Single-cell reconstruction of differentiation trajectory reveals a critical role of ETS1 in human cardiac lineage commitment
Cardiac differentiation from human pluripotent stem cells provides a unique opportunity to study human heart development in vitro and offers a potential cell source for cardiac regeneration. Compared to the large body of studies investigating cardiac maturation and cardiomyocyte subtype-specific induction, molecular events underlying cardiac lineage commitment from pluripotent stem cells at early stage remain poorly characterized.
In order to uncover key molecular events and regulators controlling cardiac lineage commitment from a pluripotent state during differentiation, we performed single-cell RNA-Seq sequencing and obtained high-quality data for 6879 cells collected from 6 stages during cardiac differentiation from human embryonic stem cells and identified multiple cell subpopulations with distinct molecular features. Through constructing developmental trajectory of cardiac differentiation and putative ligand-receptor interactions, we revealed crosstalk between cardiac progenitor cells and endoderm cells, which could potentially provide a cellular microenvironment supporting cardiac lineage commitment at day 5. In addition, computational analyses of single-cell RNA-Seq data unveiled ETS1 (ETS Proto-Oncogene 1) activation as an important downstream event induced by crosstalk between cardiac progenitor cells and endoderm cells. Consistent with the findings from single-cell analysis, chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-Seq) against ETS1 revealed genomic occupancy of ETS1 at cardiac structural genes at day 9 and day 14, whereas ETS1 depletion dramatically compromised cardiac differentiation.
Together, our study not only characterized the molecular features of different cell types and identified ETS1 as a crucial factor induced by cell-cell crosstalk contributing to cardiac lineage commitment from a pluripotent state, but may also have important implications for understanding human heart development at early embryonic stage, as well as directed manipulation of cardiac differentiation in regenerative medicine.
KeywordsCardiac lineage commitment Human pluripotent stem cells Single-cell RNA sequencing Cell-cell crosstalk ETS1 Transcription regulation
Human embryonic stem cell
Single-cell RNA sequencing
Chromatin immunoprecipitation followed by high-throughput sequencing
Unique molecular identifier
t-distributed stochastic neighbor embedding
Cardiac progenitor cells
Gene set enrichment analysis
Area under the curve
Differentiation of human embryonic stem cell (hESC) into cardiomyocytes (CMs) has been an essential model system to provide insights into the molecular mechanism of heart development [1, 2, 3, 4]. Derived CMs are a powerful tool for modeling cardiovascular diseases and drug toxicity screens and also gain widespread attention in heart regeneration . Recently, several studies used human pluripotent stem cell-derived cardiomyocytes or cardiac progenitors to repair injured myocardium in primates or even in humans [6, 7, 8]. These studies displayed considerable remuscularization and improved cardiac function, which shed light on the application of pluripotent stem cells in heart regenerative medicine. Despite these remarkable advances, adverse effects, such as arrhythmia and teratoma formation, still impede successful clinical translation, highlighting the need for a deeper understanding of the molecular paths from pluripotent stem cells to cardiomyocytes.
Compared to our knowledge of CM subtype specification and maturation at later stages of differentiation [4, 9, 10], much less is known about cardiac fate commitment at earlier stages. This was partially due to adverse effects observed in translational studies, which raised cell maturation and subtype specification as more important questions for regeneration purposes. For instance, a recent study achieved better maturation of CMs by starting physical conditioning early in the differentiation process and increasing its intensity over time . Other fronts of the protocol improvement include developing techniques to induce specific cardiomyocyte subtypes, as well as identifying early markers of cell fate prediction during differentiation . By contrast, early cardiac cell fate commitment, a critical step during heart development, is less well understood, and thus is worth exploiting for pivotal regulators of cardiac fate decisions.
Using bulk RNA-Seq technology, temporal profiling of transcriptomes during in vitro hESCs to CMs differentiation had been useful to reveal the overall dynamics of gene expression during cardiac development [1, 11]. Nevertheless, a variety of cells at different development states or from distinctly differentiated fates are mixed when performing bulk RNA-Seq, obscuring potential critical molecular events and signals taking place in cell subpopulations. Recent advancements in single-cell sequencing technology enable high-resolution observation of in vitro cardiac differentiation from hESCs to CMs dynamically [2, 4]. While cellular heterogeneity and transcriptional networks were explored in these studies, the main focus was still on possible later-stage mechanisms that lead to the immaturity of derived CMs. Therefore, to complement the missing piece of the molecular basis of cardiac lineage commitment at an early stage, we performed single-cell RNA sequencing (scRNA-Seq) across six key time points (days 0, 2, 5, 9, 14, 60) in the process of in vitro hESC-to-CM differentiation. We applied computational methods to analyze heterogeneity, to reconstruct differential trajectories, and to explore the transcription factor network around the pivotal stages of cardiac fate commitment. We demonstrated the essential role of side populations in providing a cellular microenvironment for cardiac development. We also reported the key regulatory role of the ETS1 of transcription factors in cardiac lineage specification, based on computational analysis and experimental validation. In addition, we also created an online resource for visualizing our Single-Cell data (https://hanlab.uth.edu/shiny/hESC2CM/).
Comprehensive analysis of cardiac differentiation at single-cell resolution
Reconstruction of developmental trajectory of cardiac differentiation from human embryonic stem cells
Crosstalk between endoderm cells and cardiac progenitors potentially regulates cardiac lineage commitment
ETS1 as an important downstream factor of cell-cell interaction regulating cardiac lineage commitment
ETS1 highly correlates with cardiac differentiation
ETS1 directly regulates cardiac genes to promote cardiac differentiation
The in vitro hESC-to-CM differentiation system has been a powerful tool to model human heart development. The rapid advancement of the scRNA-Seq technology enabled a high-resolution representation of gene expression dynamic during differentiation. Combining it with the state-of-art computational technologies, biological curation, and annotations, we were able to dissect the heterogeneity of cells from a single sampling point, as well as to reconstruct a trajectory from multiple sampling points across several developmental stages.
In this study, we focused on the early stages of cardiac cell fate commitment to uncover dynamically changing cell heterogeneity and important regulatory mechanism governing this process. In the past years, several protocols have been developed to obtain a relatively homogeneous CM population. For instance, one of the recent efficient monolayer-based protocols is through modulating canonical Wnt/β-catenin signaling using small molecules on hESCs at early stages [13, 33]. These highly efficient protocols hint at a homogeneous differentiation process throughout the early-middle stage of cardiac differentiation. However, using scRNA-Seq, we found that cardiac progenitor cells (CPCs) emerged alongside several side populations and showed distinct bifurcation paths on the inferred developmental trajectory at the pivotal time junction of cardiac cell fate commitment. By analyzing enriched ligand-receptor pairs in the subpopulations, we unexpectedly found that endoderm cells, one of the side populations at day 5, were deduced to function as a supporting cell population for the development of CPCs. The interaction between endoderm and mesoderm that governs the lineage specification is well known in vivo. Here, we provided evidence that this supporting model still exists in in vitro cardiac differentiation. Further experiments may be required to rank the importance of signaling pathways and ligands utilized to support cardiac lineage commitment in cell-cell interactions.
By using computational analysis of gene regulatory network and performing a loss-of-function experiment, we revealed that transcription factor ETS1 plays an essential regulatory role in mediating early CM development. We have found that both MEF2C and ETS1 were highly enriched in the CPC population and showed a significant level of co-expression with each other, which later assumed different regulatory roles in CM populations and side populations, respectively. In addition, we observed that the Ras signaling pathway was enriched in the CPC population. Interestingly, ETS1, but no other members of the ETS family, was specifically required for the migration of RAS/ERK-activated endothelial cells. These observations further indicated that cellular crosstalk between CPCs and the endoderm cell subpopulation may have activated the Ras signaling pathway in CPCs, the latter of which may further induce ETS1 expression in CPCs. Given the existing knowledge of crosstalk between endothelial cells and cardiomyocytes , and that a recent study demonstrated a favorable effect on the maturity of hESC-derived CMs by co-culturing them with endothelial cells , we proposed that ETS1 is an important factor in determining early CM fate, possibly through interaction between progenitors of CMs and side cells. These findings warrant a deeper dissection of the crosstalk between the endothelial/endoderm and cardiac lineage, which will further our understanding of heart development and diseases.
Several studies have studied the potential role of the ETS family in heart development. For instance, Schachterle et al. have identified members of the ETS family which function as enhancers to promote cardiac transcription factor expression to regulate heart development . Islas et al. found that ETS2 and MESP1 can work together to generate cardiac progenitors de novo from fibroblasts . In our scRNA-Seq based, we identified that ETS1 was highly expressed in a small portion of cardiac progenitor cells at day 5 and highly expressed in fibroblast-like CM later at day 9. Also, ETS1 exhibited direct binding towards a large set of cardiac structural genes at day 9. The transient expression of ETS1 suggested its novel regulatory role in early cardiac commitment possibly is required for stabilization of cardiomyogenic fate.
A recent single-cell study of cardiac differentiation from human pluripotent stem cells identified HOP homeobox (HOPX) as an important cardiac regulator that could be potentially dysregulated in vitro leads to maturation issues of hESC-CMs . Here, we observed a similar phenotype where HOPX was elevated in the CPC population on day 5 and was barely expressed thereafter (Fig. 5a, b). In addition to HOPX, we also observed an abrupt elevation in the expression of natriuretic peptide B (NPPB), a gene related to cardiac hypertrophy in adult heart, at an early stage of CM differentiation (T09), which dropped quickly thereafter (Fig. 1b). It is known that NPPB expression is strongly upregulated in the myocardium at embryonic and fetal stages in vivo . Inspired by these observations, one might wonder whether NPPB may be another regulator in cardiac differentiation whose dysregulation could potentially lead to maturation issues of hESC-CMs. Taking it a step forward, it would be useful to identify genes with similar expression patterns as HOPX and NPPB to help us understand the molecular basis of cardiac maturation both in vivo and in vitro.
In the study, we not only provided a rich resource and cell-cell crosstalk model supporting cardiac differentiation from hESCs in vitro at single-cell resolution, but also identified ETS1 as a pivotal regulator in early cardiac lineage commitment. These findings may further our understanding of the molecular basis of heart development and provide knowledge for developing novel strategies dictating cardiac differentiation for regeneration purposes.
In vitro hESC culture and differentiation
Human ESC lines (H1) were maintained on Matrigel-coated plates (BD Biosciences) in Essential 8 Medium (STEMCELL Technologies) and ROCK inhibitor (Selleck). The medium was changed every day. For cardiac differentiation, human ESC lines (H1) at ~ 100% confluence were incubated with a differentiation medium comprising RPMI 1640 medium (Gibco) and B27 supplement minus insulin (Invitrogen). On day 0, CHIR99021 (Selleck), a selective glycogen synthase kinase 3β inhibitor, was added to the differentiation medium (3 μM final). On day 3, the Wnt antagonist, IWR-1 (Selleck), was added to the differentiation medium (5 μM final). On day 5, the medium was removed and replaced with a differentiation basal medium without any inhibitors. On day 8, the cells were incubated with a medium consisting of RPMI 1640 medium and B27 supplement plus insulin (Invitrogen). The medium was changed every 3 days for a desired time of culture.
scRNA-Seq library preparation
The time point of selected 6 different stages are day 0 (hESC), day 2 (MES), day 5 (CP), day 9 (CM), day 14 (immature CM), and day 60 (mature CM). Cells were digested into a single-cell suspension and then filtered through a 70-μm cell strainer into 15-ml tubes. Single-cell sequencing library was generated by iCell8 platform (Takara). In brief, isolated cells were stained with a mixture of Hoechst 33342 and propidium iodide (R37610, Thermo Scientific) according to the manufacturer’s instruction. After staining, cells were washed by PBS and counted by MoxiTM Automated Cell Counter. Afterwards, the cell suspension (20,000 cells/ml) was submitted to the MultiSample NanoDispenser (MSND, Wafergen Biosystems) for single-cell preparation. The dispensed cells were then imaged with the Imaging Station, and single live cells, defined by Hoechst-positive and propidium iodide-negative staining, were selected. Selected cells were subjected to reverse transcription and first-step amplification in a Chip Cycler (Bio-Rad), and the resulting cDNA was purified and size-selected with Agencourt AMPure XP beads (A63880, Beckman Coulter). One nanogram of purified cDNA was applied to generate a sequencing library by using Nextera XT DNA sample preparation Kit (FC-131-1024, Illumina). Libraries were sequenced on the NextSeq500 sequencer (Illumina) using the 26-nt and 50-nt paired-end sequencing protocol.
scRNA-Seq data bioinformatics pre-processing
Raw sequencing reads were processed in the following four steps: (1) Only read pairs whose read 1 uniquely mapped the pre-defined barcode tag (10 nt) and UMI (14 nt) were considered as valid. (2) Read pairs were filtered by cutadapt (v1.8.1)  with the following parameters: -m 20 --trim-n --max-n 0.7 –q20. (3) Reads were then aligned to genomes of human, Escherichia. coli, mycoplasma, yeast, and adapter sequences by Bowtie2 (v2.2.4) . Contaminants were filtered by FastQ Screen (v0.5.1.4) . Clean reads were then mapped to UCSC human genome (hg19) via STAR (v2.5.2b)  and assigned to Ensembl genes41 by featureCounts .
Genes with detected expression (UMI counts ≥ 1) in at most 3 cells will be considered as not expressed and filtered. The human heart is a well-known high-energy tissue, and previous bulk RNA-Seq studies suggested that mitochondrial transcripts comprise almost 30~40% of total mRNA [53, 54]. Here, we observe cells with an expected level of mitochondrial gene expressed along with differentiation (Additional file 1: Figure S1A), so mitochondrial gene expression was not considered as a quality control measure and filtered. Single cells with less than 300 expressed genes were considered as low-quality cells and were filtered. An expression matrix of a total of 7622 single cells was retained after pre-processing.
scRNA-Seq data clustering and trajectory reconstruction
Seurat R package (v2.3.4)  was used to perform further feature selection and clustering. Because of a high dropout rate and stochastic nature of gene expression, single cells were further selected with a number of expressed genes between 1000 (low threshold) and 10,000 (high threshold). A total of 6879 cells were retained for downstream analysis. UMI counts were further normalized among different stages and log-transformed with a default scale factor in Seurat. The highly variable genes (specified in Seurat as outliers on a mean variability plot) were selected with the following parameters: “dispersion.function = LogVMR; x.low.cutoff = 0.0125; x.high.cutoff = 3; y.cutoff = 1.” The confounding factor of expressional variation coming from cell cycle was regressed out using an embedded method in Seurat under the consideration that developmental trajectory could be better inferred by removing the biological covariates (e.g., cell cycle) . Nevertheless, clustering patterns are similar between with and without regression out of cell cycle (Fig. 2a versus Additional file 1: Figure S6A), and there are no significant differences in the composition of cells in different phases among different subclusters (Pearson’ chi-squared test of independence, p > 0.05, Additional file 1: Figure S6B). Principal component analysis (PCA) was performed on the expression matrix to capture the eigenvectors. The clustering was using Louvain clustering algorithm on shared nearest neighbor (SNN) as wrapped in “FindCluster” function of Seurat. Cell clusters are defined using the top ten significant principal components under typical resolution “1.0.” t-SNE were then performed after the cell clustering to visualize the results.
The subpopulation marker genes were defined using “FindAllMarkers” function wrapped in Seurat. The Wilcoxon rank-sum test was used with multiple correction. The cutoffs for adjusted p value, log fold change, and minimum cell percentage were set as 0.01, 0.25, and 0.05, respectively.
The enrichment level of gene sets corresponded to cell types/development stages were measured by normalized area under the curve (AUC) through AUCell module of the SCENIC , and the active regulons were determined by AUCell default threshold. The AUC of each gene set was normalized across all the subpopulations across four time points. The manually curated human single cell (cell types/development stages) markers are gathered from the CellMarker database . Markers from the single-cell dataset were selected in the enrichment.
We performed cell differential trajectory reconstruction by Monocle (v2.6.4)  with the top 5000 differentially expressed genes across the subpopulations of 4 time points. To find the top 5000 differentially expressed genes, we used the “differentialGeneTest” function of Monocle and employ the model that finds genes that change as a function of inferred pseudotime. The input of monocle is the raw UMI count. The “DDRTree” algorithm was used in dimension reduction and visualization of the trajectory. The bifurcation tree from 6 time points was constructed on the bases of the cell differential trajectory of all cells using differentially expressed human transcription factor genes (n = 1253) across 6 time points.
Human CPCs at day 5 were dissociated with 0.25% trypsin (Gibco) and gently scraped off from the wells. Cells were fixed with 4% (vol/vol) paraformaldehyde for 15 min and then permeated and blocked with 0.3% (vol/vol) Triton X-100 and for 5% BSA in goat serum solution for 1 h at room temperature. The samples were then incubated with primary antibody against APOA2 (1:400), FOXA2 (1:400), MESP1 (1:400), and MEF2C (1:400) overnight at 4 °C. Next day, the samples were incubated with secondary fluoresce-labeled anti-mouse/rabbit antibody (1:500) for 1 h at room temperature. The nuclei were stained with DAPI (Invitrogen) for 5 min. Images were captured under Leica sp8 confocal microscopy.
Gene Ontology, KEGG enrichment, and GSEA analyses
We used R package clusterProfiler (v3.6.0)  to perform Gene Ontology enrichment, KEGG enrichment, and GSEA analyses for subpopulation marker genes and differentially expressed genes. Only enriched terms related to cardiac development are visualized in the figure.
Quantitatively characterizing cell-cell communications
Human transcription factor proteins were extracted from the UniProt database . In total, 1781 genes were retrieved for TF genes. A curated list of human ligand-receptor pairs (n = 2557) was retrieved from supplementary files of a previous study . In protein acting information obtained from the STRING database (9606.protein.actions.v10.5), we chose “binding” mode for interacting pairs of ligand-receptors, “activation” and “inhibition” with “combined_score >= 700” as targets as applied in a previous study . The enriched ligand-receptor pairs were retrieved by overlapping the list with marker genes of subpopulations as described in the previous section. The R version of SoptSC  was used on our data for calculating the probability matrix of signals being passed between cells and visualization. The enriched ligand-receptor pairs in VEGF signaling and their targets were used to generate Fig. 3d.
Gene regulatory network analysis
The analysis of the regulatory network and regulon activity was performed by SCENIC (pyscenic, v0.8.7) . The input to SCENIC is an expression matrix of four time points (day 2, day 5, day 9, and day 14). The MEF2C and ETS1 regulon were imputed using co-expression and binding motifs’ genomic positions specified in the GENIE3 and RcisTarget modules of the SCENIC. The normalized enrichment score (NES) of the transcription factor binding motifs (TFBS) was calculated, and NES > 3.0 was considered as significantly enriched. The gene correlation network was visualized using R package igraph . The regulon activity (measured in AUC) was analyzed by AUCell module of the SCENIC , and the active regulons were determined by AUCell default threshold.
Experimental validation of regulatory rule of ETS by knocking down ETS1 during cardiac lineage commitment
Human stem cell lines (H1) were seeded onto 6-well plates and transduced with non-targeting (shNT) or ETS1 (shETS1) lentivirus-based shRNAs. Cells expressing control or ETS1 shRNA were selected by puromycin (1 μg/ml) treatment for 2–3 days. Total RNA was extracted from cells using the GeneJet RNA Purification Kit (K0732, Thermo Scientific) 10 days after differentiation. 0.5 μg of total RNA was reverse transcribed to generate cDNA using the iScript cDNA Synthesis Kit (1708891, Bio-Rad) according to the manufacturer’s instructions. qPCRs were performed using the iTaqUniversl SYBR Green supermix (1725121, Bio-Rad) on the Bio-Rad CFX-384 or CFX-96 real-time PCR System. Actin was used for normalization.
Chromatin immunoprecipitation followed by high-throughput sequencing
Briefly, differentiated cells were fixed using 1% formaldehyde for 10 min, and 0.125 M glycine was added to stop fixation. Cells were harvested, and DNA was fragmented to 300–500 bp by sonication with a Covaris S220 sonicator. Immunoprecipitation was performed with antibodies conjugated to Dynabeads Protein G beads (1004D, Life Technologies). ChIP DNA was eluted, reverse cross-linked, extracted with phenol/chloroform, and precipitated. For ChIP-Seq, 1 ng ChIP DNA or input DNA was used to generate sequencing libraries using the Nextera XT DNA sample preparation Kit (FC-131-1024, Illumina). Libraries were sequenced on the NextSeq500 sequencer (Illumina) using the 35-nt paired-end sequencing protocol.
ChIP-Seq data was aligned to human genome (hg19) by Rsubread (v1.24.2) . Only unique mapped reads were kept. All the bam files were sorted by samtools (v0.1.19-44,428 cd) . Peaks were called by macs2 (v220.127.116.1160309)  with parameter “broad.” “BigWig” files were produced by HOMER toolkit .
LW and LH conceived and supervised the project. LW, HR, and LH designed and performed the research. HR, HX, and JL performed the data analyses. YL performed the experiments. ZR, LM, FY, PY, YY, ZZ, SL, LD, and BZ interpreted the results. HR constructed the data portal. LW, HR, and LH wrote the manuscript with input from all other authors. All authors read and approved the final manuscript.
This work was supported by the National Key R&D Program of China (2017YFA0103700), CAMS Innovation Fund for Medical Sciences (CIFMS,2018-I2M-3-002, 2016-I2M-1-015, 2017-I2M-1-003), grants (91639107, 31671542, 81722006 to LW, 81700337 to BZ) from the National Natural Science Foundation of China, grant (RR150085) from Cancer Prevention & Research Institute of Texas to CPRIT Scholar in Cancer Research (LH).
Ethics approval and consent to participate
Consent for publication
All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
- 2.Friedman CE, Nguyen Q, Lukowski SW, Helfer A, Chiu HS, Miklas J, et al. Single-cell transcriptomic analysis of cardiac differentiation from human PSCs reveals HOPX-dependent cardiomyocyte maturation. Cell Stem Cell. 2018;23:586–98.e8. https://doi.org/10.1016/j.stem.2018.09.009.CrossRefPubMedPubMedCentralGoogle Scholar
- 5.Burridge PW, Li YF, Matsa E, Wu H, Ong S-G, Sharma A, et al. Human induced pluripotent stem cell-derived cardiomyocytes recapitulate the predilection of breast cancer patients to doxorubicin-induced cardiotoxicity. Nat Med. 2016;22:547–56. https://doi.org/10.1038/nm.4087.CrossRefPubMedPubMedCentralGoogle Scholar
- 13.Sharma A, Li G, Rajarajan K, Hamaguchi R, Burridge PW, Wu SM. Derivation of highly purified cardiomyocytes from human induced pluripotent stem cells using small molecule-modulated differentiation and subsequent glucose starvation. J Vis Exp. 2015;:3–10. doi: https://doi.org/10.3791/52628.
- 14.Cai C-L, Liang X, Shi Y, Chu P-H, Pfaff SL, Chen J, et al. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev Cell. 2003;5:877–89. https://doi.org/10.1016/S1534-5807(03)00363-0.CrossRefPubMedPubMedCentralGoogle Scholar
- 18.Kattman SJ, Witty AD, Gagliardi M, Dubois NC, Niapour M, Hotta A, et al. Stage-specific optimization of activin/nodal and BMP signaling promotes cardiac differentiation of mouse and human pluripotent stem cell lines. Cell Stem Cell. 2011;8:228–40. https://doi.org/10.1016/j.stem.2010.12.008.CrossRefPubMedGoogle Scholar
- 26.Mummery C, Ward-van Oostwaard D, Doevendans P, Spijker R, van den Brink S, Hassink R, et al. Differentiation of human embryonic stem cells to cardiomyocytes: role of coculture with visceral endoderm-like cells. Circulation. 2003;107:2733–40. https://doi.org/10.1161/01.CIR.0000068356.38592.68.CrossRefPubMedGoogle Scholar
- 27.Bader A, Gruss A, Höllrigl A, Al-Dubai H, Capetanaki Y, Weitzer G. Paracrine promotion of cardiomyogenesis in embryoid bodies by LIF modulated endoderm. Differentiation. 2001;68:31–43. https://doi.org/10.1046/j.1432-0436.2001.068001031.x.CrossRefPubMedGoogle Scholar
- 32.Foulds CE, Nelson ML, Blaszczak AG, Graves BJ. Ras/mitogen-activated protein kinase signaling activates Ets-1 and Ets-2 by CBP/p300 recruitment. Mol Cell Biol. 2004;24:10954–64. https://doi.org/10.1128/MCB.24.24.10954-10964.2004.CrossRefPubMedPubMedCentralGoogle Scholar
- 33.Liu Q, Jiang C, Xu J, Zhao M-T, Van Bortle K, Cheng X, et al. Genome-wide temporal profiling of transcriptome and open chromatin of early cardiomyocyte differentiation derived from hiPSCs and hESCs. Circ Res. 2017;121:376–91. https://doi.org/10.1161/CIRCRESAHA.116.310456.CrossRefPubMedPubMedCentralGoogle Scholar
- 35.Kraut N, Frampton J, McNagny KM, Graf T. A functional Ets DNA-binding domain is required to maintain multipotency of hematopoietic progenitors transformed by Myb-Ets. Genes Dev 1994;8:33–44. http://www.ncbi.nlm.nih.gov/pubmed/8288126. Accessed 10 May 2019.
- 36.Hromas R, Orazi A, Neiman RS, Maki R, Van Beveran C, Moore J, et al. Hematopoietic lineage- and stage-restricted expression of the ETS oncogene family member PU.1. Blood. 1993;82:2998–3004. http://www.ncbi.nlm.nih.gov/pubmed/8219191. Accessed 10 May 2019.
- 37.Kanki Y, Nakaki R, Shimamura T, Matsunaga T, Yamamizu K, Katayama S, et al. Dynamically and epigenetically coordinated GATA/ETS/SOX transcription factor expression is indispensable for endothelial cell differentiation. Nucleic Acids Res. 2017;45:4344–58. https://doi.org/10.1093/nar/gkx159.CrossRefPubMedPubMedCentralGoogle Scholar
- 39.Islas JF, Liu Y, Weng K-C, Robertson MJ, Zhang S, Prejusa A, et al. Transcription factors ETS2 and MESP1 transdifferentiate human dermal fibroblasts into cardiac progenitors. Proc Natl Acad Sci U S A. 2012;109:13016–21. https://doi.org/10.1073/pnas.1120299109.CrossRefPubMedPubMedCentralGoogle Scholar
- 60.UniProt: a worldwide hub of protein knowledge. Nucleic Acids Res. 2019;47:D506–15. doi: https://doi.org/10.1093/nar/gky1049.
- 63.Csardi G, Nepusz T, others. The igraph software package for complex network research. InterJournal, Complex Syst 2006;1695:1–9.Google Scholar
- 67.Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell. 2010;38:576–89. https://doi.org/10.1016/j.molcel.2010.05.004.CrossRefPubMedPubMedCentralGoogle Scholar
- 68.Ruan H, Xu H, Wang L, Han L. Genome-wide examination of ETS1 genomic occupation from pluripotent to cardiac hESC: GEO; 2019. (GSE129986). Available from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE129986. Accessed 17 Apr 2019.
- 69.Ruan H, Wang L, Han L. Single-cell reconstruction of differentiation trajectory reveals essential dynamics in human cardiac lineage commitment: GEO; 2019. (GSE129987). Available from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE129987. Accessed 17 Apr 2019.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.