Lab. de Investigación en Biología del Desarrollo y Teratogénesis Experimental, Hospital Infantil De México Federico Gómez, Dr. Márquez 162, Col. Doctores. C.P. 06720, México D. F
Lab. de Investigación en Biología del Desarrollo y Teratogénesis Experimental, Hospital Infantil de México Federico Gómez. Dr. Márquez # 162, Col. Doctores, Del. Cuauhtémoc. CP 06720, México, D. F. México
The septal leaflet of the right atrioventricular valve is anatomically homologous in humans and chicks (Cayré et al.,1993; Aydinlioglu and Ragbetli,1998; De la Cruz and Markwald,1998). However, unlike in humans, the right septal leaflet (RSL) in chicks consists of a great fibrous ridge and multiple microleaflets and bulges (Cayré et al.,1993; De la Cruz and Markwald,1998). In both species, the leaflet is constituted by connective tissue and is directly inserted in the interventricular septum by short fine fibrous tendinous cords (TCs) without intermediation of papillary muscles (Silver et al.,1971; Cayré et al.,1993; De la Cruz and Marwald,1998). However in humans, the antero-septal commissure of the tricuspid valve is supported by the medial papillary muscle or Lancisi's muscle (Wenink,1977; Restivo et al.,1989).
In the adult human heart, the TCs are thought to be formed exclusively by fibrous connective tissue. They have been classified based on their points of attachment (Chiechi et al.,1956; Frater and Ellis,1961; Morse et al.,1984). First order TCs insert into the free edge of the leaflet. Second order TCs attach to the ventricular aspect of the leaflet a few millimeters from the free edge. Third order TCs extend from the papillary muscles to the ventricular wall.
Development of the atrioventricular (AV) valve leaflets and TCs is a complex event, as reflected by the high incidence of congenital defects involving the definitive valves (Hoffman and Kaplan,2002). The origin of the AV leaflets and their supporting structures (TCs and papillary muscles) has been a matter of controversy. The myocardial sleeve of the ventricles is thought to delaminate through an undermining process and then to be spliced into two layers, resulting in freely movable leaflets (Van Mierop et al.,1962; Van Gils,1979; Wenink et al.,1986a; Wenink,1992; Lamers et al.,1995; Oosthoek et al.,1998; De la Cruz and Markwald,1998). However, some researchers do not fully accept this idea. De Lange et al. (2004) suggested that in mouse embryos delamination is only associated with apoptosis during formation of the RSL. In contrast, Lincoln et al. (2004), studying embryonic chicken and mouse hearts, discarded the possibility of delamination in AV valvulogenesis. Morse et al. (1984), Wenink et al. (1986a,b), and Lamers et al. (1995) suggested that the internal myocardial layer of the ventricles contributes at least to some part of the connective tissue of the leaflets and gives origin to the TCs. Alternatively, Morse et al. (1984) and Oosthoek et al. (1998) pointed out that TCs, AV leaflets, and papillary muscles develop as a unit.
Pioneering in vivo labeling experiments performed by placing carbon particles or injecting India ink in the AV endocardial cushions in chick embryos (De la Cruz et al.,1983; García Peláez et al.,1984; Oosthoek et al.,1998) confirmed that the AV valve leaflets and TCs are derived from mesenchymal cells of the endocardial cushions, as had been classically described (Van Mierop et al.,1962; Ugarte et al.,1976; Chin et al.,1992; Wessels et al.,1996). More recently, this hypothesis was corroborated by genetic lineage-labeling in transgenic mice (Lincoln et al.,2004; De Lange et al.,2004). Furthermore, it has been established that formation of the endocardial cushions is mediated by endocardial-to-mesenchymal transformation (Markwald et al.,1977; Fitzharris TP,1981; Mjaatvedt and Markwald,1989; Eisenberg and Markwald,1995; Krug et al.,1995; Markwald et al.,1996; Lincoln et al.,2004; De Lange et al.,2004).
The understanding of the morphogenetic and molecular regulation of this process continues to improve (Barnett and Desgrosellier,2003; Armstrong and Bischoff,2004; Wang et al.,2005, Sakabe et al.,2006,2008). However, the origin and fate of myofibroblasts and myocardium embedded in the immature mesenchyme of the AV endocardial cushions at the beginning of AV valvulogenesis are still unclear. Some authors suggest that myocardium results from mesenchyme-to-myocardium transformation (Van den Hoff et al.,2001). In contrast, other researchers have concluded that these cells come from the myocardial ventricular wall and the interventricular septum (IVS) (Lamers et al.,1995; De Lange et al.,2004).
Almost all of this knowledge has been derived from studies of the mural AV leaflet assuming that both septal and mural leaflets have the same pattern of development. However, the supporting structures of the two types of leaflets are anatomically different and this could reflect some differences in the pattern of development. In this article, we have, therefore, investigated the morphogenetic processes involved in sculpting and maturation of the right AV septal leaflet and TCs in embryonic and post-hatching chicken hearts using scanning electron microscopic analysis and classic histological studies. Based on the fact that Nkx2.5 and MEF2C are related with respect to the early cardiomyogenic specification (Lints et al.,1993; Lough and Sugi,2000; Dodou et al.,2004; Creemers et al.,2006), the origin and fate of the myocardium embedded in the immature cushion mesenchyme at the beginning of RSL morphogenesis were studied by immunohistochemical detection of both nuclear proteins.
MATERIAL AND METHODS
Fertile Hi Line hen eggs were obtained from a local hatchery (ALPES Tehuacán, Pue, México). They were incubated at 37°C in a moist atmosphere to acquire embryos at Hamburger and Hamilton's stages (1951) 28 (4 ½ days), 30 (6½ to 7 days), 36 (10 days), 38 (12 days), 40 (14 days), 42 (16 days), 44 (18 days), 46 (21 days). One and two weeks post-hatched chicks were also obtained. Based on the labeling studies in the chick embryo, De la Cruz et al. (1983) suggested that the right horn of the inferior AV cushion gives origin to the RSL, the morphological changes of this region of the embryonic heart were studied by scanning electron microscopy (SEM), histological analysis, and immunohistochemistry.
Scanning electron microscopy
Hearts of embryos (ranging from stage 28 to 46HH) and post-hatched chickens (one and two weeks) were dissected to expose the inferior AV cushion right horn (Stages 28–40 HH) or the right surface of the posterior middle third of the IVS (Stage 42–46HH and post-hatched). Ten specimens of each age were fixed with 2.5% glutaraldehyde in 1 M sodium cacodylate, pH 7.2 for 2 h at room temperature. Subsequently, the samples were dehydrated, desiccated under liquid CO2 with a critical-point drying apparatus (Samdri 789A) and sputter coated with 35 nm of gold in a Denton Vacuum Desk 1A apparatus. Photographs were taken using a JEOL scanning electron microscope JSM 5300 at 15 kV and at a magnification of 100×.
Embryonic and post-hatched hearts at the same ages used for the SEM studies were fixed overnight in Bouin Dubosq (Humason1979). Five specimens of each age were dehydrated through graded ethanol and embedded in paraplast (Oxford Labware). Frontal serial 5 μm sections were made and stained with the Masson's trichrome technique. Photographs were taken using a Zeiss DSM 960 stereoscope equipped with an Axiocam digital camera or in an Olympus BH-2 RFCA optical microscope using a Nikon Coolpix 4500 digital camera.
Hearts of embryonic (HH stage 28–46) and post-hatched (one and two weeks) chicks were fixed overnight with 3.7% formaldehyde in phosphate buffered saline solution (PBS), pH 7.2 at 4°C.The samples were dehydrated and embedded in paraplast (Oxford Labware), serial 5 μm sections were made from five hearts of each stage and mounted onto poly-L-lysine-coated slides. After deparaffination and rehydration, the slides were rinsed with PBS and antigen retrieval was achieved by autoclaving (5 min at 20 pounds of pressure) in a target retrieval solution (Dako S2369). In the case of Nkx2.5, inhibition of endogenous peroxidase activity was attained by treatment with 3% hydrogen peroxide (5 min). After three rinses with PBS, the slides were immersed in protein block serum-free solution (Dako X0909) for 5 min and incubated with the primary antibody (goat polyclonal IgG, N-19, sc-8697, Santa Cruz, 1:100) for 20 min. They were immediately incubated with the biotinylated secondary antibody (20 min) and treated with streptavidin-HRP (20 min) and DAB+chromogen (3 min) (LSAB+System-HRP Kit IVD Dako K0679). In the case of MEF2C, after treatment with protein block serum-free solution (Dako X0909) for 5 min, the slides were incubated for 20 min with the primary antibody (rabbit polyclonal Ig G, C-21, sc-313, Santa Cruz, 1:100) and rinsed with PBS (3×). They were incubated with the alkaline phosphatase labeled secondary antibody (20 min) and fast red (1–3 min) (EnVision System-AP Kit IVD, Dako K1396). In both cases, after several washes, the preparations were counterstained with Harris' hematoxylin and mounted to take micrographs using an Olympus BH-2 RFCA optical microscope equipped with a Nikon Coolpix 4500 digital camera.
The morphological studies revealed that at stage 28HH the inferior and superior AV cushions were completely fused and in contact with the interventricular septum at the inlet region (Fig. 1A). This condition determined the presence of an interventricular orifice at the level of the outlet (Fig. 1E). The inferior AV cushion right horn had the form of an incipient ridge. It was located between the interventricular and interatrial septa and was extended from the anterior to the posterior right AV commissures (Fig. 1E). The incipient ridge was composed of loose mesenchymal tissue consisting of abundant extracellular matrix and homogeneously spaced fibroblast-like cells (Fig. 1I). Immunohistochemistry showed only the ventricular and atrial myocardium expressing Nkx2.5 (Fig. 1A,M) and MEF2C (Fig. 1Q). Mesenchymal tissue and endocardium at the incipient ridge did not express these nuclear proteins. Between stages 30HH and 38HH, the inferior AV cushion ridge became a mammillae structure (Fig. 1F–H) composed of mesenchyme which was organized into two not well delimited strata. The atrial stratum faced the right atrium. It was composed of densely packed small round cells immersed in scarce extracellular matrix. During development, this stratum gradually became more compact (Fig. 1J,K,L). The ventricular stratum was facing the right ventricle. At stage 30HH it was composed of abundant extracellular matrix in which spaced fibroblast-like cells were observed (Fig. 1J). At stage 36HH, small groups of Nkx2.5 and MEF2C expressing myocardial cells intermingled with fibroblasts-like cells were observed only at the contact zone of the ventricular stratum and the IVS (Fig. 1K,O,S). At the end of this period (stage 38HH), myocytes formed a nearly compact tissue occupying almost all of the not well defined ventricular stratum (Fig. 1L,P,T). Interestingly, from stage 30HH to the end of the study (two weeks post hatched chicks), a few endocardial cells covering the ridge and the leaflets expressed MEF2C (Figs. 1R,S,T, 2O, 3M).
By SEM it was found that the posterior third of the mammillae ridge exhibited some depressions that resulted in the progressive manifestation of two to four incipient lamellar or mammillae leaflets and TCs (Fig. 2D,E,F). Histological results showed first order endothelial TCs and second order muscular TCs, respectively, inserted in the free edge and in the ventricular face of the leaflets (Fig. 2G–I). The histological features of the atrial stratum of the mammillae ridge were completely different than those of the ventricular stratum. The atrial stratum had a lamellar or leaflet appearance. It was comprised by mesenchymal tissue with not expressing Nkx2.5 or MEF2C fibroblastic-like cells immersed in extracellular matrix (Fig. 2G–O). In contrast, the ventricular stratum was formed almost exclusively of myocardial cells expressing both Nkx2.5 and MEF2C (Fig. 2G–O). The boundary between the myocardium of the IVS and the ventricular stratum was always inconspicuous. In spite of this, from the beginning of this period (stage 40HH), the myocardium of the ventricular stratum was forming small myocardial bundles covered by endocardium that were more individualized during development (Fig. 2G–O). Simultaneously, the mesenchymal tissue of the atrial stratum directly opposing the ventricular stratum became grouped along the myocardial bundles. At the end of this period (stage 44HH), the adjoining mesenchymal and myocardial bundles had formed tendinous cord-like structures with connective tissue adjacent to the leaflets and myocardium in contact with the IVS (Fig. 2I,L,O).
Stage 46HH Embryos, One and Two Weeks Post-Hatched Chicks (Fig. 3)
By stage 46HH, well-formed lamellar and mammillar leaflets histologically similar to those found at stage 44HH were observed (Fig. 3D,G,J,M). In spite of this, the TCs presented significant histological changes during this period. The TCs inserted at the free boundary of the leaflets (first order), as well as those inserted at the ventricular aspect (second order), were composed of two types of tissues. At the leaflet insertion the TCs contained dense connective tissue not expressing Nkx2.5 or MEF2C. On the contrary, at the zone adjacent to the IVS, the TCs were composed of well organized myocardial bundles expressing Nkx2.5 and MEF2C (Fig. 3G–O). On the other hand, the TCs located at the intersection of the leaflet and IVS (third order TCs) were comprised exclusively by well differentiated myocardium (Fig. 3G,I,J,O). The three types of TCs were covered by endocardium that did not express Nkx2.5 or MEF2C. Some leaflets presented a few MEF2C expressing endocardial cells (Fig. 3M).
In this article, we used SEM, histological analysis, and immunohistochemical techniques to address two issues: (1) The pattern of development of the RSL and the accompanying TCs. (2) The origin and fate of the myocardium embedded in the ventricular stratum of the right horn of the endocardial inferior AV cushion at the beginning of valvulogenesis. Our findings indicate that the RSL and the accompanying TCs present a characteristic morphogenetic pattern in which differential growth of the cushion and myocardial ingression from the IVS are the main mechanisms involved. We did not find evidence to support myocardial delamination as reported in septal leaflets development. Furthermore, our results suggest that myocardial recruitment from the IVS, but not mesenchyme-to-myocardium transformation, participates in the development of the muscular region of the TCs adjacent to the IVS.
Development of the Right Atrioventricular Septal Valve Leaflet and Tendinous Cords
The SEM and histological studies showed that at the beginning of valvulogenesis (stages 28–38HH) the right horn of the inferior AV cushion has the form of a mesenchymal mammillae ridge (Fig. 1). The posterior third of the ridge progressively condenses and transforms into the RSL and TCs (Figs. 2, 3) with anatomical features similar to those described in the adult heart of chicken by Cayré et al., (1993). At the same time, we noted that maturation of the leaflets and TCs is determined by significant post-hatched changes in cell organization (Fig. 3). Concerning the relative contribution of endocardial cushion tissue and myocardium, similar to Chin et al., (1992) and Oosthoek et al. (1998), we observed that in the early stages the ridge is composed of two mesenchymal strata (atrial and ventricular) (Fig. 1J,K,L). However, we noted that during morphogenesis of the RSL, the mesenchymal strata underwent changes of form and cellular organization. The atrial stratum acquired a leaflet or mammillae appearance but during development the mesenchymal tissue become denser and well organized. In contrast, the originally mesenchymalized ventricular stratum became muscularized. Subsequently, incipient connective and myocardial TCs appeared in this region (Fig. 2). Initially at stage 38HH, almost the entire ventricular stratum was occupied by myocytes forming an almost compact tissue (Fig. 1 LPT). Later, at stage 40HH, the muscularized ventricular stratum and the adjacent zone of the mesenchymal atrial stratum grouped and condensed to form some cellular bundles covered by endocardium. They were constituted by mesenchyme (lying the atrial stratum) and myocardium (next to the IVS) (Fig. 2G,J,M). Between stages 42–44HH, some very thin endothelial TCs were observed in the free border of the leaflet. At the same time, the cellular bundles became tendinous cord-like structures formed by connective and myocardial tissues covered by endocardium. Those structures to contact the IVS and insert into the ventricular border of the leaflet correspond to the second order TCs (Fig. 2I). These findings suggest that the RSL and connective region of the TCs originates from the endocardial cushion atrial stratum, whereas the zone of the TCs adjacent to the IVS is formed from the ventricular stratum. However to confirm this idea another type of studies as gene mapping analysis are necessary.
Based on these findings, we do not think that myocardial–mesenchymal transformation (Lamers et al.,1995) or apoptosis (De Lange et al.,2004) is involved in development of the RSL TCs which has been described in the case of the left mural leaflet. In contrast, we speculate that first and second order TCs formation depends of the interaction between the endocardial cushion mesenchyme and the IVS myocardium. This process determines a substantial differential growth of the connective tissue compared with the myocardium. The result is an apparent retraction of myocardium in contact with the IVS. This idea partially agrees with the statement that the remnant myocardium underlying the mesenchyme of the AV endocardial cushions is initially retracted towards the papillary muscles and finally disappears (Oosthoek et al.,1998). With respect to the muscular third order TCs observed at the hinge contact zone of the leaflet and the IVS, we propose that they remain muscular because the mesenchymal tissue is not present at this zone.
The temporal analysis of the anatomical and histological changes of the right horn of the inferior AV endocardial cushion suggests that morphogenesis of the RSL is a multi-step process that can be described as follows: (1) Pre-leaflet phase (Fig. 1). Between stages 28–38HH, a mesenchymal mammillary ridge develops. (2) Primitive leaflet phase. Around stage 40HH, the posterior third of the ridge become septated by small indentations. Simultaneously, the first intercordal spaces are formed in the myocardial ventricular stratum. Immediately, each segment begins to have a leaflet appearance and incipient TCs-like structures appear (Fig. 2D,G,J,M). (3) Early maturation phase. Between stages 42–46HH, connective and myocardial cords develop (Figs. 2E,H,K,N,F,I,L,O, 3D,G,J,M). (4) Late maturation phase. After hatching, histological maturation of leaflets and TCs takes place (Fig. 3H,K,N,I,L,O).
Origin and Fate of Myocardium Embedded in the Immature Cushion Mesenchyme at the Beginning of AV Valvulogenesis
We and others using in vivo labeling and molecular mapping (De la Cruz et al.,1983; García-Peláez et al.,1984; Oosthoek et al.,1998; Lincoln et al.,2004; De Lange et al.,2004) agree that the mesenchymal cells of the AV endocardial cushions are the main lineage of the AV leaflets and TCs. In spite of this, some questions remain concerning the origin and fate of the myocardial cells embedded in the immature mesenchyme of the AV right horn of the endocardial inferior cushion at the beginning of valvulogenesis. It has been postulated that they can arise by transformation of local mesenchymal cells or by the invasion of myocardial cells (Van den Hoff et al.,1999,2001).
Regulation of cardiomyogenic specification involves transcription factors and nuclear proteins such as Nkx2.5 and MEF2C (Lints et al.,1993; Lough and Sugi,2000; Dodou et al.,2004; Creemers et al.,2006). By immunohistochemistry, we did not detect these two cardiogenic specification markers in cells inside of the mesenchyme of the right horn of the inferior cushion of the AV canal during the early stages (28–30HH) (Fig. 1M,N,Q,R). These findings indicate that at the beginning of RSL morphogenesis, myocardial-committed cells are not present in the previously mentioned region. In contrast, later in development (36–38HH), myocardial cells were identified only in the ventricular stratum near the developing IVS (Fig. 1K,L,O,P,S,T). We did not find fibroblast-like cells immersed in the extracellular matrix expressing NKX2.5 or MEF2C at any stage of development examined in this study. These findings support the idea that the myocardial cells found in the immature leaflets are recruited from the developing IVS, rather than from local mesenchymal transformation.
MEF2C and Nkx2.5 are synergistically up-regulated genes. It has been demonstrated that expression of both nuclear proteins (MEF2C and Nkx2.5) can direct the cardiomyogenic pathway by up-regulation of Brachyury T, BMP4, Nkx2.5, GATA4, cardiac α-actin and myosin heavy chain expression (Skerjanc et al.,1998). Interestingly, from stage 30HH to the end of the study, we observed few endocardial cells covering the mesenchymal ridge expressing MEF2C but not Nkx2.5 (Figs. 1R,S,T, 2O, 3M). These new and unexpected findings indicate a different MEF2C regulation pathway that is independent of Nkx2.5 and can direct a non-cardiomyogenic specification as was pointed out by Black and Olson (1999).
In human AV valves, “muscular TCs” and third order TCs composed partially or totally by myocardium have been described (Netter,1976). In the hearts of the recently hatched chicks, we observed mesenchymal first and second order TCs and occasionally found totally muscular third order TCs. These results confirm that the septal leaflet of the right AV valve in human and the RSL in chicks are anatomically homologous (Cayré et al.,1993; De la Cruz and Markwald,1998). Therefore, the description in this article of the cellular events that occur during remodeling of the cushion into mature atrioventricular RSL is essential not only to our understanding of the normal development of that leaflet, but also to our understanding of congenital defects involving the definitive AV valves in humans.
The authors are grateful to Francisco Acosta Vazquez, Mario Jáuregui Castro and María Lidia Blancas for technical assistance.