Original Article

Src family tyrosine kinases inhibit single L-type:Ca2+channel activity in human atrial myocytes

Frank Schröder a,*,1, Gunnar Klein a,1, Tanja Frank a, Michaela Bastein a, Sylvio Indris b,Matthias Karck c, Helmut Drexler a, Kai C. Wollert a

a Department of Cardiology and Angiology, Hannover Medical School, Carl-Neuberg Strasse 1, Hannover 30625, Germanyb Institute for Physical Chemistry and Electrochemistry, University of Hannover, Callinstrasse 3-3A, Hannover 30167, Germany

c Department of Cardiovascular Surgery, Hannover Medical School, Carl-Neuberg Strasse 1, Hannover 30625, Germany

Received 22 March 2004; received in revised form 1 June 2004; accepted 9 June 2004

Available online 31 July 2004

Abstract

Objective. – Tyrosine kinases (TKs) are important regulators of the L-type Ca2+ channel (LTCC) current in various cell types. However,there are no data addressing the role of TKs in the control of single LTCC activity in human atrial cardiac myocytes, where changes in LTCCgating properties have been described in a number of disease states.

Methods and results. – Single LTCC activity was recorded in isolated human atrial myocytes. The broad-spectrum TK inhibitor genisteinand the Src family-selective TK inhibitor PP1 significantly enhanced single LTCC ensemble average current, availability, and openprobability; the latter was due to significant increases of mean open time and mode 2 gating. Conversely, the tyrosine phosphatase inhibitorbisperoxo—phenanthroline-vanadate inhibited single LTCC activity, indicating that LTCC gating properties in human atrial myocytes arecontrolled by TKs and tyrosine phosphatases in a reciprocal fashion. The effects of genistein on single LTCC activity were not affected bystimulation (8Br-cAMP) or inhibition (Rp-8-CPT-cAMPS) of protein kinase A (PKA) or by inhibition of serine/threonine phosphatases typesI and IIa (okadaic acid), indicating that TKs inhibit LTCC gating in human atrial myocytes independent of PKA and phosphatases types I andIIa. However, inhibition of protein kinase C (PKC) by staurosporine or bisindolylmaleimide reversed the stimulatory effects of genistein onsingle LTCC gating properties, indicating that PKC is required for the inhibitory effect of TKs on single LTCC activity.

Conclusion. – Src family TKs inhibit single LTCC activity in human atrial myocytes via PKC-dependent, but PKA and phosphatase typesI and IIa-independent, molecular pathways.© 2004 Elsevier Ltd. All rights reserved.

Keywords: L-type Ca2+ channel; Tyrosine kinases; Human atrial cardiomyocytes

1. Introduction

Tyrosine kinases (TKs) are important signaling intermedi-ates of several cell surface receptors, such as the fibroblastgrowth factor receptors, the insulin-like growth factor(IGF-1) receptor, or the angiotensin II type 1 receptor [1–4].In many cell types, TKs play critical roles in controllingdiverse aspects of cell differentiation, proliferation, and func-tion [5]. In cardiac myocytes, TKs have been implicated inthe regulation of transmembrane ion fluxes, most notably in

the fine-tuning of the sarcolemmal L-type Ca2+ channel(LTCC) current [6–12].

In human atrial myocytes, changes in LTCC gating prop-erties have been observed in various pathological states,including subclinical hyperthyroidism [13], atrial dilation[14], and atrial fibrillation [15–18]. Since Ca2+ entry via theLTCC is crucial for excitation–contraction coupling and ac-tion potential duration [19,20], alterations in LTCC activityin atrial myocytes have been hypothesized to play a patho-physiological role in the development and persistence ofatrial fibrillation in patients [15–18].

It is important, therefore, to decipher the regulatory cir-cuits controlling LTCC activity in human atrial myocytes.Previous studies addressing the role of TKs in the regulation

1 F.S. and G.K. contribute equally to this work.* Corresponding author. Tel.: +49-511-532-3840;

fax: +49-511-532-5412.E-mail address: schroeder.f@mh-hannover.de (F. Schröder).

Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

www.elsevier.com/locate/yjmcc

© 2004 Elsevier Ltd. All rights reserved.doi:10.1016/j.yjmcc.2004.06.008

of LTCC activity have provided conflicting results, however,and it has emerged that the effects of TKs vary depending onthe species and cardiac myocyte preparation studied [6–12].Moreover, to our knowledge, single LTCC gating propertieshave never been assessed in human atrial myocytes.

We postulated that TKs influence single LTCC gatingproperties in human atrial myocytes by modulating averagecurrent amplitude, availability, and/or open probability. Inthe present study, we therefore employed the cell-attachedsingle channel patch-clamp technique in human atrial myo-cytes obtained from patients undergoing open heart surgery,(1) to examine the role of TKs in controlling single LTCCactivity, (2) to possibly identify subgroups of TKs controllingLTCC activity, and (3) to explore potential cross-talk of TKswith protein phosphatases, protein kinase A (PKA), andprotein kinase C (PKC).

2. Methods

2.1. Cell isolation

This investigation confirms with the principles outlined inthe Declaration of Helsinki and was approved by our Institu-tional Ethics Committee. All patients gave written informedconsent. Atrial myocytes were isolated from 30 patientsundergoing coronary artery bypass grafting, seven patientsundergoing aortic valve replacement, and eight patients re-ceiving combined bypass surgery and aortic valve replace-ment. All patients were in sinus rhythm at the time of theoperation. In all patients (32 male, 13 female), left ventricu-lar ejection fraction (LV-EF) was ≥ 35% as shown byechocardiography and/or angiography. Mean LV-EF was62 ± 11%. The mean age was 67 ± 11 years. 49% of thepatients were treated with angiotensin converting enzyme(ACE) inhibitors or AT1-receptor blockers, 58% withb-blockers, 31% with diuretics, and 40% with nitrates. Thesedrugs were not discontinued before surgery. Data stratifica-tion according to medication did not reveal any significanteffect of these drugs on LTCC activity (Table 1). Immedi-ately after surgical excision, a piece of right atrial tissue wasplaced in oxygenated solution A at 4 °C, containing (in mM):100 NaCl, 10 KCl, 5 MgSO4, 20 dextrose, 50 taurine, andMOPS (pH 7.4), and transferred to the laboratory within15 min. After removal of connective tissue and fat, atrialtissue was chopped into small pieces and digested for 40 minat 37 °C in solution A in the presence of collagenase(1.5 mg/ml, type CLS1, Worthington Biochemical), protease

XXIV (0.5 mg/ml, Sigma), and BSA (10 mg/ml, Sigma).After removal of the supernatant, the tissue was washed onceand incubated for 5–10 min in solution A containing collage-nase (0.5 mg/ml) and BSA (1 mg/ml). Cells were then recov-ered by centrifugation and stored at room temperature insolution B containing (in mM): 40 KCl, 20 KH2PO4,3 MgCl2, 20 KOH, 50 potassium glutamate, 20 taurine,5 EGTA, 10 dextrose, and 10 HEPES (pH 7.4).

2.2. Measurements of barium currents through singleLTCC

Cells were placed in disposable perfusion chambers filledwith 1.5–2 ml of bath solution containing (in mM): 25 KCl,2 MgCl2, 120 potassium glutamate, 1 CaCl2, 1 ATP, 2 EGTA,10 dextrose, and 10 HEPES (pH 7.4). Single channel record-ings were performed in the cell-attached configuration aspreviously described by our group [21–23] using borosilicateglass pipettes (7–10 MX) containing (in mM) 70 BaCl2,110 sucrose, 10 HEPES (pH 7.4). The cell-attached patch-clamp configuration preserves cellular integrity, eliminatespotential effects of the pipette solution on the intracellularmilieu, and provides detailed insight into LTCC gating pa-rameters [23]. Ba2+ was used as charge carrier to avoidCa2+-induced LTCC current inactivation. Ba2+ currents wereelicited at 1.66 Hz by 150 ms depolarizing command pulsesfrom –100 mV to +20 mV. Data acquisition was achieved atroom temperature at 10 kHz and filtered at 2 kHz (–3 dB,8-pole Bessel) using an Axopatch 200B amplifier andpClamp version 6.0 software (Axon Instruments, Foster City,CA). On average, LTCCs (identified by their typical unitaryconductance and voltage dependence of activation) werefound in one out of 10 patches. This success rate is due to thesmall pipettes used and probably to a low density of activeLTCC in the atrial myocytes cell membrane.

2.3. Reagents

Bisindolylmaleimide, bisperoxo-phenanthroline-vana-date (bp-vanadate), okadaic acid (NH4 salt), phorbol 12-myristate 13-acetate (PMA), 4-amino-5-(4-methylphenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (PP1) and 4-amino-7-phenylpyrazol[3,4-d]pyrimidine (PP3) were purchasedfrom Calbiochem. 8-Bromo cyclic AMP (8Br-cAMP),8-(4-chlorophenylthio)adenosine-3′,5′-cyclic monophos-phorothioate, RP-Isomer (Rp-8-CPT-cAMPS), genistein andstaurosporine were purchased from Sigma. All reagents weredissolved in water or DMSO. The concentration of DMSO in

Table 1Influence of various drugs on basal single LTCC gating properties

ACEI/ARB No ACEI/ARB b-blocker No b-blocker Nitrates No nitratesOpen probability (%) 6.1 ± 1.1 6.0 ± 0.8 5.7 ± 0.8 6.6 ± 1.1 6.0 ± 1.1 6.1 ± 0.8Availability (%) 45 ± 4 46 ± 3 44 ± 3 49 ± 4 49 ± 4 43 ± 3

ACEI denotes ACE-inhibitor; ARB, AT1-receptor blocker. P = n.s. using an unpaired t-test. ACEI/ARB vs. no ACEI/ARB; b-blocker vs. no b-blocker; nitratesvs. no nitrates.

736 F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

the bath solution did not exceed 1.1%. In a set of six experi-ments, we did not observe a significant effect of 1.1% DMSOon single LTCC activity (mean open probability, control:5.1 ± 1.4%, DMSO 5.1 ± 1.5%; availability, control:67.0 ± 8.5%, DMSO, 66.3 ± 8.5%). Genistein, PP1, PP3,bp-vanadate, and 8-Br-cAMP were added to the bath as abolus in order to obtain an optimal signal-noise ratio. At thecellular level, the concentrations of these drugs are thereforenot exactly determinable. However, we and other groups[24–27] have always observed a low variance of effect onLTCC channel gating. Therefore, we consider that drug con-centrations at the cellular level do not differ biologicallysignificant in the individual experiments. Furthermore, usinga diffusion model [28] we estimated the time a substanceneeds to reach the target concentration at the cellular level,which is in the range of minutes.

2.4. Data analysis

Linear leak and capacity currents were digitally sub-tracted using averaged currents of non-active sweeps. LTCCpeak ensemble average current, availability (fraction ofsweeps containing at least one channel opening), and meanopen probability (fraction of time spent in the open stateduring active sweeps) were corrected for the number ofchannels in the patch (n), which was estimated by dividingmaximum current amplitude by the unitary channel currentof the same channel in the same depolarizing pulse. A typicalrecording of a patch containing two channels is shown inFig. 1. Only single or double channel patches were analyzed.Peak average current was normalized by division through n.The availability was corrected by the square root method:(1 – availabilitycorr) is the nth root of (1 – availabilityuncorr),where availabilitycorr and availabilityuncorr denote correctedand uncorrected availability, respectively. Corrected openprobability was calculated on the basis of the corrected num-ber of active sweeps, i.e. total open time (in ms) within allsweeps of the ensemble current, divided by (150 ms × n ×availabilitycorr × number of test pulses). Openings and clo-sures were identified by the half-height criterion. Mean opentime represents the mean duration of single channel openingswithin a burst and is calculated from the total open time of arecording divided by the sum of the number of closures andthe number of active sweeps. Mean close time was calculated

from total close time divided by the number of closures. Firstlatency is characterized as the time from the beginning of apulse to the beginning of the first channel opening. Inactiva-tion is defined as the decrease of the ensemble channelcurrent from its peak to the end of the test pulse. Burst lengthis defined as the interval between the beginning of the firstopening and the end of the last opening during an activesweep. Since the shape of ensemble average currents variedto some extent, the area under curve (AUC) was estimatedand presented as (%) of change of the ensemble averagecurrents after drug application. Mode 2 gating was defined aschannel openings lasting longer than 4 ms.

At least 180 sweeps were recorded from the very samechannels at baseline and during steady-state conditions afterapplication of various compounds. All data are presented asmean ± SEM. Differences were analyzed by a paired t-test orby one-way ANOVA followed by Student’s t-test with Bon-ferroni correction, when appropriate. A two-tailed Pvalue < 0.05 was considered statistically significant.

3. Results

3.1. Reciprocal regulation of single LTCC activity by TKsand tyrosine phosphatases

To study the effects of TKs on single LTCC gating prop-erties in human atrial myocytes, single channel recordingswere performed in the absence or presence of genistein(5 ×10–5 M), a broad-spectrum tyrosine kinase inhibitor [5].As shown in Fig. 2A–C, genistein promoted a significantincrease in peak ensemble average LTCC current (from18 ± 6 fA to 30 ± 7 fA, change of AUC +168 ± 33%) byaugmenting single channel availability (from 31 ± 7% to43 ± 5%), and open probability (from 5.0 ± 1.4% to7.6 ± 2.0%). The increase in LTCC open probability was dueto an increase in mean open time (from 0.61 ± 0.06 ms to0.82 ± 0.13 ms) and mode 2 gating (from 13 ± 4% to18 ± 5%) (Fig. 2C). In another set of experiments, singleLTCC gating properties were assessed after application ofbp-vanadate (10–6 M), a tyrosine phosphatase inhibitor [29].Bp-vanadate significantly decreased single channel en-semble average current (from 30 ± 5% to 13 ± 3%, change of

Fig. 1. A patch containing two channels. A duplication of the unitary channel current is observed when both channels are in the open state at the same time.

737F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

AUC –63 ± 7%), availability (from 53 ± 8% to 36 ± 7%), andopen probability (from 6.6 ± 1.8% to 3.9 ± 1.6%) (Fig. 2D).These observations indicate that TKs inhibit whereas ty-rosine phosphatases enhance single LTCC activity in humanatrial myocytes.

3.2. Inhibition of single LTCC activity by Src family TKs

Since genistein inhibits soluble and membrane-boundTKs, we investigated the effects of PP1 (10–5 M), a selectiveinhibitor of Src family TKs [30], on single LTCC activity in

Fig. 2. Effects of the tyrosine kinase inhibitor genistein and the tyrosine phosphatase inhibitor bp-vanadate. (A) Representative consecutive traces from onesingle LTCC recording before (left) and after application of 5 × 10—5 M genistein (right). Ensemble average currents are depicted in the bottom row. Scale barsrepresent 20 ms and 2 pA (unitary current traces) or 20 fA (ensemble averages). (B) Exemplary single channel recording illustrating the time course of thegenistein effect. (C) Effects of genistein (Gen) on ensemble average current, availability, open probability, mean open time, and occurrence of mode 2 gating(defined as openings >4 ms within a sweep). (D) Effects of bp-vanadate (bpV, 10–6 M) on ensemble average current, availability, and open probability. Valuesare mean ± SEM from n = 7–10 experiments (paired t-test, *P < 0.05 vs. baseline).

738 F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

human atrial myocytes. As shown in Fig. 3A–C, PP1 (similarto genistein) enhanced single LTCC peak ensemble averagecurrent (from 14 ± 4 fA to 32 ± 6 fA, change of AUC+263 ± 44%), availability (from 56 ± 6% to 89 ± 3%), andopen probability (from 2.1 ± 0.9% to 4.0 ± 1.0%). Likegenistein, PP1 also significantly increased mean open time

(from 0.43 ± 0.04 to 0.61 ± 0.05 ms) and mode 2 gating (from1.7 ± 1.4% to 5.0 ± 0.8%) (Fig. 3C). By contrast, PP3, aninactive PP1 analog, had no discernible effects on singleLTCC gating properties (peak ensemble average currentfrom 26 ± 7 fA to 31 ± 9 fA, change of AUC +16 ± 31%,availability from 48 ± 9% to 51 ± 8%, and open probability

Fig. 3. Effects of the Src-specific tyrosine kinase inhibitor PP1. (A) Representative consecutive traces from a single LTCC recording before (left) and afterapplication of 10–5 M PP1 (right). Ensemble average currents are depicted in the bottom row. Scale bars represent 20 ms and 2 pA (unitary current traces) or20 fA (ensemble averages). (B) Exemplary single channel recording illustrating the time course of the PP1 effect. (C) Effects of PP1 on ensemble averagecurrent, availability, open probability, mean open time, and occurrence of mode 2 gating. (D) Effects of PP3 (10–5 M) on ensemble average current, availability,and open probability. Values are mean ± SEM from n = 7–8 experiments (paired t-test, *P < 0.05 vs. baseline).

739F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

from 6.5 ± 2.4% to 6.3 ± 2.1%) (Fig. 3D). Taken together,these data indicate that single LTCC activity in human atrialmyocytes is inhibited by Src family TKs. Table 2 compiles allsingle channel parameter as influenced by genistein and PP1.

3.3. Lack of interaction between TKs and PKA

The effects of genistein and PP1 on single LTCC gatingproperties in human atrial myocytes very much resemble theeffects observed after activation of PKA [21,22,27,31,32],raising the possibility that TKs interfere with PKA signalingpathways controlling the LTCC. We therefore analyzedwhether the effects of genistein on single LTCC activity aremodulated by inhibition (Rp-8-CPT-cAMPS) or activation(8Br-cAMP) of PKA. As shown in Fig. 4A, genistein signifi-cantly increased single LTCC peak ensemble average current(from 19 ± 8 fA to 38 ± 10 fA, change of AUC +94 ± 10%),availability (from 48 ± 11% to 60 ± 8%), and open probabil-ity (from 3.3 ± 1.0% to 6.5 ± 1.4%), in atrial myocytespre-incubated with Rp-8-CPT-cAMPS (10–5 M), suggestingthat inhibition of PKA does not contribute significantly to thesuppression of LTCC activity by TKs. As expected, 8Br-cAMP (10–3 M) significantly increased single LTCC peakensemble average current (from 26 ± 7 fA to 40 ± 12 fA,change of AUC 94 ± 10% from baseline), availability (from32 ± 5% to 47 ± 9%), and open probability (from 6.7 ± 2.0%to 8.6 ± 2.4%) (Fig. 4B,C). Subsequent application ofgenistein further increased LTCC activity (peak ensembleaverage current: 64 ± 9 fA, change of AUC from baseline+249 ± 22%, availability: 61 ± 3%, open probability:12 ± 3%) (Fig. 4B,C), providing additional support to thenotion that TKs modulate LTCC gating properties via PKA-independent signaling pathways.

3.4. TKs inhibit single LTCC activity independent fromserine/threonine phosphatases

In guinea pig cardiac ventricular myocytes, inhibition ofserine/threonine phosphatases leads to an activation of singleLTCC gating properties that is reminiscent of the effectsobserved with genistein in our system [33]. It is conceivable,therefore, that TKs inhibit LTCC activity by activating

serine/threonine phosphatases. To test for such an interac-tion, we analyzed the effects of genistein in human atrialmyocytes in the presence of okadaic acid, at a concentration(10–6 M) known to inhibit serine/threonine phosphatasestypes I and IIa [34]. As shown in Fig. 5, genistein signifi-cantly increased single LTCC peak ensemble average current(from 43 ±16 fA to 73 ± 17 fA, AUC +109 ± 35%), availabil-ity (from 52 ± 7% to 78 ± 3%), open probability (from10 ± 2% to 12 ± 2%), even in the presence of okadaic acidthus providing evidence against a direct interaction betweenTKs and tyrosine phosphatases.

3.5. PKC is required for TK inhibition of single LTCCactivity

It has recently been suggested that TKs and PKC cooper-ate in the regulation of whole-cell LTCC current in humanatrial myocytes [10]. To assess the role of PKC in the modu-lation of single LTCC gating properties by TKs in humanatrial myocytes, we analyzed whether the effects of genisteinon single LTCC activity are affected by activation (PMA) orinhibition (staurosporine and bisindolylmaleimide) of PKC(Fig. 6A,B). In human atrial myocytes where PKC was acti-vated by PMA (10–5 M), genistein still increased peak en-semble average channel current (from 20 ± 6 fA to44 ± 10 fA, change of AUC +271 ± 86%), availability (from43 ± 10% to 68 ± 8%) and open probability (from 2.8 ± 0.8%to 5.9 ± 1.8%). Remarkably, however, in the presence of10—7 M staurosporine, genistein no longer increased singleLTCC ensemble average current, availability, and open prob-ability, indicating that PKC is required for TK inhibition ofsingle LTCC gating properties in human atrial myocytes(values are before vs. after genistein in the presence ofstaurosporine: peak ensemble average channel current:39 ± 6 fA vs. 29 ± 8 fA, change of AUC –9 ± 40%; availabil-ity: 48 ± 8% vs. 36 ± 3%; open probability: before 9.2 ± 2.2%after genistein 7.7 ± 2.3%; P = n.s. for all comparisons).Similarly, genistein no longer increased single LTCC gatingproperties in the presence of 10–6 M bisindolylmaleimide(values are before vs. after genistein in the presence ofbisindolylmaleimide: peak ensemble average channel cur-rent: 47 ± 21 fA vs. 38 ± 21 fA, change of AUC –10 ± 25%;

Table 2Influence of genistein and PP1 on single LTCC gating properties

Parameter Control Genistein Control PP1Peak average current (fA) 18 ± 6 30 ± 6* 14 ± 3 32 ± 6*Open probability (%) 5 ± 1.4 7.6 ± 2.0* 2.1 ± 0.9 4 ± 1.0*Availability (%) 31 ± 7 43 ± 5* 56 ± 6 89 ± 3*Mean open time (ms) 0.6 ± 0.06 0.82 ± 0.13* 0.43 ± 0.04 0.61 ± 0.05*Mode 2 gating (%) 13.3 ± 3.8 17.5 ± 4.8* 1.7 ± 1.4 5 ± 0.8*Mean closed time (ms) 8.0 ± 2.0 7.5 ± 1.7 13.4 ± 4.6 8.4 ± 3.3Mean first latency (ms) 33.1 ± 4.8 26.7 ± 3.5 36.7 ± 7.1 17 ± 2.8Mean inactivation (%) 38.6 ± 8.8 41.8 ± 9.5 47.4 ± 6.5 37.7 ± 6.6Mean burst length (ms) 60.1 ± 13.3 63.7 ± 9.9 81 ± 15.4 98.2 ± 2.7

Definition and analyses of gating parameters are described in the Methods section. Values are mean ± SEM from n = 7–10 experiments (paired t-test,*P < 0.05 vs. baseline).

740 F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

Fig. 4. Effects of genistein in the presence of PKA modulators. (A) Effects of genistein (5 × 10—5 M) on ensemble average current, availability, and openprobability in human atrial myocytes pre-incubated with Rp-8-CPT-cAMPS (10—5 M). (B) Single consecutive traces from one representative single LTCCrecording under control conditions (left), after application of 8Br-cAMP (10–3 M) (middle), and after subsequent addition of 5 × 10–5 M genistein (right).Ensemble average currents are depicted at the bottom. Scale bars represent 20 ms and 2 pA (unitary current traces) or 50 fA (ensemble averages). (C) Bar graphsdepict single LTCC gating properties under control conditions (blank columns), after application of 10–3 M 8Br-cAMP (8Br, gray columns), and after additionaltreatment with 5 × 10–5 M genistein (8Br + Gen, black columns). Values are mean ± SEM from n = 10 experiments (differences were analyzed by one-wayANOVA followed by Student’s t-test with Bonferroni correction, *P < 0.05 vs. control; §P < 0.05 vs. 8Br-cAMP alone).

741F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

availability: 60 ± 7% vs. 55 ± 7%; open probability:7.6 ± 2.7% vs. 6.2 ± 2.8%; P = n.s. for all comparisons).

4. Discussion

The LTCC plays a fundamental role in excitation–contrac-tion coupling and in the activation of Ca2+-dependent signal-ing pathways affecting growth and survival in cardiac myo-cytes [35,36]. Not surprisingly, then, LTCC gating propertiesare tightly controlled by complex networks of protein kinasesand phosphatases [21,22,33,34,37,38]. In this regard, TKshave been recognized as important regulators of the LTCCcurrent in various cell types [6–12]. Previous studies address-ing the role of TKs in the control of LTCC activity in cardiacmyocytes have provided conflicting results, however, and ithas emerged that the effects of TKs vary depending on thespecies, the cardiac myocyte preparation (atrial vs. ventricu-lar) and the methodology used (perforated vs. rupturedpatch) [6–12]. Based on these previous data obtained bywhole-cell patch-clamp techniques, it was impossible, there-fore, to predict the role of TKs in controlling single LTCCactivity in human atrial myocytes.

Data provided in the present study indicate for the firsttime that Src family TKs inhibit single LTCC gating proper-ties in human atrial myocytes. Two pharmacological inhibi-tors were used in this study, genistein, a broad-spectrum TKinhibitor and PP1, which predominantly inhibits the Srcfamily of TKs [30]. Although “non-specific” effects ofgenistein and PP1 on LTCC activity cannot be formallyexcluded, we feel that is very unlikely, since both agentspromoted comparable effects in our system and PP3, a struc-turally related, inactive PP1 analog, did not influence LTCCactivity.

Decreases in LTCC activity by TK are mediated by acombination of inhibitory effects on single LTCC availabilityand open probability. Increases in single LTCC activity aftertreatment with genistein or PP1 were due to increases ofavailability and open probability. The latter was mediated by

a prolongation of open times and enhancement of mode2 gating. Of note, mode 2 gating is rarely observed in humanventricular cardiomyocytes [23], indicating a considerabledifference in the fine-adjustment of single LTCC gating prop-erties between atrial and ventricular tissues. The tyrosinephosphatase inhibitor bp-vanadate decreased single LTCCactivity in our study, supporting the concept of a reciprocalregulation of LTCC activity in human atrial myocytes by Srcfamily TKs and tyrosine phosphatases. Based on the obser-vation that incubation with genistein decreases whole-cellLTCC currents, it has been proposed that TKs increase LTCCactivity [7–9]. However, these previous reports need to beviewed with some caution, since genistein can inhibit theLTCC by directly binding to its extracellular surface, whichmay then override any inhibitory effects of this drug onintracellular TKs [39,40]. Importantly, the cell-attachedpatch-clamp configuration used in the present study preventsaccess of test compounds to the LTCC from the extracellularspace.

What are the molecular mechanisms whereby Src familyTKs regulate single LTCC gating in human atrial myocytes?In neurons, TKs have been shown to modulate the LTCCcurrent by phosphorylation of critical tyrosine residues thatare present, e.g., in the a1-subunit of the channel [41]. Con-ceptually, such a direct regulation of LTCC gating propertiescould also take place in (atrial) cardiac myocytes. In addition,TKs may interfere with certain signaling pathways involvedin the regulation of the LTCC current. In this regard, thestimulatory effects of genistein and PP1 on single LTCCgating properties in our study were similar to the effects seenafter activation of PKA [21,22,27,32] or inhibition ofserine/threonine-phosphatases [33], raising the possibilitythat Src family TKs inhibit LTCC activity in human atrialmyocytes by inhibiting PKA and/or activating protein phos-phatases. However, the effects of genistein were not affectedby concomitant stimulation or inhibition of PKA or by inhi-bition of phosphatases types I and IIa. Our data thereforeindicate that TKs inhibit single LTCC gating in human atrialmyocytes via PKA/phosphatase types I- and IIa-independentsignaling pathways. Certainly, these results do not excludethe possibility that TKs regulate b-adrenergic responsivenessupstream of PKA [11,12].

The PKC inhibitors staurosporine and bisindolylmaleim-ide abolished the stimulatory effects of genistein on singleLTCC gating properties in our study, indicating that PKC isrequired for TK inhibition of single LTCC gating. Our studywas not designed to elucidate where exactly this cross-talkmay appear. Recent reports suggest a complex interrelation-ship between the Src family of TKs and PKC [42,43]. PKCefor instance, has been reported to assemble with the Srcfamily member Lck leading to its phosphorylation and acti-vation [42]. On the other hand, certain PKC isoforms havebeen shown to be phosphorylated on tyrosine residues; e.g.,activation of PKCd depends upon Src kinase activation [43].In cardiac myocytes, Src kinases may assemble with focaladhesion kinase (FAK) as part of a multiprotein signaling

Fig. 5. Effects of genistein in the presence of the phosphatase inhibitorokadaic acid. Effects of genistein (Gen, 5 × 10—5 M) on ensemble averagecurrent, availability, and open probability in human atrial myocytes pre-incubated with okadaic acid (10—6 M). Values are mean ± SEM fromn = 7 experiments (paired t-test, *P < 0.05 vs. okadaic acid alone).

742 F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

Fig. 6. Effects of genistein in the presence of PKC modulators. (A) Consecutive traces from representative single LTCC recordings. Left: atrial myocytes werefirst incubated with the PKC activator PMA (10—5 M); subsequently, genistein (5 × 10–5 M) was added. Right: myocytes were incubated with the PKC inhibitorstaurosporine (10—7 M); subsequently, genistein (5 × 10—5 M) was added. Ensemble average currents are shown at the bottom. Scale bars represent 20 ms and2 pA (unitary current traces) or 20 fA (ensemble averages). (B) Effects of genistein on single LTCC gating properties. Genistein (Gen, 5 × 10–5 M) was appliedto cardiac myocytes pre-incubated with PMA (10—5 M), staurosporine (10—7 M), or bisindolylmaleimide (10—6 M). Staurosporine and bisindolylmaleimideprevented genistein-induced increases in ensemble average current, availability, and open probability. Values are mean ± SEM from n = 6–7 experiments (pairedt-test, *P < 0.05 vs. baseline).

743F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

complex [44]. In colonic smooth muscle cells, the FAK/Srckinase signaling complex has been implicated in the regula-tion of the LTCC [45]. Additionally, in cardiac myocytes,PKC may activate FAK [46]. Future studies should explorethe interactions of Src family TKs, PKC, and FAK on amolecular basis and decipher their role in the regulation ofthe LTCC in cardiac myocytes.

Limitations of the study: We cannot exclude that theagents used in our study exert “non-specific” effects beyondTK inhibition or PKC activation/inhibition. However, to ad-dress this problem, we used a panel of activators and inhibi-tors in our study. For example, genistein (broad-spectrum TKinhibitor), bp-vanadate (tyrosine phosphatase inhibitor), PP1(Src family tyrosine kinase inhibitor), and PP3 (inactive PP1analog) were used to explore the role of TKs, whereas PMA(PKC activator), staurosporine, and bisindolylmaleimide(PKC inhibitors) were employed to study the role of PKC.Importantly, experimental data obtained with these variousagents were congruent, which makes it very unlikely thatnon-specific effects explain our findings. Genistein in theconcentration used in our study is a well-known TK inhibitor[47], nevertheless, it may induce cytoskeletal changes, be-cause enzymes involved in cytoskeletal remodeling (e.g.FAK) are regulated by TKs [45]. We cannot rule out thatcytoskeletal changes contribute to the effects of TK inhibi-tors on LTCC gating properties. The fast time course of theeffect (within 2–3 min) argues against this hypothesis andsuggests a regulation at the level of the LTCC (directly or viaa regulatory protein). In the concentration used, PP1 is apotent inhibitor of the Src family of TKs, however, theactivity of other TKs, e.g. the EGF-R, is also reduced [30].Importantly, PP3, a structurally related but inactive PP1 ana-log, did not exert any significant effects of LTCC gatingproperties. PMA has been widely used as a PKC activator[48,49]. PMA also induces the expression of iNOS andactivates Ca2+-ATPase in rat liver [50]. Ca2+-ATPase activa-tion may interfere with intracellular Ca2+ handling by influ-encing Ca2+-dependent inactivation of the LTCC. However,in our study, cells were incubated with BAPTA-AM, a chela-tor of intracellular Ca2+. Furthermore, Ba2+ was used as acharge carrier of the LTCC. Finally, the effects of TK inhibi-tion were measured in the very same channel in cells pre-incubated with PMA. Thus, any effects of PMA on Ca2+-ATPase activity and intracellular Ca2+ homeostasis would beexpected to influence baseline (control) recordings as well asrecordings after TK inhibition. In rat astrocytes, staurospo-rine has been reported to destabilize the actin/myosin-basedcytoskeleton [51]. It is conceivable that this may reduceLTCC activity and diminish b-adrenergic LTCC stimulation[52]. However, in a recent study in human atrial myocytes,staurosporine did not reduce the stimulatory effect of isopro-terenol on the LTCC current [10]. In addition to staurospo-rine, we have used bisindolylmaleimide as another PKCinhibitor. In the concentration used, bisindolylmaleimide is apotent inhibitor of various subtypes of PKC [53]. Bisindolyl-maleimide does not inhibit TKs but may have an inhibitory

effect on PKA [54]. However, as shown in or study, PKAdoes not modulate the effects of TKs.

Acknowledgements

This work was supported by the Deutsche Forschungsge-meinschaft (Schr 719-1).

References

[1] Dell’Era P, Ronca R, Coco L, Nicoli S, Metra M, Presta M. Fibroblastgrowth factor receptor-1 is essential for in vitro cardiomyocyte devel-opment. Circ Res 2003;93:414–20.

[2] Detillieux KA, Sheikh F, Kardami E, Cattini PA. Biological activitiesof fibroblast growth factor-2 in the adult myocardium. Cardiovasc Res2003;57:8–19.

[3] Giovannone B, Lee E, Laviola L, Giorgino F, Clevel KA, Smith RJ.Two novel proteins that are linked to insulin-like growth factor (IGF-I)receptors by the Grb10 adapter and modulate IGF-I signaling. J BiolChem 2003;278:1564–73.

[4] Yin G, Yan C, Berk BC. Angiotensin II signaling pathways mediatedby tyrosine kinases. Int J Biochem Cell Biol 2003;35:780–3.

[5] Levitzki A, Gazit A. Tyrosine kinase inhibition: an approach to drugdevelopment. Science 1995;267:1782–8.

[6] Wang YG, Lipsius SL. Genistein elicits biphasic effects on L-typeCa2+ current in feline atrial myocytes. Am J Physiol 1998;275:H204–12.

[7] Yokoshiki H, Sumii K, Sperelakis N. Inhibition of L-type calciumcurrent in rat ventricular cells by the tyrosine kinase inhibitor,genistein and its inactive analog, daidzein. J Mol Cell Cardiol 1996;28:807–14.

[8] Katsube Y, Yokoshiki H, Nguyen L, Yamamoto M, Sperelakis N.Inhibition of Ca2+ current in neonatal and adult rat ventricular myo-cytes by the tyrosine kinase inhibitor, genistein. Eur J Pharmacol1998;345:309–14.

[9] Ogura T, Shuba LM, McDonald TF. L-type Ca2+ current in guinea pigventricular myocytes treated with modulators of tyrosine phospho-rylation. Am J Physiol 1999;276:H1724–33.

[10] Boixel C, Tessier S, Pansard Y, Lang-Lazdunski L, Mercadier JJ,Hatem SN. Tyrosine kinase and protein kinase C regulate L-type Ca2+

current cooperatively in human atrial myocytes. Am J Physiol 2000;278:H670–6.

[11] Hool LC, Middleton LM, Harvey RD. Genistein increases the sensi-tivity of cardiac ion channels to b-adrenergic receptor stimulation.Circ Res 1998;83:33–42.

[12] Belevych AE, Nulton-Persson A, Sims C, Harvey RD. Role oftyrosine kinase activity in a-adrenergic inhibition of theb-adrenergically regulated L-type Ca2+ current in guinea-pig ventri-cular myocytes. J Physiol 2001;537:779–92.

[13] Kreuzberg U, Theissen P, Schicha H, Schröder F, Mehlhorn U, deVivie ER, et al. Single-channel activity and expression of atrial L-typeCa2+ channels in patients with latent hyperthyroidism. Am J Physiol2000;278:H723–30.

[14] Le Grand B, Hatem S, Deroubaix E, Couetil JP, Coraboeuf E.Depressed transient outward and calcium currents in dilated humanatria. Cardiovasc Res 1994;28:548–56.

[15] Bosch RF, Zeng X, Grammer JB, Popovic K, Mewis C, Kühlkamp V.Ionic mechanisms of electrical remodeling in human atrial fibrillation.Cardiovasc Res 1999;44:121–31.

[16] Nattel S. Ionic determinants of atrial fibrillation and Ca2+ channelabnormalities. Cause, consequence, or innocent bystander? Circ Res1999;85:473–6.

744 F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745

[17] Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM,Nerbonne JM. Atrial L-type Ca2+ currents and human atrial fibrilla-tion. Circ Res 1999;85:428–36.

[18] Klein G, Schröder F, Vogler D, Schaefer A, Haverich A, Schief-fer B, et al. Increased open probability of single cardiac L-typecalcium channels in patients with chronic atrial fibrillation: role ofphosphatase 2A. Cardiovasc Res 2003;59:37–45.

[19] Hatem SN, Benardeau A, Rücker-Martin C, Marty I, de Chamisso P,Villaz M, et al. Different compartments of sarcoplasmic reticulumparticipate in the excitation–contraction coupling process in humanatrial myocytes. Circ Res 1997;80:345–53.

[20] Li GR, Nattel S. Properties of human atrial ICa at physiologicaltemperatures and relevance to action potential. Am J Physiol 1997;272:H227–35.

[21] Klein G, Drexler H, Schröder F. Protein kinase G reverses all isopro-terenol induced changes of cardiac single L-type calcium channelgating. Cardiovasc Res 2000;48:367–74.

[22] Schröder F, Klein G, Fiedler B, Bastein M, Schnasse N, Hill-mer A, et al. Effect of cyclic GMP-dependent protein kinase type I(PKG I) on cardiac single L-type Ca2+ channel activity from a novelPKG I transgenic mouse. Cardiovasc Res 2003;60:268–77.

[23] Schröder F, Handrock R, Beuckelmann DJ, Hirt S, Hullin R,Priebe L, et al. Increased availability and open probability of singleL-type calcium channels from failing compared with nonfailinghuman ventricle. Circulation 1998;98:969–76.

[24] Groner F, Rubio M, Schulte-Euler P, Matthes J, Khan IF, Bodi I, et al.Single-channel gating and regulation of human L-type calcium chan-nels in cardiomyocytes of transgenic mice. Biochem Biophys ResCommun 2004;314:878–84.

[25] Matthes J, Jager A, Handrock R, Groner F, Mehlhorn U,Schwinger RH, et al. Ca2+-dependent modulation of single humancardiac L-type calcium channels by the calcineurin inhibitor cyclos-porine. J Mol Cell Cardiol 2004;36:241–55.

[26] Chen-IzuY, Xiao RP, Izu LT, Cheng H, Kuschel M, Spurgeon H, et al.G(i)-dependent localization of beta(2)-adrenergic receptor signalingto L-type Ca(2+) channels. Biophys J 2000;79:2547–56.

[27] Yue DT, Herzig S, Marban E. Beta-adrenergic stimulation of calciumchannels occurs by potentiation of high-activity gating modes. ProcNatl Acad Sci USA 1990;87:753–7.

[28] Crank J. The mathematics of diffusion. 2nd ed. Oxford: ClarendonPress; 1975.

[29] Posner BI, Faure R, Burgess JW, Bevan AP, Lachance D, Zhang-Sun G, et al. Peroxovanadium compounds. A new class of potentphosphotyrosine phosphatase inhibitors which are insulin mimetics. JBiol Chem 1994;269:4596–604.

[30] Hanke JH, Gardner JP, Dow RL, Changelian PS, Brissette WH,Weringer EJ, et al. Discovery of a novel, potent, and Src family-selective tyrosine kinase inhibitor. J Biol Chem 1996;271:695–701.

[31] Yue DT, Herzig S, Marban E. b-Adrenergic stimulation of calciumchannels occurs by potentiation of high-activity gating modes. ProcNatl Acad Sci USA 1990;87:753–7.

[32] Schröder F, Herzig S. Effects of b2-adrenergic stimulation on single-channel gating of rat cardiac L-type calcium channels. Am J Physiol1999;276:H834–43.

[33] Wiechen K, Yue DT, Herzig S. Two distinct functional effects ofprotein phosphatase inhibitors on guinea-pig cardiac L-type Ca2+

channels. J Physiol 1995;484:583–92.[34] Herzig S, Meier A, Pfeiffer M, Neumann J. Stimulation of protein

phosphatases as a mechanism of the muscarinic-receptor-mediatedinhibition of cardiac L-type Ca2+ channels. Pflügers Arch 1995;429:531–8.

[35] Frey N, McKinsey TA, Olson EN. Decoding calcium signals involvedin cardiac growth and function. Nat Med 2000;6:1221–7.

[36] Fiedler B, Lohmann SM, Smolenski A, Linnemuller S, Pieske B,Schröder F, et al. Inhibition of calcineurin-NFAT hypertrophy signal-ing by cGMP-dependent protein kinase type I in cardiac myocytes.Proc Natl Acad Sci USA 2002;99:11363–8.

[37] Kamp TJ, Hell JW. Regulation of cardiac L-type calcium channels byprotein kinase A and protein kinase C. Circ Res 2000;87:1095–102.

[38] Sims C, Chiu J, Harvey RD. Tyrosine phosphatase inhibitors selec-tively antagonize b-adrenergic receptor-dependent regulation of car-diac ion channels. Mol Pharmacol 2000;58:1213–21.

[39] Chiang CE, Chen SA, Chang MS, Lin CI, Luk HN. Genistein directlyinhibits L-type calcium currents but potentiates cAMP-dependentchloride currents in cardiomyocytes. Biochem Biophys Res Com1996;223:598–603.

[40] Belevych AE, Warrier S, Harvey RD. Genistein inhibits cardiacL-type Ca2+ channel activity by a tyrosine kinase-independent mecha-nism. Mol Pharmacol 2002;62:554–65.

[41] Bence-Hanulec KK, Marshall J, Blair LA. Potentiation of neuronal Lcalcium channels by IGF-1 requires phosphorylation of the a1 subuniton a specific tyrosine residue. Neuron 2000;27:121–31.

[42] Ping P, Song C, Zhang J, Guo Y, Cao X, Li RC, et al. Formation ofprotein kinase C(epsilon)-Lck signaling modules confers cardiopro-tection. J Clin Invest 2002;109:499–507.

[43] Crosby D, Poole AW. Physical and functional interaction betweenprotein kinase C delta and Fyn tyrosine kinase in human platelets. JBiol Chem 2003;278:24533–41.

[44] Torsoni AS, Constancio SS, Nadruz Jr W, Hanks SK, Franchini KG.Focal adhesion kinase is activated and mediates the early hypertrophicresponse to stretch in cardiac myocytes. Circ Res 2003;93:140–7.

[45] Hu XQ, Singh N, Mukhopadhyay D, Akbarali HI. Modulation ofvoltage-dependent Ca2+ channels in rabbit colonic smooth musclecells by c-Src and focal adhesion kinase. J Biol Chem 1998;273:5337–42.

[46] Heidkamp MC, Bayer AL, Scully BT, Eble DM, Samarel AM. Acti-vation of focal adhesion kinase by protein kinase C epsilon in neonatalrat ventricular myocytes. Am J Physiol 2003;285:H1684–96.

[47] Akiyama T, Ishida J, Nakagawa S, Ogawara H, Watanabe S,Itoh N, et al. Genistein, a specific inhibitor of tyrosine-specific proteinkinases. J Biol Chem 1987;262:5592–5.

[48] Xiao GQ, Mochly-Rosen D, Boutjdir M. PKC isozyme selectiveregulation of cloned human cardiac delayed slow rectifier K current.Biochem Biophys Res Commun 2003;306:1019–25.

[49] Porter MJ, Heidkamp MC, Scully BT, Patel N, Martin JL,Samarel AM. Isoenzyme-selective regulation of SERCA2 geneexpression by protein kinase C in neonatal rat ventricular myocytes.Am J Physiol 2003;285:C39–47.

[50] Oishi K, Yamaguchi M. Effect of phorbol 12-myristate 13-acetate onCa2+-ATPase activity in rat liver nuclei. J Cell Biochem 1994;55:168–72.

[51] Mobley PL, Hedberg K, Bonin L, Chen B, Griffith OH. Decreasedphosphorylation of four 20-kDa proteins precedes staurosporine-induced disruption of the actin/myosin cytoskeleton in rat astrocytes.Exp Cell Res 1994;214:55–66.

[52] Malan D, Gallo MP, Bedendi I, Biasin C, Levi RC, Alloatti G.Microtubules mobility affects the modulation of L-type I(Ca) bymuscarinic and beta-adrenergic agonists in guinea-pig cardiac myo-cytes. J Mol Cell Cardiol 2003;35:195–206.

[53] Martiny-Baron G, Kazanietz MG, Mischak H, Blumberg PM,Kochs G, Hug H, et al. Selective inhibition of protein kinase Cisozymes by the indolocarbazole Go 6976. J Biol Chem 1993;268:9194–7.

[54] Toullec D, Pianetti P, Coste H, Bellevergue P, Grand-Perret T, Aja-kane M, et al. The bisindolylmaleimide GF 109203X is a potent andselective inhibitor of protein kinase C. J Biol Chem 1991;266:15771–81.

745F. Schröder et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 735–745