Lineage Tracing Identifies Dynamic Contribution of Endothelial Cells to Cardiac Valve Mesenchyme During Development

Abstract

Heart valve disease is an important cause of morbidity and mortality among cardiac patients worldwide. However, the pathogenesis of heart valve disease is not clear, and a growing body of evidence hints at the importance of the genetic basis and developmental origins of heart valve disease. Therefore, understanding the developmental mechanisms that underlie the formation of heart valves has important implications for the diagnosis, prevention, and treatment of congenital heart disease. Endothelial to mesenchymal transition is a key step in initiating cardiac valve development. The dynamic changes in the relative localization and proportion of different cell sources in the heart valve mesenchymal population are still not fully understood. Here, we used the Cdh5-CreER;R26R-tdTomato mouse line to trace endocardial cushion-derived endothelial cells to explore the dynamic contribution of these cells to each layer of the valve during valve development. This is beneficial for elaborating on the role of endocardial cells in the process of valve remodeling from a precise angle.

Introduction

Heart valve disease is an increasing health care burden, with the incidence rate continuously rising worldwide. ¹ Abnormal development of the valves usually directly or indirectly leads to various heart diseases, such as arrhythmia, atrioventricular (AV) valve regurgitation, valve prolapse, arterial stenosis, myocardial hypertrophy, heart failure, and so on. In addition, patients with congenital valve malformations are more prone to develop valve diseases later in their life. ² Therefore, understanding the development of the valves is of great significance in determining the mechanisms and developing new treatment methods for heart valve diseases.

Valve formation begins with the endothelial to mesenchymal transition (EndoMT), which determines the anatomic location of the valve within the primary heart tube. ³ EndoMT of the endothelial lining of the heart, the endocardium, forms heart valve progenitor cells. The early developing heart consists of an endocardial layer and a surrounding myocardial layer that are separated by extracellular matrix (ECM) known as cardiac jelly. At embryonic day 9 of murine development (E9), the endocardial cushions, focal expansions of the cardiac jelly at the AV and ventriculo-arterial junctions, become visible. ⁴ These endocardial cushions act as primitive valves, preventing the backflow of blood between the primitive atria and ventricle, and between the ventricles, as blood circulates in the heart. ⁴ After EndoMT occurs, mesenchymal cells migrate and proliferate, and the formed valve precursor expands into the ventricle. At E12.5, the valve precursor is formed, and then, the valve precursor gradually becomes thin and continues to elongate (E14.5). After the remodeling stage at E17.5, the endocardial cushions eventually develop into mature valve leaflets. This EndoMT event represents only the initial stage of valve morphogenesis, followed by a series of complex molecular and cellular events that are not completely understood.

Valve leaflets are composed of three layers of stratified ECM, including an elastin-rich ventricularis layer (purple), a proteoglycan-rich spongiosa layer (light blue), and a collagen-rich fibrosa layer (yellow), interspersed with valvular interstitial cells (VICs) and sheathed in a monolayer of valvular endothelial cells (Fig. 1a). ⁵ The layering of ECM is closely related to the temporal-spatial distribution of VICs. The formation of the AV valve requires integration of multiple mesenchymal building units, so an important question to be answered is which cell lineages contribute to these structures. Previous studies have shown that VICs of the AV valve seem to originate from at least four different lineages: endocardial, epicardial and epicardial-derived cells (EPDCs), neural crest cells, and bone marrow hematopoietic stem cells. ⁶ The mesenchymal cells of the inlet valve mainly originate from the endocardium and epicardium, whereas those of the outlet valve originate from the neural crest and endocardium. ⁷ Currently, the relative positioning and dynamic changes of interstitial components and proportions from different cell sources are still not completely clear.

Figure 1.

Tracing cardiac endothelial cells by Cdh5-CreER;R26R-tdTomato mice (A) Cartoon showing the microstructure of mature cardiac mitral valve. (B) Schematic figure showing the Cdh5-CreER;R26R-tdTomato knock-in strategy. (C) Schematic showing experimental strategy. (D) Cartoon figure showing endothelial cells (ECs) contribute to mesenchymal cells (MCs) through EndoMT or maintain their cell identity. (E) Whole-mount views of embryos and hearts at different embryo days. (F) Whole-mount views and CDH5 immunostaining for ventricular wall of the heart with tamoxifen treatment. (G) Whole-mount views and CDH5 immunostaining for mitral valve of the heart without tamoxifen treatment. LV, left ventricle; RV, right ventricle, scale bars: 100 μm in d, e; 1mm in g. Each image is representative of three individual biological samples.

The precise genetic manipulation of cells within living organisms has deepened our understanding of organ development and tissue regeneration. The conventional lineage tracing strategy, based on Cre-loxP recombination, is widely used to track cell fate conversion during tissue development, homeostasis, and pathogenesis.^8,9 In transgenic or Cre knock-in mice, Cre is expressed in specific cell populations driven by the specific gene promoter. Because excision at the DNA level is irreversible, the expression of reporter genes is heritable and permanent. All cells expressing Cre and their descendant cells are permanently labeled with the same fluorescent signal. This system provides a powerful and effective method for tracing specific cell populations and their descendants in vivo, allowing observation and analysis of cell proliferation, differentiation, and the movement of specific cell lineages during steady state or disease periods.

Tie2-Cre has been effectively used to visualize EndoMT during valve development; however, Tie2 is expressed in both endothelial and hematopoietic cells, leading to nonspecific detection of EndoMT. ¹⁰ In contrast, the cadherin 5 (Cdh5)-CreER lineage can more completely and specifically label valve endothelial cells in postnatal and adult stages, making it the optimal tool for EndoMT research during heart development. ¹¹ In this study, Cdh5-CreER;R26R-tdTomato mice were used to track the endothelial cell–derived endocardial cushion and described in detail the different fates of endocardial cells during valve development, as well as the dynamic proportion of mesenchymal cells derived from the endocardium.

Materials and Methods

Mice

Cdh5-CreER and R26R-tdTomato gene-edited mice were used, which were generated by Shanghai Model Organisms Center, Inc. and maintained on a C129/C57BL6/J mixed background. When Cre mice are crossed with Rosa26-loxP-Stop-loxP (R26R) reporter mice, the Cre recombinase recognizes the loxP site and excises the transcribed Stop sequence (loxP-Stop-loxP), resulting in reporter gene expression.^12,13 The mice were kept under a natural light–dark cycle and raised on a free diet, with a constant room temperature of 23C~25C. All mice were handled according to the guidelines of the Laboratory Animal Center of Jinan University.

Genomic PCR

Genomic DNA was extracted from mouse tail. Tissues were lysed in lysis buffer (100-μg/ml proteinase K) overnight at 55C and the mixture was centrifuged at the maximum speed of 20,000 × g for 8 min to obtain supernatant with genomic DNA the next day. DNA was precipitated using isopropanol, and then washed in 70% ethanol by centrifugation at 20,000 × g for 3 min. Finally, DNA was dissolved in deionized water. All the mice or embryos were genotyped using genomic PCR to distinguish knock-in allele from wild-type allele. For the Cdh5 CreER, primers 5′-TGGATAGTGAAACAGGGGCAATG-3′ and 5′-ATAGAGTATGGGGGGCTCAGCATC-3′ were used to detect the Cdh5 CreER positive allele. For the R26R-tdTomato, primers 5′-AAGGGAGCTGCAGTGGAGTA-3′ and 5′-CCGAAAATCTGTGGGAAGTC-3′ were used to detect the positive allele, and 5′-GGCATTAAAGCAGCGTATCC-3′ and 5′-CTGTTCCTGTACGGCATGG-3′ were used to detect the wild-type allele.

Tamoxifen Treatment

Healthy adult male and female gene-edited mice were randomly selected and mated in the evening, and vaginal plugs were checked the next morning.

After 8 days of vaginal plug detection, the weight of the female mice was measured to determine successful pregnancy. Tamoxifen (20 mg/ml, dissolved in corn oil, T5648; Sigma) was given by oral gavage at the indicated time points.

Whole-mount Fluorescence Microscopy

At embryonic days (E) 9.5, 10.5, 12.5, 14.5, and 17.5, mouse embryos or hearts were collected in phosphate-buffered saline (PBS) on ice and then washed gently to remove blood. They were then fixed in 4% paraformaldehyde for 15–60 min at 4C depending on tissue size. After washing several times in PBS, the tissues were placed stably at the indicated orientation on 1% agar gel for whole-mount imaging using a Leica fluorescence microscopy (M205FA). For E9.0–E10.5 embryos, images of both sagittal sides were obtained. For E12.5–E14.5 embryos, images of the heart coronal and ventral sides were collected, and the four chambers of the heart should be kept as flat as possible. After that, the embryos or hearts were dehydrated in 30% sucrose/PBS for several hours or overnight at 4C.

Immunostaining and Confocal Microscopy

After being sunk to the bottom of the sucrose solution, tissues were incubated in optimum cutting temperature compound (Sakura; Japan) for at least 1 hr, embedded in optimum cutting temperature compound at −20C, and stored at −80C. Cryosections (8 µm thick) were prepared from the optimum cutting temperature compound-embedded tissue block on the negatively charged slides and stored at −20C until use. Sections were dried at room temperature for 1 hr and were then washed in PBS for 15 min.

Tissue sections were blocked with 5% donkey serum in PBS/Triton X-100 (0.1% Triton X-100 in PBS) containing 4′,6-diamidino-2-phenylindole (Vector Laboratories, Newark, CA USA) for 30–45 min at room temperature, followed by incubation with primary antibodies (dilution ratios are as follows): CDH5 (1:200; R&D, Minneapolis, MN USA); platelet-derived growth factor receptor alpha (PDGFRA, 1:500; R&D, Minneapolis, MN USA); tdTomato (1:5000; Rockland, Philadelphia, PA USA); α-smooth muscle actin (αSMA, 1:200; Abcam, UK); and platelet-derived growth factor receptor beta (PDGFRB, 1:500; eBioscience, Waltham, MA USA) in PBS/Triton X-100 containing 2.5% donkey serum at 4C overnight. The next day, the slices were washed with PBS for 15 min to remove primary antibody and then incubated with Alexa-conjugated secondary antibodies diluted in 0.2% Triton X-100/PBS (donkey anti-goat Alexa Fluor 488, donkey anti-rabbit Alexa Fluor 488, and donkey anti-rabbit Alexa Fluor 555, 1:1000; from Invitrogen, and donkey anti-mouse Alexa Fluor 488, 1:1000; from Invitrogen, Waltham, MA USA) in the dark at room temperature for 30–50 min. The slices were washed with PBS for 15 min and then covered with an antifluorescence quencher and sealed with a coverslip. The images were taken using an Olympus confocal microscope (FV3000).

Statistical Analysis

All data are presented as mean ± SD ( $\overline{x}$ ± s). The t-test was used for analysis with a significant level of p<0.05. “ns” indicates no significant difference. Origin software was used for graphing.

Results

Characterization of the Cdh5-CreER;R26R-tdTomato System

The Cdh5-CreER;R26R-tdTomato system was designed to specifically and efficiently label mouse cardiac endothelial cells after tamoxifen treatment (Fig. 1b). A decrease in the percentage of tdTomato⁺ cells expressing the endothelial cell marker CDH5 would indicate that the labeled endocardial cells have lost their endothelial cell identity. If the tdTomato⁺ cells express VIC markers PDGFRA, PDGFRB, or αSMA, it suggests that endocardial cells have undergone transdifferentiation into different types of VICs (Fig. 1d).

To obtain endogenous labeling of as many endocardial cells as possible, tamoxifen was administered at E8.5, and samples were collected for analysis at E9.5, E10.5, E12.5, E14.5, and E17.5 (Fig. 1c). Whole-mount fluorescent images of the left side of embryos and the dorsal side of embryonic hearts showed significant tdTomato expression at these time points (Fig. 1e). Immunofluorescence staining for CDH5 on embryonic heart sections showed that almost all endothelial cells (>99%) were tdTomato⁺ (Fig. 1f), whereas only very weak fluorescence and a few tdTomato⁺ cells were observed in the control group without tamoxifen induction, suggesting that the endogenous estrogen’s effect on Cre recombinase activity was negligible (Fig. 1g).

The Spatiotemporal Distribution of CDH5⁺and tdTomato⁺Cells Reveals Different Fates of Endocardial Cells During Valvulogenesis

Heart valve development is a conserved morphogenetic process in evolution, during which a subpopulation of endocardial cells is specified to form valves. ¹⁴ In the AV canal, almost all tdTomato⁺ endocardial cells were also CDH5⁺ at E9.5 (Fig. 2a), consistent with previous studies. ¹⁵ Subsequently, some tdT⁺ cells lost endothelial marker CDH5 and migrated into the endocardial cushion at E10.5 (Fig. 2b). This indicates that EndoMT occurs at E10.5, and some endocardial cells detach from the cushion endocardium and invade the cushion mesenchyme to form the valve primordia. Therefore, endocardial cushion cells, as the valve progenitor cells, have two distinct fates from E10.5: maintaining an endothelial cell state or becoming mesenchymal cells, which is a necessary step in normal valve development.

Figure 2.

Endothelial cells in endocardial cushion labeled by Cdh5-CreER at E9.5 and E10.5 (A, B) Immunostaining for CDH5 and tdTomato on E9.5 and E10.5 mouse embryo sections. Boxed regions are magnified. DAPI (blue) stains nuclei. Scale bars: 100 μm. Each image is representative of three individual biological samples.

By immunostaining on embryonic heart sections, we found that endothelial cells derived from the endocardium almost filled the endocardial cushion at E12.5, and changed from spindle-shapes to irregular shapes with pseudopodia, while covering only the surface of the parietal leaflet and the mural leaflet (Fig. 3a). CDH5⁺ endocardial cells always covered the valve surface and maintained the endothelial phenotype at middle and late stages of embryonic development (Fig. 3b-c).

Figure 3.

Fate mapping of valvular endothelial cells during embryonic development (A) Image of endothelial cells and mesenchymal cells localization in E12.5 embryonic heart valve precursors. (B) Image of endothelial cells and mesenchymal cells localization in E14.5 embryonic heart valves. (C) Image of endothelial cells and mesenchymal cells localization in E17.5 embryonic heart valves. 1, Magnified image of tricuspid valve; 2, Magnified image of mitral valve. LV, left ventricle; RV, right ventricle; pl, parietal leaflet; sl, septal leaflet; al, aortic leaflet; ml, mural leaflet. Scale bar: 100 μm. Each image is representative of three individual biological samples.

At E14.5, tdTomato⁺ cells were only detected in the distal mesenchymal space of the mural leaflet, accounting for 25% of the total cells in the mural leaflet (Fig. 4). But in the parietal leaflet, tdTomato⁺ cells were even less than 10% in the mesenchymal space (Fig. 4), which suggests that the development of the parietal leaflet and the mural leaflet may not heavily rely on the endothelial cell lineage from the endocardium due to their anatomic differences in positioning, but rather involve contributions from other lineages.

Figure 4.

Quantification the percentage of endothelial cells during atrioventricular valve development. Data are mean ± standard deviation. * P<0.05; NS, non-significant.

Throughout the process of valve development, approximately 15–25% of CDH5⁺tdTomato⁺ double-positive cells are present on the surface of each leaflet (Fig. 4), indicating that some endocardial-derived cells are always located on the valve surface and remain relatively stable.

The Dynamic Contribution of Endocardial-derived Endothelial Cells to the Valve Fibrosa

PDGFRA is a cell surface tyrosine kinase receptor of the platelet-derived growth factor family members and has been shown to be essential for cardiac fibroblasts. ¹⁶ The PDGF receptor signaling pathway is necessary for epicardial–mesenchymal transition (EMT), but its effect on endocardial-derived mesenchymal cells in valve remodeling processes has rarely been reported.

During early embryonic development, PDGFRA⁺ cells are found throughout the endocardial cushion, and quantitative data show that most of endothelial cells undergo EndoMT and participate as mesenchymal cells in the formation of the septal leaflet of tricuspid valves and the aortic leaflet of mitral valves (Figs. 5a and 6).

Figure 5.

Valvular endothelial cells contribute to PDGFRA+ mesenchymal cells at middle and late stages of embryonic development (A) Image of endothelial cells and PDGFRA+ mesenchymal cells localization in E12.5 embryonic heart valve precursors. (B) Image of endothelial cell s and PDGFRA+ mesenchymal cells localization in E14.5 embryonic heart valves. (C) Image of endothelial cells and PDGFRA+ mesenchymal cells localization in E17.5 embryonic heart valves. 1, Magnified image of tricuspid valve; 2, Magnified image of mitral valve. LV, left ventricle; RV, right ventricle; pl, parietal leaflet; sl, septal leaflet; al, aortic leaflet; ml, mural leaflet. Scale bar: 100 μm. Each image is representative of at least three individual biological samples.

Figure 6.

Quantification the percentage of mesenchymal and endothelial cells during atrioventricular valve development. Data are mean ± standard deviation. *P<0.05; NS, non-significant.

At E14.5, PDGFRA⁺ cells are loosely distributed in the septal leaflet and aortic leaflet, while more tdTomato⁺ PDGFRA⁺ double-positive cells are observed on the ventricular side of the valve compared with other layers. About 38% of the septal leaflet fibroblasts are derived from endocardium, and this proportion remains unchanged after valve remodeling (Figs. 5b and 6).

For the parietal leaflet and the mural leaflet, mesenchymal cells were only visible in the proximal area during early heart development, and there was minimal contribution from cells derived from the endocardium (Figs. 5b and 6). At E14.5, double-positive cells appeared in their distal regions, suggesting that the surface endothelial cells in this area might have proliferated and differentiated into distant fibroblasts.

During the leaflet remodeling stage (E17.5), compared with the tricuspid valves, the mitral valves begin to exhibit clearer ECM layer differentiation. PDGFRA⁺ fibroblasts are distributed on the ventricular side, forming the fibrous layer of the valve, of which about three fourths are derived from the endocardium. The matrix surrounding these newly differentiated fibroblasts matures into highly organized fibrous connective tissue, making the valve more rigid in structure and capable of bearing increased hemodynamic loads from the beating heart (Figs. 5c and 6).

Contribution of Endocardial-derived Endothelial Cells to the Interstitial Compartment of the Atrial Layer of the AV Valve

PDGFRB is the receptor for PDGFB, which has been reported to play an important role in the EMT of the epicardium and the determination of the fate of mesenchymal cells as a potent stimulator of cell proliferation. However, there are few reports on the spatial and temporal distribution of PDGFRB⁺ cells during valve development and its relationship with endothelial cells.

In the early stages of cardiac development, PDGFRB⁺ cells were not found in the endocardial cushion (Fig. 7a). Until E14.5, PDGFRB began to be expressed to various degrees. The expression of PDGFRB is most pronounced in the aortic leaflets (Fig. 7b). About 60% of the PDGFRB⁺ mesenchymal cells are derived from the endocardium (Fig. 8).

Figure 7.

Distribution of PDGFRB+ mesenchymal cells derived from Valvular endothelial cells during embryonic development (A) Image of endothelial cells and PDGFRB+ mesenchymal cells localization in E12.5 embryonic heart valve precursors. (B) Image of endothelial cells and PDGFRB+ mesenchymal cells localization in E14.5 embryonic heart valves. (C) Image of endothelial cells and PDGFRB+ mesenchymal cells localization in E17.5 embryonic heart valves. 1, Magnified image of tricuspid valve; 2, Magnified image of mitral valve. LV, left ventricle; RV, right ventricle; pl, parietal leaflet; sl, septal leaflet; al, aortic leaflet; ml, mural leaflet; Epi, epicardium. Scale bar: 100 μm. Each image is representative of at least three individual biological samples.

Figure 8.

Quantification the percentage of mesenchymal and endothelial cells during atrioventricular valve development. Data are mean ± standard deviation. *P<0.05; NS, non-significant.

At E17.5, the PDGFRB⁺ cells almost exclusively gathered in the atrial layer of the valve, and the average proportion of double-positive cells is reduced. The septal leaflet, which has the same cell origin as the aortic leaflets, showed the opposite results (Fig. 7c). From E14.5 to E17.5, the proportion of double-positive cells increased, of which about 70% were derived from the endocardium (Fig. 8). As other source cells entered, the proportion of double-positive cells in the development of the cusps decreased in both the superior and inferior leaflets, contributing to less than 50% of the elastic protein layer (Fig. 8).

Myofibroblasts Do Not Participate in the Remodeling of Valve Leaflets

αSMA is considered as a marker for smooth muscle cells and myofibroblasts, which are unique stromal cells found in many tissues and exhibit features of both fibroblasts and smooth muscle cells. αSMA⁺ cells were detected surrounding the vascular endothelial cells located on the free walls of the left and right ventricles, with a small subset of double-positive smooth muscle cells (Fig. 9), indicating a tendency of endothelial cells differentiating into vascular smooth muscle cells, consistent with previous studies. ¹⁷

Figure 9.

Localization of αSMA+ smooth muscle cells derived from endothelial cells at E17.5 1, Magnified image of tricuspid valve; 2, Magnified image of mitral valve; 3, Right ventricular wall blood vessels; 4, Left ventricular wall blood vessels. LV, left ventricle; RV, right ventricle; pl, parietal leaflet; sl, septal leaflet; al, aortic leaflet; ml, mural leaflet; Epi, epicardium. Scale bar: 100 μm. Each image is representative of at least three individual biological samples.

Notably, αSMA expression was only detected in the connective tissue adjacent to the outflow tract, and no αSMA⁺ cells were detected in the valve leaflets at E17.5. This suggests that there is no involvement of myofibroblasts in the process of valve remodeling.

Discussion

The AV and semilunar valves in the four-chambered hearts of humans and mice share a similar microarchitecture composed of three layers of ECM interspersed with valve interstitial cells and sheathed in a monolayer of endocardial valve endothelial cells. ¹⁸ Similar to mouse valve development, there is evidence of EndoMT in human’s, involving signaling pathways related to NFATc1 (nuclear factor of activated T cells, cytoplasmic 1) and BMP2 (bone morphogenetic protein 2). ¹⁹ The difference lies in the fact that valve endocardial cushions in humans are derived from four major cushions: the posterior and anterior AV cushions, and the parietal and outflow tract septal cushions. ²⁰ Morphologically, the AV valve primordia are typically observed to appear after E12, during which time the endocardial cells are the main source of valve mesenchymal cells. Throughout mid-embryonic stages, endocardial-derived cells still comprise the majority of the mesenchyme in the aortic leaflet and the septal leaflet. In contrast, the anatomic position of the parietal leaflet and the mural leaflet is different from the two septal leaflets. Their mesenchymal cells do not originate from the endocardial cushion. In previous studies, epicardium-derived progenitor cell populations migrate from the epicardium to the valves and contribute to the valve mesenchyme of the parietal leaflet and the mural leaflet. ¹¹

Endothelial cells originating from the endocardial cushions are different from those within the heart chambers, expressing genes necessary for septum and valve development at different time patterns, such as genes encoding NFATc1 and vascular endothelial growth factor receptors.^21–23 These different gene express programs and the different signaling factors and receptors on the endocardial cushion determine the regional specificity of valve formation after EndoMT. In previous studies, a subpopulation of endocardial cells expressing NFATc1 did not undergo EndoMT but instead remained as a proliferating cell population. ²⁴ This is consistent with the results of this study, as double-positive cells retaining their endothelial cell identity have always been present on the surface of each leaflet throughout valve development. These cells act as proliferative populations to support the extension of the valve leaflets during elongation and express NFATc1 to autonomously inhibit EndoMT. ²⁴

The division of VICs and the synthesis and compartmentalization of ECM are temporally coordinated. ²⁵ In this process, we discovered that most endocardial cushion-derived mesenchymal cells, accompanied by PDGFRA signaling, are distributed within the collagen-rich fibrous layer, suggesting that most endocardial cells undergo EndoMT and differentiate into fibroblasts within the valve, while mesenchymal cells secreting elastin may be derived from other lineage cells such as EPDCs originating from the epicardium. However, the mechanisms by how endocardial cells remodel their local environment into mature leaflets and their supporting organs require further investigation. In vitro studies have shown that treatment with BMP induces Sox9, a transcription factor associated with chondrogenesis, in valve progenitor cells, followed by the expression of aggregating proteins. ²⁶ Wnt signaling is expressed in the collagen-rich fibrous layer, and Notch1 is expressed in the elastic layer. ²⁷ These data support that the diversity of valvular mesenchymal cells derived from the valvular primordium triggers ECM compartmentalization. However, further research is needed to determine the plasticity of subpopulations of VICs and their specific contributions to ECM partitioning, providing valuable insights into changes in valve structure and function in disease. There is controversy regarding whether smooth muscle cells and myofibroblasts contribute to valve development, and whether endothelial cells differentiate into myofibroblasts through EndoMT within the valves. αSMA is a marker for smooth muscle cells and myofibroblasts, ¹⁰ but it was not detected during valve developing in this study, and was only detected in vitro in studies of valve EndoMT. ²⁸ In earlier studies, smooth muscle cells were also classified as a type of cell in the AV valve. ²⁹ Recent research has shown that there are no smooth muscle lineage cells in the AV valve during mouse embryonic development, and myofibroblasts are speculated to provide support in mature valves. ¹¹ This also indicates that αSMA, as a marker for myofibroblasts, is nonspecific, and in future research, it is hoped that more specific markers or gene-edited mice can be generated to explore the possibility of myofibroblasts during EndoMT in valve development.

Establishment of cardiac organoids provides a method to study different developmental stages of the heart. The Mendjan’s group developed an in vitro self-organized cardiac organoid model in 2021 that can spontaneously form cavities and beat spontaneously by activating currently known signaling pathways involved in embryonic heart development, which can mobilize cardiac fibroblasts for repair after injury. ³⁰ In the aforementioned study, the contribution of induced pluripotent stem cells (iPSCs) is significant. iPSCs derived from normal human aortic valve leaflets were genetically manipulated via targeted gene editing to induce GATA4 mutation, and subsequently differentiated into endothelial cells. ³¹ EndoMT was interrupted by inducing GATA4 disruption, which could result in development of a bicuspid aortic valve. ³¹ However, it has not been demonstrated that cells derived from iPSCs can reproduce the formation of valve structures during development, and iPSC-ECs still do not fully recapitulate in vivo endocardial lineage. A scaffold is indispensable in the culture process of cardiac organoids, as its purpose is to simulate the ECM environment and provide an extracellular environment and structural support for cells. ³² Although endocardial endothelial cells in developing valves can undergo EndoMT and differentiate into other cell types, the mutual, interactions between endothelial cells and ECM in valves are still not fully understood, so the choice of scaffold type is also a factor to consider when standardizing the culture protocol. Valvular organoids could provide an option for future research on heart valve diseases.