Morphogenesis of the right ventricle requires myocardial expression of Gata4
http://www.100md.com
《临床调查学报》
1Cardiovascular Division and
2Center for Matrix Biology, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, USA.
3Department of Cardiology, Children’s Hospital Boston, Boston, Massachusetts, USA.
4Department of Pathology, Harvard Medical School, Boston, Massachusetts, USA.
5Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas, USA.
6Department of Pediatrics, Harvard Medical School, Boston, Massachusetts, USA.
Abstract
Mutations in developmental regulatory genes have been found to be responsible for some cases of congenital heart defects. One such regulatory gene is Gata4, a zinc finger transcription factor. In order to circumvent the early embryonic lethality of Gata4-null embryos and to investigate the role of myocardial Gata4 expression in cardiac development, we used Cre/loxP technology to conditionally delete Gata4 in the myocardium of mice at an early and a late time point in cardiac morphogenesis. Early deletion of Gata4 by Nkx2-5Cre resulted in hearts with striking myocardial thinning, absence of mesenchymal cells within the endocardial cushions, and selective hypoplasia of the RV. RV hypoplasia was associated with downregulation of Hand2, a transcription factor previously shown to regulate formation of the RV. Cardiomyocyte proliferation was reduced, with a greater degree of reduction in the RV than in the LV. Late deletion of Gata4 by Cre recombinase driven by the myosin heavy chain promoter did not selectively affect RV development or generation of endocardial cushion mesenchyme but did result in marked myocardial thinning with decreased cardiomyocyte proliferation, as well as double-outlet RV. Our results demonstrate a general role of myocardial Gata4 in regulating cardiomyocyte proliferation and a specific, stage-dependent role in regulating the morphogenesis of the RV and the atrioventricular canal.
Introduction
Congenital heart defects are the most common developmental anomaly in newborns (1). Formation of the normal heart involves the transformation of a linear heart tube into the mature 4-chambered organ. The linear heart tube grows rapidly by the proliferation of cardiomyocytes and by the addition of new cardiomyocytes to the arterial pole of the heart. These new cardiomyocytes, derived from splanchnic mesoderm adjoining the arterial pole of the heart (termed the secondary heart field), give rise to the outflow tract (OFT) and also contribute to the development of the RV (2-4). As the heart tube lengthens, it loops to the right, and the resulting structure is septated by growth and maturation of the endocardial cushions, structures derived from the endothelial lining of the heart by an epithelial-to-mesenchymal transition (EMT) (5).
Genetic dissection of heart development has shown that a hierarchy of transcription factors regulates these morphogenic events (6). Relatively independent genetic programs control the development of each heart chamber, which is consistent with the frequent occurrence of selective chamber hypoplasia in severe forms of congenital heart disease (6). For example, during mouse heart development, the transcription factor Hand2 (dHAND) is predominantly expressed in the RV, while the related transcription factor Hand1 (eHAND) is predominantly expressed in the LV (7). Deletion of Hand2 results in severe hypoplasia of the RV segment (8).
It has been found that some cases of congenital heart defects were caused by mutations in developmental regulatory genes. Mutation of the cardiac transcription factor Nkx2-5 is associated with familial congenital heart disease and has been associated with up to 4% of cases of tetralogy of Fallot and secundum atrial septal defect (9, 10). Haploinsufficiency for the transcription factor Tbx1 in patients with 22q11 chromosomal microdeletions causes OFT anomalies (11, 12), while mutation of the transcription factor Tbx5 is associated with congenital heart disease in Holt-Oram syndrome and in nonsyndromic congenital heart disease (13).
Recently, mutation of the transcription factor Gata4 was linked to congenital heart disease characterized by atrial or ventricular septal defects (14, 15). In the heart, Gata4 is expressed in cardiomyocytes and their mesodermal precursors, as well as in the endocardium and the epicardium (Figure 1, E, F, I, and J, and Figure 5, A, C, and E). Gata4 binds to and regulates expression of a number of myocardium-expressed genes (16, 17). Embryos lacking Gata4 arrested in development and died due to a defect in the visceral endoderm that caused aberrant ventral morphogenesis and cardiac bifida (18-20). Using alleles of Gata4 that cause partial loss of function, we and others have previously demonstrated that Gata4 is required for later heart development (21, 22). However, these experiments were not able to address the tissue-restricted requirements for Gata4, particularly during early stages of heart development. To evaluate the function of Gata4 within the myocardial compartment, we crossed floxed Gata4 mice with mouse lines in which expression of Cre recombinase is initiated within the myocardium at an early and a late time point in cardiac development. We found that myocardial expression of Gata4 was required in a stage-dependent manner for proliferation of cardiomyocytes, formation of the endocardial cushions, development of the RV, and septation of the OFT.
Results
Myocardial-restricted deletion of Gata4 early in cardiac development.
Previously, in order to study the temporal and spatial requirements for Gata4 in heart development, we generated a floxed Gata4 allele (Gata4flox) in which exon 2, containing the start codon and 46% of the coding sequence, is flanked by loxP sites (22). Expression of Cre recombinase resulted in deletion of exon 2, yielding a null allele (22). In order to investigate the cardiomyocyte-restricted function of Gata4 in cardiac development, we used Cre/loxP technology to conditionally ablate Gata4 in cardiomyocytes at 2 distinct time points in cardiac development. We achieved early cardiomyocyte-restricted deletion of Gata4 using Nkx2-5Cre, in which Cre recombinase expression is driven by the endogenous Nkx2-5 locus (23). We previously reported that Nkx2-5Cre initiated Cre expression around E7.5 in the cardiac progenitor pool and that robust Cre-mediated recombination occurred in cardiomyocytes by E9.5 (23). In order to further delineate the temporal and spatial pattern of Cre activity, we crossed Nkx2-5Cre mice with R26RstoplacZ mice, in which expression of lacZ requires Cre-mediated excision of a stop cassette (24). Consistent with our prior results (23), Nkx2-5Cre activated lacZ expression in most cardiomyocytes of E9.5 embryos (arrows, Figure 1, B and C). Cre-mediated expression of lacZ did not occur in the endocardium (arrowheads, Figure 1, B and C) or in the proepicardial organ (Figure 1A). Outside the heart, lacZ activity was detected in the pharyngeal endoderm and in a patchy distribution in branchial arch epithelium (Figure 1A and Figure 4, A–C).
Next, we obtained Gata4flox/flox;Nkx2-5WT/Cre (G4NK) embryos from timed matings and we confirmed loss of the Gata4 transcript in G4NK hearts by quantitative RT-PCR (qRT-PCR) and in situ hybridization. Gata4 expression was reduced 90% in G4NK hearts compared with that in controls (P < 0.05; n = 3; Figure 1D), while the expression of the closely related gene Gata6 was unaffected (Figure 1D). In situ hybridization showed that Gata4 expression in the myocardium was not detectable above background by E9.5 in G4NK embryos (compare red arrowheads in Figure 1H and Figure 1F). Surprisingly, we found that Gata4 expression in the endocardium was also reduced to background levels in G4NK embryos (compare yellow arrowheads in Figure 1H and Figure 1F). Gata5 and Gata6 expression was not altered in G4NK embryos (Supplemental Figure 3; supplemental material available online with this article; doi:10.1172/JCI23769DS1).
To further confirm the loss of myocardial and endocardial Gata4 expression, we examined Gata4 protein expression in G4NK hearts by immunofluorescence staining (Figure 1, I–L). In control embryos, robust Gata4 staining (red stain) was localized to the nucleus (blue stain) of cardiomyocytes and endocardial cells (Figure 1, I and J). However, in G4NK embryos, Gata4 signal was strongly decreased in both cardiomyocytes and endocardial cells (Figure 1, K and L). As the antibody used recognizes a peptide in the C terminus of Gata4, loss of immunoreactivity in G4NK hearts suggests that excision of the floxed region of Gata4 (exon 2, encoding the N-terminal portion of Gata4 including the start codon) resulted in loss of Gata4 protein. Loss of endocardial Gata4 expression was not due to a generalized defect in endothelial differentiation into endocardium, as the endocardial-specific marker Nfatc1 was expressed normally in G4NK hearts (Supplemental Figure 1). Because Cre activity was absent in the endocardium of Nkx2-5Cre hearts (Figure 1, B and C), loss of endocardial Gata4 expression in G4NK hearts suggests that a paracrine signal derived from the myocardium regulates endocardial Gata4 expression.
Early cardiomyocyte-restricted ablation of Gata4 in G4NK embryos resulted in lethality by E11.5 (Figure 2A). However, overall development of G4NK embryos at E9.5, as assessed by the number of somites, was not impaired (control, 20.8 ± 2.8 somites, n = 15; mutant, 21.7 ± 0.5 somites, n = 7). By E10.5, G4NK embryos were mildly delayed in overall development (control, 31.6 ± 2.2 somites, n = 19; mutant, 28.5 ± 2.2 somites, n = 6; P < 0.05) and had large pericardial effusions, a sign of heart failure (compare Figure 2B and Figure 2C).
At E9.5, most G4NK embryos could be recognized on whole-mount examination by characteristic cardiac malformations. G4NK embryos had normally positioned atria and intact atrioventricular grooves (arrowheads, Figure 2, D–F). Fifteen of 21 G4NK embryos systematically examined in whole mount displayed a single predominant ventricular chamber that connected to an OFT located toward the rostral side of the chamber (compare arrows in Figure 2G and Figure 2I). The remaining 6 of 21 G4NK embryos had normal- to mildly hypoplastic–appearing RVs that connected to normally positioned OFTs (Figure 2, H and O). On histological sections, all G4NK embryos displayed marked myocardial hypoplasia, affecting both the compact and trabecular myocardium (Figure 2, J–M). The atrioventricular and OFT endocardial cushions were small and contained few mesenchymal cells (arrowheads, Figure 2, J, K, N, and O), which indicates a defect in the EMT of endocardial cells that normally generates cushion mesenchyme. This defect was due at least in part to the aforementioned downregulation of Gata4 in the endocardium, as we have found that expression of Gata4 in the endocardium is required for EMT (J. Rivera-Feliciano et al., manuscript submitted for publication).
Myocardial Gata4 regulates formation of the RV.
Early myocyte-restricted deletion of Gata4 in G4NK embryos resulted in the formation of a single predominant ventricular chamber in the majority of embryos (Figure 2, I and P). In order to determine the molecular identity of the chamber, we examined the expression of chamber-specific markers. In normal embryos, at E9.5 the atrial chamber expressed myosin light chain 2a (MLC2a) but not MLC2v, while the ventricular chambers expressed both MLC2a and MLC2v. This expression pattern was maintained in G4NK hearts (Supplemental Figure 2), which indicates that specification of atrial and ventricular chambers occurs normally in G4NK hearts. The genes Hand1 and Tbx5 are normally expressed at higher levels in the LV than in the RV (8, 25). In G4NK hearts, the expression pattern of Hand1 and Tbx5 was not altered, with the predominant ventricular chamber preferentially expressing these markers (Figure 3, I and J, and data not shown). These data indicate that the predominant ventricular chamber in G4NK hearts is the LV. The data also demonstrate that Hand1 is not directly regulated by Gata4.
To determine whether a hypoplastic RV was incorporated into the tubular structure connecting the predominant ventricular chamber to the aortic sac in mutant embryos, we examined the expression of the OFT marker tenascin C (TNC) by in situ hybridization. In control embryos, the entire length of the OFT from the morphological RV to the aortic sac expressed TNC (Supplemental Figure 2) (26). In contrast, in G4NK embryos with a single predominant chamber, TNC was expressed in the distal end of the OFT. However, the proximal end of the outflow tube near its junction with the predominant ventricular chamber was TNC negative (Supplemental Figure 2) and likely represented a severely hypoplastic RV.
The RV hypoplasia found in G4NK hearts was highly reminiscent of the phenotype of Hand2-null embryos, which failed to form the RV (8). Hand2 is expressed in the limb buds, the branchial arches, and the myocardium of the OFT and the RV (8). By both whole-mount and section in situ hybridization, we found that Hand2 transcript levels were decreased in G4NK hearts compared with those in controls (Figure 3, A–F). The decrease in Hand2 transcript level was not due to nonspecific RNA degradation, as adjoining sections from the same embryo yielded robust signals to other probes (e.g., Hand1, Figure 3J). Decreased Hand2 levels were also not simply a consequence of hypoplasia of the RV, as Hand2 expression in the OFT was also strongly downregulated (arrows, Figure 3, E and F). Hand2 remained robustly expressed in the branchial arches of mutant embryos (arrowheads, Figure 3, A–F). These data demonstrate that normal expression of Hand2 requires Gata4 and suggest that abnormal regulation of Hand2 in G4NK mutants contributes at least in part to the selective requirement of Gata4 for RV development.
The RV and OFT are derived at least in part from cardiomyocytes generated in a secondary heart field located at the junction of the splanchnic mesoderm with the caudal surface of the OFT (2-4). In chicken embryos, this secondary heart field contains progenitor cells that express both Nkx2-5 and Gata4 (2). One hypothesis to account for the RV hypoplasia seen in G4NK mutants is that Gata4 inactivation in these progenitors impairs formation or migration of cardiomyocytes destined to contribute to the RV. To address this hypothesis, we assessed the timing and location within the secondary heart field of Gata4, -5, and -6 expression and of Cre activity driven by Nkx2-5Cre. By E9.5, when RV hypoplasia became apparent, few cells within this region had undergone recombination to express lacZ (Figure 4, A and B). At E10.5, we found that Nkx2-5Cre activation of the R26RstoplacZ reporter within the splanchnic mesoderm was more robust (Figure 4C), as has been previously reported (27). More importantly, we found that Gata4 was not expressed in splanchnic mesoderm at its junction with the OFT in wild-type embryos (arrowheads, Figure 4, E and I), while Gata5 and Gata6 were expressed in this region (arrowheads, Figure 4, F, G, J, and K). These data indicate that Gata4 inactivation by Nkx2-5Cre in the secondary heart field is not responsible for hypoplasia of the RV.
Myocardial-restricted deletion of Gata4 late in cardiac development.
In order to study the function of myocardial Gata4 expression at a late time in heart development, we used transgenic mice in which Cre recombinase expression was driven by the myosin heavy chain (MHC) promoter (MHCCre). Prior characterization of this transgene demonstrated that MHCCre catalyzed recombination only in the myocardium, starting at E9.5 and occurring in nearly all cardiomyocytes by E11.5 (28). We isolated Gata4flox/flox;MHCCre+ (G4MC) embryos at E12.5 and examined Gata4 expression by in situ hybridization and immunohistochemistry. In control embryos, Gata4 expression was readily detected in myocardium (white arrowheads, Figure 5, A, C, and E), endocardium (yellow arrowheads, Figure 5, A and E), endocardial cushions (asterisk, Figure 5A), and epicardium (blue arrowheads, Figure 5, C and E). In G4MC embryos, myocardial Gata4 expression was strongly reduced in both compact and trabecular myocardium (compare Figure 5, D and F with Figure 5, C and E), while Gata4 continued to be expressed in epicardium and endocardium (blue and yellow arrowheads, Figure 5, B, D, and F). Expression of Gata5 and Gata6 were not altered in G4MC embryos (Supplemental Figure 4).
Late cardiomyocyte-restricted deletion of Gata4 by MHCCre resulted in lethality by E14.5 (Figure 2A). Mutant embryos had marked myocardial hypoplasia and double-outlet RV (Figure 5, G–L). While early excision of Gata4 by Nkx2-5Cre resulted in the formation of a single predominant ventricular chamber in the majority of embryos, 2 well-formed ventricles were present after late Gata4 inactivation by MHCCre. In G4MC hearts, as in control hearts, the atrioventricular cushion divided the atrioventricular canal into separate RV and LV inflow tracts (asterisk, Figure 5, G and J). This result indicates that, unlike early myocardial Gata4 deletion, late deletion does not impair initial steps of atrioventricular valve morphogenesis, including generation of cushion mesenchyme and growth and fusion of the atrioventricular endocardial cushions.
Myocardial Gata4 regulates cardiomyocyte proliferation.
Hypoplasia of the myocardium was a prominent feature after both early and late myocyte-restricted Gata4 deletion. This phenotype could have been due to increased cardiomyocyte apoptosis or decreased cardiomyocyte proliferation. TUNEL staining of G4NK and G4MC embryos at E9.0 and E12.5 showed that cardiomyocyte apoptosis did not increase compared with that of control littermates (data not shown). To determine whether myocardial Gata4 inactivation decreases cardiomyocyte proliferation, we measured cardiomyocyte proliferation by BrdU labeling. After both early and late Gata4 deletion, the percentage of myocytes labeled by BrdU was significantly decreased (P < 0.05; Figure 6, A–H, and Table 1). Similar results were obtained in G4NK embryos when cardiomyocyte proliferation was measured by staining for the M phase–specific marker phosphohistone H3 (data not shown). Early Gata4 inactivation resulted in selective RV hypoplasia (Figure 2), while late Gata4 inactivation did not (Figure 5). Consistent with these phenotypes, cardiomyocyte proliferation was more severely reduced in the RV than in the LV after early Gata4 inactivation (Figure 6, A–F), but the reductions did not differ between RV and LV after late Gata4 inactivation (Figure 6, G and H). These data suggest that Gata4 is required in a cell-autonomous manner for normal proliferation of cardiomyocytes and that reduced cardiomyocyte proliferation might contribute to the selective RV hypoplasia seen in G4NK hearts.
In order to identify transcriptional targets of Gata4 that may mediate the effect of Gata4 on cardiomyocyte proliferation, we asked whether expression of myocardial genes known to be important for myocardial growth was perturbed by Gata4 deletion. cyclin D1–D3 regulate the core cell-cycle machinery in response to mitogenic stimulation (29), and in other systems Gata4 regulates proliferation by activating cyclin D2 expression (30). Bmp10 and Hop are both required for normal myocardial growth, and expression of both is regulated by Nkx2-5, which can coordinate with Gata4 to regulate gene transcription (31, 32). qRT-PCR analysis of RNA isolated from G4MC and Gata4flox/flox control hearts showed no difference in the expression of Bmp10 (Figure 6I). Expression of Hop tended to be reduced in G4MC hearts, but this did not reach statistical significance (Figure 6I). Unexpectedly, cyclin D1–D3 were not downregulated in G4MC hearts; rather, each was subtly but significantly upregulated (P < 0.05; Figure 6I). This indicates that Gata4 does not directly regulate cardiomyocyte proliferation through cyclin D1–D3. Instead, cyclin D1–D3 upregulation may be a compensatory response to altered expression of other components of the cell-cycle machinery following Gata4 deletion.
Myocardial gene expression downstream of Gata4.
Gata4 has been implicated in the expression of a large number of myocardial genes, including the transcription factors Mef2c, Carp, and Nkx2-5, each of which is essential for normal cardiac morphogenesis (33-37). To determine whether abnormalities of cardiac morphogenesis seen in G4NK or G4MC mutants could be attributed to altered transcription of these genes in vivo, we measured transcript levels by qRT-PCR on RNA isolated from control and mutant hearts (Figure 7A). We found that expression of Mef2c and Carp were not significantly altered in mutant hearts after early or late myocardial Gata4 deletion (Figure 7A). Nkx2-5 expression was reduced by 50% in G4NK hearts, but this decrease was expected due to haploinsufficiency for Nkx2-5. After late myocyte-restricted Gata4 deletion, Nkx2-5 expression was upregulated 1.7-fold in G4NK hearts compared with that of controls (P < 0.05; Figure 7A). We previously found a similar degree of Nkx2-5 upregulation in E12.5 hearts with a germline mutation that reduced Gata4 expression by 70% (22). This might reflect upregulation secondary to heart failure, since Nkx2-5 has previously been shown to be upregulated as a result of increased wall stress (38).
Gata4 has also been implicated in the expression of a number of cardiomyocyte structural genes. We measured the expression of 4 well-characterized Gata4 targets in G4MC hearts compared with that in controls. Surprisingly, we found that levels of atrial natriuretic factor (ANF), brain natriuretic peptide (BNP), and MHC expression were unaltered in G4MC hearts (Figure 7B). MHC expression was downregulated 2-fold after late myocyte-restricted Gata4 deletion (P < 0.05; Figure 7B), which is consistent with the prior finding that GATA binding sites in the proximal MHC promoter are required for promoter activity in vitro (39).
Analysis of the ANF promoter in transgenic Xenopus embryos suggests that GATA binding sites in the ANF promoter are crucial for spatially restricting ANF expression from the OFT, rather than for determining the level of expression (40). To determine whether Gata4 is required to similarly restrict the spatial distribution of ANF expression in mouse embryos, we used in situ hybridization to examine ANF expression after cardiomyocyte-restricted inactivation of Gata4. ANF is normally expressed in atrial and ventricular chamber myocardium, but not in the atrioventricular canal or OFT myocardium (41). In G4NK hearts, ANF was expressed at comparable levels in atrial and ventricular chamber myocardium. However, in 2 out of 6 G4NK embryos, ANF expression extended ectopically, well into the myocardium connecting the predominant ventricular chamber to the OFT (Figure 7C). Thus, in vivo Gata4 is necessary to consistently restrict the ANF expression domain in mice, which is consistent with the requirement for GATA sites in the ANF promoter to restrict the ANF expression domain in Xenopus (40).
Discussion
To determine the myocyte-restricted function of Gata4 in heart development, we characterized embryos in which Gata4 was inactivated at an early and a late time point in heart development. Our results demonstrate distinct temporal requirements for myocardial expression of Gata4 in the regulation of cardiac morphogenesis. Early myocardial expression of Gata4 was necessary for normal cardiac expression of Hand2 and for normal development of the RV. Early expression of Gata4 in myocardium was also necessary to support endocardial Gata4 expression and endocardial EMT. Late loss of myocardial Gata4 expression did not selectively affect RV development but did result in abnormal OFT septation, manifested as double-outlet RV. Both early and late in cardiac development, normal cardiomyocyte proliferation required expression of Gata4.
Recently, tetraploid aggregation of Gata4-null embryonic stem cells was used to bypass the early developmental arrest of Gata4-null embryos (42). As in our embryos with cardiomyocyte-restricted Gata4 deletion, tetraploid-rescued null embryos formed a thin, poorly trabeculated heart tube. However, early cardiomyocyte-restricted Gata4 inactivation resulted in a milder phenotype than in tetraploid-rescued Gata4-null embryos, since the latter had incomplete segmentation and looping of the heart tube. This indicates that they arrested earlier in heart development than did our embryos with early myocyte-restricted Gata4 inactivation (G4NK). Tetraploid-rescued Gata4-null embryos lacked a proepicardial organ. This, in conjunction with the relatively unperturbed expression of myocardial genes, led the authors to conclude that disruption of myocardial development is non–cell autonomous and perhaps due to loss of signals from the proepicardial organ (42). However, our data directly demonstrate an important cardiomyocyte-autonomous function for Gata4 in cardiomyocyte proliferation and RV formation. The G4NK phenotype is milder than the tetraploid-rescued Gata4-null phenotype, which supports additional roles for Gata4 expression in non-myocyte lineages to regulate cardiac morphogenesis. Alternatively, the difference in phenotype may reflect a later deletion of Gata4 in the G4NK hearts.
Because G4NK embryos are haploinsufficient for Nkx2-5, it is possible that reduced Nkx2-5 dosage also contributed to the phenotype observed in G4NK embryos. However, this is unlikely for several reasons. First, Nkx2-5 is not a direct regulator of RV specification or Hand2 expression (43). Second, tetraploid-rescued Gata4-null hearts (42) had abnormalities similar to, but more severe than, G4NK hearts, which indicates that these phenotypes also occur in Gata4-null embryos that have a normal dose of Nkx2-5. Third, while Nkx2-5 haploinsufficiency is associated with atrial septal defects and conduction system abnormalities (44, 45), it has not been reported to result in an observable phenotype at this early stage of heart development. Finally, the myocardial hypoplasia of G4NK embryos was also seen in G4MC embryos, which expressed slightly elevated levels of Nkx2-5. Nevertheless, we cannot exclude the possibility that Nkx2-5 haploinsufficiency contributed to the G4NK phenotype and to the differences between the phenotype of G4NK and G4MC hearts.
Gata4 regulates RV development.
Severe hypoplasia of the RV occurred in most mutant embryos after early cardiomyocyte-restricted deletion, which indicates that Gata4 plays an essential role in RV morphogenesis. Results of previous work have suggested that the OFT and at least a portion of the RV are derived from progenitors located within the secondary heart field (2-4). In chicks, cells within the secondary heart field express Gata4 and Nkx2-5 (2), and in mice, a secondary heart field enhancer has been shown to be dependent upon Gata activity (33). However, we found that in mice, Gata4 was not expressed within the secondary heart field, whereas Gata5 and Gata6 were robustly expressed in this region, which suggests that the Gata isoforms active in the splanchnic mesoderm are species specific. In addition, the Nkx2-5Cre–mediated excision in the secondary heart field occurred in a minority of cells at E9.5 (Figure 4, A and B), when hypoplasia of the RV was already apparent. These data indicate that hypoplasia of the RV in G4NK mutant embryos was not due to Gata4 inactivation in RV precursors within the secondary heart field. Inactivation of Gata4 within cells originating from the secondary heart field might occur after they have migrated into the heart tube and differentiated into cardiomyocytes; thus, our data are consistent with an important role of the secondary heart field in the development of the RV.
We found that RV hypoplasia after early myocyte-restricted Gata4 inactivation was associated with a decrease in cardiomyocyte proliferation that was more severe in the RV than in the LV. At E9.5, RV cardiomyocytes may be more dependent upon Gata4 for normal rates of proliferation than are LV cardiomyocytes. Alternatively, Gata4 inactivation in RV cardiomyocyte precursors originating from the secondary heart field might impair their migration, expansion, or differentiation, which ultimately might be manifested as a decrease in cardiomyocyte proliferation. Distinguishing these models will require mapping the fate of RV progenitors in the G4NK mutant background.
The failure of RV development after early cardiomyocyte-restricted inactivation of Gata4 was associated with downregulation of Hand2 in the RV and the OFT. Hand2 itself has been shown to be necessary for RV formation (8), which suggests that altered regulation of Hand2 in G4NK embryos may contribute at least in part to the abnormal RV morphogenesis seen in these mutant embryos. Previous work has demonstrated that cardiac expression of Hand2 is regulated by a GATA-dependent enhancer and that ablation of GATA sites within this enhancer dramatically reduces RV expression driven by this promoter in transgenic mice (46). These findings suggest that Hand2 downregulation in G4NK mice might be a direct result of decreased Gata4 binding to Hand2 regulatory elements. Interestingly, the GATA-dependent Hand2 enhancer retains expression in Gata4-null mice, which may be due to functional substitution by Gata5 and Gata6 (46). The difference in normal Hand2 enhancer activity in Gata4-null embryos versus decreased Hand2 expression in G4NK embryos might be due to the failure of Gata4-null embryos to develop to the stage at which the Hand2 enhancer requires Gata4 function, as they arrest in development before E8.5.
Our results indicate that a Gata4 transcriptional program is essential for RV formation and suggest that Hand2 likely functions downstream of Gata4 as a component of this program. Intriguingly, ablation of Hand genes was found to result in myocardial hypoplasia but was not associated with a change in cardiomyocyte proliferation (47). These results suggest that while Hand2 likely contributes to the RV hypoplasia seen in G4NK embryos, other genes that influence cardiomyocyte proliferation must also be downstream of Gata4 and contribute to the phenotype. The transcription factor Nkx2-5 was previously reported to regulate expression of Hand1 but not Hand2 (43). Hand1 is predominantly expressed in the LV and is an important regulator of myocardial growth and heart tube looping (48, 49); these findings led to the suggestion that Nkx2-5 and Hand1 selectively regulate LV formation (50). Thus, we propose a model in which Gata4 acts through Hand2 and other genetic pathways to regulate RV development, while Nkx2-5 acts in conjunction with Hand1 to regulate LV development (Figure 7D).
Severe hypoplasia of the RV was present in most, but not all, G4NK embryos. The incomplete penetrance of this phenotype was likely due to several factors. First, it is likely that subtle interembryo variability in the timing and extent of Cre-mediated Gata4 inactivation occurred, which might result in different developmental outcomes. Second, the experiments were performed in a mixed genetic background, and genetic background has a strong influence on RV morphogenesis in the setting of decreased Gata4 expression, as indicated by our finding that some Gata4 heterozygous mice had RV hypoplasia in a pure C57BL/6 background but not in a mixed genetic background (data not shown).
Downstream targets of Gata4.
Previous in vitro and in vivo promoter analysis has implicated Gata4 in the regulation of a number of genes that play important roles in cellular proliferation, heart development, and myocardial contraction (16). Using qRT-PCR, we measured the expression level of 8 genes putatively activated by Gata4 (cyclin D2, Nkx2-5, Mef2c, Carp, MHC, MHC, ANF, and BNP) and found that only MHC expression was downregulated in a manner suggestive of a required role for Gata4 in activation of gene transcription. This might have been due to functional redundancy with other Gata factors. Gata6 was expressed in myocardium at both E9.5 and E12.5, and its expression was not altered by Gata4 deletion (Figure 1D and Supplemental Figures 3 and 4). At E9.5, Gata5 was also expressed in myocardium and was unaffected by Gata4 inactivation (Supplemental Figure 3). Thus, Gata5 might have similarly mitigated the effect of Gata4 inactivation on target gene expression. In addition, Gata4 might regulate the regional expression of target genes; abnormalities of spatial distribution of a target gene might not be reflected in its overall expression level as measured by qRT-PCR. Indeed, while overall ANF expression was not altered in G4NK embryos, ANF was ectopically expressed in some of these hearts (Figure 7C). Altered expression of ANF might reflect a defect in patterning of the RV and OFT of G4NK hearts.
One of the most studied Gata4 target genes is Nkx2-5. Several GATA-dependent enhancers control the expression of Nkx2-5 (35-37). However, we found that after Gata4 inactivation by Nkx2-5Cre, Nkx2-5 expression was downregulated by only 50% at E9.5, and this degree of downregulation was expected as a result of Nkx2-5 haploinsufficiency. Late Gata4 deletion resulted in a slight upregulation of Nkx2-5, similar to the Nkx2-5 upregulation we previously observed in embryos with a germline mutation that reduced Gata4 expression by 70% (22). Thus, our present data indicate that Nkx2-5 expression after E9.5 is not dependent upon Gata4. Recently, BMP induction of Nkx2-5 expression at the onset of cardiogenesis (E7.5 in mouse gestation) was shown to require Gata4 binding to a composite GATA/Smad enhancer, which suggests that Gata4 may be important for the induction rather than the maintenance of Nkx2-5 expression (37). Alternatively, Gata5 or Gata6 may functionally substitute for Gata4 to induce and/or maintain Nkx2-5 expression.
Gata4 regulation of cardiomyocyte proliferation.
We have found that myocardial inactivation of Gata4 both early and late in heart development (in G4NK and G4MC mouse embryos, respectively) resulted in decreased cardiomyocyte proliferation. Myocardial hypoplasia in Nkx2-5 knockout (43) and in Hand knockout mice (47) was not associated with altered cardiomyocyte proliferation or apoptosis, which suggests a defect in recruitment of cardiomyocyte precursors. Thus, the mechanism of myocardial hypoplasia due to Gata4 inactivation is distinct and highlights the important role of Gata4 in promoting the proliferation of cardiomyocytes.
In pulmonary artery smooth muscle, Gata4 was found to regulate cellular proliferation, possibly through regulation of cyclin D2 (30). In other systems, Gata factors both positively and negatively regulate cellular proliferation in a manner that is dependent on the cellular environment. For instance, Gata1 inhibits proliferation and promotes terminal differentiation of erythroblasts (51), while it blocks interleukin-6–induced proliferation arrest in myeloid cells by sustaining cyclin D2 expression (52). Gene expression array analysis identified multiple cell-cycle regulators downstream of Gata1 that might mediate these effects (53). Using a focused candidate gene approach, we did not find significant downregulation of cyclin D1–D3, Hop, or Bmp10 after myocardial Gata4 inactivation. Unbiased array-based approaches will need to be taken in order to identify genes downstream of Gata4 that regulate cardiomyocyte proliferation. Identification of these genes will be important for therapeutic strategies based upon expansion of cardiomyocyte populations.
Conclusion.
Using Cre/loxP technology we have selectively inactivated Gata4 within the myocardium at early and late stages of heart development. At both time points, we demonstrated that Gata4 was an essential regulator of cardiomyocyte proliferation. Early in heart formation, myocardial inactivation of Gata4 resulted in hypoplasia of the RV, which was associated with a downregulation of Hand2 expression and a decrease in cardiomyocyte proliferation that was more severe in the RV than in the LV. These data have important implications for our understanding of the pathogenesis of congenital heart disease and for the development of therapeutic strategies that require modulation of cardiomyocyte proliferation.
Methods
Mice.
Gata4flox/flox mice were generated by gene targeting followed by Flp-mediated removal of a Kan-Neo resistance cassette, as described previously (22). In these mice, exon 2, containing the start codon and 46% of the coding sequence, is deleted upon expression of Cre recombinase. Nkx2-5Cre, MHCCre, and R26RstoplacZ alleles were described previously (23, 24, 28, 43). To obtain G4NK and G4MC embryos, timed matings were set up between Gata4WT/flox, Cre-positive and Gata4flox/flox, Cre-null mice. Noon of the day of the plug was defined as E0.5. All procedures were carried out with the approval of the Institutional Animal Care and Use Committee of Beth Israel Deaconess Medical Center and Boston Children’s Hospital.
Histology and gene expression.
Embryos were dehydrated through an ethanol series and paraffin embedded. The Gata4 probe was constructed by PCR amplification of exon 2, the exon deleted by Cre-mediated recombination. The ANF probe was a gift from C. Seidman (Harvard Medical School, Boston, Massachusetts, USA) (54). The full-length TNC probe was from an expressed sequence tag (dbEST ID 6838718). All other probes were developed as described previously (43). Section in situ hybridization was performed using 35S-labeled probes as previously described (43). Darkfield signal was colored in red and superimposed on brightfield H&GE- or DAPI-stained sections. Whole-mount in situ hybridization was performed using digoxigenin-labeled probes following previously established protocols (55).
Whole-mount X-gal staining for -galactosidase was performed as previously described (56). Immunofluorescent staining for Nfatc1 or Gata4 was performed using monoclonal antibodies (Santa Cruz Biotechnology Inc.) and the ABC method (MOM kit; Vector Laboratories) with tyramide-Cy3 as the HRP substrate. Desmin counterstaining was performed using a Desmin rabbit antibody (BioMeda Corp.) and anti-rabbit Alexa Fluor 488 (Invitrogen Corp.). Nuclei were stained with TOPRO3 (Invitrogen Corp.). Phosphohistone H3 was detected with a polyclonal antibody (Upstate) using the ABC method and tyramide-Cy3. BrdU labeling was achieved by injecting pregnant mice with 2 mg of BrdU intraperitoneally. After 2 hours, embryos were harvested, and BrdU was detected with a monoclonal antibody (Sigma-Aldrich) using the ABC method and tyramide-Cy3. Detection was facilitated by antigen retrieval with HCl and trypsin digestion per instructions of the manufacturer (Sigma-Aldrich). TUNEL staining was performed on paraffin sections using the TMR Red In Situ Death Detection kit (Roche Diagnostics).
qRT-PCR.
Hearts from E9.5 (G4NK) and E12.5 (G4MC) embryos were collected and frozen in liquid nitrogen. Hearts of each genotype (G4NK, n = 6; G4MC, n = 3) were pooled into a single sample, and RNA was prepared using the RNeasy kit with on-column DNase digestion (QIAGEN). Three samples were used per genotype. RNA integrity was verified and concentration determined using an Agilent Bioanalyzer (model 2100; Agilent Technologies). Real-time qRT-PCR was performed on an ABI 7700 Sequence Detector (Applied Biosystems) using Taqman probes (Gata4, Gata6, Nkx2-5, Carp, ANF, cyclin D1, cyclin D2, cyclin D3, Hop, and Bmp10, all from GenScript; and 18S from Applied Biosystems). For primer and probe sequences, see Supplemental Table 1.
Acknowledgments
The authors thank M. Rivera for excellent technical support and M. Zeisberg for critical reading of the manuscript. This work was funded by grants from the NIH National Heart, Lung, and Blood Institute (1 PO1 HL074734 and K08 HL004387-04). E.M. Zeisberg was funded by a Bayer fellowship grant from the German Society of Cardiology and by a grant from the Leopoldina Academy (BMBF-LPD 9901/8-105).
Footnotes
Seigo Izumo’s present address is: Novartis Institute for Biomedical Research, Cambridge, Massachusetts, USA.
Nonstandard abbreviations used: ANT, atrial natriuretic factor; BNP, brain natriuretic peptide; EMT, epithelial-to-mesenchymal transition; G4MC, Gata4flox/flox;MHCCre+; G4NK, Gata4flox/flox;Nkx2-5WT/Cre; MHC, myosin heavy chain; MHCCre; MHC promoter driving Cre recombinase expression; OFT, outflow tract; qRT-PCR, quantitative RT-PCR; TNC, tenascin C.
Conflict of interest: The authors have declared that no conflict of interest exists.
References
Hoffman, J.I. 1995. . Incidence of congenital heart disease: I. Postnatal incidence. Pediatr. Cardiol. 16::103-113.
Waldo, K.L. et al. 2001. . Conotruncal myocardium arises from a secondary heart field. Development. 128::3179-3188.
Mjaatvedt, C.H. et al. 2001. . The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238::97-109.
Kelly, R.G., Brown, N.A., and Buckingham, M.E. 2001. . The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell. 1::435-440.
Runyan, R.B., and Markwald, R.R. 1983. . Invasion of mesenchyme into three-dimensional collagen gels: a regional and temporal analysis of interaction in embryonic heart tissue. Dev. Biol. 95::108-114.
Srivastava, D., and Olson, E.N. 2000. . A genetic blueprint for cardiac development. Nature. 407::221-226.
Srivastava, D., Cserjesi, P., and Olson, E.N. 1995. . A subclass of bHLH proteins required for cardiac morphogenesis. Science. 270::1995-1999.
Srivastava, D. et al. 1997. . Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat. Genet. 16::154-160.
Benson, D.W. et al. 1999. . Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways. J. Clin. Invest. 104::1567-1573.
McElhinney, D.B., Geiger, E., Blinder, J., Benson, D.W., and Goldmuntz, E. 2003. . NKX2.5 mutations in patients with congenital heart disease. J. Am. Coll. Cardiol. 42::1650-1655.
Merscher, S. et al. 2001. . TBX1 is responsible for cardiovascular defects in velo-cardio-facial/DiGeorge syndrome. Cell. 104::619-629.
Lindsay, E.A. et al. 2001. . Tbx1 haploinsufficiency in the DiGeorge syndrome region causes aortic arch defects in mice. Nature. 410::97-101.
Basson, C.T. et al. 1997. . Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15::30-35.
Garg, V. et al. 2003. . GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature. 424::443-447.
Okubo, A. et al. 2004. . A novel GATA4 mutation completely segregated with atrial septal defect in a large Japanese family. J. Med. Genet. 41::e97.
Molkentin, J.D. 2000. . The zinc finger-containing transcription factors GATA-4, -5, and -6. Ubiquitously expressed regulators of tissue-specific gene expression. J. Biol. Chem. 275::38949-38952.
Charron, F., and Nemer, M. 1999. . GATA transcription factors and cardiac development. Semin. Cell Dev. Biol. 10::85-91.
Molkentin, J.D., Lin, Q., Duncan, S.A., and Olson, E.N. 1997. . Requirement of the transcription factor GATA4 for heart tube formation and ventral morphogenesis. Genes Dev. 11::1061-1072.
Kuo, C.T. et al. 1997. . GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11::1048-1060.
Narita, N., Bielinska, M., and Wilson, D.B. 1997. . Wild-type endoderm abrogates the ventral developmental defects associated with GATA-4 deficiency in the mouse. Dev. Biol. 189::270-274.
Crispino, J.D. et al. 2001. . Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15::839-844.
Pu, W.T., Ishiwata, T., Juraszek, A.L., Ma, Q., and Izumo, S. 2004. . GATA4 is a dosage-sensitive regulator of cardiac morphogenesis. Dev. Biol. 275::235-244.
Moses, K.A., DeMayo, F., Braun, R.M., Reecy, J.L., and Schwartz, R.J. 2001. . Embryonic expression of an Nkx2-5/Cre gene using ROSA26 reporter mice. Genesis. 31::176-180.
Mao, X., Fujiwara, Y., and Orkin, S.H. 1999. . Improved reporter strain for monitoring Cre recombinase-mediated DNA excisions in mice. Proc. Natl. Acad. Sci. U. S. A. 96::5037-5042.
Bruneau, B.G. et al. 1999. . Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev. Biol. 211::100-108.
Imanaka-Yoshida, K., Matsumoto, K., Hara, M., Sakakura, T., and Yoshida, T. 2003. . The dynamic expression of tenascin-C and tenascin-X during early heart development in the mouse. Differentiation. 71::291-298.
Xu, H. et al. 2004. . Tbx1 has a dual role in the morphogenesis of the cardiac outflow tract. Development. 131::3217-3227.
Gaussin, V. et al. 2002. . Endocardial cushion and myocardial defects after cardiac myocyte-specific conditional deletion of the bone morphogenetic protein receptor ALK3. Proc. Natl. Acad. Sci. U. S. A. 99::2878-2883.
Sherr, C.J., and Roberts, J.M. 1999. . CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 13::1501-1512.
Suzuki, Y.J. et al. 2003. . Activation of GATA-4 by serotonin in pulmonary artery smooth muscle cells. J. Biol. Chem. 278::17525-17531.
Chen, H. et al. 2004. . BMP10 is essential for maintaining cardiac growth during murine cardiogenesis. Development. 131::2219-2231.
Shin, C.H. et al. 2002. . Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell. 110::725-735.
Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., and Black, B.L. 2004. . Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development. 131::3931-3942.
Kuo, H. et al. 1999. . Control of segmental expression of the cardiac-restricted ankyrin repeat protein gene by distinct regulatory pathways in murine cardiogenesis. Development. 126::4223-4234.
Lien, C.L. et al. 1999. . Control of early cardiac-specific transcription of Nkx2-5 by a GATA-dependent enhancer. Development. 126::75-84.
Searcy, R.D., Vincent, E.B., Liberatore, C.M., and Yutzey, K.E. 1998. . A GATA-dependent nkx-2.5 regulatory element activates early cardiac gene expression in transgenic mice. Development. 125::4461-4470.
Brown, C.O. 3rd et al. 2004. . The cardiac determination factor, Nkx2-5, is activated by mutual cofactors GATA-4 and Smad1/4 via a novel upstream enhancer. J. Biol. Chem. 279::10659-10669.
Thompson, J.T., Rackley, M.S., and O’Brien, T.X. 1998. . Upregulation of the cardiac homeobox gene Nkx2-5 (CSX) in feline right ventricular pressure overload. Am. J. Physiol. 274::H1569-H1573.
Molkentin, J.D., Kalvakolanu, D.V., and Markham, B.E. 1994. . Transcription factor GATA-4 regulates cardiac muscle-specific expression of the alpha-myosin heavy-chain gene. Mol. Cell. Biol. 14::4947-4957.
Small, E.M., and Krieg, P.A. 2003. . Transgenic analysis of the atrialnatriuretic factor (ANF) promoter: Nkx2-5 and GATA-4 binding sites are required for atrial specific expression of ANF. Dev. Biol. 261::116-131.
Houweling, A.C., Somi, S., Van Den Hoff, M.J., Moorman, A.F., and Christoffels, V.M. 2002. . Developmental pattern of ANF gene expression reveals a strict localization of cardiac chamber formation in chicken. Anat. Rec. 266::93-102.
Watt, A.J., Battle, M.A., Li, J., and Duncan, S.A. 2004. . GATA4 is essential for formation of the proepicardium and regulates cardiogenesis. Proc. Natl. Acad. Sci. U. S. A. 101::12573-12578.
Tanaka, M., Chen, Z., Bartunkova, S., Yamasaki, N., and Izumo, S. 1999. . The cardiac homeobox gene Csx/Nkx2.5 lies genetically upstream of multiple genes essential for heart development. Development. 126::1269-1280.
Biben, C. et al. 2000. . Cardiac septal and valvular dysmorphogenesis in mice heterozygous for mutations in the homeobox gene Nkx2-5. Circ. Res. 87::888-895.
Tanaka, M. et al. 2002. . A mouse model of congenital heart disease: cardiac arrhythmias and atrial septal defect caused by haploinsufficiency of the cardiac transcription factor Csx/Nkx2.5. Cold Spring Harb. Symp. Quant. Biol. 67::317-325.
McFadden, D.G. et al. 2000. . A GATA-dependent right ventricular enhancer controls dHAND transcription in the developing heart. Development. 127::5331-5341.
McFadden, D.G. et al. 2005. . The Hand1 and Hand2 transcription factors regulate expansion of the embryonic cardiac ventricles in a gene dosage-dependent manner. Development. 132::189-201.
Firulli, A.B., McFadden, D.G., Lin, Q., Srivastava, D., and Olson, E.N. 1998. . Heart and extra-embryonic mesodermal defects in mouse embryos lacking the bHLH transcription factor Hand1. Nat. Genet. 18::266-270.
Riley, P., Anson-Cartwright, L., and Cross, J.C. 1998. . The Hand1 bHLH transcription factor is essential for placentation and cardiac morphogenesis. Nat. Genet. 18::271-275.
Yamagishi, H. et al. 2001. . The combinatorial activities of Nkx2.5 and dHAND are essential for cardiac ventricle formation. Dev. Biol. 239::190-203.
Weiss, M.J., Yu, C., and Orkin, S.H. 1997. . Erythroid-cell-specific properties of transcription factor GATA-1 revealed by phenotypic rescue of a gene-targeted cell line. Mol. Cell. Biol. 17::1642-1651.
Tanaka, H. et al. 2000. . GATA-1 blocks IL-6-induced macrophage differentiation and apoptosis through the sustained expression of cyclin D1 and bcl-2 in a murine myeloid cell line M1. Blood. 95::1264-1273.
Rylski, M. et al. 2003. . GATA-1-mediated proliferation arrest during erythroid maturation. Mol. Cell. Biol. 23::5031-5042.
Zeller, R., Bloch, K.D., Williams, B.S., Arceci, R.J., and Seidman, C.E. 1987. . Localized expression of the atrial natriuretic factor gene during cardiac embryogenesis. Genes Dev. 1::693-698.
Brent, A.E., Schweitzer, R., and Tabin, C.J. 2003. . A somitic compartment of tendon progenitors. Cell. 113::235-248.
Kupershmidt, S. et al. 1999. . Replacement by homologous recombination of the minK gene with lacZ reveals restriction of minK expression to the mouse cardiac conduction system. Circ. Res. 84::146-152.(Elisabeth M. Zeisberg, Qi)
2Center for Matrix Biology, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, USA.
3Department of Cardiology, Children’s Hospital Boston, Boston, Massachusetts, USA.
4Department of Pathology, Harvard Medical School, Boston, Massachusetts, USA.
5Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas, USA.
6Department of Pediatrics, Harvard Medical School, Boston, Massachusetts, USA.
Abstract
Mutations in developmental regulatory genes have been found to be responsible for some cases of congenital heart defects. One such regulatory gene is Gata4, a zinc finger transcription factor. In order to circumvent the early embryonic lethality of Gata4-null embryos and to investigate the role of myocardial Gata4 expression in cardiac development, we used Cre/loxP technology to conditionally delete Gata4 in the myocardium of mice at an early and a late time point in cardiac morphogenesis. Early deletion of Gata4 by Nkx2-5Cre resulted in hearts with striking myocardial thinning, absence of mesenchymal cells within the endocardial cushions, and selective hypoplasia of the RV. RV hypoplasia was associated with downregulation of Hand2, a transcription factor previously shown to regulate formation of the RV. Cardiomyocyte proliferation was reduced, with a greater degree of reduction in the RV than in the LV. Late deletion of Gata4 by Cre recombinase driven by the myosin heavy chain promoter did not selectively affect RV development or generation of endocardial cushion mesenchyme but did result in marked myocardial thinning with decreased cardiomyocyte proliferation, as well as double-outlet RV. Our results demonstrate a general role of myocardial Gata4 in regulating cardiomyocyte proliferation and a specific, stage-dependent role in regulating the morphogenesis of the RV and the atrioventricular canal.
Introduction
Congenital heart defects are the most common developmental anomaly in newborns (1). Formation of the normal heart involves the transformation of a linear heart tube into the mature 4-chambered organ. The linear heart tube grows rapidly by the proliferation of cardiomyocytes and by the addition of new cardiomyocytes to the arterial pole of the heart. These new cardiomyocytes, derived from splanchnic mesoderm adjoining the arterial pole of the heart (termed the secondary heart field), give rise to the outflow tract (OFT) and also contribute to the development of the RV (2-4). As the heart tube lengthens, it loops to the right, and the resulting structure is septated by growth and maturation of the endocardial cushions, structures derived from the endothelial lining of the heart by an epithelial-to-mesenchymal transition (EMT) (5).
Genetic dissection of heart development has shown that a hierarchy of transcription factors regulates these morphogenic events (6). Relatively independent genetic programs control the development of each heart chamber, which is consistent with the frequent occurrence of selective chamber hypoplasia in severe forms of congenital heart disease (6). For example, during mouse heart development, the transcription factor Hand2 (dHAND) is predominantly expressed in the RV, while the related transcription factor Hand1 (eHAND) is predominantly expressed in the LV (7). Deletion of Hand2 results in severe hypoplasia of the RV segment (8).
It has been found that some cases of congenital heart defects were caused by mutations in developmental regulatory genes. Mutation of the cardiac transcription factor Nkx2-5 is associated with familial congenital heart disease and has been associated with up to 4% of cases of tetralogy of Fallot and secundum atrial septal defect (9, 10). Haploinsufficiency for the transcription factor Tbx1 in patients with 22q11 chromosomal microdeletions causes OFT anomalies (11, 12), while mutation of the transcription factor Tbx5 is associated with congenital heart disease in Holt-Oram syndrome and in nonsyndromic congenital heart disease (13).
Recently, mutation of the transcription factor Gata4 was linked to congenital heart disease characterized by atrial or ventricular septal defects (14, 15). In the heart, Gata4 is expressed in cardiomyocytes and their mesodermal precursors, as well as in the endocardium and the epicardium (Figure 1, E, F, I, and J, and Figure 5, A, C, and E). Gata4 binds to and regulates expression of a number of myocardium-expressed genes (16, 17). Embryos lacking Gata4 arrested in development and died due to a defect in the visceral endoderm that caused aberrant ventral morphogenesis and cardiac bifida (18-20). Using alleles of Gata4 that cause partial loss of function, we and others have previously demonstrated that Gata4 is required for later heart development (21, 22). However, these experiments were not able to address the tissue-restricted requirements for Gata4, particularly during early stages of heart development. To evaluate the function of Gata4 within the myocardial compartment, we crossed floxed Gata4 mice with mouse lines in which expression of Cre recombinase is initiated within the myocardium at an early and a late time point in cardiac development. We found that myocardial expression of Gata4 was required in a stage-dependent manner for proliferation of cardiomyocytes, formation of the endocardial cushions, development of the RV, and septation of the OFT.
Results
Myocardial-restricted deletion of Gata4 early in cardiac development.
Previously, in order to study the temporal and spatial requirements for Gata4 in heart development, we generated a floxed Gata4 allele (Gata4flox) in which exon 2, containing the start codon and 46% of the coding sequence, is flanked by loxP sites (22). Expression of Cre recombinase resulted in deletion of exon 2, yielding a null allele (22). In order to investigate the cardiomyocyte-restricted function of Gata4 in cardiac development, we used Cre/loxP technology to conditionally ablate Gata4 in cardiomyocytes at 2 distinct time points in cardiac development. We achieved early cardiomyocyte-restricted deletion of Gata4 using Nkx2-5Cre, in which Cre recombinase expression is driven by the endogenous Nkx2-5 locus (23). We previously reported that Nkx2-5Cre initiated Cre expression around E7.5 in the cardiac progenitor pool and that robust Cre-mediated recombination occurred in cardiomyocytes by E9.5 (23). In order to further delineate the temporal and spatial pattern of Cre activity, we crossed Nkx2-5Cre mice with R26RstoplacZ mice, in which expression of lacZ requires Cre-mediated excision of a stop cassette (24). Consistent with our prior results (23), Nkx2-5Cre activated lacZ expression in most cardiomyocytes of E9.5 embryos (arrows, Figure 1, B and C). Cre-mediated expression of lacZ did not occur in the endocardium (arrowheads, Figure 1, B and C) or in the proepicardial organ (Figure 1A). Outside the heart, lacZ activity was detected in the pharyngeal endoderm and in a patchy distribution in branchial arch epithelium (Figure 1A and Figure 4, A–C).
Next, we obtained Gata4flox/flox;Nkx2-5WT/Cre (G4NK) embryos from timed matings and we confirmed loss of the Gata4 transcript in G4NK hearts by quantitative RT-PCR (qRT-PCR) and in situ hybridization. Gata4 expression was reduced 90% in G4NK hearts compared with that in controls (P < 0.05; n = 3; Figure 1D), while the expression of the closely related gene Gata6 was unaffected (Figure 1D). In situ hybridization showed that Gata4 expression in the myocardium was not detectable above background by E9.5 in G4NK embryos (compare red arrowheads in Figure 1H and Figure 1F). Surprisingly, we found that Gata4 expression in the endocardium was also reduced to background levels in G4NK embryos (compare yellow arrowheads in Figure 1H and Figure 1F). Gata5 and Gata6 expression was not altered in G4NK embryos (Supplemental Figure 3; supplemental material available online with this article; doi:10.1172/JCI23769DS1).
To further confirm the loss of myocardial and endocardial Gata4 expression, we examined Gata4 protein expression in G4NK hearts by immunofluorescence staining (Figure 1, I–L). In control embryos, robust Gata4 staining (red stain) was localized to the nucleus (blue stain) of cardiomyocytes and endocardial cells (Figure 1, I and J). However, in G4NK embryos, Gata4 signal was strongly decreased in both cardiomyocytes and endocardial cells (Figure 1, K and L). As the antibody used recognizes a peptide in the C terminus of Gata4, loss of immunoreactivity in G4NK hearts suggests that excision of the floxed region of Gata4 (exon 2, encoding the N-terminal portion of Gata4 including the start codon) resulted in loss of Gata4 protein. Loss of endocardial Gata4 expression was not due to a generalized defect in endothelial differentiation into endocardium, as the endocardial-specific marker Nfatc1 was expressed normally in G4NK hearts (Supplemental Figure 1). Because Cre activity was absent in the endocardium of Nkx2-5Cre hearts (Figure 1, B and C), loss of endocardial Gata4 expression in G4NK hearts suggests that a paracrine signal derived from the myocardium regulates endocardial Gata4 expression.
Early cardiomyocyte-restricted ablation of Gata4 in G4NK embryos resulted in lethality by E11.5 (Figure 2A). However, overall development of G4NK embryos at E9.5, as assessed by the number of somites, was not impaired (control, 20.8 ± 2.8 somites, n = 15; mutant, 21.7 ± 0.5 somites, n = 7). By E10.5, G4NK embryos were mildly delayed in overall development (control, 31.6 ± 2.2 somites, n = 19; mutant, 28.5 ± 2.2 somites, n = 6; P < 0.05) and had large pericardial effusions, a sign of heart failure (compare Figure 2B and Figure 2C).
At E9.5, most G4NK embryos could be recognized on whole-mount examination by characteristic cardiac malformations. G4NK embryos had normally positioned atria and intact atrioventricular grooves (arrowheads, Figure 2, D–F). Fifteen of 21 G4NK embryos systematically examined in whole mount displayed a single predominant ventricular chamber that connected to an OFT located toward the rostral side of the chamber (compare arrows in Figure 2G and Figure 2I). The remaining 6 of 21 G4NK embryos had normal- to mildly hypoplastic–appearing RVs that connected to normally positioned OFTs (Figure 2, H and O). On histological sections, all G4NK embryos displayed marked myocardial hypoplasia, affecting both the compact and trabecular myocardium (Figure 2, J–M). The atrioventricular and OFT endocardial cushions were small and contained few mesenchymal cells (arrowheads, Figure 2, J, K, N, and O), which indicates a defect in the EMT of endocardial cells that normally generates cushion mesenchyme. This defect was due at least in part to the aforementioned downregulation of Gata4 in the endocardium, as we have found that expression of Gata4 in the endocardium is required for EMT (J. Rivera-Feliciano et al., manuscript submitted for publication).
Myocardial Gata4 regulates formation of the RV.
Early myocyte-restricted deletion of Gata4 in G4NK embryos resulted in the formation of a single predominant ventricular chamber in the majority of embryos (Figure 2, I and P). In order to determine the molecular identity of the chamber, we examined the expression of chamber-specific markers. In normal embryos, at E9.5 the atrial chamber expressed myosin light chain 2a (MLC2a) but not MLC2v, while the ventricular chambers expressed both MLC2a and MLC2v. This expression pattern was maintained in G4NK hearts (Supplemental Figure 2), which indicates that specification of atrial and ventricular chambers occurs normally in G4NK hearts. The genes Hand1 and Tbx5 are normally expressed at higher levels in the LV than in the RV (8, 25). In G4NK hearts, the expression pattern of Hand1 and Tbx5 was not altered, with the predominant ventricular chamber preferentially expressing these markers (Figure 3, I and J, and data not shown). These data indicate that the predominant ventricular chamber in G4NK hearts is the LV. The data also demonstrate that Hand1 is not directly regulated by Gata4.
To determine whether a hypoplastic RV was incorporated into the tubular structure connecting the predominant ventricular chamber to the aortic sac in mutant embryos, we examined the expression of the OFT marker tenascin C (TNC) by in situ hybridization. In control embryos, the entire length of the OFT from the morphological RV to the aortic sac expressed TNC (Supplemental Figure 2) (26). In contrast, in G4NK embryos with a single predominant chamber, TNC was expressed in the distal end of the OFT. However, the proximal end of the outflow tube near its junction with the predominant ventricular chamber was TNC negative (Supplemental Figure 2) and likely represented a severely hypoplastic RV.
The RV hypoplasia found in G4NK hearts was highly reminiscent of the phenotype of Hand2-null embryos, which failed to form the RV (8). Hand2 is expressed in the limb buds, the branchial arches, and the myocardium of the OFT and the RV (8). By both whole-mount and section in situ hybridization, we found that Hand2 transcript levels were decreased in G4NK hearts compared with those in controls (Figure 3, A–F). The decrease in Hand2 transcript level was not due to nonspecific RNA degradation, as adjoining sections from the same embryo yielded robust signals to other probes (e.g., Hand1, Figure 3J). Decreased Hand2 levels were also not simply a consequence of hypoplasia of the RV, as Hand2 expression in the OFT was also strongly downregulated (arrows, Figure 3, E and F). Hand2 remained robustly expressed in the branchial arches of mutant embryos (arrowheads, Figure 3, A–F). These data demonstrate that normal expression of Hand2 requires Gata4 and suggest that abnormal regulation of Hand2 in G4NK mutants contributes at least in part to the selective requirement of Gata4 for RV development.
The RV and OFT are derived at least in part from cardiomyocytes generated in a secondary heart field located at the junction of the splanchnic mesoderm with the caudal surface of the OFT (2-4). In chicken embryos, this secondary heart field contains progenitor cells that express both Nkx2-5 and Gata4 (2). One hypothesis to account for the RV hypoplasia seen in G4NK mutants is that Gata4 inactivation in these progenitors impairs formation or migration of cardiomyocytes destined to contribute to the RV. To address this hypothesis, we assessed the timing and location within the secondary heart field of Gata4, -5, and -6 expression and of Cre activity driven by Nkx2-5Cre. By E9.5, when RV hypoplasia became apparent, few cells within this region had undergone recombination to express lacZ (Figure 4, A and B). At E10.5, we found that Nkx2-5Cre activation of the R26RstoplacZ reporter within the splanchnic mesoderm was more robust (Figure 4C), as has been previously reported (27). More importantly, we found that Gata4 was not expressed in splanchnic mesoderm at its junction with the OFT in wild-type embryos (arrowheads, Figure 4, E and I), while Gata5 and Gata6 were expressed in this region (arrowheads, Figure 4, F, G, J, and K). These data indicate that Gata4 inactivation by Nkx2-5Cre in the secondary heart field is not responsible for hypoplasia of the RV.
Myocardial-restricted deletion of Gata4 late in cardiac development.
In order to study the function of myocardial Gata4 expression at a late time in heart development, we used transgenic mice in which Cre recombinase expression was driven by the myosin heavy chain (MHC) promoter (MHCCre). Prior characterization of this transgene demonstrated that MHCCre catalyzed recombination only in the myocardium, starting at E9.5 and occurring in nearly all cardiomyocytes by E11.5 (28). We isolated Gata4flox/flox;MHCCre+ (G4MC) embryos at E12.5 and examined Gata4 expression by in situ hybridization and immunohistochemistry. In control embryos, Gata4 expression was readily detected in myocardium (white arrowheads, Figure 5, A, C, and E), endocardium (yellow arrowheads, Figure 5, A and E), endocardial cushions (asterisk, Figure 5A), and epicardium (blue arrowheads, Figure 5, C and E). In G4MC embryos, myocardial Gata4 expression was strongly reduced in both compact and trabecular myocardium (compare Figure 5, D and F with Figure 5, C and E), while Gata4 continued to be expressed in epicardium and endocardium (blue and yellow arrowheads, Figure 5, B, D, and F). Expression of Gata5 and Gata6 were not altered in G4MC embryos (Supplemental Figure 4).
Late cardiomyocyte-restricted deletion of Gata4 by MHCCre resulted in lethality by E14.5 (Figure 2A). Mutant embryos had marked myocardial hypoplasia and double-outlet RV (Figure 5, G–L). While early excision of Gata4 by Nkx2-5Cre resulted in the formation of a single predominant ventricular chamber in the majority of embryos, 2 well-formed ventricles were present after late Gata4 inactivation by MHCCre. In G4MC hearts, as in control hearts, the atrioventricular cushion divided the atrioventricular canal into separate RV and LV inflow tracts (asterisk, Figure 5, G and J). This result indicates that, unlike early myocardial Gata4 deletion, late deletion does not impair initial steps of atrioventricular valve morphogenesis, including generation of cushion mesenchyme and growth and fusion of the atrioventricular endocardial cushions.
Myocardial Gata4 regulates cardiomyocyte proliferation.
Hypoplasia of the myocardium was a prominent feature after both early and late myocyte-restricted Gata4 deletion. This phenotype could have been due to increased cardiomyocyte apoptosis or decreased cardiomyocyte proliferation. TUNEL staining of G4NK and G4MC embryos at E9.0 and E12.5 showed that cardiomyocyte apoptosis did not increase compared with that of control littermates (data not shown). To determine whether myocardial Gata4 inactivation decreases cardiomyocyte proliferation, we measured cardiomyocyte proliferation by BrdU labeling. After both early and late Gata4 deletion, the percentage of myocytes labeled by BrdU was significantly decreased (P < 0.05; Figure 6, A–H, and Table 1). Similar results were obtained in G4NK embryos when cardiomyocyte proliferation was measured by staining for the M phase–specific marker phosphohistone H3 (data not shown). Early Gata4 inactivation resulted in selective RV hypoplasia (Figure 2), while late Gata4 inactivation did not (Figure 5). Consistent with these phenotypes, cardiomyocyte proliferation was more severely reduced in the RV than in the LV after early Gata4 inactivation (Figure 6, A–F), but the reductions did not differ between RV and LV after late Gata4 inactivation (Figure 6, G and H). These data suggest that Gata4 is required in a cell-autonomous manner for normal proliferation of cardiomyocytes and that reduced cardiomyocyte proliferation might contribute to the selective RV hypoplasia seen in G4NK hearts.
In order to identify transcriptional targets of Gata4 that may mediate the effect of Gata4 on cardiomyocyte proliferation, we asked whether expression of myocardial genes known to be important for myocardial growth was perturbed by Gata4 deletion. cyclin D1–D3 regulate the core cell-cycle machinery in response to mitogenic stimulation (29), and in other systems Gata4 regulates proliferation by activating cyclin D2 expression (30). Bmp10 and Hop are both required for normal myocardial growth, and expression of both is regulated by Nkx2-5, which can coordinate with Gata4 to regulate gene transcription (31, 32). qRT-PCR analysis of RNA isolated from G4MC and Gata4flox/flox control hearts showed no difference in the expression of Bmp10 (Figure 6I). Expression of Hop tended to be reduced in G4MC hearts, but this did not reach statistical significance (Figure 6I). Unexpectedly, cyclin D1–D3 were not downregulated in G4MC hearts; rather, each was subtly but significantly upregulated (P < 0.05; Figure 6I). This indicates that Gata4 does not directly regulate cardiomyocyte proliferation through cyclin D1–D3. Instead, cyclin D1–D3 upregulation may be a compensatory response to altered expression of other components of the cell-cycle machinery following Gata4 deletion.
Myocardial gene expression downstream of Gata4.
Gata4 has been implicated in the expression of a large number of myocardial genes, including the transcription factors Mef2c, Carp, and Nkx2-5, each of which is essential for normal cardiac morphogenesis (33-37). To determine whether abnormalities of cardiac morphogenesis seen in G4NK or G4MC mutants could be attributed to altered transcription of these genes in vivo, we measured transcript levels by qRT-PCR on RNA isolated from control and mutant hearts (Figure 7A). We found that expression of Mef2c and Carp were not significantly altered in mutant hearts after early or late myocardial Gata4 deletion (Figure 7A). Nkx2-5 expression was reduced by 50% in G4NK hearts, but this decrease was expected due to haploinsufficiency for Nkx2-5. After late myocyte-restricted Gata4 deletion, Nkx2-5 expression was upregulated 1.7-fold in G4NK hearts compared with that of controls (P < 0.05; Figure 7A). We previously found a similar degree of Nkx2-5 upregulation in E12.5 hearts with a germline mutation that reduced Gata4 expression by 70% (22). This might reflect upregulation secondary to heart failure, since Nkx2-5 has previously been shown to be upregulated as a result of increased wall stress (38).
Gata4 has also been implicated in the expression of a number of cardiomyocyte structural genes. We measured the expression of 4 well-characterized Gata4 targets in G4MC hearts compared with that in controls. Surprisingly, we found that levels of atrial natriuretic factor (ANF), brain natriuretic peptide (BNP), and MHC expression were unaltered in G4MC hearts (Figure 7B). MHC expression was downregulated 2-fold after late myocyte-restricted Gata4 deletion (P < 0.05; Figure 7B), which is consistent with the prior finding that GATA binding sites in the proximal MHC promoter are required for promoter activity in vitro (39).
Analysis of the ANF promoter in transgenic Xenopus embryos suggests that GATA binding sites in the ANF promoter are crucial for spatially restricting ANF expression from the OFT, rather than for determining the level of expression (40). To determine whether Gata4 is required to similarly restrict the spatial distribution of ANF expression in mouse embryos, we used in situ hybridization to examine ANF expression after cardiomyocyte-restricted inactivation of Gata4. ANF is normally expressed in atrial and ventricular chamber myocardium, but not in the atrioventricular canal or OFT myocardium (41). In G4NK hearts, ANF was expressed at comparable levels in atrial and ventricular chamber myocardium. However, in 2 out of 6 G4NK embryos, ANF expression extended ectopically, well into the myocardium connecting the predominant ventricular chamber to the OFT (Figure 7C). Thus, in vivo Gata4 is necessary to consistently restrict the ANF expression domain in mice, which is consistent with the requirement for GATA sites in the ANF promoter to restrict the ANF expression domain in Xenopus (40).
Discussion
To determine the myocyte-restricted function of Gata4 in heart development, we characterized embryos in which Gata4 was inactivated at an early and a late time point in heart development. Our results demonstrate distinct temporal requirements for myocardial expression of Gata4 in the regulation of cardiac morphogenesis. Early myocardial expression of Gata4 was necessary for normal cardiac expression of Hand2 and for normal development of the RV. Early expression of Gata4 in myocardium was also necessary to support endocardial Gata4 expression and endocardial EMT. Late loss of myocardial Gata4 expression did not selectively affect RV development but did result in abnormal OFT septation, manifested as double-outlet RV. Both early and late in cardiac development, normal cardiomyocyte proliferation required expression of Gata4.
Recently, tetraploid aggregation of Gata4-null embryonic stem cells was used to bypass the early developmental arrest of Gata4-null embryos (42). As in our embryos with cardiomyocyte-restricted Gata4 deletion, tetraploid-rescued null embryos formed a thin, poorly trabeculated heart tube. However, early cardiomyocyte-restricted Gata4 inactivation resulted in a milder phenotype than in tetraploid-rescued Gata4-null embryos, since the latter had incomplete segmentation and looping of the heart tube. This indicates that they arrested earlier in heart development than did our embryos with early myocyte-restricted Gata4 inactivation (G4NK). Tetraploid-rescued Gata4-null embryos lacked a proepicardial organ. This, in conjunction with the relatively unperturbed expression of myocardial genes, led the authors to conclude that disruption of myocardial development is non–cell autonomous and perhaps due to loss of signals from the proepicardial organ (42). However, our data directly demonstrate an important cardiomyocyte-autonomous function for Gata4 in cardiomyocyte proliferation and RV formation. The G4NK phenotype is milder than the tetraploid-rescued Gata4-null phenotype, which supports additional roles for Gata4 expression in non-myocyte lineages to regulate cardiac morphogenesis. Alternatively, the difference in phenotype may reflect a later deletion of Gata4 in the G4NK hearts.
Because G4NK embryos are haploinsufficient for Nkx2-5, it is possible that reduced Nkx2-5 dosage also contributed to the phenotype observed in G4NK embryos. However, this is unlikely for several reasons. First, Nkx2-5 is not a direct regulator of RV specification or Hand2 expression (43). Second, tetraploid-rescued Gata4-null hearts (42) had abnormalities similar to, but more severe than, G4NK hearts, which indicates that these phenotypes also occur in Gata4-null embryos that have a normal dose of Nkx2-5. Third, while Nkx2-5 haploinsufficiency is associated with atrial septal defects and conduction system abnormalities (44, 45), it has not been reported to result in an observable phenotype at this early stage of heart development. Finally, the myocardial hypoplasia of G4NK embryos was also seen in G4MC embryos, which expressed slightly elevated levels of Nkx2-5. Nevertheless, we cannot exclude the possibility that Nkx2-5 haploinsufficiency contributed to the G4NK phenotype and to the differences between the phenotype of G4NK and G4MC hearts.
Gata4 regulates RV development.
Severe hypoplasia of the RV occurred in most mutant embryos after early cardiomyocyte-restricted deletion, which indicates that Gata4 plays an essential role in RV morphogenesis. Results of previous work have suggested that the OFT and at least a portion of the RV are derived from progenitors located within the secondary heart field (2-4). In chicks, cells within the secondary heart field express Gata4 and Nkx2-5 (2), and in mice, a secondary heart field enhancer has been shown to be dependent upon Gata activity (33). However, we found that in mice, Gata4 was not expressed within the secondary heart field, whereas Gata5 and Gata6 were robustly expressed in this region, which suggests that the Gata isoforms active in the splanchnic mesoderm are species specific. In addition, the Nkx2-5Cre–mediated excision in the secondary heart field occurred in a minority of cells at E9.5 (Figure 4, A and B), when hypoplasia of the RV was already apparent. These data indicate that hypoplasia of the RV in G4NK mutant embryos was not due to Gata4 inactivation in RV precursors within the secondary heart field. Inactivation of Gata4 within cells originating from the secondary heart field might occur after they have migrated into the heart tube and differentiated into cardiomyocytes; thus, our data are consistent with an important role of the secondary heart field in the development of the RV.
We found that RV hypoplasia after early myocyte-restricted Gata4 inactivation was associated with a decrease in cardiomyocyte proliferation that was more severe in the RV than in the LV. At E9.5, RV cardiomyocytes may be more dependent upon Gata4 for normal rates of proliferation than are LV cardiomyocytes. Alternatively, Gata4 inactivation in RV cardiomyocyte precursors originating from the secondary heart field might impair their migration, expansion, or differentiation, which ultimately might be manifested as a decrease in cardiomyocyte proliferation. Distinguishing these models will require mapping the fate of RV progenitors in the G4NK mutant background.
The failure of RV development after early cardiomyocyte-restricted inactivation of Gata4 was associated with downregulation of Hand2 in the RV and the OFT. Hand2 itself has been shown to be necessary for RV formation (8), which suggests that altered regulation of Hand2 in G4NK embryos may contribute at least in part to the abnormal RV morphogenesis seen in these mutant embryos. Previous work has demonstrated that cardiac expression of Hand2 is regulated by a GATA-dependent enhancer and that ablation of GATA sites within this enhancer dramatically reduces RV expression driven by this promoter in transgenic mice (46). These findings suggest that Hand2 downregulation in G4NK mice might be a direct result of decreased Gata4 binding to Hand2 regulatory elements. Interestingly, the GATA-dependent Hand2 enhancer retains expression in Gata4-null mice, which may be due to functional substitution by Gata5 and Gata6 (46). The difference in normal Hand2 enhancer activity in Gata4-null embryos versus decreased Hand2 expression in G4NK embryos might be due to the failure of Gata4-null embryos to develop to the stage at which the Hand2 enhancer requires Gata4 function, as they arrest in development before E8.5.
Our results indicate that a Gata4 transcriptional program is essential for RV formation and suggest that Hand2 likely functions downstream of Gata4 as a component of this program. Intriguingly, ablation of Hand genes was found to result in myocardial hypoplasia but was not associated with a change in cardiomyocyte proliferation (47). These results suggest that while Hand2 likely contributes to the RV hypoplasia seen in G4NK embryos, other genes that influence cardiomyocyte proliferation must also be downstream of Gata4 and contribute to the phenotype. The transcription factor Nkx2-5 was previously reported to regulate expression of Hand1 but not Hand2 (43). Hand1 is predominantly expressed in the LV and is an important regulator of myocardial growth and heart tube looping (48, 49); these findings led to the suggestion that Nkx2-5 and Hand1 selectively regulate LV formation (50). Thus, we propose a model in which Gata4 acts through Hand2 and other genetic pathways to regulate RV development, while Nkx2-5 acts in conjunction with Hand1 to regulate LV development (Figure 7D).
Severe hypoplasia of the RV was present in most, but not all, G4NK embryos. The incomplete penetrance of this phenotype was likely due to several factors. First, it is likely that subtle interembryo variability in the timing and extent of Cre-mediated Gata4 inactivation occurred, which might result in different developmental outcomes. Second, the experiments were performed in a mixed genetic background, and genetic background has a strong influence on RV morphogenesis in the setting of decreased Gata4 expression, as indicated by our finding that some Gata4 heterozygous mice had RV hypoplasia in a pure C57BL/6 background but not in a mixed genetic background (data not shown).
Downstream targets of Gata4.
Previous in vitro and in vivo promoter analysis has implicated Gata4 in the regulation of a number of genes that play important roles in cellular proliferation, heart development, and myocardial contraction (16). Using qRT-PCR, we measured the expression level of 8 genes putatively activated by Gata4 (cyclin D2, Nkx2-5, Mef2c, Carp, MHC, MHC, ANF, and BNP) and found that only MHC expression was downregulated in a manner suggestive of a required role for Gata4 in activation of gene transcription. This might have been due to functional redundancy with other Gata factors. Gata6 was expressed in myocardium at both E9.5 and E12.5, and its expression was not altered by Gata4 deletion (Figure 1D and Supplemental Figures 3 and 4). At E9.5, Gata5 was also expressed in myocardium and was unaffected by Gata4 inactivation (Supplemental Figure 3). Thus, Gata5 might have similarly mitigated the effect of Gata4 inactivation on target gene expression. In addition, Gata4 might regulate the regional expression of target genes; abnormalities of spatial distribution of a target gene might not be reflected in its overall expression level as measured by qRT-PCR. Indeed, while overall ANF expression was not altered in G4NK embryos, ANF was ectopically expressed in some of these hearts (Figure 7C). Altered expression of ANF might reflect a defect in patterning of the RV and OFT of G4NK hearts.
One of the most studied Gata4 target genes is Nkx2-5. Several GATA-dependent enhancers control the expression of Nkx2-5 (35-37). However, we found that after Gata4 inactivation by Nkx2-5Cre, Nkx2-5 expression was downregulated by only 50% at E9.5, and this degree of downregulation was expected as a result of Nkx2-5 haploinsufficiency. Late Gata4 deletion resulted in a slight upregulation of Nkx2-5, similar to the Nkx2-5 upregulation we previously observed in embryos with a germline mutation that reduced Gata4 expression by 70% (22). Thus, our present data indicate that Nkx2-5 expression after E9.5 is not dependent upon Gata4. Recently, BMP induction of Nkx2-5 expression at the onset of cardiogenesis (E7.5 in mouse gestation) was shown to require Gata4 binding to a composite GATA/Smad enhancer, which suggests that Gata4 may be important for the induction rather than the maintenance of Nkx2-5 expression (37). Alternatively, Gata5 or Gata6 may functionally substitute for Gata4 to induce and/or maintain Nkx2-5 expression.
Gata4 regulation of cardiomyocyte proliferation.
We have found that myocardial inactivation of Gata4 both early and late in heart development (in G4NK and G4MC mouse embryos, respectively) resulted in decreased cardiomyocyte proliferation. Myocardial hypoplasia in Nkx2-5 knockout (43) and in Hand knockout mice (47) was not associated with altered cardiomyocyte proliferation or apoptosis, which suggests a defect in recruitment of cardiomyocyte precursors. Thus, the mechanism of myocardial hypoplasia due to Gata4 inactivation is distinct and highlights the important role of Gata4 in promoting the proliferation of cardiomyocytes.
In pulmonary artery smooth muscle, Gata4 was found to regulate cellular proliferation, possibly through regulation of cyclin D2 (30). In other systems, Gata factors both positively and negatively regulate cellular proliferation in a manner that is dependent on the cellular environment. For instance, Gata1 inhibits proliferation and promotes terminal differentiation of erythroblasts (51), while it blocks interleukin-6–induced proliferation arrest in myeloid cells by sustaining cyclin D2 expression (52). Gene expression array analysis identified multiple cell-cycle regulators downstream of Gata1 that might mediate these effects (53). Using a focused candidate gene approach, we did not find significant downregulation of cyclin D1–D3, Hop, or Bmp10 after myocardial Gata4 inactivation. Unbiased array-based approaches will need to be taken in order to identify genes downstream of Gata4 that regulate cardiomyocyte proliferation. Identification of these genes will be important for therapeutic strategies based upon expansion of cardiomyocyte populations.
Conclusion.
Using Cre/loxP technology we have selectively inactivated Gata4 within the myocardium at early and late stages of heart development. At both time points, we demonstrated that Gata4 was an essential regulator of cardiomyocyte proliferation. Early in heart formation, myocardial inactivation of Gata4 resulted in hypoplasia of the RV, which was associated with a downregulation of Hand2 expression and a decrease in cardiomyocyte proliferation that was more severe in the RV than in the LV. These data have important implications for our understanding of the pathogenesis of congenital heart disease and for the development of therapeutic strategies that require modulation of cardiomyocyte proliferation.
Methods
Mice.
Gata4flox/flox mice were generated by gene targeting followed by Flp-mediated removal of a Kan-Neo resistance cassette, as described previously (22). In these mice, exon 2, containing the start codon and 46% of the coding sequence, is deleted upon expression of Cre recombinase. Nkx2-5Cre, MHCCre, and R26RstoplacZ alleles were described previously (23, 24, 28, 43). To obtain G4NK and G4MC embryos, timed matings were set up between Gata4WT/flox, Cre-positive and Gata4flox/flox, Cre-null mice. Noon of the day of the plug was defined as E0.5. All procedures were carried out with the approval of the Institutional Animal Care and Use Committee of Beth Israel Deaconess Medical Center and Boston Children’s Hospital.
Histology and gene expression.
Embryos were dehydrated through an ethanol series and paraffin embedded. The Gata4 probe was constructed by PCR amplification of exon 2, the exon deleted by Cre-mediated recombination. The ANF probe was a gift from C. Seidman (Harvard Medical School, Boston, Massachusetts, USA) (54). The full-length TNC probe was from an expressed sequence tag (dbEST ID 6838718). All other probes were developed as described previously (43). Section in situ hybridization was performed using 35S-labeled probes as previously described (43). Darkfield signal was colored in red and superimposed on brightfield H&GE- or DAPI-stained sections. Whole-mount in situ hybridization was performed using digoxigenin-labeled probes following previously established protocols (55).
Whole-mount X-gal staining for -galactosidase was performed as previously described (56). Immunofluorescent staining for Nfatc1 or Gata4 was performed using monoclonal antibodies (Santa Cruz Biotechnology Inc.) and the ABC method (MOM kit; Vector Laboratories) with tyramide-Cy3 as the HRP substrate. Desmin counterstaining was performed using a Desmin rabbit antibody (BioMeda Corp.) and anti-rabbit Alexa Fluor 488 (Invitrogen Corp.). Nuclei were stained with TOPRO3 (Invitrogen Corp.). Phosphohistone H3 was detected with a polyclonal antibody (Upstate) using the ABC method and tyramide-Cy3. BrdU labeling was achieved by injecting pregnant mice with 2 mg of BrdU intraperitoneally. After 2 hours, embryos were harvested, and BrdU was detected with a monoclonal antibody (Sigma-Aldrich) using the ABC method and tyramide-Cy3. Detection was facilitated by antigen retrieval with HCl and trypsin digestion per instructions of the manufacturer (Sigma-Aldrich). TUNEL staining was performed on paraffin sections using the TMR Red In Situ Death Detection kit (Roche Diagnostics).
qRT-PCR.
Hearts from E9.5 (G4NK) and E12.5 (G4MC) embryos were collected and frozen in liquid nitrogen. Hearts of each genotype (G4NK, n = 6; G4MC, n = 3) were pooled into a single sample, and RNA was prepared using the RNeasy kit with on-column DNase digestion (QIAGEN). Three samples were used per genotype. RNA integrity was verified and concentration determined using an Agilent Bioanalyzer (model 2100; Agilent Technologies). Real-time qRT-PCR was performed on an ABI 7700 Sequence Detector (Applied Biosystems) using Taqman probes (Gata4, Gata6, Nkx2-5, Carp, ANF, cyclin D1, cyclin D2, cyclin D3, Hop, and Bmp10, all from GenScript; and 18S from Applied Biosystems). For primer and probe sequences, see Supplemental Table 1.
Acknowledgments
The authors thank M. Rivera for excellent technical support and M. Zeisberg for critical reading of the manuscript. This work was funded by grants from the NIH National Heart, Lung, and Blood Institute (1 PO1 HL074734 and K08 HL004387-04). E.M. Zeisberg was funded by a Bayer fellowship grant from the German Society of Cardiology and by a grant from the Leopoldina Academy (BMBF-LPD 9901/8-105).
Footnotes
Seigo Izumo’s present address is: Novartis Institute for Biomedical Research, Cambridge, Massachusetts, USA.
Nonstandard abbreviations used: ANT, atrial natriuretic factor; BNP, brain natriuretic peptide; EMT, epithelial-to-mesenchymal transition; G4MC, Gata4flox/flox;MHCCre+; G4NK, Gata4flox/flox;Nkx2-5WT/Cre; MHC, myosin heavy chain; MHCCre; MHC promoter driving Cre recombinase expression; OFT, outflow tract; qRT-PCR, quantitative RT-PCR; TNC, tenascin C.
Conflict of interest: The authors have declared that no conflict of interest exists.
References
Hoffman, J.I. 1995. . Incidence of congenital heart disease: I. Postnatal incidence. Pediatr. Cardiol. 16::103-113.
Waldo, K.L. et al. 2001. . Conotruncal myocardium arises from a secondary heart field. Development. 128::3179-3188.
Mjaatvedt, C.H. et al. 2001. . The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238::97-109.
Kelly, R.G., Brown, N.A., and Buckingham, M.E. 2001. . The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell. 1::435-440.
Runyan, R.B., and Markwald, R.R. 1983. . Invasion of mesenchyme into three-dimensional collagen gels: a regional and temporal analysis of interaction in embryonic heart tissue. Dev. Biol. 95::108-114.
Srivastava, D., and Olson, E.N. 2000. . A genetic blueprint for cardiac development. Nature. 407::221-226.
Srivastava, D., Cserjesi, P., and Olson, E.N. 1995. . A subclass of bHLH proteins required for cardiac morphogenesis. Science. 270::1995-1999.
Srivastava, D. et al. 1997. . Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat. Genet. 16::154-160.
Benson, D.W. et al. 1999. . Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways. J. Clin. Invest. 104::1567-1573.
McElhinney, D.B., Geiger, E., Blinder, J., Benson, D.W., and Goldmuntz, E. 2003. . NKX2.5 mutations in patients with congenital heart disease. J. Am. Coll. Cardiol. 42::1650-1655.
Merscher, S. et al. 2001. . TBX1 is responsible for cardiovascular defects in velo-cardio-facial/DiGeorge syndrome. Cell. 104::619-629.
Lindsay, E.A. et al. 2001. . Tbx1 haploinsufficiency in the DiGeorge syndrome region causes aortic arch defects in mice. Nature. 410::97-101.
Basson, C.T. et al. 1997. . Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15::30-35.
Garg, V. et al. 2003. . GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature. 424::443-447.
Okubo, A. et al. 2004. . A novel GATA4 mutation completely segregated with atrial septal defect in a large Japanese family. J. Med. Genet. 41::e97.
Molkentin, J.D. 2000. . The zinc finger-containing transcription factors GATA-4, -5, and -6. Ubiquitously expressed regulators of tissue-specific gene expression. J. Biol. Chem. 275::38949-38952.
Charron, F., and Nemer, M. 1999. . GATA transcription factors and cardiac development. Semin. Cell Dev. Biol. 10::85-91.
Molkentin, J.D., Lin, Q., Duncan, S.A., and Olson, E.N. 1997. . Requirement of the transcription factor GATA4 for heart tube formation and ventral morphogenesis. Genes Dev. 11::1061-1072.
Kuo, C.T. et al. 1997. . GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11::1048-1060.
Narita, N., Bielinska, M., and Wilson, D.B. 1997. . Wild-type endoderm abrogates the ventral developmental defects associated with GATA-4 deficiency in the mouse. Dev. Biol. 189::270-274.
Crispino, J.D. et al. 2001. . Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15::839-844.
Pu, W.T., Ishiwata, T., Juraszek, A.L., Ma, Q., and Izumo, S. 2004. . GATA4 is a dosage-sensitive regulator of cardiac morphogenesis. Dev. Biol. 275::235-244.
Moses, K.A., DeMayo, F., Braun, R.M., Reecy, J.L., and Schwartz, R.J. 2001. . Embryonic expression of an Nkx2-5/Cre gene using ROSA26 reporter mice. Genesis. 31::176-180.
Mao, X., Fujiwara, Y., and Orkin, S.H. 1999. . Improved reporter strain for monitoring Cre recombinase-mediated DNA excisions in mice. Proc. Natl. Acad. Sci. U. S. A. 96::5037-5042.
Bruneau, B.G. et al. 1999. . Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev. Biol. 211::100-108.
Imanaka-Yoshida, K., Matsumoto, K., Hara, M., Sakakura, T., and Yoshida, T. 2003. . The dynamic expression of tenascin-C and tenascin-X during early heart development in the mouse. Differentiation. 71::291-298.
Xu, H. et al. 2004. . Tbx1 has a dual role in the morphogenesis of the cardiac outflow tract. Development. 131::3217-3227.
Gaussin, V. et al. 2002. . Endocardial cushion and myocardial defects after cardiac myocyte-specific conditional deletion of the bone morphogenetic protein receptor ALK3. Proc. Natl. Acad. Sci. U. S. A. 99::2878-2883.
Sherr, C.J., and Roberts, J.M. 1999. . CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 13::1501-1512.
Suzuki, Y.J. et al. 2003. . Activation of GATA-4 by serotonin in pulmonary artery smooth muscle cells. J. Biol. Chem. 278::17525-17531.
Chen, H. et al. 2004. . BMP10 is essential for maintaining cardiac growth during murine cardiogenesis. Development. 131::2219-2231.
Shin, C.H. et al. 2002. . Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell. 110::725-735.
Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., and Black, B.L. 2004. . Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development. 131::3931-3942.
Kuo, H. et al. 1999. . Control of segmental expression of the cardiac-restricted ankyrin repeat protein gene by distinct regulatory pathways in murine cardiogenesis. Development. 126::4223-4234.
Lien, C.L. et al. 1999. . Control of early cardiac-specific transcription of Nkx2-5 by a GATA-dependent enhancer. Development. 126::75-84.
Searcy, R.D., Vincent, E.B., Liberatore, C.M., and Yutzey, K.E. 1998. . A GATA-dependent nkx-2.5 regulatory element activates early cardiac gene expression in transgenic mice. Development. 125::4461-4470.
Brown, C.O. 3rd et al. 2004. . The cardiac determination factor, Nkx2-5, is activated by mutual cofactors GATA-4 and Smad1/4 via a novel upstream enhancer. J. Biol. Chem. 279::10659-10669.
Thompson, J.T., Rackley, M.S., and O’Brien, T.X. 1998. . Upregulation of the cardiac homeobox gene Nkx2-5 (CSX) in feline right ventricular pressure overload. Am. J. Physiol. 274::H1569-H1573.
Molkentin, J.D., Kalvakolanu, D.V., and Markham, B.E. 1994. . Transcription factor GATA-4 regulates cardiac muscle-specific expression of the alpha-myosin heavy-chain gene. Mol. Cell. Biol. 14::4947-4957.
Small, E.M., and Krieg, P.A. 2003. . Transgenic analysis of the atrialnatriuretic factor (ANF) promoter: Nkx2-5 and GATA-4 binding sites are required for atrial specific expression of ANF. Dev. Biol. 261::116-131.
Houweling, A.C., Somi, S., Van Den Hoff, M.J., Moorman, A.F., and Christoffels, V.M. 2002. . Developmental pattern of ANF gene expression reveals a strict localization of cardiac chamber formation in chicken. Anat. Rec. 266::93-102.
Watt, A.J., Battle, M.A., Li, J., and Duncan, S.A. 2004. . GATA4 is essential for formation of the proepicardium and regulates cardiogenesis. Proc. Natl. Acad. Sci. U. S. A. 101::12573-12578.
Tanaka, M., Chen, Z., Bartunkova, S., Yamasaki, N., and Izumo, S. 1999. . The cardiac homeobox gene Csx/Nkx2.5 lies genetically upstream of multiple genes essential for heart development. Development. 126::1269-1280.
Biben, C. et al. 2000. . Cardiac septal and valvular dysmorphogenesis in mice heterozygous for mutations in the homeobox gene Nkx2-5. Circ. Res. 87::888-895.
Tanaka, M. et al. 2002. . A mouse model of congenital heart disease: cardiac arrhythmias and atrial septal defect caused by haploinsufficiency of the cardiac transcription factor Csx/Nkx2.5. Cold Spring Harb. Symp. Quant. Biol. 67::317-325.
McFadden, D.G. et al. 2000. . A GATA-dependent right ventricular enhancer controls dHAND transcription in the developing heart. Development. 127::5331-5341.
McFadden, D.G. et al. 2005. . The Hand1 and Hand2 transcription factors regulate expansion of the embryonic cardiac ventricles in a gene dosage-dependent manner. Development. 132::189-201.
Firulli, A.B., McFadden, D.G., Lin, Q., Srivastava, D., and Olson, E.N. 1998. . Heart and extra-embryonic mesodermal defects in mouse embryos lacking the bHLH transcription factor Hand1. Nat. Genet. 18::266-270.
Riley, P., Anson-Cartwright, L., and Cross, J.C. 1998. . The Hand1 bHLH transcription factor is essential for placentation and cardiac morphogenesis. Nat. Genet. 18::271-275.
Yamagishi, H. et al. 2001. . The combinatorial activities of Nkx2.5 and dHAND are essential for cardiac ventricle formation. Dev. Biol. 239::190-203.
Weiss, M.J., Yu, C., and Orkin, S.H. 1997. . Erythroid-cell-specific properties of transcription factor GATA-1 revealed by phenotypic rescue of a gene-targeted cell line. Mol. Cell. Biol. 17::1642-1651.
Tanaka, H. et al. 2000. . GATA-1 blocks IL-6-induced macrophage differentiation and apoptosis through the sustained expression of cyclin D1 and bcl-2 in a murine myeloid cell line M1. Blood. 95::1264-1273.
Rylski, M. et al. 2003. . GATA-1-mediated proliferation arrest during erythroid maturation. Mol. Cell. Biol. 23::5031-5042.
Zeller, R., Bloch, K.D., Williams, B.S., Arceci, R.J., and Seidman, C.E. 1987. . Localized expression of the atrial natriuretic factor gene during cardiac embryogenesis. Genes Dev. 1::693-698.
Brent, A.E., Schweitzer, R., and Tabin, C.J. 2003. . A somitic compartment of tendon progenitors. Cell. 113::235-248.
Kupershmidt, S. et al. 1999. . Replacement by homologous recombination of the minK gene with lacZ reveals restriction of minK expression to the mouse cardiac conduction system. Circ. Res. 84::146-152.(Elisabeth M. Zeisberg, Qi)