View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Journal of the American College of Cardiology Vol. 38, No. 5, 2001 © 2001 by the American College of Cardiology ISSN 0735-1097/01/$20.00 Published by Elsevier Science Inc. PII S0735-1097(01)01558-3 Activates Extracellular Signal-Regulated Kinases Through Pyk2 and Epidermal Growth Factor Receptor in Rat Cardiomyocytes Satoko Tahara, MD,* Keiichi Fukuda, MD, PHD,† Hiroaki Kodama, MD,* Takahiro Kato, MD,* Shunichiro Miyoshi, MD, PHD,‡ Satoshi Ogawa, MD, PHD* Tokyo, Japan

ϩ OBJECTIVES We sought to determine whether potassium (K ) channel blockers (KBs) can activate extracellular signal-regulated kinase (ERK) and to characterize the upstream signals leading to ERK activation in cardiomyocytes. ϩ BACKGROUND Because KBs attenuate K outward current, they may possibly prolong the duration of action ϩ ϩ potentials, leading to an increase in calcium (Ca2 ) transient ([Ca2 ] ) in cardiomyocytes. ϩ i ϩ Elevation of intracellular Ca2 levels can trigger various signaling events. Influx of Ca2 ϩ through L-type Ca2 channels after membrane depolarization induced activation of MEK and ERK through activation of Ras in neurons. Although KBs are frequently used to treat cardiac arrhythmias, their effect on signaling pathways remains unknown. METHODS Primary cultured rat cardiomyocytes were stimulated with four different KBs—4- aminopyridine (4-AP), E-4031, tetra-ethylammonium and —and phosphorylation of ERK, proline-rich tyrosine kinase 2 (Pyk2) and epidermal growth factor receptor (EGFR) was detected. Action potentials were recorded by use of a conventional microelectrode. 2ϩ (Ca )i was monitored by the fluorescent calcium indicator Fluo-4. RESULTS E-4031, 4-AP, tetra-ethylammonium and quinidine induced phosphorylation of ERK. 2ϩ 4-Aminopyridine prolonged the duration of action potentials by 37% and increased (Ca )i by 52% at 1 mmol/l. Pre-incubation of ethyleneglycoltetraacetic acid, 1,2-bis(2- aminophenoxy)-ethane-N,N,NЈ,NЈ-tetraacetic acid tetrakis and completely blocked this phosphorylation, whereas flufenamic acid and benzamil did not. 4-Aminopyridine induced tyrosine phosphorylation of Pyk2 and EGFR, which peaked at 5 and 10 min, respectively. Cytochalasin D, AG1478 and dominant-negative EGFR strongly inhibited the phosphorylation of ERK, whereas calphostin C, calmidazolium and KN62 did not. ϩ CONCLUSIONS These findings indicate that KBs induce ERK activation, which starts with Ca2 entry ϩ through the L-type Ca2 channel in cardiomyocytes, and that EGFR and Pyk2 are involved in this activation. (J Am Coll Cardiol 2001;38:1554–63) © 2001 by the American College of Cardiology

Since the Cardiac Arrhythmia Suppression Trial was re- sarcoplasmic reticulum (1,2). blockers, ϩ ϩ ported, sodium (Na ) channel blockers have been consid- such as 4-aminopyridine (4-AP), which block K outward ered to deteriorate the prognosis of patients with cardiac currents and prolong the action potential duration (APD), ϩ 2ϩ 2ϩ arrhythmias. blockers decrease Na entry, increase Ca transient ([Ca ]i) in cardiomyocytes (3,4). ϩ which, in turn, decreases calcium (Ca2 ) influx through These data indicate that KBs could exert potentiating effects ϩ ϩ L-type Ca2 channels. Consequently, the decreased Ca2 on contractility by increasing the amplitude of intracellular 2ϩ influx has a negative inotropic effect on cardiomyocytes. In [Ca ]i. ϩ contrast, potassium (K ) channel blockers (KBs) were Mitogen-activated protein kinase (MAPK) family pro- expected to be used more frequently than before. Recent teins transduce signal from diverse receptor types, including ϩ studies have demonstrated that KBs inhibit K outward receptor protein tyrosine kinases, G-protein-coupled recep- current, which, in turn, causes prolongation of action tors and cytokine receptors, and play an important role in potentials. Prolongation of action potentials results in a the proliferation and differentiation of various cell types ϩ ϩ substantial increase in net Ca2 entry through L-type Ca2 (5,6). Among this family, extracellular signal-regulated ϩ channels, and this augmented Ca2 influx enhances myo- kinase (ERK) has been reported to be activated by various ϩ cardial contractility by increasing Ca2 release from the cytokines and growth factors, such as phenylephrine, endothelin-1, angiotensin II and leukemia inhibitory factor From the *Cardiopulmonary Division, Department of Internal Medicine; †Insti- (7), as well as by mechanical stretching in cardiomyocytes. tute for Advanced Cardiac Therapeutics; and the ‡Department of Physiology, Keio University School of Medicine, Tokyo, Japan. This study was supported by research The activation of ERK requires phosphorylation of both a grants from the Ministry of Education, Science and Culture, Japan, and Health threonine and a tyrosine residue by a dual-specificity kinase Science Research Grants for Advanced Medical Technology from the Ministry of known as MEK (MAPK/ERK kinase). Activated ERK Welfare, Japan. Manuscript received November 29, 2000; revised manuscript received June 27, translocates into the nucleus and can phosphorylate a variety 2001, accepted July 19, 2001. of nuclear transcription factors, suggesting that ERK is a JACC Vol. 38, No. 5, 2001 Tahara et al. 1555 November 1, 2001:1554–63 KB-Evoked Signals in Cardiomyocytes

current resistance of 15 to 30 M⍀, were selected. Mem- Abbreviations and Acronyms brane potentials were measured in the current clamp mode 4-AP ϭ 4-aminopyridine (MEZ-8300, Nihon Kohden, Tokyo, Japan) and stored on APD ϭ action potential duration digital magnetic tape (frequency range 0 to 20 kHz; Sony ϭ BAPTA-AM 1,2-bis(2-aminophenoxy)-ethane- Magnescale, Sony Co., Tokyo, Japan) for later analysis. N,N,NЈ,NЈ-tetraacetic acid tetrakis ϩ Ca2 ϭ calcium During measurements, the cells were field-stimulated by 2ϩ ϭ 2ϩ [Ca ]i Ca transient 7-ms pulses at 1 Hz with a step-pulse generator (SET- EGFR ϭ epidermal growth factor receptor 1201, Nihon Kohden, Tokyo, Japan). ϭ ؉ ϩ ERK extracellular signal-regulated kinase Measurement of [Ca2 ] . The [Ca2 ] transient was mon- HA ϭ hemaggulutinin i i ϩ K ϭ potassium itored by use of the fluorescent calcium indicator Fluo-4 KB ϭ potassium channel blocker (Molecular Probes, Eugene, Oregon), as described previ- LIF ϭ leukemia inhibitory factor ously (12). During measurements, the cells were field- MAPK ϭ mitogen-activated protein kinase stimulated by 7-ms pulses at 1 Hz. MEK ϭ MAPK/ERK kinase ϩ Reagents and antibodies. Potassium channel blockers, Na ϭ sodium PKC ϭ protein kinase C including 4-AP, 1-(2-[6-methyl-2-pyridil]-ethyl)-4-(4- Pyk2 ϭ proline-rich tyrosine 2 methylsulfonylaminobenzoyl)-piperidine (E-4031, Eisai, TEA ϭ tetra-ethylammonium Tokyo, Japan), tetra-ethylammonium (TEA) and quini- dine, were purchased from Sigma Chemical Co., Ltd. (Tokyo, Japan). Anti-phospho-ERK antibody was pur- key mediator of transduction of cytoplasmic signals to chased from New England Biolabs (Boston, Massachu- nuclear responses. setts). Anti-phosphotyrosine antibody (RC20H) was pur- ϩ Elevation of intracellular Ca2 levels can trigger various chased from Transduction Laboratories (Lexington, ϩ signaling events. Rosen et al. (8) reported that Ca2 influx Kentucky). Anti-Pyk2 and anti-EGFR antibodies were ϩ through L-type voltage-dependent Ca2 channels after purchased from Santa Cruz Biotechnology, Inc. (Santa membrane depolarization induced activation of MEK and Cruz, California). MAPK through activation of Ras in neurons. Moreover, Immunoprecipitation and Western blot analysis. The 2ϩ Ca influx after membrane depolarization was recently cells were lysed with a buffer, as described previously (7). reported to mediate ligand-independent transactivation of Lysates were incubated with either anti-EGFR or anti- epidermal growth factor receptor (EGFR) in PC12 cells Pyk2 antibody at 4°C for 4 h, followed by protein A- or (9,10). Subsequently, we directed our attention to KBs and G-sepharose (Sigma) for 2 h, and immunoprecipitated. investigated whether the use of KBs can activate the ERK Proteins were separated by 6% sodium dodecyl sulfate- 2ϩ pathway through Ca -dependent signaling. Moreover, we polyacrylamide gel electrophoresis (SDS-PAGE) and im- also explored the upstream signals leading to ERK activa- munoblotted with anti-phosphotyrosine antibody. Western tion by KBs. Here, we report that KBs can induce ERK ϩ blot analysis for phospho-ERK was performed, as described activation, which is initiated by Ca2 entry through L-type 2ϩ previously (7). The signals were visualized by enhanced Ca channels, and that EGFR and proline-rich tyrosine 2 chemiluminescence (Amersham Pharmacia Biotech, Buck- (Pyk2) are critically involved in KB-induced ERK activation inghamshire, England). in cardiomyocytes. Transfection of plasmids and in vitro kinase assay. Tranasfection of plasmids was performed using effectene METHODS transfection reagent (Qiagen, Hilden, Germany), according Cell culture. Primary cultures of cardiomyocytes were to the manufacturer’ s instructions. The deletion mutant prepared from the ventricles of one-day-old neonatal Wistar (533del) of human EGFR (EGFRdel) was provided by rats, as described previously (11). At 48 h after seeding, the Murasawa et al. (13). Hemaggulutinin-tagged ERK2 plas- culture medium was changed to medium containing 0.5% mid (HA-ERK2) with the SV40 promoter was provided by fetal bovine serum, incubated for 6 h, and then KBs were Minden et al. (14). HA-ERK2 was co-transfected with added. these plasmids. After 4-AP treatment, the cells were lysed, Recording of action potentials. Electrophysiologic studies and the lysates were incubated with anti-HA antibody were performed in Tyrode’s solution with 0.5% fetal bovine (Santa Cruz Biotechnology, Inc.). The immunocomplex serum, under various concentrations (0, 0.1 and 1 mmol/l) was precipitated on protein A-sepharose, resuspended in 25 of 4-AP, which was dissolved in distilled water immediately ␮l of kinase buffer and incubated with 25 ␮g of bovine before use. Cultured cells were plated on the stage of an myelin basic protein (Sigma) as a substrate at 30°C for inverted phase-contrast microscope (Diaphoto-300, Nikon, 20 min. The reaction was terminated by adding Laemmli Toyko, Japan) at 37°C. Action potentials were recorded by sample buffer. The supernatants were resolved by SDS- use of a conventional microelectrode. Glass microelectrodes PAGE, and the gel was washed with 5% trichloroacetic acid filled with potassium chloride (3 mol/l), with a direct for 30 min, dried and subjected to auto-radiography. 1556 Tahara et al. JACC Vol. 38, No. 5, 2001 KB-Evoked Signals in Cardiomyocytes November 1, 2001:1554–63

Figure 1. 4-Aminopyridine (4-AP) prolonged the action potential duration (APD) in cardiomyocytes. Action potentials of the cardiomyocytes were recorded as described in the Methods section, using a conventional glass micropippete, and the cells were treated with 4-AP (0.1 and 1 mmol/l) for 5 min. (A) Representative tracings of the action potentials. These tracings were obtained from the same cells. (B) Effect of 4-AP on the APD90. The experiments were repeated three times, and the data are presented as the mean value Ϯ SE obtained from 15 different cells. 4-Aminopyridine increased the APD by 37% at 1 mmol/l. *p Ͻ 0.01 vs. control cells. NS ϭ not significant.

Statistical analysis. The data were processed using Stat- RESULTS View J-4.5 software. All data are expressed as the mean value Ϯ SE. Comparisons among the values of all groups 4-Aminopyridine prolongs the APD in cardiomyocytes. 2ϩ (Figs. 1 to 6) were performed by one-way analysis of To begin to determine whether KBs can activate Ca - variance. For comparisons of myelin basic protein kinase mediated signaling in cardiomyocytes, we needed to deter- activities among all groups in the transfection experiments, mine precisely how KBs can affect action potentials. two-way analysis of variance was performed. The Bonfer- 4-Aminopyridine is known to inhibit Ito and Ikur, which roni method was used to determine the level of significance. may culminate in the prolongation of action potentials. We A p value Ͻ0.05 was considered statistically significant. first recorded the action potentials of cardiomyocytes stim- JACC Vol. 38, No. 5, 2001 Tahara et al. 1557 November 1, 2001:1554–63 KB-Evoked Signals in Cardiomyocytes

2ϩ Figure 2. 4-Aminopyridine (4-AP) enhances calcium transient [Ca ]i in cardiomyocytes. Cardiomyocytes were pretreated with Fluo-4 and treated with 2ϩ 2ϩ 4-AP for the indicated times. [Ca ]i was recorded as described in Methods. (A) Representative tracings of [Ca ]i. The tracings were recorded from the 2ϩ 2ϩ same cells. The cells were treated with 1 mmol/l of 4-AP for 10 min. (B) The time course of [Ca ]i. One mmol/l of 4-AP increased [Ca ]i by 52% at 10 min. The experiments were repeated three times, and the data are presented as the mean value Ϯ SE obtained from 15 different cells. *p Ͻ 0.01 and **p Ͻ 0.05 vs. control cells. ulated by various concentrations of 4-AP (0, 0.1 and cytes with either 4-AP (1 mmol/l), E4031 (2 ␮mol/l), TEA 1 mmol/l). Figure 1A shows a representative tracing of the (10 mmol/l) or quinidine (50 ␮mol/l) for up to 30 min (Fig. action potentials. Figure 1B indicates the effect of 4-AP on 3A to 3D). Interestingly, all KBs induced phosphorylation the APD90 obtained from seven separate experiments. of ERK in a time-dependent manner, which peaked at 4-Aminopyridine prolonged APD90 in a dose-dependent 15 min. We also stimulated the cells with leukemia inhib- manner. One mmol/l of 4-AP lengthened APD90 by 37%. itory factor (LIF) as a positive control agent (Fig. 3E). 2؉ 4-Aminopyridine increases intracellular [Ca ]i. Prolon- These findings indicate that KBs could induce ERK acti- gation of action potentials suggested that 4-AP may in- vation to the same extent as LIF. 2ϩ ϩ crease [Ca ]i in cardiomyocytes. To confirm this, we The KBs may affect the different K channels at different 2ϩ measured [Ca ]i in the cells stimulated by 4-AP (0.1 and concentrations and prolong the APD at certain concentra- 1 mmol/l). Figure 2A shows a representative tracing, and tions. To determine the dose dependency of KBs on ERK 2ϩ Figure 2B shows the time course of [Ca ]i of the cardio- activation, we stimulated the cells with various concentra- myocytes treated with 1 mmol/l of 4-AP. 4-Aminopyridine tions (0.01 to 1 mmol/l) of 4-AP and detected the phos- 2ϩ prominently increased [Ca ]i. 4-Aminopyridine phorylation of ERK. 4-Aminopyridine phosphorylated 2ϩ (1 mmol/l) enhanced [Ca ]i at 1 min and caused it to peak ERK in a dose-dependent manner (Fig. 3F). ؉ at 10 min, by 152 Ϯ 0.5%. Involvement of an L-type Ca2 channel-mediated in- 2؉ KBs activate ERK in cardiomyocytes. The activation of crease in [Ca ]i on KB-mediated ERK activation. ERK represents a point of convergence for growth signals Next, we investigated the molecular mechanism leading to originating from different type of receptors, such as the activation of ERK by KBs. Because KBs are not G-protein-coupled receptors, tyrosine kinase receptors and physiologic ligands, and the aforementioned findings sug- cytokine receptors (6). To address the possibility that the gest that application of KBs significantly modulates intra- ϩ use of KBs can affect intracellular signaling pathways cellular Ca2 in cardiomyocytes, we investigated the possi- 2ϩ 2ϩ through [Ca ]i in cardiomyocytes, we first investigated ble involvement of Ca in KB-mediated ERK activation. whether KBs can activate ERK. We incubated cardiomyo- We pre-incubated the cells with egtazic acid (1 mmol/l) and 1558 Tahara et al. JACC Vol. 38, No. 5, 2001 KB-Evoked Signals in Cardiomyocytes November 1, 2001:1554–63

Figure 3. Potassium channel blockers activate extracellular signal-regulated kinase (ERK) in cardiomyocytes. Cardiomyocytes were treated with (A) 4-aminopyridine (4-AP) (1 mmol/l), (B) E4031 (2 ␮mol/l), (C) (TEA) (10 mmol/l) and (D) quinidine (50 ␮mol/l) for the indicated times, and phosphorylation of ERK was detected by Western blot analysis. All potassium channel blockers induced phosphorylation of ERK, which peaked at 15 min. (E) Leukemia inhibitory factor (LIF) was used as a positive control agent. (F) 4-Aminopyridine caused dose-dependent augmentation of the phosphorylation of ERK in cardiomyocytes. The cells were treated with 0.01, 0.1 and 1 mmol/l of 4-AP for 15 min. All experiments were repeated at least four times and gave similar results.

؉ 1,2-bis(2-aminophenoxy)-ethane-N,N,NЈ,NЈ-tetraacetic Involvement of Pyk2 and EGFR in K channel blocker- acid tetrakis (BAPTA-AM) (20 ␮mol/l), and we detected induced ERK activation. The involvement of L-type ϩ ϩ ϩ the phosphorylation of ERK. Interestingly, 4-AP-induced Ca2 current and Ca2 -induced Ca2 release suggests that ϩ ERK activation was completely abrogated by egtazic acid Ca2 -binding protein may take part in this ERK activation. and BAPTA-AM (Fig. 4A). Next, we investigated the Protein kinase C (PKC), calmodulin-dependent kinases and ϩ ϩ involvement of the voltage-dependent L-type Ca2 chan- Pyk2 are Ca2 -dependent and are considered to be located ϩ ϩ nel, nonselective cation channel and Na /Ca2 exchanger. upstream of ERK in various types of cells (13,15,16). ϩ We pre-incubated the cells with the L-type Ca2 channel Therefore, we pre-incubated the cells with calphostin C ␮ blocker diltiazem (2 ␮mol/l), the nonselective cation chan- (PKC inhibitor; 1 mol/l), calmidazolium (calmodulin ϩ ϩ ␮ nel blocker flufenamic acid (100 ␮mol/l) and the Na /Ca2 inhibitor; 50 mol/l), KN62 (CaMKII and IV inhibitor; ␮ ␮ exchanger blocker benzamil (100 ␮mol/l), and we detected 10 mol/l) and cytochalasin D (10 mol/l), and we mea- ERK phosphorylation. Diltiazem almost completely inhib- sured the phosphorylation of ERK. Cytochalasin D is an ited the phosphorylation of ERK, whereas benzamil and inhibitor of actin polymerization, and other investigators flufenamic acid had no effect on this phosphorylation (Fig. have observed that this compound can inhibit the activation 4B and 4C). To confirm that the augmentation of L-type of Pyk2 (17,18). Because EGFR plays a critical role in ϩ Ca2 current directly activates this phosphorylation, we mediating ERK activation evoked by various ligands and is 2ϩ stimulated the cells with the L-type considered to be parallel to or downstream of Ca -binding Bayk8644 (10 ␮mol/l), and we detected the phosphoryla- protein, we investigated the possible involvement of EGFR. tion of ERK. Bayk8644 induced phosphorylation of ERK in We pre-incubated the cells with the specific EGFR inhib- a time-dependent manner (Fig. 4D). Figure 4E shows the itor AG1478 (250 nmol/l), and we measured ERK phos- densitometric analysis of the mean expression of phosphor- phorylation. Figure 5A and 5B shows the representative blot ylated ERK in five separate experiments. These findings analysis, and Figure 5C shows the densitometric analysis indicate that 4-AP-induced phosphorylation of ERK was obtained from four separate experiments. Calphostin C, ϩ mediated by voltage-dependent L-type Ca2 channels. calmidazolium and KN62 did not inhibit the 4-AP-induced JACC Vol. 38, No. 5, 2001 Tahara et al. 1559 November 1, 2001:1554–63 KB-Evoked Signals in Cardiomyocytes

ϩ ϩ Figure 4. Involvement of calcium (Ca2 ) and L-type Ca2 channels in 4-aminopyridine (4-AP)-induced ERK activation. (A) Cardiomyocytes were pretreated with ethyleneglycoltetraacetic acid (EG) (1 mmol/l) or BAPTA-AM (BAP) (20 ␮mol/l) for 30 min and treated with 4-AP (1 mmol/l) for 15 min. Both EG and BAPTA-AM completely abolished 4-AP-induced extracellular signal-regulated kinase (ERK) phosphorylation. (B) The cells were ϩ ϩ ϩ pretreated with the L-type Ca2 channel blocker diltiazem (Dil) (2 ␮mol/l) and the Na /Ca2 exchanger inhibitor benzamil (BZ) (100 ␮mol/l) for 5 min, and then treated with 4-AP for 15 min. Diltiazem fully inhibited ERK phosphorylation, whereas BZ did not. (C) The cells were pre-incubated with the nonselective cation channel blocker fulfenamic acid (FA) (100 ␮mol/l) and then treated with 4-AP for 15 min. Fulfenamic acid was dissolved in ϩ (vehicle). Fulfenamic acid did not attenuate 4-AP-induced ERK activation. (D) The cells were treated with the L-type Ca2 channel opener Bayk8644 (10 ␮mol/l) for the indicated times. BayK8644 induced phosphorylation of ERK. All experiments were repeated at least five times, and we obtained similar results. (E) Densitometric analysis of the phosphorylation of ERK was shown. Data are presented as the mean value Ϯ SE. *p Ͻ 0.01 vs. 4-AP alone. LIF ϭ leukemia inhibitory factor; NS ϭ not significant. activation of ERK. However, both cytochalasin D and was phosphorylated as early as 5 min and peaked at 10 min AG1478 strongly inhibited this phosphorylation. These after 4-AP stimulation (Fig. 6B). These results were repro- findings suggest that 4-AP-induced ERK activation could ducible in four separate experiment. These findings impli- be regulated through activation of EGFR and Pyk2. cate Pyk2 and EGFR in 4-AP-induced activation of ERK 4-Aminopyridine induces phosphorylation of Pyk2 and in cardiomyocytes. EGFR. To confirm whether 4-AP really activated Pyk2 or Deletion mutant of EGFR inhibits 4-AP-induced ERK EGFR, or both, we performed immunoprecipitation- activation. To confirm that EGFR is critically involved in Western blot analysis to detect tyrosine phosphorylation of 4-AP–induced ERK activation, we co-transfected HA- these two kinases. Pyk2 was phosphorylated as early as tagged ERK2 plasmid with either an EGFR deletion 2 min and peaked at 5 min (Fig. 6A). In addition, EGFR mutant or an empty plasmid and treated it with 4-AP. 1560 Tahara et al. JACC Vol. 38, No. 5, 2001 KB-Evoked Signals in Cardiomyocytes November 1, 2001:1554–63

Figure 5. Involvement of epidermal growth factor receptor (EGFR) and Pyk2 in 4-aminopyridine (4-AP)-induced extracellular signal-regulated kinase (ERK) activation. (A, B) Cardiomyocytes were pre-incubated with either KN62 (KN [calmodulin kinase II and IV inhibitor]) (10 ␮mol/l), AG1478 (AG [epidermal growth factor inhibitor]) (250 nmol/l), cytochalasin D (CyD [actin dimerization inhibitor]) (10 ␮mol/l), calphostin C (Cal [PKC inhibitor]) (1 ␮mol/l) or calmidazolium (CMZ [calmodulin inhibitor]) (50 ␮mol/l) for 30 min and then treated with 4-AP for 15 min. Leukemia inhibitory factor (LIF) was used as a positive control. AG1478 and cytochalasin D strongly blocked 4-AP-induced ERK activation, whereas the other agents did not. (C) Densitometric analysis of the inhibitor experiment. Representative blot analyses are shown. All experiments were repeated at least five times and yielded similar results. Data are presented as the mean value Ϯ SE. *p Ͻ 0.01 vs. 4-AP alone.

Activation of ERK was almost completely inhibited by the 2.5 mmol/l of 4-AP induced apoptosis in HepG2 cells in a ϩ EGFR deletion mutant (Fig. 7). These results were repro- Ca2 -dependent manner (26). This phenomenon can be ϩ duced in four separate experiments. Taken together with the explained by the fact that elevation of the intracellular Ca2 inhibitor experiment, these findings indicate that EGFR concentration activates apoptosis-inducing phosphatases, ϩ ϩ ϩ may play an important role in 4-AP-induced ERK activa- Ca2 -dependent neutral proteinase, Ca2 /Mg2 - ϩ tion. dependent endonuclease and Ca2 -dependent transglu- ϩ taminase. Thus, the effect of K channel modulation on DISCUSSION cellular function may be quite different among different cell ؉ K channel modulation and cellular function. Potassium types. channel blockers, which inhibit repolarization to prolong KB induces ERK activation. The main purpose of the the cardiac action potentials, are commonly used for the present study was to investigate whether pharmacologic ϩ ϩ treatment of cardiac arrhythmias. Meanwhile, K channels inhibition of K currents could influence the hypertrophic have been implicated in various functions in a variety of cell signaling cascade in the cardiomyocytes of rats. This study ϩ types (19–21). Many agents that inhibit K channels have clearly showed that KBs evoked ERK phosphorylation, been shown to suppress cellular proliferation (22,23), and possibly through activation of Pyk2 and EGFR in cardio- ϩ KBs have been used to suppress tumor cell proliferation myocytes, initiated by the entry of Ca2 through L-type ϩ (19,22,24). The KBs have induced apoptosis in malignant Ca2 channels. There are several lines of evidence to astrocytoma cell lines, but an anti-apoptotic effect of KBs suggest that the ERK pathway plays an important role in has also been reported (25). A recent report found that mediating cardiac hypertrophy (5–7). Various hypertrophic JACC Vol. 38, No. 5, 2001 Tahara et al. 1561 November 1, 2001:1554–63 KB-Evoked Signals in Cardiomyocytes

Figure 7. Effect of epidermal growth factor receptor (EGFR) deletion Figure 6. 4-Aminopyridine (4-AP) phosphorylates proline-rich tyrosine mutant on 4-aminopyridine (4-AP)-induced extracellular signal-regulated kinase 2 (Pyk2) (A) and epidermal growth factor receptor (EGFR) (B) in kinase (ERK) activation in cardiomyocytes. (Top) Cardiomyocytes were cardiomyocytes. Cardiomyocytes were stimulated with 4-AP for the co-transfected with HA-tagged ERK2 and either deletion mutant of indicated times, and the tyrosine phosphorylation of Pyk2 and EGFR was EGFR (pDN-EGFR), or empty plasmid (pcDNA), and treated with detected by immunoprecipitation-Western blot analysis. Endothelin 4-AP for 15 min. The immunocomplex with anti-HA antibody was (ET-1) was used as a positive control agent. Anti-p-Tyr represents precipitated on protein A-sepharose, and in vitro myelin basic protein anti-phosphotyrosine antibody. 4-Aminopyridine induced phosphorylation (MBP) kinase assay by ERK2 was performed, as described in Methods. Ϯ of Pyk2 and EGFR at 5 and 10 min, respectively, in cardiomyocytes. Data are presented as the mean value SE of the results of five separate experiments determined by densitometry. (Bottom) Densitometric analysis of the inhibitory effect of dominant-negative EGFR on 4-AP-induced stimuli can induce activation of the ERK pathways. How- myelin basic protein kinase activity of ERK. Data are presented as the Ϯ Ͻ ever, Thorborn et al. (27) reported that overexpression of mean value SE. *p 0.01 vs. 4-AP–treated cells transfected with the empty vector. pDN-EGFR ϭ dominant negative EGFR plasmid. the active forms of Raf-1 and ERK does not cause the sarcomeric organization typical of hypertrophic growth, and Iks. Tetra-ethylammonium inhibits Ikr,Ikur and, partially, Post et al. (28) reported that inhibition of MEK by Iks, but not Ito. E4031 only inhibits Ikr in cultured cardio- Ͼ ␮ PD98059 did not suppress phenylephrine-induced sarco- myocytes (29). Quinidine blocks Ikur,Ikr,Iks ( 300 mol/l) meric organization or atrial natriuretic peptide gene expres- and Ito (30). Although they have different target channels, sion. Moreover, they showed that adenosine triphosphate all of these agents can prolong the APD. Because these and activated the Raf-1/MEK/ERK cascade, agents have different chemical structures, are not physio- although neither of these reagents could cause cardiac logic ligands and have different target channels, the mech- hypertrophy. These findings indicate that activation of ERK anism of ERK activation may be caused by their common alone was not sufficient for the induction of cardiac hyper- effect of prolongation of action potentials. We suspected ϩ trophy and hypertrophic gene expression, as well as suggest that inhibition of specificK currents might not correlate that the role of this pathway was distinct from ligand to with KB-induced ERK activation. ligand. The KBs are frequently used for the treatment of Involvement of EGFR and Pyk2 in 4-AP-induced ERK ϩ cardiac arrhythmias and have never been reported to induce activation. The present study demonstrated that Ca2 - cardiac hypertrophy in vivo. The effect of KBs on the mediated signaling was critically involved in 4-AP-induced ϩ hypertrophy-mediating pathway in cardiomyocytes remains ERK activation. Elevated intracellular levels of Ca2 can unknown; however, it is possible that activation of ERK by trigger various signaling events, resulting in stimulation of ϩ KBs does not directly cause cardiac hypertrophy. the MAPK pathway. In PC12 cells, Ca2 influx after Relationship between ERK activation and specific potas- membrane depolarization by potassium chloride mediates sium currents. In this study, we used four different kinds of ligand-independent phosphorylation of EGFR in both a

KBs. 4-Aminopyridine inhibits Ito and Ikur, but not Ikr and CaMKII-dependent and -independent manner (10,31). 1562 Tahara et al. JACC Vol. 38, No. 5, 2001 KB-Evoked Signals in Cardiomyocytes November 1, 2001:1554–63

Figure 8. Model of 4-aminopyridine (4-AP)-induced extracelluarl signal-regulated kinase (ERK) activation in cardiomyocytes. Treatment with 4-AP 2ϩ 2ϩ results in augmentation of L-type calcium (Ca ) current and calcium transient ([Ca ]i), leading to the activation of proline-rich tyrosine kinase 2 (Pyk2) and EGFR. Transactivation of EGFR predominantly contributes to ERK activation. EGTA ϭ ethyleneglycoltetraaceticacid.

Pyk2 is activated in PC12 cells by extracellular stimuli that action potentials and contractility of the rat right ventricle. Gen 2ϩ Pharmacol 1996;27:379–85. increase intracellular Ca in either a PKC-dependent or ϩ 3. Tokuno T, Muraki K, Watanabe M. Effects of K channel modulators ϩ -independent manner (16). Calmodulin-dependent kinases on the relationship between action potential duration and Ca2 ϩ are activated by an increase in intracellular Ca2 levels transients in single ventricular myocytes of the guinea pig. Jpn ϩ ϩ mediated by Ca2 influx through L-type Ca2 channels J Pharmacol 1999;80:243–53. ϩ 4. Damron DS, Summers BA. enhances contraction 2 ϩ (32,33) or by Ca release from the intracellular store (34). and intracellular Ca2 transients in individual rat ventricular myocytes. Therefore, we examined the downstream pathway of 4-AP- Am J Physiol 1997;272:H350–9. ϩ induced augmentation of intracellular Ca2 leading to ERK 5. Davis RJ. The mitogen-activated protein kinase signal transduction pathway. J Biol Chem 1993;268:14553–6. activation. The present study demonstrated that Pyk2 and 6. Hunter T. Signaling: 2000 and beyond. Cell 2000;100:113–27. EGFR were involved in this activation of ERK (Fig. 8). 7. Kodama H, Fukuda K, Pan J, et al. Significance of Raf-1/MEK/ERK ϩ Because 4-AP only increases intracellular Ca2 levels, we cascade compared with JAK/STAT and PI3-K pathway in gp130- mediated cardiac hypertrophy. Am J Physiol 2000;279:H1635–44. suspected that 4-AP-induced activation of Pyk2 may be ϩ 8. Rosen LB, Ginty DD, Weber MJ, Greenberg ME. Membrane mediated, not by PKC, but only by Ca2 . The molecular depolarization and calcium influx stimulate MEK and MAP kinase via mechanism of the activation of EGFR by 4-AP remains activation of Ras. Neuron 1994;12:1207–21. 9. Zwick E, Daub H, Ullrich A, et al. Critical role of calcium-dependent unknown, but we supposed that Pyk2 might be involved in epidermal growth factor receptor transactivation in PC cell membrane EGFR activation, based on the finding that cytochalasin D, depolarization and bradykinin signaling. J Biol Chem 1997;272: but not calmidazolium, calphostin C nor KN62, inhibited 24767–70. 10. Rosen LB, Greenberg ME. Stimulation of growth factor receptor the activation of ERK. signal transduction by activation of voltage-sensitive calcium channels. Clinical implications. The KBs are known to prolong QT Proc Natl Acad SciUSA1996;93:1113–8. intervals in the clinical setting, and many reports have 11. Kodama H, Fukuda K, Pan J, et al. Leukemia inhibitory factor, a potent cardiac hypertrophic cytokine, activates the JAK/STAT path- focused on the electrophysiologic side effects of high con- way in rat cardiomyocytes. Circ Res 1997;81:656–63. centrations of KBs, such as torsade de pointes and even fatal 12. Kato T, Sano M, Fukuda K, et al. Calmodulin kinase-II, -IV and arrhythmias. To date, however, no clinical or experimental calcineurin are involved in leukemia inhibitory factor-induced cardiac hypertrophy in rats. Circ Res 2000;87:937–45. studies have examined the biologic effect of these com- 13. Murasawa S, Mori Y, Nozawa Y, et al. Angiotensin II type 1 pounds in cardiomyocytes. The present study has demon- receptor-induced extracellular signal-regulated protein kinase activa- ϩ strated that, when we consider the side effects of KBs, we tion is mediated by Ca2 /calmodulin-dependent transactivation of epidermal growth factor receptor. Circ Res 1998;82:1338–48. should bear in mind not only the electrophysiologic effects, 14. Minden A, Lin A, Karin M, et al. c-Jun N-terminal phospholyration but also the molecular effects, through various signaling correlates with activation of the JNK subgroup but not the ERK pathways. Further studies are required to determine the subgroup of mitogen-activated protein kinases. Mol Cell Biol 1994; 14:6683–8. effect of KBs on the cellular function in cardiomyocytes. 15. Abdelkarim S, Geetha G, Tina MG, Kenneth LB, Allen MS, Pamela AL. Calcium- and protein kinase C-dependent activation of the Reprint requests and correspondence: Dr. Keiichi Fukuda, In- tyrosine kinase PYK2 by angiotensin II in vascular smooth muscle. stitute for Advanced Cardiac Therapeutics, Keio University School Circ Res 1998;83:841–51. 16. Lev S, Moreno H, Martinez R, et al. Protein tyrosine kinase PYK2 of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, 2ϩ involved in Ca -induced regulation of and MAP kinase Japan. E-mail. [email protected]. functions. Nature 1995;376:737–44. 17. Brinson AE, Harding T, Diliberto PA, et al. Regulation of a calcium-dependent tyrosine kinase in vascular smooth muscle cells by REFERENCES angiotensin II and platelet-derived growth factor: dependence on calcium and the actin cytoskeleton. J Biol Chem 1998;273:1711–8. 1. Bouchard RA, Clark RB, Giles WR. Effects of action potential 18. Duong LT, Lakkakorpi PT, Nakamura I, Machwate M, Nagy RM, duration on excitation-contraction coupling in rat ventricular myo- Roden GA. PYK2 in osteoclasts is an adhesion kinase, localized in the ␣ ␤ cytes. Circ Res 1995;76:790–801. sealing zone, activated by ligation of (V) 3 integrin, and phosphory- 2. Doggrell SA, Bishop BE. Effects of potassium channel blockers on the lated by Src kinase. J Clin Invest 1998;102:881–92. JACC Vol. 38, No. 5, 2001 Tahara et al. 1563 November 1, 2001:1554–63 KB-Evoked Signals in Cardiomyocytes

ϩ 19. Rouzaire DB, Dubois JM. K channel block-induced mammalian associated with cardiac muscle cell hypertrophy. J Cell Biol 1994;126: neuroblastoma cell swelling: a possible mechanism to influence prolif- 1565–72. eration. J Physiol (Lond) 1998;510:93–102. 28. Post GR, Goldstein G, Thuerauf DJ, Glembotski CC, Brown JH. 20. Nilius B, Wohlrab W. Potassium channels and regulation of prolifer- Dissociation of p44 and p42 mitogen-activated protein kinase activa- ation of human melanoma cells. J Physiol (Lond) 1992;445:537–48. tion from receptor-induced hypertrophy in neonatal rat ventricular 21. Wang S, Melkoumain Z, Woodfork KA, et al. Evidence for an early myocytes. J Biol Chem 1996;271:8452–7. G1 ionic event necessary for cell cycle progression and survival in the 29. Guo W, Kamiya K, Toyama J. Differential effects of chronic mem- ϩ MCF-7 human breast carcinoma cell line. J Cell Physiol 1998;176: brane depolarization on the K channel activities in cultured rat 456–64. ventricular cells. Cardiovasc Res 1997;33:139–46. 22. Wondergem R, Cregan M, Strikler L, Miller R, Suttle J. Membrane 30. Himmel HM, Wettwer E, Li QI, Ravens U. Four different compo- potassium channels and human bladder tumor cells. II. Growth nents contribute to outward current in rat ventricular myocytes. Am J properties. J Membr Biol 1998;161:257–62. Physiol 1999;277:H107–18. 23. Xu B, Wilson BA, Lu L. Induction of human myeloblastic ML-1 cell ϩ 31. Zwick E, Wallasch C, Daub H, Ullrich A. Distinct calcium- G arrest by suppression of K channel activity. Am J Physiol 1 dependent pathways of epidermal growth factor receptor transactiva- 1996;271:C2037–44. ϩ 24. Lepple WA, Berweck S, Bohmig M, et al. K channels and the tion and PYK2 tyrosine phosphorylation in PC cells. J Biol Chem intracellular calcium signal in human melanoma cell proliferation. J 1999;274:20989–96. Membr Biol 1996;151:149–57. 32. Ghosh A, Greenberg ME. Calcium signaling in neurons: molecular ϩ 25. Wang L, Xu D, Dai W, Lu L. An ultraviolet-activated K channel mechanisms and cellular consequences. Science 1995;268:239–47. mediates apoptosis of myeloblastic leukemia cells. J Biol Chem 33. Bading H, Gindy DD, Greenberg ME. Regulation of gene expression 1999;274:3678–85. in hippocampal neurons by distinct calcium signaling pathways. 26. Chin LS, Park CC, Simard JM, et al. 4-Aminopyridine causes Science 1993;260:181–6. ϩ apoptosis and blocks an outward rectifier K channel in mammalian 34. Nebigil C, Malik KU. Alpha-adrenergic receptor subtypes involved in ϩ astrocytoma cell lines. J Neurosci Res 1997;48:122–7. prostaglandin synthesis are coupled to Ca2 channels through a 27. Thorborn J, Frost JA, Thorborn A. Mitogen-activated protein kinases pertussis toxin-sensitive guanine nucleotide-binding protein. J Phar- mediate changes in gene expression, but not cytoskeletal organization macol Exp Ther 1993;266:1113–24.