During cardiac progenitor formation, early cardiac transcription factors need to be present at the right time and the right dose, therefore they often exhibit dynamic gene expression changes with tight expression windows. To comprehensively investigate the role of RYBP in progenitor formation and to visualize this rapid change, we performed in vitro cardiac differentiation and analyzed the gene expression of the wild wt and Rybp ^−/− colonies (Supplementary Figure S1A) on a day-to-day basis during progenitor formation. For in vitro cardiac differentiations, we applied the hanging drop (HD) method (Figure 1A) which doesn’t involve any inducing factors that can influence cardiac lineage commitment. It allows the generation of a variety of progenitor and mature cell types of the heart including cardiac endothelial cells and cardiomyocytes (CMCs). We checked the expression of key cardiac marker genes from day 2, when the differentiation started, till day 8, which represented the late stage of progenitor formation (materials and methods).

First, we analyzed the gene expression kinetics of Mesoderm posterior 1 (Mesp1), which was previously described as the earliest cardiac transcription factor and as an essential gene for cardiac mesoderm formation (Saga et al., 1996; Saga et al., 2000). In wt cardiac cultures rapid increase in Mesp1 expression is expected as early as day 4 or day 5, and need to be downregulated later on (d7, d8). As expected, the wt cultures Mesp1 expression was increased by day 4, sustained this high expression level only for a short period till day 5, and decayed before the late progenitor stage (Figure 1B). In the Rybp ^−/− cardiac colonies, Mesp1 expression showed a significant decrease and exhibited slightly delayed kinetics compared to the wt. Next, we checked the gene expression of cardiac progenitor marker genes, which expression was expected to gradually increase throughout cardiac differentiation. Our result showed that in wt colonies, the mRNA level of late cardiac progenitor genes such as Myocyte-specific enhancer factor 2C (Mef2c), Heart- and neural crest derivatives-expressed protein 1 (Hand1), Insulin Gene Enhancer Protein ISL-1 (Isl-1) and Homeobox Protein Nkx-2.5 (Nkx2-5) increased throughout the differentiation. In the mutant cultures, we could see altered gene expression in the mRNA level of Hand1 as well, which was low in all examined time points. The level of Mef2c and Nkx2-5 mRNAs were similar in both genotypes at early time points, however from day 5/day 6, their expression in the Rybp ^−/− colonies couldn’t reach the level of the wild type. The expression level of Isl-1 was diminished only on the 6th and 7th days of cardiac differentiation in comparison to the wild type counterparts, and in the rest of the timepoints were similar (d2, d8) or higher (d3, d4, d5, Figure 1B).

To provide additional evidence, we performed western blot analysis using whole cell lysates derived from in vitro cardiac differentiation of wt and Rybp ^−/− colonies and checked the protein kinetics of the early cardiac transcription factor MESP1 and the late cardiac transcription factor ISL-1. MESP1 protein level was examined in early progenitor stage (d2, d3, d4, d5, d6, d7) while ISL-1 protein level was analyzed in early (d2, d3, d4, d5, d6) and late time points (d7, d8, d14) of in vitro cardiac differentiation. Both proteins showed decreased levels and delayed kinetics in the absence of RYBP, which was consistent with the mRNA expression levels (Supplementary Figures S1B–E).

Besides cardiomyocytes, the cardiac endothelial cells also play an important role in heart morphogenesis and directly affect the performance of the adjacent cardiomyocytes (Brutsaert, 2003). This led us to question whether the endothelial lineage commitment is also compromised in the mutant. Therefore, we monitored the expression of endothelial lineage marker genes at early (d2, d4, d6) and late (d8, d10, d14) time points of cardiac differentiation. Our results showed that the expression of early endothelial receptor tyrosine kinase Flk-1 (Kdr, Vegfr), which is essential for the differentiation of the endothelial cells, was decreased in Rybp ^−/− cardiac colonies during progenitor formation compared to the wt (Figure 1C). The expression of late endothelial markers such as Platelet endothelial cell adhesion molecule 1 (Pecam1, CD31) was also highly downregulated in the mutant cardiac cultures (Figure 1C). In addition, mutant samples exhibited reduced levels of PECAM1 positive cells in immunostained cardiac colonies (Figures 1D, E).

Taken together our data showed that the mRNA levels of key cardiac progenitor markers genes were significantly reduced in the lack of Rybp. Our results also showed that besides cardiomyocytes, the differentiation of cardiac endothelium, another mesoderm derived lineage, is also affected in the Rybp ^−/− cardiac colonies. In addition, the mRNA level of the earliest cardiac transcription factor Mesp1 and the early endothelial marker gene Kdr was highly decreased, whose expression was expected immediately after the onset of the expression of the earliest pan-mesoderm gene Brachyury. These all suggested that RYBP may have a role prior to progenitor formation.

To investigate if RYBP can have a potential role in mesoderm formation, first, we compared wild type and Rybp ^−/− whole-genome transcriptome derived from stem cells (d0), late cardiac progenitors (d8) and matured cardiomyocytes (d14) (Ujhelly et al., 2015). Although, in the lack of RYBP, transcriptomic data showed a slight reduction of mesoderm- (Brachyury, Eomes) and endoderm-specific (Gata4, Gata6, Cxcr4) gene expression, the observed changes in the late progenitor stage were not statistically significant (Figure 2A). In wt cultures during progenitor formation the expression of all three major germ layer marks gene are expected to increase and eventually decrease in later stages. To further analyze the mesoderm formation and confirm the perceived reduction in Rybp mutants, we performed qRT-PCRs every day during the time course of early (d2, d3, d4, d5, d6) and late progenitor formation (d7, d8). First, we analyzed endogenous Rybp expression in daily resolution and showed that Rybp was consistent throughout the differentiation and significant fluctuations could not been detected (Supplementary Figure S2A). However, when we compared the mesoderm marker gene expression in wt and Rybp ^−/− colonies, the T-box transcription factor Brachyury was found to be downregulated in the Rybp ^−/− cells in all examined time points (Figure 2B). Eomesodermin (Eomes, Tbr2), another T-box factor, which is also expressed in mesoderm and shares partially redundant functions with Brachyury was reduced from day 5 of cardiac differentiation. Besides T-box proteins, the anterior mesoderm gene Goosecoid (Gsc) was also significantly downregulated in the Rybp ^−/− cultures (Supplementary Figure S2B), suggesting that Rybp ^−/− colonies fail to form sufficient amount of mesoderm.

Similarly to the mesoderm markers, the mRNA level of endoderm markers genes such as SRY-Box Transcription Factor 17 (Sox17) and GATA Binding Protein 4 (Gata4) were also underrepresented in the Rybp ^−/− cultures (Figure 2B). In contrast, we found increased expression of the ectoderm marker genes such as Fibroblast Growth Factor 5 (Fgf5) and Microtubule Associated Protein 2 (Map2) in the mutant cardiac cultures (Figure 2B), which suggested that, unlike mesoderm and endoderm, the ectoderm gene expression was permitted in the Rybp ^−/− colonies.

After revealing the expression changes, we analyzed the distribution and subcellular localization of the mesoderm protein BRACHYURY in comparison to RYBP. In both cell lines BRACHYURY positive (+) cells were first detected at d3 in the edge of the early cardiac colonies. After d3, there was a rapid increase in the number of BRACHYURY + cells, peaking at d4 and starting to decrease after d5. In the Rybp ^−/− colonies, the level of BRACHYURY + cells was reduced throughout the differentiation compared to the wt (Figure 3A; Supplementary Figure S2C). These results are consistent with the previously described qRT-PCRs data and were further supported by western blot analysis (Figures 3B, C).

With regards to the subcellular localization of the proteins, we found that RYBP and BRACHYURY proteins were co-localized in the nuclei of the wild type cells at early time points (d3, d4, d5, d6), and the relative intensity of both proteins were fairly high in these cells. To quantify the number of RYBP and BRACHYURY double positive cells, when the BRCHYURY protein is present at its highest level, we performed flow cytometry using d4 cardiac differentiated samples. The analysis showed that 11.31% of the cells are RYBP+/BRACHYURY- only 4.44% of the cells are RYBP-/BRACHYURY+, and the majority, 72,27% of the cells are RYBP+/BRACHYURY+ (Supplementary Figure S2D).

Taken together, our data suggests that RYBP is required for proper formation of mesoderm and endoderm germ layers. The absence of Rybp will result in an unbalanced expression of the germ layer marker genes in favor of ectodermal lineages over mesoderm and endoderm in cardiac progenitors.

Gastrulation is a complex process which involves a series of cellular morphogenesis and cellular movements. In early embryos, the position of the cells in 3D space and the signals from the neighboring cells are also affecting the cell fate. Because of this complexity and to avoid favoring either of the germ lines, we have applied an embryoid body (EB) based 3D cell culture system and investigated the role of RYBP during the time course of EB formation (Figure 4A). EBs were generated from wt and Rybp ^−/− ES cells and spontaneously differentiated for 14 days (details in materials and methods) to form all distinctive germ layers including mesoderm and endoderm.

To reveal the spatiotemporal distribution of the mesoderm protein BRACHYURY, we performed whole mount immunocytochemistry and co-stained BRACHYURY with RYBP in d4, d7, d10, and d14 embryoid bodies (Figure 4B). In the wild type EBs, RYBP signals were detected throughout the differentiation. At d4, the majority of the cells showed medium level of RYBP signals. At d7 and d10 the intensity of RYBP signals were different as in some cells appeared higher and, in some cells, lower. At d14 most of the cells showed medium level of RYBP protein on the surface of the EBs (Figure 4B). BRACHYURY + cells were detected as early as d4 in wt EBs, reached the highest number at d7 (), and disappeared from d10 (Figure 4B). In contrast, Rybp ^−/− EBs showed decreased BRACHYURY levels throughout the differentiation. At the early progenitor stage (d4), the BRACHYURY protein was undetectable in most Rybp ^−/− EBs and only appeared from d7. In the mutants the BRACHYURY + cells persisted till d10 and only decayed by d14 (Figures 4B, D) in the mutants, demonstrating a delay in the BRACHYURY protein kinetics. The majority of the BRACHYURY + cells showed a nuclear co-localization with RYBP in wt EBs (Figure 4B; Supplementary Figures S3A, B), similarly to what we observed in two-dimensional cardiac cultures.

From d7 RYBP + cells appeared mainly in the outer layers of the EBs, where the endoderm forms, therefore we co-stained RYBP with endoderm marker GATA4 as well (Figure 4C; Supplementary Figure S4A). In wild type cells, the first GATA4+ cells were observed at d4, and the number of GATA4+ cells gradually increased during EB formation. In wt EBs, RYBP co-localized in nuclei with GATA4 from d4, and the co-localization sustained, resulting in a wavy multicellular outer GATA4+/RYBP + layer in later stages (Figure 4C; Supplementary Figure S4B). In the lack of RYBP there was a reduction in the number of GATA4+ cells throughout the time course of differentiation. Similarly to the wt EBs, GATA4+ cells started to appear at d4 in the Rybp ^−/− EBs, however at later time points (d7, d10, d14), mutant EBs failed to form proper GATA4+ exterior layers (Figure 4C). Besides the level of GATA4+ cells, we also observed decreased GATA4 signal intensities in the mutant EBs compared to the wt (Figure 4C, E white arrows).

To gain more insights into the interplay between the two germ layers, we performed BRACHYURY and GATA4 immunocytochemistry as well and analyzed the protein localization in the surface and in the middle of the EBs. Our results showed that BRACHYURY and GATA4 proteins strictly appeared in distinct cells both in the wt and in the Rybp ^−/− EBs, and neither of the proteins showed changes in their subcellular localization in the lack of RYBP (Figure 5A). In wt EBs, both BRACHYURY+ and GATA4+ cells appeared on the surface and in the middle of the EBs during progenitor formation (d4, d7). In addition, within the wt EBs, BRACHYURY + cells spread out more compared to the mutant, where we found BRACHYURY + cells to be more restricted in one area within the EBs (Figure 5A (white arrows), Supplementary Video S1).

To quantify the number of cells co-expressing RYBP and GATA4 in wt EBs, we have analyzed the constituent cells of EBs from each time point (d4, d7, d10, d14) of differentiation (Figure 5B). The analysis showed that 99.33% of the cells were RYBP + at the pluripotent stage (d0). As the differentiation proceeded, a portion of the RYBP + cell population became positive to GATA4 too, and only a small portion of the cells (4.99%) lost their RYBP positivity. The number of RYBP+/GATA4+ cells increased over time and reached 64.23% of the total cell population by the 14th day of EB differentiation.

These results showed that the absence of RYBP interfered with the number of mesoderm and endoderm cells during spontaneous EB differentiation. Mesoderm and endoderm cells lacking RYBP not only presented lower levels of BRACHYURY and GATA4 proteins than the wild type but resulted in EBs with abnormal patterning as well.

To compare our in vitro results to the in vivo stages, we re-analyzed a publicly available single-cell RNA sequencing (scRNA-Seq) dataset derived from wild type E6.5 – E8.5 mouse embryos (Mouse atlas was downloaded from Geo dataset: E-MTAB-6967) to gain information about cell populations expressing both Rybp and Brachyury. ScRNA-Seq datasets were processed, and cell populations were clustered as described in Pijuan-Sala et al. (2019a) (Supplementary Figure S5). The analysis revealed that Rybp and Brachyury could co-express as early as the primitive streak formation; one of the highest co-expressions was observed in the anterior primitive streak (Figure 6A), which is known to give rise to definitive endoderm and early somites (Loh et al., 2016). We also found Rybp and Brachyury co-expression in several mesodermal clusters, including the nascent, mixed, intermediate, somitic, caudal mesoderm populations and in neuromesodermal progenitors (NMP) (Martins-Costa et al., 2024). However, in other mesoderm populations, such as pharyngeal and paraxial mesoderm, there was no co-expression detected (Figure 6A). Next, we checked the overall gene expression of Rybp, Brachyury, and other mesoderm marker genes (Eomes, Gsc, Mesp1, Tbx6) in mesodermal clusters. In accordance with prior knowledge, scRNA-seq dot plots showed low, ubiquitous Rybp expression in all mesodermal cell populations, while the expression of Brachyury varied between clusters (Figure 6B). The highest Brachyury expression was detected in nascent mesoderm and caudal mesoderm populations. Mesp1 and Tbx6 were also expressed in all clusters, while other mesodermal markers such as Eomes and Gsc showed low or no expression in clusters derived from later developmental stages. These results further confirm our in vitro observations that Rybp and Brachyury co-express as early as the primitive streak formation, and this co-expression is sustained in nascent mesoderm, which differentiate into cardiac progenitors and eventually cardiomyocytes.

Interestingly, the caudal mesoderm cell population exhibited the second-highest co-expression in scRNA-Seq datasets. After gastrulation, the caudal mesoderm population (also called early neuromesodermal progenitors) moves posteriorly as the embryo elongates and starts to give rise to NMPs (Tzouanacou et al., 2009; Sambasivan and Steventon, 2021). It was described that both caudal progenitors and NMPs are located in the tailbud and required for proper axial elongation and somitogenesis. To gain more information on early organogenesis and check if RYBP protein is present in the tailbud, we performed RYBP immunohistochemistry (IHC) using a sagittal section of wt E8.5 mouse embryo. RYBP staining showed that besides the heart and brain (Figure 6C, black arrows). RYBP protein was detected in the tailbud and allantois of in E8.5 embryo (Figure 6C, pink arrows). RYBP protein was observed primarily in the external regions of the tailbud (Figure 6D, black arrows), while the cells in the interior part of the tailbud appeared to be negative (Figure 6D, pink arrow). The cells located in proximity to the allantois showed the strongest RYBP staining, and the detected signals gradually decreased in the proximal tailbud and vanished in the developing somites.

These results showed that Rybp and Brachyury co-expressed in the primitive streak and several mesoderm cell populations of in vivo mice. We also found that the RYBP protein is present in the tailbud, where high BRACHYURY levels are expected in developing mice.

Besides its indisputable role in gastrulation and cardiac specification, BRACHYURY plays an important role in promoting segmentation and posterior axial extension during embryonic development (Wilson and Beddington, 1996). Brachyury heterozygote mice exhibited short-tailed phenotype (Chesley, 1935), and chimeric gastruloids generated using different percentages of wt and Brachyury null mutant cells showed reduced axial elongation (Wehmeyer et al., 2022). Considering previous results, we wanted to test whether RYBP also has a role in axial elongation. To answer this question, we generated in vitro gastruloids as described in van der van den Brink et al. (2020) (Figure 7A). Gastruloids were generated by aggregation of wt and Rybp ^−/− ES cells, and axial elongation was induced by 24 h CHIR99021 (CHIR) treatment to activate the Wnt signaling pathway (details in the materials and methods). We choose to use a gastruloid model system, which is capable of somitogenesis besides posterior extension and recapitulates the trunk and tail regions of E8.5 stage mouse embryos. Morphological analysis showed that wt and Rybp ^−/− gastruloids appeared indistinguishable until day 4 (Figure 7B; Supplementary Figure S6A). However, at day 5, after we embedded the gastruloids in 10% matrigel, Rybp ^−/− gastruloids looked less developed and failed to prolong sufficiently compared to the wt (Figure 7B (white arrows), Supplementary Figures S6A, S6B). The observed phenotype resembled the phenotype of the previously published Brachyury ^−/− gastruloids (Wehmeyer et al., 2022). The anterior region of the Rybp ^−/− gastruloids also differed from the wt, as they appeared wider, darker, and fragmented (Figure 7B, white arrows; Supplementary Figures S6A, S6B).

Next, we performed whole-mount immunocytochemistry on the wt and Rybp mutant gastruloids to highlight the spatiotemporal distribution of BRACHYURY. In the early stages of gastruloid formation, we could not observe noticeable differences between wt and Rybp ^−/− gastruloids (Figures 8A, B). At day 3, in both cell line most of the cells were BRACHYURY+, while at day 4, the BRACHYURY + cells were oriented only to the elongating pole of wt and Rybp ^−/− gastruloids. In wt gastruloids at day3 and day 4 the majority of the BRACHYURY + cells were RYBP + as well. After embedding (d5), BRACHYURY positive cells were detected in the tailbud, while RYBP + cells were present throughout the exterior regions of wild type gastruloids (Figure 8C; Supplementary Figure S7B). In the mutants, we could detect fewer BRACHYURY + cells in comparison to the age-mate wt gastruloids. The position of the BRACHYURY + cells within Rybp ^−/− gastruloids did not change, they located in the caudal region of the tailbud, however the BRACHYURY + cells seemed to be restricted more to the tip of the tailbud compared to the wt. Using higher magnification and 3D projection, it was revealed that RYBP and BRACHYURY can co-localize in the nuclei of the wt cells, and the double positive cells were detected at the end of the tailbud (Figures 8C, D; Supplementary Figure S7C, Supplementary Video S2) in the same region, where caudal mesoderm cell population is expected.

In addition, we stained the gastruloids with GATA4 to monitor the endoderm layer during gastruloid formation. In early wt and Rybp ^−/− gastruloids (d3), no significant differences were observed in the amount of GATA4+ cells (Supplementary Figure S8A). At day 3, many GATA4+ cells were detected in both cell lines and the number of GATA4+ cells decreased as gastruloids started to elongate. By d4 and d5, only a few positive cells were detected in the anterior region of wt samples. In contrast, the number of GATA4+ cells at d4 and d5 was higher in the mutants, which suggests developmental arrest in Rybp ^−/− gastruloids (Supplementary Figures S8B, C and S9A, B).

To further investigate the molecular mechanism between RYBP and BRACHYURY and to determine if RYBP can interact with BRACHYURY, we amplified and cloned the Brachyury cDNA to an overexpression vector (details in materials and methods) and performed co-immunoprecipitation (CoIP). CoIP results showed that RYBP physically interacts with BRACHYURY, suggesting that the two proteins may work together in a complex to regulate gene expression (Figures 9A–C).

Our results demonstrated that similarly to BRACHYURY, the presence of the RYBP protein was indispensable for proper axial elongation as well. In addition, we showed that RYBP and BRACHYURY could co-localize in the tailbud of the wt gastruloids. Taken together these results suggest that RYBP and BRACHYURY may share a common regulatory role in different stages of early mouse embryonic development.

Mouse (129SV/Ola) R1 ES cells (mentioned as a wild type or wt) and R1 derived Rybp null mutant ES cells (mentioned as Rybp ^−/− ) were thawed on mitomycin C (Sigma, #M0503) subjected MEF feeder layer and cultured as previously described by Henry et al. (2020).

HEK293 were maintained in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, # 31966047) supplemented with 10% fetal bovine serum (Pan-Biotech, # 3702), 0.1 mM MEM Nonessential Amino Acids (Gibco, #11140-068), 1% sodium pyruvate (Gibco, #11360-039) and 50 U ml−1 penicillin/streptomycin (Gibco, #15140-122). The cells passage a 70%–80%, and the medium was changed every second day.

All the cells were cultured in humidified conditions containing 5% CO₂ at 37°C.

For in vitro cardiac differentiation, embryoid bodies (EBs) were generated by the hanging-drop (HD) method (Keller, 1995). The ES cells were dissociated from the monolayer using 0.05% Trypsin-EDTA (Gibco, #15400-054) and counted with a Bürker chamber. The cell suspension was diluted to 50 cells/μ l in a differentiation medium (ES medium without LIF), and 20 μ l droplets were dispensed to the lids of 10 cm bacterial dishes, where each droplet contained 1000 cells. The bacterial dishes were filled with Dulbecco’s phosphate-buffered saline (DPBS (1x), Gibco, #14190–094) to prevent the droplets from drying out. By reversing the lid of the dish, the cells were allowed to aggregate for 48 h with the help of gravity. At the second day, the EBs were collected and plated into a 6 cm 0.1% gelatine-coated cell culture dishes (SPL Life Sciences, #20060) for gene expression and protein studies and 24-well plate (Corning, #356230) containing 0.1% gelatine-coated coverslips for immunocytochemistry (ICC). The cells were maintained up to 8 or 14 days in a differentiation medium, which was changed every second day. The samples were collected at different time points of cardiac differentiation and ES cells were collected to represent the pluripotent stage.

Embryoid bodies were generated by the HD method in the same regard as described above in cardiac differentiation. The EBs were collected on the second day into 10 cm diameter bacteriological dishes (SPL Life Sciences, #20100) where attachment of the cells was prevented. The EBs were kept in suspension for 14 days and the medium was changed every second day. The EBs were harvested for ICC analysis on day 4, day 7, day 10, and day 14 (labeled as d4, d7, d10, d14) during EB differentiation.

Total RNA was isolated from ES cells and in vitro cardiac differentiated samples using the Gene Jet RNA Purification Kit (Thermo Scientific, #K0732) according to the manufacturer’s instructions. For a detailed examination of gene expression in the time course of progenitor formation, samples were collected every day from day 2 till day 8. To check the expression of endothelial marker genes, the samples were derived from day 2, day 4, day 7, day 10, and day 14 (d2, d4, d7, d10, d14), where d2, d4, and d7 represent the early stage and d10 and d14 represent the late stage of differentiation. cDNA synthesis was achieved with the isolated RNA using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystem, #4368814) according to the manufacturer’s instructions. qRT-PCR was performed in SYBR Select Master Mix for CFX (Applied Biosystems, #4472942) using the Roto-Gene Q PCR machine (QUIAGEN). Relative gene expression changes were determined using the ΔΔCt method. The Ct (cycle threshold) values for each gene were normalized to the expression level of Hprt (Hypoxanthine phosphoribosyl transferase I) as an internal control. To calculate the fold expression changes the values were compared to the expression of wild type d0 ES cells. The primers used in this study are listed in the Supplementary Table S1.

Protein samples for western blot analysis were collected from ES cells and from different time points of in vitro cardiac differentiation. Protein kinetics were analyzed in the time course of progenitor formation (d2-d7), and in the case of ISL-1 and BRACHYURY proteins, the time points were extended to later stages (d8, d10, d14) as well. From the designed time points total protein was extracted using RIPA-buffer supplemented with protease inhibitor (Sigma, #S8830). The concentration of the lysates was determined with Bradford assay (Bradford, 1976) and until usage the samples were stored in 6x Laemmli dye (Laemmli, 1970). Equal protein quantities (20 μg each) were loaded into a 12% sodium dodecyl sulfate-polyacrylamide, and gel electrophoresis (SDS-PAGE) was performed using Bio-Rad Mini-PROTEAN 3 Cell (#165-3301). The protein was then transferred to a polyvinylidene fluoride membrane (Immobilon-P, Merk Millipore, #IPVH00010) and blocked in 5% milk for 1 h at room temperature. Afterwards, the membranes were incubated overnight in 5% milk supplemented with primary antibodies (anti-MESP1 (Santa Cruz, #sc130461, 1:1000), anti-ISL-1 (Santa Cruz, #sc390793, 1:1000), anti-BRACHYURY (R&D Systems, #AF 2085, 1:1000), anti-FLAG M2-IgG1-HRP (Sigma, #A8592, 1:4000) antibodies and anti-GAPDH (Bio-Rad, #AHP1628T, 1:1000) was used as housekeeping protein. Goat-anti-mouse IgG-HRP conjugate (Bio Rad, #1721011, 1:2000), goat-anti-Rabbit IgG-HRP conjugate (Invitrogen, #G-21234, 1:10000) and donkey-anti-Goat IgG-HRP (Invitrogen, #A15999, 1:2000) were used as the secondary antibodies. The membranes were washed five times with TBST buffer and developed using Immobilon Western Chemiluminescent HRP Substrate (Merk Millipore, #WBKLS0500), and signals were captured with Alliance Q9 system (UviTech).

For immunofluorescence staining, in vitro cardiac cultures and EBs were fixed with 4% paraformaldehyde (PFA, Sigma-Aldrich, #P6148) for 20 min at 4°C. The cells were permeabilized in 0.2% Triton X-100 (Sigma-Aldrich, #T8787-250ML) in PBS for 20 min at RT. After permeabilization, the cells were blocked in 5% Bovine Serum Albumin (BSA) (VWR Life Science, #9048-46-8) in PBS for 1 h at RT. Sequentially the samples were incubated overnight in 5% BSA supplemented with the primary antibodies (anti-PECAM1 (R&D Systems, #AF3628, 1:1000), anti-BRACHYURY (R&D Systems, #AF2085, 1:1000), anti-DEDAF/RYBP (Merck Millipore, #AB3637, 1:1000), anti-GATA4 (Anti-GATA4 antibody, Thermo-Fisher, #14-9980-82, 1:1000)) in 4°C under gentle shaking. The next day the cells were washed five times with PBS, blocked in 5% BSA for 1 h and incubated in 5% BSA containing fluorescently labeled secondary antibodies (Alexa Fluor 488 Goat anti-Mouse (Invitrogen, #A10667, 1:2000) Alexa Fluor 488 Donkey anti-Rabbit (Invitrogen, #A21206, 1:2000), Alexa Fluor 647 Donkey anti-Goat (Invitrogen #A21447, 1:2000), Alexa Fluor 568 Donkey anti-Rabbit (Invitrogen, #A10042, 1:2000), and Alexa Fluor 488 Donkey anti-Rat (Invitrogen #A21208,1:2000)) for 1 h at 4°C. The cells were then incubated 20 min in PBS containing 4’6-diamidino-2-phenylindole (DAPI; Vector Laboratories, #H-1200, 1:2500) for nuclear visualization. The samples were washed three times with PBS and mounted in Fluoromount-G (Invitrogen, #00-4958-02). The images were taken in Olympus LSM confocal microscopy (Olympus Corporation, Japan).

Westen blot and immunocytochemistry images were quantified using ImageJ software. Western blot band intensities were normalized to GAPDH internal controls, and the intensity values were compared to wt ES cells. All immunocytochemistry standings were analyzed in triplicates derived from three independent samples and the intensity values were normalized to corresponding DAPI counterstaining and compared to wt ES cells. To quantify the 3D EBs and gastruloids we measured the merged z-stack images of the samples.

Single-cell suspension was collected from ES cells, d4 cardiac differentiated, and d4, d7, d10, and d14EBs. The samples were fixed in 70% ethanol for 20 min and washed with 0, 5% BSA. The cells were overnight stained with anti-DEDAF/RYBP (Merck Millipore, #AB3637, 1:1000) in combination with anti-BRACHYURY (R&D systems, #AF 2085, 1:1000) in cardiac samples and in combination with anti-GATA4 (Anti-GATA4 antibody, Thermo-Fisher, #14-9980-82, 1:1000) in EBs. Alexa Fluor 488 Donkey anti-Rabbit (Invitrogen, #A21206, 1:2000), Alexa Fluor 647 Donkey anti-Goat (Invitrogen #A21447, 1:2000), Alexa Fluor 647 Goat anti-Rabbit (Invitrogen, #A21244,1:2000), and Alexa Fluor 488 Donkey anti-Rat (Invitrogen #A21208,1:2000) secondary antibodies were used according to the host species. All antibodies were diluted in dilution solution [0.5% BSA and 0.5% TWEEN-20 (Sigma-Aldrich, #P9416-100ML)]. Fluorescence signals were measured of 50,000 cells/sample using Beckman Coulter CytoFLEX Flow Cytometer (Beckman Coulter, United States) with appropriate settings and compensation adjustments. The negative control was utilized to establish the negative boundary, while the fluorescence intensity and number of positive cells were measured. Subsequently, the data were analyzed using CytExpert 2.4 software.

The single-cell RNA-seq dataset derived from the different developmental stages of mouse embryos was collected from ArrayExpress (accession number: E-MTAB-6967, file name: atlas_data.tar.gz) for our analysis. The metadata generated by the original pipeline was used for our downstream analysis (Pijuan-Sala et al., 2019b), which was performed in RStudio (v.2024.04.02.) with the Seurat package (v.5.1.0). We filtered the dataset to include only cells with clearly defined cell types, then recreated the uniform manifold approximation and projection (UMAP) plot based on the clustering information and coordinates identified by Pijuan-Sala et al. (2019a) with DimPlot function from the Seurat package (v.5.1.0). The resulting 116,312 embryonic cells were classified into 37 clusters, and further analysis was performed accordingly. The bar chart was generated using the Rybp and Brachyury-positive cells of the previously identified 37 clusters. Double positive cells were classified by normalized expression value higher than 0. The plot was performed with ggplot2 (v.3.5.1). Primitive streak and mesoderm clusters were selected for gene expression pattern analysis of marker genes. The average gene expression values were obtained using the DotPlot function from the Seurat package (v.5.1.0), and a percentage cutoff of 1 was applied; the visualization was executed using ggplot2 (v.3.5.1).

Wild type E8.5 mouse embryos were fixed in 4% PFA overnight, and paraffin-embedded sections (6 μm) were mounted for staining. For immunohistochemistry, deparaffinized tissue slides were first treated with 3% H2O2 for 30 min to inactivate endogenous peroxidases, then washed with double-distilled H₂O and soaked in PBS for 5–10 min. The slides were blocked with 10% BSA and then exposed to anti-RYBP antibody (Millipore, #AB3637, 1:100) overnight at 4°C. After the excess antibody was removed, the samples were incubated with a 1:400 dilution of biotin-conjugated secondary anti-rabbit (Vector labs) antibodies for 45 min at room temperature, washed in PBS, and incubated with avidin-biotinylated enzyme complex for 45 min. The reaction was developed with a DAB kit (Vector labs) and monitored by microscopy for the proper exposure.

Gastruloids were generated as previously described in van den Brink et al. (2020). Briefly, wt and Rybp ^−/− ES cells were aggregated in ultra-low 96 well plates (Corning, #7007) containing N2B27 (Ndiff 227, #Y40002) medium. On the second day, the cells were treated with 3 μM CHIR90021 (Selleck Chemcals, #S1263) for 24 h, then the medium was changed back to N2B27. At day 4 gastruloids were collected and embedded in 10% matrigel (Corning, #356230) 90% N2B27 medium mixture in 24 well cell culture plates (Corning, #3524). The samples were fixed with 4% PFA in different time points (d2, d4, d5) during gastruloid formation and immunocytochemistry was performed as described above.

Full-length of Brachyury cDNA was amplified from d4 wt cardiac differentiated cells (where the highest Brachyury gene expression was expected) with PrimeSTAR GXL DNA Polymerase (Takara Bio, #R050Q) using 5′-TAT​AAA​GCT​TTG​TTG​GGT​AGG​GAG​TCA​AGA-3′ forward and 5′-TTT​TAA​GCT​TAT​AGA​TGG​GGG​TGA​CAC​AGG​T-3′ reverse primers. The amplified cDNA was then cloned in the prk7 overexpression vector using HindIII restriction sites. The sequence accuracy and the orientation of the cDNA construct were confirmed by restriction digestion and sequencing.

HEK293 cells were transiently transfected with 5 μg prk7-Brachyury and pcDNA3.1-Rybp-FLAG using Calcium phosphate (CaPO4) method as described in Henry et al. (2023). The cells were harvested 2 days after transfection and were washed with PBS and lysed by 10x passing the cell suspension through a G26 needle in EB buffer (50 mM HEPES (Merck, #1.12041), 150 mM NaCl (Sigma, #S-5886), 0.5 mM MgCl₂ (Roanal, #13007), 0.1% NP40 (Sigma-Aldrich, #74385), 5% glycerol (Sigma, #G-9012), 0.5 mM DTT (Thermo Scientific, #R0862), 1 mM protease inhibitor cocktail (Sigma, #S8830), 25 μM MG132 (Cayman, #CAYM10012628), 0.1 μl/mL benzonase nuclease (Merk-Millipore, #70746-3). Then the lysates were centrifuged (17,000 xg, 20 min, 4°C), and 10%f the cleared supernatant was stored as input, and the rest was used for co-immunoprecipitations (CoIP). CoIP was performed using anti-FLAG-M₂ magnetic beads (Sigma-Aldrich, #M8823) at 4°C for 90 min. Bound proteins were eluted with 200 µg/µ l 3xFLAG peptide (Sigma-Aldrich, #F4799) through competition elution, and Laemmli sample buffer was added to the eluted proteins followed by boiling for 5 min. The samples were then loaded into a 10% SDS-PAGE, and western blot analysis were carried out as described above using the indicated antibodies. For CoIP experiments anti-BRACHYURY (R&D Systems, #AF 2085, 1:1000) and anti-FLAG M2-IgG1-HRP (Sigma, #A8592, 1:4000) antibodies were used as primary antibodies donkey-anti-Goat IgG-HRP (Invitrogen, #A15999, 1:2000) were used as the secondary antibodies.

All experiments were repeated three times, and technical repeats were used as triplicates at each examined timepoints. Experiments were evaluated with Microsoft Excel by using Student’s t-test type 3. Means are standard deviation. Values of p < 0.05 were accepted as significant (*p < 0.05; **p < 0.01; ***p < 0.001).