Functional expression and pharmaceutical efficacy of cardiac-specific ion channels in human embryonic stem cell-derived cardiomyocytes

Cardiomyocytes differentiated from human pluripotent stem cells provide promising tools for screening of cardiotoxic drugs. For evaluation of human pluripotent stem cell-derived cardiomyocytes for cardiotoxicity test, in the present study, human embryonic stem cells (hESCs) were differentiated to cardiomyocytes, followed by metabolic selection to enrich the differentiated cardiomyocytes. The highly purified hESC-derived cardiomyocytes (hESC-CMs) expressed several cardiomyocyte-specific markers including cTnT, MLC2a, and α-SA, but not pluripotency markers, such as OCT4 and NANOG. Patch clamp technique and RT-PCR revealed the expression of cardiomyocyte-specific Na+, Ca2+, and K+ channels and cardiac action potential in hESC-CMs. To explore the potential use of hESC-CMs as functional cardiomyocytes for drug discovery and cardiotoxicity screening, we examined the effects of bisindolylmaleimide (BIM) (I), which inhibits native cardiac Ca2+ channels, on the Ca2+ channel activity of hESC-CMs. We observed a similar response for the BIM (I)-induced modulation of Ca2+ channels between hESC-CMs and native cardiomyocytes through L-type Ca2+ channel current. These results suggest that hESC-CMs can be useful for evaluation of pharmaceutical efficacy and safety of novel drug candidate in cardiac research.

ventricular-like cardiomyocytes, they have been reported to be relatively immature 4 . In the current study, to generate hESC-CMs for application of drug screening, we adopted hESC-CMs differentiation method utilizing small molecules that modulate the Wnt signaling pathway, in which the differentiated cardiomyocytes showed the high expression of cardiac specific markers (cTnT, α-actinin, MLC2a, and MLC2v) and function (action potential) 1 .
Although there are several cardiac markers to identify successful differentiation of stem cells into cardiomyocytes, specific ion channels expressed in cardiomyocyte can be used as well for this purpose. Cardiomyocytes express tetrodotoxin-insensitive Na + channels, L-and T-type Ca 2+ channels (I Ca ), transient outward K + (I to ) channels, ultra-rapid delayed rectifier K + (I Kur ) channels, rapid delayed rectifier K + (I Kr ) channels, slow delayed rectifier K + (I Ks ) channels, inward rectifier K + (I K1 ) channels, and ATP-sensitive K + (K ATP ) channels 5,6 . Overall, ion channels in cardiomyocytes contribute to the generation of action potential shape. Therefore, the detection of functional ion channels and action potential is essential to prove successful differentiation of stem cells into cardiomyocytes.
In this study, we explored the differentiation of human embryonic stem cells (hESCs) into cardiomyocytes by detecting cardiac marker proteins and recording cardiac specific ion channels. Furthermore, to prove the pharmaceutical efficacy of hESC-CMs in cardiac research, we investigate the effect of bisindolylmaleimide (BIM) (I), which inhibits native cardiac Ca 2+ channels 7 , on Ca 2+ channels expressed in hESC-CMs.

Results
Phenotypic characterization of cardiomyocytes differentiated from hESCs. Prior to cardiomyocyte differentiation, pluripotent hESCs were clearly defined with a tight pack of brightly-clean, round cells ( Supplementary Fig. S1a). hESC colonies expressed SOX2, OCT4, TRA-1-81, TRA-1-60 and SSEA-4, pluripotent stem cell markers, but not cardiomyocyte differentiation marker cTnT ( Supplementary Fig. S1b and c). hESCs were differentiated to cardiomyocytes and purified with the protocol summarized in Supplementary Fig. S2. At day 20, the cells exhibited a mixture of poorly-differentiated cells and well-differentiated cardiomyocytes, demonstrating morphology transition from hESCs to cardiomyocytes (Fig. 1a). To enrich well-differentiated cardiomyocytes, the differentiated cardiomyocytes were exposed to glucose-depleted lactate medium. After selection with lactate medium, spontaneous and autonomous beating of hESC-CMs was observed (Supplementary Movie S1) and cardiac enrichment was verified by flow cytometry analysis. Immunofluorescent staining showed that the differentiated cells at day 30 extensively expressed α-SA, cTnT, and MLC2a with organized sarcomere structure (Fig. 1b). Western blot analysis showed expression of the cardiomyocyte-specific markers, including α-SA, cTnT, MLC2a, and MLC2v, while OCT4 and NANOG expression decreased considerably (Fig. 1c). To further confirm the cardiomyocyte-specific phenotypes of the hESC-CMs, gene expression was evaluated by qRT-PCR. The expression of cardiomyocyte-specific markers, such as GATA4, GATA6, and cTnT, increased markedly, whereas the expression of pluripotency markers, such as OCT and NANOG, decreased significantly (Fig. 1e). In addition, flow cytometry analysis showed that the majority of cells expressed cardiomyocyte markers, such as of cTnT, MLC2a, and α-SA, but not pluripotency markers, such as TRA-1-60 and SSEA3 (Fig. 1f). However, the expression levels of α-SA and cTnT in hESC-CMs were markedly lower than those in adult human left ventricle tissues (adult hLV) (Fig. 1c). Moreover, the expression levels of Cx43, a component of cardiomyocyte gap junction, and SERCA2a, a calcium ATPase related to cardiomyocyte function 8,9 , were substantially lower in hESC-CMs than that in adult hLV. To further confirm the immaturity of hESC-CMs, we quantified the ratio of myosin heavy chain (MHC) isoforms, a well-known cardiomyocyte-specific marker, by using the qRT-PCR. In human hearts, there are more β-MHC than α-MHC at all stages; however, it has been reported that there is more α-MHC in fetal than adult hearts 10,11 . The hESC-CMs exhibited relatively lower ratio of β-MHC/α-MHC expression than adult hLV (Fig. 1d). These data suggest that hESCs can be differentiated into cardiomyocytes in vitro, but the hESC-CMs exhibit immature phenotypes.
Characterization of Na + channel activities in hESC-CMs. Voltage-gated Na + channels are typical of cardiomyocytes. Among nine subtypes, Tetrodotoxin-insensitive subtypes, such as Nav1.5 and Nav1.8, have been identified in cardiomyocytes 6 . To record the Na + current, electrical activities of K + channels were inhibited by inclusion of Cs + in the external and internal solutions. Na + currents were elicited by step depolarizing pulses from −100 to +70 mV in steps of 10 mV with holding potential of −120 mV. As shown in Fig. 2a, Na + currents were rapidly activated and then inactivated after reaching the peak. Maximal peak currents were observed in voltage of −40 mV (Fig. 2c, open circle), suggesting that the recorded currents were Na + currents. To confirm the Na + current, we recorded Na + current using similar pulse protocols, then returned potential at −40 mV for 1 s to inactivate Na + current. Figure 2b and c (closed circle) showed that the most Na + currents induced by second steps pulses were inhibited. To further prove the molecular identity of the recorded Na + currents, we compared the expression levels of Nav1.5 subtype between hESCs and hESC-CMs. As shown in Fig. 2d, the expression level of Nav1.5 subtype was greater in hESC-CMs than hESCs, supporting functional expression of Na + channels in hESC-CMs.

Characterization of Ca 2+ channel activities in hESC-CMs.
To record the L-type Ca 2+ current, electrical activities of other K + channels were reduced by addition of Cs + in the external and internal solutions, as similar to Na + current recordings. Furthermore, holding potential was maintained at −50 mV to inhibit Na + and T-type Ca 2+ channels. L-type Ca 2+ currents were evoked by applying step depolarizing pulses from −50 mV to +50 mV in steps of 10 mV at −50 mV holding potential. As shown in Fig. 3a, voltage-dependent inward currents were recorded. These currents have maximal peak current at voltage of 0 mV (Fig. 3c). To verify that recorded currents were L-type Ca 2+ currents, we applied L-type Ca 2+ channel inhibitor, Nifedipine. Treatment of hESC-CMs with 10 µM Nifedipine effectively inhibited the recorded current throughout the whole voltage range, which suggested that the recorded currents were L-type Ca 2+ current ( Fig. 3b and c). We also investigate the molecular subtypes of Ca 2+ channels. As shown in Fig. 3d, the expression levels of L-type Ca 2+ channel subtypes (Cav1.1 and Cav1.2) and T-type Ca 2+ channel subtype (Cav3.1) were highly increased in hESC-CMs. It has been reported that Cav1.2 and Cav3.1 are associated with cardiac muscle [12][13][14] . These results confirm the functional expression of Ca 2+ channels in the hESC-CMs.
Characterization of K + channel activities in hESC-CMs. To record the general K + current, we simply applied step depolarizing pulses (between −120 and +40 mV from holding potential of −70 mV) with the L-type Ca 2+ channel inhibitor Nifedipine (10 µM). As shown in Fig. 4a and b, shapes similar to native cardiac K + currents were observed. However, I K1 currents were not detectable in our recordings. We investigate the expression of representative cardiac molecular subtype, Kv11.1 (I Kr ) in hESCs and hESC-CMs. Figure 4c shows that the Recordings of action potential in hESC-CMs. In order to investigate the electrical activity of differentiated cardiomyocytes, whole-cell patch clamp technique was performed. Figure 5a shows representative traces of spontaneous action potential recorded from hESC-CMs. The analysis of action potential such as resting membrane potential, action potential duration, cell capacitance, threshold potential, and upstroke velocity was described in Fig. 5b. Although the shape of action potential did not completely correspond to the native cardiac cells, the shape is similar to the shape of action potential when blocking I K1 channels in the native cardiomyocytes. These results are consistent with the patch clamp result that I K1 currents were not detectable.

BIM (I)-induced inhibition of the L-type Ca 2+ currents on hESC-CMs.
To explore whether hESC-CMs can be used for drug screening and cardiotoxicity test, we examined the effect of BIM (I), a specific inhibitor of protein kinase C (PKC), on L-type Ca 2+ currents in hESC-CMs, since our previous report suggested that BIM (I) inhibited the L-type Ca 2+ currents expressed in native cardiac (ventricular) myocytes 7 . Application of BIM (I) (1 μM) rapidly reduced the L-type Ca 2+ currents within 1 min ( Fig. 6a and b). At 0 mV, the current density of L-type Ca 2+ channel was −6.99 ± 0.20 pA/pF under control conditions and −3.50 ± 0.30 pA/pF in the presence of BIM (I) (1 μM) (Fig. 6c). Although the current density of L-type Ca 2+ current in hESC-CMs was smaller than that of native cardiac cells, the inhibition ratio of L-type Ca 2+ channel by BIM (I) was more sensitive in hESC-CMs (50% inhibition by 1 μM BIM (I)) compared with native cardiomyocytes (47% inhibition by 3 μM BIM (I)) 7 .
To test the concentration-dependent inhibition of BIM (I) on the L-type Ca 2+ currents, various concentrations of BIM (I) (0.03, 0.1, 0.3, 1, 3, 10, 30, or 100 μM) were applied before measurement of L-type Ca 2+ currents. Currents were elicited by one-step depolarizing pulses to 0 mV from a holding potential of −50 mV. As shown in Fig. 6d, L-type Ca 2+ currents were inhibited by treatment with increasing doses of BIM (I). The dose-response curve was obtained by nonlinear least-squares fit of the Hill equation to the concentration-inhibition data with an IC 50 value of 0.71 ± 0.20 μM and a Hill coefficient (n) of 1.16 ± 0.22 (Fig. 6e). These data suggest that BIM (I) inhibited L-type Ca 2+ channel of hESC-CMs in a concentration-dependent manner.
Effect of other PKC inhibitors on L-type Ca 2+ channel in hESC-CMs. To demonstrate whether BIM (I)-induced inhibition of L-type Ca 2+ channels was due to the inhibition of PKC, we tested the effect of another PKC inhibitor, Chelerythrine on L-type Ca 2+ channel. As shown in Fig. 7a, application of Chelerythrine (3 μM) did not affect the L-type Ca 2+ current and did not alter the BIM (I)-induced inhibition of L-type Ca 2+ channel. The L-type Ca 2+ channel current densities were −6.59 ± 0.41 pA/pF in control conditions, −6.09 ± 0.28 pA/pF in the presence of Chelerythrine, and −3.31 ± 0.03 pA/pF in the presence of Chelerythrine together with BIM (I) (1 μM) (Fig. 7b). These results suggest that the inhibitory effect of BIM (I) on L-type Ca 2+ channel is not due to inhibiting PKC activity.

Discussion
Ion channels are critical for all aspects of cardiac functions, including rhythmicity and contractility, by regulating cardiac action potential. The action potential shape and duration are determined by the sum of inward and outward currents including I Na current, I to current, I Kur current, I Kr current, I Ks current, I K1 current, and I Ca current 6,[15][16][17] . In the present study, we demonstrated differentiation of hESCs into functional cardiomyocytes expressing cardiomyocyte-specific markers and ion channels. The hESC-CMs expressed functional ion channels selective for Ca 2+ , Na + , and K + ions. The results of the current study may have important implications for the potential applications of the hESC-CMs for drug screening that modulate ion channels.
L-type Ca 2+ channels are key ion channels implicated in cardiac physiology and pharmacology, and largely responsible for the action potential plateau 17 . Whole-cell Ca 2+ transients in pluripotent stem cell-derived cardiomyocytes depend on both Ca 2+ influx via L-type Ca 2+ channels and intracellular Ca 2+ store release, as previously documented in mouse 18,19 and hESC-CMs 20,21 . The importance of the L-type Ca 2+ current in generating whole-cell Ca 2+ transients in these cells was manifested by the elimination of these transients in the absence of external Ca 2+ or in the presence of Nifedipine, a selective L-type Ca 2+ channel antagonist. A similar requirement for external Ca 2+ and the consequent trans-sarcolemmal Ca 2+ influx was documented in adult cardiomyocytes [22][23][24] , hESC-CMs 21 and mouse ESC-CMs 25,26 . We have previously reported that BIM (I), a PKC inhibitor, inhibited L-type Ca 2+ channels in rat native ventricular cells with IC 50 value of 3.31 µM 7 . In this study, we obtained the IC 50 value of 0.71 µM, therefore, it is likely that the inhibitory effect of BIM (I) on L-type Ca 2+ channels in hESC-CMs is more sensitive than that of rat ventricular cells. Although the reason for the higher sensitivity of hESC-CMs to BIM (I) is not clear yet, these results suggest that hESC-CMs can be useful for drug screening based on Ca 2+ channel.
Based on the action potential phenotypes recorded in isolated cells, differentiated CMs have been reported to have a heterogeneous population classified as atrial-, nodal-, or ventricular-like 27 . Cell culture conditions and differentiation protocols may influence these action potential properties 28,29 . In our results, the action potential recorded in hESC-CMs has a short action potential duration compared with native cardiomyocytes. Moreover, we showed that the L-type Ca 2+ channel current recorded in hESC-CMs was smaller than native cardiomyocytes. Ca 2+ current is a major inward current maintaining the plateau phase of action potential in native cardiomyocytes, and calcium influx results in cytoplasmic enrichment of Ca 2+ and ventricular myocyte contraction 23 . Therefore, inhibition of Ca 2+ current shortens action potential duration and increases the heart rate. The shortening of action potential duration by reducing Ca 2+ current amplitude is commonly observed in neonatal cardiomyocytes. The action potential duration and Ca 2+ current amplitude of neonatal cardiomyocytes were shorter and smaller than those of adult in guinea-pig ventricular myocytes 26 . In addition, Ca 2+ current amplitude of neonatal cardiomyocytes was small compared with adult cardiomyocytes in rabbit and murine models 26,30 . Therefore, it is likely that hESC-CMs used in our experiment possess Ca 2+ channel characteristics similar to neonatal cardiomyocytes.
Several reports have suggested that hESC-CMs have relatively longer action potential duration than that of our recordings [31][32][33] . Although we could not address the detail reasons for the relative short action potential duration compared to the results of other study groups, it may be caused by the small Ca 2+ current amplitude. In fact, Ca 2+ current recorded in the present study is half of the Ca 2+ current that reported in previous papers. For this  ScieNtific REPORtS | 7: 13821 | DOI:10.1038/s41598-017-14198-y reason, action potential duration recorded in our group is relatively short compared to other papers (Fig. 5b). Regarding the resting membrane potential, various results were elicited differently for each group. Some studies have reported that the resting membrane potential remains stable at −60 ~ −70 mV, however, the other studies have reported that the resting membrane potential was maintained at −40 ~ −60 mV, which elicits spontaneous action potential. It is difficult to explain the maintenance of resting membrane potential at −60 ~ −70 mV even though I K1 is not present. However, other ion channels, such as K ATP and SK Ca channel, are likely to play an important role in maintaining stable resting membrane potential. This will need to be clarified through further research.
Several families of voltage-gated potassium channels are expressed in cardiomyocytes and together provide the majority of the outward current responsible for action potential repolarization 6 . I K1 currents expressed in cardiomyocytes are important in maintaining and stabilizing resting membrane potential 34,35 . Furthermore, The  I K1 currents were assigned for shaping the initial depolarization, final repolarization, and resting phases of the ventricular action potential 30,35,36 . In our recordings, however, I K1 currents were not detectable (Fig. 4) and low I k1 currents may be another reason for shorting of action potential duration. Consistently, it has been demonstrated that the density of I K1 is either absent or significantly smaller in differentiated CMs than that reported for human native ventricular CMs 27 . Since previous reports suggested that action potential duration was reduced in embryonic and neonatal stages compared with the adult stage, and I K1 currents were crucial in the generation and shape of action potential 37,38 . These results support the notion that the hESC-CMs used in our experiment could be more similar with the cardiomyocytes from embryonic or neonatal stages. BIM (I), a specific PKC inhibitor, has known to decrease L-type Ca 2+ channels activity by altering the voltage-sensitivity of channels. In fact, BIM (I) shifted the steady-state inactivation curve toward a more negative potential suggesting that BIM (I) could directly interact with L-type Ca 2+ channels 7 . Considering that L-type Ca 2+ channels play an important role in cardiac function by regulating excitation-contraction coupling, action potential duration, and intracellular Ca 2+ -dependent signaling pathways, application of BIM (I) reduces cardiac contractility, action potential duration, and PKC-dependent signaling pathways. In the present study, we applied BIM (I) on the Ca 2+ channels expressed in hESC-CMs to evaluate pharmaceutical efficacy in cardiac research. As similar results with native cardiac Ca 2+ channels, BIM (I) efficiently reduced the Ca 2+ channel expressed in hESC-CMs suggesting that hESC-CMs will be useful for cardiac functional studies, specifically Ca 2+ channels.
Overall, the human pluripotent stem cell-derived CMs present a new and rapidly developing technology with exciting applications, and with further refinements it could pave the way for the development of personalized medicine for cardiovascular diseases 27 . Although the hESC-CMs have several barriers to direct use in clinical field, but hESC-CMs used in this experiment are valuable in the field of cardiac research, specifically cardiotoxicity and the development of new pharmaceuticals. To resolve the problems that the hESC-CMs exhibit the phenotypes of neonatal cardiomyocytes, further experiments are necessary to develop technologies for differentiation of human pluripotent stem cells into functionally matured cardiomyocytes.

Differentiation of hESCs into cardiomyocytes.
To differentiate H9 hESCs into cardiomyocytes, hESCs were dissociated to single cells by treatment with Accutase at 37 °C for 7-8 min. Single cell suspension was seeded on Matrigel-coated 6-well culture plate prepared as described 1 . In brief, dissociated cells were plated in mTeSR1 medium supplemented with Y27632 (5 μM) with a density of 1.4 × 10 6 cells/well, followed by changing to fresh medium every day for 3 days with 80-90% confluence. To initiate differentiation, the media was changed to RPMI/B27 minus insulin media containing the GSK3-specific inhibitor CHIR99021 (14 μM; Selleck Chemicals, Houston, TX) for 24 h (day 0), followed by RPMI/B27 minus insulin media for 48 h (day 1-3). Differentiating cells were treated with the Wnt signaling inhibitor IWP2 (5 μM; Tocris Bioscience, Bristol, United Kingdom) in RPMI/ B27 minus insulin media for 48 h (day [3][4][5]. At day 5 and day 7, culture media was replaced with fresh RPMI/B27 minus insulin media. From day 8, culture media was replaced with RPMI/B27 medium every 3 day up to day 20. Cells were incubated in a humidified incubator at 37 °C under 5% CO 2 . Purification of hESC-CMs. In the final phase of cardiomyocyte differentiation, cells were switched to glucose-free DMEM supplemented with Lactate (4 mM) on day 30 as described 39 . Differentiating hESCs were dissociated to single cells with 0.05% trypsin-EDTA at 37 °C for 5 min, collected with RPMI supplement with 20% FBS, and seeded on 0.1% gelatin-coated dish with RPMI supplement with B27.

Flow cytometry.
To prepare for single cells, hESCs were dissociated by Accutase treatment and hESC-CMs were dissociated with 0.05% trypsin-EDTA, followed by passing through 40 µm cell strainer. Cells were fixed with 4% paraformaldehyde for 20 min, washed with PBS, and permeabilized with 1% Triton-X-100 in PBS for 10 min. Cells were blocked with 5% BSA for 2 h and incubated with primary antibodies (TRA-1-60, TRA-1-81, SSEA4, and SSEA3 as hESC markers; αSA, cTnT, and MLC2a as cardiomyocyte markers) with 1:1000 dilution and overnight incubation at 4 °C. Secondary antibodies conjugated with fluorescent dyes were incubated with 1:2000 dilutions at 4 °C for 2 h. Fluorescence intensity of stained cells was measured using a CANTO II (BD Biosciences) and analyzed using FlowJo (ver 10, Tree Star Inc.).
Reverse transcription-polymerase chain reaction (RT-PCR). Total RNA was extracted from 80-90% confluent cultures using TRIzol reagent (Sigma) and reverse transcribed into cDNA using the Reverse Transcription cDNA Kit (#RT50KN; NanoHelix Co., Ltd). cDNA in 1 μL of the reaction mixture was amplified using the Ready-2 × -Go pre-mix PCR kit (#PMD008L; NanoHelix) and 10 pmol each of sense and antisense primers. The thermal cycle profile was as follows: denaturation at 95 °C for 30 s, annealing at 54 °C for 30 s depending on the primers used, and extension at 72 °C for 30 s. Each PCR reaction was carried out for 25-30 cycles. PCR products were analyzed by 1% agarose gel with ethidium bromide electrophoresis and photographed under UV trans-illumination. The primer sequences for RT-PCR are listed in the Supplementary Table S1.
Quantitative RT-PCR (q-PCR). Quantitative RT-PCR was performed using an ABI7500 (Applied Biosystems, Foster City, CA) sequence detection system, using SYBR Green PCR Master Mix (ABS-4309155, Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. Experiments were performed in duplicate, and the data were normalized to expression of GAPDH mRNA. Data were analyzed using the Δ (Δ CT) method and normalized to GAPDH. The primer sequences for q-PCR are listed in the Supplementary  Table S2.
Electrophysiological analysis. Whole cell membrane currents were recorded using the whole-cell patch clamp technique with an EPC-8 amplifier (Medical System Corp., Darmstadt, Germany) and an NI-DAQ-7 digital interface (National Instruments, Union, CA). Patch pipettes were pulled from borosilicate capillaries (Clark Electromedical Intruments, Pangbourne, UK) using a PP-830 puller (Narishige Scientific Instrument Laboratory, Tokyo, Japan) to have a 2-3 MΩ when filled with pipette solution. The voltage signals were sampled at a rate of 1-3 kHz for K + channel recordings, 4-6 kHz for Ca 2+ channel recordings, and 15-10 kHz for Na + channel recordings.

=
where ƒ means the relative current inhibition (ƒ = 1 − I drug /I control ) at the test potential, and [D] means the drug concentration.
Statistical analysis. The results are presented as means ± S.E.M. Differences were tested for statistical significance using Student's t-test, and a value of p < 0.05 was regarded significant.