Congenital heart malformations affect almost eight of 1,000 live births (Šamánek, 2000). Bone morphogenetic proteins (BMPs) are considered to be of significant importance in the regulation of multiple aspects of cardiogenesis (for review see Brand, 2003). However, genetic evidence is limited because functional impairment of BMP signaling in mice results in embryo lethality prior to heart development, or in mice without cardiac defects (for reviews see Chang et al., 2002; Zhao, 2003). To gain insight into the potential roles of BMPs during cardiac development, we obtained a comprehensive description of the spatiotemporal expression pattern of BMP2, -4, -5, -6, and -7 mRNA during chicken heart development (Somi et al., in press). To identify known and novel BMPs that are expressed in the developing heart, we used a polymerase chain reaction (PCR) approach (Somi et al., 2003) and in silico cloning using the chicken Est databases. From these databases, we identified BMP10, determined its spatiotemporal expression pattern, and compared this to the initial description, using whole-mount in situ hybridization (Teichmann and Kessel, 2004).
In mice, BMP10 mRNA is expressed only in the developing heart. In the heart, BMP10 mRNA is expressed in the trabeculated part of the ventricles and the proximal outflow tract at days 9.0–14.5 of development. In the developing atria, BMP10 mRNA expression is observed concomitantly with the appearance of trabecules at 12.5 days of development (Neuhaus et al., 1999). An initial report on the functional disruption of BMP10 revealed embryo lethality due to severe cardiac defects, including hypoplastic ventricular wall, absence of ventricular trabeculation, and impaired contractile function (Chen et al., 2003).
MATERIALS AND METHODS
Fertilized chicken eggs were obtained from a local hatchery (Drost BV, Nieuw Loosdrecht, The Netherlands), incubated at 38.5°C in a moist atmosphere, and automatically turned every hour. After the appropriate incubation times, the embryos were isolated and staged according to Hamburger and Hamilton (HH) (Hamburger and Hamilton, 1951). The embryos were fixed in 4% paraformaldehyde dissolved in phosphate-buffered saline (PBS) for 4 hr. Nonradioactive in situ hybridization was performed on whole chicken embryos ranging from HH4 to HH18, and on serial sections ranging from HH8 to HH44, as recently described (Moorman et al., 2001).
Chicken cDNA (604156553F1), which is similar to human and mouse BMP10, was identified in the UMIST Est database (Boardman et al., 2002) and obtained from the MRC geneservice (Cambridge, UK). Cardiac Troponin I (cTnI) was used to visualize the cardiomyocytes (Houweling et al., 2002). Digoxigenine-labeled antisense RNA (Roche, Almere, The Netherlands) was produced by in vitro transcription according to the manufacturer's instructions.
Images were taken with a digital Nikon Coolpix 950 or an Olympus DP12 camera coupled to a Leica MZFLIII stereomicroscope or a Zeiss Axiophot microscope equipped with differential interference contrast (DIC) optics. A flatfield correction was applied to all images with the use of a user-written macro in PMIS 4.1 (www.gkrcc.com). Contrast and color saturation were adjusted by means of the Levels and Hue/Saturation functions of Adobe Photoshop 5.0LE.
RESULTS AND DISCUSSION
The BMP10 cDNA clone was sequence-verified, and comprised 1,641 bp. The cDNA clone encodes the mature part of BMP10 protein and 1,294 bp of the 3′ untranslated region of the mRNA. The TGF-β family ligands are translated as prepropeptide precursors with an N-terminal signal peptide followed by the prodomain and the mature domain (for review see Chang et al., 2002). The mature domain of the chicken BMP10 protein showed 89% identity (92% homology) and 87% identity (92% homology), respectively, with mouse and human BMP10 mature protein. A nucleotide comparison revealed an 80% sequence identity within this region of mouse and human BMP10. Blasting the protein sequence within Genbank showed that chicken Dorsalin-1 is the closest relative within the TGF-β superfamily showing 72% identity (85% homology). In a previous study (Chang et al., 2002), a progressive cluster multiple-sequence alignment of known mouse and human mature BMPs showed that BMP10 clusters with BMP9. A comparison of chicken BMP10 with human BMP9 revealed only 68% similarity (80% homology). Taken together, these results indicate that the identified chicken cDNA clone most probably encodes chicken BMP10.
In contrast to a previous report by Teichmann and Kessel (2004), BMP10 mRNA can be detected 12 hr earlier than that described by Techmann and Kessel (2004), during chicken development at HH10, when the linear heart tube is formed (Fig. 1A and B). BMP10 mRNA is not expressed throughout the myocardium of the entire heart, but is restricted to the myocardium of the arterial pole, i.e., anterior to the interventricular groove (Fig. 1C and D). At HH14, BMP10 mRNA is expressed in the entire heart, with the exception of the atrioventricular canal (Fig. 1E). Serial sections of HH16 hearts show that BMP10 mRNA is expressed throughout the entire myocardial wall of the heart (Fig. 1F and G). The observed expression pattern is much broader than that described by Teichmann and Kessel (2004), who observed it only in the inner myocardial layer of the developing ventricles. At this stage, BMP10 mRNA expression is not observed in the atrioventricular canal and pro-epicardium (Fig. 1F and G). At HH22, BMP10 mRNA expression starts to decrease in the outflow tract and inflow tract myocardium, and becomes confined to the trabeculated myocardium of the ventricles and atria. From HH26 onward, a transmural gradient is observed. This gradient is highest at the luminal side of the ventricles, and decreases toward the compact myocardium of the ventricular free wall and the interventricular septum, in which BMP10 mRNA is barely visible (Fig. 1H–K). Subsequently, BMP10 mRNA expression also disappears from the trabecules, being longest expressed in the most distal myocardium of the tips of the trabecules (Fig. 1L and M). Furthermore, while the ventricular expression of BMP10 mRNA decreases to nondetectable levels, intense staining remains in the trabeculated myocardium of the atria, the interatrial septum (Fig. 1J–O), and the venous valves (Fig. 1L and M), which was not previously reported (Teichmann and Kessel, 2004). At HH44, BMP10 mRNA is no longer detectable in the ventricular myocardium, but it becomes apparent in the endocardium of the ventricles (Fig. 1N and O).
In contrast to the mouse, BMP10 mRNA in chicken is not restricted to the heart, but is also expressed in the endothelium of the sinusoids of the liver from HH22 onward (data not shown). Since 9.0 days of mouse development is comparable to HH16 in the chicken (Pexieder, 1978), the onset of BMP10 mRNA expression occurs later in the mouse heart than in the chicken heart (Neuhaus et al., 1999). Nevertheless, the expression of BMP10 mRNA in the developing ventricle and outflow tract is similar in mouse and chicken. The expression of BMP10 mRNA in the atria is observed from the start of its development in chicken (Fig. 1E), whereas in mouse the expression of BMP10 mRNA is first observed at 12.5 days of development (comparable to HH27 in chicken). Because the mouse expression pattern is only described up to 14.5 days of development, being comparable to HH30, the disappearance of BMP10 mRNA from the mouse ventricular myocardium has not been observed (Neuhaus et al., 1999) and remains to be determined.
The observed expression pattern of BMP10 mRNA during heart development in mouse and chicken supports the notion that it plays a role in the development of the ventricular wall and trabecules, as previously observed in BMP10 knockout mice (Chen et al., 2003). Cardiomyocyte-specific deletion of the BMP receptor 1A (Alk3) has been reported to lead to underdevelopment of the ventricular compact myocardium and trabecules (Gaussin et al., 2002), which suggests that the myocardially produced BMP10 supports chamber development in an autocrine loop via Alk3. Furthermore, the idea that BMP10 plays a role in the initiation of chamber formation in chicken (Teichmann and Kessel, 2004) is not supported by its expression pattern, because the formation of chamber myocardium becomes apparent as early as HH9– when atrial natriuretic factor mRNA is used as a marker for the developing chamber myocardium (Houweling et al., 2002).