|Home | About | Journals | Submit | Contact Us | Français|
Bone morphogenetic proteins (BMPs) have been shown to induce ectopic expression of cardiac transcription factors and beating cardiomyocytes in nonprecardiac mesodermal cells in chicks, suggesting that BMPs are inductive signaling molecules that participate in the development of the heart. However, the precise molecular mechanisms by which BMPs regulate cardiac development are largely unknown. In the present study, we examined the molecular mechanisms by which BMPs induce cardiac differentiation by using the P19CL6 in vitro cardiomyocyte differentiation system, a clonal derivative of P19 embryonic teratocarcinoma cells. We established a permanent P19CL6 cell line, P19CL6noggin, which constitutively overexpresses the BMP antagonist noggin. Although almost all parental P19CL6 cells differentiate into beating cardiomyocytes when treated with 1% dimethyl sulfoxide, P19CL6noggin cells did not differentiate into beating cardiomyocytes nor did they express cardiac transcription factors or contractile protein genes. The failure of differentiation was rescued by overexpression of BMP-2 or addition of BMP protein to the culture media, indicating that BMPs were indispensable for cardiomyocyte differentiation in this system. Overexpression of TAK1, a member of the mitogen-activated protein kinase kinase kinase superfamily which transduces BMP signaling, restored the ability of P19CL6noggin cells to differentiate into cardiomyocytes and concomitantly express cardiac genes, whereas overexpression of the dominant negative form of TAK1 in parental P19CL6 cells inhibited cardiomyocyte differentiation. Overexpression of both cardiac transcription factors Csx/Nkx-2.5 and GATA-4 but not of Csx/Nkx-2.5 or GATA-4 alone also induced differentiation of P19CL6noggin cells into cardiomyocytes. These results suggest that TAK1, Csx/Nkx-2.5, and GATA-4 play a pivotal role in the cardiogenic BMP signaling pathway.
The heart is formed through multiple developmental steps which include the determination of the cardiac field in the mesoderm, differentiation of cardiac precursor cells, and maturation of the heart (42, 43). Many classical embryonic studies have implicated the mechanism of how and where these steps take place in the developing embryos. The vertebrate heart arises from paired mesodermal primordia that migrate to the anterior ventral midline, where they fuse and undergo terminal differentiation (18, 46). In Xenopus embryos, the cardiac field is located in the dorsal mesoderm lateral to the Spemann organizer and is specified prior to the end of gastrulation. In this relatively early step of cardiac development, inductive signals from the adjacent deep endoderm and the organizer region play a pivotal role in the determination of the cardiac field (13, 21, 23, 42, 47, 48, 59). Subsequently, the cardiac primordia always lie in close contact with the endoderm while migrating anterolaterally, and interactions between the endoderm and the overlying mesoderm are thought to be important for the promotion of cardiomyocyte differentiation in cardiac mesodermal cells. In Xenopus, the presence of the deep dorsoanterior endoderm markedly enhances the heart formation in explants of heart primordia, and the presence of both the endoderm and the organizer is necessary and sufficient to induce beating heart tissue in ventral mesoderm explants (42). In chicks, the anterior endoderm also induces the differentiation of nonprecardiac mesodermal cells into heart tissue (48). These observations also suggest that the endoderm-derived signals play a vital role both in the specification of the cardiac field and in the differentiation of determined cardiac precursor cells. However, the precise molecular mechanisms that regulate these inductive events during the formation of the heart are largely unknown at present.
Recent advances in understanding the genetic pathway of heart development have allowed us to use cardiac-restricted transcription factors as early heart-specific markers for dissecting the molecular mechanism of cardiogenesis. Among the several transcription factors implicated in cardiac development, Csx/Nkx-2.5, MEF2C, and GATA-4 have been well characterized in recent years. Csx/Nkx-2.5 is an NK-2 class homeodomain factor that was originally identified as a potential vertebrate homolog of Drosophila tinman (25, 31). The tinman gene is initially expressed in all mesodermal cells, but subsequently its expression domain is restricted to the dorsal part of the mesoderm, and later in development, expression of tinman is observed only in the dorsal vessel, an insect equivalent for the vertebrate heart (4, 5). Murine Csx/Nkx-2.5 is also predominantly expressed in the heart and in cardiac progenitor cells from the early developmental stage when two heart primordia are symmetrically situated in the anterior lateral mesoderm. The heart does not form at all in the tinman mutant of Drosophila (5), whereas in Csx/Nkx-2.5 knockout mice, the heart forms, but its development stops at the looping stage (32). MEF2C belongs to the MEF2 subfamily of MADS-box transcription factors and binds to the AT-rich element in regulatory regions of numerous muscle-specific genes (6, 29, 44). GATA-4 is a member of the cardiac GATA subfamily, which consists of GATA-4, -5, and -6, and binds to the WGATAR motif in promoter regions of cardiac- or gut-specific genes (9, 15, 22, 39, 55). MEF2C and GATA-4 are also thought to be involved in the early stage of cardiogenesis. Both of them started to be expressed in the precardiac mesoderm almost simultaneously with Csx/Nkx-2.5. Targeted disruption of MEF2C results in right ventricular dysplasia (30), and bilateral cardiac primordia fail to fuse in GATA-4−/− mice because of the ventral folding and fusion defects of the developing embryo (26, 38). Thus, Drosophila tinman, vertebrate Csx/Nkx-2.5, MEF2C, and GATA-4 are critical regulators of cardiac development and are useful molecular markers for examining effects of inductive signals from other tissues or germ layers.
In this respect, several experiments were performed by using cardiac-specific transcription factors as cardiac markers to elucidate the molecular mechanism of cardiogenesis and have demonstrated that bone morphogenetic proteins (BMPs) play a vital role in cardiac development. Initially, it was reported that expression of tinman is restricted to the dorsal part of the mesoderm by the ectodermally expressed decapentaplegic (dpp), a member of the transforming growth factor β (TGF-β) superfamily that is most closely related to vertebrate BMP-2 or BMP-4 (10). Recently, the ectopic expression of Csx/Nkx-2.5 and GATA-4 was also induced by the implantation of BMP-2-soaked beads in nonprecardiac mesoderms in chicks (49), suggesting that BMPs play a pivotal role in the induction of vertebrate cardiac development. At present, however, the precise mechanism by which BMPs induce the differentiation of cardiac precursor cells is largely unknown.
In the investigation of the molecular mechanisms of cardiomyocyte differentiation, the in vitro culture system presents a great advantage. P19 embryonal carcinoma cells are undifferentiated stem cells derived from murine teratocarcinoma (34) and differentiate into a variety of cell types representative of all three germ layers after suspension culture in the presence of several chemical inducers. When exposed to a relatively low concentration of retinoic acid (1 to 10 nM) or dimethyl sulfoxide (DMSO) (0.5 to 1%), some P19 cells differentiate into endodermal and mesodermal cells, including cardiomyocytes (8, 35). Although the P19 cell line has been widely used as a model system of cardiogenesis in vitro, its utility is limited because of its quite low efficiency of differentiation into cardiomyocytes. Recently, a clonal derivative named P19CL6 was isolated from P19 cells (17). Unlike P19 cells, this subline efficiently differentiates into beating cardiomyocytes with adherent conditions when treated with 1% DMSO. Since almost all cells differentiate into cardiomyocytes which express cardiac-specific genes, P19CL6 cells are a useful in vitro model to study cardiomyocyte differentiation (17).
In the present study, we examined the role of BMPs in the differentiation of cardiomyocytes utilizing the P19CL6 in vitro system. For this purpose, we isolated a permanent P19CL6 cell line named P19CL6noggin that stably overexpresses the BMP antagonist noggin. In contrast to parental P19CL6 cells, P19CL6noggin cells did not differentiate into beating cardiomyocytes and expression of cardiac-specific genes was not induced when treated with DMSO. Overexpression of BMP-2 or addition of BMP protein to the medium restored the ability of P19CL6noggin cells to differentiate into cardiomyocytes, suggesting that BMPs were indispensable for cardiomyocyte differentiation. The failure of P19CL6noggin cells to differentiate into cardiomyocytes was also rescued by overexpression of TAK1, a member of the mitogen-activated protein kinase kinase kinase (MAPKKK) superfamily that has been demonstrated to be involved in BMP signaling, whereas overexpression of the dominant negative form of TAK1 inhibited differentiation of parental P19CL6 cells into cardiomyocytes. Simultaneous overexpression of Csx/Nkx-2.5 and GATA-4 also rescued the differentiation defect of P19CL6noggin cells, although overexpression of Csx/Nkx-2.5 or GATA-4 alone did not. These results suggest that the MAPK pathway activated by TAK1 and two cardiac transcription factors, Csx/Nkx-2.5 and GATA-4, mediate BMP-induced cardiomyocyte differentiation.
Murine noggin cDNA was kindly provided by R. M. Harland (36). Expression plasmids encoding wild-type TAK1 and TAK1 derivatives (64), a Raf-1 mutant (37), murine GATA-4 (2), and murine MEF2C cDNA (30) were provided by H. Shibuya, R. J. Davis, D. B. Wilson, and E. N. Olson, respectively. Adenoviral vectors containing human BMP-2 cDNA (41) and expression plasmids encoding human CSX1a cDNA (53) were previously described.
P19CL6 and P19CL6noggin cells were cultured essentially as described previously (17). In brief, the cells were grown in a 100-mm tissue culture grade dish under adherent conditions with α-minimal essential medium (Gibco BRL) supplemented with 10% fetal bovine serum (JRH Bioscience), penicillin (100 U/ml), and streptomycin (100 μg/ml) (growth medium) and were maintained in a 5% CO2 atmosphere at 37°C. To induce differentiation under adherent conditions, P19CL6 and P19CL6noggin cells were plated at a density of 3.7 × 105 in a 60-mm tissue culture grade dish with the growth medium containing 1% DMSO (differentiation medium). The medium was changed every 2 days. Days of differentiation were numbered consecutively after the first day of the DMSO treatment, day 0. Natural bovine BMP cocktail (Sangi), which contains almost all types of bone-derived BMPs, including BMP-2 and BMP-4, was added into the differentiation medium at concentrations of 10 and 100 ng/ml.
Murine noggin cDNA was subcloned into pEFSAneo, which harbors the human elongation factor 1-α promoter and a neomycin resistance gene (53), and the resultant pEFSAneo-noggin was transfected into P19CL6 cells by the lipofection method (Tfx Reagents; Promega). Stable transformants were selected with 400 μg of neomycin (G418) per ml, and 12 independent cell lines were isolated.
Adenovirus-mediated BMP-2 gene delivery was performed on day 2 of differentiation. In addition to an adenoviral vector containing human BMP-2 cDNA, an adenoviral vector alone was also used as a negative control. Three microliters of virus suspension (1.0 × 1010 plaque-forming units/ml) was added into 1 ml of differentiation medium in a 60-mm culture dish. The culture dishes were incubated in a 5% CO2 atmosphere at 37°C for 2 h, and then the medium was changed. The expression vectors containing TAK1, TAK1 derivatives, a constitutively active form of Raf-1 (caRaf-1), Csx/Nkx-2.5, and GATA-4 were transfected on day 2 of differentiation according to the lipofection method as recommended (Promega).
Immunostaining with MF20, a monoclonal antibody against a sarcomeric myosin heavy chain (MHC), was performed as described previously (3) by using anti-mouse immunoglobulin G conjugated with tetramethyl rhodamine isothiocyanate as the secondary antibody. MF20-positive areas were measured on day 16 of differentiation by directly tracing the stained areas on a photograph.
Total RNA was extracted by the acid guanidine method (RNA zol B; Biotecx Laboratories, Inc.), and Northern blot analysis was performed as described below with 10 μg of total RNA for noggin, TAK1, GATA-4, MEF2C, MHC, and MLC2v. Total RNA was subjected to agarose-formaldehyde gel electrophoresis and subsequently transferred onto a Hybond N+ membrane filter (Amersham). Hybridization was carried out in 40% formamide, 5× SSPE (1× SSPE is 0.18 M NaCl, 10 mm NaPO4, and 1 mM EDTA [pH 7.7]), 5× Denhardt’s solution, 5× dextran sulfate, and 1% sodium dodecyl sulfate at 42°C overnight. The probes were labeled with [32P]dCTP by random priming (Takara). The following cDNA fragments were used as probes: the NotI/XhoI fragment of pBluescript containing murine noggin cDNA (36), the EcoRI/XbaI fragment of pEF containing murine TAK1DN cDNA (64), the EcoRI fragment of pMT2 containing murine GATA-4 cDNA (2), the EcoRI fragment of pcDNA1 containing murine MEF2C cDNA (30), the PstI fragment of pMHC25 containing rat skeletal muscle MHC cDNA (61), and the EcoRI fragment of pCRII containing a PCR product obtained by using oligonucleotide primers specific for MLC2v (32). For the analysis of BMP-2, BMP-4, and Csx/Nkx-2.5 mRNA, reverse transcription (RT)-PCR was performed. First-strand cDNA was synthesized with Superscript II reverse transcriptase and a random primer (Gibco BRL) from 5 μg of total RNA, and PCRs were performed with 1 μl of cDNA products, a 0.3 μM concentration of each oligonucleotide primer, and 1 U of Taq polymerase (Takara) in 50 μl of buffer containing 200 μM deoxynucleoside triphosphates. For BMP-2 and BMP-4, the PCR primers and regimen used were essentially as described previously (24). For Csx/Nkx-2.5, the primer sequences used were 5′-TCT CCG ATC CAT CCC ACT TTA TTG-3′ for sense and 5′-TTG CGT TAC GCA CTC ACT TTA ATG-3′ for antisense, which amplified the 3′-UTR of mouse Csx/Nkx-2.5 and were thereby specific for endogenous Csx/Nkx-2.5. PCR conditions were 94°C for 3 min, followed by 40 cycles of 94°C for 30 sec, 55°C for 30 sec, and 72°C for 1 min. PCR products were electrophoresed on 2% agarose gels and visualized by ethidium bromide staining.
To determine the role of BMPs in differentiation into cardiomyocytes, we isolated 12 independent P19CL6 clones which permanently overexpress murine noggin under the control of the human elongation factor 1-α promoter and designated them P19CL6noggin. When cultured in growth medium, both P19CL6 and P19CL6noggin cells grew well and did not differentiate into cardiomyocytes (Fig. (Fig.1).1). When 1% DMSO was added to the medium, P19CL6 cells differentiated into mononucleated, spontaneously contracting cardiomyocytes, positive for anti-sarcomeric MHC antibody MF20 (Fig. (Fig.1A).1A). As previously described (17), spontaneous beating was first observed on a limited area on day 10 (10 days after the initiation of DMSO treatment), and subsequently the majority of cells beat synchronously until around day 16 (Fig. (Fig.1B).1B). However, P19CL6noggin cells did not differentiate into MF20-positive beating cardiomyocytes after treatment with DMSO (Fig. (Fig.1A1A and B). The same results were obtained with at least six independent P19CL6noggin cell lines.
We next examined whether the failure of P19CL6noggin cells to differentiate into cardiomyocytes is due to the inhibition of BMPs. For this purpose, we first introduced human BMP-2 cDNA into P19CL6noggin cells on day 2 by adenovirus-mediated gene delivery (41). Many P19CL6noggin cells infected with BMP-2 adenovirus differentiated into mononucleated beating cardiomyocytes positive for MF20 (Fig. (Fig.1A1A and B), whereas the cells infected with control adenoviral vector did not (data not shown). We further investigated whether BMP protein directly applied to the culture media could also rescue the differentiation defect of these cells. P19CL6noggin cells incubated with 1% DMSO and 100 ng of BMP protein per ml but not with 1% DMSO and 10 ng of BMP protein per ml partially differentiated into beating cardiomyocytes (Fig. (Fig.1B),1B), implying that a sufficient amount of BMP protein was enough to overcome the inhibitory effect of noggin and restore the ability of P19CL6noggin cells to differentiate into cardiomyocytes. These results suggested that overexpression of noggin inhibited differentiation of P19CL6 cells into cardiomyocytes by blocking BMP signaling and that BMPs were required for the differentiation of P19CL6 cells into cardiomyocytes.
The expression of related genes, including cardiac-specific markers, was examined in P19CL6 cells and P19CL6noggin cells during differentiation (Fig. (Fig.2).2). Expression of BMP-2 and BMP-4 was recognized during the course of differentiation both in P19CL6 and P19CL6noggin cells by RT-PCR analysis. That BMP-2 and BMP-4 were expressed in P19CL6 cells not treated with DMSO (i.e., on day 0) suggested that BMP expression was not induced by DMSO. Northern blot analysis revealed that although noggin was not detected in parental P19CL6 cells, abundant expression of noggin mRNA was observed throughout differentiation in P19CL6noggin cells, as we expected. Cardiac transcription factors Csx/Nkx-2.5, GATA-4, and MEF2C started to be expressed on day 6 in P19CL6 cells while expression of these genes was not detected in P19CL6noggin cells during the course of the observation. Expression of MHC and MLC2v genes was detected in P19CL6 cells on day 12, while in P19CL6noggin cells, expression of these genes was not detected. These results indicated that BMPs were essential for the expression of at least some set of cardiac-specific genes.
In order to elucidate the mechanisms by which BMPs regulate differentiation into cardiomyocytes, we examined whether TAK1, a member of the MAPKKK superfamily, was involved in BMP-induced differentiation into cardiomyocytes. TAK1 has been reported to transduce BMP signaling (64). Northern blot analysis revealed that expression of endogenous TAK1 mRNA was detected throughout the differentiation, both in P19CL6 and P19CL6noggin cells (Fig. (Fig.2).2). We investigated whether and how P19CL6 and P19CL6noggin cells differentiated into cardiomyocytes after transient transfection of expression vectors containing TAK1 mutants. The P19CL6 cells transfected with the dominant negative form of TAK1 (dnTAK1) differentiated into MF20-positive beating cardiomyocytes after treatment with DMSO less efficiently than the control P19CL6 cells (Fig. (Fig.3A3A and B). This reduction of the efficiency was thought to be compatible with the transfection efficiency, which was estimated to range from approximately 40 to 70% by counting the green fluorescent protein-positive cells 1 day after the transfection of green fluorescent protein expression plasmids (data not shown). Subsequently, wild-type TAK1 (TAK1), the constitutively active form of TAK1 (caTAK1), and dnTAK1 were transfected into P19CL6noggin cells on day 2 of differentiation. Unlike the control P19CL6noggin cells, the P19CL6noggin cells transfected with TAK1 or caTAK1 partially differentiated into beating cardiomyocytes positive for MF20 in the presence of 1% DMSO (Fig. (Fig.3A).3A). The MF20-positive areas ranged from 25 to 45% (mean 33%) of the whole areas in P19CL6noggin cells transfected with TAK1 and from 42 to 68% (mean 55%) in the cells transfected with caTAK1 (Fig. (Fig.3B).3B). The differences in the areas positive for MF20 between these two transfected cells may be compatible with the previous report that the constitutively active form of TAK1 induces a TGF-β-specific promoter with greater efficiency than wild-type TAK1 in the absence of ligand stimulation (64). On the other hand, P19CL6noggin cells transfected with dnTAK1 did not differentiate into cardiomyocytes (Fig. (Fig.3A3A and B). Without DMSO treatment, differentiation into cardiomyocytes was not induced, even when TAK1 or caTAK1 was overexpressed (data not shown). Similar results were obtained in at least three independent P19CL6noggin cell lines. Furthermore, we examined whether other members of the MAPKs rescued the differentiation defect of P19CL6noggin cells. caRaf-1, a MAPKKK which transduces signals of the classical MAPK pathway, was transfected into P19CL6noggin cells 2 days after the initiation of DMSO treatment. Unlike the P19CL6noggin cells transfected with caTAK1, the cells transfected with caRaf-1 did not differentiate into beating cardiomyocytes positive for MF20 (data not shown), implying that the rescue of the ability of P19CL6noggin cells to differentiate into cardiomyocytes was specific for the MAPK pathway mediated by TAK1.
RT-PCR and Northern blot analyses revealed that, in contrast to the control P19CL6noggin cells (Fig. (Fig.4,4, lane 1), expression of Csx/Nkx-2.5, GATA-4, and MEF2C was induced in P19CL6noggin cells transfected with TAK1 or caTAK1 (Fig. (Fig.4,4, lanes 2 and 3). MHC and MLC2v were also expressed in these cells (Fig. (Fig.4,4, lanes 2 and 3). In P19CL6noggin cells transfected with dnTAK1, no expression of cardiac-specific genes was observed (Fig. (Fig.4,4, lane 4). These results indicated that overexpression of TAK1 could rescue the differentiation defect of P19CL6noggin cells with concomitant expression of cardiac-specific genes and suggested that the MAPK pathway activated by TAK1 might mediate the BMP-induced differentiation into cardiomyocytes.
Overexpression of Csx/Nkx-2.5 induces enlargement of the heart in Xenopus species (7, 11) and ectopic beating foci in zebrafish (51), suggesting that Csx/Nkx-2.5 has the potential to convert the mesodermal cells normally fated to be other cell types into cells of the cardiac lineage. It has been reported that GATA-4 is also implicated in the differentiation of P19-derived cardiomyocytes in vitro (15). Recent work with chick embryos has demonstrated that ectopic expression of BMP-2 induces the expression of both Csx/Nkx-2.5 and GATA-4 in nonprecardiac mesoderm and that the presence of noggin protein in the medium inhibits the expression of these two factors in cultured precardiac mesoderm (49). To elucidate the roles of Csx/Nkx-2.5 and GATA-4 in BMP-induced differentiation into cardiomyocytes, expression plasmids containing human CSX1a cDNA or murine GATA-4 cDNA were transfected into P19CL6noggin cells on day 2 by the lipofection method. Overexpression of Csx/Nkx-2.5 or GATA-4 alone did not induce differentiation into MF20-positive cardiomyocytes in these cells after treatment with DMSO (Fig. (Fig.5).5). Since Csx/Nkx-2.5 and GATA-4 have been reported to exhibit synergistic effects (27, 50, 54), we next overexpressed these two factors simultaneously. Cooverexpression of Csx/Nkx-2.5 and GATA-4 markedly induced differentiation into beating cardiomyocytes positive for MF20 in the presence of 1% DMSO (Fig. (Fig.5A).5A). The differentiation efficiency was estimated at approximately 50%, which was compatible with the transfection efficiency (Fig. (Fig.5B),5B), whereas without DMSO treatment, the differentiation into cardiomyocytes was not induced even when P19CL6noggin cells were transfected with both factors (data not shown). Similar results were obtained in at least three independent cell lines. These results suggested that although the induction of either Csx/Nkx-2.5 or GATA-4 was not sufficient for initiating the differentiation program of this cell line, induction of both Csx/Nkx-2.5 and GATA-4 was sufficient for BMP-mediated differentiation into cardiomyocytes in the presence of DMSO.
Expression of cardiac-specific genes was also examined on day 14 in P19CL6noggin cells with forced expression of exogenous Csx/Nkx-2.5 and/or GATA-4 (Fig. (Fig.6).6). Interestingly, expression of endogenous Csx/Nkx-2.5, MEF2C, and MLC2v but not of GATA-4 or MHC was induced in P19CL6noggin cells by overexpression of Csx/Nkx-2.5 alone (Fig. (Fig.6,6, lane 2). On the other hand, no expression of the cardiac genes tested was recognized in the cells transfected with GATA-4 alone (Fig. (Fig.6,6, lane 3). In the cells overexpressing both Csx/Nkx-2.5 and GATA-4, expression of all cardiac transcription factors and contractile protein genes was induced (Fig. (Fig.6,6, lane 4). These data suggested that although some set of cardiac genes were induced by forced expression of Csx/Nkx-2.5 alone, differentiation into beating cardiomyocytes required the cooperative effects of both Csx/Nkx-2.5 and GATA-4.
In the present study, we obtained the following results. (i) BMPs are required for the differentiation of P19CL6 cells into beating cardiomyocytes. (ii) BMP-induced differentiation into cardiomyocytes is mediated by the MAPKKK family member TAK1. (iii) Induction of two cardiac transcription factors, Csx/Nkx-2.5 and GATA-4, is required and sufficient for BMP-induced differentiation of P19CL6 cells into cardiomyocytes.
In Drosophila, dpp is expressed in the ectoderm and restricts the expression domain of tinman in the adjacent precardiac mesoderm (10). Dpp is a member of the TGF-β superfamily and is most closely related to vertebrate BMP-2 and BMP-4. In vertebrates, recent studies with chick embryos have demonstrated that BMPs are expressed in the ectoderm and the endoderm adjacent to the precardiac anterolateral mesoderm and that ectopic expression of BMPs induces ectopic expression of Csx/Nkx-2.5 and GATA-4 and differentiation into beating cardiomyocytes of nonprecardiac mesodermal cells (49). Furthermore, gene-targeting experiments have shown that normal cardiac development is impaired both in BMP-2 and BMP-4 knockout mice (62, 65). These results indicate that BMPs are required for normal cardiac development and suggest that the endoderm-derived inductive signals required for differentiation into cardiomyocytes are at least in part those from BMPs. However, the precise molecular mechanisms by which BMPs regulate cardiogenesis have been largely unknown because of the complexity in the in vivo situation. From this viewpoint, we used an in vitro system of differentiation into cardiomyocytes to dissect the cardiogenic pathways mediated by BMPs.
The secreted protein noggin was first identified as a dorsalizing factor localized in the Spemann organizer in Xenopus embryos (57). Subsequent studies have demonstrated that noggin binds specifically to BMP-2 and BMP-4 with high affinity and also to BMP-7 with lower affinity, thereby abolishing the activity of BMPs by blocking the binding of BMPs to cognate cell surface receptors (66). In fact, the effects of noggin in Xenopus embryos can be mimicked by agents that specifically block BMP signals such as dominant negative BMP receptors (14, 19, 33, 60) and antisense BMP RNAs (58). All these findings suggest that the biological function of noggin is to bind to BMPs and thereby antagonize BMP (especially BMP-2 and BMP-4) activity. Therefore, in this study, to elucidate the role of BMPs in differentiation into cardiomyocytes, we first isolated a permanent P19CL6 cell line that stably overexpresses noggin (P19CL6noggin). P19CL6noggin cells did not differentiate into beating cardiomyocytes when treated with DMSO, whereas overexpression of BMP-2 by adenoviral vectors or addition of a sufficient amount of BMP protein to the culture media rescued the differentiation defect of P19CL6noggin cells, suggesting that BMP signaling was indispensable for the differentiation of P19CL6 cells into cardiomyocytes.
Parental P19 cells differentiate into a variety of cell types, including mononucleated beating cardiomyocytes and bipolar multinucleated nonbeating myoblasts which fuse into myotubes (35). In P19CL6 cells, however, almost all the cells differentiated into beating cardiomyocytes, and skeletal muscle-like cells were not observed. On the other hand, although most P19CL6noggin cells did not differentiate into MF20-positive cells, a very small number of P19CL6noggin cells in a few cell lines differentiated into MF20-positive skeletal muscle-like cells which resembled the myocytes observed in the P19 cell aggregates (data not shown), suggesting that the MF20-positive cells observed in a reduced number of P19CL6noggin cells were skeletal muscle cells. Recently, it has been reported that when BMP-2-expressing cells are implanted adjacent to paraxial mesoderm in chick embryos formation of the somite is impaired (1). Another study has demonstrated that the myogenesis within somites is controlled by the relative levels of BMP activity and noggin-mediated anti-BMP activity (45). Noggin knockout mice exhibit a deficit in the differentiation of muscle precursor cells in somites, indicating that inactivation of BMPs is required for differentiation into skeletal muscle cells (36). These observations and our results together suggest that BMP signaling is required for differentiation into cardiomyocytes, while inhibition of BMP signaling induces differentiation into skeletal muscle cells.
TAK1 is a member of the MAPKKK superfamily and was originally identified as a molecule which complements a yeast STE11 (MAPKKK) mutant (64). TAK1 is activated by BMPs, and overexpression of wild-type TAK1 induces ventralization of Xenopus embryos, while kinase-negative TAK1 inhibits constitutively active BMP receptor-induced ventralization (52). All these results suggest that TAK1 mediates the activity of BMPs. To elucidate the mechanism by which BMPs regulate differentiation into cardiomyocytes, we examined the involvement of TAK1 in BMP-induced differentiation. Overexpression of TAK1 restored the ability of P19CL6noggin cells to differentiate into beating cardiomyocytes, whereas overexpression of the dominant negative form of TAK1 in parental P19CL6 cells reduced the differentiation efficiency into cardiomyocytes, suggesting that the TAK1-mediated MAPK pathway was involved in BMP-induced differentiation of cardiac precursor cells. We examined whether other MAPK pathways were also involved in BMP-mediated differentiation of P19CL6 cells into cardiomyocytes. Overexpression of the constitutively active form of Raf-1, a MAPKKK which transduces signals of the classical MAPK pathway, did not restore the ability of P19CL6noggin cells to differentiate into cardiomyocytes, implying that the TAK1-mediated MAPK pathway specifically rescued the differentiation defect caused by the blockade of BMP signaling. Since TAK1 has been reported to activate both the c-Jun N-terminal kinase and p38MAPK (40), it remains to be determined which member of the MAPK family is involved in TAK1-mediated differentiation into cardiomyocytes.
SMAD proteins have recently been identified and characterized as important mediators of the TGF-β superfamily signal transduction pathways (20). Among the members of SMADs, Smad1, -5, and -8 transduce signals from BMPs specifically, while Smad4 is a general partner of ligand-specific SMADs. Recent studies of Drosophila have shown that Dpp-induced tinman gene expression is positively regulated by the Smad4 homolog Medea (63), suggesting that the SMAD-mediated signal transduction pathway is also involved in the BMP-induced differentiation into cardiomyocytes. Overexpression of kinase-negative TAK1 has been reported to inhibit the Smad1-induced ventralization in Xenopus embryos (52), suggesting that the regulation by BMPs requires cooperative actions of SMADs and TAK1. We have preliminary data demonstrating that cooverexpression of Smad1 and Smad4 in P19CL6noggin cells induced differentiation into cardiomyocytes and that permanent overexpression of Smad6, an inhibitory SMAD which has been reported to block the TGF-β superfamily signal transduction, inhibited differentiation of P19CL6 cells into cardiomyocytes (39a), suggesting that SMADs also mediate BMP-induced differentiation of P19CL6 cells into cardiomyocytes. Further studies are necessary to elucidate the role of SMADs in vertebrate cardiogenesis and the cross talk between the SMAD pathway and the TAK1 pathway during differentiation of cardiac precursor cells.
Csx/Nkx-2.5 and GATA-4 are the two cardiac-enriched transcription factors that are expressed in precardiac mesoderm from the very early developmental stage, and these two factors are simultaneously induced by BMP-soaked beads in nonprecardiac mesoderms in chicks (49). In the present study, expression of Csx/Nkx-2.5 and GATA-4 was not induced in P19CL6noggin cells. Although forced expression of Csx/Nkx-2.5 or GATA-4 alone did not induce differentiation into beating cardiomyocytes, simultaneous overexpression of Csx/Nkx-2.5 and GATA-4 restored the ability of P19CL6noggin cells to differentiate into cardiomyocytes after treatment with DMSO. There may be several reasons for the need of both Csx/Nkx-2.5 and GATA-4 for the full differentiation of cardiac precursor cells. The simplest explanation is that cardiac genes induced by Csx/Nkx-2.5 are different from those induced by GATA-4, and both factors are required to activate all related genes needed for a full differentiation. Another explanation is that there is a cooperative action between Csx/Nkx-2.5 and GATA-4. In fact, Csx/Nkx-2.5 and GATA-4 showed synergistic transcriptional activation in an in vitro assay with the atrial natriuretic peptide promoter or the cardiac α-actin promoter (28, 50, 54). These explanations are not mutually exclusive and may operate at the same time. It is also noteworthy that the differentiation of cardiac precursor cells is almost normal and that the beating linear heart tube or the beating heart primordia formed in both Csx/Nkx-2.5 knockout mice and GATA-4 null mice (26, 32, 38). Taken together, these findings and the results of the present study suggest that genetic redundancies among NK-2 family members and GATA-4, -5, and -6 may partially rescue the phenotypes of the respective knockout mice. Recently, it has been reported that ectopic expression of GATA-4 in P19 cells accelerates cardiogenesis (16) and that overexpression of Csx/Nkx-2.5 in P19 cells induces differentiation into cardiomyocytes in the absence of DMSO (56). These results are partially different from ours, especially regarding the requirement of DMSO for differentiation. The understanding of the precise molecular requirement of both Csx/Nkx-2.5 and GATA-4 for the differentiation of cardiac precursor cells awaits further investigation.
In P19CL6noggin cells, overexpression of Csx/Nkx-2.5 induced the expression of MLC2v and MEF2C, suggesting that MLC2v and MEF2C were positively regulated by Csx/Nkx-2.5. These results were consistent with the previous results indicating that MLC2v expression was downregulated in Csx/Nkx-2.5 knockout mice (32) and with our recent data demonstrating that MLC2v expression was upregulated in mice overexpressing Csx/Nkx-2.5 (61a). Our present results are also compatible with those of a recent report showing that cardiac expression of Drosophila D-mef2 was positively regulated by Tinman via the Tinman-binding elements in the cardiac enhancer of the D-mef2 gene (12). In the cells overexpressing Csx/Nkx-2.5, endogenous Csx/Nkx-2.5 was also induced. The expression of the tinman gene has been reported to be positively regulated by the Tinman protein (27). These data suggested the existence of positive autoregulation by Csx/Nkx-2.5.
Expression of BMP-2 and BMP-4 was detected from day 0 (before DMSO treatment) by RT-PCR in P19CL6 cells, and overexpression of BMP-2 or TAK1 in P19CL6 cells was not sufficient to induce differentiation and expression of cardiac-specific genes in the absence of DMSO (data not shown). These results suggested that BMP signaling alone was not sufficient for cardiogenesis and that another factor(s) induced by DMSO was also essential for the differentiation of this cell line into cardiomyocytes. Recent studies have indicated that inductive signaling molecules different from BMPs are produced in the anterior endoderm and are also required for differentiation into cardiomyocytes (42, 48, 49). Although it is not known whether this anterior endoderm-derived unknown factor is the same factor which is induced by DMSO in P19 cells, it is possible that BMPs and the DMSO-induced unknown factor(s) cooperatively function to promote the differentiation of cardiac precursor cells.
A speculative diagram of the regulatory network controlling differentiation of P19CL6 cells into cardiomyocytes is shown in Fig. Fig.77 as a summary of our present study. Initially, BMPs (especially BMP-2 and/or BMP-4) transactivate the expression of two major cardiac-specific transcription factors, Csx/Nkx-2.5 and GATA-4. This transactivation is inhibited by overexpression of noggin and is mediated at least by TAK1. Subsequently, Csx/Nkx-2.5 and GATA-4 induce differentiation into cardiomyocytes cooperatively with unknown factors induced by DMSO. Some cardiac-specific genes, such as MEF2C and MLC2v, can be induced by Csx/Nkx-2.5 alone. Signals induced by DMSO are required for both the transactivation of Csx/Nkx-2.5 and GATA-4 by BMPs and the terminal differentiation into cardiomyocytes induced by these two factors (indicated by the X and Y on Fig. Fig.7),7), because neither the expression of Csx/Nkx-2.5 and GATA-4 nor terminal differentiation into cardiomyocytes was induced in the absence of DMSO. Thus, BMPs and downstream transcription factors are the central molecules of this regulatory network controlling cardiac differentiation as well as the DMSO-inducible factors. The identification of signals induced by DMSO in this system will provide new insights into the regulatory mechanisms of differentiation of cardiac precursor cells.
We thank R. M. Harland, H. Shibuya, R. J. Davis, D. B. Wilson, and E. N. Olson for providing plasmids.
This study was supported by a grant-in-aid for scientific research and developmental science research from the Ministry of Education, Science and Culture of Japan and the Program for Promotion of Fundamental Studies in Health Sciences of the Organization for Drug ADR Relief, R & D Promotion and Product Review of Japan (to I.K.).