Mesenchymal stem cells (MSCs) derived from human‐induced pluripotent stem cells (iPSCs) show superior proliferative capacity and therapeutic potential than those derived from bone marrow (BM). Ectopic expression of myocardin further improved the therapeutic potential of BM‐MSCs in a mouse model of myocardial infarction. The aim was of this study was to assess whether forced myocardin expression in iPSC‐MSCs could further enhance their transdifferentiation to cardiomyocytes and improve their electrophysiological properties for cardiac regeneration. Myocardin was overexpressed in iPSC‐MSCs using viral vectors (adenovirus or lentivirus). The expression of smooth muscle cell and cardiomyocyte markers, and ion channel genes was examined by reverse transcription‐polymerase chain reaction (RT‐PCR), immunofluorescence staining and patch clamp. The conduction velocity of the neonatal rat ventricular cardiomyocytes cocultured with iPSC‐MSC monolayer was measured by multielectrode arrays recording plate. Myocardin induced the expression of α‐MHC, GATA4, α‐actinin, cardiac MHC, MYH11, calponin, and SM α‐actin, but not cTnT, β‐MHC, and MLC2v in iPSC‐MSCs. Overexpression of myocardin in iPSC‐MSC enhanced the expression of SCN9A and CACNA1C, but reduced that of KCa3.1 and Kir2.2 in iPSC‐MSCs. Moreover, BKCa, IKir, ICl, Ito and INa.TTX were detected in iPSC‐MSC with myocardin overexpression; while only BKCa, IKir, ICl, IKDR, and IKCa were noted in iPSC‐MSC transfected with green florescence protein. Furthermore, the conduction velocity of iPSC‐MSC was significantly increased after myocardin overexpression. Overexpression of myocardin in iPSC‐MSCs resulted in partial transdifferentiation into cardiomyocytes phenotype and improved the electrical conduction during integration with mature cardiomyocytes.
- mesenchymal stem cell
Myocardin was identified as a transcription coactivator of serum response factor in mouse (Wang et al. 2001). The human myocardin gene expresses three alternative spliced mRNA variants (Torrado et al. 2003) with the longest isoform predominantly found in the heart (Du et al. 2003). Myocardin draws intensive interests because it activates the cardiac gene expression and influences heart function. The knockdown of the myocardin gene reduces the expression of cardiac markers in Xenopus and chick embryos (Wang et al. 2001; Small et al. 2005; Chen et al. 2008). Myocardin‐deficient adult mouse exhibits early onset of heart failure, dilated cardiomyopathy, and premature death within 1 week (Huang 2009) suggesting that myocardin is essential for the development of ventricular cardiomyocytes (Hoofnagle et al. 2011). Overexpression of myocardin induced the expression of various cardiac‐specific markers, such as cardiac troponin T (cTnT), MLC2a, MLC2v, α‐MHC, β‐MHC, GATA4, and sarco/endoplasmic reticulum Ca2+‐ATPase 2a (SERCA2a) in human bone marrow (BM)‐mesenchymal stem cells (MSCs) (van Tuyn et al. 2005). In addition, myocardin regulates the expression of multiple smooth muscle cell (SMC) differentiation‐related genes, including the smooth muscle (SM) α‐actin, SM22α, and SM‐calponin (Du et al. 2003; van Tuyn 2003; Wang and Wang 2003; Yoshida et al. 2003; van Tuyn et al. 2005).
Furthermore, overexpression of myocardin further improved the therapeutic potential of BM‐MSCs in a mouse model of myocardial infarction (Grauss et al. 2008), though the underlying mechanism remains unclear. Our recent studies have suggested that MSCs derived from human‐induced pluripotent stem cells (iPSCs) have more potent proliferative capacity and exhibit better therapeutic efficacy in tissue regeneration than BM‐MSCs (Lian et al. 2010). The aim of this study was to elucidate the effects of myocardin overexpression on MSC transdifferentiation, electrophysiological properties, and cellular integration of human iPSC‐MSCs with rat cardiomyocytes by examining the expression of cardiac marker genes and ion channel genes, and measuring their electrical conduction velocity. In order to obtain a thorough examination on the effect of forced myocardin expression, both lentiviral and adenoviral transduction systems were used in this study. To mimic the situation of pulsed overexpression of myocardin at the initial stage of transdifferentation, the non‐integration approach of adenoviral system was used to provide a transient forced expression of myocardin. However, the lentiviral system was used to examine the effects of a sustainable myocardin overexpression via genome integration approach.
Materials and Methods
Human iPSC‐MSCs were derived from iPSCs (iMR90)‐4 (Lian et al. 2010). Human BM‐MSCs were purchased from the Lonza Walkersville, Inc. (Walkersville, MD). The cells were cultured as monolayers in DMEM (Thermo Scientific, Waltham, MA) supplemented with 10% fetal bovine serum (FBS) (Invitrogen, Carlsbad, CA), 5 ng/mL bFGF (Invitrogen), and 5 ng/mL EGF (Peprotech, Hamburg, Germany).
Quantitative real‐time polymerase chain reactions
Total RNA was extracted from the human BM‐MSCs, human iPSC‐MSCs, human myocardin‐overexpressing iPSC‐MSCs, GFP‐overexpressing iPSC‐MSCs, or cardiomyocytes derived from the human iPSCs ([iMR90]‐4) (Lee et al. 2011), and it underwent RT‐PCR with various primers (Table 1) as detailed in our previous work (Zhang et al. 2012).
The expression of ion channels in iPSC‐MSCLVmyo and iPSC‐MSCLVGFP was examined by patch clamp techniques as detailed in our previous work (Zhang et al. 2012). In brief, the trypsinized cells were first seeded onto the glass coverslips in a four‐well plate. After a 30–45 min recovery period in the incubator, the cells were transferred into Tyrode's solution for patch clamp experiments. Using a HEKA EPC‐10 patch clamp amplifier and the PULSE v8.77 software (HEKA Instruments Inc., Southboro, MA), standard whole‐cell patch clamp recordings were performed at room temperature. The action potentials were recorded under the current (C)‐clamp at 37°C. Then, 1.5 mm thin‐walled borosilicate glass tubing (1.2 mm outer diameter, 0.69 mm internal diameter) (Sutter Instrument Co., Novato, CA) were used to prepare pipettes using a P‐97 Flaming/Brown micropipette puller (Sutter Instrument Co. Novato, CA). The typical pipette resistances were 3–4 MΩ when filled with internal pipette solution. After membrane rupture, the resting membrane potential was recorded under C‐clamp without a current input. The recorded membrane potentials were corrected for a liquid junction potential of +15.6 mV under room temperature, and +15.9 mV under 37°C. During the action potential recordings, the cells were held at 0 pA with 0.1–1 nA stimulation for 5 msec to elicit a response.
The human myocardin gene (NM_001146312.1) was first amplified via PCR from human cardiac cDNA and ligated into the pCR2.1‐TOPO shuttle vector (Invitrogen, Carlsbad, CA). The ORF was subsequently subcloned into pAD‐shuttle‐IRES‐GFP (Stratagene, La Jolla, CA) for adenovirus production. Myocardin‐ and GFP‐expressing adenoviruses were produced using the AdEasy™ XL Adenoviral Vector System (Stratagene, La Jolla, CA). The generated adenoviruses were purified using an Adeno Mini Purification ViraKit (Virapur LLC, SD, CA) before iPSC‐MSCs transduction.
Lentivirus production and transduction
Myocardin or GFP gene was inserted into lentivirus expression vector pSIN‐EF2‐Pur (Addgene, MA), and cotransfected with pCMV‐dR8.91 and pMD2G plasmid into 293FT cells (Invitrogen, Carlsbad, CA) using Lipofectamine 2000. Lentivirus were collected 3 days post transfection, filtered (0.45 μm), and used to transduce iPSC‐MSCs with polybrene (Millipore, Billerica, MA). The positively transduced cells were selected using 0.5 μg/mL puromycin‐containing medium after 3 days post transduction.
Cells seeded onto glass coverslips were washed with PBS, and fixed with 2% formalin for 30 min at room temperature. The cells were then incubated overnight with primary antibodies (1:200 dilutions in PBS) at 4°C. The primary antibodies included goat polyclonal anti‐myocardin (Santa Cruz Biotechnology Inc., Santa Cruz, CA), mouse antihuman CD29 (Stemgent, San Diego, CA), mouse monoclonal anti‐α‐actinin (sarcomeric) (Sigma‐Aldrich, Inc., St. Louis, MO), mouse anti‐myosin heavy chain (cardiac) (Upstate Biotechnology Inc., Waltham, MA), mouse monoclonal antihuman α‐actin (smooth muscle) (Dako, Copenhagen, Denmark), mouse monoclonal anti‐troponin T (cardiac) (Thermo Scientific), and rabbit monoclonal anti‐calponin‐1 (Millipore, Billerica, MA). After washing twice with PBS, the cells were incubated with the secondary antibodies for 30 min at 4°C. The secondary antibodies consisted of Alexa‐594 labeled rabbit anti‐goat IgG and Alexa‐647 labeled rabbit anti‐mouse IgG or Alexa‐594 labeled donkey anti‐goat IgG and Alexa‐647 labeled donkey anti‐rabbit IgG for myocardin and calponin costaining. After washing twice, the cells were mounted onto glass slides with mounting medium containing 4′,6‐diamidino‐2‐phenylindole dihydrochloride (DAPI) (Invitrogen). Immunocomplex was examined by fluorescent microscopy (Olympus IX81). To facilitate overlaying multicolored imaging, Alexa 647 was switched to a virtual color of green (512 nm) using Xcellence Pro version 1.1 software.
Neonatal rat cardiomyocytes isolation and multielectrode arrays recording
In order to obtain viable cardiomyocyte to assess the electric conductance of the iPSC‐MSCs, cardiomyocyte were isolated from neonatal rat hearts. The animal study protocol conforms to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health and was approved by the ethics committee of the University of Hong Kong (1851‐09). Hearts of neonatal Sprague‐Dawley rats (0–2 days old) were dissected and rinsed with modified Hanks' balanced salt solution (Invitrogen, Carlsbad, CA) (supplemented with 0.81 mmol/L MgSO4, 20 mmol/L HEPES, 100 U/mL penicillin and 100 μg/mL streptomycin, pH = 7.5 by NaOH). The ventricles were minced on ice and dissociated with 0.2% Trypsin 250 (BD Biosciences, San Jose, CA). The dissociated cells were centrifuged down and resuspended in prewarmed neonatal rat ventricular cardiomyocytes (NRVM) culture medium containing DMEM supplemented with 5% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, 1.5 mmol/L vitamin B12 (Sigma‐Aldrich), 10 mg/mL insulin (Sigma‐Aldrich), 10 mg/mL transferrin (Sigma‐Aldrich), and 0.1 mmol/L bromodeoxyuridine (Sigma‐Aldrich). The cells were then filtered (70 μm), and preplated onto a 10‐cm plate for exactly 1 h at 37°C.
The freshly isolated NRVMs were harvested, and mixed with myocardin‐ or GFP‐overexpressing iPSC‐MSCs at a ratio of 4:1, and seeded onto gelatin‐coated multielectrode arrays (MEA) plates (Multi Channel Systems, Reutlingen, Germany) at approximately 105 cells in total. The MEA plates used in this study contained 60 titanium nitride electrodes with a 200 μm interelectrode distance and 30 μm electrode diameter. The MEA plates combined with their supporting hardware allowed the simultaneous recording of extracellular electrical signals from all electrodes with one electrode as the ground. After allowing the NRVMs to settle down overnight, the electrical signals of the cell mixture were recorded and analyzed using the Cardio 2D+ software (Version 1.0.0) (Multi Channel Systems, Reutlingen, Germany).
The numeric data are presented as mean ± SEM. The statistical significance of the differences between group means was evaluated using analysis of variance (ANOVA) with post hoc Bonferroni test as appropriate. Differences with P < 0.05 were considered statistically significant.
Myocardin overexpression in human iPSC‐MSCs via lentivirus and adenovirus‐mediated transfer
Positively transduced cells (i.e., iPSC‐MSCLVmyo, iPSC‐MSCLVGFP, iPSC‐MSCADmyo, iPSC‐MSCADGFP) were verified by RT‐PCR and immunofluorescence staining. As shown in Figure 1A, the mRNA expression level of myocardin was significantly higher in iPSC‐MSCLVmyo and iPSC‐MSCADmyo than in their corresponding GFP controls. Quantitative RT‐PCR data revealed that the myocardin expression was 125‐ and 45‐fold higher after lentivirus and adenovirus‐mediated transfer, respectively, than the nontransduced cells (Fig. 1B). The myocardin protein was detected in the nucleus of iPSC‐MSCLVmyo and iPSC‐MSCADmyo, but not in their GFP controls (Fig. 2). Lentivirus transduced cells were selected with puromycin before further experiments.
Overexpression of myocardin activated cardiomyocytes and smooth muscle cells marker genes of human iPSC‐MSCs
The expression of the cardiac marker genes, including α‐MHC and GATA4, was observed in myocardin‐transduced human iPSC‐MSCs, although lower than that of the cardiomyocyte control (Fig. 1). The expression level of SERCA, SM22, and connexin43 (Cx43) did not demonstrate obvious changes upon myocardin transduction. Mef2c was significantly upregulated upon myocardin transduction (2.7‐fold) while the expression of connexin40 (Cx40) and connexin45 (Cx45) were downregulated in iPSC‐MSCLVmyo and iPSC‐MSCADmyo (all P < 0.05, Fig. 1B). Moreover, there was no detectable expression of cTnT, β‐MHC, and MLC2v in iPSC‐MSCLVmyo and iPSC‐MSCADmyo. On the other hand, iPSC‐MSCLVmyo activated the expression of the SMC marker gene MYH11. Figure 2 illustrates the positive staining for calponin, SM α‐actin, α‐actinin, and cardiac MHC, along with myocardin, in iPSC‐MSCADmyo and iPSC‐MSCLVmyo, but not in iPSC‐MSCADGFP, iPSC‐MSCLVGFP, iPSC‐MSCs, or BM‐MSCs.
Overexpression of myocardin‐modulated ion channel profile of human iPSC‐MSCs
According to the RT‐PCR results (Fig. 1), the expression of Kv4.3 (responsible for Ito) (2.2‐fold), SCN9A (for INa.TTX) (1.9‐fold), and CACNA1C (for ICa.L) (3.9‐fold) were upregulated, whereas the expression of KCNH1 (for IKDR) (2.5‐fold), KCa3.1 (for IKCa) (9.9‐fold), Kir2.1 and Kir2.2 (for IKir) (3.1‐ and 13.1‐fold, respectively) were downregulated in iPSC‐MSCLVmyo compared with that in iPSC‐MSCLVGFP(all P < 0.05). No significant difference in the expression of Clcn3 (for ICl) was observed between iPSC‐MSCLVmyo and iPSC‐MSCLVGFP.
The functional expression of the ion channels was then further examined by patch clamp in iPSC‐MSCLVmyo (n = 282) and iPSC‐MSCLVGFP (n = 101). Lentivirus‐mediated transduction combined with puromycin selection was adopted to ensure the successful transduction of myocardin or GFP in each cell for patch clamp. A current, likely ICl (Tao et al. 2007), was recorded in 26% of iPSC‐MSCLVmyo (72 of 282) and 3% of iPSC‐MSCLVGFP (3 of 101) as shown in Figure 3. The identity of ICl was confirmed using a chloride channel blocker DIDS. The current was inhibited by DIDS (150 μmol/L) in a representative iPSC‐MSCLVmyo (Fig. 3A) or iPSC‐MSCLVGFP (Fig. 3B). Figure 3A lower panel illustrates the I–V relationship curve of the DIDS‐sensitive current in iPSC‐MSCLVmyo (n = 7), by subtracting currents recorded after DIDS administration from the control current accordingly.
This inward component exhibited properties similar to IKir (Deng et al. 2006) were observed in both iPSC‐MSCLVmyo (9%, 18 of 282) and iPSC‐MSCLVGFP (12%, 11 of 101). The existence of IKir was determined using 0.5 mmol/L Ba2+, a specific blocker for IKir. As displayed in Figure 4A and B, the inward rectifying K+ current was reversibly suppressed by 0.5 mmol/L BaCl2 in both iPSC‐MSCLVmyo and iPSC‐MSCLVGFP. The I–V relationships of IKir, elicited with a 1.2 sec ramp protocol (from −120 to 0 mV) in a solution containing 5 mmol/L K+ (Tyrode's solution) or 20 mmol/L K+, and after application of 0.5 mmol/L Ba2+ in the bath solution are illustrated in a representative iPSC‐MSCLVmyo (Fig. 4A, lower panel) or iPSC‐MSCLVGFP (Fig. 4B, lower panel). Similar results were observed in 12 iPSC‐MSCLVmyo and 4 iPSC‐MSCLVGFP.
An outward current, likely BKCa current (Li et al. 2005), was activated in both iPSC‐MSCLVmyo (68%, 191 of 282) and iPSC‐MSCLVGFP (85%, 86 of 101) as shown in Figure 5A and B. The identity of BKCa was confirmed using the BKCa channel‐specific blocker paxilline (Fig. 5A and B). After administration of paxilline, the current activated at +60 mV, was significantly reduced from 2.9 ± 0.4 pA/pF to 1.0 ± 0.12 pA/pF in iPSC‐MSCLVmyo (n = 11, P < 0.001), and from 28.6 ± 8.4 pA/pF to 8.6 ± 1.9 pA/pF in iPSC‐MSCLVGFP (n = 9, P < 0.05).
Two types of paxilline‐resistant currents were recorded in iPSC‐MSCLVGFP but not in iPSC‐MSCLVmyo. As shown in Figure 5C, a paxilline‐resistant current in a representative iPSC‐MSCLVGFP was further inhibited by IKDR blocker 4‐AP. Paxilline (1 μmol/L) partially suppressed the membrane current at +60 mV from 19.2 ± 3.7 pA/pF to 9.2 ± 2.4 pA/pF. The remaining current was inhibited by 4‐AP (5 mmol/L) to 2.2 ± 1.5 pA/pF (n = 5, P < 0.05). The other type of current displayed properties of IKCa (Tao et al. 2007) was detected only in iPSC‐MSCLVGFP (6%, 6 of 101). In Figure 5D, the outward current in a representative iPSC‐MSCLVGFP was partially inhibited by paxilline (1 μmol/L), with the remaining current completely blocked by the IKCa blocker clotrimazole (1 μmol/L). Similar results were observed in five iPSC‐MSCLVGFP cells.
Two other types of currents were recorded in iPSC‐MSCLVmyo but not in iPSC‐MSCLVGFP. One was the transient outward K+ current Ito sensitive to 4‐AP (Li et al. 2005). Ito was detected in a small portion of iPSC‐MSCLVmyo (4%, 11 of 282) with a representative one shown in Figure 6A. The voltage‐dependent inactivation of Ito was assessed by the step potential at +50 mV after 1000 ms conditioning potentials from −110 to 0 mV (Fig. 6B). The Ito inactivation curve was obtained by plotting I/Imax of Ito (the peak value of Ito normalized to the maximum one) as a function of the conditioning potentials and fitting to the Boltzmann sigmoidal curve. The slope of the Ito inactivation curve was approximately 9.3 ± 1.7, and the half inactivated conditioning potential was approximately −64.0 ± 1.9 mV (n = 6). The time‐dependent recovery of Ito from inactivation was estimated with a paired‐pulse protocol (shown in the inset of Fig. 6C). The Ito recovery curve was fit to an exponential function‐one phase association with the time constant (ζ) 61.5 msec and half‐time 42.6 msec (n = 4).
The other current exclusively observed in iPSC‐MSCLVmyo was an inward current activated from approximately −30 to +60 mV and sensitive to tetrodotoxin (TTX). This phenomenon suggested the existence of INa.TTX (Li et al. 2005) in iPSC‐MSCLVmyo (4%, 12 of 282). Figure 6D demonstrates that TTX (100 nmol/L) reversibly inhibited the inward current, with limited effects on the outward current. Similar results were observed in four iPSC‐MSCLVmyo cells.
In summary, a similar functional ion channel profile (i.e., BKCa, IKir, ICl, IKDR, and IKCa) as in iPSC‐MSCs (Zhang et al. 2012) was observed in iPSC‐MSCLVGFP (85%, 12%, 3%, 56%, and 6%). Two distinct ion channels, Ito and INa.TTX were recorded in iPSC‐MSCLVmyo (68%, 9%, 26%, 4%, and 4%).
Furthermore, iPSC‐MSCLVmyo had a RMP of −25.9 ± 0.7 mV, which was more depolarizing than that of iPSC‐MSCLVGFP (−32.1 ± 1.6 mV) (P < 0.001). In addition, the cell size of iPSC‐MSCLVmyo was significantly increased upon myocardin transduction than iPSC‐MSCLVGFP (P < 0.001), as reflected by their respective average membrane capacitance (69.3 ± 2.5 pF vs. 22.4 ± 1.7 pF).
Myocardin overexpressed iPSC‐MSCs did not exhibit action potentials as mature cardiomyocytes
To assess whether overexpression of myocardin resulted in complete cardiomyocyte differentiation of iPSC‐MSCs, the action potential of iPSC‐MSCLVmyo, iPSC‐MSCLVmyo incubated in cardiomyocyte differentiation medium, iPSC‐MSCADmyo, and their GFP controls (n = 11–23) were also characterized (Fig. 7). However, none of these cells exhibited action potentials as mature cardiomyocytes.
Overexpression of myocardin significantly increased the electrical conduction velocity of human iPSC‐MSCs
The effects of myocardin overexpression on the cellular integration and conductance changes of iPSC‐MSC with cardiomyocytes were determined by measuring the electrical conduction velocity in a monolayer coculture of iPSC‐MSCADmyo or iPSC‐MSCADGFP with NRVMs on MEA recording plates (Fig. 8A). The electrical signals transmitted through the MEA plate were recorded as shown in a representative iPSC‐MSCADGFP (Fig. 8B) or iPSC‐MSCADmyo (Fig. 8C), coculture system. The electrical conduction velocities of iPSC‐MSCADmyo (n = 13; 0.13 ± 0.03 m/sec) were significantly faster than those of iPSC‐MSCADGFP (n = 15; 0.07 ± 0.01 m/sec) (Fig. 8D, P < 0.05).
Recent clinical studies have suggested that MSC is a potential cell source for cardiac regeneration in patients with myocardial infarction and heart failure (Siu and Tse 2012). The underlying mechanisms of the beneficial effect of MSC for cardiac regeneration likely mediated through multiple actions, including transdifferentiation into functional cardiomyocyte or fusion with native cardiomyocytes, paracrine effects, cell‐cell interactions, and integration with native cardiomyocytes (Williams and Hare 2011). Previous studies have shown that myocardin activates cardiac gene expression in BM‐MSCs (van Tuyn et al. 2005), and improved their therapeutic potential for myocardial infarction (Grauss et al. 2008).
This study is the first to investigate the effect of myocardin overexpression in human iPSC‐MSCs. Our previous studies have shown that iPSC‐MSCs exhibit higher proliferative potential and better therapeutic efficacy for tissue regeneration (Lian et al. 2010; Zhang et al. 2012). In this study, the myocardin overexpression in human iPSC‐MSCs resulted in partial transdifferentiation into cardiomyocyte phenotypes as reflected by the appearance of cardiac markers α‐MHC, GATA4, α‐actinin, and cardiac MHC, but not cardiac markers cTnT, β‐MHC, MLC2v, and cTnT protein. These findings were consistent with those observed in myocardin overexpression in BM‐MSC (van Tuyn et al. 2005; Grauss et al. 2008). Therefore, myocardin overexpression alone could not induce complete cardiomyocyte transdifferentiation from iPSC‐MSCs as observed in BM‐MSC. The incomplete transdifferentiation into mature cardiomyocyte phenotypes from myocardin‐transduced iPSC‐MSCs was further confirmed by the lack of mature action potentials recorded by patch clamping (data not shown). It is likely that those phenotypic changes of the MSC are due to secondary effect of multiple genes expression induced by myocardin rather than direct myocardin transcriptional activity to induce complete cardiac transdifferentiation. Moreover, our findings reveal several similar as well as different gene expression profiles related transient versus long‐term expression of myocardin using adenovirus and lentivirus transduction, respectively.
In this study, we characterized and compared the ion channel profiles of iPSC‐MSCLVmyo and iPSC‐MSCLVGFP to assess the electrophysiological effects of myocardin overexpression on iPSC‐MSCs. RT‐PCR results revealed that iPSC‐MSCLVmyo has higher expression levels of Kv4.3 (responsible for Ito), KCa1.1 (for BKCa), SCN9A (for INa.TTX), and CACNA1C (for ICa.L), but lower expression levels of KCa3.1 (for IKCa), KCNH1 (for IKDR), and Kir2.1 and Kir2.1 (for IKir) than those of iPSC‐MSCLVGFP. Both iPSC‐MSCLVmyo and iPSC‐MSCLVGFP had similar expression levels of Clcn3 (for ICl). Patch clamp results showed that BKCa, IKir, ICl, Ito, and INa.TTX were expressed in iPSC‐MSCLVmyo (68%, 9%, 26%, 4%, and 4%), compared with BKCa, IKir, ICl, IKDR and IKCa in iPSC‐MSCLVGFP (85%, 12%, 3%, 56%, and 6%). Overexpression of myocardin upregulated the functional expression of ICl channels and activated the functional expression of Ito and INa.TTX channels, but inhibited the functional expression of the IKCa IKDR and IKir channels.
The expression of the human Clcn3 gene is mainly found in the skeletal muscle and brain (Borsani et al. 1995). The inactivation of the Clcn3 gene causes severe myocardial hypertrophy and heart failure in knockout mice (Xiong et al. 2010). Clcn3 encodes the ICl channel, which is volume‐sensitive and can be activated under hypotonic conditions, as shown in human gastric epithelial cells (Jin et al. 2003), mouse BM‐MSCs (Tao et al. 2007), and bovine epithelial cells (Wang et al. 2000). In this study, the size of the iPSC‐MSCLVmyo greatly expanded compared with iPSC‐MSCLVGFP, as reflected by the increase in the average membrane capacitance from 22.4 ± 1.7 pF to 69.3 ± 2.5 pF. The percentage of cells that displayed ICl currents increased from 3% to 26%. The activation of the volume‐sensitive ICl channel possibly resulted from the increase in the cell volume of iPSC‐MSCLVmyo.
The Ito channel is encoded by Kv1.4, Kv4.2, and Kv4.3, which have been detected in the human heart (Gaborit et al. 2007). In this study, the expression of Kv4.3, but not Kv1.4 and Kv4.2, was found in iPSC‐MSCLVmyo. Therefore, the Ito current recorded in iPSC‐MSCLVmyo is likely mediated by the expression of Kv4.3. The Ito current is one of the critical components of the action potential in mature ventricular cardiomyocytes (Amin et al. 2010). 4‐AP sensitive Ito current has been observed in atrial and ventricular cardiomyocytes (Li et al. 1995, 1998). In this study, Ito current has been detected in only 4% of iPSC‐MSCLVmyo. These results suggest that overexpression of myocardin directs the transdifferentiation of iPSC‐MSCs to cardiomyocyte phenotypes, although at a very low efficiency.
In addition, myocardin can activate the expression of SMC marker genes in mouse ESCs (Du et al. 2003), fibroblast (Wang and Wang 2003), human BM‐MSCs and fibroblast (van Tuyn et al. 2005), and human ventricular scar fibroblast (van Tuyn 2003). In this study, myocardin induced the expression of SMC marker genes (i.e., SM22, MYH11, calponin, and SM α‐actin) in the iPSC‐MSCs. However, myocardin expression did not induce complete SMC transdifferentiation (Yoshida et al. 2004). Although the potential role of SMC transdifferentiation remains unclear, MSC‐derived SMC might enhance vasculogenesis.
The expression of SCN9A mRNA is significantly upregulated in iPSC‐MSCLVmyo compared with that in iPSC‐MSCLVGFP or non‐transduced cells. SCN9A might be responsible for the INa.TTX current recorded in iPSC‐MSCLVmyo (4%). The INa.TTX current has been observed in the human BM‐MSCs (Li et al. 2005) and cardiac fibroblasts (Li et al. 2009). However, no TTX‐resistant sodium current encoded by SCN5A has been recorded in the iPSC‐MSCLVmyo. The expression level of SCN5A is much higher than that of SCN9A in the human heart (Gaborit et al. 2007). Furthermore, other important ion currents, such as ICa.Land If that compose the action potential of cardiomyocytes (Amin et al. 2010) have not been detected in iPSC‐MSCLVmyo. These data again suggest that myocardin induces incomplete cardiomyocyte transdifferentiation from iPSC‐MSCs.
In addition to the induction of the functional expression of the Ito and INa.TTX channels, forced myocardin expression interferes with the expression of the IKCa and the ether à go‐go 1 potassium (responsible for IKDR) channels. The occurrence of the IKCa current (possibly encoded by the KCa3.1) in iPSC‐MSCLVGFP is approximately 6%, which is comparable with those in iPSC‐MSCs (5%, (Zhang et al. 2012)). The absence of the IKCa current in iPSC‐MSCLVmyo might have resulted from the decreased expression of KCa3.1. As we have shown previously, the ether à go‐go 1 potassium channel is critical in the regulation of the proliferation of iPSC‐MSCs (Zhang et al. 2012). The proliferation of iPSC‐MSCLVmyo has been greatly suppressed (data not shown). Myocardin possibly downregulates the expression of the ether à go‐go 1 potassium channel and thus inhibits the proliferation of iPSC‐MSCLVmyo.
With a conduction gap in an NRVM‐seeded MEA plate filled with desired cells, Pijnappels et al. (2006) measured the conduction velocities of BM‐MSCs and NRVMs to be 0.014 ± 0.004 m/sec and 0.168 ± 0.002 m/sec, respectively. Moreover, they have demonstrated that the electrical conduction of BM‐MSCs is caused by the expression of Cx43. Subsequently, they further showed that the forced myocardin expression induces the human ventricular scar fibroblasts to repair the conduction block of the NRVM field, possibly mediated by the upregulation of the expression of Cx40 and Cx45 (van Tuyn 2003). In contrast to these observations, our results showed that Cx40 and Cx45 expression, but not Cx43 were reduced upon myocardin transduction. While Cx40 and Cx45 are expressed in the embryonic heart, their expressions in the adult heart are mainly localized to pacemaker and conduction system. On the other hand, Cx43 is abundantly expressed in the myocardium. Therefore, downregulation of Cx40 and Cx45 after myocardin transduction may represent a transdifferentiation to a more mature ventricular phenotype (Verheule and Kaese 2013). Nevertheless, the conduction velocity of NRVMs cocultured iPSC‐MSCADmyo was significantly faster than that of their GFP counterparts. Therefore, forced myocardin expression increased conduction velocity possibly because of its effects on the ion channel expression of iPSC‐MSCs rather than changes in connexins expression, in particular, by activating the expression of INa. TTX. Moreover, this increase in conduction velocity might account for the superior therapeutic potential of myocardin transduced BM‐MSCs than that of BM‐MSCs themselves in the mouse model of myocardial infarction (Grauss et al. 2008). These results suggest that similar to BM‐MSCs, iPSC‐MSCs are capable of conducting the electrical signals from NRVMs. More importantly, forced myocardin expression enhances the electrical conduction of iPSC‐MSCs.
The forced myocardin expression in human iPSC‐MSCs lead to partial transdifferentiation into cardiomyocytes and SMC phenotypes, modified the expression profile of the ion channels, and improved the electrical conduction velocity during coculturing with NRVM. The changes in the electrophysiological properties of iPSC‐MSC can potentially improve their electrical coupling with native cardiomyocytes after transplantation, thus reducing the risk of proarrhythymias. However, the therapeutic effect of the ectopic expression of myocardin in iPSC‐MSCs for cardiac repair and regeneration warrants further investigation in in vivo animal models of cardiac injuries such as myocardial infarction.
Conflict of Interest
This study was supported by Small Project Funding (Project No. 201007176290 to J. C.‐Y. H.); Hong Kong Research Grant Council General Research Fund (HKU 780110M and HKU 781111M to H.‐F. T.); Hong Kong Research Grant Council Collaborative Research Fund (HKU8/CRF/09 to Q. L., C.‐W. S., H.‐F. T.); Theme‐based Research Scheme (T12‐705/11 to L. Q., C.‐W. S., H.‐F. T.); and Mr. Philip Wong Foundation Fund for Cardiac Stem Cell Research (H.‐F. T.).
These authors contribute equally to this work.
- Manuscript Received: December 13, 2013.
- Manuscript Revised: January 21, 2014.
- Manuscript Accepted: January 26, 2014.
- © 2014 The Authors. Physiological Reports published by Wiley Periodicals, Inc. on behalf of the American Physiological Society and The Physiological Society.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.