Progressive maturation in contracting cardiomyocytes derived from human embryonic stem cells: Qualitative effects on electrophysiological responses to drugs

Stem Cell Research (2010) 4, 201–213

available at www.sciencedirect.com

www.elsevier.com/locate/scr

REGULAR ARTICLE

Progressive maturation in contracting cardiomyocytes derived from human embryonic stem cells: Qualitative effects on electrophysiological responses to drugs Tomomi G. Otsuji a , Itsunari Minami a , Yuko Kurose a , Kaori Yamauchi b,c , Masako Tada a,b,⁎,1 , Norio Nakatsuji b,c a

Stem Cell and Drug Discovery Institute, Kyoto, Japan Institute for Frontier Medical Sciences, Kyoto University:53, Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan c Institute for Integrated Cell-Material Sciences, Kyoto University, Kyoto, Japan b

Received 9 November 2009; received in revised form 27 January 2010; accepted 28 January 2010 Open access under CC BY-NC-ND license.

Abstract The field of drug testing currently needs a new integrated assay system, as accurate as systems using native tissues, that will allow us to predict arrhythmia risks of candidate drugs and the relationship between genetic mutations and acquired electrophysiological phenotypes. This could be accomplished by combining the microelectrode array (MEA) system with cardiomyocytes (CMs) derived from human embryonic stem cells (hESC) and induced pluripotential stem cells. CMs have been successfully induced from both types, but their maturation process is not systematically controlled; this results in loss of beating potency and insufficient ion channel function. We generated a transgenic hESC line that facilitates maintenance of hESC-CM clusters every 2 weeks by expressing GFP driven by a cardiac-specific αMHC promoter, thereby producing a compact pacemaker lineage within a ventricular population over a year. Further analyses, including quantitative RT-PCR, patch-clamp, and MEA-mediated QT tests, demonstrated that replating culturing continuously enhanced gene expression, ionic current amplitudes, and resistance to K+ channel blockades in hESC-CMs. Moreover, temporal three-dimensional (3D) culturing accelerated maturation by restoring the global gene repressive status established in the adhesive status. Replating/3D culturing thus produces hESC-CMs that act as functional syncytia suitable for use in regenerative medicine and accurate drug tests. © 2010 Elsevier B.V. All rights reserved.

Introduction Embryonic stem cells (ESC) possess not only self-renewing potency and pluripotency but also epigenetic reprogramming ⁎ Corresponding author. Fax: +81 859 38 6210. E-mail address: [email protected] (M. Tada). 1 Present address: Chromosome Engineering Research Center, Tottori University, Nishi-cho 86, Yonago, Tottori 683-8503, Japan. 1873-5061 © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.scr.2010.01.002

activities that bestow pluripotency on somatic nuclei (Tada et al., 2001). Recently, the reprogramming factors, through which pluripotential cells have been created from somatic cells as induced pluripotential stem cells (iPSC), have been identified (Takahashi and Yamanaka, 2006; Takahashi et al., 2007; Nakagawa et al., 2008; Kaji et al., 2009; Woltjen et al., 2009). Human ESC and iPSC (hESC/hiPSC) are currently viewed as promising tools in regenerative medicine and pharmacological development, satisfying the need for both

202 scalability and physiological relevance. Significant progress toward this goal has been made in the development of a method for producing cardiomyocytes (CMs) from hESC by reassembling the extrinsic factors involved in cardiac development in a stepwise manner (Jang et al., 2008). hiPSC-derived CMs have also been successively induced (Zhang et al., 2009). It is surprising, however, that the cell's functionality is not considered a serious matter and has not yet been systematically controlled. In hearts, ion channel function develops progressively from the fetal to the adult stage (Jeck and Boyden, 1992; Obreztchikova et al., 2003). Some major mature ventricular characteristics have been demonstrated in hESC-CMs after relatively prolonged periods of culturing (Sartiani et al., 2007), but IK1 and HERG channel functions were almost absent from these cells at early stages (Sartiani et al., 2007; Satin et al., 2004). There is, in fact, no direct evidence demonstrating that hESC-CMs accurately reflect the effects of the complex interactions of molecules in native tissue cells on the cells' electrophysiological response to a given stimulus. Another major problem is that the automaticity of hESCCMs declines during long periods of culturing, resulting in the early arrest of functional maturation of the pacemaker potencies (Sartiani et al., 2007; Yanagi et al., 2007). Mature ventricular CMs in hearts lose automaticity through developmental regulation, but they are stimulated by ordered electric impulses generated by pacemakers in the sinoatrial node (SAN). A normal heartbeat appears on an electrocardiogram (ECG) as a P wave, a QRS complex, and a T wave (Ashly and Niebauer, 2004). The P wave corresponds to SANinitiated depolarization, while the QRS complex and T wave represent the depolarization and repolarization, respectively, of the ventricles. This shows that pacemaker lineages are required for mature CM clusters to generate regular impulses and contractility. Therefore, we attempted to obtain functional syncytia containing both mature pacemaker CMs and synchronously contracting ventricular CMs as a model for adult hearts. The appearance of QT interval prolongation (LQT) on ECG has been used as a marker indicating that a drug could cause arrhythmia and sudden death (Keating and Sanguinetti, 2001). To evaluate the drug-induced LQT risks of candidate compounds in vitro, a microelectrode array (MEA) system has recently been used to trace the extracellular electric activity of native heart tissue and that of hESC-CMs. The MEA recordings tracing the electric activity of beating clusters produced data resembling the familiar data output of the ECG. Prolongation of the cardiac action potential duration (APD), which can be observed in patch-clamp recordings of single cardiac cells, is reflected in the phenomenon of prolonged APD90 times that were noted between the initiation of a depolarization wave and the end of its repolarization peak, which can also be detected in MEA (Stett et al., 2003). Here, for the sake of simplicity, they have been described as relative Na+–K+ intervals. Mutation-mediated loss of function causing LQT syndrome has already been reported in seven ion channel genes, mostly in HERG channels. The HERG channels are the targets of various medications, many of which can, for this reason, lead to drug-induced LQT (Roden, 1998). This evidence shows that functional increases and decreases in HERG function directly

T.G. Otsuji et al. modulate LQT and arrhythmia risks and significantly influence the accuracy of electrophysiological measurements. Loss of function in sodium channels and in calcium channels also causes electrophysiological abnormalities on ECG (Keating and Sanguinetti, 2001). Thus, the presence of premature cardiomyocytes in the testing pool may lead to false positive or false negative LQT detections that do not directly correlate with the degree of cardiac toxicity or the cardiac dysfunction that are caused by the genetic or pathological properties of the original cells. Therefore, functional cardiomyocytes at a developmental stage comparable to that of human adult ventricular cells will be required. In a significant step toward this goal, we have defined a simple but efficient procedure that accelerates the maturation process of hESC-CMs, especially those of the pacemaker lineage. First, we generated a human transgenic hESC line expressing GFP driven by the cardiac αMHC promoter. This marker has been used as a powerful indicator of cardiomyocytes in mice (Gulick et al., 1991; Kolossov et al., 1998, 2006). In human cells, the 913-bp 5′-flanking region and the 115-bp 3′ region at the transcriptional starting point determine the constitutive expression of human αMHC in the atrium and ventricular myocardium (Tsika et al., 1990). Next, we cocultured these hESCs with visceral endoderm-like END-2 cells, which are known to efficiently induce hESCs into various sorts of CMs, including fetal ventricle-like, atrial, and pacemaker cells (Mummery et al., 2003; Passier et al., 2005). The beating potency of GFP-expressing CM colonies was successfully maintained by repeated replating culturing every 2 weeks. This extended the lives of the hESC-CMs to over one year. These long-term cultured hESC-CMs exhibited remarkably enhanced Na+, K+, and pacemaker (known as If) ion channel function, as detected by means of patch-clamp recordings and gene expression analyses. Additional threedimensional (3D) culturing of beating hESC-CM clusters for a few weeks in suspension accelerated further cell-autonomous maturation of ion channels through interminable free contraction. The functional maturation that occurred through replating culturing followed by short-term 3D culture (replating/3D) resulted in cells that responded appropriately to ion channel blockers, allowing qualified MEA-mediated detection of drug-induced Na+–K+ interval prolongation in a dose-dependent manner. Here, we demonstrate the important discovery that gene expression levels influenced by a cell's culturing conditions directly affect that cell's sensitivity to ion channel inhibitors as detected through the MEA-mediated QT test. The replating/3D culture technique presented here is a convenient way to enhance expression levels of various ion channel genes in hESC–CMs that will provide more accurate clinical responses related to cardiac action potential regulation in native tissues.

Results Cardiac differentiation of transgenic hESCs The human αMHC promoter region was isolated by means of genomic PCR, and the αMHC-EGFP expression vector was transfected into the KhES-1 hESC line (Fig. 1A). The isolated transgenic hESC clones expressed representative pluripotent-

Maturation in Beating hESC-derived Cardiomyocytes associating genes (Fig. 1B, top). KhES-1 cells rarely formed beating colonies, regardless of whether embryoid bodymediated methods (Jang et al., 2008; Yang et al., 2008) or the END-2 coculture system (Passier et al., 2005) was used. Eventually, however, we reproducibly obtained beating hESC-CM colonies using the END-2 coculture method by adding 1% knockout serum replacement (KSR) for primary 2-day cultures to a basic differentiation medium containing 20% FCS (Fig. 1A and Supplemental Fig. 1). Beating areas became visible on Day 10 (D10), one day after the first GFP+ colonies were detected (Fig. 1C). The number of GFP+ beating areas increased until D14, but their beating potency gradually decreased until it mostly vanished on D21 (Fig. 1C). To prevent immobile adhesion of hESC-CMs onto the culture dish and to stem the expansion of noncardiac cells, we mechanically transferred hESC-CM colonies onto a new culture dish; this resulted in a gradual recovery of their

203 beating potencies (Fig. 1D). As summarized in Fig. 1A, we continued transferring beating hESC-CM colonies every 2 weeks, thus achieving long-term maintenance of almost 100% automaticity in hESC-CMs for over a year. RT-PCR analysis revealed various cardiac markers already present in 1-month-old (1M) hESC-CMs, as described previously (Yang et al., 2008), including TnT and αMHC, two molecules responsible for cardiac contraction; ANP and MLC2a, atrial markers; TBX5 and TBX20, ventricular markers; NKX2.5, a downstream gene of FGF and BMP signaling; and ISL1, an early cardiac marker (Fig. 1B). Most 3M GFP+ hESC-CMs were specifically positive for antibody staining of αActinin and αMHC, two act myosin components; TnT, one of three troponin subunits; and ANP, an atrial marker (Fig. 1E). TnT in particular is used as a representative marker of cardiac cells, as it directly regulates contraction through interaction with Ca2+. Out of 148 TnT-positive cells, 59% were GFP-positive. Thus, this αMHC promoter-based hESC-CM tracing system is likely to be

Figure 1 Maturation culture of contractile GFP+ hESC-derived cardiomyocytes (hESC-CMs). (A) Establishment of the transgenic hESC line showing cardiomyocyte-specific GFP expression, induction of hESC-CMs, replating culturing under adhesive conditions (Ad), and three-dimensional culturing in suspension (Sus). P, passage number after primary 21-day differentiation. (B) RT-PCR analyses of transgenic hESCs (D0) and D37 hESC-CMs. AH, human adult hearts; FH, human fetal hearts. (C) Changes in No. of GFP+ colonies and beating colonies (n = 6). (D) Restoration of automaticity in hESC-CMs through replating (n = 3). (E) Immunostaining for αMHC, Actinin, TnT, or ANP at 3 months. Scale bars, 10 μm. (F) GFP+ colony at Day 14 and GFP+ spheroids cultured under adhesive conditions (Ad) or in suspension (Sus). Scale bars, 500 μm.

204 useful as an indicator of CM-containing cell clusters. The GFP+ hESC-CM clusters were also maintained in a culture dish with a surface resistant to cell adhesion, which enabled their maintenance as spheroids with beating potency for at least 3 months through 3D culturing (Fig. 1F). The tight cell adhesion of CMs helped maintain the cell clusters containing GFP+ CMs even after numerous replating culturing procedures.

3D culturing led to enhanced ion channel function in hESC-CM clusters In in vitro QT tests on contracting GFP+ hESC-CM clusters, our MEA system clearly traced a set of simple waves consisting of an inward Na+ and/or Ca2+ spike and a repolarization wave mostly produced by outward K+ currents (Fig. 2B, top). Fiftytwo-day-old (D52) hESC-CMs cultured in suspension (Sus) after the primary 15-day differentiation culturing were treated with the reference compound E4031, which is known to block the HERG channels involved in the rapidly activating, delayed rectifier potassium current IKr (Spector et al., 1996) (Figs. 2A and B, Sus(D52)). A 100 nM E4031 shifted the potassium component to the right within the whole waveform by inhibiting IKr channels. QT interval was shown as the Na+–K+ interval, relative to those recorded before drug treatment in each set of experiments. As seen in Fig. 2C, dose-dependent Na+–K+ interval prolongation was

T.G. Otsuji et al. recorded in the Sus(D52) hESC-CMs, while shortened Na+–K+ intervals were seen in the control Ad(D52) cells that were cultured under adhesive conditions (Ad) even at a low concentration of E4031. Thus, these two samples exhibited different electrophysiological properties, though their primary differentiation conditions and total numbers of days in culture were comparable (Fig. 2A, top). We compared gene expression levels by means of quantitative RT-PCR (qRT-PCR) at the mid-course phase of culturing. HERG1b is a shorter splicing variant of HERG known for its selective expression in CMs (Jones et al., 2004). Remarkably, 3D culturing for 14 days reproducibly enhanced the expression levels of αMHC, HERG1b, and HCN1 in Sus(D29) hESC-CMs compared to their expression levels in Ad(D29) hESC-CMs subjected to replating culturing (Fig. 2A). Similar positive effects of 3D culturing on the expression levels of HERG1b, HCN4, HCN1, and Cav3.2 in Sus(D43) cells were repressed in SusAd(D43) cells by the addition of 14 days of Ad culturing (Fig. 2D). These results show that adhesive culturing somewhat inhibits the cardiac gene regulatory network and arrests hESC-CM maturation, leading to insufficient function of HERG molecules. Thus, free contractility may immediately evoke temporally repressed transcription in hESC-CMs, accelerating the cell's functional maturation beyond developmental regulation. Such temporal changes in a cell's function can directly affect the results obtained through

Figure 2 Enhanced function in hESC-CMs as a result of three-dimensional (3D) culturing. (A) Experimental scheme for A–C. Positive effects of 14-day suspension culturing on cardiac gene expression seen in Sus(D29) hESC-CMs compared to control Ad(D29) hESC-CMs cultured in adhesion, evaluated by quantitative RT-PCR (qRT-PCR). Total culture days for each group are shown in parentheses. All values were normalized with β-actin expression and shown relative to the corresponding values in human adult hearts (AH = 100). Each graph displays the mean and standard deviation of three independent experiments. (B) E4031-induced Na+-K+ interval prolongation detected by MEA recordings of Sus(D52) hESC-CMs maintained by replating/3D-culturing. Na+–K+ intervals, corresponding to the time from the initiation of a depolarization wave to the end of its following repolarization wave (arrows). (C) Qualitative differences detected in MEA recordings. In each sample, the average times of Na+–K+ intervals are demonstrated as a relative value to the controls nontreated with E4031 (n = 6). Dose-dependent changes in relative Na+–K+ intervals of extracellular field action potential due to E4031 in Sus(D52) and control Ad(D52) hESC-CMs. (D) Re-repression through additional 14-day adhesive culturing after 3D culturing as seen in SusAd(D43) hESC-CMs through qRT-PCR.

Maturation in Beating hESC-derived Cardiomyocytes MEA recordings. No clear explanation for the cause of APD shortening induced by an IKr inhibitor E4031 has been reported. However, the fact that the mutations in the potassium ion channel genes KVLQT1, HERG, and KCNJ2 are linked not only with the long QT syndrome (corresponding to the LQT1, LQT2, and LQT7, respectively) but also with the short QT (SQT) syndrome (Schimpf et al., 2005) suggests that insufficient function in potassium ion channels may cause bidirectional phenomena depending on the degree of affected function. The fact that short APD was detected in the E4031-treated Ad(D52) that showed insufficient expression in several ion channel genes may lend credence this speculation.

Repeated replating culturing led to enhanced ion channel function in hESC-CM clusters We visualized developmental changes in the gene expression of ion channels through semiquantitative RT-PCR (Fig. 3A), and measured mRNA levels through qRT-PCR (Fig. 3B). As controls, total RNAs from human fetal and adult hearts were used. The 1.5M, 3.5M, and 5.5M hESC-CMs showed no obvious differences in expression levels of HERG, KVLQT1, and HCN1. The 8M hESC-CMs, however, exhibited much higher expression levels than adult hearts. HERG1b and αMHC showed developmental regulation in response to replating culturing of hESC-CMs. We next explored whether continuous replating culturing enhanced not only transcriptional activity but also electrophysiological phenotype in hESC-CMs. Remarkable dosedependent shifts of Na+–K+ intervals in response to E4031 were clearly recorded in 8M hESC-CMs using MEA (Fig. 3C). Shortened APD in response to E4031 was characteristic of early-stage hESC-CMs (Fig. 3D and D21). The earliest detection of E4031-induced LQT occurred in D34 hESC-CMs maintained under replating/3D culturing conditions. D110 hESC-CMs permitted clear recording of E4031-induced Na+– K+ interval prolongation at relatively low concentrations (10–100 nM) (Fig. 3D). D222 hESC-CMs did not respond to low concentrations of E4031, and maintained strong rhythmic beatings even at high concentrations of E4031 (100–200 nM) (Fig. 3E). Treatment with nifekalant and dl-sotalol, two other compounds that are known to induce LQT through their interactions with the HERG channels, induced Na+ –K+ interval prolongation in a dose-dependent manner (Fig. 3E). It is worth noting that, although dl-sotalol is well known to trigger false negative results in in vitro HERG assays (Redfern et al., 2003), the results concerning these drugs that we obtained through this MEA system using long-term replating/3D cultured hESC-CMs actually corresponded well with clinical/in vivo information. This shows that enhanced cardiac function is required for proper MEA evaluation of ion channel inhibitors.

Long-term maintenance of a pacemaker lineage in hESC-CM clusters Some HCN channels conduct cations that contribute to pacemaker function by accelerating repolarization and initiation of the next depolarization, such as Na+, Ca2+, and

205 K+. Of the four HCN genes (Biel, 2009; Accili et al., 2002), we analyzed HCN1 and HCN4 because they directly contribute to the native If currents in SAN cells by coassembling into heteromeric channels (Xue et al., 2007; Altomare et al., 2003). HCN1is the fastest channel, and HCN4 is an essential channel most highly expressed in SAN cells (Ishii et al., 1999; Liu et al., 2007; Stieber et al., 2003). In the beating hESC-CM clusters, anti-HCN4 antibody staining clearly revealed compact structures of densely packed HCN4+ CMs (Fig. 4A). This HCN4+ area was clearly isolated from the remaining part, which contained the MLC2v-positive ventricular cells (Fig. 4A). Cells in the HCN4+ area also expressed Cx43, which is involved in gap junction formation and propagation of electrical impulses in hearts (Fig. 4A) (Gros et al., 2004). Out of the TnT+ hESC-CMs dissociated from 1M, 3M, and 6M beating colonies, 29, 46, and 55%, respectively, were strongly stained with anti-HCN4 antibody (HCN4high); these cells were small and less sarcomeric, and their numbers gradually increased in the later stages (Fig. 4B). On the other hand, the HCN4-negative cells (HCN4(−))were mostly large and sarcomeric, and were often judged as polynucleic (Fig. 4B). Furthermore, HCN1 was expressed in 8M hESC-CMs at extremely high levels (Fig. 4C). This suggests that pacemaker cell masses can be maintained for long periods through replating culturing of beating clusters. Patch-clamp recordings also showed that our 8M CMs contained both ventricular cells and pacemaker cells. They were identified by their ventricular-like AP structures possessing relatively deep resting potentials (− 55 to −70 mV), steep depolarization curves, and slow repolarization process (Fig. 4D, top) and by their biological pacemakerlike shallow resting potentials (− 40 mV) and rhythmic automaticity (Fig. 4D, bottom), respectively. Pacemaker function was detected in MEA recordings in the form of zatebradine-sensitive currents. As expected, zatebradine reduced Na+ currents and spontaneous rhythm in beating 8M CM clusters; aspirin, in contrast, which we tested as a negative control for QT test, had no effect on their rhythmic firings (Fig. 4E). We next analyzed HCN currents via patch-clamp recordings in the presence of the If-specific inhibitor zatebradine in combination with BaCl2 insensitivity; BaCl2 is a known blocker of the IK1 current, another major inward current that functions under the hyperpolarization state. In addition, cell size typically influences the results of patch-clamp recordings. Thus, we evaluated the amplitudes of HCN currents by their current density values (pA/pF). We detected a functional increase in HCN channels in 8 M CMs compared to 2M CMs (Fig. 4F), which was clearly demonstrated in the I–V relationships of HCN currents (Fig. 4G). Accordingly, replating/3D culturing of hESC-CMs could ensure not only long-term maintenance of both ventricular and pacemaker lineages, but also a high level of functionality, especially in the pacemaker lineage.

Acquisition of further mature phenotypes in long-term cultured hESC-CMs by replating We next analyzed patch-clamp recordings to explore whether continuous replating culturing gave hESC-CMs the other matured cardiac electrophysiological phenotype. First, we compared D28 and D231 replating/3D cultured hESC-CMs

206

T.G. Otsuji et al.

Figure 3 Enhanced function in hESC-CMs as a result of repeated replating culturing. (A) RT-PCR analyses for cardiac gene expression at 1.5M, 3.5M, 5.5M, and 8M. (B) qRT-PCR analyses for αMHC, KVLQT1, and HERG1b expression at D28, D61, and D231. D0, undifferentiated transgenic hESCs. (C) Extracellular field action potential of an 8M hESC-CM colony treated with E4031. Arrows, indicating the end of a repolarization wave. (D) E4031-induced Na+–K+ interval prolongation in replating/3D-cultured D21, D34, and D110 hESC-CMs (n = 6). (E) Dose-dependent elongation of relative Na+–K+ intervals due to addition of E4031, nifekalant, or dl-sotalol in D222 hESC-CMs (n = 6).

(described as 1M CMs and 8M CMs, respectively). In each sample, beating spheroids were randomly separated into two populations, one for qRT-PCR analyses and one for electrophysiological recordings. Thus, the results obtained by these two methods should point to a common conclusion. The major α and β subunits of Na+ channels (INa), Nav1.5 and Navβ1, were expressed in 8M CMs more highly than in 1M CMs, but less highly than in adult hearts (Fig. 5A). The major α and β subunits of long-lasting Ca2+ channels (ICaL), Cav1.2 and Cavβ2, were increased in 8M CMs, while the α subunit of transient calcium channels (ICaT), Cav3.2, was decreased (Fig. 5B). This corresponds to a previous report of prenatal Cav3.2 reduction in mice (Harrell et al., 2007). Electric stimulation above a threshold value induces voltage-gated ion channels to open and cations to flood into the cell, as long as responsible channels are functional. Ca2+ currents sensitive to the ICaL inhibitor nifedipine were recorded through stepwise test pulses from −40 to − 10 mV,

where INa channels are inactivated and do not theoretically open. Subsequently, nifedipine-insensitive INa currents were measured through stepwise test pulses from −60 to −10 mV. As a result, the greatest developmental differences were seen in INa, where they resulted in significantly strong depolarization spikes in 8M CMs (Fig. 5C, P b 0.01). Cotreatment with nifedipine and TTX had no effect on the remaining currents, demonstrating that the TTX-resistant INa current was a major component of the INa spike in 8M CMs (Figs. 5C and D). The evidence that TTX-insensitive Na+ currents represented most of the detectable Na+ currents was ascertained by the great reduction of automatic activation of Na+ currents by the sequential treatment with lidocaine followed by nifedipine in both 1M CMs and 8M CMs (data not shown). The I–V relationships of Na+ currents and the significant increases of maximum upstroke velocity (Supplemental Table 2, Vmax, P b 0.05) clearly demonstrate such functional increases during prolonged culturing (Fig. 5E).

Maturation in Beating hESC-derived Cardiomyocytes

207

Figure 4 Long-term maintenance of pacemaker function in hESC-CMs. (A) Immunostaining for HCN4 (red) and TnT (green) in a beating colony. Arrowhead, a condensed HCN4+ body; scale bar, 20 μm. Another colony stained for the ventricular marker MLC2v (green) with HCN4 (red); scale bar, 40 μm. A gap junction formation demonstrated by Cx43 immunostaining (red) in TnT+ hESC-CMs; scale bar, 10 μm. (B) A representative HCN4-negative TnT+ CM and a representative HCN4-expressing TnT+ CM; n, nucleus; scale bar, 10 μm. (C) qRT-PCR analyses for HCN4 at 1M and 8M. (D) Spontaneous ventricular-like or pacemaker-like action potential (AP) in patchclamp recordings of 8M hESC-CMs. (E) Zatebradine (Zat)-sensitive pacemaker currents, detected in MEA recordings. Arrowheads, the remarkable shift of the second depolarization pulses. Aspirin, a negative control. (F) Zat-sensitive pacemaker currents (corresponding to the distance between blue and red lines) and BaCl2-sensitive IK1 currents (corresponding to the distance between black and blue lines) detected in patch-clamp recordings of 2M and 8M hESC-CMs after serial treatment with BaCl2 followed by treatment with Zat. Holding potential, − 50 to −140 mV for 500 ms. (G) Current-voltage (I–V) relationship of HCN channels, demonstrating developmental differences between 2M CMs (n = 3) and 8M CMs (n = 4).

Next, we compared 1M CMs and 8M CMs in terms of amplitude of outward K+ currents during repolarization. Kv1.4 mRNA expression was elevated in 8M CMs (Fig. 6A). This gene is involved in the transient outward K+ current known as Ito, which is required in rapid inactivation of the depolarization phase, mainly in ventricular cells. After the plateau of depolarization, the slow-delayed rectifier K+ channels known as IKs are continuously opened during the repolarization phase. KVLQT1, a major subunit of IKs, was highly expressed

in 8M CMs (Fig. 6A). The rapid-delayed rectifier K+ channels known as IKr exhibited activation and inactivation kinetics twice in a repolarization phase; the later deactivation is fairly important in ending the repolarization phase (Sanguinetti et al., 1995; Zhou et al., 1998). HERG1b, a major constituent of IKr, was obviously upregulated in 8M CMs (Fig. 6A). Several other ion channel genes, namely, Kir6.2, CLCN4, ARP1A2, SLC8A1, and KCNK3, were also up-regulated in 8M CMs (Fig. 6A). Although mature ventricular cells specifically express

208

T.G. Otsuji et al.

Figure 5 Enhanced function of INa channels in hESC-CMs during long-term culturing. qRT-PCR analyses for (A) Na+ and (B) Ca2+ channel genes at 1M and 8M. (C) Significant increases in voltage-dependent INa currents during culturing, detected through patchclamp recordings (P b 0.01). Cells were treated with the ICaL-specific inhibitor nifedipine. Holding potential, −60 mV for INa; − 40 mV for ICa recordings. Depolarizing test pulses, − 60 to 70 or −40 to 50 mV for 500 ms, respectively, in 10-mV increments. (D) Amplitudes of INa and ICa are expressed as current densities (pA/pF). Depolarizing test pulses, −60 or −40 to −10 mV, respectively. Each graph displays the standard error of the mean. Voltage-dependent INa and ICa currents were summarized for only the ventricular-like CMs at 1M (n = 14) and 8M CMs (n = 8) as nifedipine-resistant currents and nifedipine-sensitive currents, respectively. (E) Current–voltage (I–V) relationship of INa and ICa channels, demonstrating developmental differences between 1M CMs and 8M CMs. Depolarizing test pulses, −60 to 70 mV for INa, − 40 to 50 mV for ICa, respectively, in 10-mV increments.

Kir2.1 as an important subunit coding IK1, it was expressed at a very low level even in 8M CMs. We detected IK1 currents as BaCl2-sensitive currents in 2M CMs and 8M CMs using patchclamp recordings (Fig. 4F); however their amplitude was relatively small and the differences between the two CM populations were not significant. These data suggest that progress toward full maturity might be more time-consuming in ventricular cells than in pacemaker cells. On the other hand, enhanced cardiac gene expression was directly related to ion channel function in IKr. Obviously reduced plateau and tail outward currents (Fig. 6B, open and closed circles, respectively) were detected in patch-clamp recordings when E4031 was used to inhibit IKr in 8M CMs (P b 0.05). The major component of K+ currents in 1M CMs was chromanol 293Bsensitive IKs (Fig. 6C). All of the data obtained through patch-clamp analyses are summarized in Supplemental Table 2, which shows that maturation in various ion channel components progresses cell autonomously in vitro.

Discussion Here, we demonstrated that continuous restoration of automaticity through replating/3D culturing facilitates progressive maturation of hESC-CMs, especially in the pacemaker lineage, allowing hESC-CMs to acquire the characteristics

of matured CMs and extending their lives by over a year. Such long-term maintenance of cultured cells that exit the cell cycle itself must be rare, though the unlikely possibility that CM progenitors are capable of proliferation cannot be excluded (Anversa et al., 2007). Our technique may help to address the further important question of how nonregenerative tissues like CMs maintain their function throughout life despite exiting the cell cycle in vivo. In mice, it has been shown that transplantation of mouse GFP+ ESC-CMs into injured hearts of syngenetic mice improves heart function for at least several months (Kolossov et al., 2006). Although hESC-derived beating embryoid bodies have also functioned as ectopic pacemakers after transplantation into hearts in vivo (Kehat et al., 2004), in vitro maturation of pacemaker cells from hESCs has never been controlled to date (Yanagi et al., 2007). Our replating/ 3D culturing technique, which facilitates selective retention and progressive maturation of pacemaker lineages in hESCCM clusters, will contribute to the development of a technique for regenerating biological pacemakers from not only hESCs but also hiPSCs. The longevity of hESC-CMs demonstrated here suggests the possibility of long-tem function in recipient hearts. We have already isolated a transgenic monkey ESC line using the same construct used here, and we have maintained GFP+ monkey ESC-CMs for over a year using the replating culturing technique (data not shown). GFP-expressing monkey ESC-CMs have potential for

Maturation in Beating hESC-derived Cardiomyocytes

209

Figure 6 Progressive maturation in voltage-gated K+ channels in hESC-CMs during long-term culturing. (A) qRT-PCR of potassium and the other ion channel genes in 1M and 8M hESC-CMs. (B) Patch-clamp analyses of E4031-sensitive IKr currents at 1M (n = 7) and 8M (n = 5). Holding potential, − 40 mV; depolarizing test pulses, − 40 to + 40 mV for 500 ms in 10-mV increments. Inhibition of the IKr currents due to E4031 was clearly detected in 8M CMs (P b 0.05). ○, Repolarization peak; ●, tail current. (C) Developmental changes in K+ channels are summarized in terms of current densities. Cells were sequentially treated with E4031 followed by chromanol 293B. IKr, E4031-sensitive currents; IKs, chromanol293B-sensitive currents; IK, E4031 and/or chromanol293B-sensitive currents.

use in transplantation experiments in place of human cells so that fewer ethical issues will arise. We detected HCN4+ pacemaker lineages and If currents using patch-clamp and MEA recordings with zatebradinemediated inhibition. Pacemaker potential in SAN, which is generated by a balance between the activation of inward currents and the deactivation of outward currents (Cho et al., 2003), requires broad function of various ion channels. In 8M CMs, most of the ion channels exhibited enhanced function. Consequently, 8M CMs showed dose-dependent Na+–K+ interval prolongation in MEA recordings in response to relatively high concentrations of potassium channel inhibitors. However, early hESC-CMs showed rather shortened Na+–K+ intervals. Although the mechanism by which LQT triggers arrhythmia is uncertain, it is evident that premature function of ion channels in hESC-CMs leads to hypersensitivity to the ion channel blockades. Recently, many researchers have worked on isolating patient-specific human iPSC lines from persons carrying familial arrhythmia susceptibility in order to analyze the responsible mutations and to search for chemical compounds that enhance or reduce the symptoms. By now, more than 90 genetic mutations have been mapped in the HERG gene; this number corresponds to 45% of the total number of known

LQTS-related mutations (Splawski et al., 2000). The HERG gene encodes the α subunit that assembles with MiRP1 β subunits to form IKr potassium channels (Abbott et al., 1999; Barhanin et al., 1996; Sanguinetti et al., 1996). This shows that functional levels of HERG molecules mostly affect IKr function. As described above, only when human iPSC-CMs are mature enough to acquire basic CM functions, they will respond to specific ion channel blockades on the basis of their affected ionic channel molecules. In the future, this knowledge may enable the development of genetic diagnostic techniques by which the likelihood that a particular drug will cause arrhythmia can be evaluated. For this to be realized, a system for generating functional tissue cells comparable to native tissue cells is required. Representative cardiac markers are expressed in premature hESC-CMs and do not always indicate that the cells are sufficiently developed to perform the corresponding function. In a previous study, GFP+ CMs induced from the bone marrow-derived mesenchymal stromal cells of αMHC-EGFP transgenic mice did not show typical cardiomyocyte APs and ion currents (Rose et al., 2008). In our replating/3D culturing technique for hESC-CMs, the cardiac maturation process requires at least a few months for the MEA-mediated QT test to become valid. Clearly, diagnosis of developmental stage

210 and degree of functionality in hESC-CMs is important if hESCCMs are to be used in qualified drug tests; an inaccurate diagnosis could result in overestimation or underestimation of the likelihood that a particular drug will cause arrhythmia. Thus, accurate indicators of the developmental stages of hESC-CMs will be necessary. One potential indicator is the expression level of cardiac HERG1b, because its expression is developmentally regulated and correlated with drug sensitivity to HERG channel inhibitors, as demonstrated here. The quality of beating hESC-CM clusters is likely controlled not only by the maintenance procedure but also by the strategy used to induce cardiomyocytes from hESCs. Recently, hESC-CMs have been induced in vivo by extrinsic factors including signaling pathways; endoderm-derived signals (BMP2, FGF8, Crescent, and Shh/Ihh) induce cardiac mesoderm formation, while mesoderm-derived signals (Chordin, Noggin, Wnt, and Serrate) act as inhibitory signals (Yang et al., 2008; Brand, 2003; Zhu et al., 2008; Leschik et al., 2008). Yet CM induction efficiency may also vary depending on the ESC line. In this study, we were able to obtain beating colonies from the KhES1 hESC line only through ESC-END-2 coculture using 1% KSR. The END-2 coculture system, which is known to induce various sorts of cardiac cells (Mummery et al., 2003), resulted in a functional unit in a beating cluster. In the replating/3D culture system, END-2 also helped cells reattach onto new culture dishes or MEA probes and maintain their spherical morphology. It has been shown that cell adhesion directly regulates gene transcription in the nucleus. In fact, suspension culture enhances the levels of histone acetylation in gastric carcinoma cells compared with adherent culture through intracellular contractile-related signaling activities (Kim et al., 2005). Global acetylation through the inhibition of histone deacetylase with tricostatin A increases αMHC expression and thereby improves contractile function in cultured CMs (Davis et al., 2005). This corresponds to the evidence that short-term 3D culture of hESC-CMs immediately enhances cardiac function. Hence, we hypothesized that transient nonadhesive periods during the replating procedure might release hESC-CMs from the global repressive status established under adhesive conditions. The right combination of an induction protocol for beating cellular units, including pacing cells and contracting cells, and a system of maintenance culturing that accelerates the cell-autonomous maturation process toward full maturity in ion channels will produce CMs suitable for use in a qualified MEA system for use in drug tests.

Methods Transgenic ESC lines The promoter region of human αMHC containing the 918-bp upstream segment between the transcriptional starting site and the third exon just before the open-reading frame was amplified from genomic DNAs of KhES-1 hESCs by PCR using the following primer sets: forward primer, 5′-CCAGGCACCTGCACCCTCTGG-3′; reverse primer, 5′-(including a SalI linker) TAGTCGACCTTGGTGCTTCCCCTGGGTCAGAG-3′, and then ligated into pGEM-T Easy Vector (Promega). The HindIII/SalI fragment was connected with EGFP. The linearized αMHC-EFGP vector was electroporated into the KhES-1

T.G. Otsuji et al. line, a human ESC line that has been established from a Japanese female blastocyt. This line was selectively used, because of the stability of pluripotency during genetic manipulation. These hESCs were maintained on mitomycin C-treated mouse embryonic fibroblasts in Primate ES Cell Culture Medium (ReproCELL Inc.) supplemented with 5 ng/ml recombinant human bFGF (Suemori and Nakatsuji, 2006; Hasegawa et al., 2006). Transgenic ESC clones were selectively grown with G418 (Sigma-Aldrich). The hESC line was used in conformity with the Guidelines for Derivation and Utilization of Human Embryonic Stem Cells of the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

Induction of hESC-CMs The END-2 cells were kindly provided by Dr. Christine Mummery, which were maintained as previously described (Mummery et al., 1991). END-2 coculture was carried out as previously described with some modifications (Mummery et al., 2003; Passier et al., 2005): briefly, hESC colonies were placed onto the END-2 cell layer in 12-well culture plates containing bFGF-free ESC culture medium for a day, and then cultured in differentiation medium (DMEM/F12 containing Lglutamine, nonessential amino acids, 90 μM β-mercaptoethanol, penicillin/streptomycin, and 20% FCS). KSR was supplemented at the beginning of the hESC-CM differentiation period. Numbers of beating ESC-CM colonies were counted under a microscope. Randomly selected beating colonies were picked on a tip of micropipette, and transferred onto new gelatin-coated dishes for replating culturing or onto 35-mm dishes of Sumilon Celltight X (Sumitomo Bakelite) containing differentiation medium for 3D culturing. The replating and medium exchange procedures were repeated every 2 weeks.

Immunostaining Cells were dissociated through treatment with 80-unit Papain (Worthington Biochemical) for 1 h at 37 °C, cultured on gelatin/PLL-coated cover glasses (BD Biosciences) for a few days, and then fixed with 4% paraformaldehyde. The specimens were pretreated with 2% skim milk and incubated with the first antibodies; these were αMHC (mouse monoclonal, 1:250; Abcam), αActinin (mouse monoclonal, 1:1600; Sigma-Aldrich), TnT (mouse monoclonal, 1:100; Santa Cruz Biotechnology), ANP (mouse monoclonal, 1:500; Millipore), MLC2v (mouse monoclonal, 1:100; Biocytex), Cx43 (rabbit polyclonal, 1:100; Millipore) and HCN4 (rabbit polyclonal, 1:200; Millipore). The second antibodies were Alexa546conjugated anti-mouse IgG, Alexa488-conjugated antimouse IgG, and Alexa546-conjugated anti-rabbit IgG (1:1000; all from Molecular Probes, Invitrogen). Nuclei were visualized through DAPI. Images were captured under fluorescent microscopy and modified using the ApoTome imaging system (AxioVision, Zeiss).

RT-PCR and qRT-PCR Total RNA was isolated with an RNeasy Mini Kit (Qiagen). DNA was completely digested with DNase I (Invitrogen). Briefly,

Maturation in Beating hESC-derived Cardiomyocytes 0.5 μg of total RNA was subjected to cDNA synthesis using the SuperScript III Reverse Transcriptase provided as part of the First-Strand Synthesis System (Invitrogen). RT-PCR was performed on 2.5 μl of cDNA in a 25-μl reaction mixture of EX Taq Hot Start Version (Takara Bio) for 30 cycles. qRT-PCR was performed in triplicate on 1 μl of cDNA using Syber Green PCR Master Mix on a 7500 Real Time PCR System (Applied Biosystems). The cycling conditions were as follows: 10 min at 95 °C, 40 cycles each consisting of 15 s at 95 °C, and then 1 min at 60 °C. Total RNA from human adult heart tissue (Clontech) was used to evaluate mRNA expression levels in ESC-CM mRNAs. All values were normalized with respect to β-actin expression and shown relative to the corresponding values in human adult hearts (AH = 100). Each graph displays the mean and standard deviation of three independent experiments. Expression levels in total RNAs from human fetal heart tissue (Clontech), hESCs (D0), and END-2 were also analyzed as controls (data not shown). The entire list of primer sets used in these experiments can be found in Supplemental Online Data Table 1.

Pharmacological ion channel inhibitors Inhibitors were obtained from Sigma-Aldrich unless otherwise noted. Stock solution was prepared containing (in mM) 1 E4031, 20 nifekalant hydrochloride (Wako Pure Chemical Industries), 50 sotalol hydrochloride, and 1 tetradotoxin (TTX) in H2O. Another stock solution was prepared containing (in mM) 20 chromanol 293B, 20 zatebradine hydrochloride, 200 acetylsalicylic acid (also known as aspirin; Cayman Chemicals), and 20 nifedipine in DMSO. Both stock solutions were stored at −20 °C until use.

Patch-clamp recordings Cells were isolated from beating hESC-CMs through treatment with 0.25% trypsin/EDTA for 5 min at 37 °C. The dissociated cells were cultured for 3–7 days on gelatin/PLLcoated cover glasses. Patch-clamp recordings were performed in a temperature-controlled room at approximately 25 °C. Data from the whole cell patch-clamp configuration were recorded from spontaneously beating cells using a HEKA EPC10 amplifier (HEKA Instruments Inc.). An external solution containing (in mM) 140 NaCl, 5 KCl, 10 HEPES, 2 CaCl2, 1 MgCl2, and 10 glucose was adjusted to pH 7.2 with NaOH. A pipette solution containing (in mM) 80 K-aspartate, 50 KCl, 1 MgCl2, 10 EGTA, 3 ATP-Mg, and 10 HEPES was adjusted to pH 7.2 with KOH. In some experiments, cells were incubated with (in μM) 0.1 E4031, 4 chromanol 293B, 4 zatebradine, 1 TTX, or 4 nifedipine for 5 min. The current– voltage relationship was obtained by holding the membrane potential at −40 to −60 mV and then delivering a depolarizing pulse for 500 ms in a stepwise manner, in 10-mV increments from (in mV) −60 to 70 for INa, − 40 to 40 for ICaL, − 40 to 50 for IKs and IKr, and −50 to −140 for If and IK1.

Extracellular recordings MEA recordings were made according to the procedure described previously (Oka et al., 1999). In brief, each intact beating hESC-CM colony was placed on a gelatin-coated

211 probe (MED-P2105 or P210A, Alpha MED Sciences). After 2–4 days of culturing, chronotropic responses were assessed at a sampling rate of 20 kHz through 10-min extracellular recordings made before and after the measurement medium was replaced with a medium containing each compound. The MEA probe devices were temperature controlled at 37 °C. Cells were incubated with (in μM) 0.01–0.2 E4031, 0.1–2 nifekalant, 1–50 sotalol, 5–20 zatebradine, or 100 aspirin for at least 10 min.

Acknowledgments We thank Yasuyuki Asai for communicating the details of the electrophysiological data, and Yuji Amagai and others at SCDI for their support. This research was supported by a grant from the New Energy and Industrial Technology Development Organization (06990420-0).

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.scr.2010.01.002.

References Abbott, G.W., Sesti, F., Splawski, I., Buck, M.E., Lehmann, M.H., Timothy, K.W., Keating, M.T., Goldstein, S.A., 1999. MiRP1 forms IKr potassium channels with HERG and is associated with cardiac arrhythmia. Cell 97, 175–187. Accili, E.A., Proenza, C., Baruscotti, M., DiFrancesco, D., 2002. From funny current to HCN channels: 20 years of excitation. News Physiol. Sci. 17, 32–37. Altomare, C., Terragni, B., Brioschi, C., Milanesi, R., Pagliuca, C., Viscomi, C., Moroni, A., Baruscotti, M., DiFrancesco, D., 2003. Heteromeric HCN1-HCN4 channels: a comparison with native pacemaker channels from the rabbit sinoatrial node. J. Physiol. 549, 347–359. Anversa, P., Leri, A., Rota, M., Hosoda, T., Bearzi, C., Urbanek, K., Kajstura, J., Bolli, R., 2007. Concise review: stem cells, myocardial regeneration, and methodological artifacts. Stem Cells 25, 589–601. Ashly, E., Niebauer, J., 2004. Cardiology Explained. Remedica Medical Education and Publishing, London. Barhanin, J., Lesage, F., Guillemare, E., Fink, M., Lazdunski, M., Romey, G., 1996. K(V)LQT1 and lsK (minK) proteins associate to form the I(Ks) cardiac potassium current. Nature 384, 78–80. Biel, M., 2009. Cyclic nucleotide-regulated cation channels. J. Biol. Chem. 284, 9017–9021. Brand, T., 2003. Heart development: molecular insights into cardiac specification and early morphogenesis. Dev. Biol. 258, 1–19. Cho, H.S., Takano, M., Noma, A., 2003. The electrophysiological properties of spontaneously beating pacemaker cells isolated from mouse sinoatrial node. J. Physiol. 550, 169–180. Davis, F.J., Pillai, J.B., Gupta, M., Gupta, M.P., 2005. Concurrent opposite effects of trichostatin A, an inhibitor of histone deacetylases, on expression of alpha-MHC and cardiac tubulins: implication for gain in cardiac muscle contractility. Am. J. Physiol. Heart Circ. Physiol. 288, H1477–H1490. Gros, D., Dupays, L., Alcolea, S., Meysen, S., Miquerol, L., Theveniau-Ruissy, L.M., 2004. Genetically modified mice: tools to decode the functions of connexins in the heart-new models for cardiovascular research. Cardiovasc. Res. 62, 299–308.

212 Gulick, J., Subramaniam, A., Neumann, J., Robbins, J., 1991. Isolation and characterization of the mouse cardiac myosin heavy chain genes. J. Biol. Chem. 266, 9180–9185. Harrell, M.D., Harbi, S., Hoffman, J.F., Zavadil, J., Coetzee, W.A., 2007. Large-scale analysis of ion channel gene expression in the mouse heart during perinatal development. Physiol. Genomics 28, 273–283. Hasegawa, K., Fujioka, T., Nakamura, Y., Nakatsuji, N., Suemori, H., 2006. A method for the selection of human embryonic stem cell sublines with high replating efficiency after single-cell dissociation. Stem Cells 24, 2649–2660. Ishii, T.M., Takano, M., Xie, L.H., Noma, A., Ohmori, H., 1999. Molecular characterization of the hyperpolarization-activated cation channel in rabbit heart sinoatrial node. J. Biol. Chem. 274, 12835–12839. Jang, J., Ku, S.Y., Kim, J.E., Choi, K., Kim, Y.Y., Kim, H.S., Oh, S.K., Lee, E.J., Cho, H.J., Song, Y.H., et al., 2008. Notch inhibition promotes human embryonic stem cell-derived cardiac mesoderm differentiation. Stem Cells 26, 2782–2790. Jeck, C.D., Boyden, P.A., 1992. Age-related appearance of outward currents may contribute to developmental differences in ventricular repolarization. Circ. Res. 71, 1390–1403. Jones, E.M., Roti Roti, E.C., Wang, J., Delfosse, S.A., Robertson, G.A., 2004. Cardiac IKr channels minimally comprise hERG 1a and 1b subunits. J. Biol. Chem. 279, 44690–44694. Kaji, K., Norrby, K., Paca, A., Mileikovsky, M., Mohseni, P., Woltjen, K., 2009. Virus-free induction of pluripotency and subsequent excision of reprogramming factors. Nature 458, 771–775. Keating, M.T., Sanguinetti, M.C., 2001. Molecular and cellular mechanisms of cardiac arrhythmias. Cell 104, 569–580. Kehat, I., Khimovich, L., Caspi, O., Gepstein, A., Shofti, R., Arbel, G., Huber, I., Satin, J., Itskovitz-Eldor, J., Gepstein, L., 2004. Electromechanical integration of cardiomyocytes derived from human embryonic stem cells. Nat. Biotechnol. 22, 1282–1289. Kim, Y.B., Yu, J., Lee, S.Y., Lee, M.S., Ko, S.G., Ye, S.K., Jong, H.S., Kim, T.Y., Bang, Y.J., Lee, J.W., 2005. Cell adhesion statusdependent histone acetylation is regulated through intracellular contractility-related signaling activities. J. Biol. Chem. 280, 28357–28364. Kolossov, E., Fleischmann, B.K., Liu, Q., Bloch, W., ViatchenkoKarpinski, S., Manzke, O., Ji, G.J., Bohlen, H., Addicks, K., Hescheler, J., 1998. Functional characteristics of ES cell-derived cardiac precursor cells identified by tissue-specific expression of the green fluorescent protein. J. Cell Biol. 143, 2045–2056. Kolossov, E., Bostani, T., Roell, W., Breitbach, M., Pillekamp, F., Nygren, J.M., Sasse, P., Rubenchik, O., Fries, J.W., Wenzel, D., et al., 2006. Engraftment of engineered ES cell-derived cardiomyocytes but not BM cells restores contractile function to the infarcted myocardium. J. Exp. Med. 203, 2315–2327. Leschik, J., Stefanovic, S., Brinon, B., Puceat, M., 2008. Cardiac commitment of primate embryonic stem cells. Nat. Protoc. 3, 1381–1387. Liu, J., Dobrzynski, H., Yanni, J., Boyett, M.R., Lei, M., 2007. Organisation of the mouse sinoatrial node: structure and expression of HCN channels. Cardiovasc. Res. 73, 729–738. Mummery, C.L., van Achterberg, T.A., van den Eijnden-van Raaij, A. J., van Haaster, L., Willemse, A., de Laat, S.W., Piersma, A.H., 1991. Visceral-endoderm-like cell lines induce differentiation of murine P19 embryonal carcinoma cells. Differentiation 46, 51–60. Mummery, C., Ward-van Oostwaard, D., Doevendans, P., Spijker, R., van den Brink, S., Hassink, R., van der Heyden, M., Opthof, T., Pera, M., de la Riviere, A.B., et al., 2003. Differentiation of human embryonic stem cells to cardiomyocytes: role of coculture with visceral endoderm-like cells. Circulation 107, 2733–2740. Nakagawa, M., Koyanagi, M., Tanabe, K., Takahashi, K., Ichisaka, T., Aoi, T., Okita, K., Mochiduki, Y., Takizawa, N., Yamanaka, S., 2008. Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts. Nat. Biotechnol. 26, 101–106.

T.G. Otsuji et al. Obreztchikova, M.N., Sosunov, E.A., Plotnikov, A., Anyukhovsky, E. P., Gainullin, R.Z., Danilo, P., Yeom, Z.H., Robinson, R.B., Rosen, M.R., 2003. Developmental changes in IKr and IKs contribute to age-related expression of dofetilide effects on repolarization and proarrhythmia. Cardiovasc. Res. 59, 339–350. Oka, H., Shimono, K., Ogawa, R., Sugihara, H., Taketani, M., 1999. A new planar multielectrode array for extracellular recording: application to hippocampal acute slice. J. Neurosci. Methods 93, 61–67. Passier, R., Oostwaard, D.W., Snapper, J., Kloots, J., Hassink, R.J., Kuijk, E., Roelen, B., de la Riviere, A.B., Mummery, C., 2005. Increased cardiomyocyte differentiation from human embryonic stem cells in serum-free cultures. Stem Cells 23, 772–780. Redfern, W.S., Carlsson, L., Davis, A.S., Lynch, W.G., MacKenzie, I., Palethorpe, S., Siegl, P.K., Strang, I., Sullivan, A.T., Wallis, R., et al., 2003. Relationships between preclinical cardiac electrophysiology, clinical QT interval prolongation and torsade de pointes for a broad range of drugs: evidence for a provisional safety margin in drug development. Cardiovasc. Res. 58, 32–45. Roden, D.M., 1998. Mechanisms and management of proarrhythmia. Am. J. Cardiol. 82, 49I–57I. Rose, R.A., Jiang, H., Wang, X., Helke, S., Tsoporis, J.N., Gong, N., Keating, S.C., Parker, T.G., Backx, P.H., Keating, A., 2008. Bone marrow-derived mesenchymal stromal cells express cardiacspecific markers, retain the stromal phenotype, and do not become functional cardiomyocytes in vitro. Stem Cells 26, 2884–2892. Sanguinetti, M.C., Jiang, C., Curran, M.E., Keating, M.T., 1995. A mechanistic link between an inherited and an acquired cardiac arrhythmia: HERG encodes the IKr potassium channel. Cell 81, 299–307. Sanguinetti, M.C., Curran, M.E., Zou, A., Shen, J., Spector, P.S., Atkinson, D.L., Keating, M.T., 1996. Coassembly of K(V)LQT1 and minK (IsK) proteins to form cardiac I(Ks) potassium channel. Nature 384, 80–83. Sartiani, L., Bettiol, E., Stillitano, F., Mugelli, A., Cerbai, E., Jaconi, M.E., 2007. Developmental changes in cardiomyocytes differentiated from human embryonic stem cells: a molecular and electrophysiological approach. Stem Cells 25, 1136–1144. Satin, J., Kehat, I., Caspi, O., Huber, I., Arbel, G., Itzhaki, I., Magyar, J., Schroder, E.A., Perlman, I., Gepstein, L., 2004. Mechanism of spontaneous excitability in human embryonic stem cell derived cardiomyocytes. J. Physiol. 559, 479–496. Schimpf, R., Wolpert, C., Gaita, F., Giustetto, C., Borggrefe, M., 2005. Short QT syndrome. Cardiovasc. Res. 67, 357–366. Spector, P.S., Curran, M.E., Keating, M.T., Sanguinetti, M.C., 1996. Class III antiarrhythmic drugs block HERG, a human cardiac delayed rectifier K+ channel. Open-channel block by methanesulfonanilides. Circ. Res. 78, 499–503. Splawski, I., Shen, J., Timothy, K.W., Lehmann, M.H., Priori, S., Robinson, J.L., Moss, A.J., Schwartz, P.J., Towbin, J.A., Vincent, G.M., et al., 2000. Spectrum of mutations in long-QT syndrome genes. KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 102, 1178–1185. Stett, A., Egert, U., Guenther, E., Hofmann, F., Meyer, T., Nisch, W., Haemmerle, H., 2003. Biological application of microelectrode arrays in drug discovery and basic research. Anal. Bioanal. Chem. 377, 486–495. Stieber, J., Herrmann, S., Feil, S., Loster, J., Feil, R., Biel, M., Hofmann, F., Ludwig, A., 2003. The hyperpolarization-activated channel HCN4 is required for the generation of pacemaker action potentials in the embryonic heart. Proc. Natl. Acad. Sci. USA 100, 15235–15240. Suemori, H., Nakatsuji, N., 2006. Generation and characterization of monkey embryonic stem cells. Methods Mol. Biol. 329 N, 81–89. Tada, M., Takahama, Y., Abe, K., Nakatsuji, N., Tada, T., 2001. Nuclear reprogramming of somatic cells by in vitro hybridization with ES cells. Curr. Biol. 11, 1553–1558.

Maturation in Beating hESC-derived Cardiomyocytes Takahashi, K., Yamanaka, S., 2006. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., Yamanaka, S., 2007. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131, 861–872. Tsika, R.W., Bahl, J.J., Leinwand, L.A., Morkin, E., 1990. Thyroid hormone regulates expression of a transfected human alphamyosin heavy-chain fusion gene in fetal rat heart cells. Proc. Natl. Acad. Sci. USA 87, 379–383. Woltjen, K., Michael, I.P., Mohseni, P., Desai, R., Mileikovsky, M., Hamalainen, R., Cowling, R., Wang, W., Liu, P., Gertsenstein, M., et al., 2009. piggyBac transposition reprograms fibroblasts to induced pluripotent stem cells. Nature 458, 766–770. Xue, T., Siu, C.W., Lieu, D.K., Lau, C.P., Tse, H.F., Li, R.A., 2007. Mechanistic role of I(f) revealed by induction of ventricular automaticity by somatic gene transfer of gating-engineered pacemaker (HCN) channels. Circulation 115, 1839–1850. Yanagi, K., Takano, M., Narazaki, G., Uosaki, H., Hoshino, T., Ishii, T., Misaki, T., Yamashita, J.K., 2007. Hyperpolarization-activat-

213 ed cyclic nucleotide-gated channels and T-type calcium channels confer automaticity of embryonic stem cell-derived cardiomyocytes. Stem Cells 25, 2712–2719. Yang, L., Soonpaa, M.H., Adler, E.D., Roepke, T.K., Kattman, S.J., Kennedy, M., Henckaerts, E., Bonham, K., Abbott, G.W., Linden, R.M., et al., 2008. Human cardiovascular progenitor cells develop from a KDR+ embryonic-stem-cell-derived population. Nature 453, 524–528. Zhang, J., Wilson, G.F., Soerens, A.G., Koonce, C.H., Yu, J., Palecek, S.P., Thomson, J.A., Kamp, T.J., 2009. Functional cardiomyocytes derived from human induced pluripotent stem cells. Circ. Res. 104, e30–e41. Zhou, Z., Gong, Q., Ye, B., Fan, Z., Makielski, J.C., Robertson, G.A., January, C.T., 1998. Properties of HERG channels stably expressed in HEK 293 cells studied at physiological temperature. Biophys. J. 74, 230–241. Zhu, W., Shiojima, I., Ito, Y., Li, Z., Ikeda, H., Yoshida, M., Naito, A.T., Nishi, J., Ueno, H., Umezawa, A., et al., 2008. IGFBP-4 is an inhibitor of canonical Wnt signalling required for cardiogenesis. Nature 454, 345–349.