Cardiomyocytes derived from human embryonic stem (ES) cells are a potential source for cell-based therapy for heart diseases. We studied the effect of bone morphogenetic protein (BMP)-4 in the presence of fetal bovine serum (FBS) on cardiac induction from human H1 ES cells during embryoid body (EB) development. Suspension culture for 4 days with 20% FBS produced the best results for the differentiation of early mesoderm and cardiomyocytes. The addition of Noggin reduced the incidence of beating EBs from 23.6% to 5.3%, which indicated the involvement of BMP signaling in the spontaneous cardiac differentiation. In this condition, treatment with 12.5–25 ng/ml BMP-4 during the 4-day suspension optimally promoted the cardiomyocyte differentiation. The incidence of beating EBs at 25 ng/ml BMP-4 reached 95.8% on day 6 of expansion and then plateaued until day 20. In real-time PCR analysis, the cardiac development-related genes MESP1 and Nkx2.5 were upregulated in the EB outgrowths by 25 ng/ml BMP-4. The activation of BMP signaling in EBs was confirmed by the increase in the phosphorylation of Smad1/5/8 and by the nuclear localization of phospho-Smad1/5/8 and Smad4. The addition of 150 ng/ml Noggin considerably decreased the incidence of beating EBs and Nkx2.5 expression, and Noggin alone increased Nestin expression and neural differentiation in EB outgrowths. The cardiomyocytes induced by 25 ng/ml BMP-4 showed proper cell biological characteristics and a course of differentiation as judged from isoproterenol administration, gene expression, protein assay, immunoreactivity, and subcellular structures. No remarkable change in the extent of apoptosis and proliferation in the cardiomyocytes was observed by BMP-4 treatment. These findings showed that BMP-4 in combination with FBS at the appropriate time and concentrations significantly promotes cardiomyocyte induction from human ES cells.
- fetal bovine serum
- early mesoderm
cell-based therapy with stem cells has been extensively studied to repair damaged heart tissues (25). Embryonic stem (ES) cells established from the inner cell mass of blastocyst stage embryos possess the ability of self renewal, as well as the pluripotency to differentiate into cell lineages of all three germ layers. Several cardiac cell lineages with therapeutic significance have been induced from human ES cells, including pacemaker-like, Purkinje-like, and atrial/ventricular-like cells (11, 22). These cardiomyocytes derived from human ES cells are a potential source for cardiac regeneration.
The heart begins to develop in mesoderm during gastrulation. Two separate progenitor cell lineages have been known to participate in this process (5). The first heart field derives from the anterior lateral plate mesoderm and gives rise to the cardiac crescent. It mainly contributes to the formation of the left ventricle. The second heart field characterized by the expression of Islet-1 lies medial to the first. It primarily gives rise to the right ventricle and the outflow tract. More recently, a third cardiac progenitor cell population that expresses Tbx18 has been proposed in the proepicardial tissues of the embryonic heart (6).
Bone morphogenetic proteins (BMPs) are multifunctional cytokines belonging to the transforming growth factor-β superfamily. BMP signals play a central role in vertebral mesodermal induction (14, 39, 45), as well as in heart development (3, 27, 29). BMP-4 knockout mouse embryos die between 6.5 and 9.5 days, most of which show little or no mesodermal differentiation (39). In the chick embryos, the endoderm underlying the anterior lateral mesoderm, adjacent to the cardiac crescent, expresses BMP-2, whereas BMP-4 is expressed in the overlying ectoderm (29). Based on this positional relationship, BMP-4 is crucial for the specification of the first heart field, and it is also involved in cardiac lineage formation in the second field (25). BMP-4 induces the expression of Nkx2.5 and GATA-4 in cardiac progenitor cells, which are indispensable for differentiation into the cardiac lineage (3, 27).
To date, several growth factors, including transforming growth factor-β2 (30), Wnt11 (36), basic FGF (bFGF), and BMP-2 (16, 27), have been used to induce cardiomyocytes from mouse ES cells. Likewise, the treatment with BMP-4 enhances cardiac induction from mouse embryonic carcinoma (21) and simian ES (13) cells. Despite the identification of these and other factors and reagents (15, 24, 31, 32, 35) that facilitate cardiac induction from animal ES cells, few studies have identified their inductive roles in the human ES cell system (23, 42). Because of species differences in the mechanism of cardiac differentiation of ES cells (1), the direct application of the induction systems established in animal models to human ES cells will be difficult. To stably obtain the donor cells for the cell-based regeneration of the heart, cardiomyocyte induction methods specific for human ES cells are required.
The standard method to induce differentiated cell lineages from ES cells is to make three-dimensional cell aggregates called embryoid bodies (EBs) in suspension culture, which imitate the early period of embryogenesis (2, 11, 13, 34, 41). We have experienced that the undifferentiated human ES cells can be maintained on feeder cell layer in knockout serum replacement (KSR) medium, whereas the formation of EBs capable of generating cardiomyocytes requires FBS (7). When FBS was replaced by KSR in cultures of mouse (34), monkey (13), and human (2) ES cells, a significant decrease in the cardiac differentiation occurred. KSR appears insufficient for the development of endodermal and mesodermal germ layers in EBs (2). Although some studies assessed the effect of FBS on the cardiac induction of animal and human ES cells upon EB formation, the results have been rather controversial (2, 4, 34).
Based on the developmental role of BMP-4 in cardiomyocyte specification, in this study we attempted to determine its ability to promote cardiomyocyte differentiation from human ES cells. In combination with the effect of FBS on the formation of EBs with cardiogenic potential, the optimal concentration and timing of the administration of BMP-4 were assessed in the EB development system.
MATERIALS AND METHODS
Human ES cells and culture.
This study was approved by the Shinshu University Institutional Review Board. We used the H1 human ES cell line (37) purchased from WiCell Research Institute (Madison, WI) in accordance with The Guidelines for Derivation and Utilization of Human Embryonic Stem Cells by The Ministry of Education, Culture, Sports, Science, and Technology of Japan. The cells were cultured on a feeder cell layer of mouse embryonic fibroblasts (Invitrogen, Carlsbad, CA) inactivated with mitomycin C (Sigma, St. Louis, MO). The culture medium consisted of 80% knockout DMEM (Gibco, Grand Island, NY) supplemented with 20% KSR (Gibco), 100 μM nonessential amino acids (Gibco), 2 mM l-glutamine (Gibco), 100 μM 2-mercaptoethanol (Sigma), and 4 ng/ml bFGF (Invitrogen).
ES cells were disbursed into small clumps by incubation in a cell dissociation solution (ReproCELL, Tokyo, Japan) for 5 min. The cells were then transferred to suspension culture in a 24-well nontreated dish with a differentiation medium containing 80% knockout DMEM, 100 μM nonessential amino acids, 2 mM l-glutamine, 100 μM 2-mercaptoethanol, and specified concentrations of FBS (Hyclone, Logan, UT). The suspension cultures were maintained for 4 days unless otherwise stated. EBs were then replated onto gelatin-coated 48-well microplates with one EB per well and expanded in the differentiation medium for 60 days. FBS was present throughout the suspension, and expansion culture was at concentrations of 0, 5, 10, or 20% (Fig. 1A). Alternatively, the FBS concentration after replating was fixed at 20% irrespective of the concentration during suspension. The number and percentage of spontaneously beating EBs were determined at specified intervals. At least 24 EBs were evaluated in each experimental group, and the experiments were repeated three times for statistical analysis.
BMP-4 (R&D Systems, Minneapolis, MN) was added at 0.25, 2.5, 12.5, 25, or 125 ng/ml to the differentiation medium with 20% FBS during the suspension culture for 4 days (Fig. 1A). Alternatively, 25 ng/ml BMP-4 were added only during the expansion culture of EBs up to day 20. To determine the specificity of the response, the BMP antagonist Noggin (150 ng/ml, R&D Systems) was added to the culture medium. BMP-4 treatment was performed with four commercial sources of FBS from Hyclone, Gibco, DS Pharma Biomedical (Osaka, Japan), and MBL (Nagoya, Japan). In each batch, three cultures (24 EBs each) with or without 25 ng/ml BMP-4 were set up to gain the incidence of beating.
The response to isoproterenol was evaluated in the cardiomyocyte on day 5 of expansion. The beating rates of cardiomyocyte colonies induced by 25 ng/ml BMP-4 were estimated before and after the administration of 1 μg/ml isoproterenol (Invitrogen) into the culture media. The data were analyzed in three groups of colonies based on their beating rates before administration: low (40–59 beats/min)-, middle (60–79 beats/min)-, and high (80–99 beats/min)-rate groups.
Reorganization of isolated cardiomyocytes.
The beating colonies of cardiomyocytes induced by 25 ng/ml BMP-4 were picked up on days 7 and 14 of expansion and isolated into single cells by a treatment with 0.5 mg/ml collagenase type 2 (Sigma) in phosphate-buffered saline for 20 min. They were then replated onto glass coverslips and cultured for 14 days with the differentiation medium. After fixation, the specimens were subjected to immunocytochemical analysis with the same procedures as described in Immunocytochemistry.
Reverse transcription-polymerase chain reaction.
Total RNA from the EBs was extracted using the RNeasy kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. cDNA was synthesized from 1 μg of total RNA by SuperScript II reverse transcriptase (Invitrogen). cDNA samples were subjected to PCR amplification with a thermal cycler (MyCycler, Bio-Rad) using primer sets selective for human Oct-4, Rex-1, myosin light chain (MLC)-2a, MLC-2v, cardiac troponin I (cTnI), GATA-4, and GAPDH. The primer sets, the predicted size of the PCR product, and the annealing temperature were as shown in Table 1. The amplification conditions consisted of denaturation at 95°C for 3 min, followed by 25 cycles of denaturation at 95°C for 30 s, annealing at the temperatures specified for each of the primer sets for 30 s, and extension at 72°C for 30 s with a final 7-min extension at 72°C. The PCR products were size fractionated by 2% agarose gel electrophoresis and visualized by ethidium bromide staining.
Quantitative real-time RT-PCR.
Quantitative PCR analysis was performed for Brachyury, β-actin, MESP1, Nkx2.5, and Nestin using the Thermal Cycler Dice Real-Time System (Takara Bio, Otsu, Japan). The PCR amplification reaction mix consisted of 12.5 μl of SYBR Green PCR Master Mix (Takara Bio), 0.5 μl of 0.2 μM forward and reverse primers, 1 μl of 100 ng/μl of template cDNA, and 10.5 μl of distilled water in a total volume of 25 μl. Cycling was performed for 10 min at 95°C, followed by 40 cycles of 5 s at 95°C and 30 s at 60°C, which was the default conditions of the Thermal Cycler Dice Real-Time System software TP 800 (version 2.0). The primer sets for MESP1 and Nestin were as shown in Table 1. For Brachyury and β-actin, we used primers provided by Takara Bio. Following the manufacturer, the comparative threshold cycle method was used to analyze the data, with gene expression levels calibrated to that of the housekeeping gene β-actin. PCR was performed in triplicate for each sample, and three independent experiments were carried out.
Western blot analysis.
Whole cell lysates extracted from EBs were subjected to 10% SDS-polyacrylamide gel electrophoresis, and the separated proteins were electrophoretically transferred to polyvinylidene difluoride membranes. Blots were incubated overnight with primary antibodies against Smad1/5/8 (rabbit polyclonal, 1:200, Santa Cruz Biotechnology, Santa Cruz, CA), phospho-Smad1/5/8 (rabbit polyclonal, 1:1,000, Millipore, Temecula, CA), cTnI (clone C5, mouse IgG2b, 1:1,000, Chemicon, Billerica, MA), atrial natriuretic peptide (ANP; rabbit polyclonal, 1:1,000, Chemicon), and β-actin (clone C4, mouse IgG1, 1:1,000, Santa Cruz Biotechnology), followed by a reaction with horseradish peroxidase-coupled secondary antibody (Bio-Rad, Hercules, CA) for 1 h. Immunocomplexes were visualized with Pierce ECL reagents (GE Healthcare UK, Little Chalfont, UK).
For immunocytochemistry, EBs and EB outgrowths were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 1 h at room temperature and then rinsed three times with 20 mM phosphate-buffered saline (pH 7.4). For antigen retrieval before immunostaining, the specimens were boiled in 10 mM sodium citrate buffer (pH 6.0) for 5 min in a microwave oven and cooled down for 15 min at room temperature. To assess BMP signaling, EBs were stained with antibodies for phospho-Smad1/5/8 (rabbit polyclonal, Millipore) and Smad4 (goat polyclonal, R&D systems). The analysis of cell proliferation (mitosis) used anti-phospho-histone H3 (Ser10) antibody (rabbit polyclonal, Millipore). To characterize induced cardiomyocytes, the EB outgrowths were doubly stained with antibodies for cTnI (clone C5, mouse IgG2b, Chemicon) and for cadherin (rabbit polyclonal, Santa Cruz Biotechnology), connexin 43 (rabbit polyclonal, Sigma), or connexin 45 (rabbit polyclonal, Chemicon). Some EB outgrowths were stained with antibodies for Nestin (clone 196908, mouse IgG, R&D Systems), neuronal class III β-tubulin (clone TUJ1, mouse IgG2a, Covance, Berkeley, CA), and Map2 (rabbit polyclonal, Chemicon) to detect neural lineages. Following the primary antibody step, these samples were incubated with a mixture of appropriate secondary antibodies conjugated with Alexa Fluor 488/568 and 4′,6-diamidino-2-phenylindole dihydrochloride (Molecular Probes, Eugene, OR). The specimens were observed with a Leica TCS SP2 AOBS spectral confocal laser scanning microscope equipped with Ar, He/Ne, and blue diode lasers.
Apoptosis in EB outgrowths was detected by terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) method. After the antigen retrieval procedure as mentioned in Immunocytochemistry, EB outgrowths were processed for TUNEL using ApopTag fluorescein in situ apoptosis detection kit (Chemicon) according to the manufacturer's instructions. DNA fragments labeled with digoxigenin nucleotide were detected by sheep anti-digoxigenin antibody conjugated with FITC, and the cardiomyocytes were then detected by cTnI staining.
After being rinsed with phosphate-buffered saline at 37°C, the EB outgrowths with beating areas were fixed in 2.5% glutaraldehyde in 45 mM cacodylate-HCl (pH 7.2) overnight. After being rinsed three times in 180 mM sucrose in 80 mM cacodylate-HCl (pH 7.2) at 4°C for 3 h, the specimens were postfixed in 1% osmium tetroxide in 0.1 M sodium cacodylate buffer (pH 7.2) for 90 min at 4°C, dehydrated in a graded series of ethanol, and embedded in epoxy resin. Ultrathin sections of the specimens were stained with uranyl acetate and lead citrate and observed with a JEOL JEM-1200 transmission electron microscope at an accelerating voltage of 80 kV.
All experiments were performed at least three times, and the data are expressed as means ± SD using the statistical software StatView Tukey-Kramer method. P values < 0.05 were considered statistically significant.
Suspension culture for 4 days is optimal for mesodermal induction from human ES cells.
The efficient induction of mesoderm during EB formation is prerequisite for cardiomyocyte differentiation. Therefore, we determined the optimal period of EB formation for mesodermal differentiation in cultures with 20% FBS. The expression of Brachyury, a transcription factor transiently expressed in early mesoderm, increased continuously until day 4 of EB development, but it then suddenly fell on day 5 (Fig. 1B). Thus 4 days was the optimal period of suspension culture for mesodermal induction, and this period was used for later studies of cardiac induction.
High concentration of FBS supports EB formation with cardiogenic potential.
We next assessed the effect of FBS (Hyclone) on cardiac differentiation. The percentage of EBs with spontaneously beating cells was compared over a range of FBS concentrations varying from 0–20% during the suspension culture through the expansion (Fig. 1A, a). Alternatively, the concentration in expansion culture was fixed at 20% irrespective of the concentration during suspension (Fig. 1A, b). As a result, the highest incidence of 23.6% was obtained with 20% FBS present throughout the suspension and expansion cultures (Fig. 1, C and D). At lower concentrations of FBS in suspension culture, the incidence of beating EBs was reduced. In cultures without FBS, EB formation was remarkably curtailed and no beating cardiomyocytes were detected. When replated EBs were expanded in 20% FBS after being maintained in lower concentrations during suspension (Fig. 1A, a and b), there was no significant improvement in the incidence of beating EBs (Fig. 1D). These findings showed that high FBS concentration during suspension culture is critical to EB formation capable of generating cardiomyocytes. Because FBS in subsequent expansion culture also benefits the proliferation and differentiation of cardiomyocytes (2, 20, 34), we carried out the following experiments with 20% FBS during and after EB formation.
Cardiomyocyte differentiation involves BMP signaling in EB development.
We then tested whether BMP signaling contributed to the appearance of cardiomyocytes in EBs. BMPs are considered to be generated in EBs and/or supplied by FBS. When the BMP antagonist Noggin was added into the suspension culture with 20% FBS, while the mesodermal marker MESP1 expression was increased, the expression of the primary cardiac marker Nkx2.5 was significantly inhibited on day 4 of suspension (Figs. 1A,d and 2A). The block of BMP signaling by Noggin was also confirmed by the decrease in Smad1/5/8 phosphorylation in EBs (Fig. 3B), which mediates BMP signaling to induce the expression of Nkx2.5 and GATA-4 necessary for cardiomyocyte development (21). Noggin significantly reduced the percentage of beating EBs to less than the control with 20% FBS on day 20 of expansion (23.6% vs. 5.3%) (Fig. 2B). These results indicated that BMP signaling in this EB development system contributed to the spontaneous cardiac differentiation.
BMP-4 promotes cardiac differentiation of human ES cells.
Here we assessed the effect of the BMP-4 supplement on cardiomyocyte induction in EB development. BMP-4 at varied concentrations of 0.25–125 ng/ml was added into the incubation medium for 4 days during the suspension culture (Fig. 1A, c). The expression of MESP1 and Nkx2.5 was concentration dependent at the lower concentrations of BMP-4, but Nkx2.5 expression was significantly inhibited at 125 ng/ml (Fig. 3A). These results showed that an appropriate concentration of BMP-4 is required for cardiac differentiation of human ES cells. With BMP-4 treatment, the phosphorylation of Smad1/5/8 was increased in a concentration-dependent manner in Western blots of EB extracts (Fig. 3B), and the nuclear localization of phospho-Smad1/5/8 and Smad4 was increased in EBs (Fig. 3C), which showed the activation of BMP signaling.
To confirm the effect of BMP-4 on cardiac induction at different concentrations, the incidence of beating EBs was determined up to day 20 of EB expansion (Fig. 4). In lower concentrations of BMP-4 at 0.25 and 2.5 ng/ml, EBs with spontaneous beating appeared on day 3 after replating, and the incidence then gradually increased until day 20 (Fig. 4A). Noticeably, the percentage at 12.5 and 25 ng/ml BMP-4 rapidly increased between days 2 and 4 and reached 93.1% and 95.8% on day 6, respectively, and then plateaued until day 20 (Fig. 4A). The final incidence at 25 ng/ml, 97.7%, was about four times higher than that of the control. In contrast, the incidence at 125 ng/ml BMP-4 was lower than that of the control throughout the middle and later periods of culture. The addition of Noggin to the culture medium with 25 ng/ml BMP-4 significantly reduced the incidence of spontaneous beating (Fig. 4, B and C). Thus the optimum concentration of BMP-4 for cardiomyocyte differentiation from human ES cells was 12.5–25 ng/ml.
To determine the optimum timing for BMP exposure and the resulting differentiation, we omitted BMP-4 during the 4-day induction phase but then added 25 ng/ml into the expansion culture (Fig. 1A, e). The incidence of beating EBs was not significantly different from that of the controls (Fig. 4D), indicating that BMP-4 promotes cardiac induction only in the early stages of EB development. Thus the ability of BMP-4 to promote cardiac induction from human ES cells required both proper concentration and timing.
Interbatch differences of FBS on cardiac differentiation with or without BMP-4 were tested, comparing with three other commercial sources of FBS (Fig. 4E). The incidence of beating EBs with 20% FBS alone showed some differences between batches (Fig. 4E, a). When 25 ng/ml BMP-4 was added to FBS, the rates of incidence were comparable with each other, ranging from 94.4% to 97.2% on day 10 of expansion with no significant differences (Fig. 4E, b). These findings indicated that interbatch differences of FBS on cardiomyogenic activity can be exceeded to some extent by the addition of BMP-4 during EB development.
Validation of BMP-4-induced cardiomyocytes.
The cardiomyocytes induced by BMP-4 formed as colonies predominantly in the central part of EB outgrowths and exhibited spontaneous synchronized beating, suggesting electrical integration. In most colonies on day 5 of expansion, the administration of 1 μg/ml isoproterenol, a β-adrenergic agonist, increased the rate of beating (Fig. 5A, a), as it does for normal cardiomyocytes. The chronotropic response to isoproterenol was significant in the low (40–59 beats/min)- and middle (60–79 beats/min)-rate groups but not in the high-rate group (80–99 beats/min) (Fig. 5A, b). During a long-term culture of more than 20 days, a portion of the cardiomyocytes stopped spontaneous beating. When 1 μg/ml isoproterenol was added into culture medium, the beating resumed (not shown).
By RT-PCR analysis, the expression of cardiomyocyte-related genes GATA-4, MLC-2a, MLC-2v, and cTnI became detectable by day 7 of expansion, whereas the undifferentiation markers Oct-4 and Rex-1 completely disappeared (Fig. 5B). Western blot analysis of EBs on days 7 and 14 of expansion showed that the largest amount of myocardial proteins cTnI and ANP was detected in culture with 25 ng/ml BMP-4 (Fig. 5C). The amount of these proteins increased between days 7 and 14, during the same period in which the incidence of beating EBs reached a plateau (Fig. 4A).
By immunostaining for cTnI, cardiomyocyte colonies were frequently detected in the EB outgrowths in culture with 25 ng/ml BMP-4, whereas they were small in number in 20% FBS, and very rare, if any, in 125 ng/ml BMP-4. The differentiation of cardiomyocyte colonies induced by 25 ng/ml BMP-4 was assessed from day 7 to day 30 of expansion (Fig. 6A), the comparable periods of differentiation to the other analyses (Figs. 4–7). Cadherin-positive intercellular adhesions were obvious from the early stages. Connexin 43 was detectable by day 7 and increased subsequently, whereas connexin 45 in immature cardiomyocytes on day 7 disappeared with culture. These changes represent the course of cardiomyocyte differentiation from conduction system-like to working ones (12). cTnI-positive myofibrils developed during the culture, supporting the results of RT-PCR and Western blot analysis. To confirm the differentiation of cardiomyocytes in longer-term culture, an ultrastructural analysis was performed on day 60. The maturation of myofibrils with sarcomeres, fasciae adherens, and desmosomes was evident (Fig. 6B). With BMP-4 treatment, no remarkable change in the extent of apoptosis and proliferation was observed in the cardiomyocytes on day 7 of expansion (Fig. 6C). Taken together, these findings validated the proper characteristics of cardiomyocytes induced by BMP-4.
When the cardiomyocytes in beating colonies induced by 25 ng/ml BMP-4 were isolated into single cells at different stages of differentiation and then allowed to reaggregate in additional culture for 14 days, connexin 43 was obviously detected at the intercellular junctions of cardiomyocytes sampled from day 7 colonies (Fig. 6D, a). The reactivity was comparable with that of the cardiomyocytes in the same culture period of 21 days (Fig. 6A). In contrast, the cardiomyocytes from day 14 colonies showed an impaired reactivity for connexin 43 (Fig. 6D, b). This indicates a higher ability of early stage cardiomyocytes to gain intercellular associations.
Noggin promotes neural induction from human ES cells.
The BMP antagonism by Noggin was shown to decrease the cardiac differentiation in EB development with or without BMP-4. Noggin is generally known as a neural inducer (28). We then confirmed what occurred in EBs in the presence of 150 ng/ml Noggin. Real-time PCR of the neural progenitor marker Nestin showed that the expression was significantly increased by Noggin treatment compared with the control (Fig. 7A). This Noggin-induced expression of Nestin was clearly inhibited by the addition of 25 ng/ml BMP-4. Treatment alone with BMP-4 had no apparent effect at any concentration tested. By immunocytochemistry, Nestin-positive neural progenitor-like cells were frequently observed predominantly in peripheral parts of the EB outgrowths in cultures with Noggin, but not with BMP-4 on day 5 of expansion (Fig. 7B, a–d). On day 14 with Noggin, while cTnI-positive cardiomyocytes were absent over all the outgrowths (Fig. 7B, e), abundant neurons immunoreactive for class III β-tubulin and Map2 had been formed in the peripheral part (Fig. 7B, f), exhibiting the similar distribution pattern as the Nestin-positive cells. This distribution pattern contrasted with that of the centrally located beating cardiomyocytes in BMP-4-treated EB outgrowths (Fig. 6A). These findings showed that Noggin supports the neural differentiation in EBs in the condition employed here.
The study presented here showed that FBS supports the development of human EBs capable of generating cardiomyocytes. Indeed, the presence of FBS alone up to 20% during the suspension and expansion phases of culture was sufficient to enhance cardiomyocyte induction in human ES cells, and the removal of FBS from suspension culture resulted in a disturbed EB formation with no beating cardiomyocytes. This indicates that FBS is required for the specification of myocardial lineage during the period of EB formation. The effect of FBS on the cardiac induction in EB development seems rather controversial among reports, though this may be partly due to the variation among its lot. An inhibitory effect of FBS on cardiac induction has been reported in mouse ES cells compared with a serum-free medium with BMP-4 (4), whereas others supported its promotive effect on human (2) and BMP-4-treated mouse (34) ES cells, consistent with our study. The differentiation of mesoderm with other germ layers in human EB formation requires FBS and does not occur with serum-free culture medium (2). The present study showed that BMP(s), presumably originating from the germ layers and/or FBS, contributes to the cardiac induction to some extent, whereas in our preliminary study, the use of KSR failed to induce cardiomyocytes in human EBs even in the presence of BMP-4, bFGF, and/or activin A (unpublished data). These findings indicate that some other factors present in the serum, which promote and/or inhibit certain signaling pathways, are also necessary for the induction of cardiomyocytes via germ layer differentiation. Recently, chemically defined media have been applied for the cardiac induction from human ES cells (41, 42), because concerning the clinical application of those cardiomyocytes, difficulty exists in the use of animal sources such as FBS (19). For example, Prostaglandin I2 in combination with the p38 MAPK inhibitor SB-203580 in insulin-free and serum-free medium showed a strong positive effect on the cardiac induction in EB formation (41). Clarifying the effective factors in FBS directly and/or indirectly involved in germ layer formation and cardiomyogenesis is a potential approach to develop serum-free cardiac induction system.
We showed here that BMP-4 treatment promotes cardiomyocyte induction in serum-based human EB development. BMP-4 was required for the initial stage of cardiac induction during suspension culture. No discernible effect of BMP-4 was observed after EB replating. Thus the effect of BMP-4 on cardiac induction has a time window restricted to the early stage of EB formation when the specification into cardiac lineage occurs. In addition, we found that the maximum effect on cardiac induction was achieved at the concentrations of 12.5–25 ng/ml BMP-4. The previous studies showed that when the effect of BMP-4 at concentrations of 10, 50, and 100 ng/ml was tested, 10 ng/ml was best for the cardiac induction in monkey ES cells (13) and the treatment of human ES cells with 10 ng/ml BMP-4 for 4 days in combination with a prior treatment with activin A directly inducing cardiomyocytes (18). These reports are consistent with our findings that BMP-4 at around 12.5–25 ng/ml showed the strongest effect on cardiac induction. Lower concentrations reduced the effect, and the highest concentration (125 ng/ml) strongly inhibited cardiac induction as judged by the low incidence of beating EBs. This concentration also suppressed the expression of Nkx2.5, though the activation of Smad1/5/8 and the expression of MESP1 were high. An addition of inappropriate concentrations of BMP-4 and/or at the wrong time likely induced other cell lineages that are linked with other factors (9, 10, 13, 33, 38). Therefore, the use of BMP-4 with the optimal concentration and at the appropriate time is crucial to the efficient induction of cardiomyocytes from human ES cells.
The transient expression of Noggin has been reported in the cardiac crescent in the mouse embryo (44), suggesting its role in the early stage of murine cardiomyocyte development. The study also showed that the use of Noggin in combination with BMP-4 effectively promotes cardiac induction from mouse ES cells. In contrast, there is little information on the expression of Noggin in human heart development. In our study presented here, the addition of Noggin into culture medium decreased the incidence of beating EBs, whereas it increased the number of Nestin-positive cells. This suggests that with the concentration and administration period employed here, Noggin promotes the differentiation of human ES cells into neuroprogenitor cells rather than cardiomyocytes. The possibility cannot be excluded, however, that the concentration of Noggin used here was inappropriate for human ES cells. Optimal concentrations may facilitate cardiomyocyte differentiation. Evidence has accumulated that species differences exist in the mechanism by which the undifferentiated state of ES cells is maintained. BMP-4, through links with leukemia inhibitory factor, is important in the maintenance of the undifferentiated state in mouse ES cells (43). In contrast, leukemia inhibitory factor/STAT3 signaling does not act to maintain the undifferentiated state in human ES cells (8). BMP-4 promotes their differentiation, and the suppression of BMP signaling by Noggin has been shown to maintain the undifferentiated state via the link with bFGF (40). In our study, a transfer to the suspension culture and/or a withdrawal of bFGF appears to change the role of BMP antagonism by Noggin from the maintenance of the undifferentiated state to the promotion of neural differentiation.
The cardiomyocytes induced by BMP-4 treatment in our study had proper cell biological characteristics. The amount of cTnI and ANP proteins increased sharply between days 7 and 14. Previous studies, including ours, showed that the proliferative activity of induced cardiomyocytes remains high at this stage of in vitro differentiation (7, 20). The increase in the cardiac protein content confirms both the proliferation and maturation of cardiomyocytes in this period. At early stages of differentiation, they developed intercellular junctions to form functional colonies. A localization of the fascia adherens protein cadherin and the gap junction protein connexin 43 at the cellular junctions means that there was a mechanical and electrical connection of the cardiomyocytes, consistent with our previous report on the differentiation of ES cell-derived cardiomyocyte associations (7). Although once isolated, it was difficult for the relatively differentiated cardiomyocytes to reassociate, judging from the impaired immunoreactivity for connexin 43 in the aggregates. A previous study also showed that the ability of cardiomyocytes to integrate into host myocardial tissues after injection largely depends on their differentiation stages, in which the neonatal cardiomyocytes, but not adult, survived and formed graft myocardium in the host heart (26). Based on these facts, the use of ES cell-derived cardiomyocytes in the early stages of induction is ideal for the cell-based therapy of cardiac diseases.
In conclusion, we established an effective cardiac induction method of human ES cells by using 25 ng/ml BMP-4 and 20% FBS in suspension culture for 4 days. The addition of 150 ng/ml Noggin attenuated the inductive effect of BMP-4, and alone Noggin promoted neural differentiation. Progress in understanding the mechanism of human cardiomyocyte differentiation will help to further increase the efficiency of cardiac induction from human ES cells.
This work was supported by a Japanese Society for the Promotion of Science Fellowship 18-8454.
We gratefully thank Tadayuki Yokoyama and Sakiko Shirasawa (Bourbon Corporation Japan) for help in real-time PCR analysis. We also thank Kayo Suzuki and Dr. Kiyokazu Kametani (Research Center for Instrumental Analysis, Shinshu University) for assistance on microscopic analysis.
- Copyright © 2009 the American Physiological Society