Biochemical Journal

Research article

AW551984: a novel regulator of cardiomyogenesis in pluripotent embryonic cells

Satoshi Yasuda , Tetsuya Hasegawa , Tetsuji Hosono , Mitsutoshi Satoh , Kei Watanabe , Kageyoshi Ono , Shunichi Shimizu , Takao Hayakawa , Teruhide Yamaguchi , Kazuhiro Suzuki , Yoji Sato

Abstract

An understanding of the mechanism that regulates the cardiac differentiation of pluripotent stem cells is necessary for the effective generation and expansion of cardiomyocytes as cell therapy products. In the present study, we have identified genes that modulate the cardiac differentiation of pluripotent embryonic cells. We isolated P19CL6 cell sublines that possess distinct properties in cardiomyogenesis and extracted 24 CMR (cardiomyogenesis-related candidate) genes correlated with cardiomyogenesis using a transcriptome analysis. Knockdown of the CMR genes by RNAi (RNA interference) revealed that 18 genes influence spontaneous contraction or transcript levels of cardiac marker genes in EC (embryonal carcinoma) cells. We also performed knockdown of the CMR genes in mouse ES (embryonic stem) cells and induced in vitro cardiac differentiation. Three CMR genes, AW551984, 2810405K02Rik (RIKEN cDNA 2810405K02 gene) and Cd302 (CD302 antigen), modulated the cardiac differentiation of both EC cells and ES cells. Depletion of AW551984 attenuated the expression of the early cardiac transcription factor Nkx2.5 (NK2 transcription factor related locus 5) without affecting transcript levels of pluripotency and early mesoderm marker genes during ES cell differentiation. Activation of Wnt/β-catenin signalling enhanced the expression of both AW551984 and Nkx2.5 in ES cells during embryoid body formation. Our findings indicate that AW551984 is a novel regulator of cardiomyogenesis from pluripotent embryonic cells, which links Wnt/β-catenin signalling to Nkx2.5 expression.

  • cardiac differentiation
  • cardiomyogenesis
  • embryonal carcinoma cell
  • embryonic stem cell
  • Wnt signalling

INTRODUCTION

The heart is the first organ to form in the vertebrate embryo. Cardiac progenitor cells are derived from the mesoderm, which emerges from the primitive streak during gastrulation. Cardiac progenitor cells migrate into the anterolateral regions of the embryo to form the cardiac crescent and subsequently contribute to the myocardium and endocardium of the heart [1,2]. The in vitro differentiation of ES (embryonic stem) cells assembling into aggregates, which are called EBs (embryoid bodies), mimics early embryonic development and is commonly conducted for cardiomyogenesis [3]. Wnt/β-catenin signalling is crucial for the differentiation of ES cells into cardiomyocytes as well as for heart development [4]. Recent studies have reported that Wnt/β-catenin signalling plays biphasic roles in cardiomyogenesis. Activation of Wnt/β-catenin signalling in the early phase of EB cultures promotes cardiac differentiation, whereas late activation of Wnt/β-catenin signalling inhibits cardiac differentiation [5,6]. Extracellularly secreted Dkk-1 (Dickkopf-1) interacts with a Wnt co-receptor LRP6 (low-density lipoprotein-receptor-related protein 6) [7] and modulates cardiac differentiation owing to its antagonistic inhibition of Wnt/β-catenin signalling [46].

Heart diseases such as myocardial infarction damage cardiomyocytes and consequently lead to a significant loss of the contractile capacity of the heart [8]. To repair functions of the injured heart, a great deal of research has attempted to develop regenerative medicine using ES cell- or iPS (induced pluripotent stem)-cell-based cardiomyocytes as cell therapy products [9]. ES cells and iPS cells are pluripotent and able to differentiate spontaneously into a variety of cell types, including cardiomyocytes [3,10,11]. However, the efficiency of the current methods available for the cardiac differentiation of pluripotent stem cells is insufficient for clinical settings [3,12]. A comprehensive understanding of the mechanism involved in the cardiac differentiation of pluripotent stem cells is necessary to improve the differentiation efficiency.

In the present study, we provide a list of genes whose expression is significantly correlated with efficiency in mouse EC (embryonal carcinoma) cell cardiomyogenesis induced by DMSO. RNAi (RNA interference) experiments for these genes revealed that AW551984, 2810405K02Rik (RIKEN cDNA 2810405K02 gene) and Cd302 (CD302 antigen) effectively modulate the cardiac differentiation of both EC cells and ES cells. AW551984 transcriptionally increased in response to Wnt/β-catenin signalling stimulation during ES cell differentiation and regulated the expression of Nkx2.5 (NK2 transcription factor related locus 5), an early cardiac transcriptional factor. Our findings demonstrate that AW551984 is a novel regulator of cardiomyogenesis that links Wnt/β-catenin signalling to Nkx2.5 expression.

MATERIALS AND METHODS

EC cell culture and differentiation

P19 cells and P19CL6 cells were obtained from the A.T.C.C. and RIKEN Cell Bank respectively. Mouse EC cells were maintained in α-minimum essential medium (Invitrogen) supplemented with 10% FBS (fetal bovine serum; Cell Culture Technologies), 2 mM L-glutamine (Sigma–Aldrich), 100 units/ml penicillin and 100 μg/ml streptomycin (Invitrogen) at 37 °C in an atmosphere containing 5% CO2. To prepare P19CL6 cell sublines, P19CL6 cells were co-transfected with a DNA plasmid that encoded the EGFP (enhanced green fluorescent protein) gene (Clontech) driven by mouse Myh6 [α-MHC (myosin heavy chain)] promoter (a gift from Professor Jeffrey Robbins, Children's Hospital Medical Center, Cincinnati, OH, U.S.A.) and pcDNA3.1 (Invitrogen) using Lipofectamine™ 2000 (Invitrogen), followed by selection with 1 mg/ml G418 (Sigma–Aldrich). A total of 72 G418-resistant clones were isolated, out of which four clones (CL6G26, CL6G36, CL6G45 and CL6G52) with a wide range of cardiac differentiation efficiencies were selected for further analysis. For differentiation, cells were trypsinized and suspended in EC cell growth medium containing 1% DMSO [13]. The cells were seeded at a density of 4 × 104 cells/well on six-well plates (BD Biosciences) or Lab-Tek Chamber Slide two-well Permanox Slides (Nunc). To evaluate the differentiation efficiency, the number of nodules/cm2 that were beating spontaneously was counted under a microscope. In the GeneChip experiments, the differentiation efficiency was assessed non-parametrically by the number of beating nodules/cm2 (grade 1, <0.098/cm2 corresponding to one beating nodule/six-well plate; grade 2, 0.098 to 157/cm2 corresponding to one beating nodule/well on six-well plates; and grade 3, >157/cm2 corresponding to one beating nodule/microscopic image magnified ×200). EC cell differentiation medium was replaced every other day. The day of beginning the differentiation was set as day 0.

qRT-PCR (quantitative real-time PCR)

Total RNA was isolated from cells with an RNeasy Mini kit (Qiagen) or a BioRobot M48 Workstation (Qiagen), according to the manufacturer's instructions. One-step qRT-PCR was performed with a QuantiTect Probe RT-PCR kit (Qiagen) on an ABI Prism 7000 Sequence Detection System or a 7300 real-time PCR System (Applied Biosystems). The expression levels of target genes were normalized to those of 18S rRNA or the Gapdh (glyceraldehyde-3-phosphate dehydrogenase) transcript, which were quantified using TaqMan rRNA control reagents or TaqMan rodent Gapdh control reagents (Applied Biosystems) respectively. Probes and primers were obtained from Applied Biosystems and Sigma–Aldrich. The sequences of probes and primers used in the present study are shown in Supplementary Table S1 (at http://www.BiochemJ.org/bj/437/bj4370345add.htm.

GeneChip and biostatistical analysis

Total RNA isolated from undifferentiated EC cell strains (P19 cells, P19CL6 cells and four P19CL6 cell sublines) was converted into biotinylated cRNA using Two-Cycle Target Labelling and Control Reagents (Affymetrix). Labelled RNA was processed for microarray hybridization to MOE430A and MOE430B GeneChips (Affymetrix), which contain 22626 and 22511 probe sets (13410 and 9249 mouse RefSeq transcripts) respectively. An Affymetrix GeneChip Fluidics Station was used to perform streptavidin/phycoerythrin staining. The hybridization signals on the microarray were scanned using a GeneChip Scanner 3000 (Affymetrix) and analysed using GCOS software (Affymetrix). Normalization was performed by global scaling with the arrays scaled to a trimmed average intensity of 500 after excluding the 2% of probe sets with the highest and the lowest values. The hybridization experiments were performed in five samples of each EC cell strain. The NCBI GEO (National Center for Biotechnology Information Gene Expression Omnibus) accession number for the microarray data is GSE26875. To extract the informationally significant probe sets from the data set, we filtered probe sets using the following three steps. First, probe sets were regarded as ‘absent’ when indicated as ‘present’ by ‘absolute analysis’ using GCOS software in less than three samples from one strain. Probe sets regarded as ‘absent’ in all strains were eliminated from the data set. Secondly, when no significant difference was observed among strains using ANOVA (P≥0.05), probe sets were also eliminated from the data set. Thirdly, if the difference between the maximum and minimum mean values of probe sets in the strains was equal to or more than 2.5-fold, probe sets were used for further analysis.

After data standardization (z-scoring) of the cardiac marker genes [Nkx2.5 (NK2 transcription factor related locus 5), Gata4 (GATA binding protein 4), Mef2c (myocyte enhancer factor 2C), Myh6 (α-MHC), Myh7 (β-MHC), Mlc2a (myosin light chain 2a) and Mlc2v (myosin light chain 2v)] expressed in the differentiated EC cell strains (P19 cells, P19CL6 cells and four P19CL6 cell sublines), PCA (principal component analysis) was performed to project the data into a lower-dimensional space using SYSTAT 10.2 software. Eigenvalues of the first and second factors equal 4.43 and 1.20 and represent 63.3 and 17.1% of the total variability. The factor loading plot indicates the coefficients of all variables on the first and second principal components (Supplementary Figure S1A at http://www.BiochemJ.org/bj/437/bj4370345add.htm). The first and second principal component scores were calculated and averaged at days 8, 12, 16 and 20 for each strain (Supplementary Figures S1B and S1C). Cardiac differentiation evaluated by lag time to the onset of beating and by the number of beating nodules was shown among EC cell strains (Figure S1D). To identify probe sets related to the cardiac differentiation of EC cells, the correlation between the intensity values of the filtered probe sets and three variables (the maximum of the first principal component scores in each strain, lag time to the onset of beating and beating nodule numbers assessed non-parametrically at day 20) was determined by calculating Spearman's rank correlation coefficients [14] and their P values. Probe sets exhibiting statistically significant correlations with all three variables (P<0.05) were selected.

ES cell culture and differentiation

Mouse ES cells (R1 [15]; A.T.C.C.) were maintained with mitomycin C-inactivated MEFs (mouse embryonic fibroblasts; Millipore) on gelatin-coated dishes in DMEM (Dulbecco's modified Eagle's medium; Sigma–Aldrich) supplemented with 20% (v/v) FBS (ESGRO), 1000 units/ml LIF (leukaemia inhibitory factor; ESGRO), 1 mM sodium pyruvate (Invitrogen), 0.1 mM 2-mercaptoethanol (Invitrogen), 0.1 mM non-essential amino acids (Invitrogen), 100 units/ml penicillin and 100 μg/ml streptomycin (Invitrogen) at 37 °C in an atmosphere containing 5% CO2. Before differentiation, R1 cells were dissociated by trypsinization and separated from MEFs by pre-plating on non-coated dishes for 30 min at 37 °C. To initiate differentiation, 750 ES cells in 20 μl of ES cell growth medium without LIF were cultured in a hanging drop for 3 days. The EBs formed were transferred on to non-adherent 96-well round-bottom plates (PrimeSurface; Sumitomo Bakelite) and cultured for an additional 2 days. At day 5 of differentiation, each EB was transferred to a well on a gelatin-coated 48-well plate (Iwaki). EBs exhibiting spontaneous contraction were counted daily to calculate the percentage of beating EBs. The differentiation medium was replaced every other day. The day of starting the hanging-drop formation was set as day 0. A total of 48 EBs were observed in one group for each experiment.

siRNA (small interfering RNA) transfection

CL6G52 cells were plated at 1 × 104 cells/well in 400 μl of EC cell growth medium on 24-well plates. The next day, cells were transfected with 100 nM Stealth RNAi™ siRNA (Invitrogen) using Lipofectamine™ 2000 (Invitrogen), according to the manufacturer's instructions. After 48 h of transfection, cells were used for further analysis. R1 cells seeded at a density of 5 × 105 in 5 ml of ES cell growth medium on a gelatin-coated 60-mm-diameter dish were transfected with 50 nM Stealth RNAi™ siRNA using Lipofectamine™ RNAiMAX (Invitrogen), according to the manufacturer's instructions. After 24 h of transfection, cells were used for further analysis. Stealth RNAi™ siRNA negative control duplex (Invitrogen) was used as a control. The sequences of siRNAs used in the present study are shown in Supplementary Table S2 (at http://www.BiochemJ.org/bj/437/bj4370345add.htm).

Western blot analysis

Undifferentiated R1 cells and EBs were lysed in RIPA buffer (20 mM Tris/HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P40, 0.5% sodium deoxycholate, 0.1% SDS and 1 mM EDTA) containing protease inhibitor mixture (Roche). Cell lysates were centrifuged at 12000 g for 10 min to prepare the supernatant. Proteins were separated using SDS/PAGE, transferred on to PVDF membranes and probed with primary antibodies. The immunoreactive bands were visualized with HRP (horseradish peroxidase)-conjugated anti-(mouse IgG) or anti-(rabbit IgG) antibodies (Jackson ImmunoResearch) and an ECL (enhanced chemiluminescence) Plus Western Blotting Detecting System (GE Healthcare), and were detected using a LAS 4000 mini (Fujifilm).

Antibodies and reagents

To raise the anti-AW551984 antibody, a synthetic peptide LMPNGTPQQRQNSQKKK(C) (amino acids 654–671 of mouse AW551984) was coupled to KLH (keyhole limpet haemocyanin) and used as an antigen. Rabbits were immunized by multiple injection of the cross-linked peptide with adjuvant. The antibody was affinity-purified with antigen peptide coupled to gel (Medical and Biological Laboratories). The anti-β-actin antibody (AC15) was obtained from Sigma–Aldrich. Recombinant mouse Wnt3a and Dkk-1 was purchased from R&D Systems.

RESULTS

Identification of CMR (cardiomyogenesis-related candidate) genes

We attempted to isolate genes related to the DMSO-induced cardiomyogenesis of mouse EC cells. Our approach to identify genes for cardiac differentiation is essentially based on statistical comparison of GeneChip signal intensities of undifferentiated EC cell strains with their cardiac differentiation. The scheme highlighting the gene identification process is shown in Figure 1. To search for genes that correlated with the cardiac differentiation of EC cells, GeneChip data were obtained from undifferentiated EC cell strains and filtered by the three steps described in the Materials and methods section. The correlations of signal intensities of the filtered probe sets with three variables, i.e. (i) first principal component scores of cardiac marker gene expression, (ii) lag time to the onset of beating, and (iii) number of beating nodules at day 20, were determined by calculating Spearman's rank correlation coefficients. Finally, probe sets derived from 24 genes {Prdm5 (PR domain containing 5), AW551984, D430028G21Rik [Mavs (mitochondrial antiviral signalling protein)], 5330410G16Rik [Tmem59l (transmembrane protein 59-like)], Tmem98 (transmembrane protein 98), Ctsc (cathepsin C), F2r [coagulation factor II (thrombin) receptor], Sema3e (sema domain Ig domain short basic domain secreted 3E), Maged2 (melanoma antigen family D2), 2810405K02Rik, Rhox4b (reproductive homeobox 4B), Cd302, Fzd1 (Frizzled homologue 1), Adarb1 (adenosine deaminase RNA-specific B1), Gpaa1 (GPI anchor attachment protein 1), Chst2 (carbohydrate sulfotransferase 2), 9830115L13Rik [Zc3hav1 (zinc finger CCCH type, antiviral 1)], 9630055N22Rik [Mfsd7b (major facilitator superfamily domain containing 7B)], Ptprb (protein tyrosine phosphatase receptor type B), Gstz1 (glutathione transferase ζ1), 1110021L09Rik (RIKEN cDNA 1110021L09 gene), Tsga14 (testis-specific gene A14), Hnrnpa1 (heterogeneous nuclear ribonucleoprotein A1) and Adm (adrenomedullin)} exhibited statistically significant correlations with all of these variables, and we named these 24 genes CMR genes for convenience (Table 1 and Supplementary Table S3 at http://www.BiochemJ.org/bj/437/bj4370345add.htm). Nine of the 24 genes also had significant correlations between the intensity values of probe sets and the second principal component scores of cardiac marker gene expression (Supplementary Table S3).

Figure 1 Biostatistical-based approach for identification of CMR genes

Genome-wide expression analysis on GeneChips of undifferentiated P19 cells, P19CL6 cells and P19CL6 cell sublines (CL6G26, CL6G36, CL6G45 and CL6G52) was performed as described in the Materials and methods section. The differentiation of EC cells into cardiomyocytes was induced by stimulation with DMSO. PCA was conducted using standardized data of cardiac marker gene expression in differentiated cells, and the first principal component scores were calculated. Statistical correlation of the probe-set intensities with the first principal component scores, lag time to the onset of beating and number of nodules was determined using Spearman's rank correlation coefficient. Genes of probe sets exhibiting significant correlation with the three variables (P<0.05) were identified as CMR genes.

View this table:
Table 1 Identified CMR genes

Validation of CMR genes as regulators of cardiac differentiation in EC cells

To examine whether or not the selected 24 CMR genes function as regulators of cardiac differentiation, we first attempted to knockdown CMR gene expression in CL6G52 cells by siRNA transfection. qRT-PCR analysis verified that the siRNAs used in the present study successfully suppressed the expression of 21 CMR genes (excluding CMR11, CMR19 and CMR22) in the EC cells at 48 h after transfection (Supplementary Figure S2 at http://www.BiochemJ.org/bj/437/bj4370345add.htm). The mRNAs for CMR11, CMR19 and CMR22 were not significantly reduced by the siRNAs that we used in the present study. To determine the differentiation states of CL6G52 cells transfected with the validated siRNAs, the number of nodules beating spontaneously was counted in cells, the differentiation of which was initiated by DMSO. The number of beating nodules decreased significantly in cells transfected with siRNAs targeting CMR2–CMR5, CMR9, CMR10, CMR13, CMR18 and CMR24 (Table 2). In contrast, RNAi against CMR7, CMR12, CMR16, CMR17, CMR21 and CMR23 significantly facilitated the beating nodule development. No significant effect of CMR1, CMR6, CMR8, CMR14, CMR15 or CMR20 on beating nodule number was observed.

View this table:
Table 2 Effects of CMR gene knockdown on the number of beating nodules in EC cells

CL6G52 cells were transfected with 100 nM siRNA targeting the CMR genes or the negative control siRNA. After 48 h of transfection, differentiation of CL6G52 cells was initiated with 1% DMSO. Nodules beating spontaneously were counted per well with a microscope at day 14. Results are means±S.E.M. (n=4). Statistical significance was determined using a Student's t test (*P<0.05 compared with control).

To evaluate precisely the degree of cardiac differentiation of CL6G52 cells transfected with siRNA, we measured the mRNA levels of cardiac markers after the addition of DMSO by qRT-PCR. mRNA levels of cardiac markers [Nkx2.5, Gata4, Mef2c, Myh6 (α-MHC), Myh7 (β-MHC), Mlc2a and Mlc2v] in non-transfected CL6G52 cells changed in a time-dependent manner (Supplementary Figure S3 at http://www.BiochemJ.org/bj/437/bj4370345add.htm). Transcript expression of Myh6 (α-MHC) at day 14 was significantly decreased by the knockdown of CMR2–CMR5, CMR10, CMR13 and CMR18, whereas it was increased by knockdown of CMR12 and CMR23 (Figure 2A). Knockdown of CMR2 and CMR13 significantly inhibited Myh7 (β-MHC) expression (Figure 2B). In contrast, knockdown of CMR16, CMR17 and CMR23 markedly elevated Myh7 (β-MHC) expression. The expression of Mlc2a was significantly inhibited by knockdown of CMR2, CMR4, CMR8, CMR10, CMR20 and CMR23, and enhanced by CMR16 knockdown (Figure 2C). Knockdown of CMR2–CMR6, CMR10 and CMR13 significantly decreased MLC2v expression, whereas knockdown of CMR7, CMR12, CMR16 and CMR23 increased its expression (Figure 2D). Collectively, RNAi against 17 CMR genes exhibited significant effects on marker genes in the cardiac differentiation of EC cells.

Figure 2 Effects of CMR gene knockdown on mRNA levels of Myh6 (α-MHC), Myh7 (β-MHC), Mlc2a and Mlc2v in differentiated EC cells

CL6G52 cells were transfected with 100 nM siRNA targeting CMR genes or the negative control siRNA. After 48 h of transfection, differentiation of CL6G52 cells was initiated with 1% DMSO. At day 14, total RNA was isolated from differentiated CL6G52 cells and subjected to one-step qRT-PCR. mRNA levels were normalized to 18S ribosomal RNA levels. Expression levels of Myh6 (α-MHC) (A), Myh7 (β-MHC) (B), Mlc2a (C) and Mlc2v (D) in cells transfected with the negative control were set to 1. Results are means±S.E.M. (n=4). Statistical significance was determined using a Student's t test (*P<0.05 compared with control).

Validation of CMR genes as regulators of cardiac differentiation in ES cells

To initiate ES cell differentiation into cardiomyocytes, EBs were formed by hanging-drop culture and cultured further on non-adherent round-bottom plates. The EBs transferred to adherence culture began to exhibit spontaneous beating at day 8 and an elevated rate of beating by day 11 (Figure 3A). To examine whether or not the expression of CMR genes changed in a time-dependent manner, transcript levels of CMR genes were measured at days 0, 3, 5 and 11 by qRT-PCR. A statistically significant correlation (P<0.05) between gene expression and days was observed in CMR2, CMR4–CMR9, CMR12–CMR18 and CMR20 (Figure 3B). To verify the roles of CMR genes in the cardiac differentiation of ES cells, R1 cells were transfected with siRNAs targeting the CMR genes, the knockdown efficiency of which was confirmed (see Supplementary Figure S4 at http://www.BiochemJ.org/bj/437/bj4370345add.htm), and were subjected to EB formation 24 h after transfection. We analysed the expression of cardiac marker genes at day 8 of differentiation. RNAi against CMR2, CMR8, CMR10, CMR14, CMR15, CMR20 and CMR23 significantly inhibited the expression of both Myh6 (α-MHC) and Mlc2a, almost consistent with the case of Myh7 (β-MHC) and Mlc2v (Figure 4). In contrast, knockdown of CMR7, CMR9, CMR12, CMR21 and CMR24 significantly enhanced the expression of both Myh7 (β-MHC) and Mlc2a. Moreover, knockdown of CMR9 and CMR12 also increased levels of Mlc2v and Myh6 (α-MHC) respectively. The effects of CMR genes on the cardiac differentiation of EC cells and ES cells are summarized in Supplementary Table S4 (at http://www.BiochemJ.org/bj/437/bj4370345add.htm).

Figure 3 Time courses of beating development and CMR gene expression in ES cells during differentiation

Differentiation of R1 cells was initiated by forming EBs without LIF. (A) EBs exhibiting spontaneous beating were counted and the percentages were calculated. (B) Total RNA of R1 cells at days 0, 3, 5 and 11 was isolated and subjected to one-step qRT-PCR. mRNA levels of the CMR genes were normalized to those of Gapdh. Expression levels of cells at day 0 were set to 1. Results are means±S.E.M. (n=5). Statistical significance between gene expression and days was determined using a Spearman's rank correlation coefficient test (P<0.05) and were observed for CMR2, CMR4–CMR9, CMR12–18 and CMR20.

Figure 4 Effects of CMR gene knockdown on mRNA levels of Myh6 (α-MHC), Myh7 (β-MHC), Mlc2a and Mlc2v in differentiated ES cells

R1 cells were transfected with 50 nM siRNA targeting the CMR genes or the negative control siRNA in the presence of 1000 units/ml LIF. After 24 h of transfection, differentiation of R1 cells was initiated by forming EBs without LIF. At day 8, total RNA was isolated from differentiated R1 cells and subjected to one-step qRT-PCR. mRNA levels were normalized to those of Gapdh. Expression levels of Myh6 (α-MHC) (A), Myh7 (β-MHC) (B), Mlc2a (C) and Mlc2v (D) in cells transfected with the negative control were set to 1. Results are means±S.E.M. (n=5). Statistical significance was determined using a Student's t test (*P<0.05 compared with control).

Depletion of the CMR10 gene, 2810405K02Rik, exhibited significant inhibitory effects on cardiac differentiation in both EC cells and ES cells. Depletion of the CMR12 gene, Cd302, however, increased most of the cardiac marker genes and promoted cardiomyogenesis in both EC cells and ES cells. It is of note that depletion of the CMR2 gene, AW551984, whose function is almost unknown, markedly blocked the expression of all cardiac gene markers that we measured in both EC cells and ES cells when these cells were differentiated. These results indicate that, in our identified CMR genes, CMR2 is the most potent regulator of cardiac differentiation in EC cells and ES cells.

AW551984 regulates Nkx2.5 gene expression during cardiac differentiation of ES cells in response to Wnt3a

We attempted to analyse the function of CMR2, AW551984, in ES cell differentiation into cardiomyocytes. As shown in Figure 3(B), the mRNA level of the AW551984 gene was drastically elevated during the differentiation of ES cells. We confirmed the protein levels of AW551984 during the EB differentiation of R1 cells by immunoblotting. The band corresponding to AW551984 was almost undetectable in undifferentiated cells but clearly appeared at day 8 (Figure 5A), supporting the increase in transcript levels of AW551984 during differentiation. In addition, we counted the number of EBs exhibiting spontaneous beating on a daily basis after transfection with siRNA targeting AW551984. Consistent with the inhibited expression of cardiomyocyte markers (Figure 4), knockdown of AW551984 markedly decreased the number of EBs with beating activities (Figure 5B). These results strongly support the notion that AW551984 that is elevated during differentiation regulates the cardiac differentiation of ES cells. To identify the developmental stage regulated by AW551984, we examined whether or not AW551984 affects the expression of embryonic markers in R1 cells. No significant effect of AW551984 knockdown on Nanog and Oct3/4 (octamer-binding protein 3/4) levels was observed at day 0, indicating that AW551984 does not modulate the stemness of ES cells (Figure 5C). Moreover, neither early mesoderm marker T/brachyury nor the cardiac mesoderm marker Mesp1 (mesoderm posterior 1) were significantly affected by knockdown of AW551984 (Figure 5D). This indicated that AW551984 is not involved in the development of ES cells into cardiac mesoderm. We next examined the expression of early cardiac transcription factors Gata4, Mef2c, Nkx2.5 and Tbx5 (T-box 5) during the differentiation of ES cells. Interestingly, the RNAi against AW551984 significantly reduced the expression of Nkx2.5 at days 6 and 7, but did not affect those of Gata4, Tbx5 or Mef2c (Figure 5E). These results strongly suggest that AW551984 mainly regulates cardiac differentiation via the expression of Nkx2.5.

Figure 5 AW551984 is required for Nkx2.5 gene expression during cardiac differentiation of ES cells and is regulated by Wnt/β-catenin signalling

(A) Cell lysates from R1 cells at days 0 and 8 were subjected to Western blot analysis. AW551984 in the cell lysates was detected using the anti-AW551984 antibody. (BE) R1 cells were transfected with 50 nM siRNA targeting AW551984 or the negative control siRNA in the presence of 1000 units/ml LIF. After 24 h of transfection, differentiation of R1 cells was initiated by forming EBs. (B) EBs exhibiting spontaneous beating were counted and the percentages were calculated. Results are means±S.E.M. from three independent experiments. Statistical significance was determined using two-way repeated measures ANOVA and Bonferroni's post-hoc test (*P<0.01 compared with control). (C) After 24 h of transfection, total RNA was isolated from the cells. mRNA levels of Nanog and Oct3/4 were measured by one-step qRT-PCR and normalized to those of Gapdh. Expression levels in cells transfected with the negative control were set to 1. Results are means±S.E.M. (n=5). No statistical significance was observed using a Student's t test (P ≥ 0.05 compared with control). (D) Total RNA was isolated from the cells at days 3, 4 and 5 and subjected to one-step qRT-PCR. mRNA levels of T/brachyury and Mesp1 were normalized to those of Gapdh. Expression levels in cells that were transfected with the negative control and differentiated at day 3 were set to 1. Results are means±S.E.M. (n=5). No statistical significance on interaction between days and siRNAs was observed using two-way ANOVA (P≥0.05). (E) Total RNA of R1 cells at days 3–8 was isolated and subjected to one-step qRT-PCR. mRNA levels of Gata4, Nkx2.5, Mef2c and Tbx5 were normalized to those of Gapdh. These expression levels in cells that were transfected with the negative control and differentiated at day 3 were set to 1. Results are means±S.E.M. (n=5). Statistical significance was determined using two-way ANOVA and Bonferroni's post-hoc test (*P<0.01 compared with control). (F and G) EBs were cultured in the absence or presence of 100 ng/ml Wnt3a (F) or 500 ng/ml Dkk-1 (G) from days 2–5. Total RNA was isolated at day 5 and subjected to one-step qRT-PCR. Results are means±S.E.M. (n=5). Statistical significance was determined using a Student's t test (*P<0.01).

Wnt/β-catenin signalling has recently been reported to regulate the cardiac differentiation of mouse ES cells biphasically. Stimulation of ES cells with Wnt3a in the early phase for EB formation enhanced the cardiac differentiation associated with the development of cardiac markers such as Nkx2.5 [5,6]. Therefore we next examined whether or not Wnt3a in the early phase for EB formation influenced AW551984 expression. The addition of Wnt3a (100 ng/ml) into differentiation medium from day 2 to day 5 markedly increased the transcript of AW551984 as well as that of Nkx2.5 in differentiated ES cells at day 5 (Figure 5F). Reciprocally, the addition of Dkk-1 (500 ng/ml), an extracellular inhibitor of Wnt/β-catenin signalling from days 2–5 significantly inhibited the expression of both AW551984 and Nkx2.5 at day 5 (Figure 5G) and at day 7 (results not shown). Thus Wnt/β-catenin signalling enhanced the expression of AW551984 during EB formation, leading to the commitment of ES cells into a cardiac lineage through Nkx2.5 expression.

DISCUSSION

EC cells and ES cells are used as in vitro models of early embryonic development, and they are useful for studying the mechanism underlying the cardiac differentiation of stem cells [3]. In the present study, we have identified cardiomyogenesis-related candidate genes expressed in EC cells through comprehensive analysis. A knockdown study using siRNA revealed that the candidate genes actually influence the cardiac differentiation of EC cells and ES cells. Furthermore, AW551984, one of the candidate genes, potently regulated the cardiac differentiation via the expression of the Nkx2.5 transcript and is suggested to act downstream of Wnt/β-catenin signalling.

Several studies have attempted to identify the factors involved in the cardiac differentiation of EC cells and ES cells [12,13]. We sought to determine whether or not any additional factors existed to achieve efficient differentiation into cardiomyocytes. To address this question, we statistically compared gene expression patterns in an undifferentiated state with cardiomyogenic potential of the EC cell strains and successfully identified 24 CMR genes (Table 1). The CMR genes contained F2r, Ptprb and Adm genes, which are associated with cardiac development in vivo. Knockout mouse studies have demonstrated that the ablation of F2r [16], Ptprb [17] and Adm [18] leads to fetal bleeding, defects of angiogenesis and the heart, and cardiovascular defects respectively, indicating the high reliability of our approach in the identification of cardiomyogenesis candidate genes. It is worth noting that our screen did not isolate any cardiac transcription factors reported previously. Expression of cardiac transcription factors increases approximately in a time-dependent manner ‘after’ induction of differentiation. However, it is kept low in the undifferentiated state. We isolated genes whose expression ‘before’ induction was associated with cardiomyogenesis in EC cells. Therefore some of the CMR genes are presumed to play more significant roles at initiation or a very early stage of cardiac differentiation in EC cells, compared with cardiac transcription factors. In addition, we confirmed the effects of the CMR genes on cardiac differentiation of EC cells and ES cells through RNAi experiments. AW551984, 2810405K02Rik and Cd302 were particularly effective at modulating the cardiac differentiation of both EC cells and ES cells (Figures 2 and 4, Table 2 and Supplementary Table S4).

RNAi against AW551984 notably inhibited the cardiac differentiation of mouse EC cells and ES cells (Figures 2 and 4, and Table 2). To the best of our knowledge, we are the first to demonstrate that AW551984 is involved in the cardiac differentiation of stem cells. Furthermore, we revealed that AW551984 selectively regulates the expression of a cardiac transcription factor Nkx2.5, but not Gata4, Mef2c or Tbx5 (Figure 5E). Because Nkx2.5 is a homeobox transcription factor that is essential for the development of ventricular cardiomyocytes [19], AW551984 regulates the cardiac differentiation of ES cells mainly through Nkx2.5 activities. Ritner et al. [20] have recently reported that the expression of Nkx2.5 is shown to spike and then decreases during EB differentiation in human ES cells [20]. Expression of Nkx2.5 peaked at day 7 of differentiation in our experimental conditions, and knockdown of AW551984 inhibited Nkx2.5 expression at days 6–8 (Figure 5E). Therefore AW551984 exerts its effects on cardiomyogenesis due to modulating levels but not the peaking time of Nkx2.5 transcripts. Wnt/β-catenin signalling is known to regulate the biphasic cardiac differentiation of ES cells. Treatment of EBs with Wnt3a in the early stages of ES cell differentiation is known to facilitate cardiac differentiation following an increase in Nkx2.5 expression [5,6]. However, factors elevating early cardiac transcription factors are poorly known downstream of Wnt/β-catenin signalling. In the present study, we have demonstrated that Wnt/β-catenin signalling in the early stages of ES cell differentiation enhances the expression of AW551984 as well as that of Nkx2.5 (Figures 5F and 5G). Collectively, our results suggest that AW551984 plays a significant and critical role in the cardiac differentiation of ES cells as a novel intermediate molecule linking Nkx2.5 expression to Wnt/β-catenin signalling. The transcription-factor-binding site prediction web server (http://www.sabiosciences.com/chipqpcrsearch.php?app=TFBS) predicted that p300 acetyltransferase could regulate the gene expression of both AW551984 and the human orthologue of AW551984 VWA5A (von Willebrand factor A domain containing 5A). p300 binds to β-catenin and serves as a co-activator of β-catenin, regulating the β-catenin–Tcf4 (T-cell factor) interaction [21]. β-Catenin stabilized by Wnt signalling may inversely influence the function of p300, leading to up-regulation of AW551984 expression. Identifying the mechanism by which AW551984 regulates Nkx2.5 expression should be carried out to enhance our understanding of cardiac differentiation in the future.

AW551984 has been reported by several groups to be associated with cancer. The human orthologue VWA5A is located at chromosome 11q23-q24, which corresponds to regions of frequent loss of heterozygosity in solid tumours [22]. AW551984 was also identified as a metastasis-related gene, and knockdown of this gene by shRNA (small hairpin RNA) accelerated cell migration in a melanoma cell line [23]. Thus AW551984 had been recognized as a tumour suppressor. Interestingly, mRNA expression of AW551984 decreased in NIH 3T3 cells responding to Wnt3a [24], in direct contrast with our results with EBs treated with Wnt3a (Figure 5F). This inverse effect of Wnt3a may be explained by the difference in the cell types. Alternatively, non-canonical Wnt signalling may contribute to the decrease in AW551984 expression in NIH 3T3 cells.

2810405K02Rik has been identified as a novel type of prostamide/prostaglandin F synthase belonging to the thioredoxin-like superfamily [25]. Prostamide/prostaglandin F synthase has been reported to be abundantly expressed in the spinal cord and is thought to play an important role in the central nervous system [26]. However, its function in terms of cardiac development remains unknown.

Cd302 is a C-type lectin receptor and has been reported to be involved in cell adhesion and migration as well as endocytosis and phagocytosis [27]. Knockdown of Cd302 significantly facilitated cardiac differentiation in both EC and ES cells, although expression of the Cd302 gene in undifferentiated EC cells was positively correlated with the first principal component score and number of nodules, and was negatively correlated with lag time to the onset of beating (Table S3). This may be due to the effective timing of Cd302 suppression by RNAi on cardiomyogenesis. The direction of regulating ES cell cardiomyogenesis by Wnt/β-catenin and BMP (bone morphogenetic protein) signalling can actually be determined by the timing of the addition of ligands and their antagonists [5,28,29]. Cd302 may possibly act downstream of the Wnt/β-catenin and/or BMP signalling.

Recently, differentiated ES cells and iPS cells have been extensively explored as cell therapy products for regenerative medicine because of their pluripotency and unlimited growth. To yield the expected differentiated cells, it is necessary to increase the differentiation efficiency of ES and iPS cells. The CMR genes identified in the present study will enable the more efficient differentiation of ES cells and iPS cells into cardiomyocytes in vitro, thereby adding to the known factors that facilitate cardiac differentiation. The functional roles of the CMR gene homologues need to be clarified in human ES cells and iPS cells, the properties of which are similar to those of mouse epiblast stem cells [30]. In addition, the CMR genes may contribute not only to cardiomyogenesis, but also to cardiac organogenesis and may assist in our understanding of the mechanism of cardiac development in vivo in the future.

AUTHOR CONTRIBUTION

Satoshi Yasuda contributed to the conception and design of the study, and performed the ES cell experiments. Tetsuya Hasegawa conducted the microarray and RNAi experiments in the EC cell lines with Tetsuji Hosono. Kei Watanabe analysed the phenotypes of the EC cell lines with Kageyoshi Ono. Mitsutoshi Satoh, Shunichi Shimizu, Takao Hayakawa, Teruhide Yamaguchi and Kazuhiro Suzuki provided theoretical input and critical advice for the stem cell phenotyping. Yoji Sato led the conception and design of the study, established the CL6 cell sublines and directed the work. Satoshi Yasuda, Mitsutoshi Satoh, Takao Hayakawa, Teruhide Yamaguchi, Shunichi Suzuki and Yoji Sato also contributed to securing funding. Satoshi Yasuda wrote the manuscript together with Tetsuya Hasegawa, Kei Watanabe, Mitsutoshi Satoh, Shunichi Shimizu and Yoji Sato. The remaining authors commented on the final text.

FUNDING

This work was supported by Grants-in-Aid from the Ministry of Health, Labour and Welfare, from the Ministry of Education, Culture, Sports and Technology, and from the Japan Science and Technology Agency.

Acknowledgments

We thank K. Shimizu and T. Nishimura (Division of Environmental Chemistry, National Institute of Health Sciences, Tokyo, Japan) for technical advice.

Abbreviations: 2810405K02Rik, RIKEN cDNA 2810405K02 gene, Adm, adrenomedullin; BMP, bone morphogenetic protein; Cd302, CD302 antigen; CMR, cardiomyogenesis-related candidate; Dkk-1, Dickkopf-1; EB, embryoid body; EC, embryonal carcinoma; ES, embryonic stem; F2r, coagulation factor II (thrombin) receptor; FBS, fetal bovine serum; Gapdh, glyceraldehyde-3-phosphate dehydrogenase; Gata4, GATA binding protein 4; iPS, induced pluripotent stem; LIF, leukaemia inhibitory factor; Mef2c, myocyte enhancer factor 2C; MEF, mouse embryonic fibroblast; Mesp1, mesoderm posterior 1; MHC, mysoin heavy chain; Mlc2a, myosin light chain 2a; Mlc2v, myosin light chain 2v; Myh6, α-MHC; Myh7, β-MHC; Nkx2.5, NK2 transcription factor related locus 5; Oct3/4, octamer-binding protein 3/4; PCA, principal component analysis; Ptprb, protein tyrosine phosphatase receptor type B; qRT-PCR, quantitative real-time PCR; RNAi, RNA interference; siRNA, small interfering RNA; Tbx5, T-box 5; VWA5A, von Willebrand factor A domain containing 5A

References

View Abstract