Biochimica et Biophysica Acta 1808 (2011) 1993–1999

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r.com/ locate /bbamem

Genistein and tyrphostin AG556 inhibit inwardly-rectifying Kir2.1 channelsexpressed in HEK 293 cells via protein tyrosine kinase inhibition

De-Yong Zhang a, Wei Wu a, Xiu-Ling Deng b, Chu-Pak Lau a, Gui-Rong Li a,c,⁎a Department of Medicine, LKS Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong, Chinab Department of Physiology and Pathophysiology, Xi'an Jiaotong University School of Medicine, 76 West Yanta Road, Xi'an, Shaanxi, Chinac Department of Physiology, LKS Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong, China

⁎ Corresponding author at: L4-59, Laboratory Block,Kong, 21 Sassoon Road, Pokfulam, Hong Kong, China. Tel2855 9730.

E-mail address: grli@hkucc.hku.hk (G.-R. Li).

0005-2736/$ – see front matter © 2011 Elsevier B.V. Aldoi:10.1016/j.bbamem.2011.04.015

a b s t r a c t

a r t i c l e i n f o

Article history:Received 5 December 2010Received in revised form 13 April 2011Accepted 29 April 2011Available online 6 May 2011

Keywords:Kir2.1Inward rectifier K+ channelProtein tyrosine kinaseEpidermal growth factor receptor kinase

Previous studies reported the controversial effects that protein tyrosine kinase (PTK) inhibition could inducean up-regulation or down-regulation of Kir2.1 current. The present study investigates how the recombinanthuman Kir2.1 channels are regulated by PTKs using whole-cell patch voltage-clamp, immunoprecipitationandWestern blot, andmutagenesis approaches.We found that hKir2.1 current was reversibly inhibited by thebroad spectrum PTK inhibitor genistein and the highly selective EGFR (epidermal growth factor receptor)kinase inhibitor AG556 in a concentration-dependent manner. The inhibition of hKir2.1 channels by genisteinor AG556 was countered by the protein tyrosine phosphatase (PTP) inhibitor orthovanadate. Immunopre-cipitation and Western blot analysis revealed that tyrosine phosphorylation level of Kir2.1 channels wasreduced by genistein or AG556, and the reduction was significantly antagonized by orthovanadate. Themutation of Y242 dramatically reduced the inhibitory response to AG556. The results obtained in this studydemonstrate that hKir2.1 channels are down-regulated by PTK inhibition, suggesting that EGFR kinaseparticipates in the modulation of human cardiac excitability.

FMB, The University of Hong.: +852 2819 9513; fax: +852

l rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

The cardiac inward rectifier K+ current (IK1), present in bothventricular and atrial myocytes, has been suggested to play amajor rolein repolarization of the action potential and stabilization of the restingmembrane potential [1]. Cardiac IK1 channels are the heterotetramericassembly of different Kir2 subfamily members, including Kir2.1, Kir2.2and Kir2.3 [1,2]. Kir2.1 (KCNJ2 or IRK1) is the major candidate geneunderlying cardiac IK1 [1,3] since most of the properties of Kir2.1channels expressed in different types of cells are similar to those ofnative cardiac IK1 [2]. Dysfunction of Kir2.1 channel [4–6], includingdisruption of the channel's binding site for the membrane-associatedregulatory phospholipid phosphatidylinositol-4,5-bisphosphate (PIP2)[7–10] and impaired intracellular trafficking of Kir2.1 subunits [11] orKir2.1 mutations [4,12] have been implicated in the development oflong QT syndrome, which is characterized by abnormalities ofrepolarization that is predisposed to cardiac arrhythmias and anincreased risk of sudden death [4–6].

Protein tyrosine kinases (PTKs), including receptor PTKs (e.g. EGFRkinase, epidermal growth factor receptor kinase) and nonreceptor PTKs(e.g. Src-related kinases) are responsible for transducing key extracel-

lular signals that mediate events such as cell growth, differentiation,embryonic development, metabolism, and immune system function[13,14]. PTKs regulate different types of ion channels [14], including L-type Ca2+ current (ICa.L) [15], volume-sensitive Cl− current (ICl.vol) [16],voltage-gated K+ channels, e.g. cardiac delayed rectifier currents (IKsand IhERG) [17–20], Kv2.1 current [21], Kv4.3 current [22] as well as thevoltage-gated Na+ current (INa) [23,24]. The phosphorylation of ionchannels at tyrosine residues has been shown to up-regulate voltage-gated K+ currents [17–19,21,22,25] and INa [23,24]. However, thecontroversial results were obtained regarding the PTK regulation ofKir2.1 and/or native cardiac IK1. Wischmeyer and colleagues firstreported that native and cloned Kir2.1 channels expressed in tsA-201cells were suppressed by direct tyrosine kinase phosphorylation basedon the evidence that protein tyrosine phosphatase (PTP) inhibitorperorthovanadate suppressed the current, and the effect can becountered by the PTK inhibitor genistein [26]. Similar results wereobserved inhumanbladderurothelial cells [27] and inhumanmyoblasts[28], inwhichgenistein increasedKir2.1 current. However, in rat cardiacventricular myocytes we found that IK1 was suppressed by genistein[25]. The inhibitory effect of Kir2.1 and Kir2.3 channels by genisteinwasalso observed by Zhao and colleagues in HEK 293 cells expressing Kirchannel genes [29]. Therefore, it is unclear whether Kir2.1 channels aredown-regulated or up-regulated by tyrosine phosphorylation. Thepresent study was designed to determine this regard in HEK 293 cellsstably expressing Kir2.1 gene using whole-cell patch clamp, immuno-precipitation, Western blot, and mutagenesis approaches. We found

1994 D.-Y. Zhang et al. / Biochimica et Biophysica Acta 1808 (2011) 1993–1999

that genistein and the EGFR kinase inhibitor AG556, but not the Src-family inhibitor PP2, inhibited Kir2.1 channels via reducing tyrosinephosphorylation of the channel protein.

2. Materials and methods

2.1. Cell culture, mutagenesis and gene transfection

The established HEK 293 cell line [30] stably expressing humanKir2.1 cDNA (generously provided by Dr. Carol A. Vandenberg) wascultured in Dulbecco's modified eagle's medium (DMEM, Invitrogen,Hong Kong) supplemented with 10% fetal bovine serum and 400 μg/mlG418 (Invitrogen). The Kir2.1 channel mutants, Y242A, Y242F, Y336Aand Y366A generated using the QuickChange Site-directedMutagenesiskit (Stratagene), were transiently expressed inHEK 293 cells using 10 μlof Lipofectamine 2000™ with 4 μg of Kir2.1 mutant cDNA in pcDNA3vector. The point mutants were confirmed with DNA sequencing. Cellsused for electrophysiology were seeded on a glass cover slip.

2.2. Solution and chemicals

Tyrode solution contained (mM) NaCl 140, KCl 5.4, MgCl2 1.0, CaCl21.0, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) 10.0and glucose 10 (pH adjusted to 7.3 with NaOH). For whole-cellrecordings, the pipette solution contained (mM) KCl 20, K-aspartate110, MgCl2 1.0, HEPES 10, ethyleneglycoltetraacetic acid (EGTA) 5, GTP0.1, Na2-phosphocreatine 5 and Mg-ATP 5 (pH adjusted to 7.2 withKOH). 3-(4-chlorophenyl) 1-(1,1-dimethylethyl)-1H-pyrazolo[3,4-d]pyrimidin-4-amine (PP2) was purchased from Tocris (Bristol, UK). Allother reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA).Stock solutions of genistein (100 mM), daidzein (100 mM), tyrphostinAG556 (100 mM), AG 1295 (20 mM), and PP2 (10 mM) were madewith dimethylsulfoxide (DMSO), divided into aliquots and stored at−20 °C. Sodium orthovanadate (Sigma-Aldrich) stock solution(200 mM) was made with distilled water, and pH value was adjustedto 9.0.

2.3. Electrophysiology

Cells on a coverslip were transferred to a cell chamber (0.5 ml)mountedon the stage of an invertedmicroscope (Diaphot,Nikon, Japan)and superfused at ~2 ml/minwith Tyrode solution.Whole-cell currentswere recorded as described previously [30,31]. Briefly, borosilicate glasselectrodes (1.2-mm OD) were pulled with a Brown–Flaming puller (P-97, Sutter Instrument, Novato, CA) and had tip resistances of 2–3 MΩwhenfilledwith thepipette solution. A 3 MKCl–agar bridgewasused asthe reference electrode. The tip potential was zeroed before the patchpipette touched the cell. After a gigaohm-seal was obtained by negativesuction, the cell membrane was ruptured by applying gentle negativepressure to establish thewhole cell configuration. Series resistance (Rs)was 3–5 MΩ and was compensated by 80% to minimize voltage errors.The liquid junction potential (~12 mV) was not corrected in theexperiments. Membrane currents were measured using an EPC-10amplifier and Pulse software (Heka Elektronik, Lambrecht, Germany).Commandpulsesweregeneratedbya12-bit digital-to-analog convertercontrolled by Pulse software. Current signals were low-pass filtered at5 kHz, recorded and stored in the hard disk of an IBM compatiblecomputer. All experiments were conducted at room temperature (22–23 °C).

2.4. Immunoprecipitation and Western blotting

The immunoprecipitation and Western blotting were performedusing the procedure described previously [17–19]. Cells were treatedrespectively with 1 mM orthovanadate, 30 μM AG556, orthovanadateplusAG556, 50 μMgenistein, 1 mMorthovanadate plus 50 μMgenistein

and 10 μM PP2 for 30 min at room temperature, and centrifuged(1500 rpm for 10 min) at 4 °C. The cell pellet was then lysed with thelysis buffer containing (mM) 25 Tris, 150 NaCl, 1.0 NaF, 1.0 EDTA, 1.0VO4

−3, 1.0 phenylmethylsulfonyl fluoride, and 1%Na deoxycholate, 0.1%SDS, 1% Triton X-100, 1 μg/ml leupeptin, and 1 μg/ml aprotinin. Proteinquantification of the lysateswasmade using a protein assay reader (Bio-Rad Laboratories, Hercules, CA, USA), and diluted to equal concentra-tions. Proteinswere immunoprecipitated overnight at 4 °C using 2 μg ofanti-Kir2.1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA)and 100 μl of protein A beads (Millipore, Billerica, MA, USA). Immuno-precipitated proteins bound to pelleted protein A beads were washedthoroughly in PBS, denatured in Laemmli sample buffer, separated usingSDS-PAGE, and electroblotted onto nitrocellulose membranes. Theimmunoblots were probed with anti-phosphotyrosine antibody(1:1000, Cell signaling Technology Inc., Danvers, MA, USA) overnightat 4 °C in a blocking solution containing 5% BSA in TBS and 0.1% Tween20, and subsequently treated with goat anti-mouse IgG-HRP antibody(1:5000, Santa Cruz Biotechnology) for 1 h at room temperature. Blotswere developed with enhanced chemiluminescence (ECL, AmershamBiosciences, Sweden) and exposed on a X-ray film. The blots were thenstripped and reprobed with the anti-Kir2.1 channel antibody todetermine total Kir2.1 channel proteins. The film was scanned, imagedby a Bio-Imaging System (Syngene, Cambridge, UK), and analyzed viaGeneTools software (Syngene).

2.5. Statistical analysis

Nonlinear curve-fitting was performed using Pulsefit (HEKA) andSigmaplot (SPSS, Chicago, IL). Paired and/or unpaired Student's t-testwere used as appropriate to evaluate the statistical significance ofdifferences between two group means, and ANOVA was used formultiple groups. Group data are expressed as means±SEM. P valuesless than 0.05 were considered to indicate statistically significantdifference.

3. Results

3.1. Inhibition of Kir2.1 current by genistein

Fig. 1A shows the time course of Kir2.1 current (at −120 mV)recorded in a representative cell with a ramp protocol from −120 to0 mV from a holding potential of−40 mVevery 20 s in the absence andpresence of 30 μM genistein. Current–voltage (I–V) relation curves ofKir2.1 current are shown in right of the panel. Genistein inhibited Kir2.1current, and the inhibitory effect recovered upon washout (n=7).Similarly, voltage-dependent Kir2.1 current elicited by the voltage stepprotocol as shown in the inset was reversibly suppressed by 30 μMgenistein (Fig. 1B). This result is comparable to the previous reports innative cardiac IK1 [25] and Kir2.1 channels expressed in HEK 293 cells[29]. However, the result is different from those in tsA-201 cellsexpressing Kir2.1 gene [26], human bladder urothelial cells [27] andhuman myoblasts [28], in which genistein increases the Kir current.

To exclude possible direct channel-blocking effect by genistein, wetested whether daidzein (a PTK-inactive analog of genistein) couldsuppress Kir2.1 current. Fig. 1B shows that voltage-dependent Kir2.1current was slightly decreased by 30 μM daidzein, and the effectrecovered on washout. Fig. 1D illustrates the mean percent values ofKir2.1 current at −120 mV in the absence and presence of equimolargenistein (n=7) and daidzein (n=5). Genistein at 10, 30, and 100 μMinhibited Kir2.1 current by 8.6±3.3%, 16.3±4.1%, and 43.9±6.9%,respectively (Pb0.05 or Pb0.01 vs. control). However, daidzein had noeffect on the current at 10 μM, and a slight inhibitionwas seen at 30 and100 μM (6.3±1.3% and 9.3±1.2%, Pb0.05 vs. control). The effect ofdaidzein on Kir2.1 current wasweaker than that of equimolar genistein(Pb0.05 or Pb0.01) which suggests that the inhibitory effect ofgenistein on Kir2.1 current is likely mediated by PTK pathway.

Fig. 1. Inhibition of Kir2.1 current by genistein and daidzein. A. Time course of Kir2.1current (at−120 mV) recorded in a representative cells activated by a 2-s voltage rampfrom−120 to 0 mV. I–V relationships of Kir2.1 current at corresponding time points areshown in right of the panel. B. Voltage-dependent Kir2.1 current recorded in a typicalexperiment with the voltage protocol as shown in the inset in the absence and presenceof 30 μM genistein. Arrow indicates zero current level. C. Voltage-dependent Kir2.1current recorded in a typical experiment with the voltage protocol as shown in the insetof B in the absence and presence of 30 μM daidzein. D. Mean percent values of Kir2.1current measured at end of step protocol (−120 mV) from zero level in the absenceand presence of 10, 30, and 100 μM genistein or daidzein (n=5–7, *Pb0.05, **Pb0.01vs. control; #Pb0.05, ##Pb0.01 vs. daidzein).

Fig. 2. Genistein and orthovanadate effects on Kir2.1 current. A. Voltage-dependentKir2.1 current recorded in a representative cell with the voltage protocol as shown inthe inset during control, in the presence of 50 μM genistein, and genistein plus 1 mMorthovanadate (OV). B. Time-course of Kir2.1 current recorded in a typical experimentwith a 2-s ramp voltage protocol from −120 to 0 mV from a holding potential of−40 mV, in the absence and presence of 50 μM genistein, and genistein plus 1 mMorthovanadate. The current was measured at −120 mV from zero. I–V relationships ofKir2.1 current at corresponding time points are shown in right of the panel. C. Meanpercent values of Kir2.1 current measured at end of voltage step (−120 mV) in theabsence and presence of 50 μM genistein or genistein plus 1 mM orthovanadate. (n=6,**Pb0.01 vs. control, #Pb0.05 vs. genistein alone).

1995D.-Y. Zhang et al. / Biochimica et Biophysica Acta 1808 (2011) 1993–1999

If the inhibition of Kir2.1 current by genistein is mediated by PTKs,the effect should be countered by the PTP inhibitor orthovanadate.Orthovanadate at 1 mMhadnoeffect onKir2.1 current (n=11, data notillustratively shown); however, it antagonized the inhibitory effect ofKir2.1 current by genistein. Fig. 2A shows the voltage-dependent Kir2.1current recorded in a typical experiment. Genistein at 50 μM decreasedthe current, and the inhibitory effectwas countered by co-application of1 mM orthovanadate. Similar results were seen in the time course ofKir2.1 current recorded in a representative cell with a ramp protocolfrom −120 to 0 mV from a holding potential of −40 mV (Fig. 2B).Fig. 2C summarizes the mean percent values of Kir2.1 current in theabsence and presence of genistein or genistein plus orthovanadate.Genistein at 50 μM suppressed Kir2.1 current (at −120 mV) to 75.5±1.9% of control (360.9±40.0 pA/pF for control to 265.9±34.4 pA/pFfor genistein, n=6, Pb0.01), and the inhibitory effect was reversed to92.2±4.25% (336.9±39.1 pA/pF) by co-application of genistein and1 mM orthovanadate (n=6, Pb0.01 vs. genistein). These resultssupport the notion that the inhibition of Kir2.1 current by genistein ismediated by PTKs.

To investigate which specific PTK mediates Kir2.1 channelregulation, AG556 (a highly selective EGFR kinase inhibitor) [32,33],

AG1295 (a selective PDGFR kinase inhibitor) [34], and PP2 (a selectiveSrc-family kinase inhibitor) [35] were employed in the followingexperiments.

3.2. Inhibition of Kir2.1 current by AG556

Fig. 3A shows the voltage-dependent Kir2.1 current recorded in arepresentative cell in the absence and presence of 30 μM AG556 orAG556 plus 1 mM orthovanadate. AG556, like genistein, remarkablyinhibited Kir2.1 current, and the inhibition was almost fully counteredby co-application of orthovanadate. Fig. 3B displays the time course ofKir2.1 current recorded in typical of experiment with a 200-ms voltagestep from −40 to −120 mV every 20 s. The inhibitory effect of stepKir2.1 current by 30 μM AG556 was antagonized by 1 mM orthovana-date. I–V relationships of Kir2.1 current in the absence or presence of30 μM AG556 or AG556 plus 1 mM orthovanadate are illustrated inFig 3C. AG556 significantly decreased Kir2.1 current at −120 to−90 mV and −50 to −30 mV (n=6, Pb0.05 or Pb0.01 vs. control),and the inhibition was countered by co-application of orthovanadate(Pb0.05 or Pb0.01 vs. AG556 alone). Fig. 3D summarizes the meanpercent values of Kir2.1 current (at −120 mV) with 30 μM AG556 andAG556 plus 1 mM orthovanadate. AG556 (30 μM) inhibited Kir2.1 to81.3±2.3% of control (348.0±39.0 pA/pF of control to 281.5±38.2 pA/pF of AG556, n=6, Pb0.01), and the inhibition was reversed to 99.44±0.6% (346.6±40.5 pA/pF, Pb0.01 vs. AG556 alone) of control byorthovanadate. These results suggest that reduction of Kir2.1 current by

Fig. 3. Effects of AG556 on Kir2.1 current. A. Voltage-dependent Kir2.1 current recorded in a representative cell with a voltage protocol as shown in the inset of Fig. 1B during control,after application of 30 μMAG556, and AG556 plus 1 mM orthovanadate (OV). B. Time-course of Kir2.1 current recorded in a typical experiment with a 200-ms voltage step from−40to−120 mV from a holding potential of−40 mV, in the absence and presence of 30 μMAG556, and AG556 plus 1 mM orthovanadate. Original current traces at corresponding timepoints are shown in right of the panel. C. I–V relationships of Kir2.1 current measured at end of the voltage steps as shown in A (n=6, one way ANOVA, *Pb0.05, **Pb0.01 vs.control). D. Mean percent values of Kir2.1 current determined at end of voltage step (−120 mV) in the absence and presence of 30 μMAG556, and AG556 plus 1 mM orthovanadate(n=6, **Pb0.01 vs. control; #Pb0.05 vs. AG556 alone).

Fig. 4. Effect of AG1295 and PP2 on Kir2.1 current. A. Voltage-dependent Kir2.1 currentrecorded in a representative cell with the protocol shown in the inset in the absenceand presence of 10 μM AG1295. B. Voltage-dependent Kir2.1 current recorded in atypical experiment in the absence and presence of 10 μMPP2. C. Mean percent values ofKir2.1 current determined at end of voltage step (−120 mV) in the absence andpresence of 10 μM AG1295, or 10 μM PP2 (n=5 each group).

1996 D.-Y. Zhang et al. / Biochimica et Biophysica Acta 1808 (2011) 1993–1999

AG556 is likelymediated by EGFR kinase. Nonetheless, EGF (100 ng/ml)had no effect onbasal Kir2.1 current (n=6, data not shown), suggestingthat tyrosine phosphorylation level of the channel is at a saturated state.

3.3. Effect of PP2 and AG1295 on Kir2.1 current

Fig. 4 shows that the selective PDGFR kinase inhibitor tyrphostinAG1295 (10 μM, 20 times higher than IC50 for inhibiting PDGFRtyrosine kinase) [36], and the membrane permeable Src-family kinaseinhibitor PP2 (10 μM, 2000 times higher than IC50 for inhibiting Src-kinases) [34,35,37] had no inhibitory effect on Kir2.1 current. In thecells treated with AG1295, the mean values of Kir2.1 current at−120 mV were −354.7±36.8 and −347.9±43.3 pA/pF before andafter AG1295 treatment for 8 min (n=5, P=NS). In the cells treatedwith PP2, the mean values of Kir2.1 current were −389.2±48.5 and−380.7±44.8 pA/pF before and after PP2 treatment (n=5, P=NS).

3.4. Tyrosine phosphorylation level of Kir2.1 channels

Fig. 5A displays the confirmative data showing significantexpression of EGFR protein in HEK 293 cells with/without Kir2.1gene expression, which is consistent with previous report [38].Tyrosine phosphorylation of Kir2.1 channels was examined usingimmunoprecipitation and Western blot analysis. Fig. 5B illustratestyrosine phosphorylation level of Kir2.1 channels in the absence andpresence of 1 mM orthovanadate, 50 μM genistein, 1 mM orthovana-date plus genistein, 10 μM PP2, 100 ng/ml EGF, 30 μM AG556, 1 mMorthovanadate plus AG556. Orthovanadate and EGF had no effect ontyrosine phosphorylation level of Kir2.1 channels, suggesting thatbasal tyrosine phosphorylation level of Kir2.1 channel is at a saturatedstate. Genistein and AG556, but not PP2, significantly reduced thephosphorylation level. The reduction was antagonized by pretreat-ment (20 min) with 1 mM orthovanadate. Fig. 5C summarizes themean percent values of quantitative tyrosine phosphorylation levelsof Kir2.1 channels. The tyrosine phosphorylation level of Kir2.1channels was reduced to 74.1%±4.9% of control (n=5 experiments)in cells treated with 50 μM genistein (Pb0.01 vs. control), and theinhibition was partially reversed to 84.0±4.0% of control in the cellstreated with 1 mM orthovanadate plus genistein (P=NS vs. genistein

alone). In cells treated with 30 μM AG556, the phosphorylation levelwas 78.7±3.5% of control (Pb0.01 vs. control), and 92.4±1.6% in thecells treatedwith 1 mMorthovanadate plus AG556 (Pb0.05 vs. AG556alone). These results indicate that the inhibition of EGFR kinasereduces tyrosine phosphorylation of Kir2.1 channel protein.

Fig. 5. EGFR expression in HEK 293 cells and tyrosine phosphorylation levels of Kir2.1channels. A. Western blotting image of anti-EGFR antibody in HEK293 cells with/without Kir2.1 gene transfection. B. Upper panel, protein lysates were immunopreci-pitated with anti-Kir2.1 channel antibody, Western blots were then prepared andprobed with anti-phosphotyrosine(Tyr(P)) antibody. Lower panel, Western blots of celllysates, probed with anti-Kir2.1 channel antibody. C. Histogram summarizes the meanvalues of the relative levels of tyrosine-phosphorylated Kir2.1 channel protein analyzedby the densitometry. The amount of protein from the immunoprecipitation as in B wasnormalized to those from theWestern blots of Kir2.1 channels. Relative phosphorylatedprotein level (n=5 experiments) is expressed as a percentage of the vehicle control.**Pb0.01, *Pb0.05 (ANOVA) vs. vehicle control, #Pb0.05 vs. AG556 alone. OV,orthovanadate; EGF, epidermal growth factor.

Fig. 6. Single point mutations of Kir2.1 channels and AG556 effects. A. WT and mutant Kir2potential of −40 mV in the absence and presence of 30 μM AG556. B. Mean percent vol(−120 mV) from the zero (broken line) in the absence or presence of 3, 10, and 30 μM A**Pb0.01 vs. control or Y242A or Y242F (n=5–12 experiments).

1997D.-Y. Zhang et al. / Biochimica et Biophysica Acta 1808 (2011) 1993–1999

3.5. PTK phosphorylation sites of Kir2.1 channels

The predicted potential PTK phosphorylation sites were examinedwith the software NetPhos 2.0 (www.cbs.dtu.dk/cgi-bin). The mutantsof the top three potential PTK phosphorylation sites Y242, Y336 andY366 were generated by substituting tyrosine with alanine orphenylalanine. Fig. 6A shows the inhibitory response ofWT andmutantKir2.1 currents (recordedwith a 200-ms voltage step to−120 mV from−40 mV) to the EGFR kinase AG556 (30 μM). A slight reduction ofKir2.1 current was observed in Y242A or Y242F with 30 μM AG556,whereas a significantdecrease of the currentwas seen inWT, Y336AandY366A Kir2.1 currents. Fig. 6B illustrates themean percent values ofWTand different mutant Kir2.1 channels with 3, 10, and 30 μM AG556relative to control (before AG556 treatment). AG556 at 10 and 30 μMsignificantly inhibited Kir2.1 current (at −120 mV, n=6–7 experi-ments, Pb0.05 or Pb0.01 vs. control) in WT (by 12.7±1.9 and 22.2±2.1%), Y336A (13.5±2.1 and 20.8±2.2%) and Y366A (14.6±2.5 and21.4±2.9%), but not in mutants Y242A (4.7±2.3 and 5.4±3.3%) andY242F (3.0±4.8 and 6.6±3.7%).

The results of further immunoprecipitation and Western blotanalysis showed that tyrosine phosphorylation level was dramaticallyreduced in cells expressing in Y242F Kir2.1 mutants (Fig. 7A). Theseresults suggest that the tyrosine phosphorylation site of Kir2.1channels is located at the Y242 residue (Fig. 7B), which is comparableto that of the previous report [26,28].

4. Discussion

The present study demonstrates that the PTK inhibitor genistein orAG556, but not AG1295 and PP2, decreases activity of Kir2.1 channelsstably expressed in HEK 293 cells. The inhibitory effect is countered bythe PTP inhibitor orthovanadate. Protein tyrosine phosphorylation levelof Kir2.1 channels is reduced by the broad spectrum PTK inhibitorgenistein or EGFR kinase inhibitor AG556, and the reduced phosphor-ylation level is reversed by orthovanadate. In addition, site-directedmutation reveals that EGFR kinase is located at Y242 of Kir2.1 channels.

.1 channel currents recorded with a 200-ms voltage step to −120 mV from a holdingumes of WT and mutant Kir2.1 channel currents determined at end of voltage stepG556. Y242A and Y242F mutants were not sensitive to inhibition by AG556. *Pb0.05,

Fig. 7. Reduced tyrosine phosphorylation of Y242mutant. A. Images andmean values oftyrosine phosphorylation for WT and Y242F Kir2.1 channels with and without AG556treatment (n=3 experiments, **Pb0.01 vs. WT, #Pb0.01 vs. WT and WT plus AG556).B. Schematic model for regulation of Kir2.1 channels by EGFR kinases. Tyrosine 242residue of Kir2.1 channels is phosphorylated by EGFR kinase. Phosphorylation oftyrosine by one or more steps (dashed line) favors Kir2.1 channel activation, whereasinhibition of phosphorylation may favor channel closure. Genistein or AG556 inhibitsEGFR tyrosine kinase and reduces the phosphorylation; the inhibition is countered bythe PTP inhibitor orthovanadate (OV).

1998 D.-Y. Zhang et al. / Biochimica et Biophysica Acta 1808 (2011) 1993–1999

In addition to the activation by PIP2 via an electrostatic interaction ofthe negatively charged head groups of PIP2 with charged residues in theC-terminal region of the Kir channel [8–10], cardiac IK1 and/or Kir2.1channels are down-regulated by the β-adrenoceptor against isoproter-enol or adenylate cyclase activator forskolin via activating cAMP/proteinkinase A [1,39], and by α1-adrenoceptor agonist methoxamine viaactivating protein kinase C [40]. Acidosis also modulates IK1 in a species-dependent manner [1]. We and others have demonstrated that cardiacIK1 and/or Kir2.1 channels are regulated by PTKs [25–28], although theresponse of native or cloned Kir2.1 channels to PTK inhibitors and/ortyrosinephosphorylation is different (down-regulationorup-regulation)in various expression systems and/or cell types.

The present study has demonstrated that endogenous EGFRs arepresent in HEK 293 cells (Fig. 5A), consistent with the previousobservation [38]. Inhibition of tyrosine phosphorylation of Kir2.1channels down-regulates the current. First, Kir2.1 channels stablyexpressed in HEK 239 cells were reversibly suppressed by the broadspectrum PTK inhibitor genistein, consistent with the previousobservation in rat cardiac ventricular myocytes [25] and in HEK 293cells expressing Kir2.1 channels [29]. Second, genistein exhibits astronger inhibition of Kir2.1 current than that of the PTK-inactive analogdaidzein (Fig. 1). Third, the PTP inhibitor orthovanadate almost fullycountered the suppression of cardiac IK1 [25] or Kir2.1 current bygenistein (Fig. 2) or by the selective EGFR kinase inhibitor AG556(Fig. 3). Fourth, genistein and AG556 decreased tyrosine phosphoryla-tion level of Kir2.1 channels and the reduced phosphorylation level was

significantly antagonizedbyorthovanadate (Fig. 5).Moreover,we foundthat Y242 is the PTK phosphorylation site since the site-directedmutation of Y242 lost the response to the EGFR kinase inhibitor AG556(Fig. 6). However, the PDGFR kinase inhibitor AG1295 (10 μM, 20 timeshigher than IC50 for inhibiting PDGFR tyrosine kinase) [34] and the Src-family inhibitor PP2 (10 μM, 2000 times higher than IC50 for inhibitingSrc-family kinase) [35] had no effect on Kir2.1 channels (Fig. 4). Theseresults support thenotion that activity of human cardiac Kir2.1 channelsis up-regulatedbyphosphorylatingY242 residue in the channel by EGFRkinase. Orthovanadate and EGF had no effect on Kir2.1 current orphosphorylation, this is likely related to the saturated phosphorylationof the channels as observed in cardiac IKs [19] and hERG channels [17].This differs from native guinea pig cardiac INa [23] and Kir2.3 channelsexpressed in HEK 293 cells [41], in which ion channel proteins can besignificantly upregulated by orthovanadate or EGF application.

However, in tsA-201 cells expressing rat Kir2.1 channels, Wisch-meyer and colleagues found that the PTP inhibitor perorthovanadatedecreased the current, and the inhibitory effect was countered by thePTK inhibitor genistein. A similar effect was observed in native Kir2.1channels in rat basophilic leukocytes. Perorthovanadate failed tosuppress the mutant Y242F channels expressed in tsA-201 cells andXenopus oocytes. Moreover, Kir2.1 current was rapidly suppressed bynerve growth factors (NGF) or EGF, or insulin in Xenopus oocytes co-expressing Kir2.1 channels and growth factor receptors. These resultssuggest that an increased phosphorylation of Kir2.1 at Y242 byinhibiting PTPs or activating PTKs down-regulates Kir2.1 current inneuronal cells [26]. The tyrosine phosphorylation site Y242 is identicalto that observed in the present study (Fig. 5), although the consequenceof the channel phosphorylation is opposite. It is unknown whether thisopposite regulatory effect of Kir2.1 channels by PTKs is due to differentexpression systems, cell types and/or experimental conditions.

Both EGFR kinase and Src-related kinases modulate the α-subunithERG current of human cardiac IKr [17]. Interestingly, antagonisticregulation of swelling-activated Cl− current (ICl,swell) by Src and EGFRprotein tyrosine kinases was observed in rabbit ventricular myocytes[42,43]. PP2 stimulated, but the EGFR kinase inhibitor PD-153035inhibited ICl,swell [43]. However, Src-family tyrosine kinase regulationis not the case for recombinant cardiac IKs, Kir2.3 channel, and Kir2.1channel, because PP2 had no effect on the IKs [19], Kir2.3 [41], andKir2.1 current (Fig. 4) and also no effect on tyrosine phosphorylationof these channels. This suggests variable regulation responses of ionchannels to tyrosine kinases.

The altered property of Kir2.1 currents was observed in mutations(Fig. 6), i.e. a fast onset of the currents with voltage changes and a fastinactivation component was clearly visible in currents of Kir2.1mutants with alanine. The possible alteration of these mutations maybe related to channel general functionality rather than a specificaction on tyrosine phosphorylation. The similar phenomenonwas alsoobserved in Kir2.3 channels [41].

The Kir2.1 channel is not only responsible for repolarization of theaction potential, but also for stabilization of the resting membranepotential in cardiac myocytes [1]. The present and previous studiessupport the notion that PTK inhibitors inhibit Kir2.1 channels andcardiac IK1 [25] by EGFR kinase, which suggests that the tyrosinephosphorylation of Kir2.1 promotes the channel opening in cardiacmyocytes. The dysfunction of Kir2.1 channels can cause long QTsyndrome and increase the risk of sudden death [4–6]. Therefore,regulation of cardiac Kir2.1 channels by EGFR kinase may affect cardiacrepolarization and excitability.

Acknowledgements

The study was supported in part by the General Research Fund760306M from Research Grant Council of Hong Kong and a grant fromSun Chieh Yeh Heart Foundation of Hong Kong. De-Yong Zhang andWei Wu are supported by a postgraduate studentship from the

1999D.-Y. Zhang et al. / Biochimica et Biophysica Acta 1808 (2011) 1993–1999

University of Hong Kong. The authors thank Ms Hai-Ying Sun for theexcellent technical support, and Dr. Carol A. Vandenberg (Universityof California, Santa Barbara, CA) for generously providing us thehKir2.1 plasmid vector.

References

[1] A.N. Lopatin, C.G. Nichols, Inward rectifiers in the heart: an update on IK1, J. Mol.Cell. Cardiol. 33 (2001) 625.

[2] D. Bichet, F.A. Haass, L.Y. Jan, Merging functional studies with structures ofinward-rectifier K(+) channels, Nat. Rev. Neurosci. 4 (2003) 957.

[3] Y. Kubo, T.J. Baldwin, Y.N. Jan, L.Y. Jan, Primary structure and functional expressionof a mouse inward rectifier potassium channel, Nature 362 (1993) 127.

[4] H.J. Jongsma, R. Wilders, Channelopathies: Kir2.1 mutations jeopardize many cellfunctions, Curr. Biol. 11 (2001) R747–R750.

[5] A.S. Dhamoon, J. Jalife, The inward rectifier current (IK1) controls cardiacexcitability and is involved in arrhythmogenesis, Heart Rhythm 2 (2005) 316.

[6] G.J. Gross, IK1: the long and the short QT of it, Heart Rhythm 3 (2006) 336.[7] C. Wang, U.L. Mirshahi, B. Liu, Z. Jia, T. Mirshahi, H. Zhang, Arachidonic acid

activates Kir2.3 channels by enhancing channel-phosphatidyl-inositol 4,5-bisphosphate interactions, Mol. Pharmacol. 73 (2008) 1185.

[8] C.M. Lopes, H. Zhang, T. Rohacs, T. Jin, J. Yang, D.E. Logothetis, Alterations inconserved Kir channel-PIP2 interactions underlie channelopathies, Neuron 34(2002) 933.

[9] M. Takano, S. Kuratomi, Regulation of cardiac inwardly rectifying potassiumchannels by membrane lipid metabolism, Prog. Biophys. Mol. Biol. 81 (2003) 67.

[10] C.L. Huang, S. Feng, D.W. Hilgemann, Direct activation of inward rectifierpotassium channels by PIP2 and its stabilization by Gbetagamma, Nature 391(1998) 803.

[11] S. Bendahhou, M.R. Donaldson, N.M. Plaster, M. Tristani-Firouzi, Y.H. Fu, L.J.Ptacek, Defective potassium channel Kir2.1 trafficking underlies Andersen–Tawilsyndrome, J. Biol. Chem. 278 (2003) 51779.

[12] P.S. Lange, F. Er, N. Gassanov, U.C. Hoppe, Andersen mutations of KCNJ2 suppressthe native inward rectifier current IK1 in a dominant-negative fashion, Cardiovasc.Res. 59 (2003) 321.

[13] S.R. Hubbard, J.H. Till, Protein tyrosine kinase structure and function, Annu. Rev.Biochem. 69 (2000) 373.

[14] M.J. Davis, X. Wu, T.R. Nurkiewicz, J. Kawasaki, P. Gui, M.A. Hill, E. Wilson,Regulation of ion channels by protein tyrosine phosphorylation, Am. J. Physiol.Heart Circ. Physiol. 281 (2001) H1835–H1862.

[15] T. Ogura, L.M. Shuba, T.F. McDonald, L-type Ca2+ current in guinea pig ventricularmyocytes treated with modulators of tyrosine phosphorylation, Am. J. Physiol.276 (1999) H1724–H1733.

[16] X.L. Du, Z. Gao, C.P. Lau, S.W. Chiu, H.F. Tse, C.M. Baumgarten, G.R. Li, Differentialeffects of tyrosine kinase inhibitors on volume-sensitive chloride current inhuman atrial myocytes: evidence for dual regulation by Src and EGFR kinases,J. Gen. Physiol. 123 (2004) 427.

[17] D.Y. Zhang, C.P. Lau, Y. Wang, H.F. Tse, G.R. Li, Both EGFR kinase and Src-relatedtyrosine kinases regulate human ether-a-go-go-related gene potassium channels,Cell. Signal. 20 (2008) 1815.

[18] S. Missan, P. Linsdell, T.F. McDonald, Tyrosine kinase and phosphatase regulation ofslow delayed-rectifier K+ current in guinea-pig ventricular myocytes, J. Physiol. 573(2006) 469.

[19] M.Q. Dong, H.Y. Sun, Q. Tang, H.F. Tse, C.P. Lau, G.R. Li, Regulation of human cardiacKCNQ1/KCNE1 channel by epidermal growth factor receptor kinase, Biochim.Biophys. Acta 1798 (2010) 995.

[20] G.R. Li, M.Q. Dong, Pharmacology of cardiac potassium channels, Adv. Pharmacol.59 (2010) 93.

[21] Z. Tiran, A. Peretz, B. Attali, A. Elson, Phosphorylation-dependent regulation ofKv2.1 channel activity at tyrosine 124 by Src and by protein-tyrosine phosphataseepsilon, J. Biol. Chem. 278 (2003) 17509.

[22] P. Gomes, T. Saito, C.C. Del, A. Alioua, M. Eghbali, L. Toro, E. Stefani, Identification ofa functional interaction between Kv4.3 channels and c-Src tyrosine kinase,Biochim. Biophys. Acta 1783 (2008) 1884.

[23] H. Liu, H.Y. Sun, C.P. Lau, G.R. Li, Regulation of voltage-gated cardiac sodiumcurrent by epidermal growth factor receptor kinase in guinea pig ventricularmyocytes, J. Mol. Cell. Cardiol. 42 (2007) 760.

[24] C.A. Ahern, J.F. Zhang, M.J. Wookalis, R. Horn, Modulation of the cardiac sodiumchannel NaV1.5 by Fyn, a Src family tyrosine kinase, Circ. Res. 96 (2005) 991.

[25] Z. Gao, C.P. Lau, T.M.Wong, G.R. Li, Protein tyrosine kinase-dependent modulationof voltage-dependent potassium channels by genistein in rat cardiac ventricularmyocytes, Cell. Signal. 16 (2004) 333.

[26] E. Wischmeyer, F. Doring, A. Karschin, Acute suppression of inwardly rectifyingKir2.1 channels by direct tyrosine kinase phosphorylation, J. Biol. Chem. 273(1998) 34063.

[27] Y. Sun, M. Chen, B.H. Lowentritt, P.S. Van Zijl, K.R. Koch, S. Keay, J.M. Simard, T.C.Chai, EGF and HB-EGF modulate inward potassium current in human bladderurothelial cells from normal and interstitial cystitis patients, Am. J. Physiol. CellPhysiol. 292 (2007) C106–C114.

[28] V. Hinard, D. Belin, S. Konig, C.R. Bader, L. Bernheim, Initiation of human myoblastdifferentiation via dephosphorylation of Kir2.1 K+ channels at tyrosine 242,Development 135 (2008) 859.

[29] Z. Zhao, B. Liu, G. Zhang, Z. Jia, Q. Jia, X. Geng, H. Zhang, Molecular basis forgenistein-induced inhibition of Kir2.3 currents, Pflügers Archiv European,J. Physiol. 456 (2008) 413.

[30] D.Y. Zhang, C.P. Lau, G.R. Li, Human Kir2.1 channel carries a transient outwardpotassium current with inward rectification, Pflugers Arch. 457 (2009) 1275.

[31] M. Tian, M.Q. Dong, S.W. Chiu, C.P. Lau, G.R. Li, Effects of the antifungal antibioticclotrimazole on human cardiac repolarization potassium currents, Br. J.Pharmacol. 147 (2006) 289.

[32] T. Brenner, E. Poradosu, D. Soffer, C. Sicsic, A. Gazit, A. Levitzki, Suppression ofexperimental autoimmune encephalomyelitis by tyrphostin AG-556, Exp. Neurol.154 (1998) 489.

[33] A. Gazit, N. Osherov, I. Posner, P. Yaish, E. Poradosu, C. Gilon, A. Levitzki,Tyrphostins. 2. Heterocyclic and alpha-substituted benzylidenemalononitriletyrphostins as potent inhibitors of EGF receptor and ErbB2/neu tyrosine kinases,J. Med. Chem. 34 (1991) 1896.

[34] M. Kovalenko, A. Gazit, A. Bohmer, C. Rorsman, L. Ronnstrand, C.H. Heldin, J.Waltenberger, F.D. Bohmer, A. Levitzki, Selective platelet-derived growth factorreceptor kinase blockers reverse sis-transformation, Cancer Res. 54 (1994) 6106.

[35] J.H. Hanke, J.P. Gardner, R.L. Dow, P.S. Changelian, W.H. Brissette, E.J. Weringer, B.A. Pollok, P.A. Connelly, Discovery of a novel, potent, and Src family-selectivetyrosine kinase inhibitor. Study of Lck- and FynT-dependent T cell activation,J. Biol. Chem. 271 (1996) 695.

[36] S. Naraghi, S. Khoshyomn, J.A. DeMattia, D.W. Vane, Receptor tyrosine kinaseinhibition suppresses growth of pediatric renal tumor cells in vitro, J. Pediatr.Surg. 35 (2000) 884.

[37] X. Zhu, J.L. Kim, J.R. Newcomb, P.E. Rose, D.R. Stover, L.M. Toledo, H. Zhao, K.A.Morgenstern, Structural analysis of the lymphocyte-specific kinase Lck incomplex with non-selective and Src family selective kinase inhibitors, Structure7 (1999) 651.

[38] H.K. Kramer, I. Onoprishvili, M.L. Andria, K. Hanna, K. Sheinkman, L.B. Haddad, E.J.Simon, Delta opioid activation of the mitogen-activated protein kinase cascadedoes not require transphosphorylation of receptor tyrosine kinases, BMCPharmacol. 2 (2002) 5.

[39] S. Koumi, J.A. Wasserstrom, R.E. Ten Eick, Beta-adrenergic and cholinergicmodulation of inward rectifier K+ channel function and phosphorylation inguinea-pig ventricle51, J. Physiol. 486 (Pt 3) (1995) 661.

[40] R. Sato, S. Koumi, Modulation of the inwardly rectifying K+ channel in isolatedhuman atrial myocytes by alpha 1-adrenergic stimulation, J. Membr. Biol. 148(1995) 185.

[41] D.Y. Zhang, Y.H. Zhang, H.Y. Sun, C.P. Lau, and G.R. Li, Epidermal growth factorreceptor tyrosine kinase regulates the human inward rectifier potassium channelKir2.3 stably expressed in HEK 293 cells, Br. J. Pharmacol. (in press), doi:10.1111/j.1476-5381.2011.001424.x.

[42] D.M. Browe, C.M. Baumgarten, EGFR kinase regulates volume-sensitive chloridecurrent elicited by integrin stretch via PI-3K and NADPH oxidase in ventricularmyocytes, J. Gen. Physiol. 127 (2006) 237.

[43] Z. Ren, C.M. Baumgarten, Antagonistic regulation of swelling-activated Cl−

current in rabbit ventricle by Src and EGFR protein tyrosine kinases, Am. J. Physiol.Heart Circ. Physiol. 288 (2005) H2628–H2636.