The second heart field (SHF) is indicated to contribute to the embryonic heart development. However, less knowledge is available about SHF development of human embryo due to the difficulty of collecting embryos. In this study, serial sections of human embryos from Carnegie stage 10 (CS10) to CS16 were stained with antibodies against Islet-1 (Isl-1), Nkx2.5, GATA4, myosin heavy chain (MHC) and α-smooth muscle actin (α-SMA) to observe spatiotemporal distribution of SHF and its contribution to the development of the arterial pole of cardiac tube. Our findings suggest that during CS10 to CS12, SHF of the human embryo is composed of the bilateral pharyngeal mesenchyme, the central mesenchyme of the branchial arch and splanchnic mesoderm of the pericardial cavity dorsal wall. With development, SHF translocates and consists of ventral pharyngeal mesenchyme and dorsal wall of the pericardial cavity. Hence, the SHF of human embryo shows a dynamic spatiotemporal distribution pattern. The formation of the Isl-1 positive condense cell prongs provides an explanation for the saddle structure formation at the distal pole of the outflow tract. In human embryo, the Isl-1 positive cells of SHF may contribute to the formation of myocardial outflow tract (OFT) and the septum during different development stages.
It has been known for many years that the myocardium of the outflow tract is derived from an extracardiac region (Virágh & Challice 1973; Argüello et al. 1975; de la Cruz et al. 1977), but it is only recently that the extracardiac region contributing to the outflow tract myocardium has been described in both chick and mouse embryos (Kelly et al. 2001; Mjaatvedt et al. 2001; Waldo et al. 2001). Mjaatvedt et al. (2001) revealed that the anterior heart field (AHF) resides in the mesoderm surrounding the aortic sac, which is anterior to the primitive cardiac tube and provides myocardial cells of the outflow tract (OFT) in chick embryos. In the same year, another study demonstrated that the AHF contains the mesodermal centre of the branchial arches and splanchnic epithelial in the dorsal wall of the pericardial cavity and forms the OFT and right ventricle (RV) in mouse embryonic heart (Kelly et al. 2001). Meanwhile, another definition, secondary heart field was put forward, which is located in the pharyngeal mesoderm caudal to the OFT and only produces myocardium of the distal part of the OFT and the root of the large arteries (Waldo et al. 2001). Later, the second heart field (SHF) was defined as a progenitor cell population dorsal to the heart tube. Its cells contribute to the OFT, RV and parts of the left ventricle and inflow region (Buckingham et al. 2005; Laugwitz et al. 2008). Taken together, these studies prove that part of the heart segments originate from AHF or secondary heart field or SHF. Thus, the traditional cardiac crescent is named as the first heart field (FHF). Furthermore, it has been elucidated that the AHF, secondary heart field and SHF not only give rise to the myocardium of the atrium, part of left ventricle and interventricular septum (Cai et al. 2003; Verzi et al. 2005; Sun et al. 2007), but also differentiate into endocardium cells and smooth muscle cells (Verzi et al. 2005; Waldo et al. 2005). The identification of AHF, secondary heart field and SHF provides significant impact on our understanding of several kinds of congenital heart diseases such as double outlet right ventricle and the tetralogy of Fallot (Ward et al. 2005). However, the difference of boundary and contribution to the embryonic heart should be noticed between the SHF, AHF, and secondary heart field. The reason may come from the difference in animal species, observed stages and labeling methods. But SHF obviously defines wider scope than AHF and secondary heart field. Thus, in this study, we will choose the nomenclature SHF to discriminate with FHF. Recently, the studies mostly focus on signal regulating SHF including autocrine signal and signal from pharyngeal endoderm and cardiac neural crest (Black 2007; Dunwoodie 2007; Rochais et al. 2009; Vincent & Buckingham 2010; Zaffran & Kelly 2012). While less knowledge is available about SHF development of human embryo due to the difficulty of collecting embryos.
In this study, we intend to observe spatiotemporal distribution of SHF and its contribution to the development of the arterial pole of human embryonic heart. Our study would provide morphology foundation for the research of normal development mechanism of human embryonic heart.
Materials and methods
Collection, staging and tissue processing of embryos
Twenty-nine human embryos were collected after medically induced abortion from several hospitals of Taiyuan city in Shanxi province under the condition of patient informed consent and approval of Medical Ethics Committee of Shanxi Medical University. The embryos were fixed in an ice-cold mixture of methanol:acetone:water (2:2:1 v/v) for 24 h for immunohistochemical analysis (Ya et al. 1998). After fixation, embryos were dehydrated in a graded series of ethanol and embedded in paraffin, and serial frontal and transverse sections of 7 μm thickness were cut and mounted on poly-L-lysine-coated slides for Hematoxylin-Eosin and immunohistochemical staining. The collected embryos were graded according to the Carnegie series of developmental stages (O'Rahilly & Müller 1987). We collected 12 embryos of Carnegie stage (CS) 10 to CS13 (three embryos at each stage), 10 embryos of CS14, three embryos of CS15, and four embryos of CS16. Embryos from stage 10 to 16 correspond to 22 ± 1, 24 ± 1, 26 ± 1, 28, 32, 33 and 37 days post-ovulation, respectively.
After deparaffination, the sections were pretreated sequentially with 3% (v/v) H2O2 for 30 min and TENG-T (10 mmol/L Tris, 5 mmol/L ethylenedi-aminetetraacetic acid, 150 mmol/L NaCl, 0.25% gelatin, 0.05% Tween-20, and pH 8.0) for 15 min to reduce the endogenous peroxidase activity and non-specific antibody binding, respectively. The pretreated serial sections were incubated overnight alternatively with mouse monoclonal antibodies against Islet-1 (1:100, Isl-1, Developmental Studies Hybridoma Bank, USA), myosin heavy chain (1:1000, A4.1025, MHC; Upstate, USA) and α-smooth muscle actin (1:1000, IMMH-2, α-SMA; Sigma, USA); rabbit polyclonal antibody against Nkx2.5 (1:100; Wuhan Boster Biological Technology) and against GATA4 (1:50; Wuhan Boster Biological Technology). The sections followed by consecutive incubation with rabbit-anti mouse immunoglobulin G (IgG) (1:7500, non-commercial, gift of Professor W.H. Lamers, Department of Anatomy and Embryology, Academic Medical Center, Amsterdam, the Netherlands) (the sections stained by Nkx2.5 or GATA4 were incubated with phosphate-buffered saline (PBS) instead of rabbit-anti mouse IgG), goat-anti rabbit IgG (1:250, non-commercial, gift of Professor W.H. Lamers, Department of Anatomy and Embryology, Academic Medical Center, Amsterdam, the Netherlands), and rabbit peroxidase-antiperoxidase complex (1:750; Nordic, the Netherlands). Antibodies were diluted in PBS. All incubations were followed by three washes in PBS for 5 min each. Antibody binding was demonstrated by staining with 3,3-diaminobenzidine tetrahydrochloride and 3 mmol/L H2O2. Stained sections were taken rapidly through graded ethanols, cleared in xylene and mounted in resin.
The spatiotemporal distribution of Isl-1 positive SHF at CS10 to CS14 in human embryo
It is well known that after the cardiac tube stage, Isl-1 is a reliable marker for SHF (Cai et al. 2003; Sun et al. 2007), and Nkx2.5 and GATA4 are the key transcriptional regulators in the FHF and SHF (Dodou et al. 2004; Klaus et al. 2007; Prall et al. 2007). Hence, to investigate SHF spatiotemporal distribution pattern in human embryo, we intend to observe Isl-1, GATA4 and Nkx2.5 expression and the OFT development at the same stages.
In this study, the results showed that at Carnegie stage 10 (CS10), looped heart tube was composed of the OFT, primary ventricle, atrium and venous sinus. Its arterial pole was surrounded by the first branchial arch (Fig. 1A). The space between primitive pharynx and the dorsal wall of the pericardial cavity was filled with acellular matrix. Isl-1 positive cell stream in bilateral pharyngeal mesenchyme was continuous with the dorsal wall of the pericardial cavity (Fig. 1A, arrow and arrowhead), where transcription factors Nkx2.5 and GATA4 expressed weakly (Fig. 1B,C, arrow and arrowhead). Isl-1 positive cells also were observed in the branchial arch (Fig. 1A, asterisk). In the distal wall of the OFT, which reflected with the dorsal wall of the pericardial cavity, Isl-1 was strongly expressed (Fig. 1A, red arrow), but the markers of early differentiated cardiomyocytes, Nkx2.5 and GATA4 were weakly positive, and cardiomyocytes marker, MHC was also weakly expressed (Fig. 1B–D, red arrow). The simultaneous expression of Isl-1, Nkx2.5 and GATA4 and MHC in the distal wall of OFT indicated that Isl-1 positive cells could differentiate into cardiomyocytes to take part in the myocardium development of the distal pole of the OFT. For the mature myocardial cells of the OFT, the expression of MHC, Nkx2.5, and GATA4 were upregulated, but the Isl-1 expression was downregulated (Fig. 1A–D, blue arrow).
Between CS11 and CS12, the aortic arch arteries (AAs) 1, 2, 3 became visible and joined into the aortic sac (Fig. 2A,B). The arterial pole of the cardiac tube was still located at the branchial arch level (Fig. 2B). At CS11, the Isl-1 positive cells accumulated to form the condensed mesenchymal core of the branchial arches (Fig. 2A–C, arrowhead), which were continuous with those cells on the aortic sac ventral wall (Fig. 2B,C, arrow). Meanwhile, the aortic sac ventral wall began to express Nkx2.5 and GATA4 weakly (Fig. 2D,E, arrow). It is well known that the OFT directly connects with the aortic sac. The difference between their walls is that the former contains myocardium marked by MHC, but the latter is a mesenchymal structure. According to our results, Isl-1, Nkx2.5 and GATA4 positive expression in the aortic sac ventral wall indicate that Isl-1 positive cells may differentiate into myocardial cells to make the aortic sac wall myocardialized and added to the OFT wall. The Isl-1 positive cell stream from the mesenchymal core of branchial arches 1, 2 also was continuous with the distal wall of the OFT (Fig. 3A,B). Taken together, the condensed core of branchial arches may provide the Isl-1 positive mesenchymal cells for the aortic sac and the OFT. Compared with Isl-1 expression, Nkx2.5 or GATA4 was expressed sporadically in the branchial arches without forming the mesenchymal core (Fig. 2B–E), and was expressed weakly in the dorsal wall of the pericardial cavity (Fig. 2D,E, arrowhead).
At CS13, the shrinkage of the endoderm led to the gap formation between the pharynx and its ventral mensenchyme, which was not observed between the pharyngeal endoderm and its bilateral mesenchyme (Fig. 4A–C). Meanwhile, part of the lateral pharyngeal endoderm lost its basal membrane and some endodermal cells became small and round and dispersed into neighboring Isl-1 positive mesenchymal cells (Fig. 4C,D, frame). Differ with CS11, Isl-1 positive cells mainly distributed in the bilateral and ventral mesenchyme of the primitive pharynx (Fig. 4A,B). Isl-1 expression also was observed in the dorsal wall of the pericardial cavity, which was closely adjacent to the ventral pharyngeal mesenchyme and reflected with the OFT myocardial wall (Fig. 4A,B, arrow). At CS14, AAs 4, 6 could be identified and linked with the aortic sac (Fig. 5A). Due to the arterial pole of the embryonic heart translocated caudally to the branchial arch (Fig. 5B), the Isl-1 positive cells in mesenchymal core of the branchial arches 3, 4, 6 lost direct connection with the dorsal wall of the pericardial cavity and the OFT (Fig. 5A). At this stage, the trachea had formed from the foregut. Many Isl-1 positive cells concentrated to form a cone area in the bilateral and ventral tracheal mesenchyme (Fig. 5A, asterisk, B), which extended caudally and became small gradually (Fig. 5G, asterisk). Isl-1 positive cells were also observed in the dorsal wall of the pericardial cavity and continuous with those in the distal pole of the OFT wall (Fig. 5B, C). Sister sections staining showed that GATA4, Nkx2.5 and MHC were expressed weakly in the distal pole of the OFT wall (Fig. 5D–F). These findings suggested that the Isl-1 positive cells continued to provide myocardial cells for the OFT development, which is similar with CS13. However, no GATA4 or Nkx2.5 positive cone area was observed in ventral tracheal mesenchyme (data not shown).
SHF cells formed two condensed prongs in the OFT
It is well known that the OFT wall is composed of myocardium, endocardium and cardiac jelly between them. With development, mesenchymal cells migrate into the cardiac jelly and contribute to form endocardial cushion. In this study, accompanied with differentiation into the cardiomyocytes of the OFT, part of Isl-1 positive cells concentrated in the cardiac jelly of the caudal wall of the OFT and formed Isl-1 positive condensed cell prong (Fig. 5B, C, asterisk). Meanwhile, some Isl-1 positive cells formed another condensed cell prong in the cranial wall of the OFT (Fig. 5G, H, arrow). Hence at this stage, two prongs faced each other and extended from the distal pole to the proximal pole of the OFT (Fig. 5C, H, I). It must be ensured that the cells in the prongs are stained negatively except Isl-1 (Fig. 5C–J). At the same time, many mesenchymal cells filled with the cardiac jelly and formed the two endocardial cushions that neighbored the two prongs (Fig. 5I, asterisk). Sister sections showed that part of the cells in the cushion expressed α-SMA (Fig. 5J).
Isl-1 positive SHF cells distributed in the artery wall and the aortic-pulmonary septum
At CS15 to CS16, the Isl-1 positive cone area still was obvious (Fig. 6A asterisk). The length of the OFT did not increase any more. Aortic-pulmonary septum formed and septated the aortic sac into ascending aorta and pulmonary trunk (Fig. 6A, B). The two arterial walls contained many Isl-1 positive cells (Fig. 6A–C), and part of them were continuous with those of the cone area (Fig. 6C, arrow). A sister section showed that a few α-SMA positive smooth muscle cells were formed in the two arteries (Fig. 6D, E, arrow). We also observed the sporadic Isl-1 positive cells in the aortic-pulmonary septum, which were adjacent with the tip of the cone area (Fig. 6B).
SHF of the human embryo shows dynamic spatiotemporal distribution pattern
In chick and mouse embryos, it has been demonstrated that the SHF distributes in splanchnic mesenchyme adjacent to foregut endoderm, which contributes to the myocardium of the OFT and right ventricle (Cai et al. 2003; Verzi et al. 2005; Sun et al. 2007). However, less knowledge is available about SHF of human embryo (Abu-Issa et al. 2004). In this study, between CS10 to CS12, Isl-1 positive cells were observed in the bilateral pharyngeal mesenchyme, the mesenchymal core of the branchial arches and splanchnic mesoderm of the pericardial cavity dorsal wall. Since Isl-1 could be used to mark SHF progenitor cells after cardiac tube formation, we could conclude that during CS10 to CS12, SHF of the human embryo consists of the Isl-1 positive area we mentioned above. In this year, another study reported that Isl-1 expression could be observed outside of the endoderm of the ventral foregut after CS12 in human embryo, which is later than in mouse embryo (Golzio et al. 2012). But our results showed that Isl-1 had begun to express in bilateral pharyngeal mesenchyme and splanchnic mesoderm at CS10, corresponding to 8.5–9.0 day mouse embryo. Hence, we could draw a conclusion that the SHF distribution pattern at the early stage in human embryo is coincident with that in the mouse and chick embryo (Cai et al. 2003; Tirosh-Finkel et al. 2006; Sun et al. 2007; Nathan et al. 2008).
At CS13 and CS14, the splanchnic mesoderm of the pericardial cavity dorsal wall still showed Isl-1 positive. In the bilateral and ventral pharyngeal mesenchyme, Isl-1 positive cells dramatically increased and formed a cone area. It has been proved that SHF proliferation requires some signals regulating from pharyngeal endoderm such as sonic hedgehog (Shh) (Dyer & Kirby 2009). According to our results, at CS13, some pharyngeal endodermal cells showed morphological change and mingled with neighboring Isl-1 positive mesenchymal cells. The results suggest that pharyngeal endoderm cells might differentiate into the Isl-1 positive mesenchymal cells. That means the role of the pharyngeal endoderm for SHF proliferation is not limited to providing signals. At CS14, the Isl-1 positive mesenchymal core of branchial arches 4, 6 failed to continuous with the OFT wall. The findings indicate that the branchial arch mesenchymal core, part of SHF, does not include all of the branchial arches, but only the first three. At CS16, Isl-1 positive cone area became more obvious. The data indicate that from CS13, different with CS10 to CS12, SHF consists of a Isl-1 positive cone area and splanchnic mesoderm of the dorsal wall of the pericardial cavity. That means, consistent with the primary cardiac tube translocation to thorax, the distribution pattern of SHF also changes. Thus SHF could be thought as a dynamic structure.
SHF contributes to the development of the distal wall of the OFT in human embryo
Our previous results and other reports have revealed that in human embryo at CS14, the OFT wall is composed of MHC positive myocardium, endocardium and endocardial cushion between them. Some mesenchymal cells in the cushion show α-SMA positive (Moorman et al. 2003; Yang et al. 2009). Thus the endocardial cushion could be identified according to its morphology and α-SMA expression. Our results showed that at CS14, the two prongs were formed and faced with each other on the OFT wall. Furthermore, part of the mesenchymal cells in the prong express α-SMA. It could be confirmed that they are the OFT endocardial cushions. But at the same time, another pair of prongs was shown on the OFT wall, which coexisted with the endocardial cushion. But the prong only expressed Isl-1 and was devoid of cardiomyocyte or α-SMA positive mesenchymal cell. Hence, the function of the prong should be different with the endocardial cushion. As far as we know, the prongs have not been reported by others. Three-dimensional reconstruction shows that the distal margin of the MHC positive myocardial OFT is a saddle-shaped structure with two peaks laterally separated by a cranial and a caudal depression (Bartelings & Gittenberger-de Groot 1989; Ya et al. 1998), but the formation mechanism has not been elucidated. Our findings showed that at the distal pole, part of the OFT wall was devoid of myocardium and was accurately filled with the two prongs. It is obvious that the prong localization should be the position of the depression. Hence, our results provide an explanation for the saddle-shaped structure formation and also prove that Isl-1 positive cells take part in the distal part of development of the OFT. However, further work is required to clarify the fate of the Isl-1 positive cell in the prong.
The function of Isl-1 positive cells of SHF in human embryo
During CS10 to CS14, Isl-1 positive progenitor cells of SHF contribute to myocardial cells formation to make OFT lengthen. In CS15 and CS16 human embryo, the aortic sac was septated into ascending aorta and pulmonary trunk by aortic-pulmonary septum. Our results showed that Isl-1 positive cells distributed in the two artery walls. Meanwhile, α-SMA positive smooth muscle cells also formed in the artery wall. The findings imply that at these stages, Isl-1 positive cells of SHF might contribute to differentiate into smooth muscle cells. At the same stages, the Isl-1 positive cells also distributed in the aortic-pulmonary septum. It has been demonstrated that the septum originates from cardiac neural crest cells (Jiang et al. 2000; Engleka et al. 2005; Nakamura et al. 2006). Based on our results, only a few of cells in the septum expressed Isl-1. Thus, it needs to be studied further whether part of cardiac neural crest derived cells express Isl-1 or cells of SHF also participate in the aortic-pulmonary septum formation. As mentioned above, during CS10 to CS14, Isl-1 positive SHF provides progenitor cells for myocardium formation in the distal wall of OFT. At CS15 and CS16, Isl-1 positive cells distributed in the artery wall and aortic-pulmonary septum, which might contribute to differentiate into smooth muscle cells. Taken together, our results provide significant evidence that the distribution and differentiation potent of the Isl-1 positive SHF embody spatiotemporal specificity.
Nkx2.5 and GATA4 expression pattern in SHF of human embryo
It has been demonstrated that Nkx2.5 and GATA4 are the key transcription factors in FHF and SHF (Dodou et al. 2004; Klaus et al. 2007; Prall et al. 2007). Isl-1 is an early nodal point and could regulate Nkx2.5 and GATA4 during SHF development (Black 2007; Nathan et al. 2008). In this study, we noticed that compared with Isl-1, Nkx2.5 or GATA4 expression was confined in part of SHF such as splanchnic mesoderm of pericardial cavity dorsal wall. And also no evidence has been provided for Nkx2.5 and GATA4 positive cells contributing to the formation of ascending aorta and pulmonary trunk root wall and the aortic-pulmonary septum. In the differentiated cardiomyocytes of the OFT wall, Nkx2.5 and GATA4 kept strong expression from CS10 to CS16. Another study reported that the time window of GATA4 expression in the OFT wall of human embryo was from CS13 to CS15 (Golzio et al. 2012), which is shorter than the time in our result. In has been demonstrated that Nkx2.5 or GATA4 mutant embryos show a hypoplastic OFT (Kuo et al. 1997; Molkentin et al. 1997; Prall et al. 2007). Taken together, it might imply that Nkx2.5 and GATA4 regulation on differentiation from progenitor cell into myocardial cell of the SHF in the embryo early development. During later development, Nkx2.5 and GATA4 might promote progressive maturation of differentiated myocardial cells after migration into the OFT wall. Downregulation of Isl-1 expression in mature cardiomyocytes suggest that Isl-1 only contributes to the differentiation from the mesenchymal cell to the myocardial cell and does not play an important role for progressive maturation of differentiated myocardial cells.
This work was supported by the National Natural Science Foundation of China (30771141), the National Natural Science Foundation of China (31200899) and Shanxi Province Foundation for Returness (2008–47). We acknowledge the doctors and nurses of the Department of Gynaecology and Obstetrics, Shanxi Children's hospital for the human embryo collection. We thank Professor W. H. Lamers for his kind gift of the antibodies.
Jing Ya was responsible for the study design, and was involved in the results analysis and manuscript preparation. Yan-Ping Yang analyzed the results, performed photomicrography and wrote the initial draft of the manuscript. Hai-Rong Li performed immunohistochemistry staining of the sections and was involved in manuscript preparation. Xi-Mei Cao participated in the embryos collection and the paraffin section preparation. Qin-Xue Wang collected the human embryos. Cong-Jin Qiao contributed to the fixation and embedding of embryos.