Journal of Molecular and Cellular Cardiology

Transcription

1 Journal of Molecular and Cellular Cardiology 46 (2009) Contents lists available at ScienceDirect Journal of Molecular and Cellular Cardiology journal homepage: Original article Molecular composition and functional properties of f-channels in murine embryonic stem cell-derived pacemaker cells Andrea Barbuti, Alessia Crespi, Daniela Capilupo, Nausicaa Mazzocchi, Mirko Baruscotti, Dario DiFrancesco Department of Biomolecular Sciences and Biotechnology, Laboratory of Molecular Physiology and Neurobiology, University of Milano, via Celoria 26, Milano, Italy article info abstract Article history: Received 17 September 2008 Received in revised form 14 November 2008 Accepted 3 December 2008 Available online 11 December 2008 Keywords: Hyperpolarization-activated cation channel Embryonic stem cells Pacemaker channels β-adrenergic receptors Muscarinic receptors f-channels Mouse embryonic stem cells (mescs) differentiate into all cardiac phenotypes, and thus represent an important potential source for cardiac regenerative therapies. Here we characterize the molecular composition and functional properties of funny (f-) channels in mesc-derived pacemaker cells. Following differentiation, a fraction of mesc-derived myocytes exhibited action potentials characterized by a slow diastolic depolarization and expressed the I f current. I f plays an important role in the pacemaking mechanism of these cells since ivabradine (3 μm), a specific f-channel inhibitor, inhibited I f by about 50% and slowed rate by about 25%. Analysis of I f kinetics revealed the presence of two populations of cells, one expressing a fast- and one a slow-activating I f ; the two components are present both at early and late stages of differentiation and had also distinct activation curves. Immunofluorescence analysis revealed that HCN1 and HCN4 are the only isoforms of the pacemaker channel expressed in these cells. Rhythmic cells responded to β-adrenergic and muscarinic agonists: isoproterenol (1 μm) accelerated and acetylcholine (0.1 μm) slowed spontaneous rate by about 50 and 12%, respectively. The same agonists caused quantitatively different effects on I f : isoproterenol shifted activation curves by about 5.9 and 2.7 mv and acetylcholine by 4.0 and 2.0 mv in slow and fast I f -activating cells, respectively. Accordingly, β1- and β2-adrenergic, and M2-muscarinic receptors were detected in mesc-derived myocytes. Our data show that mesc-derived pacemaker cells functionally express proteins which underlie generation and modulation of heart rhythm, and can therefore represent a potential cell substrate for the generation of biological pacemakers Elsevier Inc. All rights reserved. 1. Introduction The pacemaker of the heart is located in the sinoatrial node (SAN), a region structurally and functionally different from the rest of the myocardium. Its function is to generate spontaneous action potentials and drive the heartbeat with a mechanism allowing fine tuning of rate by the autonomic nervous system [1 3]. Sinoatrial myocytes and autorhythmic cells in general (like embryonic ventricular myocytes) exhibit action potentials with a slow diastolic depolarization, which at the termination of an action potential drives the membrane potential to the threshold for the next one. It is established that an important role in initiating the diastolic depolarization and modulating its rate, hence the rate of spontaneous activity, is played by the pacemaker I f current [4,5]. I f flows through Hyperpolarizationactivated Cyclic Nucleotide-gated (HCN) channels, the molecular correlates of f-channels; four HCN isoforms have been identified so far which are all expressed with different densities in different regions of the heart [6 8]. Corresponding author. Tel.: ; fax: address: andrea.barbuti@unimi.it (A. Barbuti). Evidence confirming the basic function of HCN channels in the generation, development and modulation of pacemaker activity has recently been provided in mice and humans [9 14]. The specific role of f/hcn channels in pacemaking makes their properties essential in the development of new tools aiming to heart rate control, such as the biological pacemakers [15]. Overexpression of HCN channels can indeed provide a depolarizing stimulus sufficient to induce, in vitro and in vivo, quiescent myocytes to beat spontaneously [16 20]. An alternative approach proven to be effective in pacing the heart has employed human embryonic stem cell-derived autorhythmic agglomerates [21,22]. Although the mechanism by which ES cells can pace silent cardiac tissue has not been fully elucidated, it is long known that ES cells can differentiate toward a cardiac pacemaker phenotype [23 26]. While it is well established that ESC-derived cardiomyocytes express the I f current [23,26,27] and that HCN channels, whether native or overexpressed, underlie generation of spontaneous activity [28,29], a detailed analysis of the HCN isoforms expressed in ESCderived cells is still lacking. In this work, we have analyzed the HCN composition of both spontaneously beating embryoid bodies (EBs) and isolated autorhythmic cells. Further, we have analyzed the basic properties of I f, its involvement in pacemaking and its modulation by /$ see front matter 2008 Elsevier Inc. All rights reserved. doi: /j.yjmcc

2 344 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) neurotransmitters and which type of G-protein-coupled receptors underlie the I f -mediated autonomic response. 2. Materials and methods 2.1. Cell culture Mouse ES cells (D3 line, ATCC) were cultured on a feeder layer of mitomycin C (0.01 mg/ml, Sigma) -treated mouse embryonic fibroblasts (STO, ATCC). ES cells culture, differentiation and isolation protocols are detailed in the online Supplementary Methods Electrophysiology Spontaneous action potentials and I f current were recorded by the patch-clamp technique in the whole-cell configuration. For details on solution and for data analysis, see Supplementary Methods Immunofluorescence and video-confocal analysis For immunofluorescence experiments, EBs, and single cells, at various stages of differentiation were plated onto coverslips coated with poly-l-lysine solution (0.01% v/v, Sigma). For details see Supplementary Methods. Fluorescence staining was analyzed by Video Confocal microscopy (ViCo Nikon). Control experiments with secondary antibodies only were carried out for all types of antibodies used, and resulted in no staining. 3. Results It is known that ESCs can differentiate spontaneously into cardiac myocytes with electrical properties typical of either the working myocardium or the conduction system [23 25]. In order to characterize the functional and molecular features of autorhythmic, pacemakerlike cells derived from murine ESC, we induced cell differentiation by a procedure based on the formation of compact cell aggregates known as embryoid bodies (EBs), as described in the Methods. During the differentiating process, portions of the EBs started to contract spontaneously, suggesting that some of the cells had differentiated toward a cardiac pacemaker phenotype. The number of contracting EBs increased with time and by day 7+7 of differentiation, around 80% of EBs presented one or more foci of spontaneously contracting cells. The number of contracting EBs started to decrease at day 7+13 and by day 7+22 only 20% of them was still beating (Fig.1A). To study the electrical properties of spontaneously active cells, we dissected and plated the contracting portions of the EBs at various stages of differentiation (from day 7+3 to day 7+20) [24]. In Figs. 1B, C the activity recorded from a spontaneously contracting portion of an EB dissociated at day 7+ 8 is shown (expanded scale in the lower panel see also Supplementary Online Video 1). Repetitive action potentials were characterized by a pronounced diastolic depolarization phase and the absence of a plateau phase, distinctive features of action potentials of cardiac pacemaker myocytes [1]. In n =16 contracting portions of EBs, the mean maximum diastolic potential (MDP) was 60.4±1.8 mv and the mean spontaneous rate was 264 ±38 beats per minute (bpm). These values are similar to those previously reported in isolated murine sinoatrial myocytes [30,31]. HCN channels are the molecular correlates of native pacemaker f- channels, whose main role in underlying generation of diastolic depolarization and control of spontaneous rate is well established [4]. Of the four known isoforms (HCN1-4), HCN1, HCN2 and HCN4 contribute to different degrees to the I f current in heart, HCN4 being the most highly expressed isoform in pacemaker SAN myocytes [32]. We checked for the presence in ESC-derived cells of the mrna of all four known HCN subunits. RT-PCR analysis revealed that Fig. 1. Electrical activity of a spontaneously beating EB-derived cell aggregate. (A) Bar graph plotting the mean fractions of contracting EBs at various times during differentiation. Data are mean±sem with n varying from 3 to 13. (B) Top, spontaneous and regular action potentials recorded from a cell cluster dissociated from a 7+8-day EB (see also Supplementary Online Video 1) in control Tyrode solution at T=36 C. B, bottom, the same recording plotted on an expanded time scale reveals a marked slow pacemaker (or diastolic ) depolarization phase between consecutive action potentials. undifferentiated ES cells expressed the mrna of all HCN subunits; we also found expression of all HCN subunits in EBs at day 7+8 of differentiation (Supplementary Fig. 1). Since RT-PCR analysis cannot quantify levels of mrna, nor directly correlate them with the level of protein expressed, we carried out immunofluorescence experiments in whole EBs, at the various stages of differentiation, using isoform-specific HCN antibodies. In order to identify portions of EBs rich in myocytes, we co-labelled EBs with antibodies against muscular/cardiac specific proteins such as caveolin 3. We found that HCN1 and HCN4 were the only HCN subunits detectable at early (day 7+3 Figs. 2A, B), intermediate (day 7+8; Figs. 2C, D) and late stage (day 7+20, Figs. 2E, F) of differentiation. HCN staining was detected exclusively in portions of EBs positive for caveolin 3, except for the early stages where caveolin 3 was never detected (Figs. 2A, B, right) The HCN3 subunit was detected rarely and only in caveolin 3- negative portions of EBs at day 7+8 (Supplementary Fig. 2B); we never detected HCN3 in EBs at day 7+3 (data not shown) and at day (Supplementary Fig. 2D). The HCN2 signal was never detected independently of the differentiation stage (Supplementary Figs. 2A, C). In control experiments aimed to verify antibody functionality, clear staining was detected in CHO cells transiently transfected with the mhcn2 isoform (Supplementary Fig. 3). We evaluated the electrical properties of single spontaneously active cells isolated from contractile portions of EBs from early (day 7+3 to 7 +6), intermediate (day 7+7 to 7 +10) and late (day 7+12 to 7+21) stage of differentiation. Figs. 3A C illustrates the properties of a representative autorhythmic cell (see Supplementary Online Video 2). Spontaneous action potentials recorded from this cell had a pronounced slow diastolic depolarization typical of pacemaker cells expressing I f (panel A). To investigate the expression of I f, we carried out voltage clamp experiments by applying hyperpolarizing voltage steps from the holding potential of 35 mv to the range 45/

3 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) Fig. 2. HCN1 and HCN4 expression in whole EBs. Single confocal sections of whole EBs double-labelled with specific anti-hcn antibodies (left panels, red) and caveolin 3 (right panels, green) at day 7+3 (A, B), 7+8 (C, D) and 7+20 (E, F) of differentiation. Nuclei were stained with DAPI. Calibration bars 40 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 125 mv. These steps elicited a large I f component (conductance of ps/pf, panel B) which was typically almost fully blocked on hyperpolarization by 5 mm CsCl (panel C). We next studied the expression of the various HCN isoforms at the single cell level, using caveolin 3 as a muscular/cardiac differentiation marker. In agreement with the whole EBs data, we found that HCN1

4 346 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) Fig. 3. Spontaneous activity, I f current and HCN distribution in isolated ESC-derived beating cells. (A) Spontaneous action potentials recorded from an isolated ESC-derived cell (see Supplementary Online Video 2), displaying features of a typical SAN action potential. (B) Family of I f traces recorded from the same cell during an activation curve protocol consisting of voltage steps to the range 45/ 125 (20 mv increments) followed by a fully activating step to 125 mv. (C) I f traces elicited by a two-pulse protocol to 75/ 125 mv from a holding potential of 35 mv before and during perfusion with 5 mm CsCl to block I f. (D, E) Single confocal sections of isolated ESC-derived cells double-labelled with anti-hcn antibodies, as indicated (D, red) and caveolin 3 (E, green). Nuclei stained with DAPI. Calibration bars 10 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) and HCN4 signals (Fig. 3D, red) were readily detected on the membrane of cells expressing caveolin 3 (Fig. 3E, green); HCN3 was detected in a few cells, none of which co-expressed caveolin 3, while HCN2 was never detected in either caveolin 3 positive or negative cells. Immunofluorescence experiments were then performed to check whether the isoforms HCN1 and HCN4 were co-expressed in the same cells by a double-staining procedure with anti-hcn1 and anti-hcn4 antibodies. The left panels of Figs. 4A to C show a portion of a 7+8 EB in which both isoforms HCN4 (A, red) and HCN1 (B, green) were detected; merging of the two signals (C, yellow) indicates that some of the cells expressed both isoforms, while other cells expressed mostly HCN4 or HCN1. Similar experiments were then performed in isolated cells. A representative cell (from a 7+12 EB) showing partial co-localization (yellow) of HCN4 (red) and HCN1 (green) is shown in the right panels of Figs. 4D to F. Taken together, these data show that in ESC-derived myocytes, the only isoforms expressed up to detectable levels are HCN4 and HCN1, i.e. the same isoforms expressed in the SAN of different species [6 8,33,34]. In a series of similar experiments carried out onto undifferentiated ES cells, we only detected HCN4 at a low level of expression in a small fraction of cells (7 out of 64); accordingly, we did not record any substantial I f current in undifferentiated cells (n =30, see also Supplementary Fig. 4). To evaluate the contribution of I f current to spontaneous activity we investigated the action of the highly selective f-channel inhibitor ivabradine. In Fig. 5A, spontaneous action potentials recorded from a single cell in control solution and during perfusion with 3 μm ivabradine are shown. The drug slowed activity from 264 bpm to 174 bpm ( 34%) by specifically reducing the steepness of the diastolic depolarization. On average ivabradine reduced rate of ESC-derived pacemaker myocytes by 24.9±4.7% (n=7). Fig. 5B shows representative I f traces recorded at 95 mv before and during application of 3 μm ivabradine, when current reduction had reached steady-state. On average ivabradine blocked I f current by 50.0±8.3% (n=4). The identification of HCN1 and HCN4 prompted us to investigate in greater detail the kinetic properties of I f recorded from isolated ESCderived autorhythmic cells. For all stages of differentiations, we found that in some cells I f activated with slow kinetics, while in others the kinetics were clearly faster. In Fig. 6A representative recordings from two such cells are superimposed after scaling.

5 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) Fig. 4. Membrane co-localization of HCN4 and HCN1 isoforms. Left, confocal images of an EB double-labelled with anti-hcn4 (A, red) and anti-hcn1 (B, green) primary antibodies. (C), merged images showing signal co-localization (yellow). Right panels, single sections of an isolated ESC-derived cell showing the specific signals for HCN4 (D, red) and HCN1 (E, green) and after merging to see the extent of signal co-localization (F, yellow). Calibration bars 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Fig. 5. Action of ivabradine on spontaneous action potentials and I f. (A) Spontaneous action potentials recorded from an isolated ESC-derived cell before and during superfusion of 3 μm ivabradine, when rate slowing had reached steady-state. (B) I f traces elicited by stepping to 95 mv from a holding potential of 35 mv during control and after steady state block by 3 μm ivabradine. Plotting in Fig. 6B the time constants from all cells analyzed (n=20) suggested the presence of two separate families of cells with activation constants converging toward two distinct curves, a fast- (n=10, open circles) and a slow-kinetic curve (n=10, filled circles). The mean time constant curves from these two families were significantly different at all voltages (not shown, pb0.05). Although immunofluorescence data show three populations of cells, this is not necessarily in contrast with the above kinetic analysis. Our electrophysiological data show indeed that cells tend to cluster around two distinct populations, with predominance of either the slow or the fast kinetic component, but within each population there is some variability in the time constants which suggests the existence of cells with a mixed HCN1 and HCN4 expression. The bar graph in Fig. 6C shows the fraction of cells with a fast- (blank) or slow-activating (diagonal pattern) I f current at various stages of differentiation; while the fast I f component is predominant at early stages, the slow one predominates at later stages. The existence of two distinct cell populations was further supported by the difference between mean activation curves of cells from the same two families, as shown in Fig. 6D. The mean half activation voltages (V 1/2 ) were 72.1±1.3 mv (n=14, open circles) and 79.4±1.6 mv (n=11, filled circles) for cells with fast and slow activation kinetics, respectively (significantly different, pb0.05); inverse slope factors were not significantly different (7.2±0.5 mv and 8.0±0.8 mv for fast and slow I f -kinetics cells, respectively). As shown in Fig. 6E, the V 1/2 of fast and slow I f did not change significantly with the differentiation stage. On the contrary, the I f conductance density did change, increasing from 0.013±0.002 ps/pf to 0.049±0.007 and 0.077±0.009 ps/pf at

6 348 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) Fig. 6. Two types of ESC-derived pacemaker cells expressing different I f kinetics. (A) I f traces recorded at 75 mv in two different cells normalized to maximal size to highlight the different activation kinetics. (B) Plot of the voltage dependence of time constants recorded from cells with fast-activating (empty circles) and slow-activating I f (filled circles); lines are through points. (C) Bar graph showing the fraction of cells expressing a fast- (blank) or slow-activating I f (line pattern) at different stages of differentiation. (D) Plot of mean activation curves for the fast-activating (empty circles) and the slow-activating I f (filled circles). Full lines represent best fittings by the Boltzmann equation (see text for best-fitting parameters). (E) Plot of the mean V 1/2 of fast- and slow-activating I f during differentiation. (F) Bar graph of the mean I f conductance density at different times; the asterisk indicate a significant difference by one-way ANOVA. early, intermediate and late stages of differentiation, respectively (Fig. 6F, significant, pb0.05). In isolated beating cells, we measured spontaneous rate as the mean from at least 20 second recordings and averaged values for the three stages of differentiation. Mean rates were 270±67 bpm (n=7), 292±44 bpm (n=10) and 281±89 bpm (n=4) at early, intermediate and late stages of differentiation, respectively, and were not significantly different (pn0.05, one-way ANOVA). Since the observed increase in I f density should give rise to increased rate, we speculate that development may also be associated with increased density of other ion currents and specifically of K + currents, which would counteract the acceleratory action of I f density increase. Among cells dissociated at all stages, we found myocytes that fired action potentials only when stimulated (see Supplementary Fig. 5A). Action potentials from these cells differed substantially from those of autorhythmic cells and looked more like those of atrial/ventricular myocytes (mean resting potential= 73.5±7.7 mv). Correspondingly, there was essentially no I f in these cells (mean conductance density was ± ps/pf, n =6; Supplementary Fig. 5B). An important property of pacemaker cells is their I f -dependent rate modulation by autonomic neurotransmitters. β-adrenergic agonists accelerate spontaneous rate by steepening the phase 4 depolarization through a positive shift of I f current activation curve, while muscarinic agonists slow rate by the opposite process [1,5]. To investigate this mechanism in ESC-derived autorhythmic cells, we first checked if they responded to autonomic transmitters. Either cell aggregates or single cells were superfused with the β-adrenergic agonist isoproterenol (Iso, 1 μm) and/or the muscarinic agonist acetylcholine (ACh, 0.1 μm) while recording spontaneous activity. The activity of two representative cells recorded in the control solution and during superfusion with Iso (left) or ACh (right) is shown in Fig. 7A. Iso accelerated, and ACh slowed rate. In both cases the rate change was due to a modification in the slope of the slow pacemaker depolarization, where I f plays an important role, with little or no modifications of action potential shape and duration. On average, Iso (1 μm) accelerated and ACh (0.1 μm) slowed rate by 50.0±7.7% (n=4) and 12.3±5.5% (n=4), respectively (see Supplementary Online Video 3). We next studied the effects of the same agonists on I f kinetics in the two populations of cells with either fast or slow I f -activation kinetics. In Figs. 7B, C, typical fast- (B) and slow-activating (C) I f traces recorded at 75 mv are shown in control conditions and during superfusion with Iso (1 μm, left) and ACh (0.1 μm, right). We measured the shift of I f activation curve induced by the agonists according to a method described previously [35]. The two populations showed, on average, significantly different responses to agonist stimulation. Iso induced shifts of +2.7±0.6 mv (fast, n=7) and +5.9±0.4 mv (slow, n=14) while ACh induced shifts of 2.0±0.7 mv (fast, n =6) and 4± 0.5 mv (slow, n =10) (pb0.05). The response to autonomic neurotransmitters implies a functional signal transduction pathway involving activation of specific G proteincoupled receptors. We therefore investigated if ESC-derived autorhythmic cells express the same types of receptors that in native pacemaker myocytes initiate β-adrenergic and M2 muscarinic modulatory pathways. Typical results of immunofluorescence analysis of single cells performed with antibodies against β-adrenergic and muscarinic receptors are shown in Fig. 8. We found that most caveolin 3-positive cells reveal a strong expression of β 1 adrenergic (panel A) and M2 muscarinic receptors (panel C), while β 2 -adrenergic receptors are expressed less frequently and in general more weakly (panel B). 4. Discussion We have investigated the molecular and functional properties of the funny (I f ) current expressed in murine ESC-derived autorhythmic myocytes. The observation that pacemaker myocytes are formed during the development of embryoid bodies was established in early studies of embryonic stem cell differentiation. Pacemaker cells can be identified by the fact that they generate spontaneous action potentials with features similar to those of SAN cells, including the pacemaker depolarization phase of the action potential [23 25]. In native mammalian tissue, generation of pacemaker activity in the SAN is Fig. 7. Modulation of rate and the I f current by autonomic transmitters. (A) Spontaneous action potentials recorded from representative cell aggregates in Tyrode solution (ctr) and during perfusion with isoproterenol 1 μm (Iso, left) or acetylcholine 0.1 μm (ACh, right). (B, C) Cells expressing fast- (B) or slow-activating I f (C) were exposed first to Iso 1 μm (left) and then to ACh 0.1 μm (right); I f traces recorded during hyperpolarization to 75 mv from a holding potential of 35 mv.

7 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) Fig. 8. Distribution of β-adrenergic and muscarinic receptors in single ESC-derived cells. (A-C) Single confocal sections of isolated ESC-derived cells labelled with anti-β1 (A, red) or anti-β2 (B, red) adrenergic receptor antibodies (β-ar), or with anti-m2 muscarinic receptor antibody (C, red; M2-mAChR); cells were also co-imunolabelled with anti-caveolin 3 antibody (right panels, cav3, green). Nuclei stained with DAPI. Calibration bars 10 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) under control of funny channels [1,32], whose molecular correlates are mainly HCN4 subunits, with a limited species-dependent contribution of HCN1 and/or HCN2 subunits [6 8,33,34]. However the mrnas of the four known HCN isoforms are by no means exclusive of sinoatrial pacemaker cells, and are differently distributed in various cardiac regions tissues as well as in other excitable cells [36,37]. Our RT-PCR data indicate expression of all four HCN isoforms both in undifferentiated ES cells and in differentiated EBs. Since EBs recapitulate embryonic development, this result is not surprising; it is known that cells derived from all three embryonic germ layers are present in differentiating EBs, including neuronal cells and other cell types which possibly express HCN channels [38 41]. There is some variability in the HCN expression in ESC-derived myocytes according to previously published work. In one study, ESCderived cardiomyocytes were reported to express all four isoforms, with HCN2 and HCN3 representing the main isoforms [29,42], while in another study differentiating ESC were shown to express only HCN1 and HCN4 [43]. Variable results have also been published on the expression of HCN isoforms in undifferentiated ESCs. According to one report HCN1 and HCN4 are the major isoforms [43], while in another report only HCN2 and HCN3 were found [44]; in this latter study a Cs + - sensitive hyperpolarization-activated current was also recorded in about 30% of undifferentiated ES cells. In our study, PCR data indicated the presence of the mrna of all four HCN isoforms in undifferentiated cells. We failed however to record any substantial I f -like current in these cells, indicating lack of functional channels correctly inserted into cell membranes. Lack of a functional role of HCN channels in undifferentiated ES cells is consistent with the observation that these cells have resting voltage levels around 20 mv, at which no HCN channel is functional under normal conditions. The mrna of all HCN isoforms could simply reflect the presence of a small number of cells undergoing early stages of differentiation among the undifferentiated ones, which were detected due to the high resolution power of the PCR technique. The discrepancy between expression of mrna and functional proteins can also result from post-translational regulatory mechanisms such as control by micrornas (mirnas); it is known for example that expression of HCN2 is controlled by two muscle-specific mirnas (mir-i33 and mir-i), the latter controlling also expression of the HCN4 isoform [45,46]. Since RT-PCR data did not provide evidence discriminating between HCN expression at various stages of ES cell differentiation, we choose to investigate protein expression. It has been reported recently that ESC-derived flk-1 + colonies display a diffuse staining for HCN1 and HCN4 when differentiated into cardiomyocytes [28]. Our present data provide the first evidence for membrane localization of HCN channels on the plasma membrane of isolated ESC-derived cardiomyocytes and of whole EBs. We have found a clear, membrane-delimited expression only of HCN1 and HCN4 isoforms in caveolin 3-positive ESC-derived cardiomyocytes. Caveolin 3 is a structural protein of muscular/cardiac-caveolae abundantly expressed in SAN cells; we have recently shown that in isolated rabbit SAN myocytes HCN4 co-localizes and interacts with caveolin 3 and that this interaction is critically important for both proper channel function and modulation [3]. Our present data support the hypothesis that localization and interaction of HCN channels with caveolin 3 may

8 350 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) be important during development/maturation of pacemaker myocytes. In the ESC-derived cells HCN3 was detected with a very low frequency only in caveolin3-negative cells, while HCN2 was never detected. Given that HCN2 is the predominant isoform in adult ventricular myocytes [6], our results indicate that the level of expression of this isoform is below the detectable threshold during early stages of cardiac differentiation, in agreement with previous reports [47]. HCN1 and HCN4 are the isoforms most highly expressed in the SAN of various species [8,33,34]. HCN4 in particular is considered to be a specific marker of pacemaker cells of the SAN, as inferred from in situ hybridization experiments showing that this isoform is expressed very early during cardiac embryogenesis and is confined to the pacemaker region that will form the mature SAN, while it is absent from atria and ventricles [48 50]. Correspondingly, there are several indications for the role of HCN channels in pacemaker activity. For example, mice specifically lacking cardiac HCN4 die during embryogenesis [9] or have an unstable heart rhythm when the gene is reported to be turned off in adulthood according to recent work [10]; also, mutations of the HCN4 gene were reported to be linked to rhythm disturbances [11,12] and have been shown more recently to cause sinus bradycardia in humans [13,14]. Expression of HCN4 and HCN1 in ESC-derived myocytes together with caveolin3 suggests therefore that these cells have acquired a pacemaker-like phenotype typical of SAN cells. The presence of the I f current in ESC-derived cardiomyocytes has been reported previously [23,27]. Recently, the role of I f in the generation of spontaneous activity during development of embryoid bodies has been investigated [28,29,51]. To assess a direct role of I f in the generation of spontaneous activity of ESC-derived cells we have used the specific f-channel inhibitor ivabradine. The extent of inhibition of I f and rate reduction caused by ivabradine 3 μm were similar to values reported in the SAN tissue and single SAN myocytes [52 56]. We have also found two distinct populations of ESC-derived, spontaneously beating cells expressing I f currents with different kinetic properties: in one cell type I f activated relatively fast (τ=0.5 s at 75 mv) with a half-activation voltage of 72.1 mv, while in another cell type it activated more slowly (τ=1.6 s at 75 mv) with a more negative half-activation voltage ( 79.4 mv). Differences in I f kinetics between the two cell populations could arise from a variety of factors, among which different subunit composition, specific interactions with ancillary subunits, differences in phosphorylative states and in the basal concentration of camp. Clearly however, different camp levels cannot be the only cause, since this would cause similar shifting effects on activation curve and τ curve. On the contrary, while V 1/2 values differ by 7.3 mv (Fig. 6D), a much larger shift would be necessary for superimposing the τ curves plotted in Fig. 6B (approximately 25 mv on mean τ curves, data not shown). The existence of two populations of pacemaker cells expressing either fast- or slow-activating I f has been previously reported in myocytes isolated from murine SAN [30], which are also known to express higher levels of HCN1 and HCN4 than HCN2 and HCN3 [7,8]. The comparison between our data and previous data suggests therefore a close similarity between ESC-derived autorhythmic cells and mature murine pacemaker cells concerning I f expression. An important characteristic of functional pacemaker cells is their ability to interact with the modulatory pathways controlling heart rate. Since in mature pacemaker cells a fundamental mechanism of rate modulation involves the control of the I f voltage activation range by autonomic neurotransmitters [1,32], we checked if a similar mechanism also operates in ESC-derived pacemaker cells. Indeed we found that in these cells, too, isoproterenol and acetylcholine shift the I f activation curve to more positive and to more negative voltages, and are correspondingly able to accelerate and slow spontaneous rate, respectively. Interestingly, cells with fast activating I f had a significantly smaller response to adrenergic and muscarinic agonists than cells with a slow activating I f. This, and the different activation kinetics illustrated in Fig. 7, suggest that the two pacemaker cell populations identified could correspond to cells with different levels of expression of the HCN1 and HCN4 isoforms. It is indeed well established that HCN1 channels, when expressed individually, have faster kinetics and lower camp sensitivity than HCN4 channels [32,57,58]. Comparing our data with data from HCN1/HCN4 channel expression in heterologous system at 35 C [59] indicates a close similarity between the kinetics of the slow-activating I f with those of homomeric HCN4 channels, and of the fast-activating I f with those of heteromeric HCN1 HCN4 channels. For example, time constants of activation at 75 mv are 1.57 and 0.5 s for slow- and fast-activating I f and about 1.5 and 0.5 s for HCN4 homomers and HCN1 HCN4 chimeric channels, respectively [59]. Thus, our data suggest that ESC-derived cells with slow-activating I f express predominantly homomeric HCN4 channels, while cells with fast-activating I f express heteromeric HCN1 HCN4 channels. In mature pacemaker myocytes, rate of spontaneous activity is controlled by autonomic transmitters via modulation of the campdependent pathway through specific membrane receptors, such as β 1 and β 2 adrenergic receptors and M2 muscarinic receptors. Our results provide the first evidence for the expression of these receptors in ESCderived caveolin 3-positive myocytes. We have recently shown that in rabbit SAN myocytes, β 2 -ARs are largely responsible for the I f -mediated increase in rate, while β 1 -ARs activation modulates rate and the I f current to a lesser extent [3]. According to our present data in ESC-derived myocytes, on the other hand, β 2 -ARs appear to be expressed at a low level and in only a few cells. This agrees with previous observation that the isoproterenolinduced acceleration of rate in spontaneously beating EBs is diminished upon perfusion of the selective β 1 -AR antagonist CGP 20712A, but is unchanged upon perfusion of the specific β 2 -reverse agonist ICI 118,551[60]. In conclusion, our results indicate that ES cells, even at very early stages, differentiate into cardiomyocytes which comprise a subpopulation of cells with functional and molecular characteristics typical of cardiac pacemaker cells. This includes ion channels required for pacemaker activity, as well as the biochemical pathways needed for neurotransmitter-mediated frequency modulation. Following proper enrichment through more stringent culture and/or differentiating conditions aimed to increase the yield of selected pacemaker myocytes, isolation of these cells could provide a substrate suitable for generation of ESC-based biological pacemakers. Acknowledgements This work was supported by European Union (Normacor), CARIPLO / and MIUR-FIRB (RBLA035A4X) grants to DD. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi: /j.yjmcc References [1] DiFrancesco D. Pacemaker mechanisms in cardiac tissue. Annu Rev Physiol 1993; 55: [2] Boyett MR, Honjo H, Kodama I. The sinoatrial node, a heterogeneous pacemaker structure. Cardiovasc Res 2000;47: [3] Barbuti A, Terragni B, Brioschi C, DiFrancesco D. Localization of f-channels to caveolae mediates specific beta2-adrenergic receptor modulation of rate in sinoatrial myocytes. J Mol Cell Cardiol 2007;42:71 8. [4] DiFrancesco D. Serious workings of the funny current. Prog Biophys Mol Biol 2006;90: [5] Barbuti A, Baruscotti M, DiFrancesco D. The pacemaker current: from basics to the clinics. J Cardiovasc Electrophysiol 2007;18:342 7.

9 A. Barbuti et al. / Journal of Molecular and Cellular Cardiology 46 (2009) [6] Shi W, Wymore R, Yu H, Wu J, Wymore RT, Pan Z, et al. Distribution and prevalence of hyperpolarization-activated cation channel (HCN) mrna expression in cardiac tissues. Circ Res 1999;85:e1 6. [7] Marionneau C, Couette B, Liu J, Li H, Mangoni ME, Nargeot J, et al. Specific pattern of ionic channel gene expression associated with pacemaker activity in the mouse heart. J Physiol 2005;562: [8] Liu J, Dobrzynski H, Yanni J, Boyett MR, Lei M. Organisation of the mouse sinoatrial node: structure and expression of HCN channels. Cardiovasc Res 2007;73: [9] Stieber J, Herrmann S, Feil S, Loster J, Feil R, Biel M, et al. The hyperpolarizationactivated channel HCN4 is required for the generation of pacemaker action potentials in the embryonic heart. Proc Natl Acad Sci U S A 2003;100: [10] Herrmann S, Stieber J, Stockl G, Hofmann F, Ludwig A. HCN4 provides a depolarization reserve and is not required for heart rate acceleration in mice. EMBO J 2007;26: [11] Schulze-Bahr E, Neu A, Friederich P, Kaupp UB, Breithardt G, Pongs O, et al. Pacemaker channel dysfunction in a patient with sinus node disease. J Clin Invest 2003;111: [12] Ueda K, Nakamura K, Hayashi T, Inagaki N, Takahashi M, Arimura T, et al. Functional characterization of a trafficking-defective HCN4 mutation, D553N, associated with cardiac arrhythmia. J Biol Chem 2004;279: [13] Milanesi R, Baruscotti M, Gnecchi-Ruscone T, DiFrancesco D. Familial sinus bradycardia associated with a mutation in the cardiac pacemaker channel. N Engl J Med 2006;354: [14] Nof E, Luria D, Brass D, Marek D, Lahat H, Reznik-Wolf H, et al. Point mutation in the HCN4 cardiac ion channel pore affecting synthesis, trafficking, and functional expression is associated with familial asymptomatic sinus bradycardia. Circulation 2007;116: [15] Rosen MR. Biological pacemaking: a concept whose time has come...or is coming. Heart 2007;93: [16] Qu J, Plotnikov AN, Danilo Jr P, Shlapakova I, Cohen IS, Robinson RB, et al. Expression and function of a biological pacemaker in canine heart. Circulation 2003;107: [17] Plotnikov AN, Sosunov EA, Qu J, Shlapakova IN, Anyukhovsky EP, Liu L, et al. Biological pacemaker implanted in canine left bundle branch provides ventricular escape rhythms that have physiologically acceptable rates. Circulation 2004;109: [18] Potapova I, Plotnikov A, Lu Z, Danilo Jr P, Valiunas V, Qu J, et al. Human mesenchymal stem cells as a gene delivery system to create cardiac pacemakers. Circ Res 2004;94: [19] Bucchi A, Plotnikov AN, Shlapakova I, Danilo Jr P, Kryukova Y, Qu J, et al. Wild-type and mutant HCN channels in a tandem biological-electronic cardiac pacemaker. Circulation 2006;114: [20] Cai J, Yi FF, Li YH, Yang XC, Song J, Jiang XJ, et al. Adenoviral gene transfer of HCN4 creates a genetic pacemaker in pigs with complete atrioventricular block. Life Sci 2007;80: [21] Kehat I, Khimovich L, Caspi O, Gepstein A, Shofti R, Arbel G, et al. Electromechanical integration of cardiomyocytes derived from human embryonic stem cells. Nat Biotechnol 2004;22: [22] Xue T, Cho HC, Akar FG, Tsang SY, Jones SP, Marban E, et al. Functional integration of electrically active cardiac derivatives from genetically engineered human embryonic stem cells with quiescent recipient ventricular cardiomyocytes: insights into the development of cell-based pacemakers. Circulation 2005;111: [23] Maltsev VA, Rohwedel J, Hescheler J, Wobus AM. Embryonic stem cells differentiate in vitro into cardiomyocytes representing sinusnodal, atrial and ventricular cell types. Mech Dev 1993;44: [24] Maltsev VA, Wobus AM, Rohwedel J, Bader M, Hescheler J. Cardiomyocytes differentiated in vitro from embryonic stem cells developmentally express cardiac-specific genes and ionic currents. Circ Res 1994;75: [25] Hescheler J, Fleischmann BK, Lentini S, Maltsev VA, Rohwedel J, Wobus AM, et al. Embryonic stem cells: a model to study structural and functional properties in cardiomyogenesis. Cardiovasc Res 1997;36: [26] Gassanov N, Er F, Zagidullin N, Hoppe UC. Endothelin induces differentiation of ANP-EGFP expressing embryonic stem cells towards a pacemaker phenotype. FASEB J 2004;18: [27] Abi-Gerges N, Ji GJ, Lu ZJ, Fischmeister R, Hescheler J, Fleischmann BK. Functional expression and regulation of the hyperpolarization activated non-selective cation current in embryonic stem cell-derived cardiomyocytes. J Physiol 2000;523(Pt 2): [28] Yanagi K, Takano M, Narazaki G, Uosaki H, Hoshino T, Ishii T, et al. Hyperpolarization-activated cyclic nucleotide-gated channels and T-type calcium channels confer automaticity of embryonic stem cell-derived cardiomyocytes. Stem Cells 2007;25: [29] Qu Y, Whitaker GM, Hove-Madsen L, Tibbits GF, Accili EA. Hyperpolarizationactivated cyclic nucleotide-modulated HCN channels confer regular and faster rhythmicity to beating mouse embryonic stem cells. J Physiol [30] Mangoni ME, Nargeot J. Properties of the hyperpolarization-activated current (I(f)) in isolated mouse sino-atrial cells. Cardiovasc Res 2001;52: [31] Cho HS, Takano M, Noma A. The electrophysiological properties of spontaneously beating pacemaker cells isolated from mouse sinoatrial node. J Physiol 2003;550: [32] Accili EA, Proenza C, Baruscotti M, DiFrancesco D. From funny current to HCN channels: 20 years of excitation. NIPS 2002;17:32 7. [33] Tellez JO, Dobrzynski H, Greener ID, Graham GM, Laing E, Honjo H, et al. Differential expression of ion channel transcripts in atrial muscle and sinoatrial node in rabbit. Circ Res 2006;99: [34] Yamamoto M, Dobrzynski H, Tellez J, Niwa R, Billeter R, Honjo H, et al. Extended atrial conduction system characterised by the expression of the HCN4 channel and connexin45. Cardiovasc Res 2006;72: [35] Accili EA, DiFrancesco D. Inhibition of the hyperpolarization-activated current (if) of rabbit SA node myocytes by niflumic acid. Pflugers Arch 1996;431: [36] Robinson RB, Siegelbaum SA. Hyperpolarization-activated cation currents: from molecules to physiological function. Annu Rev Physiol 2003;65: [37] Cerbai E, Mugelli A. I(f) in non-pacemaker cells: role and pharmacological implications. Pharmacol Res 2006;53: [38] Moosmang S, Biel M, Hofmann F, Ludwig A. Differential distribution of four hyperpolarization-activated cation channels in mouse brain. Biol Chem 1999;380: [39] Moosmang S, Stieber J, Zong X, Biel M, Hofmann F, Ludwig A. Cellular expression and functional characterization of four hyperpolarization-activated pacemaker channels in cardiac and neuronal tissues. Eur J Biochem 2001;268: [40] Doan TN, Stephans K, Ramirez AN, Glazebrook PA, Andresen MC, Kunze DL. Differential distribution and function of hyperpolarization-activated channels in sensory neurons and mechanosensitive fibers. J Neurosci 2004;24: [41] El Kholy W, MacDonald PE, Fox JM, Bhattacharjee A, Xue T, Gao X, et al. Hyperpolarization-activated cyclic nucleotide-gated channels in pancreatic betacells. Mol Endocrinol 2007;21: [42] White SM, Claycomb WC. Embryonic stem cells form an organized, functional cardiac conduction system in vitro. Am J Physiol, Heart Circ Physiol 2005;288: H [43] van Kempen M, van Ginneken A, de I G, Mutsaers N, Opthof T, Jongsma H, et al. Expression of the electrophysiological system during murine embryonic stem cell cardiac differentiation. Cell Physiol Biochem 2003;13: [44] Wang K, Xue T, Tsang SY, Van Huizen R, Wong CW, Lai KW, et al. Electrophysiological properties of pluripotent human and mouse embryonic stem cells. Stem Cells 2005;23: [45] Xiao J, Yang B, Lin H, Lu Y, Luo X, Wang Z. Novel approaches for gene-specific interference via manipulating actions of micrornas: examination on the pacemaker channel genes HCN2 and HCN4. J Cell Physiol 2007;212: [46] Luo X, Lin H, Pan Z, Xiao J, Zhang Y, Lu Y, et al. Down-regulation of mir-1/mir-133 contributes to re-expression of pacemaker channel genes HCN2 and HCN4 in hypertrophic heart. J Biol Chem 2008;283: [47] Yasui K, Liu W, Opthof T, Kada K, Lee JK, Kamiya K, et al. I(f) current and spontaneous activity in mouse embryonic ventricular myocytes. Circ Res 2001;88: [48] Garcia-Frigola C, Shi Y, Evans SM. Expression of the hyperpolarization-activated cyclic nucleotide-gated cation channel HCN4 during mouse heart development. Gene Expr Patterns 2003;3: [49] Christoffels VM, Mommersteeg MT, Trowe MO, Prall OW, Gier-de Vries C, Soufan AT, et al. Formation of the venous pole of the heart from an Nkx2 5-negative precursor population requires Tbx18. Circ Res 2006;98: [50] Mommersteeg MT, Hoogaars WM, Prall OW, Gier-de Vries C, Wiese C, Clout DE, et al. Molecular pathway for the localized formation of the sinoatrial node. Circ Res 2007;100: [51] Mery A, Aimond F, Menard C, Mikoshiba K, Michalak M, Puceat M. Initiation of embryonic cardiac pacemaker activity by inositol 1,4,5-trisphosphate-dependent calcium signaling. Mol Biol Cell 2005;16: [52] Bois P, Bescond J, Renaudon B, Lenfant J. Mode of action of bradycardic agent, S 16257, on ionic currents of rabbit sinoatrial node cells. Br J Pharmacol 1996;118: [53] Bucchi A, Baruscotti M, DiFrancesco D. Current-dependent block of rabbit sinoatrial node I(f) channels by ivabradine. J Gen Physiol 2002;120:1 13. [54] Bucchi A, Tognati A, Milanesi R, Baruscotti M, DiFrancesco D. Properties of ivabradine-induced block of HCN1 and HCN4 pacemaker channels. J Physiol 2006;572: [55] Bucchi A, Baruscotti M, Robinson RB, DiFrancesco D. Modulation of rate by autonomic agonists in SAN cells involves changes in diastolic depolarization and the pacemaker current. J Mol Cell Cardiol 2007;43: [56] Thollon C, Bedut S, Villeneuve N, Coge F, Piffard L, Guillaumin JP, et al. Usedependent inhibition of hhcn4 by ivabradine and relationship with reduction in pacemaker activity. Br J Pharmacol 2007;150: [57] Viscomi C, Altomare C, Bucchi A, Camatini E, Baruscotti M, Moroni A, et al. C terminus-mediated control of voltage and camp gating of hyperpolarizationactivated cyclic nucleotide-gated channels. J Biol Chem 2001;276: [58] Altomare C, Terragni B, Brioschi C, Milanesi R, Pagliuca C, Viscomi C, et al. Heteromeric HCN1 HCN4 channels: a comparison with native pacemaker channels from the rabbit sinoatrial node. J Physiol 2003;549: [59] Ishii TM, Takano M, Ohmori H. Determinants of activation kinetics in mammalian hyperpolarization-activated cation channels. J Physiol 2001;537: [60] Ali NN, Xu X, Brito-Martins M, Poole-Wilson PA, Harding SE, Fuller SJ. Betaadrenoceptor subtype dependence of chronotropy in mouse embryonic stem cellderived cardiomyocytes. Basic Res Cardiol 2004;99: