Developmental Changes in Cardiomyocytes Differentiated from ...

EMBRYONIC STEM CELLS Developmental Changes in Cardiomyocytes Differentiated from Human Embryonic Stem Cells: A Molecular and Electrophysiological Approach LAURA SARTIANI,a ESTHER BETTIOL,b FRANCESCA STILLITANO,a ALESSANDRO MUGELLI,a ELISABETTA CERBAI,a MARISA E. JACONIb a

Centro Interuniversitario di Medicina Molecolare e Biofisica Applicata, University of Firenze, Firenze, Italy; Biology of Aging Laboratory, Department of Rehabilitation and Geriatrics, University Hospitals of Geneva, Geneva, Switzerland

b

Key Words. Embryonic stem cells • Cardiac differentiation • Ion current • Maturation • Ion channel subunits • Multicellular recordings

ABSTRACT Cardiomyocytes derived from human embryonic stem cells constitute a promising cell source for the regeneration of damaged hearts. The assessment of their in vitro functional properties is mandatory to envisage appropriate cardiac cell-based therapies. In this study, we characterized human embryonic stem cell-derived cardiomyocytes over a 3-month period, using patch-clamp or intracellular recordings to assess their functional maturation and reverse transcriptase-polymerase chain reaction to evaluate the expression of ion channel-encoding subunits. Ito1 and IK1, the transient outward and inward rectifier potassium currents, were present in cardiomyocytes only, whereas the rapid delayed rectifier potassium current (IKr), pacemaker current (If),

and L-type calcium current (ICa,L) could be recorded both in undifferentiated human embryonic stem cells and in cardiomyocytes. Most of the currents underwent developmental maturation in cardiomyocytes, as assessed by modifications in current density (Ito1, IK1, and ICa,L) and properties (If). Ion-channel mRNAs were always present when the current was recorded. Intracellular recordings in spontaneously beating clusters of cardiomyocytes revealed changes in action potential parameters and in response to pharmacological tools according to time of differentiation. In summary, human embryonic stem cell-derived cardiomyocytes mature over time during in vitro differentiation, approaching an adult phenotype. STEM CELLS 2007;25:1136 –1144

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION Human embryonic stem cells (hESCs) are pluripotent cells derived from in vitro fertilized-spared embryos at the blastocyst stage. Thomson et al. [1] derived the first hESC line in 1998, providing the research community with the first tool to study in vitro human differentiation processes. Several groups demonstrated the ability of hESC to differentiate into cells of the three germ layers, such as cardiac cells [2], neurons, hepatocytes, hematopoietic, and insulin-secreting cells (for review, see [3]). Cardiomyocytes (CMs) derived from hESCs are considered a promising source for cell-based therapies of the heart after damage resulting from myocardial infarction or cardiomyopathies. The appearance of spontaneously beating CMs upon differentiation of hESCs into EBs has been reported in several studies. In particular, detection of cardiac-specific molecular markers and analysis of cell structure confirmed the presence of differentiated CM within EBs [4, 5]. Initial functional studies showed that the cells displayed electrophysiological properties resembling those of CMs, such as propagation of electric signal [6, 7] and presence of nodal-, embryonic atrial-, and ventricularlike action potentials (APs), indicating the presence of CM

subtypes [6, 8]; hESC-derived CMs are also sensitive to chronotropic agents [5, 9] and display calcium transients [2, 8]. However, very little is known about the functional maturation of human cardiac cells over time. The shape of APs results from the ordered integration of several functional ionic currents. During cardiac development, expression and function of diverse relevant channel types occur over time. Indeed, studies on animal models show that channels undergo fetal and postnatal developmental changes, a complex process leading to the acquisition and maintenance of a mature cardiac electrophysiological phenotype [10 –12]. In hESC-derived CMs, sodium current (INa), which contributes to membrane excitability, was already present at early stages of differentiation (i.e., 20 –35 days), whereas the inward rectifier potassium current (IK1), which plays a role in setting the diastolic potential in mature CMs, was absent [13]. No information is available about stability and maturation of the electrophysiological cardiac phenotype for a long time. Thus, we thought it was important to assess whether hESCderived CMs undergo functional changes during in vitro development and how the electrophysiological phenotype is maintained over time. To this end, we applied electrophysiological and molecular techniques to characterize developmental

Correspondence: Marisa Jaconi, Ph.D., Department of Pathology and Immunology, Geneva Faculty of Medicine, Centre Me´dical Universitaire, 1 rue Michel-Servet, 1211 Geneva 4, Switzerland. Telephone: ⫹41-22-379-5257; Fax: ⫹41-22-379-5479; e-mail: marisa.jaconi@ medecine.unige.ch Received July 25, 2006; accepted for publication January 11, 2007; first published online in STEM CELLS EXPRESS January 25, 2007; available online without subscription through the open access option. ©AlphaMed Press 1066-5099/2007/$30.00/0 doi: 10.1634/stemcells.2006-0466

STEM CELLS 2007;25:1136 –1144 www.StemCells.com

Sartiani, Bettiol, Stillitano et al.

1137

changes of ion channels in spontaneously beating CMs for up to 3 months. Electrophysiological data, obtained with intracellular and single-cell patch-clamp recordings, and mRNA expression of ionic channels show that currents controlling AP duration and diastolic phase undergo developmentally regulated changes. The relative contribution of these channels to the generation of spontaneous AP during development was assessed with selective pharmacological tools.

MATERIALS

AND

METHODS

hESC Culture and Differentiation The hESC line H1 from WiCell Research Institute Inc. (Madison, WI, http://www.wicell.org) was cultivated following Wicell protocols. Briefly, colonies of hESC were passaged once a week on irradiated mouse embryonic fibroblasts using collagenase IV and mechanical dissociation. Propagation medium was composed of Dulbecco’s modified Eagle’s medium (DMEM)/F-12 ⫹ 20% serum replacement, 1% penicillin-streptomycin, 1% nonessential amino acid (NEAA), ␤-mercaptoethanol (Sigma-Aldrich GmbH, Buchs, Switzerland, http://www.sigmaaldrich.com), L-glutamine, and 4 ng/ml human basic fibroblast growth factor. All cell culture products were purchased from Gibco (Invitrogen AG, Basel, Switzerland, http://www.invitrogen.com), unless otherwise specified. To induce EB formation, colonies were incubated for 15 minutes with collagenase IV, rinsed with phosphate-buffered saline, and gently scraped in differentiation medium (KnockOut-DMEM ⫹20% Defined Fetal Bovine Serum [FBS; Hyclone Laboratories, Logan, UT, http://www.hyclone.com], 1% penicillin-streptomycin, 1% NEAA, L-glutamine, and ␤-mercaptoethanol), as reported elsewhere [5]. EBs were cultured for 4 days in Costar ultra-low attachment six-well plates (Corning, Schiphol-Rijk, The Netherlands, http://www.corning.com), with medium changed every 2 days. EBs were then plated on gelatin-coated 6-cm dishes, and medium was changed every 2–3 days.

Cell Isolation Undifferentiated single hESCs were isolated from confluent colonies by treatment with trypsin-EDTA for 3–5 minutes. Dissociated cells were collected, resuspended in normal Tyrode’s solution (see Solutions), and kept at room temperature. Beating clumps were dissected using a microscalpel and directly placed into the solution used to perform intracellular recordings. Alternatively, they were digested with collagenase (StemCell Technologies SARL, Grenoble, France, http://www.stemcell.com) for 20 –30 minutes at 37°C with pipetting every 5–10 minutes. Cells were then plated on gelatin- and fibronectin-coated 3.5-cm dishes in differentiation medium.

Immunofluorescence Studies Contracting areas were dissected and replated on coverslips coated with gelatin and laminin. After 3– 4 days, cells were fixed with 3% paraformaldehyde and permeabilized with 0.5% Triton X-100 (Sigma-Aldrich). Primary antibodies used in this study were mouse anti-myosin heavy chain (MHC; clone MF20, dilution 1:10; Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA, http://dshb.biology.uiowa.edu), mouse anti-␣-actinin (A7811, dilution 1:200; Sigma-Aldrich). Secondary antibodies were, respectively, goat anti-mouse-tetramethyl rhodamine isothiocyanate (Southern Biotechnology, Birmingham, AL, http://www. southernbiotech.com) or goat anti-mouse-rhodamine (Sigma-Aldrich). Nuclei were stained with TOTO-3 (Molecular Probes, Eugene, OR, http://probes.invitrogen.com). Phase contrast images and movies were taken using an inverted microscope (Eclipse TE300, Nikon, Zurich, Switzerland, http://www.europe-nikon.com) equipped with a camera (Coolpix 995; Nikon). Confocal pictures were taken with a laser scan microscope (LSM 510, Carl Zeiss, Oberkochen, Germany, http://www.zeiss.com). Stacks of images were reconstructed in three dimensions with Imaris software (Bitplane AG, Zurich, Switzerland, http://www.bitplane.com).

www.StemCells.com

RNA Isolation and Reverse TranscriptasePolymerase Chain Reaction Because spontaneously contracting CMs constitute only a small fraction of EBs, parts of EBs containing beating areas were dissected under a binocular microscope. Only dissected specimens were use to performed reverse transcriptase-polymerase chain reaction (RT-PCR; except for undifferentiated cells). Total RNAs from hESCs or from up to six dissected beating clusters were isolated using Trizol Reagent (Invitrogen). Reverse transcription was performed on 0.5–1 ␮g of total RNA with oligo-dT primers (Promega GmbH, Mannheim, Germany, http://www.promega.com), random hexamers (Promega), deoxynucleoside-5⬘-triphosphates (Promega), and Superscript II Reverse Transcriptase (Invitrogen). PCR was performed using Taq polymerase (Qiagen AG, Hombrechtikon, Switzerland, http://www1.quiagen.com). National Center for Biotechnology Information mRNA accession numbers, primer sequences, and PCR conditions are listed in supplemental online Table 1. Primers were designed on different exons, and the absence of cross-reactivity with mouse embryonic fibroblast cDNA was verified. After amplification, PCR fragments were resolved on 1.5% agarose gels containing ethidium bromide. PCRs were repeated on three independent experiments per time point, and pictures displayed (in Figs. 2–5) are representative.

Quantitative RT-PCR Expression levels of Kir2.1 and HCN1, -2, and -4 genes were further investigated using real-time quantitative RT-PCR and TaqMan (Applied Biosystems, Foster City, CA, http://www. appliedbiosystems.com) probe-based chemistry. Primer-Express Software (Applied Biosystems) was used to design primer and TaqMan-MGB probe sets for HCN genes. Primers and probes for Kir2.1 and the endogenous control genes GAPDH and Eukaryotic 18S rRNA were obtained from Applied Biosystems’ TaqMan Gene Expression Assay catalog (Kir2.1: Hs00265315_m1; GAPDH: Hs99999905_m1; 18S rRNA: 433,3760T). These assays come in a 20⫻ reaction mix, span an exon-exon junction, and are optimized to give approximately 100% efficiency. The real-time RT-PCR reactions were performed using TaqMan Universal PCR Master Mix (Applied Biosystems) in a 20-␮l reaction volume containing 20 ng of cDNA. All reactions were performed in triplicate and included a negative control. PCR reactions were carried out using an ABI Prism 7500 Sequence Detection System (Applied Biosystems). Cycling conditions were 2 minutes at 50°C, 10 minutes at 95°C, and 40 cycles of 15 seconds at 95°C and 1 minute at 60°C. Relative quantification of mRNA levels was determined by the 7500 system software, which uses only the comparative method (⌬⌬CT). To adequately assess the stability of GAPDH gene expression, we performed real-time PCR experiments in which GAPDH was considered the target gene and eukaryotic 18S ribosomal RNA the endogenous control. Eukaryotic 18S rRNA was chosen because its expression was stable over different samples (p ⫽ .41). Results show that GAPDH expression does not vary significantly during differentiation and among developmental stages (p ⫽ .39; supplemental online Fig. 1). Messenger RNA levels for ion channel genes could not be measured appropriately using 18S rRNA as endogenous control gene because, in our conditions, its amplification assay interfered with the detection of less abundant targets. This led to poor reproducibility and inaccurate quantification.

Patch-Clamp Recordings The experimental set-up for patch-clamp (whole-cell) recording and data acquisition was similar to that described previously [14, 15]. The patch-clamped cell was superfused by means of a temperaturecontrolled (37°C) micro-superfuser, allowing rapid changes of the solution bathing the cell. Patch-clamp pipettes, prepared from glass capillary tubes (Harvard Apparatus Ltd, Kent, U.K., http://www. harvardapparatus.com) by means of a two-stage horizontal puller (model P-87; Sutter Instrument, Novato, CA, http://www.sutter. com), had a resistance of 2–3 M⍀ when filled with the internal solution. Cell membrane capacitance (Cm) was measured by apply-

1138

Maturation of Human ESC-Derived Cardiomyocytes

ing a ⫾10 mV pulse starting from a holding potential of ⫺70 mV, as previously reported [14, 15]. Series resistance (Rs) and membrane capacitance were compensated to minimize the capacitive transient and routinely checked during the experiment. Only cells showing a stable Cm and Rs were included in the analysis. Properly modified Tyrode’s or pipette solutions were used to measure the different ionic currents (see Solutions). The rapid delayed rectifier outward current (IKr) was evoked by steps in the range of ⫺40 to ⫹50 mV (holding potential [HP] ⫺40 mV) to inactivate the sodium current (INa) and T-type calcium currents (ICaT). Transient outward potassium current (Ito1) was evoked by steps to ⫺40/⫹70 mV (HP ⫺70 mV); a pre-step to ⫺40 mV was used to inactivate INa current. Ito1 was measured as the difference between peak outward current at the beginning of the depolarizing step and the steady-state current at the end of the step, and normalized with respect to Cm. The funny current (If) was evoked by hyperpolarizing steps to ⫺50/⫺130 mV (HP ⫺40 mV). To evaluate steady-state values of If current, data were fitted to a monoexponential decay. The fitting allowed calculating the time constant of current activation [␶ (ms)], which was plotted as a function of the step potentials. If amplitudes were measured as the difference between the extrapolated value at the steady state and that at the beginning of the test pulse, and normalized with respect to Cm, as reported elsewhere [16]. IK1 was evoked by hyperpolarizing steps to ⫺120/25 mV (HP ⫺70 mV), measured at ⫺90 mV as barium-sensitive current, and normalized with respect to Cm. From a holding potential of ⫺50 mV, ICa,L was elicited by depolarizing steps to ⫺45/⫹55 mV. Peak ICa,L was measured as the difference between the peak inward current at the beginning of the depolarizing step and the steady-state current at the end of the step.

Intracellular Recordings Spontaneously beating EBs were placed on the Sylgard bottom (Sylgard 184 Silicone Elastomer Kit; Dow Corning, Midland, MI, http://www.dowcorning.com) of a perfusion chamber and fixed with metal pins. The chamber was thermostatically controlled at 33°C–35°C and superfusing solution (see Solutions) maintained at a constant-flow. Cellular electrical activity was recorded using standard electrophysiological techniques, as described in detail previously [17]. Briefly, the recording electrode consisted of a short Ag/AgCl pin that was partly inserted into a floating glass microelectrode containing 3 M KCl and connected to the headstage of the amplifier; an Ag/AgCl pellet served as reference electrode in the perfusion chamber. The tip resistance of the microelectrode ranged between 30 and 40 M⍀. The recording microelectrode and the reference electrode were connected through a high-input impedance amplifier (Biomedica Mangoni, Pisa, Italy, http://www. biomedicamangoni.it) interfaced with a computer. The microelectrode was slowly moved into the chamber under microscopic inspection with the use of a micromanipulator. The electrode potential was compensated to 0 in the bathing solution. Spontaneous APs were digitized by an A/D converter and analyzed off-line with Iox software (emka TECHNOLOGIES, Falls Church, VA, http://www. emkatech.com).

Solutions Normal Tyrode’s solution (in mM): NaCl, 140; KCl, 5.4; CaCl2, 1.8; MgCl2, 1.2; D-glucose, 10; and HEPES, 5 (pH 7.35 with NaOH). Modified Tyrode’s solution for If current (in mM): NaCl, 140; KCl, 25; CaCl2, 1.5; MgCl2, 1.2; BaCl2, 2; MnCl2, 2; 4-aminopyridine, 0.5; glucose, 10; and HEPES-NaOH, 5 (pH 7.35); this solution allowed the reduction of interference from other currents (i.e., ICa,L, ICa,T, IK1, and Ito1). Modified Tyrode’s solution for Ito1 current: normal Tyrode’s solution plus 0.5 mM CdCl2. Modified Tyrode’s solution for ICa,L current (in mM): TEA-Cl, 140; CsCl, 5.4; MgCl2, 1.2; CaCl2, 1.8; HEPES, 5; and glucose, 10 (adjusted to pH 7.30 with CsOH). Pipette solutions for APs, Ito1, If, and IKr (in mM): K-aspartate, 130; Na2-ATP, 5; MgCl2, 2; CaCl2, 5; EGTA, 11; and HEPESKOH, 10 (pH 7.2). Pipette solutions for ICa;L (in mM): Mg-ATP, 5;

Figure 1. Culture and differentiation of hESCs into CMs. (Aa): Phase contrast picture of hESCs growing in compact colonies on irradiated mouse feeder cells. (Ab): Cluster of spontaneously beating CMs dissected from an EB at day 30 [Ab], dotted area). (Ac, Ad): Dissected beating clusters contained MHC-positive cardiomyocytes ([Ac], day 15), as well as MHC-negative cells (D). (Ae): Higher magnification of sarcomeric organization in MHC-positive CMs (corresponding to the white square of panel [Ac]). (Af): Perinuclear sarcomeric organization of ␣-actinin within a single human CM (three-dimensional reconstruction). Scale bars ⫽ 100 (Aa–Ad) and 10 ␮m (Ae, Af). (B): Scheme of the experimental procedure for the analysis of molecular and functional properties of hESC-derived CMs. Abbreviations: CM, cardiomyocyte; hESC, human embryonic stem cell; MHC, myosin heavy chain.

EGTA, 15; TEA-Cl, 20; HEPES, 10; and CsCl 125 (pH 7.20 with CsOH). External solution for intracellular recordings (in mM): NaCl, 125; KCl, 4; NaHCO3, 25; NaH2PO4, 0.5; MgSO4, 1.2; CaCl2, 2.7; glucose, 1 (pH 7.2 when gassed with 5% CO2/95% O2).

Statistics Data are expressed as mean ⫾ SEM. Statistical analysis was performed using Student’s t test for grouped data (in case of two groups) or one-way analysis of variance (in case of multiple groups). A p value of less than .05 was considered significant.

RESULTS hESCs Differentiate into Spontaneously Beating Cardiomyocytes For propagation, hESCs grew as compact colonies on inactivated feeder cells (Fig. 1Aa). To allow differentiation, we placed colonies of hESCs in suspension in the presence of FBS. Upon aggregation and formation of EBs, spontaneously beating areas became visible from day 8 of differentiation and could maintain this automaticity for more than 3 months. Clusters of spontaneously beating CMs were dissected and replated, and


1139

Figure 2. Kv4.3, Kv1.4, and transient outward potassium current (Ito1); and HERG, HERG1b, and rapid delayed rectifier potassium current. (A): Kv4.3 mRNA was detected at all stages, whereas Kv1.4 could be detected only in CMs. (B): Ito1 measured in differentiated cardiomyocytes (CMs) at different stages; representative recordings at 12 (Ba) and 57 (Bb) days of differentiation. (C): HERG mRNA was detected both in hESCs and CMs, whereas HERG1b was specifically expressed in CMs. (D): Outward current measured in hESCs (Da) and in differentiated CMs ([Db], 80 days of differentiation). Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; hESC, human embryonic stem cell; pA, picoampere.

continued to beat (Fig. 1Ab and supplemental online movie 1). Dissected contracting clusters contained areas of MHC-positive cardiac cells (Fig. 1Ac), but also MHC-negative cells (Fig. 1Ad). At a higher magnification, organized MHC-positive sarcomeres were observed (Fig. 1Ae corresponding to the white square of Fig. 1Ac). After dissociation of beating areas, sarcomeres within single adherent CMs reorganized around the nucleus, as seen by three-dimensional distribution of ␣-actinin (Fig. 1Af). To compare the maturation state of contractile CMs over time, we decided to place CMs in two groups: early CMs, which were analyzed between days 15 and 40, and late CMs, which were analyzed between days 50 and 110 of differentiation, as represented in Figure 1B.

Repolarizing Potassium Currents Ito1 and IKr Appear at Different Stages Ito1, the calcium-independent transient outward potassium current, undergoes relevant developmental changes according to data from rat and dog models [18]. Several isoforms of ␣-subunits that assemble into the channel responsible for Ito1 have been described [19]. Therefore, we assessed by RT-PCR the expression over time of two of these isoforms, Kv1.4 and Kv4.3, which have two splice variants. Similarly to adult human heart, the shorter Kv4.3 splice variant was expressed only in late CMs (Fig. 2A), whereas the longer was homogeneously expressed in hESCs and CMs. The Kv1.4 subunit was expressed from day 25 and maintained thereafter. Kv1.4 expression was also detected in human adult heart RNA. Both early and late hESC-derived CMs displayed a transient outward current evoked by depolarizing pulses, which activated fast and exhibited a marked inactivation process (Fig. 2Ba, 2Bb). Current-voltage relation obtained by plotting peak currents revealed a voltage-dependence of activating current (data not shown). These properties and the fact that current was recorded in the presence of intracellular EGTA (which chelates calcium) suggest that this current is Ito1. Moreover, the current was completely blocked by millimolar concentrations of 4-aminopyridine, as expected for Ito1 (data not shown). The Ito1-like current present in differentiated CMs could be measured in early CMs from 12 days of differentiation and was maintained www.StemCells.com

throughout maturation. Its density, measured during a depolarizing step to 50 mV, increased during development from 4.2 ⫾ 1.2 (n ⫽ 9) to 7.7 ⫾ 2.6 picoampere/picofarad (pA/pF) (n ⫽ 9), although the difference was not statistically significant (p ⫽ .2). In contrast to CM and despite Kv4.3 mRNA expression, Ito1 was absent in undifferentiated hESC (data not shown), suggesting that the protein was either not expressed or was functionally inactive at this stage. On the other hand, a different potassium current was functionally measurable both in undifferentiated hESCs and in hESC-derived CMs. Upon depolarization (HP ⫺40 mV), hESCs and CMs displayed an outward current (Fig. 2Da, 2Db) that was sensitive to E4031 (data not shown), a selective blocker of IKr, the rapid delayed rectifier potassium current present in native CMs. Human ether-a-gogo related gene (HERG) and HERG1b, a shorter splice variant expressed in the heart and in tumor cells, constitute the channel responsible for IKr [20, 21]. RT-PCR analysis of these genes indicated the presence of HERG mRNA in all samples tested, whereas the shorter isoform HERG1b was selectively expressed in hESC-derived CMs and in adult human heart (Fig. 2C).

Temporal and Functional Changes in Diastolic Currents: Hyperpolarization-Activated and Inward Rectifier Currents If is a diastolic current present in pacemaker, atrial, and ventricular CMs [16, 22]. Four isoforms code for f-channels: hyperpolarization-activated cyclic nucleotide-gated potassium channel (HCN)-1, -2, -3, and -4 (for review, see [23]). The isoforms have different speed of activation and coassemble to form tetrameric HCN channel pores with distinct functional properties. For that reason, we investigated the expression of HCN1, -2, and -4 subunits over time. Levels of HCN1, usually expressed during embryogenesis and typical of mammal pacemaker cells (for review, see [24]), and HCN4 were roughly conserved from hESCs to early CMs, as observed in Figure 3A. Quantitatively, their expression was significantly lower in late CMs (Fig. 3B). Interestingly, HCN1 isoform is weakly expressed in adult heart (Fig. 3A) and significantly reduced in late

1140


Figure 3. HCN1, HCN2, HCN4, and fcurrent (If). (A): HCN1, HCN2, and HCN4 mRNAs were present at all time points of differentiation. (B): Quantitative mRNA levels for HCN isoforms. (C): If could be measured in undifferentiated hESCs (Ca) and in early- and late-stage CMs ([Cb]: 12 days of differentiation; [Cc]: 82 days of differentiation); the red dashed line corresponds to exponential fitting of current activation at ⫺140 mV. (D): Voltage dependence of time constant of activation at different maturation stages. Inset: Activation curves of If. In (B): ⴱ, p ⬍ .05 versus hESCs; ⴱⴱ, p ⬍ .01 l versus hESCs. In (D): ⴱ, p ⬍ .05 late CMs versus hESCs; ⴱⴱ, p ⬍ .01 late CMs versus hESCs; †, p ⬍ .05 late versus early CMs. Abbreviations: CM, cardiomyocyte; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; hESC, human embryonic stem cell; pA, picoampere; pF, picofarad.

EBs (Fig. 3B). HCN2 mRNA was rather constant in all samples tested (Fig. 3A, 3B). Figure 3C illustrates a typical family of currents evoked by hyperpolarizing steps from ⫺60 to ⫺130 mV (HP ⫺40 mV) in an undifferentiated hESC (Fig. 3Ca) and in hESC-derived CM at early (Fig. 3Cb) and late (Fig. 3Cc) stages of differentiation. Typically, the current amplitude increased and the activation rate accelerated as the potential became more negative. Currentvoltage relations (Fig. 3D, inset) indicated that the average potential of half-maximal activation was not modified by different maturation stages, being ⫺99.5 ⫾ 1.4, ⫺108.9 ⫾ 7.9, and ⫺96.7 ⫾ 6.9 mV in undifferentiated hESCs and early and late CMs, respectively. Interestingly, the activation rate of If slowed during maturation. Indeed, plotting time constant of activation versus step potential showed that the current rate of activation was faster in undifferentiated cells compared to hESC-derived CMs, and values differed significantly at ⫺130 and ⫺120 mV (Fig. 3D). IK1, the inward rectifier potassium current, is responsible for the late repolarization phase and the diastolic potential of the mature cardiac APs. Kir2.1 mRNA, which encodes for the IK1 pore subunits, was already present in undifferentiated hESCs (Fig. 4A) and was significantly upregulated in differentiated hESC-derived CMs (p ⬍ .01; Fig. 4B). However, IK1 was absent in undifferentiated cells and could be recorded only in hESCderived CMs. Representative recordings of IK1, measured at early and late stages of maturation and identified as current sensitive to barium blockade (0.5 mM BaCl2), are illustrated in Figure 4C. At ⫺90 mV, late CMs displayed a significantly higher IK1 density (early CMs, 0.6 ⫾ 0.3 pA/pF, n ⫽ 15; late CMs, 3.4 ⫾ 1.3 pA/pF, n ⫽ 12, p ⬍ .05).

Occurrence of Voltage-Dependent L-Type Calcium Current The gene CACNA1C encodes for the calcium channel ␣1C subunit, which mediates the voltage-dependent ICa,L in several tissues. Corresponding mRNA could be detected in undifferentiated hESCs (Fig. 5A) and in CMs. Consistent with these molecular findings, electrophysiological recordings showed the occurrence of ICa,L both in undifferentiated hESCs and in CMs. Current could be measured erratically in undifferentiated cells

Figure 4. Kir2.1 and inward rectifier potassium current. (A): Kir2.1 mRNA was present in undifferentiated hESCs and at all stages of differentiation. (B): Quantitative reverse transcriptase-polymerase chain reaction demonstrates upregulation of Kir2.1 expression during differentiation. (C): Background currents measured before (black line) and after (red line) superfusing cells with 0.5 mM Ba2⫹ in early- ([Ba], 24 days) and late-stage CM ([Bb], 108 days). ⴱⴱ, p ⬍ .01 versus hESCs. Abbreviations: CM, cardiomyocyte; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; hESC, human embryonic stem cell.

and consistently from small beating aggregates of CMs (Fig. 5Ba, 5Ca). Moreover, an increase of maximal current density was detected in CMs compared to hESCs; the current measured at 5 mV was ⫺0.9 ⫾ 0.13 pA/pF in hESCs (Fig. 5Bb; n ⫽ 3) and at ⫺5.8 ⫾ 1.9 mV in CMs (Fig. 5Cb; n ⫽ 4). The voltage of half maximal activation was ⫺16.7 ⫾ 1.7 and ⫺25.9 ⫾ 2.5 mV in undifferentiated and differentiated cells, respectively.

Intracellular Recordings from Beating Clusters To further elucidate the physiological role of ion channels in hESC-derived CMs, we performed intracellular recordings in spontaneously beating clusters dissected from EBs and challenged with different pharmacological tools. The average properties of APs, measured in intact spontaneously beating clusters


Figure 5. CACNA1C and L-type calcium current (ICaL). (A): CACNA1C mRNA was present in undifferentiated hESCs and at all stages of differentiation. ICaL could be measured both in hESC (Ba) and in cardiomyocytes (CMs) ([Ca], 57 days). Activation curve of ICaL density versus step potentials in hESCs (Bb) and CMs (Cb). Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; hESC, human embryonic stem cell; pA, picoampere; pF, picofarad.

at different stages, showed significant changes during differentiation. In particular, a significant increase in upstroke velocity (Vmax) and action potential duration (APD) was also observed, possibly relating to a stage-dependent increase in inward currents such as the ICa,L (Fig. 4). Vmax and APD measured at 70% repolarization were, respectively, 4.2 ⫾ 0.6 V/s and 199.7 ⫾ 11.3 ms in early CMs (n ⫽ 12) and 6.0 ⫾ 0.4 V/s and 341.4 ⫾ 47.9 ms in late CMs (n ⫽ 11; p ⬍ .05). In general, the standard deviation for APD values measured at any percentage of repolarization was significantly smaller (p ⬍ .01) in early than in late beating CMs, thus suggesting a higher variability in the latter group (supplemental online Table 2). Indeed, APD values in late beating CMs seem to cluster into two different groups with APD shorter (33%) and longer (66%) than 200 ms, possibly suggestive of atrial versus ventricular-like APs. On the other hand, the diastolic depolarization rate (DDR) was significantly flattened in late-stage clusters (9.7 ⫾ 1.0 mV/s) compared to early-stage ones (22.8 ⫾ 5.8mV/s; p ⬍ .05), and the frequency of spontaneous firing decreased, but not significantly (early CMs, 36 ⫾ 6 beats per minute; late CM, 26 ⫾ 3 beats per minute). Spontaneous rhythm and DDR showed a significant linear correlation (p ⬍ .0001; R2 ⫽ 0.57). Figure 6A shows AP recordings from spontaneously beating clusters at an early stage of maturation in control conditions (black lines) and after exposure to 10 ␮M E4031 for 30 and 60 seconds (gray and red lines, respectively). E4031 produced a striking prolongation of APD after 30 seconds, which was further enhanced at 60 seconds. Afterward, clusters of CM became unexcitable because spontaneous beating stopped and could rarely be reversed, even after prolonged wash-out (data not shown). At late stages of differentiation, E4031 depolarized membrane diastolic potential and increased the frequency of spontaneous AP (data not shown). On the whole, these data suggest a key role for IKr in the regulation of membrane potential and excitability threshold of hESC-derived CMs. The contribution of diastolic currents to DDR and spontaneous activity was evaluated by using specific blockers of IK1 www.StemCells.com

1141

Figure 6. Current sensitivity to pharmacological inhibitors. (A): Effect of 10 ␮M E4031 (rapid delayed rectifier potassium current blocker) on action potential recordings in spontaneously beating clusters at earlystage of differentiation (day 20). (B): Blockade of inward rectifier potassium current with 0.5 mM BaCl2 increased the slope of diastolic depolarization (26 days of differentiation). (C): Pacemaker current inhibition by Zat (10 ␮M) decreased the slope of diastolic depolarization and the spontaneous rate (56 days of differentiation). (D): Effect of Laci (10 ␮M) on action potential profile in an early cardiomyocyte cluster (40-day-old). Abbreviations: Ctr, control; Laci, lacidipine; Zat, zatebradine.

(0.5 mM BaCl2) and If (10 ␮M zatebradine). These compounds exerted opposite effects on DDR, in agreement with the counteracting roles of IK1 (a repolarizing current) and If (a depolarizing current). Indeed, superfusion with BaCl2 increased the slope of diastolic depolarization (Fig. 6B), whereas zatebradine decreased both DDR and the spontaneous rhythm of beating clusters (Fig. 6C). ␤-Adrenoceptor activation affects excitation-contraction coupling in the human heart [25]. In our experimental conditions, isoprenaline produced a positive chronotropic effect at 20 days of differentiation (supplemental online Fig. 2A), suggesting that the ␤-adrenergic system was already functionally mature in early CMs. A clear-cut acceleration of spontaneous discharge was also observed in CM clusters at a late stage of differentiation (supplemental online Fig. 2B). Finally, Figure 6D shows a typical effect of lacidipine (a calcium channel blocker of the dihydropyridine family) on the AP profile recorded from a 40-day-old beating cluster: ICa,L blockade by lacidipine caused a clear-cut reduction in the plateau duration and height, thus suggesting a major role for this current in controlling APD.

DISCUSSION The present study is the first to analyze a number of electrophysiological and molecular properties of CMs differentiated from hESCs over a long maturation period (i.e., 3 months). Using molecular analysis, and patch-clamp and intracellular recordings, we discovered several novel pieces of information indicating that developmentally controlled processes take place during in vitro differentiation of human CMs. After 3 months in culture, both the molecular and electrophysiological profiles of hESC-derived CMs moved toward an adult phenotype. As previously reported for mouse ESC-derived CMs [26], molecular/ functional expression of ion channels in hESC-derived CMs underwent developmental processes that were unsurprisingly

1142

Figure 7. Schematic representation of ionic currents measured in hESC and early or late CM. Developmental maturation is illustrated by the gray triangles, indicating an increase in current density, and by a dashed stripe, which denotes the slowing of activation rate for If. ⴱ, presence of INa has been described in [13]. Abbreviations: CM, cardiomyocyte; hESC, human embryonic stem cell; ICa,L, L-type calcium current; If, pacemaker current; IK1, inward rectifier potassium current; IKr, rapid delayed rectifier potassium current; INa, sodium current; Ito, transient outward potassium current.

long compared to those in mouse, which acquire a mature atrial or ventricular phenotype in approximately 3 weeks [26]. Figure 7 schematically summarizes ion currents recorded in hESCs, and early and late CMs, giving at the same time an idea of the qualitative and quantitative changes during maturation. In the sections that follow, we will discuss their relative contribution to the maturation-dependent changes of the AP profile.

Ion Currents Controlling AP Repolarization IKr was functionally present in hESC-derived CMs and clearly controlled cellular electrical properties, as extrapolated from intracellular recordings in beating clusters challenged with the selective IKr blocker E4031. Using intracellular recordings, we demonstrated that E4031, at a concentration able to completely block IKr, [27] caused a prolongation of APD much more pronounced than previously described [6], associated with depolarization of diastolic potential. This latter effect likely results from a low functional expression of IK1 and closely resembles that described for sino-atrial node cells, which also lack IK1 [28]. Overall, these results point to a prominent role of IKr in controlling the repolarization phase of hESC-derived CMs, as previously described [6]. Both of these actions— depolarization and APD prolongation—are consistent with the function of this channel in adult ventricular APs, in which it mainly contributes to the late phase of repolarization. Interestingly, at late stages of differentiation, the major effect of HERG blockade consisted of a marked depolarization, which further caused an acceleration of spontaneous rhythm, whereas APD remained unaffected. Similarly, blockade of HERG in neonatal dog ventricular tissue results in exaggerated AP prolongation compared with adult dog [11]. Thus, in immature/early hESC-derived CMs, IKr played an essential role in the regulation of APD and, therefore, in the repolarization phase. Cardiac differentiation of hESCs was associated with a selective expression of the cardiac specific isoform HERG1b, which is reported to coassemble with HERG1 to form heteromeric channels in human ventricle [21]. In this tissue, HERG channels mediate the rapid component of IKr that contributes to the repolarization phase of mammalian cardiac AP at the site of excitation-contraction coupling (i.e., the T-tubular structure) [21]. When expressed heterologously in Xenopus oocytes, mouse erg1b induces a fourfold faster deactivation rate than erg1, but both are equally blocked by E4031 [29]. Thus, we did

Maturation of Human ESC-Derived Cardiomyocytes not attempt a pharmacological dissection of the two isoforms in hESCs or CMs. Interestingly, undifferentiated cells expressed a number of mRNA transcripts for different channel proteins. However, only part of them appeared to be translated into functional proteins. In particular, an IKr-like current, sensitive to E4031 blockade, could be clearly detected in undifferentiated hESC, together with mRNA transcripts for HERG1. The functional role of IKr in these cells is unknown. Analogously to tumor cells [20], we can infer a cell-cycle-related functional expression. Other outward currents are likely to be involved in the control of APD during electrophysiological maturation. In particular, the density of Ito1, functionally expressed only in CMs, increased during maturation. This is not surprising, because data from animal models show that Ito1 channels undergo fetal and postnatal developmental changes, a process responsible for the acquisition and maintenance of a mature cardiac electrophysiological phenotype [10, 12, 30]. Similar results have been obtained previously in mouse ESC-derived CMs [31]. From a molecular point of view, human Ito1 channels were encoded mainly by Kv4.3, which was already present in undifferentiated hESCs. Kv1.4 mRNA, which is considered typical of endocardial ventricular cells [32], was expressed only in hESCderived CMs. These two isoforms were differently regulated. The simultaneous appearance of Kv1.4 transcript and of the functional current suggests that this isoform contributes to Ito1 measured in hESC-derived CMs. At present, however, we did not attempt to discriminate between the different channel isoforms. Another original finding of our work relates to the demonstration of calcium currents in hESCs and hESC-derived CMs. The functional presence of L-type calcium channels has been previously inferred in clusters of hESC-derived CMs from sensitivity to verapamil [8]. This current has been described previously in mouse ESC-derived CMs [26], but was not detected in undifferentiated hESCs [33]. Our data demonstrate that mRNA transcripts and ICaL-like currents occur in both hESC and CM; in the latter, ICaL seems to control the plateau phase, which is shortened by the selective L-type calcium antagonist lacidipine. Interestingly, molecular and functional expression of L-type calcium current seems to be enhanced upon cardiac differentiation; a larger contribution of ICaL during the plateau of action potential (characterized by high membrane resistance) may counterbalance the concomitant increase in outward potassium currents and favor the prolongation of action potential duration. However, a precise electrophysiological comparison was precluded by the low occurrence of measurable calcium currents.

Ion Currents Controlling Diastolic Potential Similarly to Ito1, IK1 seems to be a marker of cardiac differentiation and maturation, being expressed at a significantly higher density in late stage CM. These results can explain the lack of IK1 documented in other studies focusing on hESC-derived CM at early stages of differentiation [13]. Upon differentiation toward cardiac cells, we observed a clear-cut and quantitatively significant increase of Kir2.1 expression, which was in good correlation with functional expression of IK1 (supplemental online Fig. 4A). At least in canine and rabbit heart, Kir2.1 seems to be a specific ventricular isoform [34, 35], thus reinforcing the observation that differentiation toward ventricular-like myocytes occurs in our experimental conditions. The most striking consequence of IK1 blockade in beating clusters is an acceleration of the diastolic depolarization phase. This is not surprising, because in the human heart, IK1 controls the late phase of cardiac repolarization and stabilizes the diastolic potential [36].

Sartiani, Bettiol, Stillitano et al. The inward flow of sodium ions through f-channels plays a crucial physiological role in setting spontaneous rhythm of pacemaker cells and helping them to sense their autonomic control [24]. In our study, If was clearly measured in undifferentiated hESCs and CMs. Previous studies showed similar results in mouse ESCs, but failed to demonstrate If occurrence in hESCs [33]. Moreover, our results demonstrate, for the first time, that this current undergoes developmental changes during in vitro maturation of hESC-derived CMs. In fact, the kinetics of If activation were markedly slowed in late-stage CMs compared to in early ones. Accordingly, molecular data showed that the HCN1 isoform, which is reported to have the fastest activation kinetics [37], was largely expressed in hESCs and significantly reduced during cardiac differentiation and maturation. In contrast, the slower kinetic isoform HCN2 was expressed to a similar extent throughout the differentiation process, thus increasing its contribution to current properties in late CMs. A similar pattern of expression of HCN mRNA characterizes human adult heart tissue. Thus, it is conceivable that electrophysiological properties of If reflect changes in the molecular composition of f-channels, consequent to the relative contribution of HCN isoforms; indeed, a clear-cut correlation between activation kinetics and the ratio between HCN2 versus HCN1 and HCN4 was observed (supplemental online Fig. 4B). Of note, the relative distribution of HCN isoforms changed, as expected for cardiomyocytes acquiring a ventricular phenotype, because the quantitative amount of HCN1 and HCN4 (i.e., the sino-atrial node isoforms in the adult heart) decreased significantly. Interestingly enough, the rate of diastolic depolarization, which is controlled by If activation kinetics, is also significantly slowed in late-stage CMs, and the spontaneous rhythm is correspondingly decreased. The functional role of If is further supported by two observations obtained in beating CM clusters. Zatebradine, the prototype of f-channel blockers, decreased both the DDR and the spontaneous rhythm, whereas isoprenaline, which positively modulates f-channels via intracellular cAMP [38], increased both of them. Consistently with this result, preliminary observations show that If is positively modulated by isoprenaline in late CMs (supplemental online Fig. 3Aa, 3Ab), but not in early CMs (supplemental online Fig. 3Ba, Bb), in agreement with molecular data, suggesting a maturation-related dominance of HCN2. This isoform is indeed reported to be more sensitive to intracellular cAMP compared to HCN1. Our results, however, do not exclude the possibility that other mechanisms controlling pacemaker activity, such as inositol triphosphate-dependent calcium signaling [39], may contribute to isoprenaline-mediated effect in clusters of beating CMs.

Relevance and Limitations Changes occurring in hESC-derived CMs during differentiation in culture may help to assess the potential of in vitro-generated CMs for heart repair. We have identified Ito1 and IK1 (absent in hESCs) as markers of phenotypical cardiac differentiation. Early CMs have quite homogeneous APs and present a low density of Ito1 and IK1, a situation compatible with an atrial or a pacemaker phenotype. Indeed, pacemaker cells are known to have low or

REFERENCES

1143

no Ik1 and to express HCN1 [40], at variance with atrial or ventricular human cells [41]. This is confirmed by our molecular and electrophysiological data. In contrast, late-stage CMs express Ito1 and IK1 at higher densities and present a larger range of AP duration. This rather reflects differentiation into heterogeneous cardiac phenotypes, namely atrial or ventricular type. IK1 is essential for a stable and polarized diastolic membrane potential, and its absence may be arrhythmogenic in vivo [42]. Indeed, the acquisition of a suitable pattern of ion currents will be necessary for proper CM function, especially after in vivo engraftment. Under our experimental conditions, hESC-derived cardiomyocytes reach more mature phenotypes over a period of 3 months of in vitro culturing; however, they did not reach the phenotype typical of adult ventricular cardiomyocytes. Whether this is the consequence of limitations, intrinsic to cell culture conditions or to the insufficient time of observation, will require further investigations. To our knowledge, however, the functional properties of human fetal cardiomyocytes undergoing in situ development for 2–3 months are still unknown. Which CM population (i.e., early or late CMs) would be more suitable for heart repair remains an open question. This results from the lack of information concerning the possibility that hESC-derived CMs (or even earlier cardiac progenitors) (a) will maintain the phenotype acquired in vitro, or may undergo senescence processes as demonstrated for fetal cardiomyocytes in long-term cultures [43]; (b) will intrinsically pursue their maturation in vivo; or (c) will eventually modify their characteristics according to the surrounding tissue (via paracrine or electrical influence). Nevertheless, a precise characterization of the in vitro maturation of hESCs is a mandatory starting point for the understanding of their in vivo maturation and for their possible use in regenerative medicine.

ACKNOWLEDGMENTS We thank M. Stouffs and Pr. K.-H. Krause for help and fruitful discussions; D. Tirefort for technical assistance; Ph. Nouet and M. Rouzaud for irradiation of feeder cells; and Pr. A. Kleber and Dr. S. Theander for critically reading the manuscript. This work was supported by the Swiss Academy of Medical Sciences (Grant 14/03, salary to E. Bettiol), the Swiss National Science Foundation (Grant 3100-62010), and by the Italian Ministry of Education, University and Research (FIRB Grant RBNE01HLAK-009, 2001, salary to L. Sartiani; PRIN 2005062944_002). L.S. and E.B. contributed equally to this work. M.E.J. is currently affiliated with the Department of Pathology and Immunology, Faculty of Medicine, University of Geneva, Geneva.

DISCLOSURE

2

Thomson JA, Itskovitz-Eldor J, Shapiro SS et al. Embryonic stem cell lines derived from human blastocysts. Science 1998;282:1145–1147. Kehat I, Kenyagin-Karsenti D, Snir M et al. Human embryonic stem cells can differentiate into myocytes with structural and functional properties of cardiomyocytes. J Clin Invest 2001;108:407– 414.

www.StemCells.com

CONFLICTS

The authors indicate no potential conflicts of interest.

3 4

1

OF POTENTIAL OF INTEREST

5

Conley BJ, Young JC, Trounson AO et al. Derivation, propagation and differentiation of human embryonic stem cells. Int J Biochem Cell Biol 2004;36:555–567. Snir M, Kehat I, Gepstein A et al. Assessment of the ultrastructural and proliferative properties of human embryonic stem cell-derived cardiomyocytes. Am J Physiol Heart Circ Physiol 2003;285:H2355–H2363. Xu C, Police S, Rao N et al. Characterization and enrichment of cardiomyocytes derived from human embryonic stem cells. Circ Res 2002; 91:501–508.


1144

6 7 8 9 10 11

12 13 14 15

16 17 18 19 20 21 22 23 24 25

He JQ, Ma Y, Lee Y et al. Human embryonic stem cells develop into multiple types of cardiac myocytes: Action potential characterization. Circ Res 2003;93:32–39. Kehat I, Gepstein A, Spira A et al. High-resolution electrophysiological assessment of human embryonic stem cell-derived cardiomyocytes: A novel in vitro model for the study of conduction. Circ Res 2002;91:659–661. Mummery C, Ward-van Oostwaard D, Doevendans P et al. Differentiation of human embryonic stem cells to cardiomyocytes: Role of coculture with visceral endoderm-like cells. Circulation 2003;107:2733–2740. Reppel M, Boettinger C, Hescheler J. Beta-adrenergic and muscarinic modulation of human embryonic stem cell-derived cardiomyocytes. Cell Physiol Biochem 2004;14:187–196. Guo W, Kamiya K, Toyama J. Modulated expression of transient outward current in cultured neonatal rat ventricular myocytes: comparison with development in situ. Cardiovasc Res 1996;32:524 –533. Obreztchikova MN, Sosunov EA, Plotnikov A et al. Developmental changes in IKr and IKs contribute to age-related expression of dofetilide effects on repolarization and proarrhythmia. Cardiovasc Res 2003;59: 339 –350. Shimoni Y, Fiset C, Clark RB et al. Thyroid hormone regulates postnatal expression of transient K⫹ channel isoforms in rat ventricle. J Physiol 1997;500:65–73. Satin J, Kehat I, Caspi O et al. Mechanism of spontaneous excitability in human embryonic stem cell derived cardiomyocytes. J Physiol 2004;559: 479 – 496. Lonardo G, Cerbai E, Casini S et al. Atrial natriuretic peptide modulates the hyperpolarization-activated current (If) in human atrial myocytes. Cardiovasc Res 2004;63:528 –536. Sartiani L, Cerbai E, Lonardo G et al. Prenatal exposure to carbon monoxide affects postnatal cellular electrophysiological maturation of the rat heart: A potential substrate for arrhythmogenesis in infancy. Circulation 2004;109:419 – 423. Cerbai E, Pino R, Porciatti F et al. Characterization of the hyperpolarization-activated current, I(f), in ventricular myocytes from human failing heart. Circulation 1997;95:568 –571. Barbieri M, Varani K, Cerbai E et al. Electrophysiological basis for the enhanced cardiac arrhythmogenic effect of isoprenaline in aged spontaneously hypertensive rats. J Mol Cell Cardiol 1994;26:849 – 860. Jeck CD, Boyden PA. Age-related appearance of outward currents may contribute to developmental differences in ventricular repolarization. Circ Res 1992;71:1390 –1403. Nerbonne JM. Regulation of voltage-gated K⫹ channel expression in the developing mammalian myocardium. J Neurobiol 1998;37:37–59. Crociani O, Guasti L, Balzi M et al. Cell cycle-dependent expression of HERG1 and HERG1B isoforms in tumor cells. J Biol Chem 2003;278: 2947–2955. Jones EM, Roti Roti EC, Wang J et al. Cardiac IKr channels minimally comprise hERG 1a and 1b subunits. J Biol Chem 2004;279:44690–44694. Pino R, Cerbai E, Calamai G et al. Effect of 5-HT4 receptor stimulation on the pacemaker current I(f) in human isolated atrial myocytes. Cardiovasc Res 1998;40:516 –522. Biel M, Schneider A, Wahl C. Cardiac HCN channels: Structure, function, and modulation. Trends Cardiovasc Med 2002;12:206 –212. Accili EA, Proenza C, Baruscotti M et al. From funny current to HCN channels: 20 years of excitation. News Physiol Sci 2002;17:32–37. Brodde OE, Michel MC. Adrenergic and muscarinic receptors in the human heart. Pharmacol Rev 1999;51:651– 690.

26 Sachinidis A, Fleischmann BK, Kolossov E et al. Cardiac specific differentiation of mouse embryonic stem cells. Cardiovasc Res 2003;58: 278 –291. 27 Anderson CL, Delisle BP, Anson BD et al. Most LQT2 mutations reduce Kv11.1 (hERG) current by a class 2 (trafficking-deficient) mechanism. Circulation 2006;113:365–373. 28 Matsuura H, Ehara T, Ding WG. Rapidly and slowly activating components of delayed rectifier K(⫹) current in guinea-pig sino-atrial node pacemaker cells. J Physiol 2002;540:815– 830. 29 London B, Trudeau MC, Newton KP et al. Two isoforms of the mouse ether-a-go-go-related gene coassemble to form channels with properties similar to the rapidly activating component of the cardiac delayed rectifier K⫹ current. Circ Res 1997;81:870 – 878. 30 Guo W, Kamiya K, Cheng J et al. Changes in action potentials and ion currents in long-term cultured neonatal rat ventricular cells. Am J Physiol 1996;271:C93–C102. 31 Gryschenko O, Lu ZJ, Fleischmann BK et al. Outwards currents in embryonic stem cell-derived cardiomyocytes. Pflugers Arch 2000;439: 798 – 807. 32 Na¨bauer M, Beuckelmann DJ, Uberfuhr P et al. Regional differences in current density and rate-dependent properties of the transient outward current in subepicardial and subendocardial myocytes of human left ventricle. Circulation 1996;93:168 –177. 33 Wang K, Xue T, Tsang SY et al. Electrophysiological properties of pluripotent human and mouse embryonic stem cells. STEM CELLS 2005;23:1526 –1534. 34 Melnyk P, Zhang L, Shrier A et al. Differential distribution of Kir2.1 and Kir2.3 subunits in canine atrium and ventricle. Am J Physiol Heart Circ Physiol 2002;283:H1123–H1133. 35 Zobel C, Cho HC, Nguyen TT et al. Molecular dissection of the inward rectifier potassium current (IK1) in rabbit cardiomyocytes: Evidence for heteromeric co-assembly of Kir2.1 and Kir2.2. J Physiol 2003;550:365–372. 36 Silva J, Rudy Y. Mechanism of pacemaking in I(K1)-downregulated myocytes. Circ Res 2003;92:261–263. 37 Moosmang S, Stieber J, Zong X et al. Cellular expression and functional characterization of four hyperpolarization-activated pacemaker channels in cardiac and neuronal tissues. Eur J Biochem 2001;268:1646 –1652. 38 Cerbai E, Pino R, Rodriguez ML et al. Modulation of the pacemaker current If by beta-adrenoceptor subtypes in ventricular myocytes isolated from hypertensive and normotensive rats. Cardiovasc Res 1999;42:121–129. 39 Me´ry A, Aimond F, Menard C et al. Initiation of embryonic cardiac pacemaker activity by inositol 1,4,5-trisphosphate-dependent calcium signaling. Mol Biol Cell 2005;16:2414 –2423. 40 Altomare C, Terragni B, Brioschi C et al. Heteromeric HCN1-HCN4 channels: A comparison with native pacemaker channels from the rabbit sinoatrial node. J Physiol 2003;549:347–359. 41 Lonardo G, Stillitano F, Zicha S et al. Molecular basis of funny current (If) in normal and failing human heart. Circulation 2004;110(suppl): 129a–130a. 42 Miake J, Marban E, Nuss HB. Functional role of inward rectifier current in heart probed by Kir2.1 overexpression and dominant-negative suppression. J Clin Invest 2003;111:1529 –1536. 43 Ball AJ, Levine F. Telomere-independent cellular senescence in human fetal cardiomyocytes. Aging Cell 2005;4:21–30.

See www.StemCells.com for supplemental material available online.