Skip to main content
Journal of Advanced Research logoLink to Journal of Advanced Research
. 2025 Oct 28;85:441–454. doi: 10.1016/j.jare.2025.10.062

Vinculin mediates a dynamic signal flow underlying cardiac outflow tract development as revealed by single-cell and spatial transcriptomic analysis

Reeson Xu Wang a,1, Mingxuan Liang b,1, Xixin Wang a, Yan Pan b, Zhixin Li a, Chi-Chung Hui c, Kyoung-Han Kim d, Asif Javed b,⁎, Elly Sau-Wai Ngan a,⁎⁎
PMCID: PMC13316539  PMID: 41167421

Graphical abstract

graphic file with name ga1.jpg

Keywords: Cardiac neural crest cell; Outflow tract; Vinculin; Transforming growth factor-beta, extracellular matrix; Single cell transcriptomics; Spatial transcriptomics

Highlights

  • •

    Vcl is essential for CNCC-directed OFT development.

  • •

    Vcl-TGF-β-p38 directs the differentiation of CNCCs along the vascular smooth muscle lineage.

  • •

    A “digital in situ” differentiation wave of CNCCs along the OFT correlates the CNCC differentiation with their spatial patterns.

  • •

    CNCC cell state affects their interaction with SHF derivatives through the extracellular matrix, mediating OFT morphogenesis.

Abstract

Introduction

Cardiac outflow tract (OFT) development relies on the interactions among cardiac neural crest cells (CNCC) and the second heart field (SHF) cells. However, the molecular signals underlying these interactions remain largely unclear.

Objectives

Vinculin (VCL) is an adaptor protein for focal adhesions, which integrates various external signals and mediates cell–cell communications, representing a potential hub gene directing the OFT development. Here, we aimed to build a holistic map of signal flows underlying CNCC-directed OFT development using neural crest-specific Vcl knockout mouse mutants (Vcl cKO).

Methods

We validated the OFT defects of Vcl cKO mutants by performing various immunohistochemistry analyses. Subsequent high-resolution transcriptomic analysis reconstructed the differentiation trajectory of CNCCs along the vascular smooth muscle (VSMC) lineage. The spatial RNA-seq further aligned the transcriptomes of OFT cells with their spatial patterns to explore the cell interaction dynamics among OFT cells. Exploiting human pluripotent stem cell (hPSC)-based in vitro differentiation assays, the signaling pathways underlying Vcl-mediated CNCC progression along the VSMC lineage were defined.

Results

Vcl cKO exhibited malformation of pharyngeal arch arteries, OFT septation defects and hyperplastic semilunar valves, due to defects in migration and vascular smooth muscle (VSMC) lineage differentiation of CNCCs, which prohibited the subsequent myocardialization, perturbing the OFT and valve remodeling. High-resolution transcriptomic analysis revealed that CNCC-to-VSMC differentiation is orchestrated by Vcl-dependent p38 and Tgf-β signalings. Multimodal analysis integrating the single-cell and spatial transcriptomes of OFT cells further suggested a shift of signal flow among OFT cells and loss of cell-extracellular matrix interaction perturb OFT morphogenesis.

Conclusion

In sum, dynamic signal flows among CNCC and SHF cells guide OFT remodeling, in which p38-Tgf-β signaling mediates the VSMC lineage differentiation of CNCCs, and timely regulates their interactions with SHF-derived cells through ECM.

Introduction

Cardiac neural crest cells (CNCCs) originate from the dorsal neural tube between the mid-otic placode and the caudal boundary of the third somite. After they delaminate from the dorsal neural tube at approximately E8.0–9.0 in mice, CNCCs migrate through pharyngeal arches 3, 4, and 6 to populate the aortic sac and truncal OFT [1]. CNCCs in the pharyngeal arch arteries respond to developmental cues to differentiate into vascular smooth muscle cells (VSMCs), while CNCCs coordinate with the second heart field (SHF) cells at the conotruncus (CT) to regulate remodeling and separation of the outflow tract (OFT) into the aorta (Ao) and pulmonary trunk (PT) and ventricular septal (VS) closure, which take place at approximately E10.5–E14.5 in mice [2]. During the formation of OFT, several intricate cellular processes are involved. Notably, endothelial-to-mesenchymal transition (EndoMT) plays a crucial role in generating the OFT endocardial cushions, which are essential for septation of the OFT. Additionally, extracellular matrix (ECM) synthesis establishes a supportive environment conducive to both EndoMT and valvulogenesis. ECM remodeling, characterized by the dynamic restructuring of the extracellular matrix, provides structural support vital for valve morphogenesis. Mechanical sensing, or mechanotransduction, is another vital biological process in normal cardiac development, especially within the OFT. Mechanical forces such as shear stress from blood flow are detected by OFT cells and transduced into biochemical signals that govern cellular migration, differentiation, proliferation, and maturation—processes critical for proper valve formation. Given CNCCs represent the major population of progenitors for the formation of the OFT and great arteries, any failure of CNCCs to properly execute these cellular functions can lead to various congenital cardiac anomalies, including interrupted aortic arch (IAA), persistent truncus arteriosus (PTA), atrioventricular septal defect (ASV/VSD), and hyperplastic semilunar valves [3]. Recent high-resolution transcriptomic studies of CNCCs at various developmental stages have provided critical information regarding the molecular landscapes of CNCCs and their derivatives along their differentiation trajectories during heart development [[4], [5], [6]]. It has opened a new opportunity to untangle the unknowns in cardiac OFT development and the associated diseases, while a holistic map of various cell types at different cell states with the corresponding spatial information and the signal flows underpinning the dynamic interactions among CNCCs and SHF cells remain to be defined.

Vinculin (Vcl) was identified as a key component of focal adhesions (FAs) and adherens junctions (Ajs) for mediating cell-matrix and cell‒cell adhesions, respectively [7]. Since FAs are crucial for the stabilization and activation of integrins, loss of Vcl is likely to affect integrin-associated signaling and cell-ECM interactions [8,9], and it is particularly detrimental to embryo development. Vcl null mouse mutants die at embryonic day 10 with failure of the rostral neural tubule closure, aberrant forelimb development, heart malformation with reduced sizes and fewer myocytes than normal [10] and compromised cranial and spinal nerve development [11,12]. Expression analysis in avian embryos revealed that Vcl levels are selectively enhanced in migratory vagal and trunk NCCs when they enter cellular environments [13]. Vcl is highly expressed in aorticopulmonary septation, where CNCCs are transformed into ectomesenchyme, suggesting its potential roles in the septation of the cardiac OFT [14]. In addition, the dynamic distribution of Vcl at FAs is associated with the altered migration directionality and speed of CNCCs [15]. L954del and R975W are the only two VCL mutations reported to be associated with sporadic or familial dilated cardiomyopathy, where the tail region of VCL protein was affected, and that may interfere with the actin/paxillin binding of VCL [16,17]. Another novel loss-of-function mutation (M209L) in the Vinculin (VCL) gene was previously identified from a syndromic Hirschsprung patient with ASD/VSD and found that the M209L substitution interrupts FA assembly, leading to retarded migration and defective neuronal and smooth muscle lineage differentiation of human NCCs derived from a patient-specific hPSC line. Correction of this VCL mutation completely rescued the FA phenotype, accompanied by the restoration of NCC functions [18], highlighting the relevance of VCL in the development of NCCs and neurocristopathies.

In this study, we generated a mouse mutant line with NCC-specific knockout of Vcl, and the mutants exhibited various cardiac defects, suggesting that Vcl is indispensable for CNCC-directed OFT development. Analyzing the single-cell and spatial transcriptomics of Vclnull CNCCs further revealed the molecular controls underlying CNCC-directed OFT septation and valvulogenesis, defining the key signals mastering the differentiation of CNCCs along the VSMC lineage and the timely regulation of their interactions with SHF derivatives to direct the OFT development.

Methods

The detailed methods and major resources tables are provided as the Supplemental Material.

Results

Cardiac defects in Vcl mutants

To functionally demonstrate the roles of Vcl in CNCC development, we intercrossed Vclf/f and Wnt1-Cre mouse lines to generate mouse mutants with specific deletion of both alleles of Vcl gene in NCCs (Vcl KOWnt1-Cre). Down-regulation of Vcl explicitly in CNCCs in developing OFT was confirmed by immunohistochemistry (Supplementary Fig. 1). Vcl KOWnt1-Cre mice died a few hours after birth with severe cardiac defects. Approximately 60 % of Vcl KOWnt1-Cre mice could survive up to P0. All of them showed VSD and hyperplastic valves and 87.5 % of them also presented with PTA or double outlet right ventricle (DORV) due to defective septation of the OFT into the aorta and the pulmonary trunk; 62.5 % of them exhibited malformations of the great arteries, interrupted aortic arches (IAAs) (Fig. 1A). In control mice, the ascending aorta arose from the ventricle, giving rise to various branches at the aortic arch region. These branches are the right subclavian artery (RSA), right common carotid artery (RCCA), left common carotid artery (LCCA), and left subclavian artery (LSA) from right to left in anatomic orientation (Fig. 1B, left panel). On the other hand, Vcl mutants exhibited IAA type B, in which the ascending aorta was split into the RSA, RCCA, and LCCA with the LSA connected to the pulmonary trunk through the ductus arteriosus (DA) (Fig. 1B, right panel). In addition to IAA, mutants demonstrated dilated vessels, where the diameters of all the aortic branches were increased in mutants, while the aorta remained comparable to that of the control (Fig. 1C). Hematoxylin and eosin (H&E) staining of the frontal sections revealed that the longitudinal view of the aorta and the transversal view of the pulmonary trunk emerged at the same time, indicating that they are oriented in a staggered fashion in control hearts (Fig. 1D, upper panel). On the other hand, 80 % of mutant hearts exhibited PTA or DORV, the latter having both the aorta and the pulmonary trunk enter the right ventricle (Fig. 1D, lower panel).

Fig. 1.

Fig. 1

Vcl KOWnt1-Cre mice exhibited various congenital heart defects. (A) The table summarizes the most commonly seen congenital heart defects in mutant mice. (B) Gross dissection of aortic arch arteries. Asterisk indicates the discontinuity between LCCA and LSA. (C) Bar chart shows the diameters of aortic arch arteries and aorta. Hematoxylin and eosin stainings of the frontal section of the the hearts show that the mutants exhibit (D) PTA as marked by asterisk; (E) fused hyperplastic the SLV as outlined by black dotted lines; (F) VSD as marked by asterisk. IAA, interrupted aortic arch; PTA, persistent truncus arteriosus; VSD, ventricular septal defect; IAA-B, interrupted aortic arch type B; LCCA, left common carotid artery; LSA, left subclavian artery; RCCA, right common carotid artery; RSA, right subclavian artery; Ao, aorta; PT, pulmonary trunk; LV, left ventricle; RV, right ventricle; VS, ventricular septum; PV, pulmonary valve; AoV, aortic valve; DORV, double outlet right ventricle. Areas highlighted in dotted boxes are magnified as shown in the right panels. Quantitative data are represented as mean +/− SEM. P-value less than 0.05 was considered significantly different. Student t-test, two-sided.

Semilunar valves develop from endocardial cushions at the proximal ends of the aorta and pulmonary trunk, and they undergo extensive remodeling during late gestation. At E18.5, thin and elongated valve leaflets were found at the proximal end of the aorta in control hearts (Fig. 1E, upper panel), while the aorta and pulmonary trunk were parallel, with fused hyperplastic semilunar valves in the mutants (Fig. 1E, lower panel).

The ventricular septum (VS) is composed of a lower muscular part and an upper membranous part. Endocardial cushions, including aorticopulmonary septal ridges and atrioventricular cushions, contribute to the closure of the VS. In control mice, posterior to the entrance of the aorta into the left ventricle, the endocardial cushions fused and separated the aorta from the right ventricle (Fig. 1F, upper panel). However, mutants at E18.5 exhibited VSD, the aorta was located above the unclosed VS and the ventricles remained unseparated (Fig. 1F, lower panel). In summary, Vcl mutants presented a spectrum of severe abnormalities from the distal to the proximal ends of the heart, suggesting the implications of Vcl in CNCC-directed OFT development.

Perturbed migration and differentiation of CNCCs in Vcl cKO mutants

To further elucidate how the loss of Vcl influences CNCCs and causes these cardiac defects, we employed the Wnt1-Cre;RosaYFP reporter system to label the CNCCs with YFP. We did not observe a noticeable difference between control and mutant until E10.5. At E10.5, CNCCs in both control and Vcl mutants could populate the pharyngeal arch arteries (PAAs) III and IV; even fewer CNCCs were found in the mutant compared to the control and most of the mutant CNCCs remained as undifferentiated as monitored based on the expression of vascular smooth muscle cells (VSMCs) marker, smooth muscle actin (SMA) (Supplementary Fig. 2). The incomplete/delayed colonization of the PAAs by CNCCs and the defective differentiation of CNCCs into VSMCs at L-PAA IV at early gestation likely led to the interrupted and dilated aortic arch arteries observed at a later stage in the mutants.

At this stage, CNCCs also migrated into the OFT, contributing to the development of the aorta and pulmonary trunk. A significant delay in the colonization of the OFT by CNCCs and SHF (Isl1+YFP-) progenitors was observed in the mutants, as shown in the transverse sections of the OFT (Fig. 2A). In addition, two longitudinal OFT septal ridges appeared. A large number of CNCCs resided at the fusion site of the septation complex (as marked by asterisks), which was immediately beneath the separated aorta and pulmonary trunk and above the unfused septal ridges in the control heart (Fig. 2B). Although septal ridges were observed in the comparable region of the mutant hearts, fewer CNCCs were found. The septation complex was not observed (Fig. 2B), which likely perturbed OFT septation. By E12.5, aorticopulmonary septation was almost complete. Most of the control CNCCs were committed to the VSMC lineage in control hearts, while the aorta remained fused with the pulmonary trunk, and a significantly lower percentage of CNCCs was α-SMA-positive in the dilated aorta of the mutants (Fig. 2C). It is noteworthy that the apoptosis and proliferation of CNCCs in control and mutants at these stages are highly comparable (Supplementary Figs. 3 and 4). Thus, the persistent delayed population of PA and OFT by mutant CNCCs would be the result of the defects in migration and their interactions with SHF progenitor cells. In sum, the loss of Vcl in CNCCs perturbed their migration, differentiation and interactions with SHF cells, altogether leading to malformation of the great arteries and the aortic arch arteries.

Fig. 2.

Fig. 2

Delayed colonization of the OFT by CNCCs in Vcl KOWnt1-Cre mutants. (A) Transverse sections revealed the delayed migration of CNCCs and SHF progenitors in the OFT. Arrowheads and insets show the migration frontier. The migration distances of CNCCs and SHF progenitors were measured as marked by dotted lines, and the data are shown in the scatter plots (mean +/-SEM). (B) CNCCs in the OFT endocardial cushions at E11.5. White dotted lines outline the midline of OFT endocardial cushions. Asterisk marks the aorticopulmonary septal complex. (C) CNCC-derived VSMC cells were detected using antibody against SMA. SMA+ and SMA- CNCCs in the E12.5 aorta are marked by filled and empty arrowheads, respectively. The scatter plot shows the percentages of SMA+CNCCs (mean +/− SEM). OFT, outflow tract; EC, endocardial cushion; Ao: aorta. As shown in the right panels, regions highlighted in dotted boxes are magnified. P-values less than 0.05 were considered statistically significant differences. Student t-test, two-sided.

Vcl is required for CNCC-induced cushion remodeling during valve development

At E17.5, two pairs of sculpted semilunar valves (SLVs) were found at the proximal ends of the aorta and the pulmonary trunk of the control hearts, while a fused hyperplastic SLV was seen in mutants. Cushion CNCCs lying along the ventricular side of the SLV are responsible for regulating the expression of ECM for subsequent valvular stratification and remodeling [19]. In control mice, most of the cushion CNCCs aligned at the ventricular side of the SLV with some cells expressing α-SMA (Fig. 3A, upper panel), which restricted the expression of versican (Vcan) in the ventricular side of the control SLV (Fig. 3A, lower panel). Consistently, no α-SMA + CNCCs were observed in the mutants, and the mutant CNCCs failed to align at the ventricular side of the SLV, such that versican expression was found in the whole region of the SLV, including the arterial side (Fig. 3A), suggesting that the aberrant patterning and/or differentiation of CNCCs perturbs valve remodeling.

Fig. 3.

Fig. 3

Dysfunction of CNCCs caused malformation of SLV and ventricular septal defect in Vcl KOWnt1-Cre mutants. (A) The presence of activated α-SMA+ CNCCs and the expression of versican (Vcan) in SLVs at E17.5 were detected using immunohistochemistry. The valve leaflets are outlined by white dotted lines. (B) Disrupted myocardialization in SLVs of Vcl mutants. Immunostaining images showed distribution of CNCCs and α-SMA+ smooth muscle cells. Solid arrowheads indicate the cardiomyocytes with extended lamellipodia and filopodia in the control while open arrowheads indicate the cardiomyocytes with round morphology in mutants. (C) Failure in myocardialization in ventricular septum of Vcl KOWnt1-Cre;RosaYFP mice at E13.5. Asterisk indicates the communication between left and right ventricles. CNCCs and mycardiomyocytes are marked by YFP (green) and SMA (red), respectively. Solid arrowheads indicate the cardiomyocytes with extended lamellipodia and filopodia in control while open arrowheads indicate the cardiomyocytes with round morphology in mutants. White dotted lines indicate the fusion site of septal ridges. (D) Constrained myocardialization in the aorticopulmonary septal ridges of Vcl KOWnt1-Cre;RosaYFP mice at E12.5. Distributions of CNCCs (green) and α-SMA+ cardiomyocytes (red) in the proximal OFT septum are shown. Ao, aorta; AoV, aortic valve; PV, pulmonary valve; SLV, semilunar valve; a.s., arterial side; v.s. ventricular side. Regions highlighted in the dotted boxes are magnified as shown in the middle and/or right panels. The scatter plots show the corresponding quantitative data (mean+/− SEM). P-value less than 0.05 was considered significantly different. Student t-test, two-sided.

Invasion of myocardial cells into the cushions is essential for subsequent valve development. In the control, it was found at the root region of the SLV, which was populated by CNCCs. On the other hand, fewer CNCCs reached the root region with no obvious myocardialization in the mutants, where the myocardial cells flanking the SLV exhibited a round morphology, showing no lamellipodia for invasion into cushions (Fig. 3B). The defective myocardialization and VSMC differentiation of cushion CNCCs likely perturbed valve maturation in mutants.

Vcl-deficient CNCCs fail to induce myocardialization during VS closure

The presence of CNCCs and their subsequent apoptosis are important aspects of the regulation of myocardialization during the separation of the aorta from the right ventricle and the closure of the VS [[20], [21], [22]]. At E13.5, CNCCs reached the proximal OFT cushions beneath the aorta and on the top of the VS and myocardialization was initiated for the closure of the VS, where α-SMA+ myocardial cells with elongated striking lamellipodia invaded the conotruncus septum. In mutants, fewer CNCCs reached the top of the VS and the myocardial cells flanking the cushions were not able to invade and showed a round morphology (Fig. 3C). As a result, the aorta connected to the right ventricle, and the VS remained open.

Conotruncal septation occurred a day earlier and at E12.5, CNCCs resided in the middle of the conotruncal septation complex between the separating aorta and pulmonary trunk and started differentiating into VSMCs, and myocardialization was initiated, as evidenced by the presence of α-SMA+ cardiomyocytes invading the CNCCs (Fig. 3D, upper panel). Although aggregated CNCCs were observed in the middle of the septation complex and the apoptosis of CNCCs remained largely unaffected in the mutants (Supplementary Fig. 5), cardiomyocytes failed to protrude into the endocardial cushion properly, and no α-SMA-expressing CNCCs were found (Fig. 3D, lower panel). These data suggest that the perturbed myocardialization in the VS and proximal OFT likely represents a potential cause underlying the septation defect, leading to PTA and VSD.

Reconstruction of the differentiation paths of CNCCs reveals that Vcl is required for MES-to-VSMC lineage differentiation

Vcl cKO displayed a broad spectrum of OFT defects. Among those, the molecular mechanisms contributing to the defect in myocardialization remain largely uninvestigated. It was observed at E13.5 and was associated with abnormalities in VSMC lineage differentiation of CNCCs and interactions with the SHF derivatives. Based on the public scRNA-seq dataset of CNCCs at various stages along OFT development [4], E13.5 represents an excellent stage of studying VSMC differentiation, where most of the CNCCs have been committed and started differentiating to VSMC, covering the key stages of VSMC lineage differentiation. Therefore, we sequenced CNCCs that were isolated and FACS-enriched from the OFTs of 6 control and 7 Vcl mutant (Vcl KOWnt1-Cre;RosaYFP) embryos at E13.5 using single-cell RNA sequencing (scRNA-seq) with 10X Genomics, to further investigate how the loss of Vcl alters the molecular dynamics of CNCCs along their differentiation trajectories. In parallel, we performed 10x Visium spatial transcriptomics (ST) analysis on sagittal sections of E13.5 embryos to elucidate how changes in signal flows among various OFT cells contribute to the OFT defects as observed in the mutants. Cross-referencing our scRNAseq and ST-RNAseq datasets, we anticipated that the molecular mechanisms underlying how Vcl regulates CNCC development and their interactions with their neighborhood could be defined. A schematic summarizing of the analysis pipeline is shown in Fig. 4A.

Fig. 4.

Fig. 4

High-resolution transcription analysis reveals immature differentiation of CNCCs in Vcl mutant at E13.5. (A) Schematic diagram summarizing the experimental pipelines for transcriptomic analyses and validations. (B) tSEN plots of 6916 cardiac neural crest cell (CNCCs) in control and mutant with normalized cell number, colored by clusters. Canonical markers expressed in each cluster are shown. (C) Bar chart shows the proportions of cells at different cell state in control and mutant. (D) The lineage cladogram shows the hierarchical relationship of the clusters based on the average expression of the top 1500 variable features. (E) RNA velocity analysis shows the dynamic differentiation path of VSMC and MES subtypes in control and mutant CNCCs. (E) RNA velocity analysis shows the dynamic differentiation path of control and mutant CNCCs. (F) Boxplots show the expression of Sox9 and Acta2 at different cell states. Each box represents the interquartile range (IQR, the range between the 25th and 75th percentile) with the mid-point of the data, whiskers indicate the upper and lower value within 1.5 times the IQR. Tests marked by “***” represent they are statistically different (two-sided Wilcoxon rank-sum test) from the controls of P values < 0.001. Immunohistochemistry for the detection of CNCC derivatives at early- and late-cell state expressing mesenchymal (Sox9+) and SMC (Sox9-α-SMA+) markers, respectively, in (G) semilunar valves (SLV) and (H) ventricular spectrum (VS) at E13.5 control and Vcl mutants. Scatter plots show the quantitative data as mean +/− SEM. Student t-test, two sided. CNCCs at undifferentiation (Sox9+YFP+), intermediate (Sox9-YFP+) and advanced (Sox9-α-SMA+YFP+) cell state along the SMC lineage are marked by filled arrowheads, open arrowheads and arrows, respectively. Volcano plot shows the genes significantly up- (red) and down- (blue) regulated in (I) Proliferation of CNCCs was examined using the immunofluorescence staining with anti-Ki67 antibody (Red). Solid arrowheads mark the proliferating CNCCs in SLVs at E13.5 and E17.5. (K) Increased percentage of Tbx20+ CNCCs in SLV of Vcl mutants at E13.5 and E17.5. SLV: semilunar valves; AoV, aortic valve; PV, pulmonary valve; VS, ventricular septum; v.s., ventricular side.

A total of 6564 high-quality single-cell transcriptomes of CNCCs (1623 control and 4941 Vcl mutant, Supplementary Data 1) were obtained after stringent quality control. We clustered the 6383 OFT CNCCs into 6 distinct cell clusters (Fig. 4B & Supplementary Data 2). Consistently, Vcl was down-regulated in all cell clusters in mutants (Supplementary Fig. 6). We found two (c2 & c3) neural- and four (c1, c4–6) nonneural lineage-distinct clusters, where the four nonneural clusters were grouped according to their cell state. c1 represents the progenitors with the highest proliferative potential based on the expression of cell cycle markers (Ube2c, Cdc20, Ccnb1). c4 and c5 are the intermediate mesenchymal (MES) populations expressing mesenchymal markers (Sox9, Lum, Pdgfra). c6 comprised CNCCs at the most mature cell state corresponding to vascular smooth muscle cells (VSMC) that expressed smooth muscle cell-associated contractile proteins (Acta2+ and Tagln+) (Fig. 4B). The proportion of each cluster in control and mutant was estimated. An expansion of the intermediate MES population (c5) (control: 21.2 % and mutant: 26.9 %) with a concomitant reduction in the most mature VSMC population (c6) (5.6 % in control versus 4.1 % in mutant) was observed in the mutants, while the proportions of the other clusters in control and mutant were highly comparable (Fig. 4C). It may suggest that the MES-to-VSMC differentiation is interrupted in the mutants. Lineage relationship of these clusters is further illustrated in the cladogram (Fig. 4D). The RNA velocity and expression profiling analyses also indicated that the majority of cells in c6 were mature VSMCs, which lost the mesenchymal signature (Sox9-, Pdgfra-, Lum-) with concomitant expression of the VSMC marker Acta2 (α-SMA) (Fig. 4E & F). Compared to the control, c6 cells in the Vcl mutant OFTs still retained the mesenchymal signature (Sox9+, Acta2Low) (Fig. 4F), suggesting the immature nature of the mutant CNCCs. Consistent with these observations, immunofluorescence staining revealed that significantly more immature (Sox9+) CNCCs and fewer mature VSMCs (c6 cells, Sox9−α SMA+) were found in both the SLV (Fig. 4G) and VS (Fig. 4H) of the Vcl mutant at E13.5 than in the control. Concordantly, a significantly higher percentage of CNCCs remained proliferative (Ki67+) and expressed Tbx20 (proliferative precursor populations [23]) in the fused SLV of the mutants when compared to that in the control (Fig. 4I & J), indicating that most of the CNCCs remained uncommitted in the SLV of the mutants. These immature CNCCs remained proliferative while no obvious cell death of valvular cells was detected (Supplementary Fig. 7), leading to the hyperplastic valves as seen in the mutants.

Perturbed transforming growth factor-beta (Tgf-β) pathway in Vclnull CNCCs during the MES-to-VMSC transition

Our functional data suggested perturbed VSMC differentiation of CNCCs in the Vcl mutant, and the scRNA-seq data indicated that more mutant CNCCs remained in the intermediate cell state. These observations raised the possibility that the mutant CNCCs may be functionally compromised or immature, such that they fail to direct OFT development. We next sought to identify the interrupted genes and pathways during the formation of c6 by comparing the transcriptomes of c6 in the Vcl mutant to the control. A total of 45 upregulated and 201 downregulated differentially expressed genes (DEGs) were identified in c6 in Vcl mutants, and various transcription factors implicated in NCC development, such as the Notch pathway (Hey1, Hes1), activator protein 1 (AP-1) (Fosb/Jun) and Nfia (a modulator of myogenic lineage differentiation), were downregulated in Vcl mutants (Fig. 5A & Supplementary Data 3, p < 0.05, FDR < 0.05). To identify the biological pathways underlying the OFT defects, we performed Gene Ontology (GO) enrichment analysis with the down-regulated DEGs. Among the enriched pathways, genes associated with “aorta morphogenesis”, “mesenchymal cell differentiation”, “outflow tract morphogenesis”, “extracellular matrix (ECM) organization”, “integrin-mediated signaling pathway” and “response to Tgf-β” were found to be the most relevant based on their biological functions. These data highlighted that loss of Vcl perturbs the cross-talk among the Tgf-β signaling pathway, integrins, and the extracellular matrix [9] and interrupts the interactions between CNCCs and their neighborhoods, leading to aberrant OFT development (Fig. 5B). c5 cells (Sox9- and Actalow) likely represented the CNCCs at the intermediate cell stage, bridging the early Sox9+ mesenchymal progenitors (c4) and the mature VSMC (c6, Sox9- and Actahigh). When compared to the control cells, a total of 49 and 484 genes were found to be up- and downregulated, respectively, in c5 mutant cells (Supplementary Data 4). Additional transcription factors that direct progenitor progression (Elk3) [24] and migration (Zfp) (Fig. 5C), the Wnt pathway, the stress-activated MAPK (p38/JNK) cascade and the biological pathways associated with cell proliferation and migration/organization (Fig. 5D) were downregulated in the mutant cells in this cell state. It is noteworthy that the stress-activated MAPK (p38/JNK) cascade, which is downstream of the Tgf-β pathway, was also downregulated in c5, raising the possibility that dysregulation of the Tgf-β pathway affects the transition of intermediate cells to mature VSMC (from c5 to c6).

Fig. 5.

Fig. 5

Perturbation of Integrin-TGF-β-p38 pathway Vcl-/- CNCCs. Volcano plot shows the genes significantly up- (red) and down- (blue) regulated in (A) c6 and (B) c5 in Vcl mutant compared to the control. Top differentially expressed transcription factors (TFs) are labelled. Barplot shows the disrupted pathways in (C) c6 and (D) c5 in Vcl-/- CNCCs. (E) Schematic diagram of Vcl-Integrin-TGF-β-p38 pathway. (F) the semilunar valves and (G) ventricular septum of the mutants as monitored based on the expression of phospho-Smad2 (pSmad2) using immunohistochemistry. Quantitative data are represented as mean+/- SEM and shown in the scatter plots. P-value less than 0.05 was considered significantly different. Student t-test, 2-sided. AV, aortic valve; PV, pulmonary valve; SLV, semilunar valve. White dotted lines outline the SLVs. White solid and open arrowheads mark the positively and negatively stained CNCCs in the valves, respectively. (H) Reduced expression of Itgb1 in Vcl cKO (Vcl KOWnt1-Cre;RosaYFP) at E13.5 was revealed by RNAscope and immunohistochemistry on sagittal sections. SLV: semilunar valves; PV, pulmonary valve; VS, ventricular septum; v.s., ventricular side.

The integrins-ECM-mediated Tgf-β pathway has been well established [8,9], in which Vcl serves as an adaptor protein stabilizing integrins [25] (Fig. 5E). To examine whether the integrins-ECM-mediated Tgf-β pathway is interrupted in Vcl mutants, we monitored Tgf-β pathway activity based on the expression of phosphorylated Smad2 (pSmad2). We found that the percentages of CNCCs expressing pSmad2 in the SLVs (Fig. 5F) and VS (Fig. 5G) of the mutants were significantly lower than those in control embryos. Concordantly, the expression of Itgb1 was down-regulated in the mutants (Fig. 5H), suggesting that loss of Vcl destabilizes Itgb1 and perturbs the Tgf-β pathway in CNCCs in Vcl mutant OFT.

Loss of Vcl in CNCCs interrupts the CNCC-SHF interaction in the OFT

Vcl is essential for cell–cell interactions and our scRNA-seq data showed perturbation of integrins-ECM in Vcl mutant OFT cells. Thus, we also examined whether loss of Vcl in CNCCs perturbs the communications between CNCCs and their neighborhoods, which may interrupt the OFT modeling. To this end, we performed ST-seq analysis on sagittal sections of E13.5 embryos using replicate tissue sections approximately 16 µm apart. We sequenced the samples to a median depth of 171,916,081 reads (interquartile range 153,220,340–200,680,799), which yielded a mean of 6629 genes and 25,122 unique molecular identifiers (UMIs) per spot (Supplementary Fig. 8A). CNCC-specific down-regulation of Vcl in mutant cells was consistently observed. Its expression in SHF derivatives remained unchanged in the mutant (Supplementary Fig. 8B). The data were normalized and clustered using Seurat, and 10 ST clusters were found covering the heart region (Supplementary Fig. 8C). We then further defined the OFT region and the corresponding boundary by detecting CNCC derivative spots based on YFP expression. We found that the YFP+ spots in five ST clusters (Clusters 3, 30, 53, 56 and 96) highly aligned with the OFT region as defined histologically (Fig. 6A & Supplementary Fig. 8D). Therefore, the downstream analyses were restricted to these spots, where the YFP-positive (YFP+) and YFP-negative (YFP-) cells within these clusters were considered CNCC and SHF derivatives, respectively. We then integrated and jointly analyzed the replicate sections from the control (C1 & C2) and mutant (M1 & M2) groups to cover two independent regions of the OFT. scRNA-seq data-based deconvolution was then used to transfer the cell annotations in the scRNA-seq data to the ST data to estimate the composition of each spot in the ST data. From this analysis, most of the MES and VSMC populations of CNCCs as defined by scRNA-seq data could be detected on the sections, including the two small populations of neural derivatives of CNCCs, neurons and Schwann cells. Canonical markers for MES cells (Lumhigh Pdgfrahigh) and VSMCs (Mhy11high Cxcl12high) were used to predict the cell types in each cluster, where MES cells and VSMCs were found in Clusters 56 & 96 and Clusters 30 & 53, corresponding to the proximal and distal OFT, respectively. Cells in Cluster 3 coexpressed the early VSMC (Cxcl12high) and MES markers, representing the cells at the intermediate stage (Supplementary Fig. 9). Consistently, the CNCC-derived MES and VSMC populations were largely localized at the proximal and distal OFT, respectively, suggesting a sequential wave of differentiation along the OFT from the distal to the proximal regions (Fig. 6B). Although a very similar spatial distribution of MES and VSMC populations in the control and mutant OFTs was observed, in line with our data from the immunohistochemical analyses, more MES cells were found in the distal OFT of the Vcl cKO mutant, and 8 % versus 20 % of the total OFT CNCCs belonged to MES in control and mutant sections, respectively (Fig. 6B & C). The cell–cell communications among CNCC and SHF derivatives within the OFT were then predicted based on the expression of ligand‒receptor (L–R) pairs using CellChat [26]. The Vcl cKO group showed a weaker overall interaction among OFT cells, as evidenced by the lower number and strength of interactions (Fig. 6D). The involved pathways and the corresponding communication probabilities between OFT cells in the control and mutant were further illustrated in the bubble plot, where the communications among CNCC derivatives were almost completely abolished. In control, Ptn-Sdc1/3 and Itgb1/angpt2 represented the most significant contributing pathways for the interactions among CNCC derivatives and bridging CNCC and SHF derivatives, respectively, while the mutant cells lost most of these interactions (Fig. 6E and Supplementary Data 5). In addition, examination of the potential interaction dynamics among VSMC, MES and SHF derivative clusters revealed that interactions largely occur between VSMCs and SHF derivatives in the control OFT, with a total of 21 contributory pathways that were significantly interrupted in the mutant, including Fn1 (Fibronectin) and Ptn (Pleiotrophin), where their overall interacting pathway activities were almost completely abolished in the mutants (Fig. 6F and Supplementary Data 6). The Fn1 pathway activity was mainly driven by Fn1-Sdc4 and Fn1-Itga8/Itgb1 interactions. Fn1, Sdc4 and Itga8/Itgb1 encode fibronectin, syndecan proteoglycan and integrins, respectively, and the interactions among ECM and integrins proteins are essential for CNCC migration and VSMC differentiation in the OFT [[27], [28], [29]]. Ptn-Sdc3 and Ptn-Sdc4 represented the potential L-R pairs underpinning Ptn signaling for VSMC-to-SHF signal flow in control, this signaling was not detected in the mutant (Supplementary Data 6). On the other hand, the Fn1 signals were bidirectional, bridging CNCCs and SHF derivatives in control, while a shift in dynamic interactions of a greater role of MES was observed in the mutant, where the VSMC-SHF derivatives interaction became insignificant (Fig. 6G). Spatial maps shown in Fig. 6H further illustrated the distribution of gene expression levels in a binary format across a representative control and Vcl cKO sample. Most of the VSMCs exhibited downregulation of both Ptn and Sdc3 in the distal OFT of Vcl cKO. In contrast, Ptn and Sdc3 were predominantly expressed in MES in the proximal OFT region, indicating a loss of VSMC-SHF interaction and/or a shift in signaling patterns. A similar change in expression pattern was observed for Fn1 and Sdc1, with decreased levels of both in VSMCs located within the distal OFT of Vcl cKO, accounting for the retarded signal flow between VSMC and SHF. Furthermore, Sdc1 expression was mainly detected in MES cells situated in more proximal regions, causing a shift in the SHF-directed Fn1-Sdc1 signal from the VSMCs to MES cells, in concordance with the data shown in Fig. 6G. These were concurrent with the loss of the Ptn expression (Fig. 6I), and aberrant expression of Fn1 (Fig. 6J) in Vcl cKO mutant, as illustrated by the RNAscope assay and immunohistochemistry analyses, respectively. The expression of Fn1 was detected in Sox9- CNCC derivatives (as marked with open arrowheads) in the atrioventricular valves of control at E13.5, while it was mainly expressed in Sox9+ CNCC derivatives in mutants, corresponding to the immature OFT cells (Fig. 6J). Taken together, loss of Vcl perturbed the signal flows among various OFT cells, implying the potential interruption in CNCC-SHF communications.

Fig. 6.

Fig. 6

Loss of Vcl perturbs interactions among OFT cells. (A) Spatial projection of the five OFT clusters on the sagittal section of a control embryo at E13.5. (B) Cell-type prediction overlay on ST spots showing the MESCNCC, VSMCCNCC and SHF populations defined based on the YFP expression and inferred from scRNA-seq on the replicate sagittal sections (that were approximately 16 µm apart from each other) of control (C1 & C2) and Vcl mutant (M1 & M2) embryos at E13.5. (C) Stacked plot shows the proportion of cells inferred for the replicate sections of control and Vcl cKO. (D) Bar charts show the number and strength of cell–cell communications between OFT cells as predicted by CellChat. (E) Bubble plot shows the communication probabilities among OFT cells. (F) The overall signaling pattern of the top four signaling pathways underlying the interactions between VSMCCNCC and SHF. (G) Chord diagram shows the interactions between different cell clusters through Fn1-Sdc1 signaling network in control and mutant OFT cells. (H) Reduced expressions of Ptn and Itgb1 in Vcl cKO (Vcl KOWnt1-Cre;RosaYFP) at E13.5 were revealed by RNAscope and immunohistochemistry on sagittal sections. (I) Expression of Fn1 in Sox9- VSMCCNCC (open arrowheads) and Sox9+ MESCNCC (filled arrowheads) was detected in control and Vcl cKO, respectively, as revealed by immunohistochemistry. RV, right ventricle.

VCL-mediated TGF-β and p38/JNK pathways direct VSMC lineage differentiation of CNCCs as revealed by human pluripotent stem cell (hPSC) based-CNCC model

To demonstrate the direct involvement of Vcl in VSMC differentiation, we established an in vitro model using human pluripotent stem cells (hPSCs). First, an inducible CRISPR/Cas9 (iCas9)-hPSC line was generated to specifically knock down VCL in CNCCs, where the expression of Cas9 protein was induced by the addition of doxycycline (DOX) (Supplementary Fig. 10). A two-step differentiation protocol was used to model CNCC-MES-VSMC differentiation. iCas9-hPSCs were first directed to the NCC lineage by dual-SMAD inhibitors (LDN193189 and SB431542) with a WNT agonist (CHIR99021) and then caudalized with retinoic acid (RA) to obtain posterior/vagal NCCs [18,30]. The VSMC differentiation was then induced by culturing the posterior/vagal NCCs in VSMC medium with TGF-β (hereafter posterior/vagal NCCs are referred to as hCNCCs). The VCL gene was knocked down in hCNCCs by transfecting cells with VCL-targeting gRNAs and in the presence of doxycycline (Fig. 7A). As shown in Fig. 7B, when cultured in VSMC medium, hCNCCs (SOX10+) were committed to the mesenchymal lineage and expressed SOX9 on Day 14 and then became VSMCs, where SMA+ VSMCs emerged at approximately Day 21 of differentiation. VCL knockdown CNCCs (DOX+) migrated significantly slower than the wild-type (DOX−) cells as demonstrated in the wound healing assay (Fig. 7C). Subsequent Western blot consistently showed a concomitant reduction in Integrin (ITGB1) expression in VCL knockdown CNCCs (Fig. 7D), indicating that VCL is essential for the integrin expression in both mouse and human CNCCs. Furthermore, significantly fewer FAs were detected in VCL knockdown CNCCs as monitored based on the expression of PAXILLIN, which is primarily located at FAs (Fig. 7E). Down-regulation of integrins and failure in forming FAs account for the retarded migration of VCL knockdown CNCCs. Our data from mouse studies suggested that Vcl mediates VSMC lineage differentiation via TGF-β and p38/JNK pathways. Therefore, we next sought to demonstrate the direct contribution of VCL in mediating these pathways and VSMC differentiation of CNCCs. Phosphorylation of p38 was upregulated when hCNCCs were directed to the VSMC lineage driven by the addition of TGF-β (Fig. 7F), while knocking down VCL abolished VSMC differentiation (Fig. 7G), accompanied by the downregulation of phospho-p38 (Fig. 7H). In line with these observations, the addition of a p38 antagonist (BIRB 796) significantly inhibited the VSMC differentiation of hCNCCs, while the addition of a p38 agonist (LX-4) rescued the VSMC differentiation defect of VCL knockdown cells (Fig. 7I). The VCL-TGF-β-p38 pathway was further illustrated in the Western blot analyses, where Tgf-β- induced phosphorylation of both p38 and pSmad2 required the presence of VCL (Fig. 7J). As evidenced in the in vivo and in vitro models, the VCL-TGF-β-p38 pathway is indispensable for VSMC lineage differentiation of CNCCs.

Fig. 7.

Fig. 7

VCL is required for Integrin-TGF-β-p38/SMAD2 signalings for CNCC development. (A) Schematic shows the human iPSC-based model of SMC differentiation from CNCCs. An inducible-iPSC line was used, where the expression of Cas9 was induced by the addition of doxycycline. Co-transfection of VCL-specific gRNA drove the deletion of VCL gene in CNCCs. (B) Immunocytochemistry with antibodies against CNCC (SOX10), MES (SOX9),) and SMC (α-SMA) markers. (C) Wound healing assay revealed that VCL knock-down cells exhibit retarded migration. (D) Western blot analyses of ITGB1 and VCL. ACTIN was used as the loading control. The relative expression levels were shown in the bar charts (mean +/− SEM from 3 independent experiments). P-value less than 0.05 was considered significantly different. Student t-test, 2-sided. (E) Immunocytochemistry with antibody aganist Paxillin and counterstained with Phalloidin. Bar chart shows the percentage of FAs as marked by Paxillin over the total number of CNCCs. Data shown in the bar charts are mean +/− SEM from 3 independent experiments. P-value less than 0.05 was considered significantly different. Student t-test, 2-sided. Western blot analyses of phospho-p38, total p38 and VCL (F) in NCCs at Days 0, 14 and 21 during the SMC differentiation and (H) in VCL knock-down cells at Day 21 of differentiation. ACTIN was used as the loading control. The relative expression levels are shown in the bar charts (mean +/− SEM from 3 independent experiments). P-value less than 0.05 was considered significantly different. Student t-test, 2-sided. Immunocytochemistry shows the expression of SOX9 and α-SMA (G) in control and VCL knock-down cells and (I) in presence of p38 agonist/antagonist. Data shown in the bar charts are represented as mean +/− SEM. Student t-test, 2-sided. (I) The VCL-TGF-β-p38 pathway in NCCs was demonstrated using Western blot analysis based on the expressions of VCL, phospho-Smad2, phospho-p38.

Discussion

Vcl mutants exhibited a series of abnormalities in the heart. IAA-B was seen in approximately 60 % of the mutants at P0, with a discontinuity in the aortic arch between the left subclavian artery and left common carotid artery. This is likely attributed to the incomplete/delayed colonization of the fourth L-PAA by CNCCs in Vcl mutants, which may lead to regression of the artery, resulting in IAA-B. In addition, the migration defect in both CNCC and SHF lineages into the OFT likely perturbs the subsequent OFT remodeling. Nearly 90 % of the mutants showed an unsplit trunk, or PTA, resulting from the failure to separate the aorticopulmonary septum into the aorta and pulmonary trunk at its proximal end of the OFT. Septation of the OFT involves 3 major steps: endocardial cushion enlargement, fusion of the endocardial cushions, and myocardialization of the fused endocardial cushions to form the muscular septum. Endocardial cushion enlargement relies on the endothelial-to-mesenchymal transition (EndoMT) and proliferation of mesenchymal cells. Both EndoMT (Supplementary Fig. 11) and the proliferation (Supplementary Fig. 4) of CNCC derivatives at the OFT cushion and septal ridges remained largely unaffected in the mutants at E10.5–E12.5. Nevertheless, the mutant CNCCs failed to properly populate the septal ridges and differentiate into VSMCs, which was associated with defective myocardialization of the fused endocardial cushions, leading to septation defects. These problems also caused SLV and ventricular septation defects. On the other hand, the formation of SLV valves requires the expansion of valve primordia and subsequent remodeling and stratification. Downregulations of Sox9 and Tbx20 in cardiac valvular cells are needed for directing CNCCs to the VSMC lineage and suppressing their chondrogenic differentiation potential [31] and for the maturation of myocardial cells during heart development [23], respectively. The persistently high expression levels of Sox9 and Tbx20 in the valvular cells of the mutant likely impaired the VSMC lineage differentiation of CNCCs and perturbed valve patterning. Regarding ventricular septation, all mutants showed ventricular septal defects at P0. This defect could be traced back to E14.5, at which the VS was completely closed in control, while the OFT cushion was not properly formed at the top of the VS in most of the mutants at this stage due to insufficient numbers of CNCCs. In a few cases, the OFT cushions were properly developed and filled with CNCCs, but no myocardialization was induced, indicating the dysfunction of mutant CNCCs.

All these data indicated that the loss of Vcl interrupts the function of CNCCs, particularly the MES-to-VSMC differentiation, to direct OFT morphogenesis. To further unveil the molecular mechanisms underlying CNCC differentiation and how it affects their interactions with the neigbouring cells, perturbing the OFT development, single cell and spatial transcriptomes of control and mutant cells were compared. First, we performed an unsupervised clustering of the CNCC derivatives using scRNA-seq data, in which cells were clustered unbiasedly and based on their cell state during the VSMC differentiation of CNCCs. This approach allowed us to identify the DEGs associated with the differentiation defect of the mutant CNCCs. From this analysis, we identified 4 major cell populations of CNCCs on the VSMC lineages, including the two populations (c1 & c4) in the immature cell state with high Sox9 expression and two Sox9-negative populations, the intermediate CNCCs (c5) and the mature VSMCs expressing high levels of Acta2 (αSMA) (c6). By defining the DEGs on c6, we further found that the Tgf-β/integrin/extracellular matrix pathways are dysregulated in the mutant cells. The Tgf-β/Smad signaling pathway has been suggested as a mediator of the transition of valvular mesenchymal cells into their activated form. Consistently, immunofluorescence staining showed that the expression of pSmad2 was reduced in the mutant. Building the differentiation trajectory of VSMC lineage differentiation further linked the early molecular events with the mouse phenotypes, by which c5 was defined as the intermediate cell state during the MES-VSMC transition and dysregulation of the p38 pathway might interrupt the MES-VSMC progression of Vcl mutant CNCCs.

How the perturbed VSMC differentiation of CNCCs alters the cross-talks among CNCCs and SHF derivatives was further delineated by examining the interaction dynamics of the OFT cells using ST RNA-seq. ST RNA-seq depicted the anatomical expression pattern of OFT cells and aligned their transcriptomes with their spatial positions across each sagittal section. Through an integrative analysis with the single-cell transcriptome data, we further resolved CNCC-derived MES cells and VSMCs as identified by scRNA-seq by their spatial profiles. A “digital in situ” visualization of the differentiation wave of CNCCs along the OFT was also established by correlating the pseudotime of the differentiating CNCCs with their spatial patterns. More importantly, the ST data gave us unprecedented anatomical resolution of physical and molecular interactions between OFT cells. Based on the expression of the L-R pairs with their spatial constraints, the interaction dynamics among the OFT cells were defined. Multilevel analyses of ST data further highlighted the importance of Vcl in coordinating interactions among CNCC and SHF derivatives via various ECM (Tnc, Sdc1, Fn1) proteins and integrins during OFT morphogenesis, reinforcing our findings from the DEG analysis of the scRNA-seq data with the enrichment of the Tgf-β-integrins-ECM pathway. Deleting Vcl in CNCC lineage disrupted the interaction of CNCCs with ECM, leading to the absence of the Tgf-β signal. This disturbance affected SMC differentiation of CNCCs, which in turn altered cell morphology, impacting cell–cell junctions and communication SHF. In addition, our ST data also suggested that VSMCs are the key player directing communications with SHF cells, so the immature nature of Vcl-/- CNCCs likely hindered CNCC-SHF coordination, such as myocardialization and prohibited OFT morphogenesis.

Given that there is no proper inducible Wnt1-Cre mouse line available to delineate the stage-specific roles of Vcl along VSMC lineage differentiation in vivo, the complementary use of scRNA-seq and hPSC model represents an alternative approach to demonstrate the direct involvement of Vcl in VSMC lineage differentiation of CNCCs. Exploiting the hPSC-based model with a drug-inducible CRISPR system, we recapitulated CNCC-to-MES-to-VSMC differentiation in vitro and knocked down VCL specifically at the NC stage. It directly demonstrated VCL is indispensable for p38 and Tgf-β pathways for directing the MES-to-VSMC cell state transition during the VSMC lineage differentiation of CNCCs. Although it is known that p38 and Tgf-β signaling pathways are crucial for VSMC differentiation, at what cell stages these signalings are involved and how these signaling pathways were activated in vivo during the OFT development remains largely unclear. Our data from scRNA-seq and iPSC studies suggest that the activation of p38 pathway is followed by Tgf-β signaling during the later stage of CNCC development for directing the MES-to-VSMC transition. In addition, we also demonstrated the potential implication of CNCC-ECM interaction in activation of the p38 and Tgf-β signaling pathways where Vcl is a key molecule bridging the integrins, ECM and cellular signalings to mediate the CNCC development.

In this study, we focused on OFT development, where Vcl mediated Tgf-β signaling pathways to control VSMC lineage differentiation. Activation of Tgf-β signaling in CNCCs becomes detectable only at around E13.5. Thus, it may suggest that other signaling pathways may be involved to control VSMC lineage differentiation of CNCCs for the formation of the aorta that happens at E9.5-E10.5 before the onset of Tgf-β signaling. More work should be performed with embryos at earlier developmental stages to further address the molecular mechanisms underlying the great artery development. Given the significance of Vcl in neural crest and OFT development, future genetic screening of patients with neurocristopathies should be considered to assess its potential clinical implications.

In summary, NCC-specific loss of Vcl resulted in cardiac defects similar to those in the CNCC ablation model, indicating the indispensable roles of Vcl in CNCCs and OFT development. Vcl possesses at least two roles, (1) in mediating p38 and Tgf-β signalings to direct the MES-to-VSMC differentiation of CNCCs and (2) in cell-matrix interactions which are essential not only for the VSMC differentiation of CNCCs, but also their interaction with SHF derivatives and their migration to control the OFT remodelling. As a key adaptor protein of FAs, Vcl is essential for the signal flows among various OFT cells. Understanding the molecular events underlying the OFT defects, as seen in Vcl mutants, provides unprecedented insights into the signaling pathways directing CNCC development and cell communications among various OFT cells during OFT remodeling.

Ethic statement

All experiments were performed in accordance with procedures approved by the committee on the Use of Live Animals, the University of Hong Kong (CULATR 4168-16 and CULATR 5701-21). Our study examined male and female animals, and similar findings are reported for both sexes.

Author contributions

R.X.W. and X.W. performed the functional studies. M.L., A.Y.P., and Z.L. performed bioinformatics analyses. K.H.K., C.C.H., A.J., and E.S.W.N. supervised the whole project and prepared the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We thank Drs. Frank Lai and Kathy Lui for establishing the iCas-iPSC line and conducting the in vitro studies. Confocal imaging and RNA sequencing were performed using equipment maintained by the Li Ka Shing Faculty of Medicine Faculty Core and the Centre for PanorOmic Sciences, the University of Hong Kong, respectively. The work described in this paper was substantially supported by seed funding from the University of Hong Kong and the GRF grant from the Research Grants Council of Hong Kong Special Administrative Region, China Hong Kong (HKU17110818 & 17107322) to E.S.W.N.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jare.2025.10.062.

Contributor Information

Asif Javed, Email: javed@hku.hk.

Elly Sau-Wai Ngan, Email: engan@hku.hk.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary Data 1
mmc1.xlsx (9.9KB, xlsx)
Supplementary Data 2
mmc2.xlsx (122.8KB, xlsx)
Supplementary Data 3
mmc3.xlsx (187.3KB, xlsx)
Supplementary Data 4
mmc4.xlsx (225.9KB, xlsx)
Supplementary Data 5
mmc5.xlsx (11.1KB, xlsx)
Supplementary Data 6
mmc6.xlsx (18.8KB, xlsx)
Supplementary Data 7
mmc7.docx (48.7KB, docx)
Supplementary Data 8
mmc8.pdf (2.3MB, pdf)
Supplementary Data 9
mmc9.pdf (138.2KB, pdf)

References

  • 1.Jiang X.B., Rowitch D.H., Soriano P., McMahon A.P., Sucov H.M. Fate of the mammalian cardiac neural crest. Development. 2000;127(8):1607–1616. doi: 10.1242/dev.127.8.1607. [DOI] [PubMed] [Google Scholar]
  • 2.Laugwitz K.L., Moretti A., Caron L., Nakano A., Chien K.R. Islet1 cardiovascular progenitors: a single source for heart lineages? Development. 2008;135(2):193–205. doi: 10.1242/dev.001883. [DOI] [PubMed] [Google Scholar]
  • 3.Keyte A., Hutson M.R. The neural crest in cardiac congenital anomalies. Differentiation. 2012;84(1) doi: 10.1016/j.diff.2012.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chen W., Liu X., Li W., Shen H., Zeng Z., Yin K., et al. Single-cell transcriptomic landscape of cardiac neural crest cell derivatives during development. EMBO Rep. 2021;22(11) doi: 10.15252/embr.202152389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Liu X., Chen W., Li W., Li Y., Priest J.R., Zhou B., et al. Single-cell RNA-seq of the developing cardiac outflow tract reveals convergent development of the vascular smooth muscle cells. Cell Rep. 2019;28(5):1346–61 e4. doi: 10.1016/j.celrep.2019.06.092. [DOI] [PubMed] [Google Scholar]
  • 6.De Bono C., Liu Y., Ferrena A., Valentine A., Zheng D., Morrow B.E. Single-cell transcriptomics uncovers a non-autonomous Tbx1-dependent genetic program controlling cardiac neural crest cell development. Nat Commun. 2023;14(1):1551. doi: 10.1038/s41467-023-37015-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Bays J.L., DeMali K.A. Vinculin in cell-cell and cell-matrix adhesions. Cell Mol Life Sci. 2017;74(16):2999–3009. doi: 10.1007/s00018-017-2511-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Chen X., Wang H., Liao H.J., Hu W., Gewin L., Mernaugh G., et al. Integrin-mediated type II TGF-beta receptor tyrosine dephosphorylation controls SMAD-dependent profibrotic signaling. J Clin Invest. 2014;124(8):3295–3310. doi: 10.1172/JCI71668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Munger J.S., Sheppard D. Cross talk among TGF-β signaling pathways, integrins, and the extracellular matrix. Cold Spring Harbor Perspect. Biol. 2011;3(11) doi: 10.1101/cshperspect.a005017. a005017-a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Zemljic-Harpf A.E., Ponrartana S., Avalos R.T., Jordan M.C., Roos K.P., Dalton N.D., et al. Heterozygous inactivation of the vinculin gene predisposes to stress-induced cardiomyopathy. Am J Pathol. 2004;165(3):1033–1044. doi: 10.1016/S0002-9440(10)63364-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Marg S., Winkler U., Sestu M., Himmel M., Schonherr M., Bar J., et al. The vinculin-DeltaIn20/21 mouse: characteristics of a constitutive, actin-binding deficient splice variant of vinculin. PLoS One. 2010;5(7) doi: 10.1371/journal.pone.0011530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Xu W., Baribault H., Adamson E.D. Vinculin knockout results in heart and brain defects during embryonic development. Development. 1998;125(2):327–337. doi: 10.1242/dev.125.2.327. [DOI] [PubMed] [Google Scholar]
  • 13.Duband J.L., Thiery J.P. Spatial and temporal distribution of vinculin and talin in migrating avian neural crest cells and their derivatives. Development. 1990;108(3):421–433. doi: 10.1242/dev.108.3.421. [DOI] [PubMed] [Google Scholar]
  • 14.Okagawa H., Nakagawa M., Shimada M. Immunolocalization of vinculin in the heart of the early developing rat embryo. Anat Rec. 1996;245(4):699–707. doi: 10.1002/(SICI)1097-0185(199608)245:4<699::AID-AR10>3.0.CO;2-K. [DOI] [PubMed] [Google Scholar]
  • 15.Xu X., Francis R., Wei C.J., Linask K.L., Lo C.W. Connexin 43-mediated modulation of polarized cell movement and the directional migration of cardiac neural crest cells. Development. 2006;133(18):3629–3639. doi: 10.1242/dev.02543. [DOI] [PubMed] [Google Scholar]
  • 16.Olson T.M., Illenberger S., Kishimoto N.Y., Huttelmaier S., Keating M.T., Jockusch B.M. Metavinculin mutations alter actin interaction in dilated cardiomyopathy. Circulation. 2002;105(4):431–437. doi: 10.1161/hc0402.102930. [DOI] [PubMed] [Google Scholar]
  • 17.Vasile V.C., Ommen S.R., Edwards W.D., Ackerman M.J. A missense mutation in a ubiquitously expressed protein, vinculin, confers susceptibility to hypertrophic cardiomyopathy. Biochem Biophys Res Commun. 2006;345(3):998–1003. doi: 10.1016/j.bbrc.2006.04.151. [DOI] [PubMed] [Google Scholar]
  • 18.Lai F.P., Lau S.T., Wong J.K., Gui H., Wang R.X., Zhou T., et al. Correction of hirschsprung-associated mutations in human induced pluripotent stem cells via clustered regularly interspaced short palindromic repeats/cas9, restores neural crest cell function. Gastroenterology. 2017;153(1):139–53 e8. doi: 10.1053/j.gastro.2017.03.014. [DOI] [PubMed] [Google Scholar]
  • 19.Jain R., Engleka K.A., Rentschler S.L., Manderfield L.J., Li L., Yuan L.J., et al. Cardiac neural crest orchestrates remodeling and functional maturation of mouse semilunar valves. J Clin Invest. 2011;121(1):422–430. doi: 10.1172/JCI44244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lakkis M.M., Epstein J.A. Neurofibromin modulation of ras activity is required for normal endocardial-mesenchymal transformation in the developing heart. Development. 1998;125(22):4359–4367. doi: 10.1242/dev.125.22.4359. [DOI] [PubMed] [Google Scholar]
  • 21.Plein A., Calmont A., Fantin A., Denti L., Anderson N.A., Scambler P.J., et al. Neural crest-derived SEMA3C activates endothelial NRP1 for cardiac outflow tract septation. J Clin Investig. 2015;125(7):2661–2676. doi: 10.1172/JCI79668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Waller B.R., McQuinn T., Phelps A.L., Markwald R.R., Lo C.W., Thompson R.P., et al. Conotruncal anomalies in the trisomy 16 mouse: an immunohistochemical analysis with emphasis on the involvement of the neural crest. Anat Rec. 2000;260(3):279–293. doi: 10.1002/1097-0185(20001101)260:3<279::AID-AR65>3.0.CO;2-2. [DOI] [PubMed] [Google Scholar]
  • 23.Shelton E.L., Yutzey K.E. Tbx20 regulation of endocardial cushion cell proliferation and extracellular matrix gene expression. Dev Biol. 2007;302(2):376–388. doi: 10.1016/j.ydbio.2006.09.047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Rogers C.D., Phillips J.L., Bronner M.E. Elk3 is essential for the progression from progenitor to definitive neural crest cell. Dev Biol. 2013;374(2):255–263. doi: 10.1016/j.ydbio.2012.12.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bays J.L., Peng X., Tolbert C.E., Guilluy C., Angell A.E., Pan Y., et al. Vinculin phosphorylation differentially regulates mechanotransduction at cell-cell and cell-matrix adhesions. J Cell Biol. 2014;205(2):251–263. doi: 10.1083/jcb.201309092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Jin S.Q., Guerrero-Juarez C.F., Zhang L.H., Chang I., Ramos R., Kuan C.H., et al. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1) doi: 10.1038/s41467-021-21246-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Christensen G., Herum K.M., Lunde I.G. Sweet, yet underappreciated: proteoglycans and extracellular matrix remodeling in heart disease. Matrix Biol. 2019;75–76:286–299. doi: 10.1016/j.matbio.2018.01.001. [DOI] [PubMed] [Google Scholar]
  • 28.Mittal A., Pulina M., Hou S.Y., Astrof S. Fibronectin and integrin alpha 5 play essential roles in the development of the cardiac neural crest. Mech Dev. 2010;127(9–12):472–484. doi: 10.1016/j.mod.2010.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Warkala M., Chen D.Y., Ramirez A., Jubran A., Schonning M., Wang X., et al. Cell-extracellular matrix interactions play multiple essential roles in aortic arch development. Circ Res. 2021;128(3):E27–E44. doi: 10.1161/CIRCRESAHA.120.318200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Zhou D., Feng H., Yang Y., Huang T., Qiu P., Zhang C., et al. hiPSC modeling of lineage-specific smooth muscle cell defects caused by TGFBR1(A230T) variant, and its therapeutic implications for loeys-dietz syndrome. Circulation. 2021;144(14):1145–1159. doi: 10.1161/CIRCULATIONAHA.121.054744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Arai H.N., Sato F., Yamamoto T., Woltjen K., Kiyonari H., Yoshimoto Y., et al. Metalloprotease-dependent attenuation of bmp signaling restricts cardiac neural crest cell fate. Cell Rep. 2019;29(3) doi: 10.1016/j.celrep.2019.09.019. 603-16 e5. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data 1
mmc1.xlsx (9.9KB, xlsx)
Supplementary Data 2
mmc2.xlsx (122.8KB, xlsx)
Supplementary Data 3
mmc3.xlsx (187.3KB, xlsx)
Supplementary Data 4
mmc4.xlsx (225.9KB, xlsx)
Supplementary Data 5
mmc5.xlsx (11.1KB, xlsx)
Supplementary Data 6
mmc6.xlsx (18.8KB, xlsx)
Supplementary Data 7
mmc7.docx (48.7KB, docx)
Supplementary Data 8
mmc8.pdf (2.3MB, pdf)
Supplementary Data 9
mmc9.pdf (138.2KB, pdf)

Articles from Journal of Advanced Research are provided here courtesy of Elsevier

RESOURCES