Tbx20 Interacts With Smads to Confine Tbx2 Expression to the Atrioventricular Canal
Rationale: T-box transcription factors play critical roles in the coordinated formation of the working chambers and the atrioventricular canal (AVC). Tbx2 patterns embryonic myocardial cells to form the AVC and suppresses their differentiation into chamber myocardium. Tbx20-deficient embryos, which fail to form chambers, ectopically express Tbx2 throughout the entire heart tube, providing a potential mechanism for the function of Tbx20 in chamber differentiation.
Objective: To identify the mechanism of Tbx2 suppression by Tbx20 and to investigate the involvement of Tbx2 in Tbx20-mediated chamber formation.
Methods and Results: We generated Tbx20 and Tbx2 single and double knockout embryos and observed that loss of Tbx2 did not rescue the Tbx20-deficient heart from failure to form chambers. However, Tbx20 is required to suppress Tbx2 in the developing chambers, a prerequisite to localize its strong differentiation-inhibiting activity to the AVC. We identified a bone morphogenetic protein (Bmp)/Smad-dependent Tbx2 enhancer conferring AVC-restricted expression and Tbx20-dependent chamber suppression of Tbx2 in vivo. Unexpectedly, we found in transfection and localization studies in vitro that both Tbx20 and mutant isoforms of Tbx20 unable to bind DNA attenuate Bmp/Smad-dependent activation of Tbx2 by binding Smad1 and Smad5 and sequestering them from Smad4.
Conclusions: Our data suggest that Tbx20 directly interferes with Bmp/Smad signaling to suppress Tbx2 expression in the chambers, thereby confining Tbx2 expression to the prospective AVC region.
The complex multichambered heart of vertebrates arises from a simple, rapidly elongating tubular structure through a coordinated program of cellular differentiation and proliferation and tissue morphogenesis. Highly localized processes of further myocardial differentiation and increased proliferation within the growing heart tube mediate the formation of the atrial and ventricular chambers. Regions separating and bordering the developing chambers, the atrioventricular canal (AVC) and the outflow tract (OFT), retain low proliferation rates and slow impulse conduction and resist differentiation in chamber myocardium.1
Functional analyses in the mouse revealed that members of the T-box family of transcription factors participate in myocardial patterning and cardiac compartmentalization2,3 Both Tbx5 and Tbx20 are activated in the early cardiac field by bone morphogenetic protein (Bmp) signaling4 and act as transcriptional activators that cooperate with other conserved cardiac transcription factors including Nkx2.5 and Gata-binding proteins to activate expression of chamber-specific genes such as ANF (Nppa) and connexin40 (Cx40).3,5–7 Mice homozygous mutant for Tbx20 establish a heart tube with a primary myocardial phenotype but fail to undergo looping morphogenesis and to initiate chamber formation.5,8–10 Tbx5 acts independently of Tbx20 and maintains posterior domains of the heart.7
Tbx2 encodes a transcriptional repressor that suppresses differentiation and the chamber-specific gene program.2 Individual loss of Tbx2 results in a locally restricted gain of a chamber myocardial program in the AVC,11,12 whereas forced expression of Tbx2 in the early heart tube leads to an early cardiac developmental arrest and a complete failure of chamber formation.13 Tbx2 competes with activating T-box proteins such as Tbx5 and Tbx20 for binding to conserved T-box–binding elements (TBEs) in promoters of chamber specific genes.6 Regionally restricted induction of Tbx2 by Bmp signaling in the AVC and OFT4,14 may therefore underlie the inhibition of the chamber myocardial gene program and maintenance of the primary myocardial phenotype in these areas.
Notably, expression of Tbx2 is precociously activated in the cardiac crescent and heart tube of Tbx20-deficient embryos and may thus be responsible for the observed block in cardiac chamber differentiation in these embryos.8–10 This suggests a primary role for Tbx20 in repressing Tbx2 for the progression to a multichambered heart. Here, we present genetic experiments in the mouse that further decipher the molecular pathways that underlie localized formation of the chambers and the AVC.
Animal care was in accordance with national and institutional guidelines.
An expanded Methods section is available in the Online Data Supplement at http://circres.ahajournals.org.
A Minimal Regulatory Region for Restriction of Tbx2 Expression to the AVC
Precocious expression of Tbx2 in the cardiac crescent of Tbx20-deficient embryos suggested a primary role for Tbx20 in restricting Tbx2 expression and AVC formation.8–10 To gain insight into the molecular mechanism operated by Tbx20 to confine Tbx2 expression to the developing AVC, we functionally examined the Tbx2 regulatory region (Figure 1). We first tested a 6-kbp genomic fragment previously shown to mediate AVC expression15 and found that enhanced yellow fluorescent protein (Eyfp) reporter gene expression driven by this fragment in transgenic animals (Tbx2[−5.6/+0.316]Eyfp) recapitulated endogenous cardiac expression of Tbx2 in the AVC and OFT (Figure 1A and 1B). This fragment also conferred ectopic expression in Tbx20-deficient hearts (Figure 1C). Deletion analysis showed that a 0.9-kbp fragment, located 2.3 kbp upstream of the transcriptional start site, in combination with a minimal promoter piece of 0.6 kbp (Tbx2[−3.2/−2.3,−0.314/+0.316]Eyfp) recapitulated cardiac expression of Tbx2 in wild-type and upregulation in Tbx20-deficient embryos faithfully (Figure 1A through 1C). The previously identified TBE recognized by Tbx2010 is absent from this construct and, therefore, is neither required for AVC and OFT-specific expression nor Tbx20-mediated repression. Recent reports have pinpointed the relevance of conserved Foxn and additional TBE sites in the zebrafish tbx2b promoter for activation of the gene in the AVC.16 Moreover, β-catenin could regulate Tbx2 by direct binding to a Lef/Tcf site. Genomic Tbx2 fragments deprived of these deeply conserved sites (Tbx2[−3.2/−2.3,0.063/+0.316]Eyfp and Tbx2[LEFmut,−0.314/0.316]Eyfp, respectively) still drove reporter gene expression to the AVC and OFT, making it unlikely that T-box factors, Lef/Tcf proteins, or Fox transcription factors regulate this Tbx2 genomic fragment (Figure 1A and 1B; Online Figure I). Further truncation of the 0.9-kbp region revealed that a 380-bp Tbx2 subfragment (−2.7/−2.3) in conjunction with a minimal promoter (−0.063/+0.316) in the Tbx2(−2.7/−2.3,−0.063/+0.316)Eyfp construct is sufficient to completely recapitulate expression of Tbx2 in the AVC and OFT (Figure 1A and 1B). The 380-bp Tbx2 genomic fragment (−2.7/−2.3) contains a large number of putative Smad-binding sites (SBEs),17 supporting a role of Bmp/Smad signaling in the regulation of cardiac Tbx2 expression (Figure 1D). Taken together, this analysis indicates that a small SBE-containing enhancer in Tbx2 is sufficient to drive AVC expression in vivo.
Regulation of Cardiac Tbx2 Expression by Bmp/Smad Signaling and Tbx20
In in vitro reporter assays, we observed a strong activation of the Tbx2 upstream regulatory region (Tbx2[−5.6/+0.316]Luc) on coexpression of Smad5 and constitutively active Bmp receptor Alk3 (Alk3CA)18 and a reduction to basal activity in the presence of increasing concentrations of Tbx20 (Figure 2A and 2B). Deletion analysis of the Tbx2 promoter fragment revealed the absolute requirement of an 0.9-kbp NheI/AflII fragment (Tbx2[−3.2/−2.3]) for Bmp/Smad-dependent activation of the promoter (Figure 2A; Online Figure II). This fragment on its own mediated only weak activation, whereas inclusion of the adjacent 5′ region in the construct Tbx2(−3.9/−2.3,−0.314/+0.316)Luc gave strong Bmp/Smad-dependent activation (Figure 2A; Online Figure II). We identified several SBEs in the phylogenetically conserved 3′ region of the Tbx2(−3.9/−3.2) fragment (Online Figure I). We assume that these sites normally cooperate with SBEs in the Tbx2(−3.2/−2.3) fragment. Sufficiency of the Tbx2(−3.2/−2.3) fragment to confer AVC/OFT expression of Eyfp in vivo (Figure 1A and 1B) may thus result from multimerization of this fragment by tandem integration of the reporter construct in the genome. In support of this hypothesis, its trimerization in the construct Tbx2([−3.2/−2.3]x3,−0.314/+0.316)Luc strongly increased the Bmp/Smad-dependent activation (Figure 2A; Online Figure II, H).
Removal of previously identified TBEs10 did not affect the repression activity of Tbx20 on the promoter (Figure 2A; Online Figure II, A). This may suggest the presence of cryptic DNA-binding sites for Tbx20. Alternatively, repression by Tbx20 may not be mediated by DNA binding but by protein interaction. To test the latter, we constructed point mutants of Tbx20 that do not exert specific DNA binding anymore (Online Figure III). Unexpectedly, these mutant Tbx20 proteins still repressed transactivation of the Tbx2(−5.6/+0.316) fragment by Bmp/Smad signaling both in the presence of transfected Smad5 (Figure 2B) and in the absence of transfected, ie, in the presence of endogenous Smad5 only (Figure 2C). All deletion constructs of this Tbx2 genomic fragment that were activated by Bmp/Smad signaling to variable degree were repressed by the DNA binding–deficient Tbx20 protein, whereas this Tbx20mut protein had no effect on the constructs that were not activated by Bmp/Smad signaling. Thus, Smad activation is required to observe the DNA-binding independent Tbx20 effect (Online Figure II). Repression achieved by the DNA binding–deficient mutant of Tbx20 was lower than with the wild-type protein suggesting the coexistence of DNA binding–dependent and –independent mechanisms of repression for Tbx20 protein in the regulation of Tbx2 in this assay (Figure 2B and 2C). Together, the in vivo and in vitro analyses of the Tbx2 promoter argue that temporal and spatial confinement of cardiac Tbx2 expression is supported by Tbx20-dependent repression of Bmp/Smad-mediated transcriptional activation.
Tbx20 Inhibits Transcriptional Activation of Bmp/Smad-Dependent Promoters in a DNA Binding–Independent Manner
We wondered whether the repressive effect of Tbx20 on Bmp/Smad-dependent transcriptional activation might be of a more general nature. We tested minimal fragments of Msx2 and Id1 promoters known to be activated by Smad binding in transactivation experiments in NIH3T3 cells (Figure 3A through 3C).19,20 Both Tbx20 wild-type and Tbx20 DNA-binding mutant proteins repressed Bmp/Smad-mediated activation of the Msx2 and Id1 promoter in a dose-dependent manner. As in the case of the Tbx2 promoter, repression by the DNA-binding deficient mutant form of Tbx20 did not reach the level of the wild-type protein, suggesting that the wild-type protein in addition regulates transcription directly. Interestingly, the DNA-binding deficient mutant of Tbx20 exerted very weak repression of Bmp/Smad-mediated activation of the Id1 promoter in the absence of exogenous Smad5 (Figure 3B and 3C), suggesting a role as direct DNA-dependent transcriptional repressor of Tbx20 in this context. Wild-type Tbx20 protein acted as a DNA-dependent transcriptional activator of the Nppa promoter6 that did not respond to Bmp/Smad signaling. The Tbx20 mutant protein did not show a transactivation in this context (Figure 3D). Introduction of a potent transcriptional activator into eukaryotic cells can suppress the transcription of a cointroduced target gene most likely by titration of 1 or more general transcription factors that might be in limiting supply. Because very high doses of Tbx20 protein activated the Nppa promoter to the same degree as lower doses (Figure 3E), we deem it unlikely that the repressive effect of Tbx20 on the Bmp/Smad-dependent promoters that use the same pGL-based reporter plasmids can be explained by such a general squelching mechanism. Together, these experiments suggest that Tbx20 can regulate transcription of target genes in multiple ways: as an inhibitor of Bmp/Smad-mediated transactivation (Tbx2), as a transcriptional activator of Bmp/Smad-independent promoters (Nppa), and most likely also as a direct transcriptional repressor.
Tbx20 Binds to Regulatory Smad Proteins
DNA binding–independent inhibition of Bmp/Smad-mediated transactivation by Tbx20 may rely on physical interaction and/or functional interference with the transcriptional activators Smad5 or Smad1. In vitro pull-down assays with glutathione S-transferase (GST)-Tbx20 fusion proteins showed that Tbx20 directly binds to regulatory Smad5 and the closely related Smad1 proteins but not to co-Smad4 and the inhibitory Smads, Smad6, and Smad7 (Figure 4A). Binding was mediated by the T-box of Tbx20 as shown by GST pull-down assays with fusion constructs of GST and various Tbx20 protein fragments (Figure 4A). Tbx20 binding to Smad1 and Smad5 also occurred in a cellular context as shown by cotransfection/coprecipitation experiments in HeLa cells (Figure 4B and 4C). Tbx20 bound to the phosphorylated form of Smad5 (Figure 4D). Because of the lack of a specific antibody against unphosphorylated Smad1/5/8 we cannot exclude that Tbx20 binds to this form of the protein as well. However, cotransfection experiments of Smad5/Smad1 and Tbx20 expression constructs and detection by immunofluorescence confirmed that Tbx20 predominantly localized to the nucleus and did not mediate cytoplasmic–nuclear shuttling of the unphosphorylated form of Smad5 and Smad1 (Figure 4E and 4F). Together, these data suggest that Tbx20 binds to phospho-Smad1/5 in the nuclear compartment. Smad5 also bound to Tbx2 and Tbx5 in a cellular system, suggesting a more general nature of interaction of regulatory Smad proteins with T-box transcription factors (Online Figure IV). However, Tbx-Smad interaction may not necessarily influence Smad-dependent transcriptional activation because we did not detect an effect of Tbx5 expression on Bmp/Smad-dependent activation of the Tbx2 promoter in reporter assays in vitro (data not shown).
Inhibition of Smad1/5-Smad4 Complex Formation by Tbx20
Because Smad4 is a necessary cofactor for nuclear translocation and transcriptional activation by Smad1/Smad5, we investigated whether Tbx20 binding to Smad1/Smad5 competes with Smad1/5-Smad4 complex formation. These experiments revealed a decrease of coimmunoprecipitated Smad4.HA protein in the presence of increasing amounts of Tbx20 (Figure 5A). In an alternative assay, we transfected expression constructs for Myc-tagged Smad1 and hemagglutinin (HA)-tagged Smad4 into HeLa cells and precipitated Smad1/Smad4 complexes by anti-Myc antibodies. Addition of in vitro translated Tbx20 protein to resuspended immunocomplexes resulted in complete release of Smad4 from the complex (Figure 5B). In a reverse experiment, we observed that addition of in vitro translated Smad4 protein to resuspended immunocomplexes resulted in release of Smad5 from the complex (Figure 5C). Finally, we eased the suppressive effect of DNA binding–deficient Tbx20 protein on Bmp/Smad-mediated activation of the Tbx2 regulatory fragment by increasing the concentration of Smad4 in the cellular system (Figure 5D). Thus, Tbx20 effectively competes with Smad4 for binding to Smad1/Smad5 explaining the DNA-independent inhibition of Bmp/Smad-mediated activation of target promoters including Tbx2.
Tbx20 Is Required for Chamber Formation Independently From Tbx2
Previous analysis has shown that ectopic expression of Tbx2 in the early heart tube leads to arrest of cardiogenesis and chamber differentiation.13 To test the hypothesis that ectopic Tbx2 expression is responsible for this observed arrest in Tbx20-deficient hearts,8–10 we generated embryos with combined deficiencies of Tbx2 and Tbx20 by interbreeding double heterozygous animals. Similar to Tbx20 single mutants, Tbx20lacZ/lacZ;Tbx2Cre/Cre embryos were severely growth retarded and died at embryonic day (E)10.5 because of hemodynamic failure (data not shown). At E9.5, Tbx20 mutant embryos featured a straight and short tubular heart (Figure 6A through 6D). In Tbx20/Tbx2 double mutants, the heart tube appeared morphologically more varied from being straight and short like in the Tbx20 single mutants to being more extended and inflated with some rightwards looping. On histologically stained sagittal sections, both myocardium and endocardium appeared homogenously thin throughout the Tbx20lacZ/lacZ;Tbx2Cre/Cre heart tube. This is in contrast to Tbx20lacZ/lacZ hearts, where a thick layer of cardiac jelly filled the space between myocardium and endocardium. Cardiac jelly production was not associated with endocardial EMT and cushion formation in the Tbx20lacZ/lacZ linear heart tube as revealed by histological inspection (Figure 6E through 6H) as well as expression analysis of the cushion marker Sox9 (Figure 6I through 6L). This suggests that ectopic Tbx2 mediates the induction of cardiac jelly formation in Tbx20-deficient hearts.
Molecular analysis using markers with differential expression along the linear heart tube (Tbx5, Myh7) confirmed that anterior-posterior patterning occurred normally in Tbx20lacZ/lacZ;Tbx2Cre/Cre hearts (Figure 6M through 6T). Chamber myocardium, however, was not formed, as shown by absence of expression of Nppa (Anf) and Cited1 (Figure 6U through 6B′). Bmp2 is expressed in the primary myocardium of the AVC in the wild type.14 In the tubular hearts of single and double mutant embryos, Bmp2 expression is restricted to the posterior part of the primitive ventricle suggesting that a molecular AVC domain has been established in these embryos (Figure 6C′ through 6F′). In summary, loss of Tbx2 does not rescue the Tbx20-deficient heart from developmental arrest at the linear heart tube stage, defining a role of Tbx20 as regulator of cardiac chamber formation independent from its repression of Tbx2.
Our genetic and biochemical analyses suggest that Tbx20 function may couple and localize cardiac compartmentalization into chambers and AVC by suppressing the Bmp/Smad-signaling pathway and Tbx2 activation in the entire heart tube (Figure 7). The future AVC, however, expresses Bmp2 at levels sufficient to overcome the suppressive function of Tbx20 to activate Tbx2.
Tbx20 Restricts Tbx2 to the Prospective AVC by Attenuating Bmp/Smad Signaling
Expression analyses, embryological manipulation in the chick and genetic ablation experiments in the mouse have revealed a range of Bmp-dependent processes during amniote heart development.21 Bmp signaling is broadly activated in the precardiac mesoderm and the cardiac crescent and is required for myocardial differentiation of progenitor cells in the first heart field and for proliferation and recruitment of second heart field cells to the forming OFT.20,22 Ablation of Bmpr1a in early mesoderm resulted in a complete failure to establish cardiomyocytes expressing conserved core cardiogenic factors including Tbx5, Tbx20, Nkx2.5, and Gata family members, implicating them as downstream mediators of early Bmp signaling. After cardiac specification, Bmp signaling is redeployed for specification of the AVC. Conditional deletion of Bmp2 showed that Bmp signaling is required to establish an AV myocardium and to induce endocardial EMT.14 Furthermore, Bmp2 regulates expression of the transcriptional repressor Tbx2 in the AVC.4,14 Our analysis of the regulatory region of the Tbx2 gene identified a genomic fragment that is sufficient to direct AVC expression in vivo. This fragment is rich in SBEs and responsive to Bmp/Smad signaling in vitro, strongly arguing that Tbx2 is a direct target gene of this pathway in the heart.
Expression of Tbx2 in the early heart tube prevents chamber formation,13 illuminating the necessity to prevent premature activation of Tbx2 by the first wave of Bmp/Smad signaling in the cardiac crescent, the heart tube and the prospective chambers (Figure 7). In Tbx20 mutant hearts, Tbx2 is induced in the developing cardiac crescent and throughout the linear heart tube,8–10 demonstrating that temporal and spatial restriction of Tbx2 to the developing AVC is not achieved by positive regulatory inputs, but by Tbx20-mediated inhibition of broad activation in regions outside the AVC. Absence of TBE sites from a minimal genomic Tbx2 regulatory region, sufficient to recapitulate cardiac Tbx2 expression in wild-type and Tbx20 mutant embryos, strongly argues against Tbx20 acting as a transcriptional repressor in this context, as indicated previously by chromatin immunoprecipitation experiments.10 It is formally possible that Tbx20 acts indirectly as a DNA-dependent transcriptional repressor of an inducer of Tbx2, including Bmp2. However, confinement and low-level expression of Bmp2 in the posterior region of the primitive heart tube in Tbx20-deficient embryos does not support this model. Alternatively, Tbx20 may transcriptionally activate a repressor of Tbx2 expression outside the AVC, compatible with its known biochemical function as a transactivator as shown for the Nppa promoter. However, again, no evidence for the existence of such a repressor has been obtained from our in vivo analysis of the Tbx2 promoter.
We unexpectedly found that Tbx20 binds to the phosphorylated forms of Smad1 and -5 in the nucleus and sequesters them from binding to the co-Smad4, abolishing the formation of transcriptionally active P-Smad1/5-Smad4 complexes. Hence, transcriptional modulation of target gene expression by T-box transcription factors may not only rely on the presence of TBE sites. Direct binding and sequestration of transcriptional regulators by the conserved T-domain suggests another level of complexity of transcriptional regulation and establishes the T-box as a versatile interface both for DNA and protein interaction. Indeed, previous evidence has suggested that the T-box represents a common interface for binding to numerous classes of DNA-binding domains.2,3 Notably, binding of T-box proteins to Smad proteins as shown in this study is not without precedence. In Xenopus, Brachyury, the prototypical member of the T-box family, interacts with Smad1 to activate expression of the transcriptional repressor Xom. In this case, Smad binding is conferred by an N-terminal region of XBra and not the T-box.23 In addition, phosphorylated Smad1 interacts with Tpit in the corticotroph AtT-20 cell line to prevent the transcriptional activation of the POMC gene by a Tpit/Pitx1 complex, suggesting a third mode of target gene regulation by T-box/Smad interaction.24 Finally, and similar to the findings of this study, Tbx1 binding to Smad1 was found to interfere with Smad1-Smad4 complex formation and Smad signaling.25 Thus, Tbx-Smad interaction seems to be widespread but may have different context-dependent functional implications on gene regulation.
Tbx20-mediated sequestration of regulatory Smads may be one of several mechanisms that synergize to shut off Bmp/Smad signaling after cardiac specification of the lateral plate mesoderm. In Nkx2.5 mutant hearts, Bmp/Smad signaling is dramatically augmented and expanded, suggesting that Nkx2.5 represses Bmp/Smad signaling.26 Smad6, which is expressed in the cardiac crescent, stably binds to activated type I receptors and competes with co-Smad4 for receptor activation.27 Moreover, similar and likely in addition to Tbx20, Smad6 specifically competes with Smad4 for binding to receptor-activated Smad1, yielding an inactive Smad1/Smad6 complex.28 Because Tbx20, Nkx2.5, and Smad6 are targets of Bmp signaling in the cardiac crescent, they may be part of a concerted feed-back inhibition loop.
Tbx20 and Tbx2 Regulate Chamber Versus AVC Development
Previous analyses by a number of research groups revealed the crucial role of Tbx20 in cardiac chamber formation in vertebrates.5,8–10 Tbx2 expression, normally restricted to the AVC,4,6 was expanded into the entire heart tube of Tbx20-deficient embryos arguing that chambers have been lost at the expense of an AVC. Because ectopic expression of Tbx2 in the prechamber heart tube also inhibited chamber formation, similar to the phenotype in Tbx20 mutants,13 it was hypothesized that derepression of Tbx2 fully explains the cardiac phenotype in Tbx20-deficient embryos. Our analysis of Tbx20/Tbx2 double mutant embryos revealed that loss of Tbx2 does not rescue the failure to form chambers. We assume that Tbx20 similar to Tbx5 promotes chamber formation as a DNA-dependent transcriptional activator of chamber-specific genes such as Nppa. Ectopic expression of the strong transcriptional repressor Tbx2 in a wild-type embryo is likely to block this process by its ability to compete with Tbx20 (and Tbx5) binding to conserved T-binding sites in promoters of these genes and prevent their activation. Hence, ectopic expression of Tbx2 in the Tbx20 mutant heart does not further enhance the loss of Tbx20 phenotype. On the other hand, Tbx2 may require Tbx20 to suppress chamber differentiation, providing an alternative explanation for the failure to rescue the Tbx20 mutant phenotype in Tbx2/Tbx20 double mutants.
We thank Corrie de Gier-de Vries for technical contributions and advice; M. van Roon of the Academic Medical Center GGM facility for generating mice; and Michael Oelgeschläger, Kai Jiao, Peter ten Dijke, Sangita Chauhan, and Nisar Malek for reagents.
Sources of Funding
This work was supported by grants from The Netherlands Organization for Scientific Research (Vidi grant 864.05.006 to V.M.C.); the European Community’s Sixth Framework Programme contract (“HeartRepair” LSHM-CT-2005-018630 to V.M.C., A.F.M.M., and A.K.); and the German Research Foundation for the Cluster of Excellence REBIRTH (from Regenerative Biology of Reconstructive Therapy) at Hannover Medical School (A.K.).
Moorman AF, Christoffels VM. Cardiac chamber formation: development, genes, and evolution. Physiol Rev. 2003; 83: 1223–1267.
Stennard FA, Harvey RP. T-box transcription factors and their roles in regulatory hierarchies in the developing heart. Development. 2005; 132: 4897–4910.
Takeuchi JK, Mileikovskaia M, Koshiba-Takeuchi K, Heidt AB, Mori AD, Arruda EP, Gertsenstein M, Georges R, Davidson L, Mo R, Hui CC, Henkelman RM, Nemer M, Black BL, Nagy A, Bruneau BG. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and motoneuron development. Development. 2005; 132: 2463–2474.
Habets PE, Moorman AF, Clout DE, van Roon MA, Lingbeek M, van Lohuizen M, Campione M, Christoffels VM. Cooperative action of Tbx2 and Nkx2.5 inhibits ANF expression in the atrioventricular canal: implications for cardiac chamber formation. Genes Dev. 2002; 16: 1234–1246.
Stennard FA, Costa MW, Lai D, Biben C, Furtado MB, Solloway MJ, McCulley DJ, Leimena C, Preis JI, Dunwoodie SL, Elliott DE, Prall OW, Black BL, Fatkin D, Harvey RP. Murine T-box transcription factor Tbx20 acts as a repressor during heart development, and is essential for adult heart integrity, function and adaptation. Development. 2005; 132: 2451–2462.
Singh MK, Christoffels VM, Dias JM, Trowe MO, Petry M, Schuster-Gossler K, Burger A, Ericson J, Kispert A. Tbx20 is essential for cardiac chamber differentiation and repression of Tbx2. Development. 2005; 132: 2697–2707.
Cai CL, Zhou W, Yang L, Bu L, Qyang Y, Zhang X, Li X, Rosenfeld MG, Chen J, Evans S. T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis. Development. 2005; 132: 2475–2487.
Aanhaanen WT, Brons JF, Dominguez JN, Rana MS, Norden J, Airik R, Wakker V, de Gier-de Vries C, Brown NA, Kispert A, Moorman AF, Christoffels VM. The Tbx2+ primary myocardium of the atrioventricular canal forms the atrioventricular node and the base of the left ventricle. Circ Res. 2009; 104: 1267–1274.
Harrelson Z, Kelly RG, Goldin SN, Gibson-Brown JJ, Bollag RJ, Silver LM, Papaioannou VE. Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development. 2004; 131: 5041–5052.
Ma L, Lu MF, Schwartz RJ, Martin JF. Bmp2 is essential for cardiac cushion epithelial-mesenchymal transition and myocardial patterning. Development. 2005; 132: 5601–5611.
Kokubo H, Tomita-Miyagawa S, Hamada Y, Saga Y. Hesr1 and Hesr2 regulate atrioventricular boundary formation in the developing heart through the repression of Tbx2. Development. 2007; 134: 747–755.
Chi NC, Shaw RM, De Val S, Kang G, Jan LY, Black BL, Stainier DY. Foxn4 directly regulates tbx2b expression and atrioventricular canal formation. Genes Dev. 2008; 22: 734–739.
Korchynskyi O, ten Dijke P. Identification and functional characterization of distinct critically important bone morphogenetic protein-specific response elements in the Id1 promoter. J Biol Chem. 2002; 277: 4883–4891.
Brugger SM, Merrill AE, Torres-Vazquez J, Wu N, Ting MC, Cho JY, Dobias SL, Yi SE, Lyons K, Bell JR, Arora K, Warrior R, Maxson R. A phylogenetically conserved cis-regulatory module in the Msx2 promoter is sufficient for BMP-dependent transcription in murine and Drosophila embryos. Development. 2004; 131: 5153–5165.
Monteiro RM, de Sousa Lopes SM, Korchynskyi O, ten Dijke P, Mummery CL. Spatio-temporal activation of Smad1 and Smad5 in vivo: monitoring transcriptional activity of Smad proteins. J Cell Sci. 2004; 117: 4653–4663.
van Wijk B, Moorman AF, van den Hoff MJ. Role of bone morphogenetic proteins in cardiac differentiation. Cardiovasc Res. 2007; 74: 244–255.
Klaus A, Saga Y, Taketo MM, Tzahor E, Birchmeier W. Distinct roles of Wnt/beta-catenin and Bmp signaling during early cardiogenesis. Proc Nat Acad Sci U S A. 2007; 104: 18531–18536.
Prall OW, Menon MK, Solloway MJ, Watanabe Y, Zaffran S, Bajolle F, Biben C, McBride JJ, Robertson BR, Chaulet H, Stennard FA, Wise N, Schaft D, Wolstein O, Furtado MB, Shiratori H, Chien KR, Hamada H, Black BL, Saga Y, Robertson EJ, Buckingham ME, Harvey RP. An Nkx2–5/Bmp2/Smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell. 2007; 128: 947–959.
Hata A, Lagna G, Massague J, Hemmati-Brivanlou A. Smad6 inhibits BMP/Smad1 signaling by specifically competing with the Smad4 tumor suppressor. Genes Dev. 1998; 12: 186–197.