Enhancement of pacing function by HCN4 overexpression in human pluripotent stem cell-derived cardiomyocytes

The number of patients with bradyarrhythmia and the number of patients with cardiac pacemakers are increasing with the aging population and the increase in the number of patients with heart diseases. Some patients in whom a cardiac pacemaker has been implanted experience problems such as pacemaker infection and inconvenience due to electromagnetic interference. We have reported that overexpression of HCN channels producing a pacemaker current in mouse embryonic stem cell-derived cardiomyocytes showed enhanced pacing function in vitro and in vivo. The aim of this study was to determine whether HCN4 overexpression in human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) can strengthen the pacing function of the cells. Human HCN4 was transduced in the AAVS1 locus of human induced pluripotent stem cells by nucleofection and HCN4-overexpressing iPSC-CMs were generated. Gene expression profiles, frequencies of spontaneous contraction and pacing abilities of HCN4-overexpressing and non-overexpressing iPSC-CMs in vitro were compared. HCN4-overexpressing iPSC-CMs showed higher spontaneous contraction rates than those of non-overexpressing iPSC-CMs. They responded to an HCN channel blocker and β adrenergic stimulation. The pacing rates against parent iPSC line-derived cardiomyocytes were also higher in HCN4-overexpressing iPSC-CMs than in non-overexpressing iPSC-CMs. Overexpression of HCN4 showed enhancement of If current, spontaneous firing and pacing function in iPSC-CMs. These data suggest this transgenic cell line may be useful as a cardiac pacemaker.


Introduction
The number of patients with bradyarrhythmia is increasing with the increase in the number of patients with heart diseases associated with aging. The treatment for bradyarrhythmia is mechanical pacemaker implantation, but there are remaining issues such as battery replacement due to battery drain, infection, electromagnetic interference and rate response. Biological pacemakers generated by gene transfer or cell transplantation are expected to be used as a new treatment for resolving these problems.
TBX18 gene transduction by an adenoviral vector has been shown to convert ventricular myocytes into sinoatrial node-like cells in vitro and in vivo. [1,2] However, adenoviral vector-based gene therapy decays over time because adenoviral vectors are suitable for transient high expression and have immunogenicity that causes immune cells to eliminate reprogrammed cells. Therefore, this method is useful for temporary pacing but not for long-term use.
Several methods for the induction of sinoatrial nodelike pacemaker cells derived from human pluripotent stem cells (PSCs) have been reported. Yechikov et al. [3] increased the induction efficiency of sinoatrial node-like cells from 10-15% to 20-30% by adding a TGFβ receptor inhibitor, SB431542, in addition to the Wnt inhibitor IWR1 after cardiac mesoderm induction during cardiac induction of human induced pluripotent stem cells (iPSCs). Protze et al. [4] found that NKX2-5-negative cells among cardiomyocytes induced by BMP4 and Activin A from human embryonic stem cells (ESCs) exhibited sinoatrial node-like gene expression patterns. Furthermore, they described a method in which 80% of the induced cardiomyocytes become sinoatrial node-like cells when BMP4, retinoic acid, and FGF receptor inhibitor (PD173074), and SB431542 were added in addition to a Wnt inhibitor during the cardiac mesoderm stage. Although this method is very useful, it is likely that the concentrations of various growth factors and inhibitors need to be adjusted for each cell line. Liu et al. [5] reported that the induction efficiency of sinoatrial nodelike cells was 40-50% by adding BMP4, PD173074, and a retinoic acid receptor inhibitor, BMS189453, during the cardiac progenitor stage. Liang et al. [6] found that 70% of the induced cardiomyocytes were sinoatrial nodelike cells by removing the Wnt inhibitor after mesoderm induction, but the cardiomyocyte yield was as low as 20%. Pezhouman et al. [7] reported that adjustments of the cell seeding density and the concentration of a GSK3 inhibitor, CHIR99021, during mesoderm induction resulted in the generation of NKX2-5-negative TBX5-positive cardiac progenitor cells that gave rise to podoplaninpositive sinoatrial node-like cells with high efficiency. However, no functional analysis has been performed. Zhao et al. [8] transduced TBX3 with a lentiviral vector during the cardiac mesoderm stage, but the proportion of sinoatrial node-like cells was only 20%. For each method, there is room for improvement in induction efficiency, yield and simplicity.
Recently, designer cells with functions added by genetic modification, such as cells used in chimeric antigen receptor (CAR) T-cell therapy, are expected to become new tools for cell-based therapy [9,10]. Previously, we have transduced HCN4 into mouse embryonic stem cell-derived cardiomyocytes (ESC-CMs) to intensify their pacing function in vitro and in vivo [11,12]. HCN4 encodes a hyperpolarization-activated cyclic nucleotidegated potassium channel (HCN channel) producing a pacemaker current. It has been reported that human PSC-derived cardiomyocytes (PSC-CMs) express HCN4 and have some pacing potency that would be lost along maturation [13][14][15][16]. HCN4 overexpression seems simpler than other methods to generate pacemaker cells from human PSCs, however, this strategy have not been tested with human PSCs. In addition, the electrophysiological properties are different between rodent cardiomyocytes and human cardiomyocytes [17]. In this study, we investigated whether HCN4 overexpression in human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) improves pacemaker function in vitro.

Method summary
The strategy to generate HCN4-overexpressing iPSC-CMs is summarized in Fig. 1.
PCR primers are shown in Table 1.

Cell culture
Human iPSCs generated from dermal fibroblasts obtained from a healthy subject with retroviral vectors were used in this study [11,18]. The cells were dissociated with StemPro Accutase Cell Dissociation Reagent  [19].
The medium was changed with StemFit without Y-27632 the next day and cells were passaged once a week.

Cardiac differentiation of human iPSCs
Our cardiac differentiation protocol was based on that previously reported by Lian et al. [20]. Human iPSCs were dissociated into single cells with StemPro Accutase Cell Dissociation Reagent and seeded on 0.5 µg/cm 2 iMatrix-511 silk-coated 12-well plates (Cat#. 3513, Corning) at 3 × 10 5 /cm 2 in StemFit AK02N supplemented with 10 µmol/L Y-27632. The medium was changed with a medium without Y-27632 every day. Four days after plating, at day 0, cells were treated with 6 µmol/L CHIR99021 (Cat#.  Table 1.

Transmission electron microscopy (TEM)
HCN4-overexpressing iPSC-CMs were seeded on a Matrigel-coated 35 mm dish. The cells were fixed by chemical fixation and TEM was performed by Tokai Electron Microscopy (Aichi, Japan).

Statistical analysis
All data are expressed as means ± standard deviation. 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 using SPSS ver. 24. Values of P < 0.05 were considered significant.

Induction of HCN4-overexpressing cardiomyocytes
Cardiomyocytes were efficiently induced from HCN4overexpressing iPSCs with a commonly used cardiac differentiation protocol using a GSK3 inhibitor and a Wnt inhibitor (Additional file 1). Electron micrographs showed muscle fibers and mitochondria, which were consistent with cardiomyocytes (Fig. 3A). Overexpressed HCN4 protein was confirmed by immunostaining (Fig. 3B). EGFP expression indicating transgene expression was detected in cardiomyocytes 100 days after differentiation (Fig. 3C, (Additional files 2 and 3).
Enhancement of the I f current was also confirmed by the patch clamp technique (Fig. 3D and E).

Gene expression profile of HCN4-overexpressing cardiomyocytes
HCN4 mRNA level in HCN4-overexpressing iPSC-CMs was 30-times higher than that in non-overexpressing iPSC-CMs. There was no significant difference in were not upregulated in HCN4-overexpressing iPSC-CMs (Fig. 4A). Protein expression of NKX2-5 and MLC2v was also checked with immunostaining (Fig. 4B).

Enhancement of spontaneous firing in HCN4-overexpressing cardiomyocytes
HCN4-overexpressing cardiomyocytes showed a significantly higher rates of spontaneous firing and beating than that in non-overexpressing cardiomyocytes: 13.2 ± 1.7/15 s versus 7.8 ± 0.8/15 s ( Fig. 5A and B, (Additional files 4 and 5). Additionally, the frequency of spontaneous contraction was suppressed by ivabradine and was promoted by

Discussion
In this study, we generated HCN4-overexpressing cardiomyocytes from human iPSCs, and showed that HCN4 overexpression in human cardiomyocytes evoked a higher spontaneous firing frequency than that of control cardiomyocytes, resulting in enhanced pacing function in vitro. PSC-CMs are immature and lack I k1 current to maintain quiescent membrane potential and have a relatively high maximal diastolic potential (− 60 mV), which may not be sufficient to activate HCN channels [24]. However, a significant increase in I f was observed in HCN4-overexpressing cardiomyocytes at − 45 and − 65 mV as shown in Fig. 3E. In addition, the spontaneous contraction frequency was suppressed by the HCN channel inhibitor ivabradine (Fig. 5C), suggesting that the enhancement of diastolic depolarization by increased I f current caused by HCN4 overexpression led to an enhanced pacing effect of iPSC-CMs. So far, many groups, including us, have reported an increase in the frequency of sponteneous firing and the addition of a pacing function by transduction of HCN genes in mouse ESC-CMs, mesenchymal stem cells or the left ventricle [11,[25][26][27][28][29][30][31][32]. Results in this study are consistent with the findings of previous studies. As previously reported [23,33], it was possible to confirm long-term expression not only in iPSCs but also in iPSC-CMs by transduction of HCN4 into the AAVS1 safe harbor locus. We have also experienced loss of EGFP expression during differentiation in cardiomyocytes and non-cardiomyocytes derived from EGFP-positive iPSC lines without homologous recombination of the transgene (data was not shown).
HCN4-overexpressing cardiomyocytes were induced very efficiently since the commonly used efficient cardiac differentiation protocol using a GSK3 inhibitor and a Wnt inhibitor and the cardiomyocyte purification method can be applied [20,21,34]. Therefore, our strategy for generating cardiomyocytes with enhanced pacing function is simple and feasible compared to previously reported methods for generating sinoatrial node-like cells from human PSCs using complicated combinations of growth factors and compounds [4][5][6][7][8].
In the embryonic heart, HCN4 expression is first seen in the cardiac crescent, the source of the left ventricle and atria, and as it develops, it becomes localized to the conduction system, including the sinoatrial node [35,36]. In our study, there was no significant upregulation of sinoatrial node marker genes or working myocyte marker genes in HCN4-overexpressing iPSC-CMs as far as we examined with qPCR. This suggests that constitutive HCN4 overexpression did not change the cardiomyocyte subtype into sinoatrial node-like cells. The proportions of ventricular, atrial and nodal type in human PSC-CMs vary depending on the differentiation protocols and cell lines [20,37,38]. That is, the baseline cardiomyocyte phenotype might affect the spontaneous firing frequency and pacing function also when HCN4 is overexpressed. For better integration between transplanted cells and the recipient myocardium, HCN4-overexpressing ventricular myocytes might be useful for transplantation into the ventricle, while HCN4-overexpressing atrial myocytes might be better for transplantation into the atria.

Study limitations
Only in vitro studies have been conducted. It has not been investigated how much pacing can be done in vivo. Since the connexin expression patterns are different between the atrium and ventricles or between healthy and diseased hearts [39], the pacing ability in vivo might be affected by various conditions.
The firing frequency probably depends on the expression level of transduced HCN4. Although EGFP expression was detected for more than 3 months in vitro in this study, the maintenance of transgene expression over a longer period should be checked, especially in vivo.
In addition, the possibility of tachycardia is a concern because some groups have reported that transplantation of a large number of pluripotent stem cell-derived cardiomyocytes transplanted into an infarcted heart resulted in ventricular tachyarrhythmias [40][41][42]. However, it might be possible to control the pacing rate with drugs because spontaneous firing responded to an HCN channel inhibitor, ivabradine, in this study. Actually, Nakamura et al. also showed an anti-arrhythmic effect of ivabradine on iPSC-CM engraftment arrhythmias [43]. Cardiac Troponin I promoter-driven human HCN4 transgenic mice showed similar cardiac morphology at Spontaneous firing and beating rates in human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs). A Action potential configurations in HCN4-overexpressing (left) and non-overexpressing (right) iPSC-CMs. B Comparison of spontaneous beating rates in HCN4-overexpressing and non-overexpressing iPSC-CMs (n = 6 in each) (The data are shown as means ± standard deviation.). C Responses to 10 µmol/L ivabradine and 1 µmol/L isoproterenol in HCN4-overexpressing iPSC-CMs (n = 8 in each) (The data are shown as means ± standard deviation) birth but cardiac dilatation after birth due to dysregulation of calcium homeostasis and increased myocardial apoptosis [44]. Therefore, it is necessary to investigate the effect of long-term HCN4 overexpression on transplanted cells in vivo in the future.
Small animals such as mice and rats have a faster heart rate than humans (> 350 bpm). Even if atrioventricular block is performed to create a bradycardia model, its heart rate is more than 120 bpm (Saito et al. Int Heart J. 2018), making it difficult to evaluate the pacing function of human cells. Therefore, in vivo studies need to use large animals such as monkeys or pigs that have been treated with immunosuppressive drugs. In addition, tumorigenic potential should be assessed.

Conclusions
We generated HCN4-overexpressing human iPSC-CMs showing enhancement of I f current and pacing ability. Application of these cells with added pacing function to a biological pacemaker is expected in the future.