Retinoic acid signaling modulation guides in vitro specification of human heart field-specific progenitor pools

Zawada, Dorota; Kornherr, Jessica; Meier, Anna B.; Santamaria, Gianluca; Dorn, Tatjana; Nowak-Imialek, Monika; Ortmann, Daniel; Zhang, Fangfang; Lachmann, Mark; Dreßen, Martina; Ortiz, Mariaestela; Mascetti, Victoria L.; Harmer, Stephen C.; Nobles, Muriel; Tinker, Andrew; De Angelis, Maria Teresa; Pedersen, Roger A.; Grote, Phillip; Laugwitz, Karl-Ludwig; Moretti, Alessandra; Goedel, Alexander

doi:10.1038/s41467-023-36764-x

Download PDF

Article
Open access
Published: 03 April 2023

Retinoic acid signaling modulation guides in vitro specification of human heart field-specific progenitor pools

Nature Communications volume 14, Article number: 1722 (2023) Cite this article

5218 Accesses
9 Citations
121 Altmetric
Metrics details

Subjects

Abstract

Cardiogenesis relies on the precise spatiotemporal coordination of multiple progenitor populations. Understanding the specification and differentiation of these distinct progenitor pools during human embryonic development is crucial for advancing our knowledge of congenital cardiac malformations and designing new regenerative therapies. By combining genetic labelling, single-cell transcriptomics, and ex vivo human-mouse embryonic chimeras we uncovered that modulation of retinoic acid signaling instructs human pluripotent stem cells to form heart field-specific progenitors with distinct fate potentials. In addition to the classical first and second heart fields, we observed the appearance of juxta-cardiac field progenitors giving rise to both myocardial and epicardial cells. Applying these findings to stem-cell based disease modelling we identified specific transcriptional dysregulation in first and second heart field progenitors derived from stem cells of patients with hypoplastic left heart syndrome. This highlights the suitability of our in vitro differentiation platform for studying human cardiac development and disease.

Single-cell analysis of cardiogenesis reveals basis for organ-level developmental defects

Article 24 July 2019

Generation of heart-forming organoids from human pluripotent stem cells

Article 10 November 2021

A genome-wide CRISPR screen identifies BRD4 as a regulator of cardiomyocyte differentiation

Article 23 February 2024

Introduction

The heart is the first organ that forms during embryogenesis and its functionality depends on the proper spatiotemporal assembly of multiple progenitor cell populations that ultimately give rise to the diverse cell types within the specialized cardiac structures.

Lineage tracing and clonal cell analyses in mice have identified two main sources of cardiovascular progenitors in the developing mesoderm, known as the first and second heart fields (FHF and SHF), which show differential contribution to specific heart compartments¹. A growing body of evidence suggests that the specification of FHF and SHF progenitors already occurs during gastrulation depending on their position within the primitive streak, which impacts the signaling cues they receive^2,3. Moreover, besides having discrete molecular signatures at primitive streak stages, progenitors that later contribute to the ventricles or to the outflow tract (OFT) and atria leave the streak at different time points and form distinct regions of the cardiogenic mesoderm⁴. The heterogeneity of the heart fields is further highlighted by the recent discovery of a multipotent cardiac progenitor that generates cardiomyocytes (CMs) of the left ventricle (LV), atria and atrioventricular canal (AVC) as well as cells of the epicardium. This so-called juxta-cardiac field (JCF) resides in a distinct region within the FHF in close proximity to the extraembryonic tissue^5,6.

The patterning of the primitive streak within the embryo occurs in response to specific signaling cues including Wnt, BMP, and Activin signaling⁷. Stimulation and inhibition of these signaling pathways in a temporally controlled manner can be used to direct in vitro differentiation of human pluripotent stem cells (hPSCs) into a posterior or anterior primitive streak-like fate, with cardiovascular progenitor cells arising from the latter⁸. Formation of the anterior-posterior boundaries of the cardiogenic mesoderm is, in part, controlled by retinoic acid (RA) signaling⁹, which restricts the border of the anterior SHF (aSHF—ultimately giving rise the outflow tract (OFT) and the right ventricle (RV))¹⁰ while maintaining its posterior component (posterior SHF (pSHF)—which generates the atria and the sinus venosus). Posteriorization of the SHF through RA signaling has been utilized during cardiac differentiation of hPSCs to promote the formation of atrial CMs and epicardial cells^11,12,13,14. Notably, the addition of low-dose RA during early mesoderm formation results in the generation of hPSC-derived cardiovascular progenitors that can self-organize into three-dimensional (3D) structures comprising of CMs and endocardial cells, as reported recently¹⁵. On a transcriptional level, these progenitors resemble murine FHF cells, which are the first to migrate from the primitive streak during mid-streak stages, form the primitive heart tube, and later give rise to the LV^4,15. Such effect of RA during early in vitro cardiogenesis could reflect its inhibitory role on aSHF development providing a permissive environment for FHF lineage commitment, which has yet not been investigated in detail.

Our study aimed at systemically defining how RA impacts human cardiovascular progenitor specification during early mesoderm formation. We approached this by combining genetic labeling, transcriptomics, ex vivo human-mouse embryonic chimeras, and in vitro disease modeling. Our findings indicate that a wide spectrum of mesoderm-derived cardiovascular progenitor cells including progenitors of the JCF can be differentiated from hPSCs by modulating RA signaling in early mesoderm-like cells. These progenitors contribute to the expected cardiac structures when injected into the cardiac crescent of ex vivo cultured mouse embryos, highlighting the similarity to their native counterparts. Moreover, they allowed us to uncover heart field-specific differentiation defects in an in vitro disease model of hypoplastic left heart syndrome (HLHS), a congenital heart disease that primarily affects the LV and left ventricular outflow tract (LVOT). Overall, this study presents a versatile toolbox for the generation of defined cardiovascular progenitor pools from hPSCs and shows its applicability for studying chamber-specific heart disease.

Results

RA signaling and cardiovascular progenitor specification

To investigate the influence of RA dosage and timing on the appearance and characteristics of early human cardiovascular progenitors in the cardiogenic mesoderm, we utilized a growth-factor based protocol for the directed differentiation of hPSCs towards CMs, which is distinct from other in vitro CM differentiation protocols utilizing RA to generate atrial CMs^11,13,14 (Fig. 1a). In the absence of RA, mesoderm induction (TBXT) was detected within the first 24 h of differentiation, followed by the expression of early cardiac progenitor markers (NKX2.5, ISL1) at day 3, shortly after Wnt signaling inhibition (Fig. 1b). Concomitant upregulation of BMP4, FGF10, CXCR4, and LGR5 along with TBX1 and WNT5A at day 5 suggested that these cells adopted a fate resembling cardiovascular progenitors of the aSHF (Fig. 1c, d; Supplementary Fig. 1a)^16,17,18,19. Interestingly, addition of RA from day 1.5 to day 5.5 resulted in a dose-dependent downregulation of aSHF markers at the progenitor state (days 4–5) and an upregulation of genes related to a posterior fate, such as WNT2 and HOXB1, suggesting a posteriorization of the cardiac progenitor pool^17,20,21 (Fig. 1d). This is in line with observations during cardiac development in the mouse, where RA is critical in establishing the anterior-posterior axis within the heart fields²². The expression level of the pan cardiac progenitor marker NKX2.5 was comparable between the different treatments reflecting a similar efficiency in cardiac progenitor specification (Supplementary Fig. 1a). Notably, we also detected a strong upregulation of TBX5, which at the equivalent stage in vivo marks cardiac progenitors of the murine FHF¹⁷, as well as THBS4, which is highly enriched in FHF progenitor cells in the mouse embryo at embryonic day 7.75 (E7.75)⁶ (Fig. 1d). During further differentiation, we observed earlier and higher expression of sarcomeric genes such as TNNT2 and MYL3 in presence of RA compared to differentiation without RA, which was accompanied by a faster downregulation of ISL1 (Fig. 1e, Supplementary Fig. 1b). In the first 24 h after RA application ISL1 expression remained similar in both conditions suggesting that RA does not directly downregulate ISL1 but rather induces changes in the transcriptional network leading to its rapid downregulation (Fig. 1e). In concordance with these transcriptional changes, beating foci emerged on average two days earlier after addition of RA (Fig. 1f). Notably, rapid downregulation of ISL1 expression and rapid generation of functional CMs are among the characteristic features of FHF cardiovascular progenitors, which are necessary to form the primitive heart tube in vivo⁴.

**Fig. 1: RA signaling modulation impacts cardiovascular progenitor specification.**

At day 30, expression levels of TNNT2 as well as the percentage of cTnT⁺ cells were comparable between differentiations with and without RA, suggesting efficient CM differentiation in both conditions (Supplementary Fig. 1c, d). Without RA, cells expressed high levels of ventricular CM genes such as MYL2, IRX4, and IRX3, while atrial and nodal CM markers such as NR2F2, KCNA5, KCNJ3, and SHOX2 were absent (Fig. 2a; Supplementary Fig. 1d). In addition, these cells expressed transcripts typical of CMs found in the OFT, such as LTBP3 and RSPO3¹⁸ (Fig. 2a), providing further evidence that this differentiation condition induces an aSHF-like fate. Upon RA addition, we detected a dose-dependent change in expression of FHL2 (Fig. 2a), which was reported to be enriched in left ventricular CMs of adult human hearts²³. At low dosage of RA (0.5 μM), FHL2 expression increased with time of treatment, but it declined again at higher dose (1 μM) and prolonged treatment. We observed similar dynamics for the ventricular marker MYL2. Genes associated with CMs of the OFT showed an inverse relation to the FHL2 expression pattern with the lowest expression levels at 0.5 µM RA for 4 days (Fig. 2a). Atrial CM markers increased only marginally at low dosage of RA, but a more substantial expression was observed upon higher dose and longer treatment (Fig. 2a, Supplementary Fig. 1d). Taken together, these results suggest that the occurrence of FHF-like progenitors is tightly controlled by timing and dosage of RA addition.

**Fig. 2: RA signaling modulation guides differentiation into distinct cardiomyocyte subtypes.**

To confirm our findings on the transcriptomic level we performed immunofluorescence analysis for the ventricular specific myosin light chain isoform—MLC2v, and atrial isoform—MLC2a (Fig. 2b) as well as flow cytometry-based quantification of cells expressing the atrial specific protein NR2F2 (Fig. 2c, Supplementary Fig. 1e). Without addition of RA, most of the cells were exclusively positive for MLC2v and we observed only around 10% of NR2F2⁺ cells, confirming that most cells acquired a ventricular-like fate (Fig. 2b, c). This did not change upon addition of RA up to 1 µM for 4 days (Fig. 2b, c). However, longer exposure to RA (1 µM for 10 days) resulted in an increase of MLC2a⁺ CMs (Fig. 2b) as well as NR2F2⁺ cells (Fig. 2c), suggesting partial atrial fate acquisition in line with previous reports¹¹.

To further explore the identity of the CMs generated with and without RA, we performed electrophysiological examination and focused on the highest RA dose that did not lead to a significant increase in atrial fate acquisition (1 µM for 4 days). In line with our findings on the transcriptomic and protein level, this analysis showed that the majority of CMs obtained in both conditions had indeed ventricular identity (Fig. 2d, e; Supplementary Data 1). Without RA, over a quarter of cells showed “intermediate” action potential characteristics, suggesting that they have not fully differentiated yet (Fig. 2d, e; Supplementary Data 1). However, among the cells classified as ventricular action potential properties such as the mean diastolic potential and amplitude were comparable between the two groups (Fig. 2f). Upon RA addition, we did observe a slight increase in cells with a nodal-like action potential (Fig. 2d) which is in line with the observed increase in SHOX2 expression (Supplementary Fig. 1d), probably reflecting a permissive environment for sinus nodal commitment²⁴.

Based on these findings, we hypothesized that a time- and dose-limited exposure to RA during early mesoderm formation is not efficient in inducing a pSHF-like fate resulting in atrial CMs, but rather promotes the specification of progenitors resembling a FHF-like fate, resulting in CMs with a (left) ventricular identity. Since the intermediate dose of RA (0.5 µm, 4 days) was sufficient to suppress aSHF genes and induce FHF commitment (Fig. 1d) we decided to use this dose in further experiments.

To track the appearance of distinct cardiovascular progenitor pools and enable enrichment of specific subpopulations for in-depth analysis, we generated a double reporter human embryonic stem cell (hESC) line expressing eGFP and nuclear mCherry (H2B-mCherry) under the control of the endogenous NKX2.5 and TBX5 locus, respectively (ES03 TN cell line, Fig. 3a). We confirmed proper reporting of the fluorescent markers using live imaging (Fig. 3b), qPCR, and immunofluorescence staining (Supplementary Fig. 2a, b). During differentiation, the modified hESC line displayed only a slight reduction in expression levels of NKX2.5 and TBX5 compared to the parental line, most likely due to partial compensation of the loss of transcription from the targeted gene allele by the untargeted allele (Supplementary Fig. 2c). In addition, the expression patterns of key cardiac differentiation genes and the percentage of cTnT⁺ cells were comparable between the reporter and parental line as well as an unrelated human induced pluripotent stem cell (hiPSC) line, suggesting that the reporter line is not affected by relevant haploinsufficiency of the targeted transcription factors (Supplementary Figs. 2c, 3a–c). Without addition of RA, eGFP⁺ cells (NKX2.5⁺) appeared from day 3 while mCherry⁺ (TBX5⁺) cells were virtually absent until day 8 (Fig. 3c; Supplementary Fig. 4a). eGFP⁺ cells maintained ISL1 expression while progressively upregulating the CM marker cTnT from day 5 (Fig. 3d). With addition of RA, eGFP⁺ cells were first detected at day 3 and both mCherry⁺ and eGFP⁺/mCherry⁺ populations emerged from day 4 (Fig. 3c; Supplementary Fig. 4b). In this condition, we observed a rapid downregulation of ISL1 and upregulation of cTnT in all populations (Fig. 3d; Supplementary Fig. 4c, d), again suggestive of a FHF-like fate.

RA signaling patterns the cardiogenic mesoderm

To gain further insights into the formation of cardiac progenitors in response to limited RA exposure, we performed single-cell RNA sequencing (scRNA-seq) on early mesodermal cells at day 1.5 (right before the addition of RA) as well as on individual cardiovascular progenitor populations at day 4.5 (Fig. 3e). Using flow cytometry-based sorting, we isolated eGFP⁺ (NKX2.5) cells obtained in the absence of RA and the three populations with differential eGFP (NKX2.5) and mCherry (TBX5) expression emerging in the presence of RA (Fig. 3e; Supplementary Fig. 5a). Low dimensionality embedding (UMAP) of all cells merged resulted in the formation of five distinct groups: early mesodermal cells (day 1.5), endothelial/endocardial progenitor cells (E-PCs), endodermal cells, and cardiovascular progenitor cells derived in the absence (CPC-RA−) or presence of RA (CPC-RA+) (Fig. 3f; Supplementary Fig. 5b; Supplementary Data 2).

Cells at day 1.5 represented a homogenous cell population expressing mesodermal genes such as TBXT and MESP1 (Supplementary Fig. 5c). The expression of genes such as EOMES, GSC, and MIXL1 suggested that these cells match a mid/anterior primitive streak fate⁸ (Supplementary Fig. 5d). Integration of this data with a recently published scRNA-seq dataset of a gastrulating mouse embryo confirmed that these cells correspond to mid/late primitive steak cells, which give rise to LV and RV CMs in the mice⁴ as opposed to the no-bud/early bud stage streak cells, which generate atrial CMs (Supplementary Fig. 5e). In line with this, the expression profile of RA receptors (RXRs) and co-receptors (RARs) was comparable between cells at day 1.5 and the mid/late primitive streak cells in the mouse (Supplementary Fig. 6a). Moreover, we observed high expression of the cytochrome member CYP26A1, which has been reported to mark a subpopulation of the cardiogenic mesoderm prone to giving rise to ventricular cardiomyocytes¹¹ (Supplementary Fig. 6b).

The cardiovascular progenitors derived in the absence of RA (CPC-RA−) separated into two main clusters based on their proliferative state (cluster 3, non-proliferative; cluster 13, proliferative; Fig. 3f; Supplementary Fig. 6c, d), both showing expression of aSHF genes such as ISL1, FGF10, and BMP4 (Supplementary Fig. 6c). Two small additional clusters were shared with the RA+ condition: cluster 12 was characterized by high expression of PDGFRA, indicating undifferentiated cardiovascular progenitor cells, and cluster 17 was enriched in genes related to smooth muscle cells (SMCs) such as TAGLN, NPPB, LOXL2, TMSB4X, and CALD1 (Fig. 3f–h).

In the presence of RA, the cardiovascular progenitor cells (CPC-RA+) formed several clusters (Fig. 3f). Clusters 0 and 1 showed expression of THBS4 as well as cardiac sarcomere protein genes such as MYL3, and MYL7 most likely resembling progenitors of the FHF that are in the process of converting to CMs (Fig. 3g). Clusters 4, 7, and 10 represented the proliferative equivalents of clusters 0, and 1 (Supplementary Fig. 6c, d). Notably, among the CPC-RA+ we observed two small additional clusters (cluster 14 and its proliferative counterpart cluster 18) that showed a gene expression profile (MAB21L2, HOXB6, HOXB5, BNC2) resembling that of the recently described JCF, which is a distinct subset of the FHF in mice⁶ (Fig. 3g). Cluster 9 was closely related to cluster 14 (Fig. 3i) and was characterized by expression of TBX5, as well as posterior transcripts such as WNT2 and HOXB1 while also expressing low levels of the JCF marker MAB21L2 (Fig. 3g).

The cluster identities were further evaluated by integrating the cells from the non-proliferative CPC clusters (clusters 0, 1, 3, 9, 12, 14, and 17) with the scRNA-seq datasets of Tyser et al.⁶ as well as Zhang et al.⁵ covering stages of mouse embryonic heart development that include JCF emergence⁶. This analysis confirmed the FHF- and aSHF-like characteristic of the above-mentioned clusters and showed that clusters 14 and 9 are closely related with the mouse JCF cluster me5 described by Tyser et al (Fig. 3i). Notably, these cells were also very similar to the cardiac progenitor clusters CP6 and CP7 described by Zhang et al., which are related to the late extraembryonic mesoderm in mice and have a comparable transcriptional signature with the JCF⁵. This suggests that cluster 9 as well as cluster 14 both represent JCF-like cells (Fig. 3i). Cluster 9 was mainly derived from TBX5⁺/NKX2.5⁻ CPCs (~75%; Fig. 3h), which is in line with findings in the mouse where the JCF arises from a TBX5⁺/NKX2.5⁻ progenitor population⁶. Cluster 14 was mainly derived from TBX5⁺/NKX2.5⁺ cells and expressed low levels of NKX2.5 (Supplementary Fig. 6e). This suggests that cluster 9 likely represents an early JCF population and cluster 14 a further committed JCF population that, due to the cardiogenic differentiation conditions, has started to express NKX2.5 and acquire a cardiomyogenic fate. Progenitors residing in the JCF in the mouse have the potential to contribute to the cardiomyocytic as well as the epicardial lineage⁶. Our data suggests that a similar population could exist during human cardiogenesis.

ITGA8 allows isolation of human JCF-like progenitor cells

To further explore the potential of this putative JCF cell population, we aimed at identifying a surface marker allowing the specific isolation of these cells. Comparative differential gene expression analysis between the mouse dataset of Tyser et al.⁶ and ours revealed TNC, AHNAK, and ITGA8 as promising candidates showing co-expression with other key JCF markers (MAB21L2, HOXB5, HOXB6, HAND1, BNC2; Supplementary Data 3)⁶. Among these, ITGA8 showed the most restricted expression pattern (Fig. 4a, b; Supplementary Fig. 7a, b). Flow cytometry analysis confirmed the presence of a small population of ITGA8⁺ cells from day 4 to day 6 (Fig. 4c, d). This population initially expressed only TBX5 (mCherry) but quickly upregulated NKX2.5 (eGFP) (Fig. 4e, f), which is in line with observations in the murine JCF⁶. Cells sorted based on the presence of ITGA8 showed significantly higher expression of JCF-specific markers compared to unsorted cells (MAB21L2, BNC2, HAND1; Fig. 4g; Supplementary Fig. 7c). After cultivating ITGA8⁺ sorted cells in myocardial differentiation conditions, we obtained cTnT⁺ CMs (Fig. 4h, i). On the other hand, when exposed to epicardial differentiation conditions, the cells formed a tight epithelial layer stained positive for the epithelial markers ZO1 and CK18 as well as the epicardial markers TCF21 and WT1 (Fig. 4h, j).

**Fig. 4: Human JCF-like progenitors expressing ITGA8 differentiate into epicardial and myocardial cells.**

To assess the fate potential of the ITGA8⁺ population at a clonal level, we took advantage of hiPSCs constitutively expressing eGFP and sorted ITGA8⁺ and ITGA8^- cells emerging at day 5 of differentiation in the presence of RA (0.5 μM, 4d). We plated these cells at very low density with equivalent, unsorted day 5 CPCs generated from un-labeled hiPSCs (~5 eGFP⁺ cells with 20,000 eGFP⁻ cells/well of a 12 well chamber slide) and differentiated them for 10 days in culture conditions that allow for simultaneous generation of epicardial cells and CMs (Fig. 4k). As compared to the eGFP⁺/ITGA8^- fraction, eGFP⁺/ITGA8⁺ cells showed a higher propensity to give rise to epicardial cells, while maintaining the ability for CM differentiation (Fig. 4l, m; Supplementary Data 4). While the vast majority of eGFP⁺/ITGA8⁺ clones differentiated into one lineage, around 10% of them gave rise to both epicardial cells and CMs (Fig. 4m, Supplementary Fig. 7d, Supplementary Data 4). This data provides further proof that a bipotent JCF progenitor pool giving rise to CMs and epicardial cells also exists during human cardiogenesis and can be enriched using ITGA8 as a membrane marker.

CPC-RA+ and RA− give rise to distinct cardiac cell populations

To characterize the descendants of the cardiovascular progenitors emerging in our system, cells differentiated with and without RA were sorted at day 4.5 based on eGFP (NKX2.5) and mCherry (TBX5) expression and further cultured as 3D aggregates to allow more physiological conditions for differentiation. Aggregates were then collected at day 30 for scRNA-seq (Fig. 5a). Cells from day 30 formed in UMAP three distinct groups: endoderm-derived cells, non-myocytic cardiac cells, and CMs (Fig. 5b, c; Supplementary Fig. 8a–c; Supplementary Data 5).

**Fig. 5: CPC-RA- and CPC-RA+ give rise to distinct cardiac cell types.**

The CMs separated into four main groups: ventricular-like CMs (clusters 0, 1, 2, 3, 4) OFT-like CMs (cluster 6), AVC-like CMs (cluster 13), and proliferating CMs (clusters 8, 10, 12) (Fig. 5d; Supplementary Fig. 8b, c).

The transcriptional profiles of ventricular-like CMs derived from the CPC-RA+ (clusters 0, 2, 3 and 4) and CPC-RA− (cluster 1) were overall quite similar, in line with data from human in vivo embryonic hearts showing that transcriptional differences between ventricular CMs located in the LV vs. RV are subtle^25,26 (Fig. 5d, e; Supplementary Data 6). Among the genes that show a preferential expression in either the LV or the RV in vivo²⁵, we observed higher levels for TBX5 and GJA1 in the LV-like clusters (0, 2, 3, 4) as compared to the RV-like cluster (1), while the expression levels for VCAN and GPRIN3 was higher in the RV-like cluster (1) as compared to the LV-like clusters (0, 2, 3, 4) (Supplementary Fig. 9a, b). Notably, the expression level of HAND1 was highest in the OFT-CM cluster 6 (Supplementary Fig. 9a). This is in line with reports in the literature showing that HAND1 is expressed not only in LV CMs, but also in OFT and AVC CMs^1,17. OFT-like CMs (cluster 6) were almost exclusively found as derivatives of CPC-RA− (Fig. 5e) and expressed markers of human conoventricular CMs, such as BMP2 and RSPO3¹⁸, as well as SMC markers (e.g., ACTA2, TAGLN, COL1A2) and only low expression levels of MYL2, according with the transcriptional profile of OFT-CMs in mice²⁷ (Fig. 5d; Supplementary Fig. 9c). Moreover, they did not express TBX5, which is consistent with data from human fetal tissue showing absence of TBX5 in OFT structures²⁸ (Supplementary Fig. 8c). We also identified a small cluster 14 which was enriched in CPNE5 and IGFBP5 transcripts, and could correspond to a transitional population of nodal CMs occurring during in vitro differentiation of hiPSC^29,30 (Fig. 5d, Supplementary Fig. 9c). This cluster was almost exclusively derived from the CPC-RA+ in line with the increased number of nodal cells upon RA exposure observed in the electrophysiology assay (Fig. 2d).

The identity of the OFT-like and ventricular-like CMs was confirmed on a transcriptome-wide level using enrichment analysis for gene signature scores derived from human and mouse in vivo heart samples^25,31 (Fig. 5f).

Subclustering of the non-myocytic cells, which had in common a high expression level of VIM (Fig. 5c, h), revealed the presence of endothelial/endocardial-like cells (APLN, CDH5), smooth muscle-like cells (TAGLN, ACTA2, CALD1), fibroblast-like cells (COL1A1, THY1, LUM), myofibroblasts (MYOZ2, MYL2), and valvular-like cells^32,33,34 (SOX9, POSTN; Fig. 5g–i; Supplementary Data 7). In addition, we detected cells expressing genes associated with endothelial to mesenchymal transition (EndoMT), such as ZEB2 and LUM (Fig. 5i). While endothelial/endocardial-like cells and smooth muscle-like cells were found as derivatives of all populations, valvular-like cells were almost exclusively derived from aSHF-like progenitors (CPC-RA−; Fig. 5j). This is in agreement with previous reports in the mouse showing that valve formation mainly relies on contribution from the aSHF^35,36.

Overall, these findings confirmed that the differences observed at the cardiovascular progenitor state translated into the formation of distinct progeny reflecting FHF-like (CPC-RA+) vs aSHF-like (CPC-RA−) fate potential. Notably, the clusters at day 30 of differentiation arising from the sorted populations (eGFP⁺, mCherry⁺, double mCherry⁺/eGFP⁺) of the CPC-RA+ (0, 2, 3, 4) where highly similar in their transcription profiles (Fig. 5e; Supplementary Data 5), suggesting that the different subpopulations of the CPC-RA+ sorted at day 4.5 (eGFP⁺, mCherry⁺, double mCherry⁺/eGFP⁺) do not represent distinct progenitor pools but are more likely reflecting different stages of progression on the same differentiation trajectory. In addition, we found that the differentiation efficiency into the myocytic lineages, in both RA− and RA+ conditions, was similar between sorted (RA+:~87–96% TNNT2⁺ CMs, ~4–12% FN1⁺ non-myocytic cell types; RA−: 84.7%; and 15.3%, respectively) and unsorted cells (RA+:~91–93% of cTnT⁺ CMs, ~5–10% of FN1⁺ non-myocytic cell types; RA−: ~83–92%; and ~13–17%, respectively), indicating that the negative (mCherry−/eGFP−) population present at day 4.5 does not seem to significantly impact the differentiation outcome (Supplementary Fig. 10a–c). Taken together, these findings suggest that the differentiation protocol described in this manuscript can be utilized independently of the fluorescent marking of the hESC-line, providing a substantial advantage over previously described methods to isolate heart-field specific populations³⁷.

Human CPCs contribute to distinct structures in mouse hearts

To further investigate the lineage commitment and functionality of in vitro derived cardiovascular progenitors, we injected CPC-RA+ or CPC-RA− into the heart region of developing mouse embryos at the cardiac crescent stage, where the FHF and SHF are located in close proximity³⁸ before initiation of visible contraction. These embryos were then subjected to whole-embryo ex vivo culture for 24 or 48 h (Fig. 6a). The embryos’ heart developed normally, including beginning of contraction, formation of the linear heart tube, its looping and onset of chamber formation (Fig. 6b–d). Since separation of the heart compartments in these early stages of heart development is still incomplete, we defined rather broadly delineated cardiac areas (OFT/RV, RV/LV, LV, IFT) to evaluate the regional contribution of the injected CPCs, as shown in Supplementary Movie 1.

**Fig. 6: Human CPC-RA+ and CPC-RA− integrate within distinct regions of the developing mouse embryo heart.**

The injected human cardiovascular progenitors, identified by human nuclear antigen (HNA) immunofluorescence staining, from both conditions (CPC-RA+ and CPC-RA−) successfully integrated into the host myocardium (Fig. 6e–g; Supplementary Fig. 11a; Supplementary Data 8). Immunofluorescence analysis in whole-mounted cleared embryos and in sectioned embryos showed that CPC-RA+ contributed mainly to the LV, with minor additions to the inflow-tract (IFT; Fig. 6f, g). No contribution to the OFT/RV region was observed, aside from one single cell in one out of eight embryos (Fig. 6f, g). Conversely, CPC-RA− contributed mainly to the OFT/RV, and we did not find any of these cells in the LV (Fig. 6f, g). Immunofluorescence staining for cTnT indicated that cardiovascular progenitor cells from both conditions differentiated into CMs, with sarcomere organization of early myofibrils similar to the host CMs (Fig. 6h, Supplementary Fig. 11b). Notably, the percentage of human non-myocytic cells (HNA⁺/cTnT⁻) within the cardiac region was significantly higher after injection of CPC-RA− as compared to CPC-RA+ (Fig. 6i; Supplementary Data 8), most likely reflecting the broader lineage potential of the CPC-RA− population.

This data not only confirms our in vitro findings on the identity of CPC-RA+ and CPC-RA− at day 4.5 but also highlights that the cells at this stage are already committed to contributing to their respective cardiac structures.

CPC-RA+ and RA− show disparate phenotypes in HLHS disease model

To evaluate the utility of our fate-committed cardiac progenitors in studying human disease, we employed them as an in vitro model of HLHS, which is characterized by underdevelopment of the LV and the LVOT and is one of the most severe congenital heart defects³⁹. We previously reported that HLHS patient-specific hiPSCs show defects in the differentiation of early cardiac progenitors resulting in abnormal cardiomyocyte subtype lineage specification and maturation⁴⁰. However, the differentiation protocol used in that study had not allowed the investigation of heart field-specific differences.

Differentiating two of the previously reported HLHS hiPSC lines (see Methods for details on line ID and genotype) into FHF-like (CPC-RA+) and aSHF-like progenitors (CPC-RA−), we observed, in both conditions, a downregulation of the key CPC marker NKX2.5 (Fig. 7a), which is in agreement with our previous findings⁴⁰. This was paralleled by a substantial reduction of TNNT2 levels in CPC-RA+, suggestive of impaired myocytic commitment, and a significant dysregulation of the aSHF-marker FGF10 in the CPC-RA−, together indicating that altered CPC-fate acquisition in HLHS seems to affect both, the FHF-like (CPC-RA+) as well as the aSHF-like (CPC-RA−) population (Fig. 7a). During further differentiation, HLHS patient-specific hiPSCs gave rise to CMs in both conditions as shown by immunofluorescence analysis of cTNT and qPCR for TNNT2, although at lower efficiency (in particular in the CPC-RA− conditions) as compared to the control (Fig. 7b, Supplementary Fig. 11c). In addition, these CMs showed much less, if any, expression of MLC2v, suggesting impaired attainment of ventricular identity (Fig. 7c). qPCR analysis confirmed a reduction of ventricular cardiomyocyte transcripts (MYH7, MYL2; Fig. 7d) and revealed increased MYL7/MYL2 and TNNI3/TNNI1 isoform ratio as indicator of cardiac immaturity (Fig. 7d), corroborating our previous findings on maturation defect in HLHS patient-specific CMs⁴⁰. Notably, HLHS CMs arising from the CPC-RA+ also displayed lower levels of the LV-specific marker FHL2 (Fig. 7d). At the same time, they exhibited higher expression of genes related to OFT-CMs (RSPO3) rendering their transcriptional profile similar to the patient-specific CMs arising from the CPC-RA−, indicative of a failed commitment to a LV CM fate (Supplementary Fig. 11c). We also observed dysregulation of OFT-CM transcripts (RSPO3) in the CMs derived from the CPC-RA−, which could likely be a consequence of the altered FGF10 level in the progenitors, since this gene is crucially required in the aSHF for proper OFT formation in mice^41,42,43. The fact that both progenitor pools (CPC-RA+ and CPC-RA−) and derived CMs appear affected is consistent with observations in vivo showing that CMs from the LV as well as the RV of HLHS patient share similar transcriptional defects^40,44.

**Fig. 7: Human CPC-RA+ and CPC-RA− generated from HLHS patients-derived hiPSCs show heart-field specific defects in vitro and ex vivo.**

To further evaluate the fate commitment of HLHS-derived CPCs, we injected the cells into the heart region of developing mouse embryos at the cardiac crescent stage and subjected those to whole-embryo ex vivo culture for 24 or 48 h. Immunofluorescence analysis indicated that HLHS-derived CPCs still contributed preferentially to their respective heart compartments (CPC-RA+ to the LV, CPC-RA− to the RV/OFT). However, their segregation to specific cardiac areas was less defined as compared to the control, with cells from both conditions being also present in other heart compartments (Fig. 7e–g; Supplementary Fig. 11d). In addition, we observed a significant reduction in the percentage of human HNA⁺/cTnT⁺ cells after injection of the CPC-RA+ as compared to the control (43% vs. 78%, p = 0.0087), suggesting impaired CM differentiation in line with the findings from the 2D culture (Figs. 6i, 7h; Supplementary Fig. 11e; Supplementary Data 8). Interestingly, this was not the case in the embryos injected with the HLHS-derived CPC-RA−, where we detected a similar proportion of human HNA⁺/cTnT⁺ cells as compared to the control (29% vs. 34%) (Figs. 6i, 7h; Supplementary Fig. 11e; Supplementary Data 8). As far as the limited number of embryos in each group allows, this suggests that the differentiation defect seen in the 2D system for this population can be, at least partially, compensated in the ex vivo setting in line with observations in patients where RV cardiomyocytes, although altered on the transcriptional level, form a heart chamber^40,44.

Taken together, these findings suggest that HLHS patient-specific hiPSCs fail to acquire a FHF-like fate at the progenitor level, resulting in perturbed CMs without proper LV CM identity. Notably, they also show a defect in aSHF progenitor differentiation resulting in CMs with significant dysregulation of ventricular and OFT CM markers, providing a potential link to the LVOT defects observed in HLHS patients. This highlights how heart-field specific progenitors can increase the resolution of in vitro disease models and provide additional insights in human congenital heart disease.

Discussion

Due to the limited accessibility of early human embryonic tissue samples, our knowledge on the emergence of human cardiovascular progenitors and their contributions to the developing heart relies largely on findings from model organisms and in vitro model systems. In the current study, we describe a versatile in vitro differentiation platform that allows the generation and isolation of various cardiac progenitor pools from hPSCs by modulating RA-mediated signaling. RA signaling is one of the key signaling pathways in the developing embryo and is critically required for the organization of the trunk and the organogenesis of various tissues of all three germ layers, including the hindbrain, the eye, the spinal cord, the pancreas, and the heart^45,46. It is first detected at late primitive streak stages in the paraxial mesoderm⁴⁷ and plays an important role in defining the posterior boundary of heart fields^46,48 as well as the proper specification of the pSHF that gives rise to the IFT¹⁰.

The latter has been exploited previously showing that addition of RA during in vitro differentiation of hPSCs towards CMs generates atrial CMs that are suitable for use in drug screening^11,13,14. In our study we did not observe an upregulation of atrial CM markers after exposure to low dosages of RA. Instead, the majority of cardiovascular progenitors found in our differentiation conditions had acquired a FHF-like fate rapidly giving rise to ventricular-like CMs. Upon exposure of higher dosage of RA for longer periods of time, we did detect a slight increase of atrial cardiomyocytes, however not comparable to protocols targeted at generating atrial CMs⁴⁹. Differently to our approach, the aforementioned studies have used an alternative cardiac differentiation protocol¹³ most likely resulting in distinct mesodermal progenitor subsets. There is an increasing body of evidence, that the commitment of mesodermal progenitors to form specific cardiac structures is already laid out during primitive streak formation^4,50. Mesodermal progenitors that leave the streak early (at mid/late stage)—to which the mesodermal progenitors generated with our protocol show resemblance— preferentially give rise to ventricular cardiomyocytes, while later cells (no-bud/early bud stage) generate atrial cardiomyocytes⁴. In addition, the dosage of RA applied in our protocol is lower, and it has been reported previously that low doses of RA during hPSC differentiation improve differentiation efficiency but do not affect cardiac subtype specification^8,12.

Taken together this shows that timing and dosage of RA as well as the type of mesodermal progenitors exposed results in distinct differentiation outcomes highlighting the pleiotropic roles of RA during development⁵¹.

In addition to the “classical” first and second heart field progenitors, we identified a subset of cells that resembled progenitors of the recently discovered JCF, which contribute to the epicardium as well as myocardium in mice^5,6. We uncovered that the surface protein ITGA8, which is expressed during development on epicardial cells in the mouse and the chicken^52,53, can be used to enrich these cells using live cell sorting. Sorted JCF-like progenitors readily differentiated into epicardial as well as cardiomyocytic progeny, proposing the presence of a similar progenitor population during human cardiogenesis. Remarkably, this population was only present upon exposure to RA, suggesting that RA-mediated signaling might be critically involved in the generation of the JCF. Whether the JCF-like cells arising in our differentiation conditions represent the entire JCF population or only a subset remains elusive. Further studies on the origin of this population, its relation to the extraembryonic tissue, its exact contributions to the developing heart as well as its heterogeneity are needed to shed further light on this aspect.

The toolbox presented in this study can be used to differentiate hPSCs into various mesodermal cardiac progenitor pools that contribute to heart formation, including aSHF, FHF, as well as the JCF, offering novel approaches for deciphering human cardiac development and for studying congenital heart disease. Combined with the current advances in self-organizing 3D cardiac tissue cultures^15,54,55,56, it could provide the opportunity to generate even more complex heart structures and gain unprecedented insights into human cardiac development and disease.

Methods

This study was approved by the Ethics Commission of the Technical University of Munich (TUM) Faculty of Medicine (# 447/17S).

hESC/hiPSC cell culture

Human iPSCs were generated using the CytoTune-iPS 2.9 Sendai Reprogramming Kit (Invitrogen; A16157) as previously described^40,57. The following hiPSC lines were used in differentiation experiments: hPSCreg MRIi003-A (hiPSC/CTRL), hPSCreg MRIi003-A-8 (hiPSC eGFP⁺), MRIi018-A (HLHS-1), DHMi003-A (HLHS-2). HLHS-1 carries de novo mutations in DENND5B, SYBU and BAI2, and HLHS-2 carries de novo mutations in MACF1, NFDUFB10, MYRF, AIM1L. Details on de novo calling, genetic background and clinical phenotype have been previously reported⁴⁰. Authorization to use the hESC line ES03 (ES03/CTRL) (hPSCreg ESIBIe003) generated by ES Cell International Pte Ltd in Singapore was granted by the Central Ethics Committee for Stem Cell Research of the Robert Koch Institute to AM (AZ 3.04.02/0131). Karyotype analysis was performed at the Institute of Human Genetics of the Technical University of Munich using standard methodology.

All hESC and hPSC lines were maintained in E8 medium containing 0.5% Penicillin/Streptomycin (15140-122, ThermoFisher Scientific) on Geltrex (A1413302, ThermoFisher Scientific) coated dishes under standard conditions (37 °C, 5% CO₂). Cells were non-enzymatically passaged every 4 days using 0.5 mM EDTA (AM9912, ThermoFisher Scientific) in PBS(−/−). To promote survival, Thiazovin (SML1045, Sigma-Aldrich) was added at a concentration of 2 μM for 24 h after passaging.

Generation of the reporter cell line by gene targeting

The method of gene targeting applied here was developed by the Stanley and Elefanty lab in Melbourne Australia^58,59. In brief, ES03 cells were grown on MEF feeders in KSR + bFGF (233-FB, R&D) and passaged enzymatically using trypsin-like enzyme (TrypLE, 12605010, ThermoFIsher Scientific) for at least three passages before electroporation. Ten million cells were suspended in a total of 800 µl of ice-cold PBS containing 20 µg of linearised targeting vector in a 4 mm gap cuvette. Electroporation was performed at 250 V and 500 µF. Cells were pipetted in pre-warmed medium and centrifuged at 259 × g for 3 min. The cell pellet was resuspended in warm KSR containing 12 ng/ml bFGF and spread over nine 6-cm dishes containing antibiotic resistant MEF feeders. 5 days after plating, medium was changed to KSR + FGF containing G418 (A1720, Sigma-Aldrich). About 7 days after the start of selection, colonies were picked and split in two wells, a DNA well and a maintenance well. Cells in the DNA well were lysed for at least 3 h at 55 °C (Lysis buffer: 100 mM TrisHCl, pH 8.0, 200 mM NaCl, 5 mM EDTA, pH 8.0, 0.2% (w/v) SDS powder (L4509, Sigma-Aldrich), 200 μg/ml proteinase K (P2308, Sigma-Aldrich) and DNA was isolated using isopropanol precipitation. The DNA pellet was washed once in 70% Ethanol and after air-drying resuspended in 50–100 µl TE buffer. Long-range PCR over the 5’ homology arm was used to screen for correct integration.

The resistance cassette can potentially interfere with expression of the reporter gene^60,61. Therefore, removing the resistance cassette ensures that reporting of locus activity of the targeted gene is faithful and specific. The design we used for the targeting vectors includes rox sites flanking the puromycin resistance cassette. These rox sites can be recombined with Dre, the D6 site-specific DNA recombinase⁶² similar to the Cre/loxP system. To avoid issues with genomic integration of the expression plasmid, we chose to transduce the DRE protein directly using a fragment of the HIV TAT protein, which has been previously used as a cell-penetrating peptide^63,64. The TAT-DRE fusion protein was provided by Dr. Marko Hyvönen from the Department of Biochemistry (University of Cambridge). Cells grown on feeders were incubated with 2 µM TAT-DRE/CRE in serum free medium for 6 h. After this treatment, the cells were placed back in KSR medium supplemented with bFGF. Removal of the resistance cassette was evaluated using primers flanking the resistance cassette. At the end of the process, cells were clonally derived by sorting single cells into 96 well plates in order to ensure a homogeneous population of targeted cells (and excision of the resistance cassette). Correct clones were identified by long range PCR on genomic DNA for targeting and the absence of the resistance cassette.

Cardiac differentiation

Cardiac differentiation protocol was adapted from Mendjan et al.⁸, Hofbauer et al.¹⁵. Briefly, cells were plated on Geltrex-coated 24-well plates at a density of 200,000 cells/well in E8 medium containing 2 μM Thiazovin (day −1). The next day, medium was replaced with CDM-BSA medium—consisting of 1:1 DMEM/F-12 with Glutamax (31331028, ThermoFisher Scientific) and IMDM (21980032, ThermoFisher Scientific) containing 0.1 g/ml BSA (A9647, Sigma-Aldrich), 15 μg/ml transferrin (10652202001, Roche), 1% chemically defined lipid concentrate (11905031, ThermoFisher Scientific), ~0.46 mM (0.004%) of thioglycerol (T1753, Sigma-Aldrich)—supplemented with 10 ng/ml BMP4 (314-BP, R&D), 1.5 μM CHIR99021 (4423, R&D), 50 ng/ml Activin A (SRP3003, Sigma-Aldrich), 30 ng/ml bFGF (233-FB, R&D) and 5 μM LY 294002 hydrochloride (1130, R&D). After 40 h, the medium was replaced with CDM-Meso medium consisting of CDM-BSA supplemented with 10 ng/ml BMP4, 8 ng/ml bFGF, 10 μg/ml insulin (11376497001, Roche/Sigma-Aldrich), 5 μM IWP2 (72122, Stem Cells). In addition, the medium was supplemented or not with RA (R2625, Sigma-Aldrich) in varying concentrations between 0.5 and 1 μM as indicated. CDM-Meso was changed the next 4 days in 24 h intervals. Depending on the treatment, RA was added to media for 2 or 4 days. On day 5 and 6, the medium was replaced with CDM-BSA supplemented with 10 ng/mL BMP4, 8 ng/ml bFGF, and 10 μg/ml insulin. From day 7 on, the medium was replaced every second day with CDM-Maintenance medium consisting of CDM-BSA supplemented with 10 μg/ml insulin.

FACS and flow cytometry analysis of live cells

For flow cytometry based sorting and analysis of live cells, cells between day 0 and day 8 of differentiation were washed two times with PBS(−/−), dissociated with Accutase (A11105013, ThermoFisher Scientific 5 min, 37 °C), centrifuged at 300 × g for 5 min, resuspended in 2% FCS in PBS, filtered through a 40 μm filter and subjected to sorting procedure on a FACS Aria III Cell Sorter (BD) or subjected to live flow cytometry analysis procedure on a Cytoflex S (Beckman-Coulter). DAPI (D3571, ThermoFisher Scientific) staining was used to discriminate dead and live cells (final concentration 0.01 ng/μl). Cells were sorted based on mCherry and eGFP expression and collected into 50–100% FCS in PBS, unless otherwise indicated. Analysis was performed using Kaluza software (v2.1) Beckman-Coulter).

3D culture of sorted cells

After sorting, mCherry⁺, eGFP⁺, double eGFP⁺/mCherry⁺, ITGA8⁺, eGFP⁺/ITGA8⁺, eGFP⁺/ITGA8⁻ cells were centrifuged, resuspended in CDM-BSA, counted, centrifuged again, and resuspended in CDM-Meso containing 0.5% Penicillin/Streptomycin, 10 μM of Rock Inhibitor Y- 27632 (688000, Calbiochem) and supplemented or not with RA in varying concentrations as indicated at a concentration of 100,000 cells per 200 μl per well of a U-shaped 96-well plate previously coated with 5% Poly(2-hydroxyethyl methacrylate (P3932, Sigma-Aldrich). Plates were centrifuged for 2 min at 300 × g and transferred to an incubator. From the next day on, the cardiac differentiation protocol was applied as described above. At day 7 aggregates were transferred into the wells of a 48-well plate previously coated with 5% Poly(2-hydroxyethyl methacrylate) and put on a shaker. From this point on CDM-Maintenance was supplemented with 50 ng/μl VEGF (293-VE, R&D).

Epicardial differentiation

For epicardial differentiation, day 4.5 FACS sorted ITGA8⁺ cells were re-plated on 0.1% gelatin-coated wells of 12-well chamber slides (81201, Ibidi) in density of 40,000 cells per 1 cm² and subjected to modified protocol Bao et al.⁶⁵. Briefly, cells were re-plated in LaSR medium consisting of Advanced DMEM/F12 containing Glutamax, Ascorbic Acid (0.1 mg/ml; A5960, Sigma) and 0.5% Penicillin/Streptomycin (LaSR) with addition of 1% FCS and 10 μM of Rock Inhibitor Y- 27632. On day 6 and day 7 the medium was replaced with LaSR supplemented with 3 μM CHIR99021. From day 8 onwards LaSR medium was replaced daily till day 12.

Epicardial–myocardial differentiation

For simultaneous epicardial and myocardial differentiation, day 4.5 eGFP^- cells were replated on Geltrex-coated wells of 12-well chamber slides (81201, Ibidi) and 96-well plates (89646, Ibidi) in density of 40 000 cells per 1 cm². Immediately, FACS-sorted eGFP⁺/ITGA8⁺ and eGFP⁺/ITGA8^- cells were added in the number of 5–10 cells per well resulting in 0–5 cells per well on the next day. Cells then were subjected to modified cardiac differentiation protocol described above. Modulations of BMP, RA and WNT signaling cues were described previously to favor either myocardial or epicardial lineages²⁹ and thus we decided to use low levels of WNT signaling inhibition with concomitant addition of BMP4 and RA. Briefly, cells were re-plated in CDM-BSA media supplemented with 20 ng/ml BMP4, 8 ng/ml bFGF, 10 μg/ml insulin, 2.5 μM IWP2, 1 μM RA with addition of 0.5% Penicillin/Streptomycin, 1% FCS, 10 μM of Rock Inhibitor Y- 27632. On day 6 and day 7 the medium was replaced with CDM-BSA supplemented with 20 ng/mL BMP4, 8 ng/ml bFGF, 10 μg/ml insulin, and 1 μM RA with addition of 0.5% Penicillin/Streptomycin. From day 8 onwards CDM-Maintenance with addition of 0.5% Penicillin/Streptomycin media was replaced daily till day 15.

Single-cell dissociation

Cells were dissociated using Accutase up to day 8. From day 8 on cells or aggregates were subjected to papain-based dissociation. Briefly, cells/aggregates were washed two times with 2 mM EDTA. Dissociation was carried out using a papain solution prepared as described by Fischer et al.⁶⁶. Cells/aggregates were incubated with papain solution for 20–40 min at 37 °C (the optimal dissociation time was previously determined for each cell line/time point to obtain a single cell suspension without compromising cell quality and survival). Aggregates were incubated on a shaker at 37 °C. After that time, a solution containing 1 mg/ml trypsin inhibitor (T9253, Sigma-Aldrich) was added. Cells/aggregates were dissociated by pipetting, transferred to a tube containing PBS(−/−), centrifuged at 300 × g for 5 min and resuspended in appropriate solutions depending on the downstream analysis.

Cardiomyocytes replating for immunofluorescence

At day 30 of cardiac differentiation, media was changed to EB2 media consisting of DMEM/F12 (21331-020, ThermoFisher Scientific), 2% FBS, 1% L-glutamine (25030-081, ThermoFisher Scientific), 1% non-essential aminoacids (11140-050, ThermoFisher Scientific), 0.5% Penicillin/Streptomycin and 0.1 mM beta-mercaptoethanol and changed every second day. At day 40 of cardiac differentiation, cells were subjected to papain-based dissociation as described above. Cells were replated on fibronectin (F1141, Sigma-Aldrich) coated wells of 12-well chamber slides in a density of 140,000–200,000 cells per 1 cm² in EB20 media consisting of DMEM/F12, 20% FBS, 1% L-glutamine, 1% non-essential aminoacids, 0.5% Penicillin/Streptomycin and 0.1 mM beta-mercaptoethanol. The next day media was changed for EB2 and changed daily till day 45.

RNA isolation, reverse transcription PCR (RT-PCR), and quantitative real-time PCR (qPCR)

For RNA isolation cells/aggregates were dissociated as described above. For time course analysis of day 0 to day 8 RNA collection was performed 2 h after media change. In case of aggregates, 3–5 aggregates were collected from each differentiation for RNA collection. Cell pellets were lysed and RNA was isolated using the Absolutely RNA Microprep Kit (400805, Agilent Technologies), Absolutely RNA Nanoprep Kit (400753, Agilent Technologies) or Rneasy Microkit (74004, Qiagen), depending on the cell number. For RNA isolation of ITGA8⁺ sorted cells, after sorting and centrifugation in PBS, cell pellets of 20,000 cells were lysed. 0.3–0.5 µg of RNA was used to synthesize cDNA with the High Capacity cDNA Reverse Transcription kit (4368813, Applied Biosystems). 40 ng was used to synthesize cDNA from ITGA8⁺ cells with the SuperScript IV Vilo Master Mix (11756050, ThermoFisher Scientific). Gene expression was quantified by qPCR using 1 µl cDNA, the Power SYBR Green PCR Master Mix (4367659, Applied Biosystems), the primers listed in Supplementary Data 9 and a 7500 Real-Time PCR System (Applied Biosystems). Gene expression levels were quantified relative to GAPDH expression using the ΔCt method, unless otherwise indicated.

Single-cell RNA sequencing (scRNA-seq)

For scRNA-seq at day 1.5, cells were washed two times with PBS, dissociated with Accutase (3–5 min, 37 °C), centrifuged at 300 × g for 5 min, resuspended in 0.04% BSA in PBS, filtered through a 40 μm filter and counted. For scRNA-seq at day 4.5 of differentiation, cells were subjected to sorting as described above. Cells were sorted based on mCherry and eGFP expression and collected into 10% FCS in PBS. Then cells were centrifuged, washed in 0.04% BSA in PBS, centrifuged again, filtered through a 40 μm filter and resuspended in 0.04% BSA in PBS for counting. For scRNA-seq at day 30, aggregates were dissociated with papain as described above, filtered through a 40 μm filter, centrifuged at 200 × g for 3 min and resuspended in 0.04% BSA in PBS for counting. After counting, 10,000 cells for each sample were processed using the Chromium Single Cell 3’ Library & Gel Bead Kit v3.1 (1000075, 10x Genomics), Chromium Single Cell B Chip Kit (1000073, 10x Genomics), and Chromium i7 Multiplex Kit (220103, 10x Genomics) to generate Gel Bead-In-EMulsions (GEMs) and single-cell sequencing libraries. Libraries were pooled and sequenced using the NextSeq 500/500 (Ilumina; High Output v2 kit 75 cycles v2.5 flow cell) with 28 cycles in read1 for the 10x barcodes and UMIs in and 8 cycles i7 index read and 58 cycles for cDNA in read2 with a read depth of at least 20,000 pair reads per cell. All day 4.5 or day 1.5 and day 30 samples subjected to scRNA-Seq procedure came from one differentiation (n = 1) and were processed at the same day together. Day 4.5 or day 1.5 and day 30 libraries were pooled and sequenced together.

The Cell Ranger pipeline (v6.1.1) was used to perform sample demultiplexing, barcode processing and generate the single-cell gene counting matrix. Briefly, samples were demultiplexed to produce a pair of FASTQ files for each sample. Reads containing sequence information were aligned using the reference provided with Cell Ranger (v6.1.1) based on the GRCh37 reference genome and ENSEMBL gene annotation. PCR duplicates were removed by matching the same UMI, 10x barcode and gene were collapsed to a single UMI count in the gene-barcode UMI count matrix. All the samples were aggregated using Cell Ranger with no normalization and treated as a single dataset. The R statistical programming language (v3.5.1) was used for further analysis. Count data matrix was read into R and used to construct Seurat object (v4.0.1)⁶⁷. The Seurat package was used to produce a diagnostic quality control plots and select thresholds for further filtering. Filtering method was used to detect outliers and high numbers of mitochondrial transcripts. These pre-processed data were then analyzed to identify variable genes, which were used to perform principal component analysis (PCA). Statistically significant PCs were selected by PC elbow plots and used for UMAP analysis. Clustering parameter resolution was set to 1 for the function FindClusters() in Seurat. For sub-clustering analysis we used clustree package (v0.4.3). The clustering analysis of the day 30 dataset resulted in three clusters that were characterized solely by very low read counts (5, 15, 17) and were removed from further analyses.

All DEGs were obtained using Wilcoxon rank sum test using as threshold p-value ≤ 0.05. We used adjusted p-value based on Bonferroni correction using all features in the dataset. For the cell type-specific analysis, single cells of each cell type were identified using FindConservedMarkers function as described within Seurat pipeline. For all the gene signatures analyzed, we used a function implemented in yaGST R package (v2017.08.25) (https://rdrr.io/github/miccec/yaGST/)⁶⁸.

To compare our dataset with published scRNA-sequencing datasets we identified common genes with datasets from humans²⁵ based on Homo sapiens gene symbols. For datasets from mouse^4,5,6 IDs were converted to Homo sapiens gene symbols according to the ortholog list from Ensemble (only one-to-one orthologues were considered⁶⁹). The datasets were combined using the CCA workflow implemented in the Seurat package (v4.1.1)⁶⁷ based on 10 dimensions and 2000 anchor features. After integration the Euclidean distances between the different clusters were calculated based on the averaged, corrected, normalized counts followed by hierarchical clustering using Ward’s hierarchical agglomerative clustering method⁷⁰. The results were plotted as a dendrogram.

For evaluating lowly expressed transcripts as in the case of RA receptors (RXRs) and co-receptors (RARs) missing values were imputed using adaptively thresholded Low-Rank Approximation (ALRA) with standard settings⁷¹. In case imputed values are displayed the figure panels are labeled accordingly.

Flow cytometry

Intracellularly/intranuclearly stained cells

For intracellular/intranuclear staining for flow cytometry cells were dissociated, counted, and distributed in equal numbers per sample in 15 ml tubes (typically 2 × 10⁶ cells per sample per tube). Cells were fixed with 4% PFA for 7 min at room temperature (RT; 500 µl/10⁶ cells), then centrifuged 3 min at RT at 400 × g. Next cells were washed three times with PBS (shaked for 5 min and centrifuged 5 min at 400 × g between each wash). Then cells were stored in 2% FCS in PBS+/+ at 4 °C or incubated right away with blocking/permeabilization buffer containing 10% FCS, 0.1% Triton-X-100 (Sigma, T8707), 0.1% saponin in PBS(+/+) (for intranuclear staining) or 10% FCS, 0.1% saponin in PBS+/+ (for intracellular/membrane staining) for 1 h at RT on a shaker (1 ml/10⁶ cells). Then cells were centrifuged and incubated with the primary antibody (antibodies and dilutions used are provided in Supplementary Data 10) diluted in 1% FCS, 0.1% saponin with or without 0.1% Triton-X-100 in PBS+/+(500 µl/10⁶ cells), overnight at 4 °C on a shaker. Then cells were washed three times with 0.1% saponin in PBS+/+ with or without 0.1% Triton-X-100 for a total of 45 min on a shaker (cells were centrifuged between washes). Then the secondary antibody (Supplementary Data 11) was added diluted 1:500 in 1% FCS, 0.1% saponin in PBS+/+ with or without 0.1% Triton-X-100, (500 µl/10⁶cells) and cells were incubated for 1 h at RT, protected from light, on a shaker. After that, cells were washed three times with 0.1% saponin in PBS+/+ with or without 0.1% Triton-X-100 for a total of 45 min on a shaker. Next, cells were resuspended in 2% FCS in PBS+/+ (100 µl/10⁶ cells), passed through a 40 µm strainer and subjected to analysis on a CytoFlex S or Gallios (Beckman Coulter). Analysis was performed using Kaluza software (v2.1) (Beckman Coulter). No-primary antibody, no-secondary antibody, IgG antibody controls were performed.

Surface protein-stained cells

For flow cytometry analysis of surface protein-stained cells, cells were dissociated, counted, and distributed in equal numbers per sample in 15 ml tubes (typically 5 × 10⁶ cells per sample per tube). Samples were incubated with ITGA8-Alexa-Fluor-647-conjugated (ITGA8-APC) or IgG1-APC-conjugated (IgG-APC) antibodies (Supplementary Data 10) diluted in the FACS buffer containing 2% FCS in PBS (10 µl antibody/100 µl buffer/10⁶ cells or 1:50, respectively (Supplementary Data 10) for 30 min on ice. Then cells were washed three times with FACS buffer and resuspended in FACS buffer for sorting. Before sorting DAPI was added to samples at a final concentration of 0.01 ng/µl to discriminate dead and live cells. Cells were sorted into tubes containing 1 ml of 50% FCS in PBS, washed with PBS, centrifuged, and resuspended in lysis buffer for RNA extraction or cell culture media for further differentiation.

Immunofluorescence staining of cells

Cells were fixed with 4% PFA for 15 min at RT, washed 3 times with PBS, blocked for 1 h in 0.1% Triton-X-100 PBS containing 3% BSA and stained using primary antibodies (antibodies and dilutions used are provided in Supplementary Data 10) diluted in 0.1% Triton-X-100 PBS containing 0.5% BSA overnight at 4 °C. Samples were then washed three times for 5 min with 0.1% Triton-X-100 PBS and incubated with secondary antibodies (Supplementary Data 11) diluted 1:500 in 0.1% Triton-X-100 PBS containing 0.5% BSA for 1 h at RT. Specimens then were washed three times with 0.1% Triton-X-100 PBS. Nuclei were detected with 5 μg/ml Hoechst 33258 (94403, Sigma-Aldrich) (5 min). Slides were washed once again with PBS, covered with mounting medium and a cover slip and stored at 4 °C.

Aggregates embedding, cryosectioning, and immunofluorescence staining

In preparation for cryosectioning and immunofluorescence staining aggregates were subjected to modified protocol of Lancaster & Knoblich⁷². Briefly, aggregates were transferred to wells of a 24-well plate containing 1 ml PBS+/+ using a cut 1000 μl pipette tip. PBS+/+ was replaced with 4% PFA, and aggregates were incubated for 30 min at room temperature. After that time, aggregates were washed three times for 5 min with PBS+/+. Then, 30% sucrose (Sigma-Aldrich, S9378) in PBS+/+ was added for overnight in 4 °C. For aggregates embedding, the 30% in PBS+/+ sucrose on aggregates was replaced with warmed up to 37 °C 1 ml 10% sucrose and 7.5% gelatin in PBS+/+ solution and allowed to equilibrate for 15 min at 37 °C on aggregates. Meanwhile, the bottom of a cryomold was covered with 400 μl sucrose/gelatin solution and placed at 4 °C to solidify. Next aggregates were transferred to molds using a cut 1000 μl tip, and placed for 3–5 min at 4 °C, then the cryomold was filled with 500 μl sucrose/gelatin solution and placed at 4 °C for 20 min to solidify. For aggregates freezing, such prepared molds were placed in cold 2-methylbutane (M32631, Sigma-Aldrich) for 1–2 min and stored at −80ÅãC. Molds were cryosected using Cryostat (Microm HM 560, ThermoFisher Scientific) into 12–16 μm slices on poly-L-lysine coated slides (J2800AMNZ, ThermoFisher Scientific), dried for 30 min at room temperature and stored at −80 °C. For immunofluorescence analysis, slides were dried for 30 min at room temperature after taking out of −80 °C freezer, fixed with 4 °C PFA for 10 min at room temperature, washed three times for 5 min with PBS+/+. Then permeabilized with 0.25% Triton X-100 for 15 min at room temperature, washed again three times for 5 min with PBS, blocked with 3% BSA in 0.05% Tween-20 (P2287, Sigma-Aldrich) in PBS+/+(PBST) for 1 h. Then primary antibodies (Supplementary Data 10) diluted in 0.5% BSA in PBST were added for overnight at 4 °C. Next day, slides were washed five times for 10 min with PBST. Then secondary antibodies (Supplementary Data 11) diluted in 0.5% BSA in PBST were added and slides were incubated for 2 h at room temperature. After that time, slides were washed five times for 10 min with PBST. Next, Hoechst 33258 at 5 μg/mL diluted in PBS was added for 15 min at room temperature. Slides were washed once with PBS, covered with mounting medium and cover slip and stored at 4 °C.

Mouse-human embryo chimeras

Timed mating and embryo dissection

All animal experiments were performed in accordance with German animal protection laws and EU ethical guidelines (Directive 2010/63/EU). For this study, we used a mouse line with C57/B6 background. C57BL/6 male (2–6 months) and female (2–3 months) mice were housed at 20–24 °C, 45–60% humidity, and a dark/light cycle of 12/12 h. A day after overnight mating, mice were separated and checked for the presence of a vaginal plug (ED0.5). On the desired day of embryonic development, the status of gestation was evaluated and pregnant females were sacrificed. Pregnant females sacrificed for embryo collection within the first two-thirds of gestation were reported to the inspection authority as organ collection and did not require any additional ethical approvement. Embryos were dissected out of the uterus and placed in a dish with pre-warmed dissection medium (5% FCS, 1% Pen/Strep, 20 mM HEPES in DMEM (ThermoFisher Scientific, 31966). Embryos were carefully removed from the decidua and Reichert’s membrane. Attention was paid to not remove the ectoplacental cone or destroy the yolk sack. Damaged embryos were not used further. Embryos used for further experiments were not disaggregated for sex.

Injection and whole embryo culture ex utero

Unsorted CPCs at day 4.5 of differentiation were dissociated using Accutase as described above, resuspended in the CDM-BSA media and injected into the heart region of dissected mouse embryos at the cardiac crescent stage. At this stage, structures within the mouse embryo are small and FHF and SHF regions appear in the close proximity enabling targeting the whole heart-forming region with a single injection. Approximately 20–50 cells per embryo were introduced via a glass capillary of 20 µm inner diameter. Operated and stage-matched unoperated embryos were placed into media containing 50% rat serum (S2150, Biowest) and 50% DMEM for 24 h. Medium was changed after 24 h to media containing 75% rat serum and 25% DMEM (adapted from Aguilera-Castrejon et al.⁷³, Pliszek et al.⁷⁴; similarly to other reports⁷⁵). Embryos were cultured using a rotating incubator system (BTC Engineering) at 5% CO₂ and 5–20% O₂ depending on their developmental stage (adapted from Aguilera-Castrejon et al.⁷³, Pliszek et al.⁷⁴; similarly to other reports⁷⁵). At the end of the culture, embryos were removed from the incubator, dissected from the yolk sac and amnion, and evaluated in terms of heart beat and overall morphological development.

Whole-mount immunofluorescence staining and clearing of embryos

After evaluation, properly developed embryos were transferred to tubes and fixed in 4% PFA at RT for 1–2 h depending on size. Next, embryos were washed in PBS, and incubated in blocking/permeabilization solution containing 10% FBS, 0.1% Triton X-100 in PBS on a shaker at RT for 4 h. Embryos were then incubated with primary antibodies (Supplementary Data 10) diluted in 1% FBS, 0.1% Triton X-100 in PBS on a shaker at 4 °C overnight; after which they were 3× washed in 0.1% Triton X-100 in PBS for 3 h in total; and incubated with secondary antibodies (Supplementary Data 11) and Hoechst dye diluted in 1% FBS, 0.1% Triton X-100 in PBS on a shaker at RT for 2 h. After another round of washing the embryos were cleared using a protocol adapted from Masselink et al⁷⁶. Briefly, embryos were sequentially transferred into dehydration solutions containing 30%, 50%, 70% and 2 × 100% 1-Propanol in PBS (pH adjusted to 9.0–9.5 using trimethylamine) and incubated on a shaker at 4 °C. Incubation times were adjusted considering the embryos’ size (at least 4 h per dehydration step). After complete dehydration, the embryos were stored in ethyl cinnamate.

Immunofluorescence staining of cryosections

Embryos were subjected to a sucrose gradient (5–20%) followed by embedding in a 1:1 mixture of Tissue-Tek O.C.T. (Sakura, 4583) and 20% sucrose (Sigma-Aldrich, S9378), and frozen in a bath of 2-methylbutane chilled with liquid nitrogen. Samples were stored at −80 °C or cryo-sectioned into 8 µm slices transferred to polysine-coated slides. Sections were fixed in 4% PFA at RT for 10 min. After washing with PBS, samples were incubated in solution containing 10% FBS, 0.1% Triton X-100 in PBS at RT for 1.5 h. Next, they were incubated with primary antibodies (Supplementary Data 10) diluted in 1% FBS, 0.1% Triton X-100 in PBS at 4 °C overnight. The next day, samples were washed 3x with 0.1% Triton X-100 in PBS for 15 min total and incubated with secondary antibodies (Supplementary Data 11) and 5 μg/ml Hoechst 33258 diluted in 1% FBS, 0.1% Triton X-100 in PBS at RT for 1 h. After another round of washing, sections were covered with mounting medium and a cover slip and stored at 4 °C.

Microscopy and image analysis

All images of stained differentiated cells were acquired using using confocal laser scanning microscopy (Leica Microsystems, SP8). Images were assigned with pseudocolours and processed with Leica Application Suite X software (v3.5.7.23225) from Leica Microsystems. Pictures were processed only in terms of brightness and contrast. Changes were applied equally across the entire image, and to all samples imaged. Cleared embryos were transferred into an 8 Well Glass Bottom µ-Slide (80827, ibidi) and whole embryos imaged using confocal laser scanning microscopy (Leica Microsystems, SP8). Using the 3D reconstruction and section view tools from Leica Application Suite X software (v3.5.7.23225) heart compartments were determined to be OFT/RV, RV/LV, LV, and IFT region and integration of human cells into the host myocardium was verified, respectively. Sections of embryos were imaged using a THUNDER system (Leica Mircrosystems, 11525679). Images were captured using the Leica K5 camera (Leica Mircrosystems, 11547112), a 40× objective (Leica Mircrosystems, 11506203) and the filter cube set (Leica Mircrosystems, 11525480). To quantify the contribution of human cells to the heart of injected embryos, HNA⁺ cells were counted and the region of integration of HNA⁺/cTnT⁺ cells was noted. OFT and IFT could be clearly distinguished by their morphology and its right/cranial or caudal position, respectively. OFT and RV at the analyzed stage of heart development are not as clearly separated therefore combined in the category OFT/RV. Human cells which were not clearly found in left or right ventricle were assigned to the category RV/LV. Images were analyzed with Image J (Version v1.53a).

Patch clamp electrophysiological recordings and data analysis

Cardiomyocyte monolayers were washed twice with PBS and dissociated to single cells by adding 0.25% Trypsin-EDTA (ThermoFisher Scientific; 25200072) and incubating at 37 °C for 2–3 min. If clumps of cells were still present, the cells were incubated for a further 2–3 min. The cells were then pipetted gently six times to ensure dissociation and FBS was added to neutralize the activity of trypsin. Dissociated cells were centrifuged at 130 × g for 3 min and resuspended in RPMI-1640 media (ThermoFisher Scientific; 11875093) supplemented with Penicillin/Streptomycin (ThermoFisher Scientific; 15140122), ITS-G (ThermoFisher Scientific; 41400045), chemically defined lipid concentrate (ThermoFisher Scientific, 11905031) and 1-Thioglycerol (M6145) (RI medium) containing 20% FBS. Cells were then seeded onto glass coverslips pre-coated with laminin (Sigma-Aldrich; L2020). Twenty-four hours later, cells were inspected for attachment and the medium was changed to RI medium. Cells were patch clamped 3–7 days after seeding (medium was changed every 3 days). Whole-cell current-clamp recording was carried out at room temperature (22 ± 2 °C) using an Axopatch 200B amplifier (Molecular Devices, USA). Pipette (Intracellular) solution contained (mM): 110 K-D-gluconate, 20 KCl, 10 NaCl, 2 EGTA, 10 HEPES, 1 MgCl₂, 0.3 GTP, and 2 MgATP (pH 7.4 with KOH). Extracellular solution (Tyrode’s) contained (mM): 135 NaCl, 5.4 KCl, 5 HEPES, 1 MgCl₂, 0.33 NaH₂PO₄, 2 CaCl₂ and 10 Glucose (pH 7.4 with NaOH). Pipette resistance, when filled with intracellular solution, was ~2–3.5 MΩ. Action potentials were Liquid Junction Potential (LJP) corrected (LJPc). The LJPc (13.4 mV) was calculated using the Clampex Junction Potential Calculator (Molecular Devices, USA). AP (Action Potential) properties were analyzed using Clampfit software (v10.4 and v10.7) (Molecular Devices, USA). Cardiomyocytes based on their action potentials were classified into 4 groups: Ventricular-like cardiomyocytes (V-CMs), Atrial-like cardiomyocytes (A-CMs); Nodal-like cardiomyocytes (N-CMs) and Intermediate (ventricular)-like cardiomyocytes (I-CMs) based on APD90/50 ratio. I-CMs = APD90/APD50 ratio between 1.4 and 1.8. V-CMs = APD90/APD50 ratio between 1.0 and 1.4. A-CMs = APD90/APD50 ratio > 1.8; N-CMs: were classified based on the typical shape of action potentials and signs of spontaneous depolarization.

Statistics

Statistical analysis was performed with GraphPad Prism version 5 and 8 (La Jolla California, USA). Bar graphs indicate the mean ± SEM with all data points displayed separately, unless otherwise indicated. Data from two experimental groups were compared either by unpaired Student’s t-test or by Mann-Whitney-Wilcoxon test depending on the assumed distribution. For more than two experimental groups one-way analysis of variance (ANOVA) was used first, followed by Student’s t-test or by Mann-Whitney-Wilcoxon for groupwise comparisons in case of a statistically significant result from the overall analysis. A p-value < 0.05 was considered statistically significant, unless otherwise indicated.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The scRNA-seq data generated in this study have been deposited in the GEO database under accession code GSE197660. The raw scRNA-seq data by Zhang et al.⁵ is available in the GEO database under accession code GSE176306. The processed data was downloaded from the University of California, Santa Cruz (UCSC) cell browser https://cells.ucsc.edu/?ds=chi-10x-mouse-cardiomyocytes. The raw scRNA-seq data by Tyser et al.⁶ is available in the European Nucleotide Archive database under study PRJEB14363, and ArrayExpress, under accession E-MTAB-7403. The processed data was downloaded from the Cancer Research UK Cambridge Institute https://content.cruk.cam.ac.uk/jmlab/mouseEmbryonicHeartAtlas/. The raw and processed scRNA-seq data by Ivanovitch et al.⁴ is available in the GEO database under accession code GSE153789. The raw and processed scRNA-seq data by Cui et al.²⁵ is available in the GEO database under accession code GSE106118. All the data generated in this study are either deposited in the above-mentioned repository or are provided in the Supplementary Information and Source Data file. Source data are provided with this paper.

References

Meilhac, S. M. & Buckingham, M. E. The deployment of cell lineages that form the mammalian heart. Nat. Rev. Cardiol. 15, 705–724 (2018).
Article PubMed Google Scholar
Devine, W. P., Wythe, J. D., George, M., Koshiba-Takeuchi, K. & Bruneau, B. G. Early patterning and specification of cardiac progenitors in gastrulating mesoderm. Elife 3, e03848 (2014).
Article PubMed PubMed Central Google Scholar
Lescroart, F. et al. Defining the earliest step of cardiovascular lineage segregation by single-cell RNA-seq. Science 359, 1177–1181 (2018).
Article ADS CAS PubMed PubMed Central Google Scholar
Ivanovitch, K. et al. Ventricular, atrial, and outflow tract heart progenitors arise from spatially and molecularly distinct regions of the primitive streak. PLoS Biol. 19, e3001200 (2021).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Q. et al. Unveiling complexity and multipotentiality of early heart fields. Circ. Res. 129, 474–487 (2021).
Article CAS PubMed PubMed Central Google Scholar
Tyser, R. C. V. et al. Characterization of a common progenitor pool of the epicardium and myocardium. Science 371, eabb2986 (2021).
Article CAS PubMed Google Scholar
Witman, N., Zhou, C., Grote Beverborg, N., Sahara, M. & Chien, K. R. Cardiac progenitors and paracrine mediators in cardiogenesis and heart regeneration. Semin. Cell Dev. Biol. 100, 29–51 (2020).
Article PubMed Google Scholar
Mendjan, S. et al. NANOG and CDX2 pattern distinct subtypes of human mesoderm during exit from pluripotency. Cell Stem Cell 15, 310–325 (2014).
Article CAS PubMed Google Scholar
Bertrand, N. et al. Hox genes define distinct progenitor sub-domains within the second heart field. Dev. Biol. https://doi.org/10.1016/j.ydbio.2011.02.029 (2011).
Article PubMed PubMed Central Google Scholar
Ryckebusch, L. et al. Retinoic acid deficiency alters second heart field formation. Proc. Natl Acad. Sci. USA 105, 2913–2918 (2008).
Article ADS CAS PubMed PubMed Central Google Scholar
Lee, J. H., Protze, S. I., Laksman, Z., Backx, P. H. & Keller, G. M. Human pluripotent stem cell-derived atrial and ventricular cardiomyocytes develop from distinct mesoderm populations. Cell Stem Cell 21, 179–194 (2017).
Article CAS PubMed Google Scholar
Wiesinger, A., Boink, G. J. J., Christoffels, V. M. & Devalla, H. D. Retinoic acid signaling in heart development: application in the differentiation of cardiovascular lineages from human pluripotent stem cells. Stem Cell Rep. 16, 2589–2606 (2021).
Article CAS Google Scholar
Devalla, H. D. et al. Atrial‐like cardiomyocytes from human pluripotent stem cells are a robust preclinical model for assessing atrial‐selective pharmacology. EMBO Mol. Med. 7, 394–410 (2015).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Q. et al. Direct differentiation of atrial and ventricular myocytes from human embryonic stem cells by alternating retinoid signals. Cell Res. 21, 579–587 (2011).
Article CAS PubMed Google Scholar
Hofbauer, P. et al. Cardioids reveal self-organizing principles of human cardiogenesis. Cell 184, 3299–3317.e22 (2021).
Article CAS PubMed Google Scholar
Andersen, P. et al. Precardiac organoids form two heart fields via Bmp/Wnt signaling. Nat. Commun. 9, 3140 (2018).
Article ADS PubMed PubMed Central Google Scholar
Soysa, T. Yde et al. Single-cell analysis of cardiogenesis reveals basis for organ level developmental defects. Nature 572, 120 (2019).
Article ADS PubMed PubMed Central Google Scholar
Sahara, M. et al. Population and single-cell analysis of human cardiogenesis reveals unique LGR5 ventricular progenitors in embryonic outflow tract. Dev. Cell 48, 475–490.e7 (2019).
Article CAS PubMed Google Scholar
High, F. A. et al. Murine Jagged1/Notch signaling in the second heart field orchestrates Fgf8 expression and tissue-tissue interactions during outflow tract development. J. Clin. Invest. 119, 1986–1996 (2009).
CAS PubMed PubMed Central Google Scholar
Tian, Y. et al. Characterization and in vivo pharmacological rescue of a Wnt2-Gata6 pathway required for cardiac inflow tract development. Dev. Cell 18, 275–287 (2010).
Article CAS PubMed PubMed Central Google Scholar
Roux, M. & Zaffran, S. Hox genes in cardiovascular development and diseases. J. Dev. Biol. 4, 14 (2016).
Article PubMed PubMed Central Google Scholar
Sirbu, I. O., Zhao, X. & Duester, G. Retinoic acid controls heart anteroposterior patterning by downregulating Isl1 through the Fgf8 pathway. Dev. Dyn. 237, 1627 (2008).
Article CAS PubMed PubMed Central Google Scholar
Litviňuková, M. et al. Cells of the adult human heart. Nature. https://doi.org/10.1038/s41586-020-2797-4 (2020).
Article PubMed PubMed Central Google Scholar
Christoffels, V. M., Smits, G. J., Kispert, A. & Moorman, A. F. M. Development of the pacemaker tissues of the heart. Circulation Res. 106, 240–254 (2010).
Article CAS PubMed Google Scholar
Cui, Y. et al. Single-cell transcriptome analysis maps the developmental track of the human heart. Cell Rep. 26, 1934–1950.e5 (2019).
Article CAS PubMed Google Scholar
Asp, M. et al. A spatiotemporal organ-wide gene expression and cell atlas of the developing human heart. Cell 179, 1647–1660.e19 (2019).
Article CAS PubMed Google Scholar
Liu, X. et al. Single-cell RNA-Seq of the developing cardiac outflow tract reveals convergent development of the vascular smooth muscle cells. Cell Rep. 28, 1346–1361.e4 (2019).
Article CAS PubMed Google Scholar
Hatcher, C. J., Goldstein, M. M., Mah, C. S., Susan Delia, C. & Basson, C. T. Identification and localization of TBX5 transcription factor during human cardiac morphogenesis. Dev. Dyn. 219, 90–95 (2000).
Article CAS PubMed Google Scholar
Wiesinger, A. et al. A single cell transcriptional roadmap of human pacemaker cell differentiation. Elife 11, e76781 (2022).
Article PubMed PubMed Central Google Scholar
Goodyer, W. R. et al. Transcriptomic profiling of the developing cardiac conduction system at single-cell resolution. Circ. Res. 125, 379–397 (2019).
Article CAS PubMed PubMed Central Google Scholar
Li, G. et al. Transcriptomic profiling maps anatomically patterned subpopulations among single embryonic cardiac cells. Dev. Cell 39, 491–507 (2016).
Article CAS PubMed PubMed Central Google Scholar
Mikryukov, A. A. et al. BMP10 signaling promotes the development of endocardial cells from human pluripotent stem cell-derived cardiovascular progenitors article BMP10 signaling promotes the development of endocardial cells from human pluripotent stem cell-derived cardio. Cell Stem Cell 28, 1–16 (2021).
Article Google Scholar
Pucéat, M. Embryological origin of the endocardium and derived valve progenitor cells: from developmental biology to stem cell-based valve repair. Biochim. Biophys. Acta - Mol. Cell Res. 1833, 917–922 (2013).
Article Google Scholar
Neri, T. et al. Human pre-valvular endocardial cells derived from pluripotent stem cells recapitulate cardiac pathophysiological valvulogenesis. Nat. Commun. 10, 1–14 (2019).
Article CAS Google Scholar
Verzi, M. P., McCulley, D. J., De Val, S., Dodou, E. & Black, B. L. The right ventricle, outflow tract, and ventricular septum comprise a restricted expression domain within the secondary/anterior heart field. Dev. Biol. 287, 134–145 (2005).
Article CAS PubMed Google Scholar
Crucean, A. et al. Re-evaluation of hypoplastic left heart syndrome from a developmental and morphological perspective. Orphanet J. Rare Dis. 12, 138 (2017).
Article CAS PubMed PubMed Central Google Scholar
Zhang, J. Z. et al. A human iPSC double-reporter system enables purification of cardiac lineage subpopulations with distinct function and drug response profiles. Cell Stem Cell 24, 802–811.e5 (2019).
Brade, T., Pane, L. S., Moretti, A., Chien, K. R. & Laugwitz, K. L. Embryonic heart progenitors and cardiogenesis. Cold Spring Harb. Perspect. Med. 3, a013847 (2013).
Article PubMed PubMed Central Google Scholar
Anderson, R. H., Spicer, D. E. & Crucean, A. Clarification of the definition of hypoplastic left heart syndrome. Nat. Rev. Cardiol. 18, 147–148 (2021).
Article PubMed Google Scholar
Krane, M. et al. Sequential defects in cardiac lineage commitment and maturation cause hypoplastic left heart syndrome. Circulation 144, 1409–1428 (2021).
Article PubMed PubMed Central Google Scholar
Cai, C.-L. et al. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889 (2003).
Article CAS PubMed PubMed Central Google Scholar
Sun, K. et al. Mutations in fibroblast growth factor (FGF8) and FGF10 identified in patients with conotruncal defects. J. Transl. Med. 18, 1–13 (2020).
Google Scholar
Watanabe, Y. et al. Role of mesodermal FGF8 and FGF10 overlaps in the development of the arterial pole of the heart and pharyngeal arch arteries. Circ. Res. 106, 495–503 (2010).
Article CAS PubMed Google Scholar
Liu, X. et al. The complex genetics of hypoplastic left heart syndrome. Nat. Genet. 49, 1152–1159 (2017).
Article CAS PubMed PubMed Central Google Scholar
Duester, G. Retinoic acid synthesis and signaling during early organogenesis. Cell 134, 921–931 (2008).
Article CAS PubMed PubMed Central Google Scholar
Ghyselinck, N. B. & Duester, G. Retinoic acid signaling pathways. Development 146, dev167502 (2019).
Article CAS PubMed PubMed Central Google Scholar
Rossant, J., Zirngibl, R., Cado, D., Shago, M. & Giguere, V. Expression of a retinoic acid response element-hsplacZ transgene defines specific domains of transcriptional activity during mouse embryogenesis. Genes Dev. 5, 1333–1344 (1991).
Article CAS PubMed Google Scholar
Keegan, B. R., Feldman, J. L., Begemann, G., Ingham, P. W. & Yelon, D. Retinoic acid signaling restricts the cardiac progenitor pool. Science 307, 247–249 (2005).
Article ADS CAS PubMed Google Scholar
Devalla, H. D. et al. Atrial-like cardiomyocytes from human pluripotent stem cells are a robust preclinical model for assessing atrial-selective pharmacology. EMBO Mol. Med. 7, 394–410 (2015).
Article CAS PubMed PubMed Central Google Scholar
Yang, D. et al. Modeling human multi-lineage heart field development with pluripotent stem cells. Cell Stem Cell 29, 1382–1401.e8 (2022).
Article CAS PubMed Google Scholar
Niederreither, K. & Dollé, P. Retinoic acid in development: towards an integrated view. Nat. Rev. Genet. 9, 541–553 (2008).
Article CAS PubMed Google Scholar
Ogawa, T., Li, Y., Lua, I., Hartner, A. & Asahina, K. Isolation of a unique hepatic stellate cell population expressing integrin α8 from embryonic mouse livers. Dev. Dyn. 247, 867–881 (2018).
Article CAS PubMed PubMed Central Google Scholar
So, H. P., Dokic, D. & Dettman, R. W. Communication between integrin receptors facilitates epicardial cell adhesion and matrix organization. Dev. Dyn. 237, 962–978 (2008).
Article Google Scholar
Lewis-Israeli, Y. R. et al. Self-assembling human heart organoids for the modeling of cardiac development and congenital heart disease. Nat. Commun. 12, 5142 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Drakhlis, L. et al. Human heart-forming organoids recapitulate early heart and foregut development. Nat. Biotechnol. 39, 737–746 (2021).
Article CAS PubMed PubMed Central Google Scholar
Silva, A. C. et al. Co-emergence of cardiac and gut tissues promotes cardiomyocyte maturation within human iPSC-derived organoids. Cell Stem Cell 28, 2137–2152.e6 (2021).
Article CAS PubMed Google Scholar
Gramlich, M. et al. Antisense-mediated exon skipping: a therapeutic strategy for titin-based dilated cardiomyopathy. EMBO Mol. Med. 7, 562–576 (2015).
Article CAS PubMed PubMed Central Google Scholar
Costa, M. et al. A method for genetic modification of human embryonic stem cells using electroporation. Nat. Protoc. 2, 792–796 (2007).
Article CAS PubMed Google Scholar
Davis, R. P. et al. Generation of human embryonic stem cell reporter knock-in lines by homologous recombination. Curr. Protoc. Stem Cell Biol. Chapter 5, Unit 5B 1 1 1-34 (2009).
Pham, C. T. N., Macivor, D. M., Hug, B. A., Heusel, J. W. & Ley, T. J. Long-range disruption of gene expression by a selectable marker cassette. Proc. Natl Acad. Sci. USA 93, 13090–13095 (1996).
Article ADS CAS PubMed PubMed Central Google Scholar
Scacheri, P. C. et al. Bidirectional transcriptional activity of PGK-neomycin and unexpected embryonic lethality in heterozygote chimeric knockout mice. Genesis 30, 259–263 (2001).
Article CAS PubMed Google Scholar
Sauer, B. & McDermott, J. DNA recombination with a heterospecific Cre homolog identified from comparison of the pac-c1 regions of P1-related phages. Nucleic Acids Res. 32, 6086–6095 (2004).
Article CAS PubMed PubMed Central Google Scholar
Frankel, A. D. & Pabo, C. O. Cellular uptake of the tat protein from human immunodeficiency virus. Cell 55, 1189–1193 (1988).
Article CAS PubMed Google Scholar
Fawell, S. et al. Tat-mediated delivery of heterologous proteins into cells. Proc. Natl Acad. Sci. USA 91, 664–668 (1994).
Article ADS CAS PubMed PubMed Central Google Scholar
Bao, X. et al. Directed differentiation and long-term maintenance of epicardial cells derived from human pluripotent stem cells under fully defined conditions. Nat. Protoc. 12, 1890–1900 (2017).
Article CAS PubMed PubMed Central Google Scholar
Fischer, B. et al. A complete workflow for the differentiation and the dissociation of hiPSC-derived cardiospheres. Stem Cell Res. 32, 65–72 (2018).
Article CAS PubMed Google Scholar
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587.e29 (2021).
Article CAS PubMed PubMed Central Google Scholar
Pagnotta, S. M. Competitive gene set and regulon tests. R package version 2017.08.25. (2017).
Cunningham, F. et al. Ensembl 2022. Nucleic Acids Res. 50, D988–D995 (2022).
Article CAS PubMed Google Scholar
Murtagh, F. & Legendre, P. Ward’s hierarchical agglomerative clustering method: which algorithms implement ward’s criterion? J. Classif. 31, 274–295 (2014).
Article MathSciNet MATH Google Scholar
Linderman, G. C. et al. Zero-preserving imputation of single-cell RNA-seq data. Nat. Commun. 13, 1–11 (2022).
Article Google Scholar
Lancaster, M. & Knoblich, J. Generation of cerebral organoids from human pluripotent stem cells. Nat. Protoc. 9, 2329–2340 (2014).
Article CAS PubMed PubMed Central Google Scholar
Aguilera-Castrejon, A. et al. Ex utero mouse embryogenesis from pre-gastrulation to late organogenesis. Nature 593, 119–124 (2021).
Article ADS CAS PubMed Google Scholar
Piliszek, A., Kwon, G. S. & Hadjantonakis, A. K. Ex utero culture and live imaging of mouse embryos. Methods Mol. Biol. 770, 243–257 (2011).
Article CAS PubMed PubMed Central Google Scholar
Mascetti, V. L. & Pedersen, R. A. Human-mouse chimerism validates human stem cell pluripotency. Cell Stem Cell 18, 67–72 (2016).
Article CAS PubMed PubMed Central Google Scholar
Masselink, W. et al. Broad applicability of a streamlined ethyl cinnamate-based clearing procedure. Dev 146, dev166884 (2019).
Article Google Scholar

Download references

Acknowledgements

We would like to acknowledge Birgit Campbell, Christina Scherb, and Marco Crovella for their technical assistance, Gabrielle Lederer (Cytogenetic Department, TUM) for karyotyping, Dr. Rupert Öllinger (TUM, Germany) for sequencing, Dr. David Elliott for sharing the ES03 and ES03-NKX2.5^eGFP cell lines, Drs. Ed Stanley and Andrew Elefanty (MCRI, Australia) for advice in construct design and gene targeting, and Dr. Sasha Mendjan for advice and discussion. This work was supported by the European Research Council (ERC) (grant 788381 to A.Mo. and grant 261053 to K-.L.L.), the Else-Kroener-Fresenius Stiftung (EKFS, to A.G.), the German Research Foundation (grant GO3220/1-1 to A.G.; Transregio Research Unit 152 to A.Mo. and K-.L.L.; Transregio Research Unit 267 to A.Mo., K-.L.L., and P.G.), the German Centre for Cardiovascular Research (DZHK) (grant FKZ 81Z0600601 to A.Mo. and K-.L.L.; grant 81X3600607 to J.K.), the Fondazione Umberto Veronesi (to G.S.).

Funding

Open Access funding enabled and organized by Projekt DEAL.

Author information

These authors contributed equally: Dorota Zawada, Jessica Kornherr, Anna B. Meier, Gianluca Santamaria.

Authors and Affiliations

First Department of Medicine, Cardiology, Klinikum rechts der Isar, Technical University of Munich, School of Medicine and Health, Munich, Germany
Dorota Zawada, Jessica Kornherr, Anna B. Meier, Gianluca Santamaria, Tatjana Dorn, Monika Nowak-Imialek, Fangfang Zhang, Mark Lachmann, Maria Teresa De Angelis, Karl-Ludwig Laugwitz, Alessandra Moretti & Alexander Goedel
German Center for Cardiovascular Research (DZHK), Munich Heart Alliance, Munich, Germany
Dorota Zawada, Jessica Kornherr, Anna B. Meier, Gianluca Santamaria, Tatjana Dorn, Monika Nowak-Imialek, Fangfang Zhang, Mark Lachmann, Maria Teresa De Angelis, Karl-Ludwig Laugwitz & Alessandra Moretti
Regenerative Medicine in Cardiovascular Diseases, First Department of Medicine, Klinikum rechts der Isar, Technical University of Munich, School of Medicine and Health, Munich, Germany
Dorota Zawada, Jessica Kornherr, Anna B. Meier, Gianluca Santamaria, Tatjana Dorn, Monika Nowak-Imialek, Fangfang Zhang & Alessandra Moretti
Department of Experimental and Clinical Medicine, University “Magna Graecia”, Catanzaro, Italy
Gianluca Santamaria & Maria Teresa De Angelis
Department of Surgery, University of Cambridge, Cambridge, UK
Daniel Ortmann & Mariaestela Ortiz
Wellcome-MRC Cambridge Stem Cell Institute, Jeffrey Cheah Biomedical Centre, University of Cambridge, Cambridge, UK
Daniel Ortmann
German Heart Center Munich, Department of Cardiovascular Surgery, Institute Insure - Technical University of Munich, School of Medicine and Health, Munich, Germany
Martina Dreßen
Bristol Heart Institute, Bristol Medical School, Translational Health Sciences, Bristol, UK
Victoria L. Mascetti
School of Physiology, Pharmacology and Neuroscience, Faculty of Life Sciences, University of Bristol, Bristol, UK
Stephen C. Harmer
Clinical Pharmacology & Precision Medicine, William Harvey Research Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, UK
Muriel Nobles & Andrew Tinker
Department of Obstetrics and Gynecology, Stanford School of Medicine, Stanford University, Stanford, USA
Roger A. Pedersen
Georg-Speyer-Haus, Institute for Tumor Biology and Experimental Therapy, Frankfurt am Main, Germany
Phillip Grote
Institute of Cardiovascular Regeneration, Centre for Molecular Medicine, Goethe University, Frankfurt am Main, Germany
Phillip Grote
Department of Surgery, Yale University School of Medicine, New Haven, USA
Alessandra Moretti
Department of Cell and Molecular Biology, Karolinska Institute, Stockholm, Sweden
Alexander Goedel

Authors

Dorota Zawada
View author publications
You can also search for this author in PubMed Google Scholar
Jessica Kornherr
View author publications
You can also search for this author in PubMed Google Scholar
Anna B. Meier
View author publications
You can also search for this author in PubMed Google Scholar
Gianluca Santamaria
View author publications
You can also search for this author in PubMed Google Scholar
Tatjana Dorn
View author publications
You can also search for this author in PubMed Google Scholar
Monika Nowak-Imialek
View author publications
You can also search for this author in PubMed Google Scholar
Daniel Ortmann
View author publications
You can also search for this author in PubMed Google Scholar
Fangfang Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Mark Lachmann
View author publications
You can also search for this author in PubMed Google Scholar
Martina Dreßen
View author publications
You can also search for this author in PubMed Google Scholar
Mariaestela Ortiz
View author publications
You can also search for this author in PubMed Google Scholar
Victoria L. Mascetti
View author publications
You can also search for this author in PubMed Google Scholar
Stephen C. Harmer
View author publications
You can also search for this author in PubMed Google Scholar
Muriel Nobles
View author publications
You can also search for this author in PubMed Google Scholar
Andrew Tinker
View author publications
You can also search for this author in PubMed Google Scholar
Maria Teresa De Angelis
View author publications
You can also search for this author in PubMed Google Scholar
Roger A. Pedersen
View author publications
You can also search for this author in PubMed Google Scholar
Phillip Grote
View author publications
You can also search for this author in PubMed Google Scholar
Karl-Ludwig Laugwitz
View author publications
You can also search for this author in PubMed Google Scholar
Alessandra Moretti
View author publications
You can also search for this author in PubMed Google Scholar
Alexander Goedel
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

D.Z., A.G., A.B.M., and A.Mo. conceived the study, interpreted the data, and wrote the manuscript. D.Z. designed and performed most of the experiments using hPSCs, including maintenance and differentiation of hPSCs, cell sorting, molecular assays, immunostainings, single cell libraries preparation, and analyzed data. G.S. and A.G. performed bioinformatic analyses. J.K. performed mouse embryo injections, ex vivo culture, and embryo analysis. D.O. and M.O. generated the ES03-TN line and provided conceptual advice. M.N-.I. supported maintenance and differentiation of hPSCs and preparation of cells for downstream assays. M.L. performed some molecular assays and immunostainings. V.L.M. supported further development of differentiation protocol. T.D. supported analysis and data interpretation, and provided conceptual advice. S.C.H., M.N., A.T. performed patch clamp experiments. F.Z. generated hiPSC eGFP⁺ line. M.D. generated HLHS-2 line. M.T.D.A. supported some data interpretation. R.A.P. provided conceptual advice. P.G. provided conceptual advice and financial support. A.Mo., A.G., and K-.L.L. conceived and supervised the study and provided financial support. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Karl-Ludwig Laugwitz, Alessandra Moretti or Alexander Goedel.

Ethics declarations

Competing interests

D.O. is currently an employee at Bit Bio Ltd. (United Kingdom) and holds stock options. R.A.P. is an advisor at Meatable NV (Netherlands) and holds stock options, is co-founder of DefiniGen Ltd. (United Kingdom) and holds shares, and is an advisor at Bit Bio Ltd. (United Kingdom) and holds stock options. A.G. receives consultancy fees from SmartCella (Sweden). The remaining authors declare that they have no competing interests.

Peer review

Peer review information

Nature Communications thanks Stephane Zaffran and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Supplementary Data 7

Supplementary Data 8

Supplementary Data 9

Supplementary Data 10

Supplementary Data 11

Supplementary Movie 1

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as 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 images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Zawada, D., Kornherr, J., Meier, A.B. et al. Retinoic acid signaling modulation guides in vitro specification of human heart field-specific progenitor pools. Nat Commun 14, 1722 (2023). https://doi.org/10.1038/s41467-023-36764-x

Download citation

Received: 02 June 2022
Accepted: 15 February 2023
Published: 03 April 2023
DOI: https://doi.org/10.1038/s41467-023-36764-x

This article is cited by

HAND factors regulate cardiac lineage commitment and differentiation from human pluripotent stem cells
- Huixin Guo
- Chengwen Hang
- Yi-Han Chen
Stem Cell Research & Therapy (2024)
Engineered platforms for mimicking cardiac development and drug screening
- Madison Stiefbold
- Haokang Zhang
- Leo Q. Wan
Cellular and Molecular Life Sciences (2024)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.