Endocardial TIE1 synergizes with TIE2 to regulate the atrial internal muscular network assembly

Kai Ding; Beibei Xu; Xinhao Yu; Xiwen Jia; Taotao Li; Xin Shen; Junda Li; Xudong Cao; Yahui Liu; Zhen Zhang; Yulong He

doi:10.7554/eLife.111156.1

eLife Assessment

This study reports the relative importance of Tie1 and Tie2 signaling for atrial versus ventricular trabeculation. It is an important study and is one of the few works to date that have carefully and simultaneously analyzed these two processes. In line with a previous study in zebrafish, the authors demonstrate key differences between atrial and ventricular trabeculation. While the imaging and quantitative data were conducted with solid and validated methodology throughout the manuscript, the work would benefit from more rigourous approaches where Tie1/2 signaling is disrupted prior to the onset of atrial/ventricular trabeculation, to allow for a more direct comparison.

https://doi.org/10.7554/eLife.111156.1.sa4

Significance of findings

important: Findings that have theoretical or practical implications beyond a single subfield

landmark
fundamental
important
valuable
useful

Strength of evidence

solid: Methods, data and analyses broadly support the claims with only minor weaknesses

exceptional
compelling
convincing
solid
incomplete
inadequate

During the peer-review process the editor and reviewers write an eLife assessment that summarises the significance of the findings reported in the article (on a scale ranging from landmark to useful) and the strength of the evidence (on a scale ranging from exceptional to inadequate). Learn more about eLife assessments

Abstract

Atrial cardiomyopathy is characterized by altered atrial structures and the genetic basis underlying the disorders remains inadequately explored. TIE1 variants or loss of function mutations were reported in a subset of lymphedema patients, and it is unknown whether the patients have also cardiac defects in addition to the lymphatic abnormality. We show in this study that endothelial Tie1 and Tek are highly expressed in endocardial cells of atria by the single cell RNA-seq analysis. TIE1 deficiency led to the disruption of atrial morphogenesis with minor defects in the ventricles. The bulk RNA-seq analysis of hearts at the four-chambered stage revealed that gene transcripts related to endothelial cell development and cardiac trabeculation were reduced in the Tie1 mutant mice compared with littermate controls. This was further confirmed by the RNA-seq analysis of atria and ventricles separately, showing more trabecular genes downregulated in the atria including Tek, upon the loss of Tie1. Consistent with the scRNA-seq data, we found that Tie1 and Tek transcripts were higher in atria than in ventricles. Furthermore, the endothelial deletion of Tek resulted in the defective formation of cardiac trabeculae, particularly in atria. Consistently, the loss of Tie1 combined with one null allele of Tek disrupted both atrial and ventricular trabeculation. Surprisingly, defects with the atrial chamber morphogenesis were already detectable 48 hours later upon the induced endothelial loss of Tie1 plus the Tek heterozygous deletion, implying a critical role of endocardium in the organization of the atrial internal muscular network. The synergy of TIE1 and TIE2 in the remodeling of atrial trabeculae was further confirmed at the postnatal stage, while TIE1 insufficiency alone had no obvious effect. Together, findings from this study imply that TIE1 is differentially required for the atrial and ventricular development and acts in synergy with TIE2 to regulate the endocardial cell-coordinated atrial internal muscular network assembly.

Introduction

The endocardium, as a specialized type of endothelial cells (EC) lining the cardiac lumen, plays an important role in vertebrate heart morphogenesis ^1,2. Cardiac trabeculation relies on the coordinated interaction of endocardium and myocardium ^3,4. The endocardium senses mechanical forces and provides signals for the myocardial morphogenesis ⁵. When endocardial cells fail to form buds, the assembly and extension process of trabeculae are disrupted ^2,6. Defective trabecular formation could lead to embryonic lethality or congenital heart disease ^7–9. Cell lineage studies have also shown that the endocardium is an important source of the coronary vascular endothelial cells, and gives rise to most mesenchymal cells that constitute the cardiac cushions and subsequently contribute to the heart valves ¹⁰. Atrial internal muscular networks, mainly pectinate muscles running in an anterolateral direction, are found on the anterior surface of both right and left atrial walls and corresponding auricles. In contrast to the better understanding of ventricular trabecula morphogenesis, the dynamic process and regulatory networks of atrial trabeculation are inadequately elucidated. Likewise, the atrial cardiomyopathy is characterized by altered atrial structure as well as function, and pathological mechanisms underlying the disorders are largely unknown.

Cardiac morphogenesis is a complex process involving multiple pathways to orchestrate the reciprocal paracrine signaling between the endocardial and other cardiac cells. Specifically, myocardial VEGFA activates VEGFR2 signaling for coronary vascular growth and also participate in the regulation of cardiac trabeculation ¹¹. Endocardial cell-derived NRG1 regulates myocardium development via its receptors ERBB2/4 in cardiomyocytes. Mice null for Nrg1 or its receptors display hypoplastic ventricular wall lacking normal trabeculation ¹². The aberrant NOTCH1 signaling was found to disrupt the heart morphogenesis ^13,14. NRG1 has been show to regulate VEGFA expression in the myocardium, and VEGFA-VEGFR2 signaling is required for the NOTCH1 pathway restriction in endocardium during trabeculation. NOTCH1 signaling was shown to promote extracellular matrix (ECM) degradation during the formation of endocardial projections while NRG1 promoted myocardial ECM synthesis for trabecular rearrangement during the process of cardiac trabeculation ¹⁵. It has recently been shown that endocardium and myocardium interacted directly through tunneling nanotube–like structures extending from cardiomyocytes to endocardial cells, contributing to NOTCH1 activation ¹⁶. Factors involved in the synthesis or degradation of cardiac jelly, a layer of extracellular matrix between endocardium and myocardium, also play essential roles in the process of chamber morphogenesis, including glycosaminoglycan proteins such as versican and ADAMTS family proteins ^17,18. The ANGPT1-TIE2 pathway plays a crucial role in cardiovascular development ^19–22. ANGPT1 is expressed by cardiomyocytes, particularly in atrial myocardium and ventricular trabeculae, and participates in the atrial morphogenesis by regulating the spatiotemporal degradation of cardiac jelly ¹⁹. The endocardial specific deletion of Tek resulted in the fewer but thicker ventricular trabeculae as well as impaired endocardial sprouting, potentially via the paracrine suppression of retinoic acid signaling and proliferation in trabecular cardiomyocytes ²³.

Endothelial TIE1 shares high homology with TIE2 and could form TIE1/TIE2 heterodimers ²⁴. TIE1 has been shown to participate in the regulation of blood vascular and lymphatic network formation and also aortic valve development ^25–29. It has recently shown that TIE1 and TIE2, via the PI3K-AKT-mediated signaling pathway for the protein stability of COUP-TFII, regulate venous specification ^28,30. Loss of the transcription factor COUP-TFII was reported to disrupt both the vein and atrial development ^31–33. Tie1 mutations or variants have also been linked to lymphatic disorders ^34,35. It remains to be characterized whether Tie1 mutations could affect heart structure and function, and how TIE1 participates in the process of cardiac wall morphogenesis. To investigate the role of TIE1 and TIE2 in the process of endocardial cell-mediated chamber formation, we employed genetic mouse models targeting Tie1, Tek (encoding TIE2) or both at different developmental stages. Based on the analysis of the published single cell RNA sequencing data (scRNA-seq) by Feng et al. ³⁶ and the RNA-seq analysis of atria and ventricles from the Tie1 mutant and littermate control mice, we found in this study that Tie1 and Tek expression were higher in atrial than ventricular endocardial cells. TIE1 is differentially required for the atrial and ventricular trabeculation, acting in synergy with TIE2.

Materials and Methods

Mouse models

All animal procedures performed conform the guidelines from Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes; and the experiment protocols were approved by the Animal Care Committee of Soochow and Nanjing University Animal Center (MARC-AP#YH2/SUDA20250507A03). All the mice used in this study were housed in a SPF (specific pathogen free) animal facility with a 12/12 hours dark/light cycle and were free to food and water access. Normal mouse diet (Suzhou Shuangshi Experimental Animal Feed Technology, Co, Ltd) and cage bedding (Suzhou Baitai Laboratory Equipment, Co, Ltd) were used. Two genetically modified mouse models targeting Tie1 gene were employed in this study. One mouse line targets TIE1 intracellular kinase domain (ICD), with exon 15 and exon 16 floxed, Tie1ICD^Flox/Flox³⁷. The other line is a knockout first mouse model established from EUCOMM embryonic stem cells (EPD0735-3B07) targeting Tie1 gene (Tie1^tm^1a^/tm^1a), in which targeting cassette is recombined downstream of exon 7 (with exon 8 floxed) ²⁸. Tek knockout mouse model was generated as previously reported ³⁸ and was crossbred with Tie1^ΔICD/ΔICD mouse model to obtain Tie1ICD and Tek double knockout mice (Tie1^ΔICD/ΔICD;Tek^+/-). To generate mice with endothelial cell–specific gene deletion mouse models, we employed the Cdh5-Cre^ERT² mouse line ³⁹. In all the phenotype analysis, wildtype or heterozygous littermates were used as controls. For the genotyping of Tie1 and Tek mouse lines, the primers used were as previously described ^28,37. The genetic background of Tie1^tm^1a^/tm^1a is on C57BL/6 N, and the other lines are on C57BL/6J or C57BL/6J/SV129. Mice were sacrificed by asphyxiation with rising concentration of carbon dioxide gas, followed by cervical dislocation and tissue collection.

Induced gene deletion

Induction of gene deletion was performed as previously described by the tamoxifen treatment ^28,38,40. Briefly, pregnant mice were treated with tamoxifen (Sigma-Aldrich, T5648-5G) at E10.5-E11.5, E12.5-14.5 or E12.5-15.5 (1 mg/per mouse for 2-4 consecutive days by intraperitoneal injection) and embryos were collected for analysis at the specified stages. For the postnatal studies, new-born pups were treated by four daily intragastric injections of tamoxifen starting from postnatal day 1 (P1, 60 μg/per mice daily). Tissues were collected for analysis at postnatal day 21. The genotypes of the Tie1/Tek knockout and control mice are as follows: Tie1ICD^Flox/-;Cdh5-Cre^ERT² named as Tie1ICD^iECKO, Tek^Flox/-;Cdh5-Cre^ERT² named as Tek^iECKO, Tie1ICD^Flox/-;Cdh5-Cre^ERT²;Tek^+/- named as Tie1ICD^iECKO;Tek^+/-.

Embryo preparation

To investigate the developmental process of cardiac chamber formation, female mice were mated in the late afternoon and vaginal plugs were checked in the morning of the following day. The embryonic stages are estimated considering midday of the day on which the vaginal plug is present as embryonic day 0.5 (E0.5). Tail tips from mutant and control embryos were used for genotyping, and heart tissues from the embryos were collected for the subsequent histology and RNA analysis. For the RNA preparation, heart tissues were harvested and placed in cold PBS. The residual blood from the hearts was cleaned and heart tissues were snap-frozen in liquid nitrogen. For the separate analysis of atria and ventricles, heart tissues were collected and dissected before snap-freezing.

Immunostaining

For the frozen tissue section staining, tissues were collected and fixed in 4% paraformaldehyde for 2 hours at 4°C, followed by the incubation in 20% sucrose overnight before being embedded in optimal cutting temperature (OCT) compound. Frozen sections (10 µm) were used for the immunostaining analysis of cardiac chamber structures and coronary vessels. For the visualization of atrial internal muscular network-trabeculae structures, the whole-mount immunostaining was performed. Heart tissues were fixed in 4% paraformaldehyde overnight at 4°C, followed by blocking with 3%(w/v) skim milk in PBS-TX (0.3% Triton X-100) and incubating with the primary antibodies. The antibodies used were: hamster anti-mouse PECAM1(MAB1398Z; Millipore); rat anti-mouse CD31 (553370; BD Pharmigen), rat anti-mouse Endomucin (14-5851; eBioscience), goat-anti-mouse DLL4 (AF1389; R&D) and goat anti-human TIE1 (AF619; R&D). Appropriate Alexa 488(Invitrogen), Cy3(Jackson ImmunoReaearch) conjugated secondary antibodies were used. All fluorescently labeled samples were mounted and analyzed with a confocal microscope (Olympus Flueview 3000) or Leica MZ16F fluorescent dissection microscope. For comparison, the parameters were kept consistent for all the confocal microscopic imaging in this study.

Bulk RNA sequencing analysis of atria and ventricles

For the bulk RNA-seq analysis of heart tissue, the whole hearts from mutant and control embryos were collected. Atria and ventricles were separated and tissues from two embryos with the same genotype (Tie1^tm^1a^/tm^1a and control) were pooled for the RNA preparation and further analysis. According to the manufacturer’s instructions, total RNA was extracted from the tissues using Trizol (Invitrogen). The procedures for the RNA-seq analysis were performed as previously reported ²⁸. Briefly, sequencing was performed on DNBSEQ platform with PE150 (read length, BGI-Shenzhen, China), and the sequencing data were processed according to the predefined analysis pipeline. The differential expression analysis was conducted using the R package DESeq2 (v 1.42.1). Functional enrichment analysis was performed using clusterProfiler (v4.10.1) and org.Mm.eg.db (v 3.18.0). The hallmark gene sets used in the GSEA (Gene Set Enrichment Analysis), including the hallmarks for trabeculation and EC migration, are mainly defined according to the transcriptome atlas of murine endothelial cells, the molecular signatures database (mh.all.v2024.1.Mm.symbols) and the related GO terms ^28,41,42. The raw and processed data of RNA-seq analysis will be deposited in GEO after the publication.

Single cell RNA-seq data analysis

The single-cell RNA sequencing data by Feng et al. ³⁶ were processed and analyzed with the Seurat package (v5.2.1) following the same parameters and criteria as described to ensure consistency and comparability. Major cell types were annotated based on established marker genes: pan-cardiomyocyte (Ttn), atrial cardiomyocytes (Sln), ventricular cardiomyocyte (Myl2), pan-endothelial cells (Pecam1), endocardial cells (Npr3), coronary endothelial cells (Fabp4), epicardial cells (Wt1, Tbx18, and Aldh1a2), fibroblast-like cells (Postn), mural cells (Pdgfrb), red blood cells (Hba-a1), macrophages (Adgre1). Following the identification of endocardial cells, subsequent analysis was performed using the Seurat package and normalized by the “LogNormalize” method, highly variable genes were identified with the ‘FindVariableFeatures” function (selection.method = “vst”, nfeatures = 2000). The optimal number of principal components (dims = 1:22) was determined by the ElbowPlot method. Cells were then clustered using the “FindNeighbors” and “FindClusters” functions (dims= 1:16, resolution = 0.5). Finally, Unsupervised clustering of cells will be performed by “RunUMAP” and gene expression was visualized with “FeaturePlot”. Ligand–receptor interactions between cardiomyocytes and endocardial cells during the embryonic stage (E10.5-E18.5) were identified using the CellChat package (v1.6.1). Atrial and ventricular cell subsets were analyzed separately using the mouse ligand-receptor database (CellChatDB.mouse) and communication probabilities were computed (min.cells = 10) to identify significant interactions.

Quantitative analysis of cardiac wall and trabeculae

For the quantification of trabecular structures in atria and ventricles, the EMCN positive areas was measured. The PECAM1 positive area was quantified to indicate the coronary vessels within the ventricular wall. The cardiac wall thickness was also measured as described in the supplemental figure 2. For the quantitative analysis of dorsal coronary veins, images were taken under a fluorescence microscope and kept constant near the venous sinus for all of the samples. Horizontal lines were evenly laid on the images, and the diameters of veins crossed with these lines were measured and analyzed using Image Pro Plus (Media Cybernetics, Inc., Bethesda, MD). The ratio of heart to embryo length was used to represent heart size. The starting point for the heart measurement was the junction of aorta and pulmonary artery, and the junction of left and right ventricles at the apex as the end point. All the quantification was measured and analyzed by using Image Pro Plus. Furthermore, we also performed the measurement of cardiac trabecular complexity by the Fractal analysis. Fractal dimensions were derived from confocal images of hearts from Tie1 mutants and control mice. Quantification was carried out using ImageJ with the Fraclac plugin (including 6 grid orientations).

Statistical analysis

Data analyses were performed using GraphPad Prism (version 8.0). For the 2-group comparison, the unpaired t test was performed with Welch correction if data passed the D’Agostino-Pearson normality test, or the unpaired nonparametric Mann-Whitney U test was applied. Data are expressed as mean ± SD. All statistical tests were 2-sided.

Results

Higher TIE1 and TIE2 in atrial than ventricular endocardial cells

To dissect the transcriptional dynamics of key endothelial regulators during the murine heart development, we analyzed the published single-cell transcriptomic dataset spanning the murine embryonic development (E10.5-E18.5) ³⁶. We found that Tie1, and Tek were highly expressed in endocardial cells (Pecam1⁺, Npr3⁺) while the coronary EC gene such as the coronary endothelial marker Fabp4 was lowly expressed in endocardial cells (Fig. 1A-B, Supplemental Fig. 1A-C). Interestingly, we found that the transcript levels of Tie1, and Tek were higher in the atrial endocardial cells than those of ventricles, especially at embryonic stages up to E14.5 when the structures of four-chambered hearts were established (Table 1). In contrast, it appears that the Angpt1 transcript level was higher in atrial than ventricular cardiomyocytes (Fig. 1A-B, Table 1).

Analysis of TIE1 expression in endocardial cells by the single cell RNA-seq data and its role in cardiac trabeculation during embryogenesis.
**A-B**. Analysis of the single-cell RNA-seq data of embryonic hearts from E10.5 to E18.5 by Feng et al. Nat Commun 2022 (DOI: 10.1038/s41467-022-35691-7). UMAP visualization of 10 distinct cell clusters in the murine hearts during embryogenesis (E10.5 to E18.5, A). *Tie1* and *Tek* are highly expressed in the endocardial cell population (*Npr3*⁺) as shown in feature plots. *Angpt1* are highly expressed in the atrial cardiomyocytes (B). Black arrows point to the coronary EC population (*Fabp4*⁺). Red arrows point to the endocardial cell population (*Npr3*⁺). Blue arrows point to the atrial cardiomyocytes. The quantitative results are shown in Table 1. C-J. Analysis of heart trabeculae of *Tie1^tm*^1a^/tm^1a and littermate controls by immunostaining for EMCN (endomucin; green), PECAM1 (platelet endothelial cell adhesion molecule 1; red) and DAPI (blue) at E12.5 (C-D) and E14.5 (G-H). Note that *Tie1^tm*^1a^/tm^1a mice displayed defective trabeculation in the atrium (white arrows) and sparse trabecular structures in the ventricles (white arrows) at E12.5, with more severe defects by E14.5, lacking atrial trabeculae (white arrows) and sparser ventricular trabeculae (white arrows). Asterisks in C and G point to the delayed formation of interventricular septum in *Tie1^tm*^1a^/tm^1a mice. Arrows point to the trabecula-associated endocardium. E-F. Quantification of the EMCN-positive area in the atrial and ventricular trabeculae at E12.5 (E, LA: *Tie1^tm*^1a^/tm^1a: 0.63±0.33, n=6; *Control*: 1.00±0.22, n=8; P=0.054. RA: *Tie1^tm*^1a^/tm^1a: 0.77±0.26, n=6; *Control*: 1.00±0.13, n=8; P=0.075. LV: *Tie1^tm*^1a^/tm^1a: 0.86±0.16, n=6; *Control*: 1.00±0.046, n=8; P=0.054. RV: *Tie1^tm*^1a^/tm^1a: 0.91±0.11, n=6; *Control*: 1.00±0.12, n=8; P=0.13.). Quantification of the ventricular wall thickness (F, LV: *Tie1^tm*^1a^/tm^1a: 24.87±3.70 µm, n=6; *Control*: 27.67±4.52 µm, n=8; P=0.41. RV: *Tie1^tm*^1a^/tm^1a: 26.01±8.42 µm, n=6; *Control*: 32.84±11.21 µm, n=8; P=0.28.). I-J. Quantification of the EMCN-positive area in the atrial and ventricular trabeculae at E14.5 (I, LA: *Tie1^tm*^1a^/tm^1a: 0.40±0.16, n=7; *Control*: 1.00±0.079, n=7; P=0.00060. RA: *Tie1^tm*^1a^/tm^1a: 0.36±0.17, n=7; *Control*: 1.00±0.092, n=7; P=0.00060. LV: *Tie1^tm*^1a^/tm^1a: 0.81±0.22, n=7; *Control*: 1.00±0.033, n=7; P=0.036. RV: *Tie1^tm*^1a^/tm^1a: 0.77±0.23, n=7; *Control*: 1.00±0.040, n=7; P=0.026.). Quantification of the PECAM1-positive area in ventricular walls (J, LV: *Tie1^tm*^1a^/tm^1a: 0.91±0.15, n=4; *Control*: 1.00±0.013, n=6; P=0.56. RV: *Tie1^tm*^1a^/tm^1a: 0.64±0.096, n=4; *Control*: 1.00±0.024, n=6; P=0.0095.). The quantification data of mutant mice was normalized against that of littermate control mice. In E-J, statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three independent technical replicates. LA, Left atria, RA, Right atria, LV, Left ventricle, RV, Right ventricle, Scale bar: Scale bar: 400 μm in C and G, 50 μm in D and H.

The transcript levels of *Tie1*, *Tek* in atrial and ventricular endocardial cells and *Angpt1* in atrial and ventricular cardiomyocytes during the embryonic stage from E10.5-E18.5 (based on the published scRNA-seq data, Feng et al., Nat Commun 2022; DOI: 10.1038/s41467-022-35691-7).
Values are represented as means ± SD and the number referred to cell counts.

To investigate mechanisms underlying the murine atrial trabeculation, we first analyzed the dynamic processes of cardiac chamber morphogenesis by the immunostaining analysis of endomucin (EMCN) for trabeculae and PECAM1 for coronary vessels (embryonic day 10.5-E18.5). The quantification methods for atrial and ventricular trabecular areas, wall thickness and vessel density were shown in Supplemental Fig. 2A. As shown in Supplemental Fig. 2B, the heart tube became looped to form the preliminary chamber structures at E10.5, including the initiation of interventricular septum. While there appeared to have some wavy edges or the budding of endocardium in atria (E10.5), trabecular structures were readily detectable in ventricles at this stage. Therefore, atrial wall morphogenesis occurs later than that of ventricles, which initiates the trabeculation process around E8.5 when endocardial cells migrate through cardiac jelly to form anchor points with compact myocardium ¹⁵. We found that pectinate muscles, defined as the atrial trabeculae, became detectable in the atria at E12.5 (Supplemental Fig. 2B). The formation of interventricular septum was completed by E14.5, establishing the four-chambered heart. By E16.5 or later stages (Supplemental Fig. 2C), the atrial and ventricular trabeculae gradually formed a complex network. As shown in Supplemental Fig. 2D-G (Supplemental Table 1), the trabecular areas in both the atria and ventricles gradually increased, being greater in the right atria (RA) than that in the left atria (LA, Supplemental Fig. 2D). There was no obvious difference in the trabecular areas between the left and right ventricles (Supplemental Fig. 2E). The ventricular wall thickness increased rapidly after E12.5, with the left ventricular wall being thicker than that of the right ventricular wall by E18.5 (Supplemental Fig. 2F), accompanied by the increase of coronary vessel density (Supplemental Fig. 2G).

Differential requirement of TIE1 in atrial and ventricular trabeculation

To study the role and mechanism underlying TIE1 function in cardiac chamber morphogenesis, we employed two genetically modified mouse models targeting Tie1, including Tie1^tm^1a^/tm^1a and Tie1^ΔICD/ΔICD. The effects of Tie1 gene deletion on atrial and ventricular trabeculae were analyzed at different developmental stages (including E12.5, E14.5 and E17.5). TIE1 deficiency led to an impaired trabeculation in the atria, characterized by a decrease of trabecular structures at E12.5, while the formation of ventricular trabeculae was not obviously affected at the same stage (Fig. 1C-D). Shown in Fig. 1E was the quantification of trabecular areas in atria and ventricles. The thickness of left and right ventricular walls at this stage was shown in Fig. 1F. Consistently, the atrial trabeculae were almost absent in the Tie1 mutant mice at E14.5, while trabeculae became sparse in both left and right ventricles (Fig. 1G-I). There was also a trend of decrease in the coronary vessel density upon TIE1 deficiency (Fig. 1J). By the later stages of embryonic development (E17.5), TIE1 deficiency resulted in more severe cardiac defects, including a significant reduction in heart size, absence of atrial trabeculae, and disorganized trabecular structures in ventricles, together with a decrease in the ventricular wall thickness as well as the coronary vessel density (Fig. 2A-F). The defective atrial trabecula formation was also validated in another Tie1 mutant line targeting its intracellular kinase domain (Tie1^ΔICD/ΔICD, Supplemental Fig. 3A-E). Consistent with the above results, TIE1 deficiency resulted in the lack of atrial trabeculation while trabeculae in ventricles became sparse in Tie1 mutants at E14.5. There was also a trend of decrease in ventricular wall thickness as well as the coronary vessel density in the Tie1^ΔICD/ΔICD mice. Consistently, the atrial trabecular defects could also be reflected by the fractal analysis of cardiac trabecular complexity in atria from Tie1 mutant and control mice at E14.5 and E17.5 using the Fraclac Plugin in ImageJ, while the relatively minor defects in the ventricles were not detected by the method (Tie1^tm^1a^/tm^1a, Supplemental Fig. 4; Supplemental Table 2).

Suppression of cardiac trabeculation-related gene expression by the loss of TIE1.
**A-B**. Analysis of heart trabeculae of *Tie1^tm*^1a^/tm^1a and littermate controls by immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) at E17.5. Note that the ventricular trabeculae of *Tie1^tm*^1a^/tm^1a mice become sparse compared with those of control mice (white arrows). Arrows point to the trabecula-associated endocardium. **C-F**. Quantification of the EMCN-positive area in the atrial and ventricular trabeculae (C, LA: *Tie1^tm*^1a^/tm^1a: 0.56±0.16, n=6; *Control*: 1.00±0.22, n=8; P=0.0043. RA: *Tie1^tm*^1a^/tm^1a: 0.57±0.33, n=6; *Control*: 1.00±0.24, n=8; P=0.026. D, LV: *Tie1^tm*^1a^/tm^1a: 0.86±0.18, n=6; *Control*: 1.00±0.082, n=8; P=0.054. RV: *Tie1^tm*^1a^/tm^1a: 0.61±0.18, n=6; *Control*: 1.00±0.10, n=8; P=0.00070). Quantification of the PECAM1-positive area in ventricular walls (E, LV: *Tie1^tm*^1a^/tm^1a: 0.66±0.14, n=6; *Control*: 1.00±0.11, n=8; P=0.00070. RV: *Tie1^tm*^1a^/tm^1a: 0.55±0.10, n=6; *Control*: 1.00±0.088, n=8; P=0.00070.). Quantification of the ventricular wall thickness (F, LV: *Tie1^tm*^1a^/tm^1a: 101.97 ± 22.44 µm, n=6; *Control*: 174.09 ± 22.84 µm, n=8; P=0.0013. RV: *Tie1^tm*^1a^/tm^1a: 87.14±26.20 µm, n=6; *Control*: 137.35±24.06 µm, n=8; P=0.0080.) at E17.5. Except the ventricular wall thickness, the quantification data of mutant mice was normalized against that of littermate control mice. **G-L**. RNA-seq analysis of whole hearts of *Tie1^tm*^1a^/tm^1a and littermate controls (E14.5, n=4 per group). G. PCA analysis showed that *Tie1^tm*^1a^/tm^1a and *control* groups were divided into two clusters. H. Volcano plot visualization of the differentially expressed genes (DEGs, P-value ≤ 0.05, |Log2Foldchange| > 0). I. Heatmap of significantly downregulated cardiac trabeculation genes (P-value ≤ 0.05). J. GSEA analysis showed the decrease of enrichment in trabeculation, venous development, and endothelial cell migration pathways. **K-L**. GO analysis revealed that gene related to cardiac chamber morphogenesis and trabeculae development, extracellular matrix, cell adhesion and endothelial migration/differentiation were significantly downregulated (K), while cardiac contraction and apoptosis related genes were significantly upregulated (L). In C-F, statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three independent technical replicates. LA: Left atria, RA: Right atria, LV: Left ventricle, RV: Right ventricle; Scale bar: 400 μm in A, 50 μm in B; c-c: cell-cell, p-m: plasma membrane in K.

Consistent with the previous observation ²⁸, the lack of TIE1 also disrupted the coronary vein formation at the later embryonic stages (E14.5 and E17.5), including the decrease of vein diameter and the increase of vein-associated angiogenic sprouting (Supplemental Fig. 5A-D).

Alteration of atrial versus ventricular gene regulatory network upon TIE1 deficiency

To explore the mechanism underlying TIE1 in heart development, we performed the RNA sequencing (RNA-seq) analysis of hearts to examine the alteration of cardiac transcriptome upon TIE1 deficiency. Lack of TIE1 led to a dramatic change in the overall transcriptional profile (Fig. 2G-L), including the suppression of gene expression related to the endocardial / endothelial development and cardiac chamber morphogenesis, such as Tek (normalized by Pecam1, Tie1^tm^1a^/tm^1a: 0.89 ± 0.031, n=4; WT: 1.00 ± 0.12, n=4; P = 0.057.). Specifically, pathways related to extracellular matrix, cell adhesion and endothelial cell migration were downregulated (Fig. 2K), while the cardiac contraction and cell apoptosis pathways were upregulated (Fig. 2L). This was further confirmed by the GSEA analysis using the custom gene sets (Fig. 2J). The gene sets included the following entries based on the GO database (HALLMARK_TRABECULATION, HALLMARK_EC_MIGRATION) and the fifty-three entries from the previous publication ²⁸. Consistent with previous reports ²⁸, TIE1 deficiency led to the downregulation of venous genes. In addition, key regulatory factors for cardiac trabecular development were downregulated, including Adgrg6 (adhesion G protein-coupled receptor G6), Nrg1 (neuregulin 1), and Notch1 (Notch receptor 1) and Dll1 (Delta Like Canonical Notch Ligand 1). Notably, we found that the expression of the TIE receptor family member Tek was also significantly reduced (Fig. 2I), suggesting that TIE1 acts, at least partially, via TIE2 in the regulation of cardiac trabecular morphogenesis.

To better visualize the trabecular structures, particularly the atrial internal muscular network, we performed the wholemount immunostaining of the hearts at the four-chambered stage. As shown in Fig. 3A, TIE1 deficiency led to the absence of the atrial trabecular network, with minor defects in the ventricles including a trend of sparse trabeculae (the 3D views shown in Supplemental Fig. 6). To further investigate the differential requirement of TIE1 in atrial and ventricular trabeculation, we collected atria and ventricle tissues separately from Tie1 mutant and control mice (E14.5) for the RNA-seq analysis. We found that the number of differentially expressed genes was greater in atria than in the ventricles after the loss of TIE1 (Fig. 3B-G). Similar to the results from the whole heart RNA-seq analysis, the transcript levels of genes related to atrial development was decreased. Although a similar trend of decrease in genes involved in the cardiac morphogenesis were also observed in the ventricles, the enrichment levels and the number of genes involved were much lower in ventricles than in atria. Specifically, 17 trabecula-related genes were downregulated in the atria, while there was only 8 in the ventricles. Notably, Tek and Dll1 were downregulated in both the atria and ventricles while Notch1 only in atria of Tie1 mutant mice (Fig. 3F-G; Supplemental Fig. 3F-G; Supplemental Table 3 and Supplemental Table 4). Interestingly, the expression of endothelial Tie1, Tek and Notch1 was higher in atria than in ventricles (Fig. 3H, Supplemental Fig. 1D, Supplemental Table 5). The top GO terms enriched for the up- or down-regulated genes were related to the biological processes of heart development, including those in trabecular formation, extracellular matrix organization and binding, intercellular adhesion and endothelial cell migration in both atria and ventricles (Fig. 3I-J). In addition, the analysis of scRNA-seq data for the ligand-receptor interactions showed that the atrial_EC (endocardial cells) and atrial_CM (cardiomyocytes) exhibited stronger ANGPT1-TIE2 signaling compared with those in ventricles (Supplementary Fig. 1E). This may account for the differential requirement of TIE1 and also TIE2 in the process of atrial versus ventricular trabeculation.

Differential effect of *Tie1* deficiency on atrial versus ventricular trabeculation related gene network.
A. Visualization of heart trabeculae structure of *Tie1^tm*^1a^/tm^1a and littermate controls by whole-mount immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) at E14.5. Note that *Tie1* deletion resulted in the loss of atrial trabecular network with minor defects with ventricular trabecular structures (white arrows). Arrows point to the trabecula-associated endocardium. **B-J**. RNA-seq analysis of atrium and ventricles of *Tie1^tm*^1a^/tm^1a and littermate controls (E14.5, n=3 per group). B. PCA analysis results showed clear separation between the atria and ventricles of *Tie1^tm*^1a^/tm^1a and control mice. C. Differential expression analysis (P-value ≤ 0.05, |log2(Foldchange)| ≥ 0.15) identified over 693 down-regulated and 948 up-regulated atrial genes, while 144 down-regulated and 190 up-regulated genes in ventricles. **D-E**. Volcano plots of atrial (D) and ventricular (E) differential expressed genes (DEGs). **F-G**. Heatmap showed significantly downregulated (P-value ≤ 0.05) heart trabeculation genes in atrium (F) and ventricles (G), including *Tek*, *Dll1* and *Notch1*. H. Quantitation of *Tie1* and *Tek* transcript levels (normalized by *Pecam1*) in atrium and ventricles of *Tie1^tm*^1a^/tm^1a and littermate controls (*Tie1*: Ventricles: 0.72 ± 0.021, n = 5; Atria: 1.00 ± 0.026; P = 0.0079. *Tek*: Ventricle: 0.55 ± 0.024; Atria: 1.00 ± 0.048; P = 0.0079). **I-J**. GO enrichment analysis of the DEGs in atrium and ventricles including the heart trabecula morphogenesis, extracellular matrix, cell adhesion and endothelial migration related process down-regulated while genes related to the cardiac contraction, apoptosis process up-regulated. There was a more obvious enrichment detected in the atria. In H, statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three technical replicates. LA: left atria, RA: right atria, LV: left ventricle, RV: right ventricle; Scale bar: 50 μm in A; c-c: cell-cell, p-m: plasma membrane in I-J.

Stronger ANGPT1-TIE2 interactions among atrial endocardium and myocardium

Using the published scRNA-seq data ³⁶, we also analyzed the ligand-receptor interactions among endocardial cells and cardiomyocytes in atria and ventricles respectively during the embryonic stage (E10.5-E18.5). We found that the ANGPT1-TIE2 signaling between atrial cardiomyocytes and endocardial cells was stronger than that in ventricles (Supplemental Fig. 1E). Consistently, the crucial role of endothelial TIE2 in the atrial trabeculation was validated in this study by employing a conditional knockout mouse model targeting Tek (also known as Tie2 gene; Tek^Flox/-; Cdh5-Cre^ERT²; named Tek^iECKO). As the complete knockout of Tek gene leads to embryonic lethality before E10.5, we first performed the induced endothelial Tek deletion starting from E12.5 (Fig. 4A-B). We found that TIE2 insufficiency led to an obvious decrease of trabeculae in the atrium 48 h later (E14.5), while there were no obvious defects observed in the ventricular trabeculae at this stage (Fig. 4C-E). The coronary vessel density in the ventricular wall as well as the wall thickness, including left and right ventricles, were not obviously altered when examined 48 h later (Fig. 4F). In separate experiments with the induced gene deletion from E12.5 to E14.5, embryos were collected for the analysis at 72 hours (E15.5) or later (E17.5). The formation of atrial trabeculae was completely missing as shown in Fig. 4A, G-H. The defective trabecula formation was also detected in the ventricles, especially at later embryonic stages (E17.5, Fig. 4A, G-H). TIE2 insufficiency also led to the defective formation of the coronary veins as previously shown in skin and mesentery ³⁰ (Supplemental Fig. 5E, G).

Impaired atrial trabecular formation after TIE2 insufficiency.
A. Tamoxifen intraperitoneal (i.p.) administration and analysis scheme. **B-D**. Analysis of heart trabeculae of *Tek^iECKO* and littermate controls by immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) 48 hours later (tamoxifen treatment starting at E12.5). **E-F**. Quantification of the EMCN-positive area in the atrial and ventricular trabeculae (E, LA: *Tek^iECKO*: 0.60±0.20, n=9; *Control*: 1.00±0.097, n=8; P=0.00010. RA: *Tek^iECKO*: 0.64±0.23, n=9; *Control*: 1.00±0.22, n=8; P=0.0047. LV: *Tek^iECKO*: 0.85±0.15, n=9; *Control*: 1.00±0.057, n=8; P=0.020. RV: *Tek^iECKO*: 0.90±0.16, n=9; *Control*: 1.00±0.089, n=8; P=0.13.), quantification of the PECAM1-positive area in ventricular walls (F, LV: *Tek^iECKO*: 0.92±0.25, n=7; *Control*: 1.00±0.048, n=6; P=0.81. RV: *Tek^iECKO*: 1.22±0.53, n=7; *Control*: 1.00±0.13, n=6; P=0.28.) and quantification of the ventricular wall thickness (F, LV: *Tek^iECKO*: 62.35±11.70 µm, n=9; *Control*: 62.59±10.26 µm, n=8; P=0.96. RV: *Tek^iECKO*: 63.87± 9.19 µm, n=9; *Control*: 58.87±11.02 µm, n=8; P=0.32.) at E14.5. Except the ventricular wall thickness, the quantification data of mutant mice was normalized against that of littermate control mice. **G-H**. Analysis of heart trabeculae of *Tek^iECKO* and littermate controls by immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) 72 hours after the induced gene deletion or later (tamoxifen treatment starting at E12.5). Arrows point to the trabecula-associated endocardium. In E-F, statistical analysis was performed using the unpaired t test. Values are represented as means ± SD of at least three independent technical replicates. LA: left atria, RA: right atria, LV: left ventricle, RV: right ventricle; Scale bar: 500 μm (embryo) and 400 μm (heart) in B, 400 μm in C, 50 μm in D, G, H.

Synergy of TIE1 and TIE2 in atrial trabeculation

To investigate the synergistic effects of TIE1 and TIE2 in the regulation of cardiac morphogenesis, we generated a doubly knockout model targeting Tie1 and Tek (Tie1^ΔICD/ΔICD; Tek^+/-). There were no obvious defects observed with the atrial or ventricular structures upon Tie1 deficiency alone at E10.5 (Fig. 5A-B). Intriguingly, loss of Tie1 combined with the heterozygous deletion of Tek suppressed both atria and ventricle development (E10.5), including reduced heart size, abnormal atrial chamber formation, and fewer trabecular structures in the ventricles (Fig. 5C-D), leading to embryonic lethality before E12.5.

Developmental requirement of TIE1 and TIE2 in atrial and ventricular chamber morphogenesis.
**A-D**. Analysis of heart trabeculae of *Tie1*^ΔICD/ΔICD, *Tie1*^ΔICD/ΔICD;Tek^+/- and littermate controls by immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) at E10.5. Note that there were no obvious defects observed with the atrial or ventricular structures upon *Tie1* deficiency alone at E10.5, while loss of *Tie1* combined with the heterozygous deletion of *Tek* suppressed both atria and ventricle development. White arrows point to the trabecular structures. Quantification of the EMCN-positive area in the ventricular trabeculae at E10.5, and the quantification data of mutant mice was normalized against that of littermate control mice (B, LV: *Tie1*^ΔICD/ΔICD: 0.86±0.11, n=4; *Control*: 1.00±0.14, n=5; P=0.29. RV: *Tie1*^ΔICD/ΔICD: 0.79±0.19, n=4; *Control*: 1.00±0.15, n=5; P=0.19. D, *Tie1*^ΔICD/ΔICD;Tek^+/-: 0.64±0.22, n=6; *Control*: 1.00±0.16, n=7; P=0.0047). E. For the induced gene deletion, tamoxifen intraperitoneal (i.p.) administration (starting from E10.5) and analysis scheme (at E14.5-E16.5). F. The deletion efficiency of TIE1 was examined by the immunostaining of skins from *Tie1ICD^iECKO;Tek^+/-* and littermate control mice at E16.5 for PECAM1 (green) and TIE1 (red). **G-H**. Analysis of heart trabeculae of *Tie1ICD^iECKO;Tek^+/-* mice and littermate controls by immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) at E14.5. Note that the atrial trabeculae structure was missing, while the ventricular trabeculae became sparse (white arrows). Arrows point to the trabecula-associated endocardium. In B and D, statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three independent technical replicates. LA: left atria, RA: right atria, LV: left ventricle, RV: right ventricle. Scale bar: 200 μm in A(left) and C(left), 30 μm in A(right) and C (right), 400 μm in G, 200 μm in F and 50 μm in H.

To further explore the synergy of TIE1/TIE2 in the heart development at later stages of embryonic development, we generated an inducible mouse model with the endothelial deletion of Tie1 combined with Tek heterozygous knockout (Tie1ICD^Flox/- ;Cdh5-Cre^ERT²;Tek^+/-, named Tie1ICD^iECKO;Tek^+/-). The induced gene deletion was performed by the tamoxifen treatment starting from E10.5, and mice were sacrificed for analysis after 96 hours (E14.5). The gene deletion efficiency was confirmed by the immunostaining for PECAM1 plus TIE1. TIE1 insufficiency combined with the heterozygous loss of TIE2 led to the suppression of atrial trabeculation, while the trabeculae at both right and left ventricle became sparse and irregular (Fig. 5E-H). Furthermore, when the induced gene deletion was performed at E12.5, the defective atrial trabeculation were already detectable in the mutant mice 48 hours later (E14.5 Tie1ICD^iECKO;Tek^+/-; Fig. 6 A-D). Consistent with the observation with the endothelial Tek deletion for 48 hours, there was no obvious defects with the ventricular trabeculation, coronary vessel density and ventricular wall thickness (Tie1ICD^iECKO;Tek^+/-; Fig. 6D-E, Table 2). The disruption of cardiac trabeculation became more obvious when analyzed 72-96 hours later, particularly in atria but less severe with ventricles (Tie1ICD^iECKO; Tek^+/-, E15.5-E16.5; Fig. 6 F-G, Table 3). There was also a trend of decrease in coronary vessel density and ventricular wall thickness as shown at later stages (E16.5, Fig. 6 F, H, Table 3). In addition, the TIE1/TIE2 insufficiency also resulted in the disruption of cardiac vein development as shown in (Supplemental Fig. 5F, H), as previously reported in skin and retina ²⁸.

Synergistic role of TIE1 and TIE2 in the atrial trabeculation.
A. Tamoxifen intraperitoneal (i.p.) administration and analysis scheme. **B-H**. The induced gene deletion by tamoxifen started at E12.5. Analysis of heart trabeculae of *Tie1ICD^iECKO;Tek^+/-*and littermate controls by immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) at E14.5 (**B-C**) and E16.5 (F). Note that the atrial trabeculae were absent and the ventricular trabeculae were slightly sparse in the doubly mutant mice at E14.5 (**B-C**), and the phenotype became more severe at E16.5 (F). Quantification of the EMCN-positive area in the atrial and ventricular trabeculae (D, E14.5; G, E16.5), the PECAM1-positive area in ventricular walls and the ventricular wall thickness (E, E14.5; H, E16.5). Arrows point to the trabecula-associated endocardium. The quantification data of mutant mice was normalized against that of littermate control mice as shown in Table 2 and Table 3. Arrows in C and F point to the trabecula-associated endocardium. In D, E, G and H, statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three independent technical replicates. LA: left atria, RA: right atria, LV: left ventricle, RV: right ventricle. Scale bar: 400 μm in B, 50 μm in C and F.

Quantitation of the trabecular area (EMCN⁺), blood vessel density in ventricular wall (PECAM1⁺) and ventricular wall thickness in *Tie1ICD^iECKO;Tek^+/-* and littermate controls (at E14.5, 48 hours after the induced gene deletion).
Statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three independent technical replicates.

Quantitation of the trabecular area (EMCN⁺), coronary vessel density in ventricular wall (PECAM1⁺) and ventricular wall thickness in *Tie1ICD^iECKO;Tek^+/-* and littermate controls 96 hours after the induced gene deletion (E16.5; tamoxifen treatment from E12.5-E13.5).
Statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test. Values are represented as means ± SD of at least three independent technical replicates.

Requirement of TIE receptors for the postnatal atrial trabecular remodeling

The role of TIE1 and its synergy with TIE2 in cardiac trabecular development was further confirmed at the neonatal stage (Tie1ICD^iECKO or Tie1ICD^iECKO; Tek^+/-) with the induced gene deletion from postnatal day 1 (P1-4) and the analysis performed at P21. The Tie1 deletion was confirmed by the vascular defects in retinas of Tie1ICD^iECKO or Tie1ICD^iECKO; Tek^+/-mice as previously reported (Fig. 7B, E) ²⁸. TIE1 insufficiency alone had no obvious impact on the heart weight and also the atrial or ventricular trabeculation analyzed within three weeks (Fig. 7 A-D). Consistent with the studies at embryonic stages, the induced postnatal gene deletion (Tie1ICD^iECKO;Tek^+/-) showed the retarded heart development by the quantification of heart-body weight (Fig. 7 F, G), and showed abnormal atrial trabeculation while the trabeculae in the ventricles had no obvious defects (Fig. 7F, H). There was a significant decrease of trabecular structures (white arrows in Fig. 7H) in atria of Tie1ICD^iECKO; Tek^+/- mice compare with that of the control mice (LA, Tie1ICD^iECKO; Tek^+/-: 0.69±0.16, n=11; Control: 1.00±0.28, n=16; P=0.0022. RA, Tie1ICD^iECKO; Tek^+/-: 0.77±0.15, n=11; Control: 1.00±0.11, n=16; P=0.0002). The effects of TIE1 and TIE2 deficiency or insufficiency on heart trabecular morphogenesis were summarized in Fig. 7I. Based on the previous findings about the TIE1/TIE2 in venous EC specification ^28,30, we propose that TIE1 acts in synergy with TIE2 in the specification of atrial endocardial cells (Fig. 7I). TIE1 deficiency or TIE1/TIE2 insufficiency results in a more severe defects with the atrial trabeculation, suggesting the differential requirement of TIE1 and TIE2 in the endocardium-mediated organization of the atrial internal muscular network.

Abnormal atrial trabecular remodeling after the postnatal deletion of *Tie1* combined with one null allele of *Tek*.
A. Tamoxifen intragastric (i.g.) administration and the analysis scheme. **B-H**. Analysis of blood vessels in the retinas of *Tie1ICD^iECKO* (B), *Tie1ICD^iECKO;Tek^+/−*(E) and *control* mice by whole-mount immunostaining for PECAM1 (green), DLL4 (red) at P21. Note that hemangioma-like vascular tufts observed in *Tie1ICD^iECKO* and *Tie1ICD^iECKO;Tek^+/−*mice. The retinal vascular defects were used to confirm the efficiency of induced gene deletion. Analysis of heart trabeculae in *Tie1ICD^iECKO* (C), *Tie1ICD^iECKO;Tek^+/−*(F) and control mice by the immunostaining for EMCN (green), PECAM1 (red) and DAPI (blue) at P21. TIE1 insufficiency alone had no obvious effect on heart size (D, *Tie1ICD^iECKO*: 0.54±0.021, n=7; *Control*: 0.54±0.014, n=5; P=0.91.) and atrial trabecular structures (C). The *Tie1ICD^iECKO;Tek^+/-* mice displayed smaller hearts (G, *Tie1ICD^iECKO;Tek^+/-*: 0.49±0.086, n=8; *Control*: 0.59±0.10, n=11; P=0.035.) and abnormal atrial trabecular morphology (**F,H**). White arrows in H point to the trabecula-associated endocardium in the atria. I. Schematic illustration of TIE1, TIE2 and their synergy in cardiac chamber morphogenesis. TIE1 deficiency, TIE2 or TIE1/TIE2 insufficiency results in a more severe defect with the atrial trabeculation than that of ventricles. Based on the previous findings about the TIE1/TIE2 in venous EC specification ^28,30 and findings from this study, we propose that TIE1 acts in synergy with TIE2 in the specification and remodeling of atrial endocardium, and that there is a differential requirement of TIE receptors in atrial versus ventricular wall morphogenesis. In D and G, statistical analysis was performed using the unpaired nonparametric Mann-Whitney U test (D) and the unpaired t test (G). Values are represented as means ± SD of at least three independent technical replicates. Scale bar: 50 μm in B and E, 1 mm in C and F, 200 μm in H.

Discussion

Developmental defects or pathological remodeling of atria may result in the atrial dysfunction leading to the heart failure ⁴³. In contrast to the better understanding of ventricular chamber morphogenesis, the dynamic process and regulatory networks of the atrial internal muscular network formation (known as atrial trabeculation) are inadequately elucidated. TIE1 variants or loss of function mutations were reported in a subset of lymphedema patients ^34,35, and it is unknown whether the patients have also cardiac defects in addition to the lymphatic abnormality. We show in this study that the atrial trabeculation occurred later than in ventricles. Based on the analysis of the published scRNA-seq data by Feng et al. ³⁶ and the bulk RNA-seq analysis of atria and ventricles separately in this study, we showed that Tie1 and its homologue Tek were highly expressed in atrial endocardial cells than those of ventricles. By employing the genetically modified mouse models targeting Tie1, Tek and both, we found that TIE1 was differentially required in the process of atrial versus ventricular trabeculation and acted in synergy with TIE2 in the process of cardiac trabeculation, especially in atria.

Endocardial versus vascular effects of TIE1

Cardiac chamber morphogenesis involves the coordinated endocardial-myocardial interaction governed by a plethora of cardiovascular regulators ². It is worth noting that most of the reported endothelial regulators have important roles in both heart and vascular development. ANGPT1 has been shown to participate in the regulation of atria and cardiac vein development ^19,44. We have demonstrated that TIE1 and TIE2 are required for the vein specification via regulating the protein stability of COUPTFII ³⁰. Disruption of COUPTFII led to the defective atrial and venous formation ^31,33. Defective cardiac venous formation was also observed in this study in Tie1 or Tek mutant mice at later embryonic stages. It is therefore important to dissect whether cardiac defects in the mutants are a direct effect or secondary to the coronary vascular defects. By performing a serial analysis of the developing hearts during embryogenesis (including E10.5, E12,5, E14.5, E16.5, E17.5), we found in this study that TIE1 deficiency resulted in an impaired atrial trabeculation readily detectable at embryonic day 12.5 (E12.5), while there were only minor defects observed with the ventricular trabeculae at the same stage. In addition, Tek deficiency leads to the embryonic lethality before E10.5 ^30,45. Interestingly, the defective atrial trabeculae formation was detected within 48 hours after the induced deletion of endothelial Tek (starting from E12.5), while there were no obvious defects with coronary vessels at the same stage. This was further confirmed in the Tie1/Tek doubly mutant mice with defective atrial trabeculation detected within 48 hours after the induced deletion of Tie1 combined with one null allele of Tek. Lack of TIE1 did not produce obvious developmental defects in mutant embryos by E10.5 as previously reported ^26,28. Therefore, the differential requirement of TIE1 may be related to the specification of atrial versus ventricular endocardium, which involves a synergistic role of TIE2, as previously reported in vein development ^28,30. Based on the above findings, it is likely that TIE1 and TIE2 are directly required for the endocardial cell-mediated organization of atrial trabecular structures. In addition, defects with the ventricular trabeculation and a trend of thinner ventricle walls were also observed at later stages after the Tie1 deficiency or induced Tie1/Tek insufficiency. The cardiac trabeculation contributes to the thickening and maturation of the ventricular wall by being incorporated into the compact myocardium ^4,8. It has been reported that endocardial cells contribute to the formation of coronary vascular network formation ¹⁰. The defective ventricular wall morphogenesis may also be related to the decreased coronary vessel growth in Tie1 or Tie1/Tek mutant mice as observed in this study.

Synergy of TIE1 and TIE2 in cardiac trabeculation

Loss of TIE1 alone did not produce an obvious defect with atrial chamber formation at initial stages of heart development (e.g. E10.5). However, the loss of Tie1 combined with the heterozygous deletion of Tek disrupted the atrial and ventricle development at the same stage (E10.5). The synergy of TIE1 and TIE2 in heart development was further verified at later embryogenesis as well as the postnatal studies. The defects with the atrial trabeculation were already detectable 48 hours later in the induced knockout of Tie1 plus the heterzygous Tek deletion (starting at E12.5, Tie1ICD^iECKO; Tek^+/-). This was further confirmed at the neonatal with the induced gene deletion from postnatal day 1 and the analysis 3-weeks later stage (Tie1ICD^iECKO; Tek^+/-), while TIE1 insufficiency alone (Tie1ICD^iECKO) had no obvious impact on the atrial or ventricular trabecula development. TIE1 and TIE2 could form heterodimers, participate in the formation of endothelial cell-cell adherens junction as well as the EC migration and survival ^21,22,46,47. Furthermore, TIE1 and TIE2 act in a synergistic manner to specify the vein formation via regulating the protein stability of the venous factor COUPTFII ^28,30. As COUPTFII is crucial for the atrial development ^31,33, this could provide some insights into the synergy of TIE1/TIE2 in the regulation of cardiac trabeculation especially in atria, likely involving COUPTFII-mediated regulation of the atrial endocardial cell specification.

Differential requirement of TIE receptors in atria and ventricles

So far, little is known about mechanisms underlying the differential regulation of atrial versus ventricular endocardial development. We found in this study that Tie1 and Tek were expressed at higher levels in atrial endocardial cells than those of ventricles by the RNA-seq analysis at the single cell level or the separate analysis of atria and ventricles. By the analysis of scRNA-seq data from embryonic hearts for the ligand-receptor interactions, we showed that the atrial endocardial cells and cardiomyocytes exhibited stronger ANGPT1-TIE2 signaling compared with those in ventricles. This suggests that the differential requirement of TIE1 and TIE2 in cardiac chamber morphogenesis may, at least partly, be related to the differential expression levels of TIE receptors in atrial versus ventricular endocardial cells. In addition, we also found that Dll1 was mainly expressed by endocardial cells from the analysis of single cell RNA-seq data ³⁶, and that Dll1 was among the most downregulated genes in the hearts from Tie1 deficient mice. A similar trend of decreased expression with Notch1 was also detected in atria but not in ventricles of Tie1 deficient mice. As NOTCH1 signaling promotes the extracellular matrix degradation during the formation of endocardial projections that are critical for individualization of trabecular units ^13,15, this may also contribute to the differential requirement of TIE1 as well as TIE2 in atrial versus ventricular trabeculation.

In summary, findings from this study show that endothelial TIE1 is crucial for the atrial trabecular formation and play synergistic roles with TIE2 in the process of cardiac chamber morphogenesis. Tie1 deficiency resulted in the significant decrease of Tek, Dll1 and Notch1. This suggests that the differential requirement of TIE receptors in the process of atrial and ventricular trabeculation may involve the DLL1-NOTCH1 pathway. Mechanisms underlying the differential regulation of atrial versus ventricular formation during the heart development warrants further investigation. Knowledge in this direction will facilitate the development of novel interventions for the atria-related cardiomyopathy.

Data availability

RNA sequence data will be deposited in GEO and information provided later.

Acknowledgements

The authors thank the staff in Animal facility of Soochow University for technical assistance.

Additional information

Funding

This work was supported by grants from the National Key R&D Program of China (2021YFA0805000), the National Natural Science Foundation of China (82470518, 82401544), China Postdoctoral Science Foundation (CPSF, 2024M762300), the Natural Science Foundation of Jiangsu Province (BK20240786), the Project of State Key Laboratory of Radiation Medicine and Protection (No. GZN120 20 02), and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

Funding

MOST | National Key Research and Development Program of China (NKPs) (2021YFA0805000)

Yulong He

MOST | National Natural Science Foundation of China (NSFC) (82470518)

Yulong He

MOST | National Natural Science Foundation of China (NSFC) (82401544)

Xudong Cao

China Postdoctoral Science Foundation (2024M762300)

Taotao Li

JST | Natural Science Foundation of Jiangsu Province (Jiangsu Natural Science Foundation) (BK20240786)

Xudong Cao

Project of State Key Laboratory of Radiation Medicine and Protection (GZN120 20 02)

Yulong He

Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD)

Yulong He

Additional files

Supplemental Figures and Tables

Supplemental Figure 6 for Fig. 3A Control

Supplemental Figure 6 for Fig. 3A Tie1tm1a

References

1.
1. Meilhac SM
2. Buckingham ME
2018The deployment of cell lineages that form the mammalian heartNat Rev Cardiol 15:705–724https://doi.org/10.1038/s41569-018-0086-9 PubMed Google Scholar
2.
1. Tian Y
2. Morrisey EE
2012Importance of myocyte-nonmyocyte interactions in cardiac development and diseaseCirc Res 110:1023–1034https://doi.org/10.1161/CIRCRESAHA.111.243899 PubMed Google Scholar
3.
1. Moorman AF
2. Christoffels VM
2003Cardiac chamber formation: development, genes, and evolutionPhysiol Rev 83:1223–1267https://doi.org/10.1152/physrev.00006.2003 PubMed Google Scholar
4.
1. Tran YTH
2. Saha D
3. Del Monte-Nieto G
2025Cardiac trabeculation in vertebrates: Convergent evolution or evolutionary adaptations associated with heart complexity?Seminars in cell & developmental biology 172https://doi.org/10.1016/j.semcdb.2025.103622 PubMed Google Scholar
5.
1. Haack T
2. Abdelilah-Seyfried S
2016The force within: endocardial development, mechanotransduction and signalling during cardiac morphogenesisDevelopment 143:373–386https://doi.org/10.1242/dev.131425 PubMed Google Scholar
6.
1. Rhee S
2. Paik DT
3. Yang JY
4. Nagelberg D
5. Williams I
6. Tian L
7. Roth R
8. Chandy M
9. Ban J
10. Belbachir N
11. et al.
2021Endocardial/endothelial angiocrines regulate cardiomyocyte development and maturation and induce features of ventricular non-compactionEur Heart J 42:4264–4276https://doi.org/10.1093/eurheartj/ehab298 PubMed Google Scholar
7.
1. Captur G
2. Syrris P
3. Obianyo C
4. Limongelli G
5. Moon JC
2015Formation and Malformation of Cardiac Trabeculae: Biological Basis, Clinical Significance, and Special Yield of Magnetic Resonance Imaging in AssessmentThe Canadian journal of cardiology 31:1325–1337https://doi.org/10.1016/j.cjca.2015.07.003 PubMed Google Scholar
8.
1. Gunawan F
2. Priya R
3. Stainier DYR
2021Sculpting the heart: Cellular mechanisms shaping valves and trabeculaeCurrent opinion in cell biology 73:26–34https://doi.org/10.1016/j.ceb.2021.04.009 PubMed Google Scholar
9.
1. Samsa LA
2. Yang B
3. Liu J
2013Embryonic cardiac chamber maturation: Trabeculation, conduction, and cardiomyocyte proliferationAm J Med Genet C Semin Med Genet :157–168https://doi.org/10.1002/ajmg.c.31366 PubMed Google Scholar
10.
1. Meilhac SM
2. Buckingham ME
2018The deployment of cell lineages that form the mammalian heartNature reviews Cardiology 15:705–724https://doi.org/10.1038/s41569-018-0086-9 PubMed Google Scholar
11.
1. Wu B
2. Zhang Z
3. Lui W
4. Chen X
5. Wang Y
6. Chamberlain AA
7. Moreno-Rodriguez RA
8. Markwald RR
9. O’Rourke BP
10. Sharp DJ
11. et al.
2012Endocardial cells form the coronary arteries by angiogenesis through myocardial-endocardial VEGF signalingCell 151:1083–1096https://doi.org/10.1016/j.cell.2012.10.023 PubMed Google Scholar
12.
1. Lai D
2. Liu X
3. Forrai A
4. Wolstein O
5. Michalicek J
6. Ahmed I
7. Garratt AN
8. Birchmeier C
9. Zhou M
10. Hartley L
11. et al.
2010Neuregulin 1 sustains the gene regulatory network in both trabecular and nontrabecular myocardiumCirc Res 107:715–727https://doi.org/10.1161/CIRCRESAHA.110.218693 PubMed Google Scholar
13.
1. Grego-Bessa J
2. Luna-Zurita L
3. del Monte G
4. Bolos V
5. Melgar P
6. Arandilla A
7. Garratt AN
8. Zang H
9. Mukouyama YS
10. Chen H
11. et al.
2007Notch signaling is essential for ventricular chamber developmentDev Cell 12:415–429https://doi.org/10.1016/j.devcel.2006.12.011 PubMed Google Scholar
14.
1. Watanabe Y
2. Kokubo H
3. Miyagawa-Tomita S
4. Endo M
5. Igarashi K
6. Aisaki K
7. Kanno J
8. Saga Y
2006Activation of Notch1 signaling in cardiogenic mesoderm induces abnormal heart morphogenesis in mouseDevelopment 133:1625–1634https://doi.org/10.1242/dev.02344 PubMed Google Scholar
15.
1. Del Monte-Nieto G
2. Ramialison M
3. Adam AAS
4. Wu B
5. Aharonov A
6. D’Uva G
7. Bourke LM
8. Pitulescu ME
9. Chen H
10. de la Pompa JL
11. et al.
2018Control of cardiac jelly dynamics by NOTCH1 and NRG1 defines the building plan for trabeculationNature 557:439–445https://doi.org/10.1038/s41586-018-0110-6 PubMed Google Scholar
16.
1. Miao L
2. Lu Y
3. Nusrat A
4. Fan G
5. Zhang S
6. Zhao L
7. Wu CL
8. Guo H
9. Huyen TLN
10. Zheng Y
11. et al.
2025Tunneling nanotube-like structures regulate distant cellular interactions during heart formationScience 387:eadd3417https://doi.org/10.1126/science.add3417 PubMed Google Scholar
17.
1. Lockhart M
2. Wirrig E
3. Phelps A
4. Wessels A
2011Extracellular matrix and heart developmentBirth Defects Res A Clin Mol Teratol 91:535–550https://doi.org/10.1002/bdra.20810 PubMed Google Scholar
18.
1. Kern CB
2. Wessels A
3. McGarity J
4. Dixon LJ
5. Alston E
6. Argraves WS
7. Geeting D
8. Nelson CM
9. Menick DR
10. Apte SS
2010Reduced versican cleavage due to Adamts9 haploinsufficiency is associated with cardiac and aortic anomaliesMatrix biology : journal of the International Society for Matrix Biology 29:304–316https://doi.org/10.1016/j.matbio.2010.01.005 PubMed Google Scholar
19.
1. Kim KH
2. Nakaoka Y
3. Augustin HG
4. Koh GY
2018Myocardial Angiopoietin-1 Controls Atrial Chamber Morphogenesis by Spatiotemporal Degradation of Cardiac JellyCell Rep 23:2455–2466https://doi.org/10.1016/j.celrep.2018.04.080 PubMed Google Scholar
20.
1. Qu X
2. Harmelink C
3. Baldwin HS
2019Tie2 regulates endocardial sprouting and myocardial trabeculationJCI Insight 5https://doi.org/10.1172/jci.insight.96002 PubMed Google Scholar
21.
1. Augustin HG
2. Koh GY
3. Thurston G
4. Alitalo K
2009Control of vascular morphogenesis and homeostasis through the angiopoietin-Tie systemNature reviews Molecular cell biology 10:165–177https://doi.org/10.1038/nrm2639 PubMed Google Scholar
22.
1. Saharinen P
2. Eklund L
3. Alitalo K
2017Therapeutic targeting of the angiopoietin-TIE pathwayNature reviews Drug discovery 16:635–661https://doi.org/10.1038/nrd.2016.278 PubMed Google Scholar
23.
1. Qu X
2. Harmelink C
3. Baldwin HS
2019Tie2 regulates endocardial sprouting and myocardial trabeculationJCI Insight 5https://doi.org/10.1172/jci.insight.96002 PubMed Google Scholar
24.
1. Leppanen VM
2. Saharinen P
3. Alitalo K
2017Structural basis of Tie2 activation and Tie2/Tie1 heterodimerizationProc Natl Acad Sci U S A 114:4376–4381https://doi.org/10.1073/pnas.1616166114 PubMed Google Scholar
25.
1. Puri MC
2. Rossant J
3. Alitalo K
4. Bernstein A
5. Partanen J
1995The receptor tyrosine kinase TIE is required for integrity and survival of vascular endothelial cellsThe EMBO journal 14:5884–5891https://doi.org/10.1002/j.1460-2075.1995.tb00276.x PubMed Google Scholar
26.
1. Shen B
2. Shang Z
3. Wang B
4. Zhang L
5. Zhou F
6. Li T
7. Chu M
8. Jiang H
9. Wang Y
10. Qiao T
11. et al.
2014Genetic dissection of tie pathway in mouse lymphatic maturation and valve developmentArterioscler Thromb Vasc Biol 34:1221–1230https://doi.org/10.1161/ATVBAHA.113.302923 PubMed Google Scholar
27.
1. Qu X
2. Tompkins K
3. Batts LE
4. Puri M
5. Baldwin S
2010Abnormal embryonic lymphatic vessel development in Tie1 hypomorphic miceDevelopment 137:1285–1295https://doi.org/10.1242/dev.043380 PubMed Google Scholar
28.
1. Cao X
2. Li T
3. Xu B
4. Ding K
5. Li W
6. Shen B
7. Chu M
8. Zhu D
9. Rui L
10. Shang Z
11. et al.
2023Endothelial TIE1 Restricts Angiogenic Sprouting to Coordinate Vein Assembly in Synergy With Its Homologue TIE2Arterioscler Thromb Vasc Biol 43:e323–e338https://doi.org/10.1161/ATVBAHA.122.318860 PubMed Google Scholar
29.
1. Qu X
2. Violette K
3. Sewell-Loftin MK
4. Soslow J
5. Saint-Jean L
6. Hinton RB
7. Merryman WD
8. Baldwin HS
2019Loss of flow responsive Tie1 results in ImpairedAortic valve remodelingDev Biol https://doi.org/10.1016/j.ydbio.2019.07.011 PubMed Google Scholar
30.
1. Chu M
2. Li T
3. Shen B
4. Cao X
5. Zhong H
6. Zhang L
7. Zhou F
8. Ma W
9. Jiang H
10. Xie P
11. et al.
2016Angiopoietin receptor Tie2 is required for vein specification and maintenance via regulating COUP-TFIIeLife 5:e21032https://doi.org/10.7554/eLife.21032 PubMed Google Scholar
31.
1. Pereira FA
2. Qiu Y
3. Zhou G
4. Tsai MJ
5. Tsai SY
1999The orphan nuclear receptor COUP-TFII is required for angiogenesis and heart developmentGenes Dev 13:1037–1049https://doi.org/10.1101/gad.13.8.1037 PubMed Google Scholar
32.
1. You LR
2. Lin FJ
3. Lee CT
4. DeMayo FJ
5. Tsai MJ
6. Tsai SY
2005Suppression of Notch signalling by the COUP-TFII transcription factor regulates vein identityNature 435:98–104https://doi.org/10.1038/nature03511 PubMed Google Scholar
33.
1. Wu SP
2. Cheng CM
3. Lanz RB
4. Wang T
5. Respress JL
6. Ather S
7. Chen W
8. Tsai SJ
9. Wehrens XH
10. Tsai MJ
11. et al.
2013Atrial identity is determined by a COUP-TFII regulatory networkDev Cell 25:417–426https://doi.org/10.1016/j.devcel.2013.04.017 PubMed Google Scholar
34.
1. Michelini S
2. Ricci M
3. Veselenyiova D
4. Kenanoglu S
5. Kurti D
6. Baglivo M
7. Fiorentino A
8. Basha SH
9. Priya S
10. Serrani R
11. et al.
2020TIE1 as a Candidate Gene for Lymphatic Malformations with or without LymphedemaInternational journal of molecular sciences 21https://doi.org/10.3390/ijms21186780 PubMed Google Scholar
35.
1. Brouillard P
2. Murtomaki A
3. Leppanen VM
4. Hyytiainen M
5. Mestre S
6. Potier L
7. Boon LM
8. Revencu N
9. Greene A
10. Anisimov A
11. et al.
2024Loss-of-function mutations of the TIE1 receptor tyrosine kinase cause late-onset primary lymphedemaJ Clin Invest 134https://doi.org/10.1172/JCI173586 PubMed Google Scholar
36.
1. Feng W
2. Bais A
3. He H
4. Rios C
5. Jiang S
6. Xu J
7. Chang C
8. Kostka D
9. Li G
2022Single-cell transcriptomic analysis identifies murine heart molecular features at embryonic and neonatal stagesNature communications 13:7960https://doi.org/10.1038/s41467-022-35691-7 PubMed Google Scholar
37.
1. Shen B
2. Shang Z
3. Wang B
4. Zhang LQ
5. Zhou F
6. Li TT
7. Chu M
8. Jiang HJ
9. Wang Y
10. Qiao T
11. et al.
2014Genetic Dissection of Tie Pathway in Mouse Lymphatic Maturation and Valve DevelopmentArterioscl Throm Vas 34:1221–1230https://doi.org/10.1161/Atvbaha.113.302923 PubMed Google Scholar
38.
1. Chu M
2. Li TT
3. Shen B
4. Cao XD
5. Zhong HY
6. Zhang LQ
7. Zhou F
8. Ma WJ
9. Jiang HJ
10. Xie PC
11. et al.
2016Angiopoietin receptor Tie2 is required for vein specification and maintenance via regulating COUP-TFIIeLife 5https://doi.org/10.7554/eLife.21032 PubMed Google Scholar
39.
1. Okabe K
2. Kobayashi S
3. Yamada T
4. Kurihara T
5. Tai-Nagara I
6. Miyamoto T
7. Mukouyama YS
8. Sato TN
9. Suda T
10. Ema M
11. et al.
2014Neurons limit angiogenesis by titrating VEGF in retinaCell 159:584–596https://doi.org/10.1016/j.cell.2014.09.025 PubMed Google Scholar
40.
1. Cao X
2. Xu B
3. Li X
4. Li T
5. He Y
2021A Genetically Engineered Mouse Model of Venous Anomaly and Retinal Angioma-like Vascular MalformationBio Protoc 11:e4117https://doi.org/10.21769/BioProtoc.4117 PubMed Google Scholar
41.
1. Kalucka J
2. de Rooij L
3. Goveia J
4. Rohlenova K
5. Dumas SJ
6. Meta E
7. Conchinha NV
8. Taverna F
9. Teuwen LA
10. Veys K
11. et al.
2020Single-Cell Transcriptome Atlas of Murine Endothelial CellsCell 180:764–779https://doi.org/10.1016/j.cell.2020.01.015 PubMed Google Scholar
42.
1. Liberzon A
2. Birger C
3. Thorvaldsdottir H
4. Ghandi M
5. Mesirov JP
6. Tamayo P
2015The Molecular Signatures Database (MSigDB) hallmark gene set collectionCell Syst 1:417–425https://doi.org/10.1016/j.cels.2015.12.004 PubMed Google Scholar
43.
1. Fatkin D
2. Huttner IG
3. Johnson R
2019Genetics of atrial cardiomyopathyCurrent opinion in cardiology 34:275–281https://doi.org/10.1097/HCO.0000000000000610 PubMed Google Scholar
44.
1. Arita Y
2. Nakaoka Y
3. Matsunaga T
4. Kidoya H
5. Yamamizu K
6. Arima Y
7. Kataoka-Hashimoto T
8. Ikeoka K
9. Yasui T
10. Masaki T
11. et al.
2014Myocardium-derived angiopoietin-1 is essential for coronary vein formation in the developing heartNature communications 5:4552https://doi.org/10.1038/ncomms5552 PubMed Google Scholar
45.
1. Sato TN
2. Tozawa Y
3. Deutsch U
4. Wolburg-Buchholz K
5. Fujiwara Y
6. Gendron-Maguire M
7. Gridley T
8. Wolburg H
9. Risau W
10. Qin Y
1995Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formationNature 376:70–74https://doi.org/10.1038/376070a0 PubMed Google Scholar
46.
1. Saharinen P
2. Eklund L
3. Miettinen J
4. Wirkkala R
5. Anisimov A
6. Winderlich M
7. Nottebaum A
8. Vestweber D
9. Deutsch U
10. Koh GY
11. et al.
2008Angiopoietins assemble distinct Tie2 signalling complexes in endothelial cell-cell and cell-matrix contactsNat Cell Biol 10:527–537https://doi.org/10.1038/ncb1715 PubMed Google Scholar
47.
1. Fukuhara S
2. Sako K
3. Minami T
4. Noda K
5. Kim HZ
6. Kodama T
7. Shibuya M
8. Takakura N
9. Koh GY
10. Mochizuki N
2008Differential function of Tie2 at cell-cell contacts and cell-substratum contacts regulated by angiopoietin-1Nat Cell Biol 10:513–526https://doi.org/10.1038/ncb1714 PubMed Google Scholar

Article and author information

Author information

Kai Ding
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
ORCID iD: 0009-0008-1443-0163
- Equal contribution
Beibei Xu
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
ORCID iD: 0009-0000-0861-5593
- Equal contribution
Xinhao Yu
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Xiwen Jia
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Taotao Li
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Xin Shen
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Junda Li
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Xudong Cao
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Yahui Liu
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
Zhen Zhang
Pediatric Translational Medicine Institute and Pediatric Congenital Heart Disease Institute, Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China, Shanghai Collaborative Innovative Center of Intelligent Medical Device and Active Health, Shanghai University of Medicine & Health Sciences, Shanghai, China
ORCID iD: 0000-0002-9898-054X
Yulong He
Cyrus Tang Hematology Center, Collaborative Innovation Center of Hematology, National Clinical Research Center for Hematologic Diseases, State Key Laboratory of Radiation Medicine and Protection, Cam-Su Genomic Resources Center, Suzhou Medical College of Soochow University, Suzhou, China
ORCID iD: 0000-0002-0099-3749
- For correspondence: heyulong@suda.edu.cn

Author Notes

Competing interests: No competing interests declared

Version history

Preprint posted: February 28, 2026
Sent for peer review: April 7, 2026
Reviewed Preprint version 1: May 22, 2026

Cite all versions

You can cite all versions using the DOI https://doi.org/10.7554/eLife.111156. This DOI represents all versions, and will always resolve to the latest one.

Copyright

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

views: 0
downloads: 0
citations: 0

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Significance of findings

Strength of evidence

Abstract

Introduction

Materials and Methods

Mouse models

Induced gene deletion

Embryo preparation

Immunostaining

Bulk RNA sequencing analysis of atria and ventricles

Single cell RNA-seq data analysis

Quantitative analysis of cardiac wall and trabeculae

Statistical analysis

Results

Higher TIE1 and TIE2 in atrial than ventricular endocardial cells

Analysis of TIE1 expression in endocardial cells by the single cell RNA-seq data and its role in cardiac trabeculation during embryogenesis.

The transcript levels of Tie1, Tek in atrial and ventricular endocardial cells and Angpt1 in atrial and ventricular cardiomyocytes during the embryonic stage from E10.5-E18.5 (based on the published scRNA-seq data, Feng et al., Nat Commun 2022; DOI: 10.1038/s41467-022-35691-7).

Differential requirement of TIE1 in atrial and ventricular trabeculation

Suppression of cardiac trabeculation-related gene expression by the loss of TIE1.

Alteration of atrial versus ventricular gene regulatory network upon TIE1 deficiency

Differential effect of Tie1 deficiency on atrial versus ventricular trabeculation related gene network.

Stronger ANGPT1-TIE2 interactions among atrial endocardium and myocardium

Impaired atrial trabecular formation after TIE2 insufficiency.

Synergy of TIE1 and TIE2 in atrial trabeculation

Developmental requirement of TIE1 and TIE2 in atrial and ventricular chamber morphogenesis.

Synergistic role of TIE1 and TIE2 in the atrial trabeculation.

Quantitation of the trabecular area (EMCN+), blood vessel density in ventricular wall (PECAM1+) and ventricular wall thickness in Tie1ICDiECKO;Tek+/- and littermate controls (at E14.5, 48 hours after the induced gene deletion).

Quantitation of the trabecular area (EMCN+), coronary vessel density in ventricular wall (PECAM1+) and ventricular wall thickness in Tie1ICDiECKO;Tek+/- and littermate controls 96 hours after the induced gene deletion (E16.5; tamoxifen treatment from E12.5-E13.5).

Requirement of TIE receptors for the postnatal atrial trabecular remodeling

Abnormal atrial trabecular remodeling after the postnatal deletion of Tie1 combined with one null allele of Tek.

Discussion

Endocardial versus vascular effects of TIE1

Synergy of TIE1 and TIE2 in cardiac trabeculation

Differential requirement of TIE receptors in atria and ventricles

Data availability

Acknowledgements

Additional information

Funding

Funding

Additional files

References

Article and author information

Author information

Kai Ding#

Beibei Xu#

Xinhao Yu

Xiwen Jia

Taotao Li

Xin Shen

Junda Li

Xudong Cao

Yahui Liu

Zhen Zhang

Yulong He

Author Notes

Version history

Cite all versions

Copyright

Metrics

Quantitation of the trabecular area (EMCN⁺), blood vessel density in ventricular wall (PECAM1⁺) and ventricular wall thickness in Tie1ICD^iECKO;Tek^+/- and littermate controls (at E14.5, 48 hours after the induced gene deletion).

Quantitation of the trabecular area (EMCN⁺), coronary vessel density in ventricular wall (PECAM1⁺) and ventricular wall thickness in Tie1ICD^iECKO;Tek^+/- and littermate controls 96 hours after the induced gene deletion (E16.5; tamoxifen treatment from E12.5-E13.5).

Kai Ding

Beibei Xu