Enhancement of Spontaneous Activity by HCN4 Overexpression in Mouse Embryonic Stem Cell-Derived Cardiomyocytes - A Possible Biological Pacemaker

Background Establishment of a biological pacemaker is expected to solve the persisting problems of a mechanical pacemaker including the problems of battery life and electromagnetic interference. Enhancement of the funny current (I f) flowing through hyperpolarization-activated cyclic nucleotide-gated (HCN) channels and attenuation of the inward rectifier K+ current (I K1) flowing through inward rectifier potassium (Kir) channels are essential for generation of a biological pacemaker. Therefore, we generated HCN4-overexpressing mouse embryonic stem cells (mESCs) and induced cardiomyocytes that originally show poor I K1 currents, and we investigated whether the HCN4-overexpressing mESC-derived cardiomyocytes (mESC-CMs) function as a biological pacemaker in vitro. Methods and Results The rabbit Hcn4 gene was transfected into mESCs, and stable clones were selected. mESC-CMs were generated via embryoid bodies and purified under serum/glucose-free and lactate-supplemented conditions. Approximately 90% of the purified cells were troponin I-positive by immunostaining. In mESC-CMs, expression level of the Kcnj2 gene encoding Kir2.1, which is essential for generation of I K1 currents that are responsible for stabilizing the resting membrane potential, was lower than that in an adult mouse ventricle. HCN4-overexpressing mESC-CMs expressed about a 3-times higher level of the Hcn4 gene than did non-overexpressing mESC-CMs. Expression of the Cacna1h gene, which encodes T-type calcium channel and generates diastolic depolarization in the sinoatrial node, was also confirmed. Additionally, genes required for impulse conduction including Connexin40, Connexin43, and Connexin45 genes, which encode connexins forming gap junctions, and the Scn5a gene, which encodes sodium channels, are expressed in the cells. HCN4-overexpressing mESC-CMs showed significantly larger I f currents and more rapid spontaneous beating than did non-overexpressing mESC-CMs. The beating rate of HCN4-overexpressing mESC-CMs responded to ivabradine, an I f inhibitor, and to isoproterenol, a beta-adrenergic receptor agonist. Co-culture of human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) with aggregates composed of mESC-CMs resulted in synchronized contraction of the cells. The beating rate of hiPSC-CMs co-cultured with aggregates of HCN4-overexpressing mESC-CMs was significantly higher than that of non-treated hiPSC-CMs and that of hiPSC-CMs co-cultured with aggregates of non-overexpressing mESC-CMs. Conclusions We generated HCN4-overexpresssing mESC-CMs expressing genes required for impulse conduction, showing rapid spontaneous beating, responding to an I f inhibitor and beta-adrenergic receptor agonist, and having pacing ability in an in vitro co-culture system with other excitable cells. The results indicated that these cells could be applied to a biological pacemaker.


Introduction
The establishment of a biological pacemaker is expected to solve the persisting problems of a mechanical pacemaker including the problems of battery life, lead breaks, infection, electromagnetic interference, appearance, and heart rate response during exercise.
Cardiac pacemaker activity originates in the sinus node. Spontaneous diastolic depolarization in phase 4 of the action potential is initiated in the sinus node, and the electrical impulse is conducted through the atria to the atrioventricular node. The sinus node can generate impulses faster than those generated in other areas. Different kinds of ionic currents are involved in the formation of spontaneous diastolic depolarization of the sinus node [1][2][3][4][5][6].
Among them, the funny current (I f ) flowing through hyperpolarization-activated cyclic nucleotide-gated potassium (HCN) channel is robustly present in the sinus node [7]. Since the HCN4 isoform is mainly expressed in the sinus node [8,9] and contains the cAMP-binding domain, adrenergic stimulation followed by intracellular cAMP accumulation increases the heart rate through augmentation of I f currents [7,10]. We previously reported a mutation of HCN4 found in a patient suffering from sick sinus syndrome (SSS) [11]. Other groups also reported an association between SSS and mutation of HCN4 [12,13]. Therefore, HCN4 is thought to be the key pacemaking ion channel [8].
In contrast, the inward rectifier K + current (I K1 ), which maintains resting membrane potential in working myocytes and antagonizes spontaneous activity, is negligibly small in sinus node cells [14,15]. Interestingly, genetic suppression of I K1 can give rise to pacemaker activity in ventricular myocytes [16]. Despite the presence of I K1 , Purkinje fibers show I f -dependent pacemaking, but the pacemaking is slow and not robust. Thus, one of the most important factors responsible for the rapid and robust pacemaking of the sinus node is the absence of I K1 [1,17].
That is, enhancement of I f flowing through HCN channels and attenuation of I K1 flowing through K ir channels are required for pacemaker cells to undergo spontaneous diastolic depolarization. Therefore, we generated HCN4-overexpressing mouse embryonic stem cells (mESCs) and induced cardiomyocytes that originally show weak I K1 currents to achieve two prerequisites for forming spontaneous diastolic depolarization, large I f currents and small I K1 currents, and investigated whether the HCN4-overexpressing mESC-derived cardiomyocytes (mESC-CMs) function as a biological pacemaker in vitro.

Plasmid construction
Vertebrate HCN4 proteins are highly conserved [11] and we used the rabbit Hcn4 in mESCs due to the traceability. CAG promoter-IRES-EGFP construct in pCAGIG (Addgene #11159) and kanamycin/neomycin resistance gene (Kan R /Neo R ) in pIRES2-AcGFP1 vector (Clontech) were amplified via the polymerase chain reaction (PCR) method using PrimeSTAR GXL DNA polymerase (TaKaRa), purified using a QIAquick PCR Purification Kit (Qiagen), and ligated using an In-fusion HD enzyme (Clontech), i.e., pCAGIG-Kan R /Neo R . Rabbit Hcn4 cDNA in the pCI vector (previously reported and kindly provided by Dr Takano) [9,11] was amplified via the PCR method and ligated with pCAGIG-Kan R /Neo R , i.e., pCAGIG-rabbit Hcn4-Kan R /Neo R . PCR primers are shown in Table 1.

Nucleofection
CGR8 cells were harvested using 0.05% trypsin/EDTA. Two μg of a non-linearized vector was used for nucleofection (Amaxa Nucleofector Ⅱ; A-023, which is optimized for a nucleofection program for mouse ESCs) and HCN4-overexpressing ESCs were selected using a medium containing 400 μg/ml G418 (Roche Applied Science) for 7 days. Three stable clones that were resistant to G418 and were EGFP-positive were selected and expanded.

Electrophysiology
The funny current (I f ) was recorded at room temperature by using the perforated patch-clamp technique. Cells were superfused with a bath solution containing (in mM): 132 NaCl, 4.8 KCl, 2.0 CaCl 2 , 1.2 MgCl 2 , 1.0 BaCl 2 , 2.0 MnCl 2 , 5.0 D-glucose, and 10 Hepes; pH 7.4. Pipettes (2-4 MO resistances) were filled with a pipette solution containing (in mM): 110 K-aspartate, 5.0 K 2 -ATP, 11 EGTA, 1.0 CaCl 2 , 1 MgCl 2 , and 5 Hepes; pH 7.2. Then 0.3 mg/mL Amphotericin B was added to the pipette solution to achieve patch perforation (10-20 MO; series resistance). The I f current was activated by a standard activation protocol. I f currents through activated HCN4 channels were obtained during hyperpolarizing test pulses of 5 seconds between -45 and -125 mV in 20 mV increments from a holding potential of -35 mV.
Action potentials (APs) were also measured with the perforated patch-clamp technique. mESC-CMs were dissociated using 0.25% Trypsin/0.02% EDTA for 5minutes and resuspended in high-glucose DMEM supplemented with 20% FBS, 50 μM 2-ME, and NEAA. Then 1 x 10 5 /cm 2 cells were replated on Matrigel-coated cover glasses and incubated for 48 hours. AP recordings were performed on monolayer cardiomyocytes. APs were measured respectively by the

Cardiac differentiation of mESCs
Emryoid bodies (EBs) were formed by cultivating 500 mESCs with 0.5 mmol/L 2-O-alpha-Dglucopyranosyl-L-ascorbic acid (AA-2G; Hayashibara Biochemical Labs) and without LIF in a hanging drop for 5 days (culture day 0 to day 5) [18]. On day 5, EBs were collected and plated on a 0.1% gelatin-coated dish with a medium containing 0.25 mM AA-2G and 10 μM IWR-1-endo (WAKO). On day 7, the medium was exchanged a medium consisting of modified Eagle's medium (MEM; GIBCO), Insulin-Transferrin-Selenium-A supplement (100×; ITS-A supplement; GIBCO), and 100 μg/mL kanamycin. The medium was changed every other day after plating on the dishes. On day 14, the medium was changed to no glucose DMEM (Gibco) with 4 mM L-sodium lactate (Sigma-Aldrich) [19]. Until day 21, the medium was changed every other day. On day 21, EBs were treated with 0.25% Trypsin/EDTA (invtrogen) at 37°C for 5 minutes and dissociated. Dissociated cells were resuspended in 3 ml of medium and loaded onto a discontinuous Percoll (GE Healthcare) gradient, containing 20 mM HEPES and 150 mM NaCl. The gradient consisted of 3 ml of a 40.5% Percoll layer over 3 ml of a 58.5% Percoll layer. After centrifugation at 1,500 xg for 30 minutes, cell layers were apparent. Cells at a 58.5% layer were collected [20]. The purified cells were resupended in high-glucose DMEM supplemented with 20% FBS, 1% NEAA, and 100 μM 2-ME.

Generation of human induced pluripotent stem cells (hiPSCs)
To investigate pacing ability of HCN4-overexpressing mESC-CMs, we used a co-culture system with human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs).
Human dermal fibroblasts (HDFs) were obtained from abdominal skin of 41-year-old healthy Japanese male, the corresponding author of this article, by punch biopsy. The skin biopsy sample was dissected into 10 evenly sized pieces. The dissected pieces were plated in 6-well plates containing DMEM supplemented with 10% FBS and 0.1 mg/mL kanamycin and placed in a 37°C incubator. The medium was changed every 3 days. After reaching confluence, the cells were dissociated with trypsin/EDTA and passaged.
Cardiac Differentiation of hiPSCs hiPSCs were dissociated into single cells with StemPro Accutase Cell Dissociation Reagent (GIBCO) and seeded on Matrigel-coated dishes at 1 x 10 5 /cm 2 in TeSRE8 medium supplemented with 10 μM Y27632 (WAKO). Five days after plating, at day 0, cells were treated with 6 μM CHIR99021 (Cayman) and 50 μM AA-2G in RPMI medium (GIBCO) supplemented with B27-insulin supplement (GIBCO) and kanamycin. After 24 hours, at day 1, the medium was changed to RPMI supplemented with B27-insulin and 50 μM AA-2G. On day 3, 5 μM IWR-1-endo (WAKO) was added and was removed during the medium change on day 5. Cells were maintained in RPMI supplemented with B27 supplement (GIBCO) starting from day 7, with the medium was changed every 2 or 3 days. On day 28, cardiomyocyte purification was performed by the Percoll gradient procedure mentioned above.

Reverse Transcription-PCR (RT-PCR) and quantitative PCR
Total RNA from mESC-CMs was extracted using a Trizol Plus RNA Purificarion Kit (Invitrogen). Complementary DNA was synthesized from 1 μg of total RNA using a QunatiTect Reverse Transcription Kit (QIAGEN) as prescribed in the manual and subjected to PCR amplification.
Taq DNA polymerase (Roche Applied Science) was used for RT-PCR and PCR products were subjected to electrophoresis in 2% agarose gels and stained with ethidium bromide. RT-PCR experiments were performed twice in each of the 3 cell lines. SYBR Green PCR Master Mix and Applied Biosystems 7300 Real-Time PCR Systems (Applied Biosystems) were used for quantitative PCR (q-PCR). The q-PCR data were processed with a standard curve method. PCR primers are shown in Table 1. The q-PCR experiments were performed in triplicate in each 3-cell line.

Counting beating rates of mESC-CMs
The purified mESC-CMs were plated on a laminin (Sigma Aldrich)-coated 96-well plate at 4 x 10 4 cells/well in high-glucose DMEM supplemented with 20% FBS, 1% NEAA, and 100 μM 2-ME and were incubated for 4 days.
We counted spontaneous beating frequencies and examined responses to ivabradine and isoproterenol.
Counting beating rates of hiPSC-CMs in a co-culture system with mESC-CMs Aggregates containing 1 x 10 4 mESC-CMs were made by using the hanging drop method for 3 days. The purified hiPSC-CMs were plated on a Matrigel-coated 24-well plate at 2 x 10 4 cells/ well and maintained in RPMI supplemented with B27. Three days after plating of hiPSC-CMs, the aggregate of mESC-CMs was plated on the well layered with hiPSC-CMs. After 7 days, synchronized beating rates of hiPSC-CMs away from aggregates were counted.

Statistical analysis
All data are expressed as means ± SD. Statistical analysis was performed by student's t test for unpaired data or one-way ANOVA with comparison of different groups by Dunnett's post hoc test. Values of P < 0.05 were considered to be significant.

Generation of mESC lines stably overexpressing rabbit Hcn4
A CAG promoter-rabbit Hcn4-IRES-EGFP/SV40 promoter-neomycin resistance gene plasmid vector (HCN4/EGFP vector) (Fig 1A) or CAG promoter-IRES-EGFP/SV40 promoter-neomycin resistance gene plasmid vector (EGFP vector) was transfected in mESCs, and colonies of cells that were resistant to G418 and were EGFP-positive were selected (Fig 1B and S1 Fig). We generated 3 mESC lines in each group (HCN4/EGFP or EGFP-stably-transfected mESCs group). All 3 selected HCN4/EGFP-stably-transfected mESC lines expressed rabbit Hcn4 (S2 Fig). As shown by immunostaining, RT-PCR and q-PCR, expression levels of Oct4 (Figs 1C and 2B and S2 and S3 Figs) and Nanog (Fig 2B and S2 and S3C Figs), which are essential for maintaining the self-renewing and undifferentiated state in mESCs, were not influenced by HCN4 overexpression in any of the 3 cell lines.

Establishment of purified HCN4-overexpressing mESC-CMs
We generated 3 mESC lines in each group. In all mESCs, cardiac differentiation was performed well via EB formation with or without HCN4 overexpression. Cardiomyocytes were purified by changing culture media to glucose-free and lactic acid-supplemented media. Most of the collected cells spontaneously beat and expressed cardiac sarcomere proteins as assessed by immunostaining: non-overexpressing mESC-CMs, 91.6 ± 6.5% α-actinin positive and 90.2 ± 6.7% troponin I-positive; HCN4 overexpressing mESC-CMs, 86.2 ± 1.3% α-actinin positive and 88.8 ± 6.9% troponin I-positive. Proportions of α-actinin, troponin I, myosin light chain (MLC)-2v and MLC-2a-positive cells were not significantly different between the non-overexpression and HCN4 overexpression groups (Fig 2A).
RT-PCR showed that undifferentiated markers (Oct4 and Nanog) had disappeared and that cardiac markers (Nkx2.5 and Tnnt2) were positive in all 3 mESC-CM lines in each group (HCN4/EGFP or EGFP-stably-transfected mESC-CM group) (Fig 2B and S2 Fig). Expression of the Cacna1h gene, which encodes T type calcium channel and generates diastolic depolarization in the sinoatrial node, was also confirmed. Additionally, genes required for impulse conduction including Connexin40, Connexin43, and Connexin45 genes, which encode connexins forming gap junctions, and the Scn5a gene, which encodes sodium channels, are expressed in the cells (Fig 2B and S2 Fig). q-PCR showed that rabbit Hcn4 mRNA levels were not significantly different among the 3 HCN4/EGFP mESC-CM lines (S2B Fig). HCN4-overexpressing mESC-CMs (HCN4/EGFP mESC-CMs) expressed a 3-times higher level of Hcn4 than did non-overexpressing mESC-CMs (EGFP mESC-CMs) (Fig 3A). Both HCN4-overexpressing and non-overexpressing mESC-CMs expressed lower levels of Kcnj2, which is involved in I K1 maintaining resting membrane potential, than did an adult mouse ventricle (Fig 3B).

Discussion
We established HCN4-overexpressing mESC-CMs as a candidate for a biological pacemaker. We generated the cells in order to achieve two prerequisites for forming spontaneous diastolic depolarization: large I f currents and small I K1 currents. The cells have five specific abilities to become an appropriate biological pacemaker. (1) HCN4-overexpressing mESC-CMs expresses high levels of Hcn4 and Cacna1h genes and a low level of the Kcnj2 gene. (2) The cells show large I f currents and rapid spontaneous beating, in other words, rapid impulse generation. (3) Additionally, genes required for impulse conduction including Connexin40, Connexin43, and Connexin45 genes, which encode connexins forming gap junctions and the Scn5a gene, which encodes sodium channels, are expressed in the cells. (4) Furthermore, the beating of the cells responds to an I f inhibitor and beta-adrenergic receptor agonist. (5) HCN4-overexpressing mESC-CMs have pacing ability in an in vitro co-culture system with other excitable cells.
Several investigators have reported overexpression of HCN channels as a strategy for generating a biological pacemaker. Injection of the Hcn1-ΔΔΔ, Hcn2 or Hcn4 gene [22][23][24][25][26][27] and transplantation of Hcn2-or Hcn4-overexpressing mesenchymal stem cells [28,29] have been reported. Furthermore, Tbx18 gene transfer in mature ventricular cardiomyocytes increases HCN4 channel expression and yields pacemaker activity [30]. We also generated HCN4-overexpressing mESC-CMs. The cells show large I f currents. Enhancement of I f currents flowing through HCN channels is a core strategy for generating a biological pacemaker. Mouse ESCderived myocytes expressed I f currents [31]. Since HCN4 expression gradually decreases in the late stage of cardiac differentiation [10,32], there is concern about whether pacemaker function can be maintained in the long term. Therefore, we overexpressed HCN4 in ESC-CMs. Several investigators also isolated or induced sinoatrial node-like cells that express HCN4 from ESCs [33][34][35]. Although it is not clear which method is the most useful, our HCN4-overexpressing system is an effective and easy method to obtain a large I f current and a large amount of cells.
Not only enhancement of I f current through HCN channels but also attenuation of I K1 current and presence of other currents through T-type Ca 2+ channels are required for diastolic depolarization [36,37]. Additionally, subsequent propagation to the surrounding working myocardium through connexins and Na v 1.5 channel is necessary as a pacemaker [38,39]. Our HCN4-overexpressing mESC-CMs also expressed these channels.
Ivabradine is an I f inhibitor and reduces the firing rate of pacemaker cells [40,41], and it has already been applied in a clinical setting [42]. HCN4-overexpressing mESC-CMs showed responses to ivabradine. This drug could regulate tachyarrhythmia caused by hyperexcitability of these cells. In addition, HCN4-overexpressing mESC-CMs showed responses to isoproterenol, a β1and β2-adrenoreceptor agonist, and might be useful to achieve heart rate response during exercise.
A sufficient number of cells is necessary to engraft well and capture the surrounding myocardium [28]. Since pluripotent stem cells have a strong self-renewal property, the use of pluripotent stem cell-derived cardiomyocytes is reasonable. Furthermore, hiPSCs are thought to be able to solve the likelihood of immune rejection [43,44]. Thus, this HCN4-overexpressing method might be applicable to hiPSC-CMs.
Recently, Inada et al reported that the spatial heterogeneous nature of the sinus node is important for its normal functioning and that the presence of Na channel and connexin 43 in the periphery may be essential for the node to be able to drive the atrial muscle [38,39]. Mouse ESC-CMs include a heterogeneous population: nodal-like, atrial myocyte-like and ventricular myocyte-like cells. Working myocyte-like cells from ESCs originally express Na channel and connexins. However, it is not obvious that spatial sorting would occur if ESC-CMs were injected in vivo. Further studies are needed to clarify this point.
Transplantation of human ESC-CMs into the ventricle of a complete atrioventricular block model animal has been reported [45]. Our experiment was only an in vitro experiment, and we need to evaluate the efficacy of these cells in bradycardia model animals.

Conclusion
We generated HCN4-overexpresssing mESC-CMs showing rapid spontaneous beating, responses to drugs and improved pacing ability in vitro. The results indicated that these cells could be applied to a biological pacemaker.  Fig 1C). Bar = 50 μm. B. Immunofluorescent staining of OCT4 (a and b) and nuclear DNA staining by Hoecst (c and d) in EGFP mESC-2 and 3 (immunofluorescent staining in EGFP mESC-1 shown in Fig 1C). Bar = 50 μm. C. q-PCR showed that Nanog and Oct4 mRNA levels were not significantly different in all mESC lines with or without HCN4 overexpression.