1521-0103/361/2/209–218$25.00 https://doi.org/10.1124/jpet.116.238287THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 361:209–218, May 2017Copyright ª 2017 by The American Society for Pharmacology and Experimental Therapeutics

Dual Mechanism for Inhibition of Inwardly RectifyingKir2.x Channels by Quinidine Involving Direct PoreBlock and PIP2-interference

Christoph Koepple, Daniel Scherer, Claudia Seyler, Eberhard Scholz, Dierk Thomas,Hugo A. Katus, and Edgar ZitronDepartment of Cardiology, Medical University Hospital Heidelberg, Heidelberg, Germany (C.K., D.S., C.S., E.S., D.T., H.A.K.,E.Z.); DZHK (German Centre for Cardiovascular Research), partner site Heidelberg/Mannheim, University of Heidelberg,Heidelberg, Germany (C.S., E.S., D.T., H.A.K., E.Z.); and Department for Hand-, Plastic- and Reconstructive Surgery, BGUnfallklinik Ludwigshafen, University of Heidelberg, Heidelberg, Germany (C.K.)

Received October 13, 2016; accepted February 8, 2017

ABSTRACTClass IA antiarrhythmic drug quinidine was one of the firstclinically used compounds to terminate atrial fibrillation and actsas multichannel inhibitor with well-documented inhibitory effectson several cardiac potassium channels. In the mammalian heart,heteromeric assembly of Kir2.1–2.3 channels underlies IK1current. Although a low-affinity block of quinidine on Kir2.1 hasalready been described, a comparative analysis of effects onother Kir2.x channels has not been performed to date. Therefore,we analyzed the effects of quinidine on wild-type and mutantKir2.x channels in the Xenopus oocyte expression system.Quinidine exerted differential inhibitory effects on Kir2.x chan-nels with the highest affinity toward Kir2.3 subunits. Onset ofblock was slow and solely reversible in Kir2.2 subunits. Quinidineinhibited Kir2.x currents in a voltage-independent manner. By

means of comparative Ala-scanning mutagenesis, we furtherfound that residues E224, F254, D259, and E299 are essential forquinidine block in Kir2.1 subunits. Analogously, quinidinemediatedKir2.3 inhibition by binding corresponding residues E216, D247,D251, and E291. In contrast, Kir2.2 current block merely involvedcorresponding residue D260. Using channel mutants with altered(phosphatidylinositol 4,5-bisphosphate PIP2) affinities, we wereable to demonstrate that high PIP2 affinities (i.e., Kir2.3 I214L)correlate with low quinidine sensitivity. Inversely, mutant channelsinteracting only weakly with PIP2 (i.e., Kir2.1 K182Q, and L221I) areprone to a higher inhibitory effect. Thus, we conclude that inhibitionof Kir2.x channels by quinidine ismediated by jointmodes of actioninvolving direct cytoplasmic pore block and an impaired channelstabilization via interference with PIP2.

IntroductionOriginally derived from cinchona bark, class IA antiar-

rhythmic drug quinidine was the first clinical used compoundto terminate atrial fibrillation and still is in widespread usetoday (Grace and Camm, 1998). Nevertheless, compared withnovel class IC antiarrhythmics, quinidine has been associatedwith an increased mortality and unfavorable side effects.Consequently, novel compounds such as class IC and classIII antiarrhythmics have been developed and replaced quin-idine in the management of atrial fibrillation. Adverse sideeffects, however, were regularly based on a compromised renaland biliary clearance of digoxin as quinidine competitivelyinhibits multidrug resistance protein 1 (Doering, 1979; Frommetal., 1999). Thus, proarrhythmic side effects of quinidinemaybecircumvented when its administration is combined with verap-amil. Only recently, quinidine has regained clinical interest as

one of the most effective agents in the pharmacological manage-ment of inherited malignant arrhythmias such as Brugadasyndrome, idiopathic ventricular fibrillation, and short-QTsyndromes (Kaufman, 2007; Viskin et al., 2007). Electrophysio-logically, quinidine acts as a broad multichannel blocker withwell-documented inhibitory effects on several potassium chan-nels, including Ito, IKur, IKATP, IKr, IKs, IKACh, and IK1 (Tamargoet al., 2004).Cardiac inwardly rectifying current IK1 is essential for

stabilizing the resting membrane potential of cardiomyocytes(Lopatin and Nichols, 2001). Several lines of evidence suggestthat at the molecular level heteromeric assembly of inwardlyrectifying potassium channels Kir2.1, Kir2.2, and Kir2.3underlie human cardiac IK1 currents (Schram et al., 2003).Kir2.1 subunits are abundantly expressed in the entiremyocardium and heteromeric Kir2.1/Kir2.2 channels underliehuman ventricular IK1 current (Schram et al., 2003). Incontrast, Kir2.3 subunits are stronger expressed in humanatria (Wang et al., 1998; Melnyk et al., 2005).Cardiac IK1 may constitute a promising target for pharma-

cological suppression of atrial and ventricular fibrillation. Forinstance, increased IK1 outward currents due to gain-of-function

Financial support for this study was granted by Deutsche Forschungsge-meinschaft (DFG, Projects ZI1177/1-1 to Dr. Zitron and SCHO1350/2-1 to Dr.Scholz) and the University of Heidelberg: Postdoc Fellowship Program of theFaculty of Medicine to Dr. Scherer.

https://doi.org/10.1124/jpet.116.238287.

ABBREVIATION: IV, current-voltage.

209

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

mutations of Kir2.1 underlie short-QT syndrome Type 3 (Prioriet al., 2005). These patients are characterized by a shortenedQTinterval and a strong predisposition for atrial or ventricularfibrillation and can effectively be treated with quinidine. In linewith this pathophysiology, it has been demonstrated thatfamilial atrial fibrillation is linked to a gain-of-function muta-tion of Kir2.1 subunits (Xia et al., 2005). Furthermore, Noujaimet al. (2007) provided direct evidence that increased IK1 currentsmight stabilize reentrant tachycardias as overexpressing Kir2.1subunits generated fast and stable rotors in a mouse model.In silico approaches identified several aromatic residues

within the cytoplasmic pore region of Kir2.1 (E224, E299,D255, D259, and F254) to serve as potential binding positionsof cationic amphipathic drugs that have been experimentallyvalidated by Ala-scanning mutagenesis (Rodriguez-Menchacaet al., 2008; Noujaim et al., 2010; Lopez-Izquierdo et al.,2011b; van der Heyden et al., 2013). Noujaim et al. (2011)demonstrated that the inhibitory effect of quinidine on Kir2.1ismediated by an acute pore block of Kir2.1 subunits involvingamino acid residue E224, F254, and D259.To date, there are no data available regarding other potential

mechanisms of action of quinidine or further effects on otherKir2.x subtypes. Therefore, we studied the effects of quinidineon all cardiac wild-type and mutant Kir2.x potassium chan-nels in the Xenopus oocyte expression system to fully elucidatethe molecular properties that may underlie the inhibition ofIK1 current by quinidine.

MethodsDNA Constructs and Heterologous Expression. Kir2.x cDNA

constructs were kindly provided by Dr. Barbara Wible (Cleveland,OH) and Dr. Carol Vandenberg (Santa Barbara, CA). ComplementarymRNA was synthesized from Kir2.x cDNA with the mMESSAGEmMACHINE in vitro transcription kit (Thermo Scientific, Dreieich,Germany) using T7 Polymerase (Kir2.1 and Kir2.2 constructs) or T3Polymerase (Kir2.3 constructs). Kir2.x mutants were generated byinverse polymerase chain reactionwith syntheticmutant oligonucleotide

primers using the QuikChange Site-Directed Mutagenesis Kit (Strata-gene, La Jolla, CA) as previously described (Karle et al., 2002). The Kir2.x-Kir2.y heteroconcatemers (Fig. 1) used in this study were kindlyprovided by Dr. Regina Preisig-Müller (Marburg, Germany). All muta-tions and DNA constructs were confirmed by sequencing (SeqLab,Goettingen, Germany).

By using a Nanoject automatic injector (Drummond, Broomall, PA),50 nl per oocyte of cRNA (50 to 500 ng/ml) was injected into stage V andVI defolliculated oocytes. Experiments were performed 1 to 3 daysafter injection. All experiments performed in this study are inaccordance with theGuide for the Care andUse of Laboratory Animalspublished by the US National Institutes of Health (NIH publicationno. 85–23, revised 1996).

Electrophysiology and Data Analysis. Currents were recordedusing the two-microelectrode voltage-clamp technique in Xenopuslaevis oocytes using a commercially available amplifier (WarnerOC-725A, Warner Instruments, Hamden, CT) and pCLAMP software(Axon Instruments, Foster City, CA) for data acquisition and analysis.Data were low-pass filtered at 1 to 2 kHz (23 dB, four-poleBesselfilter) before digitalization at 5 to 10 kHz. During the experi-ments no leak subtraction was performed. Statistical data arepresented as mean 6 S.E.M. In all figures, n represents the numberof cells recorded. Statistical significance was evaluated using Stu-dent’s t test. For more than two groups, one-way analysis of variancewas used instead. Differences were considered to be significant ifthe P value was ,0.05. For statistical analysis and figures OriginPro2015 (b9.2.214, OriginLab Corp., Northampton, MA) was used. Theconcentration-response curves were fitted with the Hill equation: I/I05 I0/(11X/IC50)

n, I/I0 being the relative current, I0 the unblockedcurrent amplitude, X the drug concentration, IC50 the concentrationfor half maximal block, and n the Hill coefficient.

Solutions and Drug Administration. Voltage clamp measure-ments of Xenopus oocytes were performed in a K1 solution containing(in mM) 5 KCl, 100 NaCl, 1.5 CaCl2, 2 MgCl2, and 10 HEPES. The pHof this solution was corrected to 7.4 with 1 M NaOH. Electrodes werefilledwith 3MKCl solution. Allmeasurementswere carried out at roomtemperature (20°C). Xenopus oocytes were incubated in the drugsolution. Recordings were made before incubation and 30 minutesafterward.Quinidine (Sigma-Aldrich, Steinheim,Germany)wasdilutedin dimethylsulfoxide to obtain a 100 mM stock solution and stored at220°C. Aliquots of the stock solutionwere dissolved in the bath solutionto the desired concentrations on the day of the experiments.

ResultsQuinidine Inhibits Kir2.3 Currents. We initially char-

acterized the effects of quinidine on homomeric Kir2.3 sub-units. Channels were heterologously expressed in Xenopuslaevis oocytes and currents were recorded using the voltage-clamp technique. To measure Kir2.x currents, a standardizedvoltage protocol was applied. From a holding potential of280 mV, 400-ms test pulses ranging from 2120 to 140 mVwere applied in 10-mV increments. Inward current ampli-tudes at 2120 mV were used to quantify inhibitory effects,unless stated otherwise. In Fig. 2, representative currenttraces under control conditions (Fig. 2A) and after 30 minutesof incubation with 1 mM quinidine (Fig. 2B) are shown. Thecorresponding current-voltage (IV) curves are shown in Fig.2C. Under control conditions we observed a slight currentrun-up by 9 6 5.9% (n 5 7, data not shown) as previouslydescribed (Karle et al., 2002). Dose-response curves wereobtained accordingly. Relative effects of quinidine at concen-trations ranging from 5 mM to 1 mM were compared inXenopus oocytes heterologously expressing Kir2.3 channelsand yielded an IC50 of ∼72 mM (n 5 4–16, Fig. 2D).

Fig. 1. Schematic illustration of Kir2.x–Kir2.y heteroconcatemers. Kir2.x-2.y heteroconcatemers were generated by concatenating the mRNA ofdifferent Kir2 subunits. C-terminal domains of Kir2.x subunits were fusedto the N terminus of Kir2.y subunits (Preisig-Muller et al., 2002).Corresponding heteroconcatemeric Kir2.x-2.y constructs were heterolo-gously expressed in the oocyte expression system. For a functional channelpore the association of two heteroconcatemers is required.

210 Koepple et al.

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

Fig. 2. Inhibition of Kir2.3 currents by quinidine. Representative Kir2.3 current traces under control conditions and after 30 minutes of incubation with1 mM quinidine are shown in (A) and (B). Corresponding current-voltage curves are displayed in (C). Dose-response experiments revealed a half-maximal inhibiting concentration of quinidine at approximately 72 mM for Kir2.3 subunits (D) (n = 4–16). The relative effects of quinidine on inwardcurrent amplitudes are plotted as a function of time in (E) and (F). Cells were treated with 100 mM quinidine (E) (n = 10) and were subsequentlyreperfused with extracellular bath solution for 45 minutes, respectively (F) (n = 10). Control experiments were performed accordingly with extracellularK+ solution (n = 4, each).

Inhibitory Effects of Quinidine on hKir2.x Channels 211

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

Onset and wash-out of the inhibitory effect of quinidine onKir2.x subunits were investigated using a repetitive voltageclamp protocol at start-to-start intervals of 30 seconds over atime period of 45 minutes. Starting from a holding potentialof 280 mV test pulses to 2120 mV and 2110 mV for 400 mswere applied to elicit large inward currents. Relative in-hibition of Kir2.3 currents during wash-in and wash-out ofquinidine at a concentration of 100 mM are shown in Fig. 2, Eand F, respectively. Onset of block was slow and approximatedsteady-state conditions after 45minutes (32.86 2.4% of initialcurrent amplitudes, n 5 10). Correspondingly, recovery frominhibition was also slow. Even after 45 minutes of reperfusionwith bath solution, the current block still persisted as seen inFig.2F (71.966.3%of initial current amplitudes,n5 10). For controlexperiments, the chamber was continuously perfused withextracellular control solution (n 5 4), which resulted in arelative current amplitude of 104.96 4.2% (n 5 4) and 105.86 3.8% (n5 4) after 45minutes and 90minutes, respectively.Inhibition of Kir2.1 and Kir2.2 Currents by Quinidine.

Next, we studied the effects of quinidine on homomeric Kir2.2and Kir2.1 currents to compare electrophysiological character-istics in different cardiac Kir2.x subunits. Concentration-response relations were obtained analogously to those ofKir2.3 as described above. Under control conditions, Kir2.1and Kir2.2 channels showed a current run-up by 2.4 6 7%(n 5 7) and 6.5 6 8.3% (n 5 5), respectively (data not shown).Representative IV relations of current recordings before andafter administration of 2mMquinidine are shown in Fig. 3A forKir2.1 subunits and in Fig. 3E for Kir2.2 subunits. Dose-response characteristics were acquired as described aboveand yielded IC50 values of approximately 290 mM for Kir2.1(n5 5–13) and 370 mM for Kir2.2 (n5 4–7) channels (Fig. 3, Band F).Contrary to Kir2.3, neither Kir2.1 nor Kir2.2 currents

reached steady-state conditions after 45 minutes. We did notlengthen incubation times, because according to our experienceXenopus oocytes do not tolerate longer experiments. Kir2.1currents andKir2.2 currentswere reduced to 46.764.1% (n5 4)and 63.9 6 6.4% (n 5 4), respectively (Fig. 3, C and G).Interestingly, quinidine-mediated current inhibition was solelyreversible in Kir2.2 subunits and reached 122.4 6 9% of initialcurrent amplitudes (n 5 4) (Fig. 3H). In contrast, inhibitoryeffects on Kir2.1 current persisted even after 45 minutes ofquinidine wash-out as shown in Fig. 3D (66.4 6 8.8%, n 5 4).Oocytes were continuously perfused with extracellular bathsolution to obtain time-matched controls. After 45 minutes and90 minutes, relative currents increased to 100.1 6 0.1% and103.3 6 0.07% or 103.0 6 1.4% and 107.8 6 0.08% (n5 4 each)for Kir2.1 or Kir2.2, respectively.Inhibition of Heteroconcatemeric Kir2.x-2.y Chan-

nels. As stated earlier, heterotetrameric assembly of Kir2.1and Kir2.2 is the main molecular correlate of human ventric-ular IK1 current, whereas Kir2.3 is more abundantly expressedin atrial myocardium (Liu et al., 2001; Gaborit et al., 2007).Coexpression of Kir2.1 and Kir2.2 subunits in Xenopus oocytesresults in distinctive biophysical current characteristics moreclosely resembling human IK1 (Schram et al., 2003). Tominimize random association artifacts of coinjected subunits,we solely injected mRNA of heteroconcatemeric Kir2.x-2.ychannels (Fig. 1) as previously described (Preisig-Mulleret al., 2002). Under control conditions, Kir2.1-2.2, Kir2.1-2.3,andKir2.2-2.3 currents increased by 13.66 5.4%, 186 7%, and

16.6 6 2.6%, respectively (n 5 4 each, data not shown).IV curves of representative current recordings before and afteradministration of quinidine are shown in Fig. 4, A, C, and E.Dose-response relationships were assessed analogously tothe experiments described above and yielded IC50 values of470, 250, and 190 mM for Kir2.1-2.2 (n 5 4–6), Kir2.1-2.3(n 5 4–8), and Kir2.2-2.3 (n 5 4–8) subunits, respectively(Fig. 4, B, D, and F).Comparison of Maximum Inhibitory Effects of Quin-

idine on Kir2.x Channels. The maximum subunit-specificand relative inhibition of Kir2.x or Kir2.x-2.y currents atsupramaximal concentrations of quinidine are listed inTable 1. In comparison, Kir2.3 currents were blocked to thegreatest extent, followed by Kir2.2 and Kir2.1 currents. Theeffects of quinidine on Kir2.x-2.y heteroconcatemers were farless pronounced. Heteroconcatemeric channels containingthe Kir2.3 subunits, however, were prone to a greater blockcompared with Kir2.1-2.2 heteroconcatemers.Inhibition of Kir2.x Currents by Quinidine is Voltage

Independent. To examinewhether quinidine inhibits Kir2.xcurrents in a voltage-dependent manner, inhibitory effects werecompared at voltages ranging from 2120 to 0 mV. Inhibitoryeffects between280and260mVwere excluded fromtheanalysisdue to small current amplitudes. We found that the inhibition ofKir2.x currents by quinidine is not significantly different at theanalyzed voltages (Fig. 5). Relative current inhibition by 1 mMquinidine ranged from 72.76 8.6% at250mV to 50.66 9.7%at 0mV inKir2.2 currents (n5 6) and 64.56 7% at250mV to45.3 6 17.3 at 220 mV in Kir2.1 currents (n 5 4). The relativeinhibitory effect of 0.1 mM quinidine on Kir2.3 channels was68.2 6 4.2% at 250 mV and 59.5 6 2.9% at 2100 mV (n 5 7),respectively.Functional Significance of the Cytoplasmic Pore

Domain for the Inhibitory Effects of Quinidine. Severalnegatively charged amino acid residues within the cytoplas-mic pore domain of Kir2.1 (E224, F254, and D259) have beenassociated with an acute block by quinidine (Noujaim et al.,2011). Furthermore, chloroquine, structurally similar toquinidine, blocks the permeation pathway by interacting elec-trostatically with residues Kir2.1 D255 and E299 (Rodriguez-Menchaca et al., 2008; Noujaim et al., 2010). We performedsequence alignments to identify corresponding residues in otherKir2.x subunits (Fig. 6). Via Ala-scanning mutagenesis weexperimentally validated their role in quinidine-mediatedKir2.x inhibition. Unfortunately, the influence of Kir2.1 D255on channel inhibition could not be verified, because heterologousexpression in Xenopus oocytes resulted in nonfunctional cur-rents. Summary data of inhibitory effects of 1, 0.5, and 0.1 mMquinidine on Kir2.1, Kir2.2, and Kir2.3 pore mutants comparedwith corresponding effects in wild-type channels (Kir2.1 wildtype: 49 6 2.1%, n 5 13; Kir2.2 wild type: 51.6 6 3.3%, n 5 7;Kir2.3 wild type: 58.6 6 2%, n 5 15) are shown in Fig. 7. Theinhibitory effect of quinidine was significantly reduced in allAla-mutated Kir2.1 and Kir2.3 channels (Fig. 7, A and C).Relative current was inhibited by 22.3 6 2.2% in Kir2.1E224A (n 5 5) and by 5.1 6 4.8% in Kir2.1 D259A (n 5 4).Furthermore, in Kir2.1 E299A and in Kir2.1 F254A, currentincreases by 8.26 4.5% (n5 6) and by 3.86 3.3% (n5 5) wereobserved, respectively. Also, all studied Ala-mutated Kir2.3subunits exhibited significantly reduced effects of quinidine,resulting in relative current reductions by 33.9 6 1.4% (n 5 6)inKir2.3 E216Aand 45.162.7% (n5 6) inKir2.3D247A and in a

212 Koepple et al.

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

current increase of 0.07 6 2.6% (n 5 4) in Kir2.3 D251A and2.3 6 2.9% (n 5 6) in Kir2.3 E291A channels. Interestingly,only mutation D260A altered the effect on Kir2.2 channelswith a current increase of 11.56 5.6% (n5 5) after quinidineapplication (Fig. 7B).

PIP2-Mediated Inhibition of Kir2.x Subunits byQuinidine. Glycerophospholipid PIP2 is essential formembrane-associated signaling. As an important cofactor forchannel function and lipid rafts (Suh and Hille, 2008), ithas previously been recognized to participate in high-affinity

Fig. 3. Inhibition of Kir2.1 and Kir2.2currents by quinidine. Typical current-voltage relations of Kir2.1 and Kir2.2current traces before and after a30-minute incubation period with 2 mMquinidine are shown in (A) and (E),respectively. Kir2.1 and Kir2.2 dose-response experiments yielded IC50 valuesof approximately 290 mM (B) (n = 5–13)and 370 mM (F) (n = 4–7). Summary dataof wash-in/wash-out experiments with1 mM quinidine are shown in (C) and(D) and (G) and (H) for Kir2.1 and Kir2.2currents, respectively. Extracellular bathsolution was used as control (n = 4, each).

Inhibitory Effects of Quinidine on hKir2.x Channels 213

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

interactions with inwardly rectifying ion channels and partic-ularly with Kir2.1 subunits (Rohacs et al., 2003; Logothetiset al., 2007). Various Kir2.x inhibitors possess cationic amphi-philic properties, thereby interfering with membranous PIP2

distributions and effectively abrogating channel stabilization(Ponce-Balbuena et al., 2009; Lopez-Izquierdo et al., 2011a,b).

As a cationic amphiphilic drug we postulated that quinidineinterferes with PIP2-mediated channel activation. Thus, wegenerated channel mutants with low (Kir2.1 K182Q, andL222I) and high (Kir2.3 I214L) endogenous PIP2 affinities(Xie et al., 2008) and compared the extent of quinidine-mediated blocks as shown in Fig. 8. Mutation K182Q as well

Fig. 4. Inhibitory effects of quinidine on heteroconcatemeric Kir2.x-2.y channels. Representative current-voltage curves of the quinidine block onheteroconcatemeric Kir2.x-2.y channels are shown in (A), (C), and (E). Dose-response analysis revealed a half-maximal inhibitory concentration ofapproximately 470 mM for Kir2.1-2.2 (B) (n = 4–6), 250 mM for Kir2.1-2.3 (D) (n = 4–8), and 190 mM for Kir2.2-2.3 heteroconcatemers (F) (n = 4–8).

214 Koepple et al.

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

as L222I of Kir2.1 increased its sensitivity toward 200 mMquinidine because thewild-type current reduction of 18.263.8%(n5 5) is further increased to 68.96 3.3% (n5 6) and 316 1.1%(n5 5), respectively. In contrast, the inhibitory effect of 100 mMquinidine inKir2.3wild-type channels (58.661.9%,n5 16) wascompletely abolished in Kir2.3 I214L channels with a totalcurrent increase of 9.2 6 3.6% (n 5 7).

DiscussionInhibitory effects of quinidine on cardiac IK1 current arewell

known and may contribute to its antifibrillatory efficacy(Salata and Wasserstrom, 1988). However, apart from pre-vious investigations focusing on Kir2.1 subunits, no fur-ther analyses of the drug’s inhibitory effect on other Kir2.xchannels were performed before this study. Here, we showthat quinidine differentially inhibits Kir2.x channels with thehighest potency toward Kir2.3 subunits, being approximatelyfivefold more sensitive than homomeric Kir2.1 or Kir2.2channels. In addition, quinidine inhibits Kir2.3 currents by a

three- to fourfold greater degree compared with homomericKir2.1 or Kir2.2 currents. In line with these results, hetero-concatemeric Kir2.1-2.3 and Kir2.2-2.3 currents exhibit anapproximately twofold higher sensitivity toward quinidinecompared with Kir2.1-2.2 heteroconcatemers. Biophysical cur-rent characteristics, however, are not altered by quinidine.Onsetof inhibition is slow and effects are solely reversible in Kir2.2currents. Furthermore, quinidine inhibits Kir2.x currents in avoltage-independentmanner. By application of channelmutantswith altered PIP2 affinities and Ala-scanning mutagenesis ofamino acidswithin the channel permeation pathway, we provideevidence that the inhibitory effect of quinidine is dependent on adual mode of inhibitory action: centrally blocking the channelpermeation pathway as well as interfering with PIP2-mediatedchannel stabilization.Pharmacological IK1 inhibition might constitute a novel

approach for antiarrhythmic treatment strategies (Rees andCurtis, 1993). Patients suffering from short-QT syndromeType 3 and familial atrial fibrillation might benefit from aselective IK1 or even Kir2.x-specific inhibition. However, thelack of selective Kir2.x inhibitors simultaneously suited forin vivo application renders the validation of potential benefitsdifficult. Interestingly, with b3-adrenoceptor antagonist SR59230A, our group could recently identify a selective Kir2.xinhibitor thatmay serve as a potentialmodel substance for thedevelopment of IK1 modulators (Kulzer et al., 2012).Chronic atrial fibrillation, chronic atrioventricular block,

and ischemic heart disease are linked to complex regionalchanges in IK1/Kir2.x expression profiles (Mustroph et al.,2014). It has been shown that the increased expression ofKir2.1 contributes to reentrant tachycardias by stabilizinghigh-frequency atrial and ventricular rotors (Noujaim et al.,2007; Girmatsion et al., 2009; Sekar et al., 2009). Interest-ingly, the inhibition of IK1 might constitute an effectivetherapeutic approach to the treatment of reentrant arrhyth-mias (Warren et al., 2003; Pandit et al., 2005; Kharche et al.,2014).However, IK1 dysfunction as a result of chronic heart failure

or long-QT syndrome has been recognized to destabilize theresting membrane potential of cardiomyocytes and increasetheir susceptibility for tachyarrhythmias. Furthermore, ani-mal studies provide evidence that pulmonary vein cardiomyo-cytes exhibit reduced IK1 and Kir2.3 current densities (Ehrlichet al., 2003; Melnyk et al., 2005), which may at least partlyexplain their increased arrhythmogenic potential (Cherryet al., 2007).Quinidine was the first antiarrhythmic drug used clinically

to terminate atrial fibrillation. As a multichannel blocker,quinidine may display a favorable electrophysiological profileby normalizing the effective refractory period while simulta-neously preventing early afterdepolarizations by reducingCa21 and Na1 inward currents. Because distinctive intrinsicaffinities toward different ion channels, quinidine primarilyprolongs the cardiac action potential at low concentrations byinhibiting IKr. As the concentration further increases, addi-tional effects on other cardiac potassium channels and sodiumchannels are observed that consequently shorten the actionpotential duration (Paul et al., 2002; Tamargo et al., 2004; Wuet al., 2008).Blocking mechanisms of several cationic amphiphilic drugs

such as chloroquine have been deciphered by means ofmolecular modeling and Ala-scanning mutagenesis. Several

TABLE 1.Maximum inhibitory effect of quinidine on homomeric Kir2.x and hetero-concatemeric Kir2.x-2.y channelsData given in mean 6 S.E.M. for maximum relative inhibitions of Kir currents after30 minutes of incubation with quinidine (see text for details).

Channel Concentration Relative Current to Control S.E.M. n

Kir2.1 2 mM 42% 3.9% 5Kir2.2 2 mM 29% 5.2% 4Kir2.3 1 mM 10% 1.4% 4Kir2.1-2.2 2.5 mM 34% 3% 5Kir2.1-2.3 2.5 mM 22% 1.8% 5Kir2.2-2.3 2 mM 31% 4.6% 4

Fig. 5. Voltage independent effect of quinidine on Kir2.x channels.Current measurements before and after 30 minutes of incubation withquinidine were compared to determine voltage dependency of thequinidine block. Results are shown for the voltages between 2120 mVand 0 mV. Currents of 806 10 mV were excluded from the analysis due tosmall and unrepresentative amplitudes. Relative current block of Kir2.xchannels did not show significant differences in the analyzed voltagerange (n = 4–7).

Inhibitory Effects of Quinidine on hKir2.x Channels 215

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

amino acid residues within the central pore of Kir2.1 channelshave been identified as critical binding sites for thesecompounds (Rodriguez-Menchaca et al., 2008). Kir2.1 E224,F254, D259, have been verified experimentally to be crucial forthe inhibitory effect of quinidine (Noujaim et al., 2011). Herewe demonstrate that, analogously to chloroquine, amino acidresidue E299 is involved in the quinidine-mediated inhibitionof Kir2.1 channels. Additionally, the effect of quinidine onKir2.2 and Kir2.3 channels was evaluated. Increased IK1

density in patients with chronic atrial fibrillation (Boschet al., 1999; Workman et al., 2001; Zhang et al., 2005) mayrender atrial Kir2.3 subunits as promising novel targets forantiarrhythmic therapies. We identified homologous residuesin Kir2.2 and Kir2.3 subunits and consequently assessed theirrole in quinidine-mediated current inhibition. Inhibitoryeffects of quinidine were altered in all Kir2.3 mutants. Thesedata resemble the situation in Kir2.1 channels andmay eitherbe based on a decentralization of quinidine within thepermeation pathway or a sterically intact blocking positionbut weakened interactions with the remaining residues. Re-gardless of the molecular basis, we here demonstrate thatKir2.3 E216, D247, D251, and E291 are essential aminoresidues for quinidine-mediated block. Interestingly Kir2.2subunits fundamentally differ from this pattern. Solely Kir2.2D260A was found to abrogate quinidine block and maydirectly interact with quinidine. Further investigations withmultiple substituted residues are needed to clarify whetherthe unchanged residues in the remaining Kir2.2 mutants oryet unidentified residues are involved in their unalteredinhibition.Onset of current inhibition was slow. Solely the effect on

Kir2.3 currents reached steady-state conditions after 45 min-utes. Interestingly, upon wash-out inhibition of Kir2.2 currentswas fully reversed and it is intriguing to speculate that thisobservation might be due to fewer binding sites of quinidineand/or weakened interaction of the drug within the channelpore. Nevertheless, the significance of homomer-specific inhib-iting kinetics in vivo is questionable considering the lowmyocardial expression of Kir2.2 subunits on the one hand andthe heteromeric nature of human IK1 on the other (Wang et al.,1998; Schram et al., 2002; Melnyk et al., 2005). In line withthese considerations, Salata andWasserstrom (1988) were ableto demonstrate that quinidine inhibits canine IK1 currentsirreversibly for a period of 45 minutes.Noujaim et al. (2011) reported a diminished current in-

hibition of outward currents after heterologous expression ofKir2.1 channels in HEK293 cells. This observation not onlycontradicts our presented detailed analysis of Kir2.x channelsbut also earlier reports from animal models suggesting a

voltage-independent IK1 block (Salata and Wasserstrom,1988). We here provide evidence that endogenous polyaminesand quinidine mediate Kir2.x current blocks by interactingwith the same residues that line the functional channelpore (Baronas and Kurata, 2014). Contrary to polyamines,quinidine, however, inhibits Kir2.x subunits in a voltage-independent manner. This discrepancy may suggest anadditional mode of action by which quinidine can facilitatecurrent inhibition. Considering the subunit-specific sensitiv-ities and kinetics, we hypothesized a combined mechanism forKir2.x inhibition by quinidine involving direct pore block ofthe permeation pathway and additional inhibition by inter-ference with PIP2-mediated channel stabilization. To validatethis hypothesis, we generated previously described Kir2.xmutants with altered PIP2 affinities (Xie et al., 2008). Wenoticed that high PIP2 affinities (Kir2.3 I214L) correlate withsmaller inhibitory effects. Inversely, channels interactingweakly with PIP2 (Kir2.1 K182Q, and L221I) are prone to astrong inhibitory effect. These data suggest that quinidineinhibits Kir2.x channels by a combined mechanism involv-ing direct pore block and impaired open channel gating byPIP2-interference. Although concordant changes in Kir2.1K182Q/L221I and Kir2.3 I214L currents might suggest thatthe intrinsic affinity in these channels toward quinidine isunaltered, we acknowledge the fact that these data are obtainedat single concentrations. Lopez-Izquierdo et al. (2011b), noted arightward shift inmefloquine inhibition in corresponding Kir2.3I213L channels, suggesting a decreased IC50 value. The corre-sponding dose-response curve shows simultaneously a change ofslope that might be an indicator of altered mefloquine bindingcharacteristics. A similar alteration could complicate the in-terpretation of our presented results. Further experiments atmultiple quinidine concentrations are needed to fully excludethe possibility of an altered Hill coefficient in mutant Kir2.xchannels.A similar blocking mechanism was previously proposed for

the antimalarial agent quinacrine (Lopez-Izquierdo et al.,2011a). However, to the best of our knowledge, quinidine is thefirst classic IK1-inhibiting antiarrhythmic compound in whichthis complex mode of action may be an essential part of itsproven antifibrillatory efficacy.Xenopus oocytes are ideally suited for longer current

recordings as performed in this study. Because of their lowleakage of currents as well as their vigorous nature, they offera couple of advantages over mammalian cells, rendering themuseful to study pharmacological properties of ion channels.However, anatomic characteristics such as yolk sacks andvitellinemembranesmay reduce effective drug concentrationswithin the cell. Hence, IC50 values may differ up to 30-fold

Fig. 6. Alignment of amino acid sequences of human Kir2.x subunits. Specific amino acid residues that line the cytoplasmic pore of Kir2.1 subunitsinteract with cationic amphiphilic drugs. Residues reported to be involved in quinidine- and chloroquine-mediated Kir2.1 current inhibition are markedby asterisks (Kir2.1 E224, F254, and D259) and obelisks (D255 and E299), respectively. Sequence alignments indicate that they are highly conservedwithin the Kir2.x subfamily.

216 Koepple et al.

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

compared with native cardiomyocytes or transfected cell lines(Madeja et al., 1997). As recordings of native IK1 in cardio-myocytes require pharmacological inhibition of other ion

channels, nonspecific interactions with intrinsic post-translational pathways cannot be excluded and are ageneral concern when interpreting native electrophysio-logical data. However, it has been demonstrated thatquinidine blocks human atrial IK1 currents in a voltage-dependent manner with IC50 ranging from 4.1 to 42.6 mM(Nenov et al., 1998). Noujaim et al. (2011) reported thatquinidine inhibits Kir2.1 in mammalian cells in a dose-dependent manner with an IC50 of 57 mM, being approxi-mately sixfold lower than our result from Xenopus oocytes.In humans, antiarrhythmic plasma concentrations ofquinidine fall within the range of 2–5 mg/ml correspondingto 6–15 mM (Fremstad et al., 1979; Yang et al., 2009) andmay even be higher in patients with an individual suscep-tibility based on renal or hepatic dysfunction, drug inter-actions, sex, age, and pharmacogenomics. Thus, on thedata presented here, we hypothesize that an antifibrilla-tory IK1 block within therapeutic plasma concentrationsmay primarily be based on the compound’s predominantinfluence on atrial Kir2.3 subunits. In conjunction withother highly sensitive atrial-specific potassium currentslike Ito, IKur, and IKACh (Tamargo et al., 2004), quinidinemay exhibit a favorable pharmacological profile for thesuppression of supraventricular tachyarrhythmias mainlyby prolonging refractory periods in human atria.In the present study we characterized themolecular basis of

quinidine’s inhibitory effects on human Kir2.x channels. Wefound that quinidine exerts differential effects on Kir2.xchannels with the highest affinity toward Kir2.3 subunits.Inhibition of Kir2.x channels is mediated by joint modes ofaction involving direct cytoplasmic pore block and impairedchannel gating by interference with PIP2 stabilization. Thesedata add to the understanding of the antifibrillatory efficacy ofquinidine and may contribute to the development of newantiarrhythmic IK1 antagonists.

Fig. 7. Quinidine exerts differential effects on Kir2.x cytoplasmic poremutants. Inhibitory effects of quinidine were significantly reduced in allKir2.1 (A) and Kir2.3 pore mutants (C). Interestingly, Kir2.2 channelsdiffered from these patterns. As shown in panel B only in Kir2.2 D260Amutants the inhibitory effect was significantly altered compared withwild-type channels with a current increase of 11.5 6 5.6% (n = 5).Quinidine concentration used for Kir2.1, Kir2.2 and Kir2.3 pore mutantswere 1 mM, 0.5 mM and 0.1 mM, respectively.

Fig. 8. Quinidine blocks Kir2.x currents in a PIP2-dependent manner.Channel mutants that have less affinity to endogenous PIP2 weresignificantly more sensitive to the inhibitory effect of 200 mM quinidine(68.96 3.3%, n = 6 for Kir2.1 K182Q and 316 1.1%, n = 5 for Kir2.1 L221I)compared with Kir2.1 wild-type channels (18.2 6 3.8%, n = 5). Kir2.3I214L enhances the affinity to PIP2, and the effects of quinidine on thesechannels were similar to those under control conditions.

Inhibitory Effects of Quinidine on hKir2.x Channels 217

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from

Acknowledgments

The authors thank Christine Jeckel for excellent technicalassistance.

Authorship Contributions

Participated in research design: Koepple, Scherer, Seyler, Scholz,Thomas, Katus, and Zitron.

Conducted experiments: Koepple.Performed data analysis: Koepple and Seyler.Wrote or contributed to the writing of the manuscript: Koepple and

Zitron.

References

Baronas VA and Kurata HT (2014) Inward rectifiers and their regulation by en-dogenous polyamines. Front Physiol 5:325.

Bosch RF, Zeng X, Grammer JB, Popovic K, Mewis C, and Kühlkamp V (1999) Ionicmechanisms of electrical remodeling in human atrial fibrillation. Cardiovasc Res44:121–131.

Cherry EM, Ehrlich JR, Nattel S, and Fenton FH (2007) Pulmonary veinreentry–properties and size matter: insights from a computational analysis. HeartRhythm 4:1553–1562.

Doering W (1979) Quinidine-digoxin interaction: Pharmacokinetics, underlyingmechanism and clinical implications. N Engl J Med 301:400–404.

Ehrlich JR, Cha TJ, Zhang L, Chartier D, Melnyk P, Hohnloser SH, and Nattel S(2003) Cellular electrophysiology of canine pulmonary vein cardiomyocytes: actionpotential and ionic current properties. J Physiol 551:801–813.

Fremstad D, Nilsen OG, Storstein L, Amlie J, and Jacobsen S (1979) Pharmacoki-netics of quinidine related to plasma protein binding in man. Eur J Clin Pharmacol15:187–192.

Fromm MF, Kim RB, Stein CM, Wilkinson GR, and Roden DM (1999) Inhibition ofP-glycoprotein-mediated drug transport: A unifying mechanism to explain the in-teraction between digoxin and quinidine [seecomments]. seecomments Circulation99:552–557.

Gaborit N, Le Bouter S, Szuts V, Varro A, Escande D, Nattel S, and Demolombe S(2007) Regional and tissue specific transcript signatures of ion channel genes in thenon-diseased human heart. J Physiol 582:675–693.

Girmatsion Z, Biliczki P, Bonauer A, Wimmer-Greinecker G, Scherer M, Moritz A,Bukowska A, Goette A, Nattel S, Hohnloser SH, et al. (2009) Changes inmicroRNA-1 expression and IK1 up-regulation in human atrial fibrillation. HeartRhythm 6:1802–1809.

Grace AA and Camm AJ (1998) Quinidine. N Engl J Med 338:35–45.Karle CA, Zitron E, Zhang W, Wendt-Nordahl G, Kathöfer S, Thomas D, Gut B,Scholz E, Vahl CF, Katus HA, et al. (2002) Human cardiac inwardly-rectifying K1channel Kir(2.1b) is inhibited by direct protein kinase C-dependent regulation inhuman isolated cardiomyocytes and in an expression system. Circulation 106:1493–1499.

Kaufman ES (2007) Quinidine in short QT syndrome: an old drug for a new disease.J Cardiovasc Electrophysiol 18:665–666.

Kharche SR, Stary T, Colman MA, Biktasheva IV, Workman AJ, Rankin AC, HoldenAV, and Zhang H (2014) Effects of human atrial ionic remodelling by beta-blockertherapy on mechanisms of atrial fibrillation: a computer simulation. Europace16:1524–1533.

Kulzer M, Seyler C, Welke F, Scherer D, Xynogalos P, Scholz EP, Thomas D, BeckerR, Karle CA, Katus HA, et al. (2012) Inhibition of cardiac Kir2.1-2.3 channels bybeta3 adrenoreceptor antagonist SR 59230A. Biochem Biophys Res Commun 424:315–320.

Liu GX, Derst C, Schlichthörl G, Heinen S, Seebohm G, Brüggemann A, Kummer W,Veh RW, Daut J, and Preisig-Müller R (2001) Comparison of cloned Kir2 channelswith native inward rectifier K1 channels from guinea-pig cardiomyocytes.J Physiol 532:115–126.

Logothetis DE, Jin T, Lupyan D, and Rosenhouse-Dantsker A (2007) Phosphoinositide-mediated gating of inwardly rectifying K(1) channels. Pflugers Arch 455:83–95.

Lopatin AN and Nichols CG (2001) Inward rectifiers in the heart: an update on I(K1).J Mol Cell Cardiol 33:625–638.

López-Izquierdo A, Aréchiga-Figueroa IA, Moreno-Galindo EG, Ponce-BalbuenaD, Rodríguez-Martínez M, Ferrer-Villada T, Rodríguez-Menchaca AA, van derHeyden MA, and Sánchez-Chapula JA (2011a) Mechanisms for Kir channelinhibition by quinacrine: acute pore block of Kir2.x channels and interferencein PIP2 interaction with Kir2.x and Kir6.2 channels. Pflugers Arch 462:505–517.

López-Izquierdo A, Ponce-Balbuena D, Moreno-Galindo EG, Aréchiga-Figueroa IA,Rodríguez-Martínez M, Ferrer T, Rodríguez-Menchaca AA, and Sánchez-ChapulaJA (2011b) The antimalarial drug mefloquine inhibits cardiac inward rectifier K1channels: evidence for interference in PIP2-channel interaction. J CardiovascPharmacol 57:407–415.

Madeja M, Musshoff U, and Speckmann EJ (1997) Follicular tissues reduce drugeffects on ion channels in oocytes of Xenopus laevis. Eur J Neurosci 9:599–604.

Melnyk P, Ehrlich JR, Pourrier M, Villeneuve L, Cha TJ, and Nattel S (2005)Comparison of ion channel distribution and expression in cardiomyocytes of caninepulmonary veins versus left atrium. Cardiovasc Res 65:104–116.

Mustroph J, Maier LS, and Wagner S (2014) CaMKII regulation of cardiac K chan-nels. Front Pharmacol 5:20.

Nenov NI, Crumb WJ, Jr, Pigott JD, Harrison LH, Jr, and Clarkson CW (1998)Quinidine interactions with human atrial potassium channels: developmental as-pects. Circ Res 83:1224–1231.

Noujaim SF, Pandit SV, Berenfeld O, Vikstrom K, Cerrone M, Mironov S, ZugermayrM, Lopatin AN, and Jalife J (2007) Up-regulation of the inward rectifier K1 cur-rent (I K1) in the mouse heart accelerates and stabilizes rotors. J Physiol 578:315–326.

Noujaim SF, Stuckey JA, Ponce-Balbuena D, Ferrer-Villada T, López-Izquierdo A,Pandit S, Calvo CJ, Grzeda KR, Berenfeld O, Chapula JA, et al. (2010) Specificresidues of the cytoplasmic domains of cardiac inward rectifier potassium channelsare effective antifibrillatory targets. FASEB J 24:4302–4312.

Noujaim SF, Stuckey JA, Ponce-Balbuena D, Ferrer-Villada T, López-Izquierdo A,Pandit SV, Sánchez-Chapula JA, and Jalife J (2011) Structural bases for the dif-ferent anti-fibrillatory effects of chloroquine and quinidine. Cardiovasc Res 89:862–869.

Pandit SV, Berenfeld O, Anumonwo JM, Zaritski RM, Kneller J, Nattel S, and JalifeJ (2005) Ionic determinants of functional reentry in a 2-D model of human atrialcells during simulated chronic atrial fibrillation. Biophys J 88:3806–3821.

Paul AA, Witchel HJ, and Hancox JC (2002) Inhibition of the current of heterolo-gously expressed HERG potassium channels by flecainide and comparison withquinidine, propafenone and lignocaine. Br J Pharmacol 136:717–729.

Ponce-Balbuena D, López-Izquierdo A, Ferrer T, Rodríguez-Menchaca AA, Aréchiga-Figueroa IA, and Sánchez-Chapula JA (2009) Tamoxifen inhibits inward rectifierK1 2.x family of inward rectifier channels by interfering with phosphatidylinositol4,5-bisphosphate-channel interactions. J Pharmacol Exp Ther 331:563–573.

Preisig-Müller R, Schlichthörl G, Goerge T, Heinen S, Brüggemann A, Rajan S, DerstC, Veh RW, and Daut J (2002) Heteromerization of Kir2.x potassium channelscontributes to the phenotype of Andersen’s syndrome. Proc Natl Acad Sci USA 99:7774–7779.

Priori SG, Pandit SV, Rivolta I, Berenfeld O, Ronchetti E, Dhamoon A, Napolitano C,Anumonwo J, di Barletta MR, Gudapakkam S, et al. (2005) A novel form of shortQT syndrome (SQT3) is caused by a mutation in the KCNJ2 gene. Circ Res 96:800–807.

Rees SA and Curtis MJ (1993) Specific IK1 blockade: a new antiarrhythmic mecha-nism? Effect of RP58866 on ventricular arrhythmias in rat, rabbit, and primate.Circulation 87:1979–1989.

Rodríguez-Menchaca AA, Navarro-Polanco RA, Ferrer-Villada T, Rupp J, Sachse FB,Tristani-Firouzi M, and Sánchez-Chapula JA (2008) The molecular basis of chlo-roquine block of the inward rectifier Kir2.1 channel. Proc Natl Acad Sci USA 105:1364–1368.

Rohács T, Lopes CM, Jin T, Ramdya PP, Molnár Z, and Logothetis DE (2003)Specificity of activation by phosphoinositides determines lipid regulation of Kirchannels. Proc Natl Acad Sci USA 100:745–750.

Salata JJ and Wasserstrom JA (1988) Effects of quinidine on action potentials andionic currents in isolated canine ventricular myocytes. Circ Res 62:324–337.

Schram G, Pourrier M, Melnyk P, and Nattel S (2002) Differential distribution ofcardiac ion channel expression as a basis for regional specialization in electricalfunction. Circ Res 90:939–950.

Schram G, Pourrier M, Wang Z, White M, and Nattel S (2003) Barium block of Kir2and human cardiac inward rectifier currents: evidence for subunit-heteromericcontribution to native currents. Cardiovasc Res 59:328–338.

Sekar RB, Kizana E, Cho HC, Molitoris JM, Hesketh GG, Eaton BP, Marbán E,and Tung L (2009) IK1 heterogeneity affects genesis and stability of spiral waves incardiac myocyte monolayers. Circ Res 104:355–364.

Suh BC and Hille B (2008) PIP2 is a necessary cofactor for ion channel function: howand why? Annu Rev Biophys 37:175–195.

Tamargo J, Caballero R, Gómez R, Valenzuela C, and Delpón E (2004) Pharmacologyof cardiac potassium channels. Cardiovasc Res 62:9–33.

van der Heyden MA, Stary-Weinzinger A, and Sanchez-Chapula JA (2013) Inhibitionof cardiac inward rectifier currents by cationic amphiphilic drugs. Curr Mol Med13:1284–1298.

Viskin S, Antzelevitch C, Marquez MF, and Belhassen B (2007) Quinidine: a valuablemedication joins the list of ’endangered species’. Europace 9:1105–1106.

Wang Z, Yue L, White M, Pelletier G, and Nattel S (1998) Differential distribution ofinward rectifier potassium channel transcripts in human atrium versus ventricle.Circulation 98:2422–2428.

Warren M, Guha PK, Berenfeld O, Zaitsev A, Anumonwo JM, Dhamoon AS, Bagwe S,Taffet SM, and Jalife J (2003) Blockade of the inward rectifying potassium currentterminates ventricular fibrillation in the guinea pig heart. J Cardiovasc Electro-physiol 14:621–631.

Workman AJ, Kane KA, and Rankin AC (2001) The contribution of ionic currents tochanges in refractoriness of human atrial myocytes associated with chronic atrialfibrillation. Cardiovasc Res 52:226–235.

Wu L, Guo D, Li H, Hackett J, Yan GX, Jiao Z, Antzelevitch C, Shryock JC,and Belardinelli L (2008) Role of late sodium current in modulating the proar-rhythmic and antiarrhythmic effects of quinidine. Heart Rhythm 5:1726–1734.

Xia M, Jin Q, Bendahhou S, He Y, Larroque MM, Chen Y, Zhou Q, Yang Y, Liu Y, LiuB, et al. (2005) A Kir2.1 gain-of-function mutation underlies familial atrial fibril-lation. Biochem Biophys Res Commun 332:1012–1019.

Xie L-H, John SA, Ribalet B, and Weiss JN (2008) Phosphatidylinositol-4,5-bisphosphate (PIP2) regulation of strong inward rectifier Kir2.1 channels:multilevel positive cooperativity. J Physiol 586:1833–1848.

Yang F, Hanon S, Lam P, and Schweitzer P (2009) Quinidine revisited. Am J Med122:317–321.

Zhang H, Garratt CJ, Zhu J, and Holden AV (2005) Role of up-regulation of IK1 inaction potential shortening associated with atrial fibrillation in humans. Cardio-vasc Res 66:493–502.

Address correspondence to: C. Koepple, Department of Cardiology, MedicalUniversity Hospital Heidelberg, Im Neuenheimer Feld 410, D-69120 Heidel-berg, Germany. E-mail: christoph.koepple@gmail.com

218 Koepple et al.

at ASPE

T Journals on January 28, 2020

jpet.aspetjournals.orgD

ownloaded from