphatase (De VSB) acuvation on R3A-PCr (proxial C-erninus) roltage ULV) curves right ia response t ramp pales {irom —100 to 30 mV, 0235 Vis) and to ssuined (HEK295T) cele tvnsestd with TWIK-tlaed two pote domain K” channels (TREKS) and Dr-VSP. Pie Dr-¥SP by sustained depoaczation (20 mV) indices plate increase and subsequent decree of outward cu ents in wiletype (WT) TREK AL the end of ach ‘experiment 10M arachidonic aid (AA) wa applied to ‘confi fll activation NT, normal Tyrode's soltien reals Ulan @ ~40 ni) before () snd ding depolai {ono 20m (nal peak (and ate decease (0), Ope bare, WTWVSP- soi bars, RRRS""=AVVSP or RRR" ‘VSP; hatched bars, KRKO™TA/ VSP or RRA) sp =" = 001, +P = 000), Downloaded fom jours phy solng-orpjurnalapel (112134090060) on ly 8, 2023,‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION statistically significant. All statistical analyses were done with Origin S soliware (OriginLab, Northampton, MA\ RESULTS Ramp-like depolarizations from —100 to 30 mV (0.26 Vis) were applied at every 10 sto HEK2937 cells expressing wild-type (WT) BTREKs and pCt mutants, and the current-voltage relations (EV curves) of W-C currents were drawn. Consistent with our Taboratory's previous reports (32, 33), WI BTREK-1 and hTREK-2 commonly showed a spontaneous increase in outward currents afer dialyzing the cytosol with ATP-ree pipette solution. In contrast, RRR"°*A (hTREK-1) and RRR'S*’A (STREK-2) showed insignificant spontaneous activation (Fig. 2). In the mu- tants of more distal wiple cationic residues (KRK' "A in BTREK-1 and RRR'”A in hTREK-2), activation by ATP-free solution occurred similar to that in WT. At the end of each Fig 4, Inuct activation by acidic inacellolar pH (pH human TWIK lated tworpore domain K Top: representative inward tothe cytoplasmic change by. ATP. whle activation by pl, 5.0 was intact. KRK™" acovation by acidic pH, sizer to WT. Initia! spontan ofeach experiment, coniplete ition by polyzclysine (PLL the maxim cureat uncer pl, 5:0 without ATP (i pi BIREK2 (0), "*°P = 0001 in RIAD cument ace of WT sowing spontaneous te Downloaded from journals py proximal channel (XTREK)-1 and pt mutants under inside-out -0) path clamp at ~00 mV with symmetesl KC oltions ation ater the meninaneexeision, reverie inhibi by 5 mM MgATP app ‘td the acvaion by pH 5.0. RRES"™A of RTREK-] and RRR™ oe activtions of KRK™ 10 gh) was conte. ogy -rgjoumal spell (112134090, ca recording, 10 »M AA was applied to induce maximum activation of TREKS (Jaa). The amplitudes of Jag at ~40 mV in hTREK-1 were 1,345 +744, 1,581 = 444, and 1,360 + 943 pA for WT (n= 9), RRR™A (n = 9), and KRK™*SA (n= 8), respec- tively. In BTREK-2, Ja, amplitudes at —40 mV_ were 07 * 15.1, 1,398 = 140, and 1,362 = 164 pA for WT (n = 2), RRR'S*"A' (n= §), and RRR'”A (n = 5), respectively The similar amplitudes of I4a from the different groups sug~ gested that the membrane expression and the responses to AA were intact in the mutants tested in this study. For comparison between groups, normalized currents (I/Ia) ate summarized in Fig. 2, C and D. Next, dephosphorylation of PIP2 was induced by activation of Dr-VSP coexpressed with TREKs. Ramp pulses from — 100 to 30 mV with ~80 mY of holding voltage were applied every 10's. After confirming stable W-C current with 3 mM. mins) mutants and B: representative cute traces of wildtype (WT) ‘A of MTREK-2 showed neghgible basal euent witout sgnieast, WoAaad RRR'™"A were truncated inthe representative traces, At he etd ‘nd D: stony ofeurent amplitade st ~60 mV normalized to IS aEC38 the pipette, holding voltage was changed from —80 to 20 mV, and this induced biphasic changes in the outward current in’ WT TREKS. Both the holding current at 20 mV and FV curves showed initial increase and subsequent decrease to below the initial level (Fig. 3, A and D). In contrast, RRR™™A. (xTREK-1) and RRR!**"A (ATREK-2) showed no significant increase at the initial period of depolarization to 20 mV Fig. 3, B, D, and B). The neutralized mutants of the more distal polybasic motif in KTREK-1 (KRK"***"A) and hTREK-2 (RRR"™”"A) showed biphasie changes similar to those of WT TREKs (Fig. 3, ¢ In LO mode of patch clamp recordings at ~60 mV, signif- icant activation of WT TREKs was always observed after ‘making a membrane excision into ATP-free KCI bath solution. ‘The application of ATP (3 mM MgATP) to the side bath solution inhibited the inward current of recording (nmax.zo) in a reversible manner, as previously reported (32, 33). Also, consistent with the known property of ‘TREKs, acidification ofthe eytoplasmic side pH (pI) from 7.0 to 5.0 (pil. 5) increased Freak... With or without ATP (Fig. 4 C and D, open bars). Finally, the application of PLL at the end of the experiment completely suppressed fresco. indicating the positive effects of PIP: and anionic phospholipids (Fig. 4). By contrast, RRR™°2A (ATREK-1) and RRR™*"A.(STREK~2) showed insignificant spontaneous activation from negligible lev- cls of basal current and were not affected by ATP application (Fig. 4, C and D, solid bars), Nevertheless, activation by pl, 5 was intact in both RRR™°"PA and RRR™™"A (Fig. 4). In KRK™#"A (hTREK-1) and RRR"””*A (MTREK-2), sponta- ‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION neous activation by I-O configuration, the ATP-sensitivity, and pH, 5 activation were well preserved, similar to those in WT (Fig. 4, C and D, hatched bars). Previous studies have shown stimulatory effects of PA, anionic phospholipids with the head group smaller than that of PIP: (3, 4, 7). Therefore, we examined whether the positive effect of PA was altered in RRR™°7A (hTREK-1) and RRR'*"A (HTREK-2). In LO configuration, the activation of WT TREKs by 5 uM PA was observed in both the presence and absence of ATP (Fig. 5, A and D). By contrast, in RRR and RRR'557A, application of PA did not increase Irasxe1o, irrespective of ATP, whereas the activation by acidic pH, was consistently observed (Fig. 5, B and E). Sum- muarized results of the effect of PA are shown as bar graphs in Fig. 5, C and F ‘Next, we examined whether the temperature and the stretch, sensitivities of KTREK-2 were altered in RRR***"A. In W-C patch clamp with ATP in the pipette solution, ramplike pulses were applied at different temperatures (Fig. 6, A and B), Maximum activation was induced by 10 pM AA (qq). Al- though the normalized W-C current (IH/,,) of RRR™™'A was lower at 26 and 30°C than that of WT, the normalized current at 40°C was similar to that of WT (Fig. 6C) ‘The mechanosensitivity of KTREK-2 was analyzed in the cell-attached configuration, where negative pressure through a patch pipette was increased from 0 to 14.7 and then to ~29.4 mmllg for 60-70 s, respectively (Fig. 7, A and B). At the end of experiment, 10 ¢M AA was applied to confirm the maxi- mum level of ATREK-2 in the C-A recording. The averaged Fig. 5, No activation of R3A-pC1 (proximal Cermins) by phosphatidic aid (PAY, A and De representative current traces of inside-out (LO) patch clamp in vwldaype (WT) TWIKcrelated two-pore domain K channels (TREKS) at ~60 mV’ with syn ytne Band E: representative current races showing oo reaponse of R3A-pCt mutans to PA. Cand F sunnaty SuMPA.incspecive of ATP. PLL, poly cof inward K" curent al ~60 mV normalized (othe coment sctvated by aise treslular pt (pi (1 pl, § ® ~®0 mV) im WT and RBA-PC (RRRUWA: F) oP = 0.05; "P TREK] (WTREK-1) (RRRO™A; C) and Downloaded tom journals py soloy- ofp journal apell (112 154000. sical KCI solutions, Tae inward K eure wae inereared by ot 0) on Duly 8, 2023,‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION A TREK-2, WT 150-80 -60 -40 -20 Vim) currents during the last 10 s of each pressure level were calculated and normalized to the same duration of averaged currents in the presence of AA. The normalized channel activity showed stretch-dependent activation in WT hTREK~ In RA, statistical significance was confirmed at ~29.4 but not at —14,7 mmlg (Fig. 7C, solid bars), In addition, activity at ~29.4 mmlg as well as at 0 mmlg was significantly lower in RRR™™7A than those in WT (Fig. 70). Negative pressure higher than ~29.4 mmllg was not applied due to the instability of the patch membrane ‘As mentioned in the nvrronucios, we have recently found that neutralizing mutation of the post-M4 Lys (KA) in BTREK-2 and hTREK-I (K"°A) markedly increases the basal activity with loss of ATP inhibition, and no further activation by acidie pH; (30). Representative I-O and W-C patch-clamp recordings of K"°A hTREK-2 are also shown in Fig. 8, A an B. However, the additional K”°A mutation in RRR™"A (KC ARRR™*"A) did not recover the suppressed channel activity, imespective of ATP application (Fig. 8C). Activation by acidic pH, in I-0 patch clamp (Fig. 8©) and the activation by AA in the W-C patch clamp (Fig. 8C) were consistently observed for K™°ARRR"*"A (Fig. 8, G and H, open bats). Similar to E of mouse TREK-1, Ein the post-Mé region of KTREK-2 is the eritical residue for activation by acidic pH, (Fig. 1, A and C).Itis known that its Ala substitution (EA) leads to a fully active state, with no further activation by acidic pH, and no inhibition by ATP (30). Interestingly, the combined mutation of E™ and RRR"*7 (E™=A/RRR'*7A) revealed higher basal activity without pH, or ATP sensitivity similar to the property of "A (Fig. 8, E and G, solid bars), Consistent with FO recordings, the high W-C current of E'PARRR™™"A did not show an increase with ATP-free pipette solution and negligible further activation by 10 uM AA. (ig. 8, F and H, solid bars) Downloaded tom journals py soloy- ofp journal apell (112 154000. C319 ig. 6, Temperature-dependent activation of widype (7 human TWik-rlsted two-pore domain K el QATREK)-2 and R3ApCtpvoxin C-terminus) tnd current-voltage (1) curves of whe cel (WO) patch clamp in WT ATREK2 (A) and R3A-PCT (RRR™™TA, B) that shows teperstredependest ine ‘cease of eurent amplitudes, 10 nM arachidonic wed (AA) was applied for maximum setivation. NT, nora ‘Tyrode's volition. C: summary ofeurent amples st “so mv) in WT (open bars n= 7) and RBA-PCt Gola bars, = 3), Aldiough RRROA showed 1 Dldes than WT at 26 and 30°C: the amples at HO'C mere not diferent “P< 0.05; ***P = 0.01, Gly is the simplest form of proteinogenic a.a. and provides flexibility to the protein structure. There is a single Gly (G"? of hTREK-1, G of hTREK-2) between the post-Mé region and R3-pCt, When substituting the G?™ with Ala (GA), IO recordings of hTREK-2 showed tonic high activity, similar to that observed with E™*A without spontaneous activation after making I-O configuration (Fig. 9). Perplexingly, the responses to pl, and ATP were opposite to those of WT, with partial inhibition at pH, 5.0 and slight activation by 3 mM ATP (Fig. 9, A and C). In the W-C recording of GA TREK-2 without ATP in the pipette solution, the amplitude of the initial current was already high without spontancous increase and was not farther increased by 10 uM AA (Fig. 9, B and D, n = 6). DISCUSSION ‘Our present study demonstrated that R3-pCt is critical for the higher activity of TREKs under moderately decreased PIP; condition, As previously reported, the Lys in the post-M4 region may conversely regulate TREKs under higher PIP, Because combined mutation of the inhibitory Lys (K™°A RRR’™7A of hTREK-2) did not recover the suppressed activ- ities in R3A-pCt (Fig. 8), the interaction between R3-pCt and PIP; appeared (o be the prerequisite factor for TREK activation by partial decrease in PIP Intriguingly, we also found that the single mutation of Gly (ie, GMA in TREK-2) between the oppositely acting motifs dramatically increased the basal ac- tivity and reversed the direetion of responses to ATP and pH, ‘One may argue thatthe biphasic effects of PIP: on TREKS, particularly activation by ATP removal inthe cytoplasmic side, may be mediated by a partial decrease in PIP2 and increas PIP (ic, phosphatidylinositol 4’-monophosphate, PI4P). How- ever, our previous study demonstrated that PI4P alone does not affect TREK-2 activity, but shows inhibitory effect only in the 0) on Duly 8, 2023,320 7. Lower mechanoeeasiivty of R3A- CK (proximal C-terminus) tan, wil-ype sain K* channel (BTREK)-2. A and B: ep feseattve. cure traces of cellataced A) patch-clamp recordings in WT (A) and RRR" (B) BTREK-2, The patch emivane voltage wat held at 60 mV Ader symmetneal KCI condition, Negative resure Tyough the patch pipete was ib ‘teaaed from Oto 14.7 untig and 29.4 ‘mg for 10st the end ofeach exper: ‘ent, 10 yM arachidonic acid (AA) wat ‘plied for normalization of the sustchle pendent activation (eurent ree nol shown CC summary of the nomalizeé curents Ing, @ ~60 nV) of WT (open bars,m = 6) sod RRR™"A (oid bares m= 6). 4P = 005; +4? < a0! presence of ATP (21). Thus, it is more likely that PIP; generated from PIP and ATP by PI kinases plays an essential role in the membrane lipid-dependent regulation of TREKs Gig. 1B). Another possibility was the conversion of PIP: into phosphatidylinositol (3,4,5)-trisphosphate (PIP) by ATP and ISK. However, a previous quantitative analysis has demon- strated that the relative amount of PIP, is very low among anionic phospholipids in the plasma membrane (22). In this respect, the physiological role of PIP, as a direct regulator of ‘TREKS is unlikely Role of R3-pCt for positive regulation by anionic phospholipids. Based on the present results and our previous study (30), we present a schematic summary of the pCtsite- dependent regulation of TREKs by PIP3, ATP, and pH, (Fig. 10, A-C). The key is that the loss in electrostatic binding of R3-pCt by mutation or excessive depletion of PIP. severely impairs the basal activity of TREKs (Fig. 10D). Moreover. itis proposed that the increased activity with partial decrease in PIP; or by neutralization of post-Md Lys (K**"A) commonly requires the membrane-bound state of R3-pCt (Fig. 10B). Itcan be assumed that R3-pCt may have higher affinity to PIP than post-M4 cationic residue(®) ‘Although not included in the schematic model, the stimula tory effect of PA on TREK: also required the cationic charges of R3-pCt. It was notable thatthe activation of WT TREKs by Downloaded tom journals py soloy- ofp journal apell (112 154000. ‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION PA was independent of ATP (Fig. 5). Although a recent study by Cabanos et al. (3) suggested a counteractive competition between PA and PIP: to the inhibitory post-M4 region of TREKs, our present study suggested that activation by PA required R3-pCt that is located slightly distal to the inhibitory post-Mé region Other poly-basic motif distal to R3-pCt [eg KRK*** (STREK-1) and RRR"””? (bTREK-2)] appeared to play insig- nificant roles because their neutralization did not alter ATP- sensitive basal activities and the intial activation of TREK by Dr-VSP (Figs. 2-4). However, R3-pCt may not be solely responsible for the positive influence of anionie phospholipids on TREKs, because the unspecific scavenging with PLL. still abolished the high basal current of E™”A/RRR'™*"A (Fig. 82). Although the R3-pCt seemed to be the main critical site for the activation by anionic phospholipids, further investigation of other cationic residues in the cytoplasmic and the intraceliular sides of transmembrane regions may reveal additional PIP,~ dependent regulatory sites. Effects of other stimuli on R3A-pCt activity, Different from the effects of PIPs and PA, R3-pC1 neutralizing mutants were normally activated by acidic pH, and AA. Although not shown here, we have also confirmed that W-C current of R*°7A (XTREK.2) was effectively activated by 2-APB or external acidic pH (pH. 6.0), similar to that of WT. Temperature 0) on Duly 8, 2023,‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION Fig. $,Propeits of oman TWIKcled two pore domain K* chen Claminu) A tepesenve caren! tac of nest 0) Ghats ten igh cet yao ‘ we6o nV in ROUARRO A The coi iy eure of WC pach clamp wih ATPte pipe ssn Re tropes cent ate 1 pal foes of Woe pt lap with ATP pipe soln in "AR Tinea lin @ 240 mV om We econ tn RO MRRR™A sensitivity of R'¥*7A decreased slightly, but activation by a higher temperature was basically preserved. The strong activation of RRR"™™’A by AA w observed in the present study. According to a previous study consistently Downloaded from journals py (QTREK)-2 with combined mt patch clamp at “6D mV in OA so further activation by aaehidonicaei (AA). NT, normal Tyrode's rlston, Crepreretaive caren peutalization of K° did not recover bas hclamp st 60 mV in E*ARRR"™A showing large baal eaten wilhot A, showing tonic high eurent without further activation by AA. G: sary of a bas) oe ogy -rgjoumal spell (112134090, C321 of post resides and R3-pCt (proximal ig igh basal cent without responses to mM MgATP ‘eof 0 patch clamp iy alliough the setivaion by acide pH. was preserved. D: A showing ne spontaneous activation, albeit intact activation by 10 uM AA, strom pi, 0 ang ATP. FV 1 topen brs) and E®"AIRRR™"A (oid bas), Harmar of normalized "NIRRRO™"A (oid bate) truncated deletion of TREK-1 Ct up to Tie, proximal to the R3-pC, showed robust activation by AA, whereas further tuuncation of the post-Mé region abolished the AA effect (25) Furthermore, clustered deletion of K™KTKEE™* in TREK~ IS aE322 Fig. 9. Tone high activity of human TWIK-related two-pore domain K chan (Cerminoe) A epreseattive current trace of nsice-ot (1-0) pate clamp in holding vole). Application of 3 aM ATP and aide intracellular pl (9H) 5.0 B: representative caret arachidonic acid (AA), C. summary of normalized eurenls (0 ATP @ ~60 mV (@ 40 mi) tom the W-E recordings in G™*A, *P = 005, "P= ODL abolished activation by AA, but their neutralization (substita- tion with neutral a.) did not alter activation by AA (15). ‘These reports indicate the importance of the post-MA structure rather than R3-pCt for the activation by polyunsaturated fatty acids such as AA. However, different from the effect of pli, and PIP;, specific residues for AA effects have not been pinned down in TREKs yet Intriguingly, the activation of ATREK-2 by negative pres- sure was consistently attenuated in RRR"**"A, suggesting that electrostatic binding of late pCt (ie., R3-pC) to the plasma membrane could facilitate gating by membrane stretch. Recent studies on mechanosensitive gating in TREKS have also sup posed the membrane-bound state of pCt in their molecular structural models (1, 8, 18, 23, 28, 34). In addition, we have previously reported that stretch-sensitive phospholipase activa- tion lowers PIP; to the optimal range for activation, thereby facilitating the activation of TREK-2 (21, 33), In this respect, the attenuated mechanosensitivity of R"™7A may be partly due to the loss of PIP2 sensitivity Interactions between R3-pCt and post-M4 regulatory sites. Tn contrast to the severely suppressed basal activity, acidic endent activation was intact in R3A-pCt (Fig. 4), Sin hhas been identified forthe first time in mouse TREK-I asthe critical pH, sensor (14), it has been consistently observed that its, neutralizing mutation (€.g., E”°A of BTREK-2) induces 4 ‘maximum activation (5, 30), In the early hypothesis of TREK-I regulation, Chemin etal. (5) proposed that titration of Glu would facilitate the positive interaction between PIP and the post-Mé polybasic motif (Fig. 14). However, our present results strongly suggest that the pH, sensitivity of TREKs is indepen dent ofthe charged state of R3-pCt. Infact, additional mutation ltage(-¥) carves of whole cell (WC) reconnge io Downloaded tom journals py soloy- ofp journal apell (112 154000. ‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION Je (STREK)-2 with mutation of Gly between pos-MA and R3-pCt (proximal “A showing high basal cureat (symametical KCl coadition with ~60 mV of induced paral inteate and decrease of the inward K” current, respectively A that shows tonic high seavity wilh! further activation by 10 ut om te LO recordings in A.D: summary of normalized cum an of Glu (E*A/RRR"*7A) could reverse the negligible basal activity of RRR*"A to the maximum level without further activation by AA (Fig. 8) Instead ofthe facilitation of positive inleraction via R3-pCt, the inhibitory interaction between post-Mé K"!? (K) and PIP, may be relieved by titration or neutralizing mutation of Glu (Fig. 10C). However, it has to be conceded that the intact pH; sensitivity of K??A/ RRR™*7A (Fig. 8) was not consistent with the one-dimensional model In addition to the evident role of titratable Glu (.g., B™ of, TREK-2), pH, sensitivity of TREKs may also be alfected by the charged states of PIP2, A previous “1P-NMR spectroscopy has revealed that negative charges of PIP2 could be reduced from —5 to ~3 by acidifying the pH from 8 to 5 (29). As the first protonation of PIP; could occur at the physiological range of pH, (29), the activation of TREKs by acidic pH, may partly reflect a putative reduction in the inhibitory interaction tween PIP, and post-Mé Lys, However, the potential involve- ment of PIP2 titration in the acidic pH, sensitivity of ion channels has not been determined yet Precise three-dimensional structure of TREKs, including Ct, would be helpful to fully understand the complex regula- tory mechanisms. In a recent report of TREK-1 crystal stn ture, Lolicato et al. (18) showed the a-helical structure of pC. However. the information depicted in the paper did not yet provide helpful mechanistic insight for our present results of the counteracting regulation between post-M4 residues. and R3-pCt. Moreover, the intriguing possibility of Gly (G'™) as the flexible hinge for the opposite regulation was not consid~ exed in the study by Lolicato et al. (18). According to our present study, the saturated activity of G?*A may be due to the putatively rigid achelical pCt. In other words, flexibility of 0) on Duly 8, 2023,‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION 323 A T™4 T™4 B Decreased PIP K2304, Fig 10, Scheme summary of man TWIKcrelated owo- a 330 pote domain K~ channel QETREK)-2 regulation. Drawings 2* Og! FRRSST eof TREK? dine wi TMG aC Te tical egslatry esgves (Kana E™) so post Man E382 G34 [RS-pCt(geosimal C-erminus) residues ate indeated with colored ieles (ed and gray eeles for the clonic and ‘bionic amino aciés, respectively). Green ele, puaive Hinge G'™ between pose TMU and R3-pCt. A low-actvty . . fale of BTREK-? wih eaively high phosphaylinone c Acidic pHi E320 or G44, 4S-baphospbate (PIP) bound wih! K™™ and. R3-pCt (Con), Bbigh-acuity sate with deceased PIPs or wath neutralization of posta Lys (KA: High). Cbighe etsy rate with tration of !, neulzing mation (7A), of deteaved fexsbilty (GA), Doss of acuty ; by neutralization of RRR" (Closed a ta G34 Proton D R357 G may allow post-M Lys-dependent inhibitory regulation with ATP, ic, higher intrinsic PIP: (Fig. 10C). In addition to the tonic high activity of GA, opposite responses to ATP. and pl; 5.0 indicate the importance of understanding the actual shape of pCt a-helix. While preparing this manuscript, an intriguing paper was published that demonstrated the regulatory role of triple Arg residues in the basal activity of hTREK-I (transcription variant 2) 27), The triple Arg residues correspond to R®!** and R=" of hTREK-1 and-2 in our study (transcription variant 1). In the study comparing the roles of pCt between TREKS and TRAAK, Soussia et al. (27) have suggested the tiple Arg residues of TREKs as a ctitical motif for the higher basal activity of TREKs than TRAAK, which contains QRA in the corresponding sites. It was notable that Soussia et al. used an ATP-free pipette solution for W-C patch clamp, which would have enhanced the basal activities of TREKS, but not those of TRAAK lacking triple Arg (ie., R3-pCD. They also demon- strated that the inhibition of TREKS by excessive VSP stimu- lation was abolished by swapping the tiple RRR with QRA, hich are the corresponding residues of TRAAK. However, activations of R3-pCt mutants by acidic pH, temperature, and AUP-Cell Physiol do:10.1152Ijpcelt 00417 2018 - wor aipecl on Downloaded fom journals physiology -orgJounalgpeat (112.134 096.080) 330 33209) RRRI5S- 7A membrane strech were not investigated in the study by Soussia et al. (27). Although their study focused on the differential responses between TREKs and TRAAK in terms of the gating mode (up- or down-states and channel activation), our present study provides experimental clues to understand the regulatory interaction between post-M4 (K™® and E?) and R3-pCt via G** hinge Fig. 10). ‘Taken together, site-directed mutagenesis of pCt in TREKs and their patch-clamp analysis demonstrated intriguing roles of R3-pCt interacting with the post-M4 regulatory domain. More- over, the dramatic functional changes in G"**A. propose that putative “flexibility” in pCt achelix is also critical for the counteractive regulations by the early and late domains of pCt. Future investigations of TREKs and elucidating the pCt struc ture would expand our understanding of TREKs and their perplexing gating mechanisms GRANTS “This work was supported in part by National Research Foundation of Korea (NRE) Grant funded by the Korean goveroment (NRF-20ISRIASA2025968 and NRF-2018R 1D1A 1807088996). uly 8, 2023,324 DISCLOSURES: No cont of interest, nancial or otherwise, ave declared by the authors AUTHOR CONTRIBUTIONS LW. abd SK. conesived and designed research, JW. VK, HUN. and DAIS. pesformed expermens: |W. and Y KJ. analyzed dala: 1W,, YK, YHZ.THN, DHS. and STK inerpreed results of experiments, 1W. YI and DHS. popared gues; LW. dated manuscript, YZ, LIN and SR. edited and tevited manuscripts YILZ. and SK, approved final ‘essoa of manuecript, REFERENCES 1. Aryal P, Jareratanachat V, Clausen MY, Schewe M, MeClenaghan. Argent L, Conrad LI, Dong YY, Pike ACW, Carpenter EP, Baukrowite 7, Sansom MSP, Tucker SJ. Buayer-medited ster ‘wansiions contol mechanotensivity ofthe TREK-2 K2P chasse. Struc fare 25: 108118 62, 2017, do10 1016s 2017 03.006. 2. Beltran L, Beltran M, Aguado A, Gisselmann G, Hatt HL 2-Amino- tshoxydipenyl borate seats the mechanically gated human KCNK. ‘hanaels KCNK 2 CTREK- 1), KCNK 4 CTRAAK), an KCNK 10 (TREK 3). Front Pharmacol $63, 3013. doi:10 3380fphat 201300063, 4. Cabanes C, Wang M, Han X, Hansen SB. A soluble fuoreseen binding say reveals PFs antagonism of TREK-I channels. Cell Rep 20° 1287 1294, 2017. doi:10 lot} eetep. 2017 07034 4 Chemin J, Patel AJ, Delmas P, Sachs F, Lazdunski M, Honore E- Regulation of the mechano-gated KOP channel TREK-I by membrane phospholipids. Curr Top Membr 59: 155-170, 2007. do 10 1016/8 1065~ 5425406339007 5. Chemin J, Patel AJ, Duprat ¥, Lauritzen I, Lardunski M, Honoré B. ‘A phospholipid sensor contols mechanogating of the K™ chanael TREK-1 EMBO J 26 44-5, 2005, do 10,1038) eho, 7600854 6, Chemin J, Patel AJ, Duprat¥, Sachs F, Lazdunski M, Honore E. Up- ad dow-egulaton of the mechanoguted Kir channel TREK] by PLP fd other membrane phospholipids. Pagers Arch 455: 7-103, 2007 foi: 10 Looroota007-0950.2 1. Comoslio Y, Levty J, Kinsler MA, Lesage F, Isacff EY, Sandor G. Phosphlipaze D2 specifically regulates TREK potassium channels va ‘ze! interaction and local prodecuon of phosphaticic acid. Proc Net! ‘Asad Sey USA 111. 18547-13952, 2014, da101073/pnas 1407160111 8. Dong YY, Pike ACW, Mackenzie A, MeClenaghan C, Aryal P, Dong 1, Quigly A, Grieben M, Goubin'S, Mukhopadhyay S, Ruda GF. Glausen MV, Cao L, Bremhan PE, Burgess-Brown NA, Sansom MSP, Tacker SJ, Carpenter EP. K2P channel gating mechanisms revealed by structures of TREK-2 and a complex with Prozac. Science 347 1256 1239, 015. doi: 101 126cienee 1261512 9, Raye Crirjik G, Molecular background of leak K* cutets two- ote domain potassium chanbels. Pye! Rey 90: 559-405, 2010. do 0, 1132p 000282008, 10, Gamper N, Rohacs T. Phosphoinositide sensitivity of ion channels. a ‘unetional perspective, Subeel Bockem $9: 249333, 2012, do 1010017 978-94-007-3013-1-10, 11, Haneen SB, Lipid agonism: The PIP2 paradigm of ligand-gated ion channels. Biochim Blophys Acta 1451: 620-628, 2015~ do: 10.1016, balip 2015 01011. 12, Hille B, Dickson KJ. Kruse M, Vivas O, Suh BC. Phosphoinosides regulate jon chansel. Biochim’ Biophys Acta 1851. Si4-856, 2015, 44x10 1016 bap 2014.08 010, 13, Honoré E The neszonal background KOP channels: fous on TREKI. Nat Rey Neurovet 951-261, 2067, do: 101038n21, 14, Honoré k, Maingret F, Lazdunski M, Patel AJ. An intracellular proton sensor commands lipid abd mechano-gatig of the K* channel TREK-L Eug0 21: 2968-0916, 2002 ds 10 109sVembojed38 15, Kim Y, Gnatenco C, Bang H, Kim D. Locabration of TREK.2 K anne! domains that regulate channel Lnetics an sensitivity to pressure, Saty acids and pl. Pflugers Arch 442: 952-960, 2001, do 10 1007) 008240160696, 16, Lesage F, Terrenore C, Romey G, Lardunski M. Human TREK, «2? domain roechano-enstive K" channel wih rauliple regulations by poly ‘nsatrstes fay seid, Iysophoepholpids, and G, Gh and Gy protein AUP-Cel Phil i 0 Sil 00812018 wo ipl on Downloaded frm joumals physiology oop journal apell (112154090 2. 2, 2. 25, 26. ». a a. M. ‘TRIPLE ARG-DEPENDENT TREK CHANNEL REGULATION ‘coupled receptors. J Bio! Chem 275: 28598-28405, 2000, do:10 1074) she 003429000 Logothetis DE, Petrou VI, Aduey SK, Mahajan R, Channelopathies linked to plasva membrane phosphoinosisdes. Pflugers Arch 460. 321— 341, 2010, doi10 1007-00825-010-0828 Lolicato M, Arvigonl C, Morl T, Sekioka Y, Bryant C, Clark KA, ‘Minor DL Ir- Kir 1 (TREK-1>-tvstor complenes reveal a cryptic selectivity ltr binding site. Nature 547. 364-868, 2017. doi 10 10387 ate, ‘MaingretF, Patel AJ, Lesage , Lazdunski M, Honoré E. Mechano- 0: acid _sbmolaion, wo interactive moder of aetivaion of the TREK: Potassium channel J Blal Chem 274: 26691-26696, 1999. doi:10.1074/ e278 38 25681, Mathie A, Veale EL. Two-pore domain potassium chansls: potenti ‘beraprie get forthe teatment of psn: Pfarere Arch 467: 931-943, 2015, dis0 1007/.00424-014-16553, ‘Nam JH, Lee HS, Nguyen YI, Kang TM, Lee SW, Kim HY, Kim SJ, Farm YE, Kim SJ. Mechanorenstive activation of K° channel via hombolipae Cindiced depletion of pospatdylinasitl 4, Spor hate in B lymphocytes, J Physiol 582. 977-090, 2007, doil0 1117 Shi 20072843. Nasuhogla C, Feng S, Mao J, Yamamoto M, Yin HL, Earnest S, Barslko B,ibansi JP, Hilgemann DW. Noncadescive snags of Phombaginodtdes and ofr anon phospholipids by anomsnebngs Iguperormance ligid chromatography wh suppesed conduc tection. Ano! Biochem, 301" 243-238, 2002 dor 10 loose 201 Sis ‘Niemeyer Ml, Cid LP, Gonsiles W, Seplveda FV. Gang, regain. and sructre a K2P K> channels in vance concordat Pharmacol 50: 309-317, 2016, do-10.1124)mo 176105895, Nod J, Sando G, Lesage F Moleult eglatonsgoviing TREK and ‘TRAAK chante! funetons, Chanel (Aurtin) 02-808, 201, do 0 AMeliehun 551668, Patel AJ, Honoré Hy Maingret F, Lesage F, Fink M, Duprat F, LLardunski MA anil vo pote domain tichano-gaed Slike K henel, EMBO 17283-1290, 1998 do 10 109tenbo17 15-983 Sandor G. Bell SC, Lace EY. Opal roing of a ypani imbrane interaction that oglates he TREKI chanel. Proc Nell Acad Sei USA 10k 260-2610, 2011 dos 101073pma 015788108, Sousa IB, Choveau FS, Blin S, Kits BJ, eliangel S, Chatelain FC, Kang D, Biche. Lege F- Astagonisic effet of stole domain tn the baal atvy of polyenadal passim channels: Pron ot New oct 1-301, 2018, do'0 3389201800301, “Treplow W, Klin MI The membrane-bound sae of KAP potash hanel. dm Chem Sc 132: 6145-8151, 2010 do 102101021915 van Pardon PA, de Kruif B, Ouwerkork R, Wirtz KW. Polyphor Phsinostdes snag chrge ntrlzaton a the pyle pl ange EIPAMR study. Biochim Biohye Acta #77. $16-219, 1986. dot 0 Totsioans-270186)90137-9. ‘Woo J, Jun YK, Zhang YH, Nam JH, Shin DH, Kiss J. esiestion of encl sine seid in th proximal terminal of TREK-2 K cane! for acvaion by acide Bs abd ATP-dependst aibion. Pflaers Arch $70, 527-337, 2018 dor 10 100700050017-2072-1 ‘Woo J, Kim HJ, Park KS, Shin DH, Zhang YH, Nam JH, Kim ST ‘osiiy of TREK“2 Kv channel ometbane Ps abd Bk: FASEE J 28, Suppl I: LBS, 2018. do 1010966, 26 | -sopleen: BS ‘Woo ), Shin DH, Kiss HJ, Yoo HY, Zhang YH Nam JH, Kim WK, Kim i, lhibiuon of TREK-2 K* chanel by PlasP an inne mode regulation by. itaclhlar ATP via poepbatdynostal. Kinase Phogrs Arch 404: 1389-1402, 2016, do 10 100160824 01618470, “Aveng H, Nam JH, Pang B, Shin DE, Kim JS, Chan YS, Pask JW, ang i Kim WK, Baron YE, Kim J. Isstieion of the large condita backgeouné Kr chase in nous B eel at TREK2. Am 7 Phyl Cel Phot 297, CL88-C197, 2008, do 10 1152p 0032 om. “Zhuo RG, Peng P, Liv XY, Van HT, Xo JP, Zheng JQ, Wei XL, Ma XX. Allonene coupling between pronimal Creominae tod ecity ‘esis failed by tbe movement of tasumerbrane segment 4 10 ‘TREK.2 channel. Se Rep 6° 21248, 2016. dos 10 L0SHeep2 1288 ot Duy 8, 2023