186 Research Article

IntroductionRegulation of the nuclear Ca2+ concentration ([Ca2+]) plays a keyrole in the regulation of gene expression. In cardiac myocytes, thereis a transient rise in the cytoplasmic [Ca2+] during each heartbeatcaused by Ca2+ influx through voltage-dependent sarcolemmalCa2+ channels during the action potential and the ensuing Ca2+-induced Ca2+ release from the sarcoplasmic reticulum (SR) throughryanodine (Ry) receptor Ca2+ release channels. This cytoplasmic[Ca2+] transient (CaT) serves to induce contraction and thuscouples the electrical to the mechanical activity of the heart(excitation-contraction coupling). The increase in cytoplasmic[Ca2+] is thought to transmit passively to the nucleus by Ca2+

diffusion through nuclear pore complexes (Genka et al., 1999;Tanaka et al., 1996a; Tanaka et al., 1996b). Thus, each cytoplasmicCaT elicits a nuclear CaT. According to this scheme, the nuclear[Ca2+] in cardiac myocytes is regulated passively and is entirelydependent on changes in the cytoplasmic [Ca2+]. Whether thenuclear [Ca2+] can also be regulated actively and, if so, by whichmechanisms, remains unknown.

In addition to Ry receptors, cardiac myocytes also expressinositol (1,4,5)-trisphosphate [Ins(1,4,5)P3] receptors, albeit atmuch lower densities (Lipp et al., 2000). Interestingly, theexpression of Ins(1,4,5)P3 receptors is greater in atrial than inventricular myocytes (Lipp et al., 2000), suggesting a moreprominent role for Ins(1,4,5)P3 in the atrial myocardium. Atrial

Ins(1,4,5)P3 receptors are found in the cytoplasm as well as inthe perinuclear area (Bare et al., 2005; Mackenzie et al., 2002;Yamada et al., 2001). G-protein-coupled receptors activating thephospholipase C (PLC)-Ins(1,4,5)P3 cascade thus have thepotential to modulate the cytoplasmic and nuclear [Ca2+] throughIns(1,4,5)P3-dependent Ca2+ release from the SR and nuclearenvelope, respectively. Elegant studies by Remus et al. havedemonstrated directly, using fluorescent Ins(1,4,5)P3 biosensors,that endothelin-1 (EDN1, hereafter referred to as ET) causestime-dependent increases in both cytoplasmic and nuclearIns(1,4,5)P3 concentration ([Ins(1,4,5)P3]) (Remus et al., 2006).It has not been shown so far, however, whether these increases in[Ins(1,4,5)P3] can elicit relevant alterations in cytoplasmic andnuclear [Ca2+] through Ins(1,4,5)P3-dependent Ca2+ release. Inparticular, alterations in nuclear CaTs in beating adult cardiacmyocytes following activation of the PLC-Ins(1,4,5)P3 cascadehave not been demonstrated to date. Such a mechanism wouldoffer the myocyte a means to regulate nuclear [Ca2+]independently from the action-potential-induced cytoplasmicCaT during excitation-contraction coupling and, therefore, wouldconstitute the basis for the concept of cardiac excitation-transcription coupling (Wu et al., 2006). Therefore, wecharacterised ET-induced changes in cytoplasmic and nuclearCaTs in electrically stimulated adult rabbit atrial myocytes andevaluated the possible involvement of PLC-Ins(1,4,5)P3

Nuclear Ca2+ plays a key role in the regulation of geneexpression. Inositol (1,4,5)-trisphosphate [Ins(1,4,5)P3)] mightbe an important regulator of nuclear Ca2+ but its contributionto nuclear Ca2+ signalling in adult cardiomyocytes remainselusive. We tested the hypothesis that endothelin-1 enhancesnuclear Ca2+ concentration transients (CaTs) in rabbit atrialmyocytes through Ins(1,4,5)P3-induced Ca2+ release fromperinuclear stores. Cytoplasmic and nuclear CaTs weremeasured simultaneously in electrically stimulated atrialmyocytes using confocal Ca2+ imaging. Nuclear CaTs weresignificantly slower than cytoplasmic CaTs, indicative ofcompartmentalisation of intracellular Ca2+ signalling.Endothelin-1 elicited a preferential (10 nM) or a selective (0.1nM) increase in nuclear versus cytoplasmic CaTs. This effectwas abolished by inhibition of endothelin-1 receptors,

phospholipase C and Ins(1,4,5)P3 receptors. Fractional Ca2+

release from the sarcoplasmic reticulum and perinuclear storeswas increased by endothelin-1 at an otherwise unaltered Ca2+

load. Comparable increases of cytoplasmic CaTs induced by �-adrenoceptor stimulation or elevation of extracellular Ca2+

could not mimic the endothelin-1 effects on nuclear CaTs,suggesting that endothelin-1 specifically modulates nuclearCa2+ signalling. Thus, endothelin-1 enhances nuclear CaTs inatrial myocytes by increasing fractional Ca2+ release fromperinuclear stores. This effect is mediated by the coupling ofendothelin receptor A to PLC-Ins(1,4,5)P3 signalling and mightcontribute to excitation-transcription coupling.

Key words: Endothelin, Inositol (1,4,5)-trisphosphate, Calcium,Nucleus, Myocyte

Summary

Endothelin-1 enhances nuclear Ca2+ transients inatrial myocytes through Ins(1,4,5)P3-dependent Ca2+

release from perinuclear Ca2+ storesJens Kockskämper1,2,*, Lea Seidlmayer1, Stefanie Walther1, Kristian Hellenkamp1, Lars S. Maier1 andBurkert Pieske1,2

1Department of Cardiology and Pneumology, University Medicine Göttingen, Germany2Department of Cardiology, Medical University of Graz, Auenbruggerplatz 15, A-8036 Graz, Austria*Author for correspondence (e-mail: jens.kockskaemper@meduni-graz.at)

Accepted 25 October 2007Journal of Cell Science 121, 186-195 Published by The Company of Biologists 2008doi:10.1242/jcs.021386

Jour

nal o

f Cel

l Sci

ence

187Endothelin enhances nuclear Ca2+

signalling and Ins(1,4,5)P3-dependent Ca2+

release in the regulation of nuclear [Ca2+].

ResultsIdentification and characterisation of nuclearCaTs in electrically stimulated atrial myocytesAs we sought to detect ET-induced changesin cytoplasmic and nuclear Ca2+ regulation,changes in the cytoplasmic and nuclear [Ca2+]had to be measured simultaneously withsufficient spatial and temporal resolution. Fig.1A illustrates a series of two-dimensionalconfocal images of an atrial myocyte during anelectrically stimulated CaT. The myocyteexhibited quite uniform resting fluorescence (a),indicating that Fluo-4 was distributed mainlyin the cytoplasm. Following stimulation,intracellular [Ca2+] started to increase in thesubsarcolemmal space and then propagated tothe centre of the cell (b-k). This pattern is typicalfor atrial myocytes lacking T-tubules(Kockskamper et al., 2001; Mackenzie et al.,2001). During the inward spread of the CaT, anellipsoidal region in the centre of the myocytewas characterised by a delayed increase influorescence (f-k). During the decay of the CaT,the same region exhibited higher fluorescencethan the surrounding parts of the cell (m-q).Thus, both rise and decay of the CaT weredelayed in this central region, which presumablyrepresented the nucleus, as suggested byprevious studies on cardiomyocyte Ca2+

signalling (Genka et al., 1999; Tanaka et al.,1996a; Tanaka et al., 1996b).

For a detailed kinetic analysis, CaTs ofsubcellular regions corresponding to the wholecytoplasm (‘wc’, black), the subsarcolemmalregions (‘ss’, green), the nucleus (‘nuc’, red) andtwo adjacent central areas of identical size (‘ct’,blue) were obtained, and these are plotted in Fig.1B (left). The subsarcolemmal CaTs had thelargest amplitudes and peaked first. The twocentral areas exhibited CaTs with loweramplitudes. They peaked ~60 mseconds laterthan the subsarcolemmal CaTs, but their decaykinetics were almost identical to thesubsarcolemmal decay. As the CaT from thewhole cytoplasm represents the sum of allcytoplasmic sites, it exhibited spatiotemporalcharacteristics that were in between those of thesubsarcolemmal and central sites. By contrast,the nuclear CaT displayed the lowest amplitudeand peaked 83 mseconds later than the centralCaTs. Fluorescence in the nucleus still rose whenthe central CaTs had already started to decline(vertical line). Furthermore, during the decayingphase, fluorescence in the nucleus became greaterand the kinetics of decay slower than in each of the cytoplasmicareas. Average values of peak systolic [Ca2+], time-to-peak and thetime constant � for transient decay are shown in Fig. 1B (right).Subsarcolemmal CaTs had significantly higher peaks and shorter

times-to-peak than central CaTs, whereas their time constants fordecay were identical. Nuclear CaTs, by contrast, exhibitedsignificantly lower peaks, longer times-to-peak and larger timeconstants for decay than both central and subsarcolemmal areas.

Fig. 1. Simultaneous confocal imaging of the cytoplasmic and nuclear [Ca2+] transient (CaT) inan electrically stimulated rabbit atrial myocyte. (A) 2D confocal images of an atrial myocyte.Numbers indicate milliseconds after electrical stimulation. (B, left) CaTs of distinct subcellularregions, as indicated in the schematic representation of the cell (same cell as in A): black trace:whole cytoplasm (wc); green traces: subsarcolemmal regions (ss); red trace: nucleus (nuc); bluetraces: central regions adjacent to the nucleus (ct). The vertical dashed line marks the peak of theCaTs of the two central regions. (B, right) Average values of peak systolic [Ca2+], time-to-peakand the time constant � for decay from 13 atrial myocytes. Colours refer to the subcellular regionsas shown in the cell scheme (bar, 5 �m). *, P<0.05 versus subsarcolemmal regions; #, P<0.05versus central regions. (C) A pair of atrial myocytes (a,b). Left: Fluo-4 fluorescence (Ca2+) duringthe decaying phase of an electrically stimulated CaT. Right: same cells after cessation ofstimulation and loading with Syto-16 (DNA). Because of the large increase in fluorescencecaused by Syto-16 loading, laser energy was attenuated to 10% of the initial value. Bars, 10 �m.The results in C are representative for a total of seven cells studied.

Jour

nal o

f Cel

l Sci

ence

188

Similar results were obtained using conventional line-scan imaging(data not shown).

To verify whether the ellipsoidal region in the centre of themyocyte characterised by a delayed CaT indeed represented thenucleus, myocytes were first loaded with Fluo-4 and electricallystimulated to record the CaT. Afterwards, stimulation was switchedoff and the cells were loaded additionally with the fluorescentnucleic acid dye Syto-16. Fig. 1C (left) illustrates an image of apair of myocytes obtained during the decay of the CaT. Fig. 1C(right) shows the same cells after cessation of stimulation andloading with Syto-16. Clearly, Syto-16 stained two regions in thecell centres that were essentially identical to the regions ofincreased Fluo-4 fluorescence during the CaT decay. Similarobservations were made in five other cells. These results confirmthat the ellipsoidal regions with delayed CaTs indeed represent thenuclei of atrial myocytes.

Taken together, these data indicate that cytoplasmic and nuclearCaTs in atrial myocytes can be measured simultaneously and thatthe nuclear CaTs are characterised by distinct kinetics. Thissuggests that the nucleus represents a cell compartment with itsown [Ca2+]-regulating properties.

The relative amplitude of the nuclear CaT varies between atrialmyocytesWithin a given atrial myocyte, the amplitude of the nuclear CaTrelative to the cytoplasmic CaT was constant (data not shown). Acomparison between myocytes, however, revealed that the amplitudeof the nuclear CaT relative to the cytoplasmic CaT was quite variable(Fig. 2). Fig. 2A shows three pairs of cytoplasmic (black) and nuclear(red) CaTs from three different atrial myocytes. In each case, the

Journal of Cell Science 121 (2)

amplitudes were normalised to the amplitude of the cytoplasmicCaT. The nuclear CaT could be smaller, equal to or higher than thecytoplasmic CaT. The kinetics of the nuclear CaTs, however, wereconsistently slower than the kinetics of the cytoplasmic CaTs, andthis was true for all atrial myocytes studied (n=95). Fig. 2B showsa histogram of the distribution of the ratio of systolic [Ca2+] in thenucleus to systolic [Ca2+] in the cytoplasm obtained from a total of95 atrial myocytes. In most cells (75%), this ratio was <1. However,in 25% of the atrial myocytes studied, the nuclear to cytoplasmicratio was >1. On average, the peak F/F0 (see Materials and Methods)in the nucleus amounted to 88±2% of the peak F/F0 in the cytoplasm(n=95). The fact that different atrial myocytes exhibited variablerelative amplitudes of the nuclear CaT provides further evidence forthe notion that the nucleus is capable of regulating [Ca2+]independently from cytoplasmic [Ca2+]. The reason for thedifferences in relative nuclear CaT amplitude between cells,however, is not known at present. We speculate that it might berelated to variations of the Ca2+ content of perinuclear and SR Ca2+

stores and/or variations of the Ca2+ release from these stores. In thecontext of this study, it is important to note that neither the absolutenor the relative amplitude of the nuclear CaT affected the ability ofET to enhance nuclear CaTs (see Fig. 5).

ET augments cytoplasmic and nuclear CaTs and alters theirkineticsIn cardiac myocytes, ET activates G-protein-coupled receptors tostimulate the PLC-Ins(1,4,5)P3 signalling cascade (Endoh et al.,1998; Sugden, 2003). Previous studies have shown that ET canincrease global CaTs in atrial myocytes of various species,including rat (Bootman et al., 2007; Mackenzie et al., 2002), cat(Zima and Blatter, 2004), mouse (Li et al., 2005) and human(Meyer et al., 1996). In rabbit atrial myocytes, we found that ETelicited a time- and concentration-dependent increase inintracellular CaTs (n=58, [ET] ranging from 0.1 to 50 nM).Maximal increases were observed 10-20 minutes after applicationof 10 nM ET (data not shown). Therefore, ET-induced changes incytoplasmic and nuclear CaTs were studied 10 minutes afterapplication of the peptide and compared with those of pre-ETcontrols. The effects of ET on cytoplasmic and nuclear CaTs areillustrated in Fig. 3. Fig. 3A shows two-dimensional fluorescenceimages of an atrial myocyte during individual CaTs before (left,control) and during exposure of the cell to 10 nM ET (right). ETincreased both the cytoplasmic and nuclear [Ca2+]. To quantify andcompare the increases in cytoplasmic and nuclear [Ca2+], regionsof interest corresponding to the whole cytoplasm and the nucleus,respectively, were analysed. Fig. 3B shows four consecutive CaTsfrom the cytoplasm (black) and the nucleus (red) before and duringET exposure. Clearly, ET increased CaTs in both regions. Notably,however, the increase in nuclear CaTs was larger than incytoplasmic CaTs. Average values are shown in Fig. 3C (n=24).ET increased systolic [Ca2+] in the cytoplasm from 6.75±0.44 F/F0

to 8.74±0.66 F/F0 or to 130±5% of the control and in the nucleusfrom 6.07±0.75 F/F0 to 8.94±1.12 F/F0 or to 148±5% of the control(all P<0.01). The ratio of nuclear to cytoplasmic peak [Ca2+](nuc/cyto ratio) was elevated from 0.85±0.05 to 0.97±0.06 or to115±3% of the control (all P<0.01).

ET also altered the kinetics of the CaT (Fig. 3D, n=24). Thetime-to-peak CaT was significantly increased in the cytoplasmfrom 134±6 mseconds to 163±7 mseconds (P<0.01) but wasunaltered in the nucleus (~220 mseconds). This reduced the delaybetween peak [Ca2+] in the cytoplasm and nucleus from 84±9

Fig. 2. The nuclear CaT amplitude varies between atrial myocytes. (A) Pairs ofcytoplasmic (black) and nuclear (red) CaTs from three different atrialmyocytes. For better comparison, CaTs were normalised to the cytoplasmicCaT. (B) Histogram showing the distribution of the ratio of nuclear tocytoplasmic systolic [Ca2+] obtained under control conditions from a total of95 atrial myocytes studied.

Jour

nal o

f Cel

l Sci

ence

189Endothelin enhances nuclear Ca2+

mseconds to 60±10 mseconds (P<0.05). The time constant for CaTdecay was decreased both in the cytoplasm and nucleus from236±25 mseconds to 205±17 mseconds (P<0.05) and from 350±32mseconds to 262±20 mseconds (P<0.01), respectively. Thus, ETprolonged the time-to-peak in the cytoplasm, reduced thedifference between peak [Ca2+] in the cytoplasm and nucleus andaccelerated CaT decay in the cytoplasm and nucleus.

Fig. 4 illustrates the effects of ET on cytoplasmic and nuclearCaTs when recorded using conventional confocal line-scanimaging. Compared with our Nipkow disc-based 2D imagingsystem, line-scan (1D) imaging offers higher temporal resolution(0.77 mseconds per scan line versus 8.33 mseconds per 2D image)but provides less spatial information. To record part of thecytoplasmic and part of the nuclear CaT simultaneously with thistechnique, the scan line was positioned perpendicular to thelongitudinal axis of the myocyte crossing the nucleus, as illustratedin the schematic drawing of the cell in Fig. 4A. The line-scan imagesand the cytoplasmic (black) and nuclear (red) CaTs obtained fromthe regions shown next to the line-scan images are presented in Fig.4A before (top) and following (bottom) ET exposure. Similar towhat was found using 2D imaging (Fig. 3), ET increasedcytoplasmic and nuclear CaTs and altered their kinetics in a specificmanner. In a total of five myocytes (Fig. 4B), ET increased systolic[Ca2+] in the cytoplasm by 22±9% and in the nucleus by 46±7%(P<0.01). Time-to-peak was significantly increased in thecytoplasm from 46±7 mseconds to 99±16 ms (P<0.05) but remainedunchanged in the nucleus (156±9 ms versus 164±23 ms, P=not

significant). Finally, the time constant for decay was unaltered inthe cytoplasm (193±23 mseconds versus 186±21 mseconds; P=notsignificant) and decreased from 372±43 mseconds to 264±21mseconds in the nucleus (P<0.01). These results compare well withthe data obtained by 2D confocal imaging (see Fig. 3) and confirmthat ET augments CaTs preferentially in the nucleus.

The ability of ET to enhance nuclear CaTs is independent ofthe amplitude of the nuclear CaTThe absolute as well as the relative amplitude of the nuclear CaTvaried between cells (see Fig. 2). This raised the question regardingwhether the effect of ET on nuclear CaTs might be dependent oneither the absolute or relative amplitude of the nuclear CaT beforeET stimulation. Fig. 5 shows the ET-induced increase in nuclearCaTs from 24 atrial myocytes as a function of the absoluteamplitude (expressed as systolic nuclear F/F0; Fig. 5A) and therelative amplitude of the nuclear CaT (expressed as the ratio ofsystolic nuclear to systolic cytoplasmic F/F0; Fig. 5B) before ETexposure. There was no clear correlation between the ET effect onnuclear CaTs and nuclear CaT amplitudes in either case, indicatingthat the ability of ET to enhance nuclear CaTs was independent ofthe initial nuclear CaT amplitude.

Increases in cytoplasmic CaTs are not sufficient to enhancenuclear CaTsThe ET-induced enhancement of nuclear CaTs might have beencaused by specific nuclear actions of ET or simply be secondary

Fig. 3. ET enhances nuclear CaTs. (A) 2DFluo-4 fluorescence images of an atrialmyocyte during an electrically stimulated CaTbefore (Control) and during exposure to 10 nMET. Bar, 20 �m. (B) Cytoplasmic (blacktraces) and nuclear (red traces) CaTs recordedbefore (left) and during (right) ET treatment.(C) The means ± s.e.m. of ET-induced changesin systolic [Ca2+]. Nuc/Cyto, ratio of nuclear tocytoplasmic [Ca2+]. (D) ET-induced changes inCaT kinetics. (Left) Normalised original CaTsbefore and during (blue) ET exposure. (Right)The means ± s.e.m. of time-to-peak and thetime constant for CaT decay. The data wereobtained from 24 atrial myocytes.

Jour

nal o

f Cel

l Sci

ence

190

to increases in cytoplasmic CaTs. To examine this possibility, wevaried the amplitude of cytoplasmic CaTs either by variations inextracellular [Ca2+] or by exposure of the myocytes to the �-adrenoceptor agonist isoproterenol (ISO). To facilitate directcomparison, cells were treated with ISO concentrations (2-10 nM)eliciting increases in cytoplasmic CaTs that were almost identicalto the ET-induced increases [ISO: 126±5% (n=10) versus ET:130±5% (n=24), P=not significant, Fig. 6B]. Fig. 6A illustratescytoplasmic (black) and nuclear (red) CaTs of two atrial myocyteschallenged with either ET (10 nM) or ISO (10 nM). Both ET andISO elicited a comparable increase in the cytoplasmic CaT. OnlyET, however, was able to cause a relatively larger increase in thenuclear CaT. ISO, by contrast, did not elicit an increase in thenuclear CaT, despite the augmentation of the cytoplasmic CaT.Similar results were obtained when the cytoplasmic CaT waselevated by increasing extracellular [Ca2+] from 2 mM to 4 mM(data not shown). Average results of the three protocols arepresented in Fig. 6B,C. The normalised increases in systolic [Ca2+]in the cytoplasm (black) and the nucleus (red) evoked by ET, ISOand 4 mM extracellular Ca2+ are shown in Fig. 6B, whereas therespective alterations in the nuclear:cytoplasmic ratio are displayed

Journal of Cell Science 121 (2)

Fig. 4. Line-scan imaging of the ET-induced enhancement of nuclear CaTs.(A) Original line-scan images and normalised fluorescence traces of thecytoplasmic (black) and nuclear (red) CaTs in an atrial myocyte before(Control) and following a 10 minute exposure to 10 nM ET. The cell schemeillustrates the position of the scan line. Bars (at right), 5 �m. (B) Averagevalues of the effects of 10 nM ET on peak systolic [Ca2+] (normalised to thepre-ET control), time to peak and the time constant � for decay in thecytoplasm (black) and the nucleus (red). Means ± s.e.m. of five atrialmyocytes.

Fig. 5. The ET-induced enhancement of nuclear CaTs depends neither on thebasal systolic [Ca2+] in the nucleus nor on the ratio of nuclear to cytoplasmicpeak [Ca2+]. The ET-induced increase in peak [Ca2+] in the nucleus as afunction of basal systolic [Ca2+] in the nucleus (A) or the ratio of nuclear tocytoplasmic systolic [Ca2+] (B) before ET exposure. The data are from 24atrial myocytes challenged with ET.

Fig. 6. Elevation of cytoplasmic CaTs is not sufficient to enhance nuclearCaTs. (A) Original recordings of cytoplasmic (black) and nuclear CaTs (red) oftwo atrial myocytes before and during exposure to 10 nM ET (left) or 10 nMISO (right). (B) Average values of peak systolic [Ca2+] in the cytoplasm(black) and in the nucleus (red) following application of ET (n=24), ISO(n=10) or 4 mM extracellular Ca2+ (‘4Ca’; n=6). (C) Normalised ratio ofnuclear to cytoplasmic [Ca2+] following application of ET (n=24), ISO (n=10)or 4 mM extracellular Ca2+ (n=6). In B and C, values were normalised to theinitial control (=100%). **, P<0.01 versus initial control.

Jour

nal o

f Cel

l Sci

ence

191Endothelin enhances nuclear Ca2+

in Fig. 6C. ET (n=24), ISO (n=10) and 4 mM extracellular Ca2+

(n=6) increased cytoplasmic [Ca2+] to 130±5%, 126±5% and114±3% of the initial control, respectively. Only ET, however,elicited a relatively larger increase in nuclear [Ca2+] (to 148±5%of the control). By contrast, both ISO and 4 mM extracellular Ca2+

caused increases in nuclear [Ca2+] (to 125±5% and 115±3% of thecontrol, respectively) that were essentially identical to the increasesin cytoplasmic [Ca2+]. These subcellular differences in theelevations of cytoplasmic and nuclear CaTs are also reflected in thenuclear:cytoplasmic ratio obtained during the peak of the transient(Fig. 6C). ET increased this ratio by +15±3%. In the presence ofISO or 4 mM extracellular Ca2+, however, it remained unaffected,indicating that there was no additional increase in nuclear [Ca2+]beyond the increase in cytoplasmic [Ca2+] under these conditions.Taken together, these data demonstrate that the enhancement ofnuclear CaTs is a specific action of ET. It cannot be mimicked byelevations of cytoplasmic CaTs by either ISO or 4 mM extracellularCa2+. Thus, elevations in cytoplasmic CaTs are not sufficient toincrease nuclear CaTs to a larger extent than cytoplasmic CaTs.

Selective increases of nuclear CaTs by low concentrations ofETFig. 7A illustrates cytoplasmic (black) and nuclear (red) CaTs of amyocyte challenged with 0.1 nM ET. Before exposure to ET, theamplitudes of cytoplasmic and nuclear CaTs were essentiallyidentical. Following application of ET, cytoplasmic CaTs remainedunchanged. Nuclear CaTs, however, clearly rose above baselinelevels (dashed red line). Similar observations were made in fiveadditional myocytes. The average data for changes in cytoplasmicand nuclear CaTs elicited by 0.1 nM ET are presented in Fig. 7B.Cytoplasmic CaTs were unaffected by 0.1 nM ET. Ten minutesafter exposure to 0.1 nM ET, the systolic [Ca2+] in the cytoplasmwas essentially identical (104±4%, P=not significant) to that of theinitial control. By contrast, nuclear CaTs were significantlyincreased by this low concentration of ET. After 10 minutes, thesystolic [Ca2+] in the nucleus was increased to 109±4% (P<0.05)of the initial control. Consequently, ET (0.1 nM)significantly increased the nuclear:cytoplasmic ratio by5±1% (Fig. 7B, right). The results show that, at 0.1 nM,ET increased nuclear CaTs selectively withoutelevating cytoplasmic CaTs.

The ET-induced increases in cytoplasmic andnuclear CaTs are mediated by ET receptorscoupling to PLC-Ins(1,4,5)P3 signallingTo elucidate the signalling pathway underlying the ET-induced enhancement of nuclear CaTs, the followinginhibitors were tested: BQ-123 (0.25 �M) – targetingthe ET receptor A (EDNRA, hereafter referred to ETA);U-73122 (3 �M) – targeting PLC; and 2-aminoethoxy-diphenylborate (2-APB; 3 �M) and xestospongin C (5�M) – both targeting Ins(1,4,5)P3 receptors. Myocyteswere incubated with xestospongin C (>40 minutes)before ET treatment, whereas the other inhibitors wereapplied acutely 10 minutes before ET exposure. BQ-123 (n=10) and 2-APB (n=16) did not affect basalCaTs both in the cytoplasm (BQ-123: 101±3% of thecontrol; 2-APB: 99±3% of the control) and in thenucleus (BQ-123: 103±3% of the control; 2-APB:99±3% of the control). U-73122 (n=8), by contrast,reduced basal CaTs to 82±6% of the control in the

cytoplasm and to 82±6% of the control in the nucleus (both P<0.05;data not shown).

Fig. 8A shows cytoplasmic (black) and nuclear (red) CaTs oftwo atrial myocytes before and during application of BQ-123 (left)or 2-APB (right), inhibitors of ETA and of Ins(1,4,5)P3 receptors,respectively. Neither BQ-123 nor 2-APB altered cytoplasmic ornuclear CaTs. When ET was applied in the presence of BQ-123 or2-APB, cytoplasmic and nuclear CaTs remained essentially

Fig. 8. Signalling pathway underlying the ET-induced enhancement of nuclear CaTs.(A) Original recordings of cytoplasmic (black) and nuclear CaTs (red) before and afterapplication of 0.25 �M BQ-123 (left) or 3 �M 2-APB (right) and following additionalapplication of 10 nM ET. (B) Means ± s.e.m. of peak cytoplasmic and nuclear [Ca2+]following exposure to ET for 10 minutes in the absence and presence of BQ-123 (n=10),U-73122 (n=7), 2-APB (n=12) and xestospongin C (n=7).

Fig. 7. Low concentrations of ET augment nuclear CaTs in the absence ofincreases in cytoplasmic CaTs. (A) Original recording of cytoplasmic (black)and nuclear (red) CaTs of an atrial myocyte before and during exposure to0.1 nM ET. (B) Changes of peak systolic [Ca2+] in the cytoplasm (left) andnucleus (right) induced by 0.1 nM ET. (C) Normalised ratio of nuclear tocytoplasmic [Ca2+] following application of 0.1 nM ET. In B and C, values arenormalised to the initial control (=100%). Means ± s.e.m. of six atrialmyocytes (N.S.=not significant).

Jour

nal o

f Cel

l Sci

ence

192

unaffected. Fig. 8B shows average data of the changes incytoplasmic and nuclear CaTs induced by exposure to 10 nM ETfor 10 minutes in the absence and presence of BQ-123, U-73122,2-APB and xestospongin C. Each inhibitor tested attenuated orabolished the ET-mediated increases in cytoplasmic and therelatively larger increases in nuclear CaTs. In addition, eachinhibitor also abolished the ET-induced alterations in the kineticsof cytoplasmic and nuclear CaTs (data not shown). Thus, theintracellular signalling pathway causing the ET-induced increasesin cytoplasmic and nuclear CaTs involved activation of ETA

receptors that couple to PLC and Ins(1,4,5)P3.

Increased fractional Ca2+ release from SR and perinuclearstores underlies the ET-induced increase in cytoplasmic andnuclear CaTsThe results obtained so far suggested that Ins(1,4,5)P3-dependentCa2+ release from SR and perinuclear stores contributed to the ET-induced increases in cytoplasmic and nuclear CaTs. To characterisein more detail the role of Ca2+ release from these intracellular Ca2+

stores for the ET-induced increases in cytoplasmic and nuclearCaTs, Ca2+ store contents were estimated by rapid application of20 mM caffeine, and fractional Ca2+ release (FCR) from the storeswas calculated as the ratio of the amplitude of electrically evokedCaTs to the amplitude of the caffeine-induced CaTs (see Materialsand Methods for further details). Fig. 9A illustrates an originalrecording of cytoplasmic (black) and nuclear (red) CaTs of an atrialmyocyte exposed to 1 nM ET. Two electrically evoked CaTsfollowed by a caffeine-induced CaT are shown first in the absenceand then in the presence of ET. ET elicited an increase in the

Journal of Cell Science 121 (2)

cytoplasmic CaT (+44%) and an even more pronounced increasein the nuclear CaT (+97%). Application of caffeine, however,revealed that neither the SR nor perinuclear Ca2+ store content waselevated by ET. Fractional Ca2+ release from the SR increased from0.53 to 0.92 and fractional Ca2+ release from perinuclear storesincreased from 0.38 to 0.86. Average data from seven myocytesconfirm these observations (Fig. 9B). The Ca2+ contents of bothstores were unchanged, whereas fractional Ca2+ release wassignificantly augmented by ET (SR: from 0.65±0.05 to 0.78±0.07;perinuclear: from 0.52±0.06 to 0.75±0.08; n=7, both P<0.05). TheET-induced increase in FCR (�FCR) from perinuclear stores(0.23±0.06) was significantly larger than the increase in FCR fromthe SR (0.13±0.04; n=7, P<0.05).

For comparison, analogous experiments were conducted withISO (10 nM, n=7). ISO increased cytoplasmic and nuclear CaTsby the same extent (~50%; data not shown). The increase incytoplasmic CaTs was associated with a 13±4% elevation of SRCa2+ content (Fig. 9B). Furthermore, ISO increased FCR both fromSR and perinuclear regions by the same amount (SR: 0.22±0.07;perinuclear: 0.21±0.06, Fig. 9B). Thus, ISO induced alterations inSR and perinuclear Ca2+ handling that were distinctly differentfrom the ET-induced alterations.

DiscussionThe physiological and pathophysiological relevance ofIns(1,4,5)P3 signalling in cardiac myocytes has long beenenigmatic. Recently, some elegant studies have led to novelinsights into the putative roles of Ins(1,4,5)P3 in cardiac myocytes.For example, ET-induced Ins(1,4,5)P3 signalling generatesarrhythmogenic alterations in Ca2+ signalling in atrial andventricular myocytes (Bootman et al., 2007; Li et al., 2005;Mackenzie et al., 2002; Proven et al., 2006; Zima and Blatter,2004). In addition, most recent work has implicated Ca2+ releasethrough Ins(1,4,5)P3 receptors from the nuclear envelope into thenucleoplasm in Ca2+-dependent activation of transcription andhypertrophy in adult cardiac myocytes, a concept referred to asexcitation-transcription coupling (Bare et al., 2005; Wu et al.,2006). The present study, however, is the first to demonstratedirectly active modulation of nuclear CaTs during normalexcitation-contraction coupling in cardiac myocytes byIns(1,4,5)P3-dependent Ca2+ release from perinuclear stores. Ourmajor new findings are: (1) ET increases cytoplasmic and nuclearCaTs in beating adult atrial myocytes through activation of ETA

receptors coupling to PLC-Ins(1,4,5)P3 signalling. (2) The increasein nuclear CaTs is larger than the increase in cytoplasmic CaTs. Atlow concentrations, ET causes selective increases in nuclear CaTs.(3) The enhancement of nuclear CaTs is specific for ET-inducedIns(1,4,5)P3 signalling and elicited by increased fractional Ca2+

release from perinuclear stores presumably mediated by Ca2+

release through Ins(1,4,5)P3 receptors in the nuclear envelope.These results thus provide a sound framework for the cellular andsubcellular mechanisms underlying a compartmentalised increasein nuclear [Ca2+] in cardiac myocytes evoked by ET.

Mechanisms underlying the ET-induced increases incytoplasmic and nuclear CaTsAccording to our results, ET elicited increases in both cytoplasmicand nuclear CaTs through activation of ETA receptor coupling toPLC-Ins(1,4,5)P3 signalling. Interestingly; inhibition of the PLC-Ins(1,4,5)P3 pathway completely suppressed the ET effects oncytoplasmic and nuclear CaTs. This is somewhat surprising, given

Fig. 9. ET increases fractional Ca2+ release from SR and perinuclear stores.(A) Original recording of cytoplasmic (black) and nuclear (red) CaTs before(left) and after (right) ET application. Each time, two electrically stimulatedCaTs and a caffeine-induced CaT are shown. The ET effect on electricallystimulated CaTs shown here was the largest effect recorded in this series ofexperiments. Still there was no increase in Ca2+ load of either SR orperinuclear (PN) stores, indicating that ET augmented the CaT by increasedfractional release rather than elevation of Ca2+ store content. (B) Averagevalues of Ca2+ store content (left, normalised to control) and fractional Ca2+

release (FCR) from SR and PN stores (middle) before and followingapplication of ET (n=7) or ISO (n=7). Right, ET- and ISO-induced increases infractional release from SR and PN stores (�FCR). Means ± s.e.m. of sevenatrial myocytes for each case.

Jour

nal o

f Cel

l Sci

ence

193Endothelin enhances nuclear Ca2+

that ET activates various signalling cascades (Sugden, 2003), andsuggests that PLC-Ins(1,4,5)P3 signalling is the major pathwayutilised by atrial myocytes to elevate cytoplasmic and nuclearCaTs. Our findings are in line with a recent study in which an ET-induced increase in cytoplasmic and nuclear [Ins(1,4,5)P3] incardiac myocytes was demonstrated directly by means of novelFRET-based Ins(1,4,5)P3 biosensors (Remus et al., 2006).Furthermore, previous studies have shown the expression ofIns(1,4,5)P3 receptors in the cytoplasm as well as in the nuclearenvelope of cardiac myocytes (Bare et al., 2005; Lipp et al., 2000;Mackenzie et al., 2002; Yamada et al., 2001). Thus, it is likely thatET increased the [Ins(1,4,5)P3] in the cytoplasm and nucleus andaugmented CaTs in the two compartments through Ca2+ releasefrom Ins(1,4,5)P3 receptors located in the SR and the nuclearenvelope, respectively. Ins(1,4,5)P3-dependent Ca2+ release fromperinuclear stores into the nucleoplasm has been observed beforein rat neonatal myocytes (Ibarra et al., 2004; Luo et al., 2007), incat atrial myocytes (Zima et al., 2007) and in nuclei isolated fromrat skeletal myotubes (Cardenas et al., 2005) or rat heart (Zima etal., 2007). The most elegant and direct evidence for independentIns(1,4,5)P3-mediated nuclear Ca2+ signalling comes from a recentstudy on permeabilised atrial myocytes and isolated cardiac nuclei(Zima et al., 2007). In that study, it was shown that Ins(1,4,5)P3

and the potent Ins(1,4,5)P3 receptor agonist adenophostin were ableto release Ca2+ from the nuclear envelope directly into thenucleoplasm, suggesting that nuclear Ins(1,4,5)P3 receptors arelocated predominantly on the inner membrane of the nuclearenvelope (Zima et al., 2007). A further study provided evidence forthe involvement of the ET-PLC-Ins(1,4,5)P3 pathway in localisednuclear Ca2+ signalling in cultured rabbit and mouse ventricularmyocytes (Wu et al., 2006). An Ins(1,4,5)P3-mediated increase innuclear [Ca2+], however, could not be detected (Wu et al., 2006).Thus, in conjunction with most recent work on rat atrial myocytes[see fig. 3 in Bootman et al. (Bootman et al., 2007)], the presentstudy is the first to demonstrate directly an Ins(1,4,5)P3-mediated[Ca2+] increase in the nucleus of adult beating cardiac myocytes.The results suggest that localised Ins(1,4,5)P3-mediated increasesin nuclear [Ca2+] might occur in intact cardiac myocytes duringnormal excitation-contraction coupling.

The ET-induced alterations in the kinetics of cytoplasmic andnuclear CaTs are also consistent with Ins(1,4,5)P3-dependent Ca2+

release being the underlying mechanism. As shown by Zima andBlatter, the duration of Ins(1,4,5)P3-mediated Ca2+ release eventsin atrial myocytes is approximately three times longer than Ryreceptor-2-mediated Ca2+ release events (Zima and Blatter, 2004).This might explain the increase in the time-to-peak of thecytoplasmic CaT following ET exposure. Because of the muchslower kinetics of nuclear CaTs, such an effect might be obscuredand not detectable in the nucleus. Moreover, Ca2+/calmodulin-dependent protein kinase II (CaMKII) is coupled to Ins(1,4,5)P3

receptors (Bare et al., 2005) and is responsible for the frequency-dependent acceleration of CaT decay (DeSantiago et al., 2002),suggesting that the ET-induced acceleration of cytoplasmic andnuclear CaT decay was mediated by Ins(1,4,5)P3-dependentstimulation of CaMKII activity.

ET increases nuclear CaTs selectivelyLow concentrations of ET elicited selective increases in nuclearCaTs, meaning that increases in nuclear CaTs occurred in theabsence of increases in cytoplasmic CaTs. This finding has twoimportant implications: first, it demonstrates that ET can elicit a

compartmentalised increase in nuclear [Ca2+]; and second, ETmight alter nuclear Ca2+ signalling and, therefore, presumably alsoCa2+-dependent gene expression without causing relevant changesin cytoplasmic Ca2+ signalling. The question remains how ET-induced Ins(1,4,5)P3 signalling can alter the nuclear [Ca2+] withoutaltering cytoplasmic Ca2+ signalling. One possible explanation forthis apparent paradox is that nuclear Ins(1,4,5)P3 receptors mightexhibit either an increased density or an increased sensitivity toIns(1,4,5)P3. This means that, at the same concentration ofIns(1,4,5)P3, nuclear Ins(1,4,5)P3 receptors might be activated,whereas cytoplasmic Ins(1,4,5)P3 receptors are not. AsIns(1,4,5)P3 receptors are modulated by a variety of factors(Bultynck et al., 2003), differences in Ins(1,4,5)P3 sensitivitybetween cytoplasmic and nuclear Ins(1,4,5)P3 receptors could becaused by subcellular differences in the concentrations or activitiesof these modulatory factors. Alternatively, the local [Ins(1,4,5)P3]near nuclear Ins(1,4,5)P3 receptors might be higher than in thevicinity of cytoplasmic Ins(1,4,5)P3 receptors. A preferentialincrease of perinuclear [Ins(1,4,5)P3] could be caused by activationof nuclear PLC through nuclear ET receptors (Bkaily et al., 2003;Boivin et al., 2003). This explanation, however, appears less likelyfor two reasons: first, it is unclear how extracellular ET couldactivate intracellular ET receptors and, second, ET-inducedelevation of the [Ins(1,4,5)P3] in cardiomyocytes occurs first in thecytoplasm and with a delay in the nucleus (Remus et al., 2006),suggesting that Ins(1,4,5)P3 is generated in the cytoplasm beforediffusing into the nucleus.

The nuclear CaT consists of two componentsAnother important finding of the current study is that the nuclearCaT exhibits distinct kinetics and that it consists of twocomponents, a passive and an active component. The passivecomponent is probably mediated by cytoplasmic Ca2+ diffusingthrough nuclear pores to increase nuclear [Ca2+]. Several lines ofevidence support this notion. First, nuclear pore complexes arereadily permeable to Ca2+ ions. Second, the nuclear CaT lagsbehind the cytoplasmic CaT. This delay is most likely caused byslow diffusion of cytoplasmic Ca2+ through the nuclear porecomplexes. In addition, the effective diffusion of Ca2+ within thenucleus is significantly slower than in the cytoplasm (Soeller et al.,2003), and this could contribute to the slower kinetics of thenuclear CaT. Third, in the absence of Ins(1,4,5)P3 production, anyincreases in cytoplasmic CaTs are followed by increases in nuclearCaTs of identical magnitude. During ET treatment, however, anadditional active component comes into play. This activecomponent is reflected by the ET-induced extra increase in nuclearCaTs. It is inhibited by U-73122, 2-APB and xestospongin Cand, therefore, is mediated by PLC-Ins(1,4,5)P3 signalling.Determination of the Ca2+ load of perinuclear stores revealed that,during ET exposure, the Ca2+ store content was unchanged.Fractional release from perinuclear stores, however, was increasedgreatly by ET. Taken together, these data imply that ET increasesnuclear CaTs by increasing fractional Ca2+ release from the nuclearenvelope through recruitment of Ins(1,4,5)P3 receptor-mediatedCa2+ release.

Physiological and pathophysiological implicationsThe ET receptor density and plasma levels are elevated incardiovascular diseases, including human heart failure (Pieske et al.,1999). Nuclear Ca2+ is a crucial factor for the regulation of geneexpression. The ET-induced increases in nuclear [Ca2+]

Jour

nal o

f Cel

l Sci

ence

194

characterised in the present study, therefore, are expected to altercardiac gene expression, consistent with the fact that ET exertshypertrophic actions in the heart. There is clear experimentalevidence for the functional relevance of Ins(1,4,5)P3-dependentCa2+ release for gene expression in striated muscle. In isolatednuclei of skeletal myotubes, Ins(1,4,5)P3-dependent Ca2+ releaseinto the nucleoplasm causes phosphorylation of the transcriptionfactor cAMP response element binding protein (Cardenas et al.,2005). Moreover, in cardiac myocytes, Ins(1,4,5)P3 receptors andCaMKII colocalise in the nucleus (Bare et al., 2005), andIns(1,4,5)P3-dependent Ca2+ release into the nucleoplasm has beenimplicated in CaMKII-mediated phosphorylation of histonedeacetylases and de-repression of gene expression (Wu et al., 2006).Remarkably, our finding that low concentrations of ET causeselective increases in nuclear CaTs emphasises that the nuclear[Ca2+] can be regulated independently from the cytoplasmic [Ca2+]in beating cardiac myocytes during normal excitation-contractioncoupling. Furthermore, it suggests that the hypertrophic effects ofET might occur in the low concentration range found in cardiactissue and precede the onset of the inotropic and arrhythmogeniceffects of ET caused by alterations in cytoplasmic [Ca2+].

Materials and MethodsAtrial myocytesAtrial myocytes from adult rabbit hearts were isolated by a standard collagenase-based Langendorff perfusion protocol as described previously for ventricularmyocytes (Schillinger et al., 2000). Isolated myocytes were plated on glass-bottomedculture dishes and allowed to attach to the glass bottom for >45 minutes. The cellisolation procedure was in accordance with national and international animal careguidelines.

Confocal Ca2+ imagingFast, two-dimensional (2D) confocal [Ca2+] imaging was performed as describedpreviously (Kockskamper et al., 2001) using a confocal imaging setup (VisiTechInternational, UK) consisting of an inverted microscope (Nikon) equipped with a�40 oil-immersion objective lens (N.A. 1.3), a Nipkow dual disc-based confocalunit and an ICCD camera with a temporal resolution of 120 Hz. Myocytes wereloaded with Fluo-4 by a 20-25 minute incubation in Tyrode’s solution containing 8�M Fluo-4/AM. Fifteen minutes were allowed for de-esterification. Myocytes wereplaced on the stage of the microscope, superfused with Tyrode’s solution by meansof a gravity-driven superfusion pipette and field-stimulated at 0.7 Hz. Fluo-4 wasexcited by 488 nm light from an argon-ion laser. The fluorescence emission wascollected at wavelengths >505 nm. Changes of the [Ca2+] are expressed as changesof background-corrected normalised fluorescence, F/F0, where F denotes Fluo-4fluorescence and F0 resting fluorescence at the beginning of an experiment.

Line-scan imaging of atrial CaTs was performed using a confocal microscope(Zeiss LSM 5 Pascal) equipped with a �40 oil-immersion objective lens (N.A. 1.3).A zoom factor of six was used in all recordings, yielding a spatial resolution of 0.07�m. The scan time was 0.77 mseconds per line. Myocytes were treated as describedabove for 2D Ca2+ imaging (20-25 minutes loading with 8 �M Fluo-4/AM; 0.7 Hzstimulation). Fluo-4 was excited by the 488 nm line of an attached argon ion laserand fluorescence was collected at >515 nm.

Staining of the nucleusThe nucleus was stained by the nucleic acid dye Syto-16 (Molecular Probes) andimaged using 2D confocal microscopy. Syto-16 stock solution (1 mM in DMSO)was diluted in Tyrode’s solution (1:100) to yield a concentration of 10 �M. 40 �l ofthis solution were then added directly to the bath solution (2 ml) containing themyocytes, thus yielding a final concentration of 0.2 �M Syto-16. Syto-16fluorescence was imaged in quiescent cells following recording of CaTs with Fluo-4 using identical excitation and emission wavelengths of 488 nm and >505 nm,respectively. The fluorescence of Syto-16-loaded cells was much brighter than Fluo-4 fluorescence. Therefore, after recording of Fluo-4 fluorescence, the laser intensitywas reduced by 90% by means of a neutral-density filter in the excitation light pathbefore recording Syto-16 fluorescence. This method allowed imaging of the nuclear[Ca2+] signal and identification of the nucleus by Syto-16 staining in the same cells.

Measurement of the Ca2+ load of SR and perinuclear stores anddetermination of fractional Ca2+ releaseThe Ca2+ content of SR and perinuclear stores was measured using rapid applicationof 20 mM caffeine (Bassani et al., 1993). Caffeine quickly depletes both SR and

Journal of Cell Science 121 (2)

perinuclear stores, with time-constants in the range of ~150 mseconds and ~200mseconds; respectively (Wu and Bers, 2006). Myocytes were superfused withTyrode’s solution and electrically stimulated (0.7 Hz) for >2 minutes. Stimulationwas switched off and caffeine was applied rapidly, i.e. within ~1 second (as verifiedin separate experiments by monitoring the fluorescence change induced by switchingthe superfusate from normal Tyrode’s solution to a fluorescein-containing Tyrode’ssolution). The amplitude of the caffeine-induced CaT in the cytoplasm was taken asthe SR Ca2+ load and the amplitude in the nucleus as the Ca2+ load of perinuclearstores. Fractional Ca2+ release from the SR was then calculated as the ratio of theamplitude of the electrically stimulated cytoplasmic CaT to SR Ca2+ load. In thesame way, fractional Ca2+ release from perinuclear stores was calculated as the ratioof the amplitude of the electrically stimulated nuclear CaT to the Ca2+ load ofperinuclear stores. Following application of caffeine under control conditions, thesuperfusate was switched back to caffeine-free Tyrode’s solution and stimulation wasresumed. Myocytes were then exposed to ET for 10 minutes before application ofcaffeine was repeated to determine the Ca2+ load of, and fractional Ca2+ release from,the SR and perinuclear stores in the presence of ET.

Solutions and drugsTyrode’s solution comprised: 140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2,10 mM HEPES, 10 mM D-glucose, pH 7.35 (NaOH). Endothelin-1 (ET), BQ-123(ETA antagonist), 1-[6-((17�-3-methoxyestra-1,3,5(10)-trien-17-yl)-amino)hexyl]-1H-pyrrole-2,5-dione (U-73122; PLC inhibitor), 2-aminoethoxy-diphenylborate (2-APB; Ins(1,4,5)P3 receptor blocker), xestospongin C (Ins(1,4,5)P3 receptor blocker),caffeine and isoproterenol (ISO) were from Calbiochem or Sigma.

StatisticsThe data are presented as mean ± s.e.m. Differences between data sets were evaluatedby ANOVA or Student’s t-test and considered significant when P<0.05.

We are grateful for financial support from the Ernst und BertaGrimmke-Stiftung (J.K. and B.P.) and the DeutscheForschungsgemeinschaft [DFG PI 414/1 and KFO 155, TP5 (L.S.M.)and TP6 (B.P. and J.K.)].

ReferencesBare, D. J., Kettlun, C. S., Liang, M., Bers, D. M. and Mignery, G. A. (2005). Cardiac

type 2 inositol 1,4,5-trisphosphate receptor: interaction and modulation bycalcium/calmodulin-dependent protein kinase II. J. Biol. Chem. 280, 15912-15920.

Bassani, J. W., Bassani, R. A. and Bers, D. M. (1993). Twitch-dependent SR Caaccumulation and release in rabbit ventricular myocytes. Am. J. Physiol. 265, C533-C540.

Bkaily, G., Choufani, S., Sader, S., Jacques, D., d’Orleans-Juste, P., Nader, M.,Kurban, G. and Kamal, M. (2003). Activation of sarcolemma and nuclear membranesET-1 receptors regulates transcellular calcium levels in heart and vascular smoothmuscle cells. Can. J. Physiol. Pharmacol. 81, 654-662.

Boivin, B., Chevalier, D., Villeneuve, L. R., Rousseau, E. and Allen, B. G. (2003).Functional endothelin receptors are present on nuclei in cardiac ventricular myocytes.J. Biol. Chem. 278, 29153-29163.

Bootman, M. D., Harzheim, D., Smyrnias, I., Conway, S. J. and Roderick, H. L.(2007). Temporal changes in atrial EC-coupling during prolonged stimulation withendothelin-1. Cell Calcium 42, 489-501.

Bultynck, G., Sienaert, I., Parys, J. B., Callewaert, G., De Smedt, H., Boens, N.,Dehaen, W. and Missiaen, L. (2003). Pharmacology of inositol trisphosphate receptors.Pflugers Arch. 445, 629-642.

Cardenas, C., Liberona, J. L., Molgo, J., Colasante, C., Mignery, G. A. and Jaimovich,E. (2005). Nuclear inositol 1,4,5-trisphosphate receptors regulate local Ca2+ transientsand modulate cAMP response element binding protein phosphorylation. J. Cell Sci. 118,3131-3140.

DeSantiago, J., Maier, L. S. and Bers, D. M. (2002). Frequency-dependent accelerationof relaxation in the heart depends on CaMKII, but not phospholamban. J. Mol. Cell.Cardiol. 34, 975-984.

Endoh, M., Fujita, S., Yang, H. T., Talukder, M. A., Maruya, J. and Norota, I. (1998).Endothelin: receptor subtypes, signal transduction, regulation of Ca2+ transients andcontractility in rabbit ventricular myocardium. Life Sci. 62, 1485-1489.

Genka, C., Ishida, H., Ichimori, K., Hirota, Y., Tanaami, T. and Nakazawa, H. (1999).Visualization of biphasic Ca2+ diffusion from cytosol to nucleus in contracting adultrat cardiac myocytes with an ultra-fast confocal imaging system. Cell Calcium 25, 199-208.

Ibarra, C., Estrada, M., Carrasco, L., Chiong, M., Liberona, J. L., Cardenas, C., Diaz-Araya, G., Jaimovich, E. and Lavandero, S. (2004). Insulin-like growth factor-1induces an inositol 1,4,5-trisphosphate-dependent increase in nuclear and cytosoliccalcium in cultured rat cardiac myocytes. J. Biol. Chem. 279, 7554-7565.

Kockskamper, J., Sheehan, K. A., Bare, D. J., Lipsius, S. L., Mignery, G. A. andBlatter, L. A. (2001). Activation and propagation of Ca (2+) release during excitation-contraction coupling in atrial myocytes. Biophys. J. 81, 2590-2605.

Li, X., Zima, A. V., Sheikh, F., Blatter, L. A. and Chen, J. (2005). Endothelin-1-inducedarrhythmogenic Ca2+ signaling is abolished in atrial myocytes of inositol-1,4,5-trisphosphate (IP3)-receptor type 2-deficient mice. Circ. Res. 96, 1274-1281.

Jour

nal o

f Cel

l Sci

ence

195Endothelin enhances nuclear Ca2+

Lipp, P., Laine, M., Tovey, S. C., Burrell, K. M., Berridge, M. J., Li, W. and Bootman,M. D. (2000). Functional InsP3 receptors that may modulate excitation-contractioncoupling in the heart. Curr. Biol. 10, 939-942.

Luo, D., Yang, D., Lan, X., Li, K., Li, X., Chen, J., Zhang, Y., Xiao, R. P., Han, Q.and Cheng, H. (2007). Nuclear Ca (2+) sparks and waves mediated by inositol 1,4,5-trisphosphate receptors in neonatal rat cardiomyocytes. Cell Calcium doi:10.1016/j.ceca.2007.04.017.

Mackenzie, L., Bootman, M. D., Berridge, M. J. and Lipp, P. (2001). Predeterminedrecruitment of calcium release sites underlies excitation-contraction coupling in rat atrialmyocytes. J. Physiol. 530, 417-429.

Mackenzie, L., Bootman, M. D., Laine, M., Berridge, M. J., Thuring, J., Holmes, A.,Li, W. H. and Lipp, P. (2002). The role of inositol 1,4,5-trisphosphate receptors in Ca(2+) signalling and the generation of arrhythmias in rat atrial myocytes. J. Physiol. 541,395-409.

Meyer, M., Lehnart, S., Pieske, B., Schlottauer, K., Munk, S., Holubarsch, C., Just,H. and Hasenfuss, G. (1996). Influence of endothelin 1 on human atrial myocardium-myocardial function and subcellular pathways. Basic Res. Cardiol. 91, 86-93.

Pieske, B., Beyermann, B., Breu, V., Loffler, B. M., Schlotthauer, K., Maier, L. S.,Schmidt-Schweda, S., Just, H. and Hasenfuss, G. (1999). Functional effects ofendothelin and regulation of endothelin receptors in isolated human nonfailing andfailing myocardium. Circulation 99, 1802-1809.

Proven, A., Roderick, H. L., Conway, S. J., Berridge, M. J., Horton, J. K., Capper, S.J. and Bootman, M. D. (2006). Inositol 1,4,5-trisphosphate supports thearrhythmogenic action of endothelin-1 on ventricular cardiac myocytes. J. Cell Sci. 119,3363-3375.

Remus, T. P., Zima, A. V., Bossuyt, J., Bare, D. J., Martin, J. L., Blatter, L. A., Bers,D. M. and Mignery, G. A. (2006). Biosensors to measure inositol 1,4,5-trisphosphateconcentration in living cells with spatiotemporal resolution. J. Biol. Chem. 281, 608-616.

Schillinger, W., Janssen, P. M., Emami, S., Henderson, S. A., Ross, R. S., Teucher, N.,Zeitz, O., Philipson, K. D., Prestle, J. and Hasenfuss, G. (2000). Impaired contractile

performance of cultured rabbit ventricular myocytes after adenoviral gene transfer ofNa (+)-Ca (2+) exchanger. Circ. Res. 87, 581-587.

Soeller, C., Jacobs, M. D., Jones, K. T., Ellis-Davies, G. C., Donaldson, P. J. andCannell, M. B. (2003). Application of two-photon flash photolysis to revealintercellular communication and intracellular Ca2+ movements. J. Biomed. Opt. 8, 418-427.

Sugden, P. H. (2003). An overview of endothelin signaling in the cardiac myocyte. J. Mol.Cell. Cardiol. 35, 871-886.

Tanaka, H., Kawanishi, T., Kato, Y., Nakamura, R. and Shigenobu, K. (1996a).Restricted propagation of cytoplasmic Ca2+ oscillation into the nucleus in guinea pigcardiac myocytes as revealed by rapid scanning confocal microscopy and indo-1. Jpn.J. Pharmacol. 70, 235-242.

Tanaka, H., Kawanishi, T., Matsuda, T., Takahashi, M. and Shigenobu, K. (1996b).Intracellular free Ca2+ movements in cultured cardiac myocytes as shown by rapidscanning confocal microscopy. J. Cardiovasc. Pharmacol. 27, 761-769.

Wu, X. and Bers, D. M. (2006). Sarcoplasmic reticulum and nuclear envelope are onehighly interconnected Ca2+ store throughout cardiac myocyte. Circ. Res. 99, 283-291.

Wu, X., Zhang, T., Bossuyt, J., Li, X., McKinsey, T. A., Dedman, J. R., Olson, E. N.,Chen, J., Brown, J. H. and Bers, D. M. (2006). Local InsP3-dependent perinuclearCa2+ signaling in cardiac myocyte excitation-transcription coupling. J. Clin. Invest. 116,675-682.

Yamada, J., Ohkusa, T., Nao, T., Ueyama, T., Yano, M., Kobayashi, S., Hamano, K.,Esato, K. and Matsuzaki, M. (2001). Up-regulation of inositol 1,4,5 trisphosphatereceptor expression in atrial tissue in patients with chronic atrial fibrillation. J. Am. Coll.Cardiol. 37, 1111-1119.

Zima, A. V. and Blatter, L. A. (2004). Inositol-1,4,5-trisphosphate-dependent Ca (2+)signalling in cat atrial excitation-contraction coupling and arrhythmias. J. Physiol. 555,607-615.

Zima, A. V., Bare, D. J., Mignery, G. A. and Blatter, L. A. (2007). IP3-dependent nuclearCa2+ signalling in the mammalian heart. J. Physiol. 584, 601-611.

Jour

nal o

f Cel

l Sci

ence