Cardiac tissue model

Abstract

The present invention provides a cardiac tissue model consisting of at least 60% cardiac cells, the cardiac cells surrounding an inner cavity, the cardiac cells being selected from cardiomyocytes, endocardial cells and epicardial cells, a method for generating such a tissue model, and uses thereof.

Description

Technical Field

[0001] The present invention relates to the field of generating heart tissue models.

Background Art

[0002] The heart is the first functional organ to form in the developing human embryo. The survival of the embryo depends on whether myocardial progenitor cells self-organize into a four-chambered heart and then mature. Self-organization in biology refers to the ability of cells to self-assemble when differentiating into structures such as tissues and organs in vivo or in vitro under conditions similar to those in vivo. Thus, the ability of cells to self-organize under appropriate conditions is extremely important for the functioning of tissues and organs. The lack of self-organization models and physiological models hinders progress in understanding human heart development and regeneration, the etiology of congenital defects and cardiovascular diseases, and the actual physiological and toxicological effects of drugs, and has a major impact.

[0003] Organoids, which are stem cell-derived self-organizing tissue-like structures, are currently revolutionizing biomedical research as they reproduce important aspects of organ development, physiology, and disease. Organoids have been induced from stem cells that mimic aspects of the formation of physiological tissue structures such as the small intestine, large intestine, stomach, liver, lung, brain, epidermis, kidney, retina, esophagus, bladder, placenta, etc. However, in this field, i) the complexity of organoids, the difficulty of controlling their structure and shape, and ii) the fact that reproducing individual organoids in vivo is far from their development due to the robustness of the living body, etc. are major issues.

[0004] International Publication No. 2019 / 174879(A1) describes an artificial heart tissue organoid grown into a multilayer aggregate containing a large number of non-cardiac cells, such as foregut endoderm cells. In this model, only a few small endodermal cavities develop, and it fails to reproduce the large cavities that develop into the four large chambers found in a natural heart. Furthermore, the separation of cardiomyocytes and endocardium, and the formation of gaps (myocardial jelly) seen in the vertebrate heart tube are not observed. Therefore, such an organoid does not reproduce the important aspect of the in vivo condition of a developing heart, which is basically composed only of cardiac cells. Long-term development and maturation of this tissue model is probably impossible because the accumulation of non-cardiac cells prevents the formation of an in vivo-like structure with at least one large cavity that develops into a ventricle.

[0005] Mendjan et al., Cell Stem Cell (2014) Vol.15, pp.310-325, disclose a study on early mesoderm differentiation in two-dimensional culture, but they were unable to obtain cardiac organoids with a lumen.

[0006] Halloin et al., Stem Cell Reports (2019) Vol.13, pp.366-379, disclose a myocardial suspension culture of ventricular-like cardiomyocytes with high strain purity, but the lumen is not described.

[0007] Ma et al., Nature Communications (2015) Vol. 6, p. 7413, describe tissue in which microventricles were formed using PEG-patterned polystyrene microstructures to confine cell proliferation. Thus, these microventricles are artificial structures created by artificial cell proliferation interference and do not reproduce the natural cavity formation that occurs in vivo through self-organization during early cardiac development. For example, an independent endocardial layer is missing, such as the space between cardiomyocytes and the endocardium (myocardial jelly).

[0008] There remains a need to provide organoid-like tissue models that mimic self-organization, such as that found in vivo, and reproduce the early development of the heart—particularly ventricular development. This invention provides such tissue models and methods for generating them. [Overview of the project]

[0009] The present invention relates to a cardiac tissue model comprising at least 60% cardiac cells or at least 50% cardiac cells, independently of any additional cells in the vascular tissue system of the tissue model, wherein the cardiac cells surround the lumen and are selected from the group consisting of cardiomyocytes, endocardial cells (also called myocardial endothelial cells), and epicardial cells.

[0010] The present invention further provides a method for generating a cardiac tissue model, comprising the following steps: a) providing pluripotent stem cells, preferably in a two-dimensional culture; b1) inducing mesodermal differentiation in the presence of a WNT activator and / or a GSK3-β inhibitor, wherein the WNT activator and / or GSK3-β inhibitor, along with an optional PI3 kinase inhibitor, and / or one or a combination thereof of FGF2, activin A, or BMP4, are present in an amount sufficient to differentiate the pluripotent stem cells such that at least 90% of the pluripotent stem cells lose their pluripotency within 40 hours after the start of induction, thereby generating aggregates of mesodermal cells; or b2) inducing mesodermal differentiation in the presence of a WNT activator and / or a GSK3-β inhibitor, and further comprising a PI3 kinase inhibitor and / or one or a combination thereof of FGF2, activin A, or BMP4 A method for developing a cardiac tissue model is provided, comprising: generating aggregates of mesoderm cells by agglutinating cells without binding them to the culture vessel in a low-adhesion three-dimensional culture in the presence of or a combination thereof; and differentiating the mesoderm cells from step b) into cardiac mesoderm cells in a low-adhesion three-dimensional culture, and for at least 3 days, preferably 3 to 7 days, in the presence of cardiomyocyte differentiation factors (preferably BMP4, FGF2, insulin), and in the absence of a WNT activator and / or in the presence of a WNT antagonist, the cells binding to each other instead of the culture vessel to form aggregates of the cells, thereby forming cardiac mesoderm and lumen. Both indicators "b1" and "b2" are referred to as "b".

[0011] The present invention further provides a cardiac tissue model obtained by the method of the present invention.

[0012] The present invention further provides screening and testing methods for observing the effects of a test compound or screening compound, or an overexpressed or underexpressed gene, using the tissue model or method of the present invention.

[0013] The present invention further provides a kit for carrying out the method of the present invention. This kit may include i) a WNT activator and / or a GSK3-β inhibitor, ii) a PI3 kinase inhibitor, and iii) a low-adhesion cell culture vessel.

[0014] The present invention also provides a container plate comprising at least 10 compartments, each containing a tissue model according to the present invention. The tissue models may be substantially the same developmental stage.

[0015] All embodiments of the present invention are described in accordance with the following detailed description, and all preferred embodiments are similarly relevant to all embodiments, aspects, methods, cardiac tissue models, organoids, applications, and kits. For example, a kit or its components may be used in or suitable for the methods of the present invention. Any component used in the described methods may be part of a kit. An inventive tissue model or organoid may be the result of an inventive method or may be used in an inventive method and application. The preferred and detailed descriptions of the methods of the present invention can also be read regarding the suitability of the organoids or tissue models of the present invention obtained or used. All embodiments can be combined with each other unless otherwise stated. [Brief explanation of the drawing]

[0016] [Figure 1] Figure 1 outlines the basic cardiac organoid protocol, starting from pluripotent stem cells and progressing through aggregation in a 96-well plate, chamber formation, and the beating cardiomyocyte stage. The actual time progression is shown section by section. [Figure 2]Figure 2 shows cardiac organoids derived from hPSCs (WTC MYL7-GFP) at day 7.5, induced with CHIR99021 at a concentration of 8 μM. Left: Whole-mount bright-field image showing organoid size. Right: Frozen sections stained with DAPI show the relationship between organoid and cavity size. B: Aggregates of mesoderm cells detached from 1000 hPSCs (WTC MYH10-GFP) show efficient elimination of pluripotency after 40 hours of mesoderm induction (using 8 μM CHIR99021), due to upregulation of the primitive streak marker T and downregulation of the pluripotency marker SOX2. C: Endogenous expression of the cardiomyocyte-specific marker MYL7 is shown 10 days after culturing cardiac organoids (WTC MYL7-GFP). [Figure 3] Figure 3 shows a section of an organoid at day 3.5, A: Absence of endodermal markers SOX17 and EOMES. Scale 200 μm. [Figure 4-1]Figure 4 shows: A: Schematic diagram and bright-field image of a protocol for generating cardiac organoids containing cardiomyocytes (CM) and endothelial cells (EC) in separate layers. Arrows indicate the formed cavities. B: Multiple cardiomyocyte organoids showing expression of MYL7-GFP (green) and CDH5-Tomato (red) in separate layers. C: Flow cytometry results quantitatively showed that the organoid composition was, on average, 53% EC and 41% CM. D: Section of an organoid showing the presence of a lumen surrounded by a ring of CM expressing MYL7-GFP and HAND1, surrounded by a layer of EC expressing CD31. E: Heatmap of vst counts (vst: dispersive stabilization transformation) of different EC and pluripotent stem cells showing HOX gene expression. Analysis of ECs: 2D human cardiac microvascular endothelial cells (HCMECs), 2D ECs differentiated using a forward differentiation protocol, 3D CDH5-tomato positive ECs from cardiac organoids, 2D human umbilical vein endothelial cells (HUVECs), 2D ECs differentiated using the protocol of Patsch et al. (Example 3), and 3D ECs from vascular organoids (Wimmer et al., see Example 3). HOX gene expression in ECs from cardiac organoids shows very similar signs compared to HCMECs. Sections of CHIR99021 (CHIR) at F:6 μM and 4 μM show the inner ring of ECs expressing CD31 under low (4 μM) CHIR conditions. Cleaved caspase 3 staining suggested that the level of apoptosis was low under the 6 μM CHIR condition, but increased inside the EC ring (CD31+) under the 4 μM CHIR condition. VST = Dispersion stabilization conversion. [Figure 4-2]Figure 4 shows: A: Schematic diagram and bright-field image of a protocol for generating cardiac organoids containing cardiomyocytes (CM) and endothelial cells (EC) in separate layers. Arrows indicate the formed cavities. B: Multiple cardiomyocyte organoids showing expression of MYL7-GFP (green) and CDH5-Tomato (red) in separate layers. C: Flow cytometry results quantitatively showed that the organoid composition was, on average, 53% EC and 41% CM. D: Section of an organoid showing the presence of a lumen surrounded by a ring of CM expressing MYL7-GFP and HAND1, surrounded by a layer of EC expressing CD31. E: Heatmap of vst counts (vst: dispersive stabilization transformation) of different EC and pluripotent stem cells showing HOX gene expression. Analysis of ECs: 2D human cardiac microvascular endothelial cells (HCMECs), 2D ECs differentiated using a forward differentiation protocol, 3D CDH5-tomato positive ECs from cardiac organoids, 2D human umbilical vein endothelial cells (HUVECs), 2D ECs differentiated using the protocol of Patsch et al. (Example 3), and 3D ECs from vascular organoids (Wimmer et al., see Example 3). HOX gene expression in ECs from cardiac organoids shows very similar signs compared to HCMECs. Sections of CHIR99021 (CHIR) at F:6 μM and 4 μM show the inner ring of ECs expressing CD31 under low (4 μM) CHIR conditions. Cleaved caspase 3 staining suggested that the level of apoptosis was low under the 6 μM CHIR condition, but increased inside the EC ring (CD31+) under the 4 μM CHIR condition. VST = Dispersion stabilization conversion. [Figure 4-3]Figure 4 shows: A: Schematic diagram and bright-field image of a protocol for generating cardiac organoids containing cardiomyocytes (CM) and endothelial cells (EC) in separate layers. Arrows indicate the formed cavities. B: Multiple cardiomyocyte organoids showing expression of MYL7-GFP (green) and CDH5-Tomato (red) in separate layers. C: Flow cytometry results quantitatively showed that the organoid composition was, on average, 53% EC and 41% CM. D: Section of an organoid showing the presence of a lumen surrounded by a ring of CM expressing MYL7-GFP and HAND1, surrounded by a layer of EC expressing CD31. E: Heatmap of vst counts (vst: dispersive stabilization transformation) of different EC and pluripotent stem cells showing HOX gene expression. Analysis of ECs: 2D human cardiac microvascular endothelial cells (HCMECs), 2D ECs differentiated using a forward differentiation protocol, 3D CDH5-tomato positive ECs from cardiac organoids, 2D human umbilical vein endothelial cells (HUVECs), 2D ECs differentiated using the protocol of Patsch et al. (Example 3), and 3D ECs from vascular organoids (Wimmer et al., see Example 3). HOX gene expression in ECs from cardiac organoids shows very similar signs compared to HCMECs. Sections of CHIR99021 (CHIR) at F:6 μM and 4 μM show the inner ring of ECs expressing CD31 under low (4 μM) CHIR conditions. Cleaved caspase 3 staining suggested that the level of apoptosis was low under the 6 μM CHIR condition, but increased inside the EC ring (CD31+) under the 4 μM CHIR condition. VST = Dispersion stabilization conversion. [Figure 5] Figure 5 shows bright-field (BF), MYL7-GFP, CDH5-Tomato, and merged images of organoids differentiated under sunitinib (100 nM) treatment designed as A: "+sunitinib" or the standard protocol ("normal"). The images show more homogeneous MYL7-GFP expression under the "+sunitinib" condition and the absence of CDH5-Tomato expression under both conditions. Each organoid image is 2500 × 2500 μm. [Figure 6-1]Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-2] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-3] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-4] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-5] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-6] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-7] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-8] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-9]Figure 6 is a representative image of the quantification strategy applied to flow cytometry data at the end of differentiation (day 7.5). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophore, "Normal" designates the standard differentiation protocol, and "+ Sunitinib" refers to organoids treated with sunitinib (100 μM). B: Quantification of the flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes were boosted to 96% by sunitinib treatment under the standard protocol. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low at approximately 2% under the standard protocol and similarly 3% under the "+ Sunitinib" condition. [Figure 6-10] Figure 6 is a representative image of the quantification strategy applied to flow cytometry data at the end of differentiation (day 7.5). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophore, "Normal" designates the standard differentiation protocol, and "+ Sunitinib" refers to organoids treated with sunitinib (100 μM). B: Quantification of the flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes were boosted to 96% by sunitinib treatment under the standard protocol. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low at approximately 2% under the standard protocol and similarly 3% under the "+ Sunitinib" condition. [Figure 6-11] Figure 6 is a representative image of the quantification strategy applied to flow cytometry data at the end of differentiation (day 7.5). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophore, "Normal" designates the standard differentiation protocol, and "+ Sunitinib" refers to organoids treated with sunitinib (100 μM). B: Quantification of the flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes were boosted to 96% by sunitinib treatment under the standard protocol. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low at approximately 2% under the standard protocol and similarly 3% under the "+ Sunitinib" condition. [Figure 6-12] Figure 6 is a representative image of the quantification strategy applied to the flow cytometry data at the end of differentiation (day 7.5). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophore, "Normal" designates the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). B: Quantification of the flow cytometry data showed that about 60% of MYL7-GFP-expressing cardiomyocytes were boosted to 96% by sunitinib treatment under the standard protocol. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low at about 2% under the standard protocol and similarly 3% under the "+Sunitinib" condition. [Figure 6-13] Figure 6 is a representative image of the quantification strategy applied to the flow cytometry data at the end of differentiation (day 7.5). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophore, "Normal" designates the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). B: Quantification of the flow cytometry data showed that about 60% of MYL7-GFP-expressing cardiomyocytes were boosted to 96% by sunitinib treatment under the standard protocol. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low at about 所2% under the standard protocol and similarly 3% under the "+Sunitinib" condition. [Figure 6-14] Figure 6 is a representative image of the quantification strategy applied to the flow cytometry data at the end of differentiation (day 7.5). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophore, "Normal" designates the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). B: Quantification of the flow cytometry data showed that about 60% of MYL7-GFP-expressing cardiomyocytes were boosted to 96% by sunitinib treatment under the standard protocol. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low at about 2% under the standard protocol and similarly 3% under the "+Sunitinib" condition. [Figure 6-15] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-16] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-17]Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 6-18] Figure 6 shows representative images of the quantification strategies applied to flow cytometry data at the end of differentiation (day 7.5) (A). "WT CTRL" refers to wild-type cells from organoids that do not express any fluorophores, "Normal" specifies the standard differentiation protocol, and "+Sunitinib" refers to organoids treated with sunitinib (100 μM). (B) Quantification of flow cytometry data showed that approximately 60% of MYL7-GFP-expressing cardiomyocytes under the standard protocol were boosted to 96% by sunitinib treatment. Quantification also showed that cells expressing the endothelial marker CDH5-Tomato were low, at approximately 2% under the standard protocol and similarly at 3% under the "+Sunitinib" condition. [Figure 7] Figure 7 shows: A: Representative images of organoid sections containing cardiomyocytes (CM) and endothelial cells (EC) in separate layers. The bright-field (BF) image shows the cellular tissue, and the MYL7-GFP (CM) and CD31 (EC) images show the lumen surrounded by the CM layer, indicating that the cavity is sequentially separated from the EC layer by other cavities. B: Representative images of organoid sections containing cardiomyocytes (CM) and endothelial cells (EC) in separate layers. The BF image shows the cellular tissue, and staining with the CM marker TN NT2 shows overlap with other CM markers such as MYL7. Scale bar = 200 μm. [Figure 8]Figure 8 shows that cardiac organoids from hPSCs (WTC MYL7-GFP) at a concentration of 4 μM CHIR99021 during induction exhibit lining of the internal cavity by endothelial (CD31+) cells. [Figure 9] Figure 9, a combined image of bright-field (BF), MYL7-GFP, and CDH5-Tomato images, shows the two-layer structure of cardiomyocytes (MYL7-GFP) and endothelial cells (EC, CDH5-Tomato) when different Wnt inhibitors (IWR-1 (IWR), XAV939 (XAV), and IWP2 (IWP)) were used during differentiation at two CHIR99021 concentrations (5 μM (CHIR5) and 6 μM (CHIR6)). The image also includes the "IWP-VEGF" state where EC is absent. Each organoid image is 2000 × 2000 μm. [Figure 10-1]Figure 10 shows: A. Schematic diagram of epicardial differentiation. B. Heatmap of vst counts (vst: variance-stabilizing transformation) at different stages of epicardial differentiation. Pluripotency markers are downregulated after mesoderm induction, but the expression of cardiac mesoderm-specific markers is upregulated over time. The expression of epicardial-specific markers is observed between epicardial identification and maintenance. C. Expression of epicardial markers TCF21 and TBX18 in the epicardia and fibronectin-reforming epicardia is compared with pluripotency as seen by qPCR. D. Epicardial cells associated with the end of differentiation (day 8.5) show high positivity for the epicardial marker WT1 (Wilms tumor protein), as seen by antibody staining. Scale bar: 100 μM. E. After differentiating replicated epicardium into SMCs (smooth muscle cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and PDGF-BB, the resulting cells were highly positive for the SMC markers α-SMA (α-smooth muscle actin) and calponin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. F. When replicated epicardium was differentiated into CFs (cardiac fibroblast-like cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and FGF, the resulting cells were highly positive for the CF markers DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2) and vimentin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. [Figure 10-2]Figure 10 shows: A. Schematic diagram of epicardial differentiation. B. Heatmap of vst counts (vst: variance-stabilizing transformation) at different stages of epicardial differentiation. Pluripotency markers are downregulated after mesoderm induction, but the expression of cardiac mesoderm-specific markers is upregulated over time. The expression of epicardial-specific markers is observed between epicardial identification and maintenance. C. Expression of epicardial markers TCF21 and TBX18 in the epicardia and fibronectin-reforming epicardia is compared with pluripotency as seen by qPCR. D. Epicardial cells associated with the end of differentiation (day 8.5) show high positivity for the epicardial marker WT1 (Wilms tumor protein), as seen by antibody staining. Scale bar: 100 μM. E. After differentiating replicated epicardium into SMCs (smooth muscle cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and PDGF-BB, the resulting cells were highly positive for the SMC markers α-SMA (α-smooth muscle actin) and calponin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. F. When replicated epicardium was differentiated into CFs (cardiac fibroblast-like cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and FGF, the resulting cells were highly positive for the CF markers DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2) and vimentin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. [Figure 10-3]Figure 10 shows: A. Schematic diagram of epicardial differentiation. B. Heatmap of vst counts (vst: variance-stabilizing transformation) at different stages of epicardial differentiation. Pluripotency markers are downregulated after mesoderm induction, but the expression of cardiac mesoderm-specific markers is upregulated over time. The expression of epicardial-specific markers is observed between epicardial identification and maintenance. C. Expression of epicardial markers TCF21 and TBX18 in the epicardia and fibronectin-reforming epicardia is compared with pluripotency as seen by qPCR. D. Epicardial cells associated with the end of differentiation (day 8.5) show high positivity for the epicardial marker WT1 (Wilms tumor protein), as seen by antibody staining. Scale bar: 100 μM. E. After differentiating replicated epicardium into SMCs (smooth muscle cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and PDGF-BB, the resulting cells were highly positive for the SMC markers α-SMA (α-smooth muscle actin) and calponin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. F. When replicated epicardium was differentiated into CFs (cardiac fibroblast-like cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and FGF, the resulting cells were highly positive for the CF markers DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2) and vimentin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. [Figure 10-4]Figure 10 shows: A. Schematic diagram of epicardial differentiation. B. Heatmap of vst counts (vst: variance-stabilizing transformation) at different stages of epicardial differentiation. Pluripotency markers are downregulated after mesoderm induction, but the expression of cardiac mesoderm-specific markers is upregulated over time. The expression of epicardial-specific markers is observed between epicardial identification and maintenance. C. Expression of epicardial markers TCF21 and TBX18 in the epicardia and fibronectin-reforming epicardia is compared with pluripotency as seen by qPCR. D. Epicardial cells associated with the end of differentiation (day 8.5) show high positivity for the epicardial marker WT1 (Wilms tumor protein), as seen by antibody staining. Scale bar: 100 μM. E. After differentiating replicated epicardium into SMCs (smooth muscle cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and PDGF-BB, the resulting cells were highly positive for the SMC markers α-SMA (α-smooth muscle actin) and calponin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. F. When replicated epicardium was differentiated into CFs (cardiac fibroblast-like cells) for 12 days in the presence of TGF-β, insulin, L-ascorbic acid, and FGF, the resulting cells were highly positive for the CF markers DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2) and vimentin. Scale bar: 1 mm, composite and zoomed images are 1 mm × 1 mm. [Figure 11-1]Figure 11 shows: A. Schematic diagram of the epicardial co-culture system in which myocardial organoids are differentiated (dpf: days after fusion). B. Condition of myocardial organoids co-cultured with epicardial tissue, 2 days after fusion. The epicardial tissue expresses the epicardial marker WT1 (Wilms tumor protein), while the myocardial organoids show high positivity for cTNT (cardiac troponin-T). Scale bars: 100uM (b') and 50uM (b''). C. Epicardial cells (marked with TdTomato expression) migrate to the cardiomyocyte layer (marked with arrows) and then differentiate, subsequently marked with SM22 (tagrin) and VIM (vimentin). Specimen fixed at 24dpf. Scale bar: 20uM. D. Epicardial cells form the outer layer of cardiac organoids and migrate to the cardiomyocyte layer (epidural tissue is indicated by TdTomato expression). Cells are positive for fibroblast-specific Col1A2. The specimen is fixed at 8 dpf. Scale bar: 50 μM. E. Epicardial cells migrate to the pericardial layer (epidamidal cell expression is indicated by TdTomato expression). The cells are positive for fibroblast-specific DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2). The specimen is fixed at 7 dpf. Scale bar: 20 μM. F. Epicardial cells form the outer layer of myocardial organoids and also migrate to the myocardial layer (epidamidal cell expression is indicated by TdTomato expression). The cells are positive for fibroblast-specific decorin. The specimen is fixed at 16 dpf. Scale bar: 50 μM. G: Schematic diagram of a three-layered cardiac model with an outer layer of epicardial cells, an intermediate layer of myocardial cells, and an inner layer of endocardial cells (cardiac endothelial cells). [Figure 11-2]Figure 11 shows: A. Schematic diagram of the epicardial co-culture system in which myocardial organoids are differentiated (dpf: days after fusion). B. Condition of myocardial organoids co-cultured with epicardial tissue, 2 days after fusion. The epicardial tissue expresses the epicardial marker WT1 (Wilms tumor protein), while the myocardial organoids show high positivity for cTNT (cardiac troponin-T). Scale bars: 100uM (b') and 50uM (b''). C. Epicardial cells (marked with TdTomato expression) migrate to the cardiomyocyte layer (marked with arrows) and then differentiate, subsequently marked with SM22 (tagrin) and VIM (vimentin). Specimen fixed at 24dpf. Scale bar: 20uM. D. Epicardial cells form the outer layer of cardiac organoids and migrate to the cardiomyocyte layer (epidural tissue is indicated by TdTomato expression). Cells are positive for fibroblast-specific Col1A2. The specimen is fixed at 8 dpf. Scale bar: 50 μM. E. Epicardial cells migrate to the pericardial layer (epidamidal cell expression is indicated by TdTomato expression). The cells are positive for fibroblast-specific DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2). The specimen is fixed at 7 dpf. Scale bar: 20 μM. F. Epicardial cells form the outer layer of myocardial organoids and also migrate to the myocardial layer (epidamidal cell expression is indicated by TdTomato expression). The cells are positive for fibroblast-specific decorin. The specimen is fixed at 16 dpf. Scale bar: 50 μM. G: Schematic diagram of a three-layered cardiac model with an outer layer of epicardial cells, an intermediate layer of myocardial cells, and an inner layer of endocardial cells (cardiac endothelial cells). [Figure 11-3]Figure 11 shows: A. Schematic diagram of the epicardial co-culture system in which myocardial organoids are differentiated (dpf: days after fusion). B. Condition of myocardial organoids co-cultured with epicardial tissue, 2 days after fusion. The epicardial tissue expresses the epicardial marker WT1 (Wilms tumor protein), while the myocardial organoids show high positivity for cTNT (cardiac troponin-T). Scale bars: 100uM (b') and 50uM (b''). C. Epicardial cells (marked with TdTomato expression) migrate to the cardiomyocyte layer (marked with arrows) and then differentiate, subsequently marked with SM22 (tagrin) and VIM (vimentin). Specimen fixed at 24dpf. Scale bar: 20uM. D. Epicardial cells form the outer layer of cardiac organoids and migrate to the cardiomyocyte layer (epidural tissue is indicated by TdTomato expression). Cells are positive for fibroblast-specific Col1A2. The specimen is fixed at 8 dpf. Scale bar: 50 μM. E. Epicardial cells migrate to the pericardial layer (epidamidal cell expression is indicated by TdTomato expression). The cells are positive for fibroblast-specific DDR2 (Discoidin Domain Receptor Tyrosine Kinase 2). The specimen is fixed at 7 dpf. Scale bar: 20 μM. F. Epicardial cells form the outer layer of myocardial organoids and also migrate to the myocardial layer (epidamidal cell expression is indicated by TdTomato expression). The cells are positive for fibroblast-specific decorin. The specimen is fixed at 16 dpf. Scale bar: 50 μM. G: Schematic diagram of a three-layered cardiac model with an outer layer of epicardial cells, an intermediate layer of myocardial cells, and an inner layer of endocardial cells (cardiac endothelial cells). [Figure 12-1] Figure 12 shows that A: the diameter of organoids (WTC iPS cells) at day 3.5 treated with the BMP inhibitor noggin and LDN193189 was quantified, and it can be seen that the diameter was significantly reduced by the inhibitor treatment. B: Organoids (H9 embryonic stem cells) at day 3.5 of wild-type (WT) control cells were compared with HAND1 knockout (KO) cells. Quantification shows that the diameter size was significantly reduced in the KO organoids. [Figure 12-2]Figure 12 shows that A: the diameter of organoids (WTC iPS cells) at day 3.5 treated with the BMP inhibitor noggin and LDN193189 was quantified, and it can be seen that the diameter was significantly reduced by the inhibitor treatment. B: Organoids (H9 embryonic stem cells) at day 3.5 of wild-type (WT) control cells were compared with HAND1 knockout (KO) cells. Quantification shows that the diameter size was significantly reduced in the KO organoids. [Figure 12-3] Figure 12 shows that A: the diameter of organoids (WTC iPS cells) at day 3.5 treated with the BMP inhibitor noggin and LDN193189 was quantified, and it can be seen that the diameter was significantly reduced by the inhibitor treatment. B: Organoids (H9 embryonic stem cells) at day 3.5 of wild-type (WT) control cells were compared with HAND1 knockout (KO) cells. Quantification shows that the diameter size was significantly reduced in the KO organoids. [Figure 13] Figure 13 illustrates cardiac organoids ("cardioids") and some of their applications. Left: Cardioid products and their use in high-throughput methods. Center: Different cardioids at different stages, some with cardiomyocytes or epicardium cells on the inner liner and epicardium cells on the outside, and some without. The figure below shows their use as a damage model due to cryogenic injury, along with fibroblasts infiltrating the injury site. Right: Modeling of the ventricle. [Figure 14-1]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 14-2]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 14-3]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 14-4]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 14-5]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 14-6]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 14-7]Figure 14 shows the formation of ventricular-like structures in vitro. (A) Cardiomyocyte differentiation protocol: WNT signaling activation (by CHIR99021), WNT signaling inhibition (by IWP-2 / IWR-1 / XAV-939), and PI3K signaling inhibition (by LY294002). (B) Representative full-mount image (day 5.5) of a high-throughput differentiation approach showing robust generation of pulsating structures including cavities in three biological replications (iPSC WTC line). (C) Cryosection of a cardioid at day 7.5 showing cavities and expression of the CM-specific marker TNNT2. Scale bar: 200 μm. (D) Quantification of TNNI1-GFP+ cells in a cardioid at day 7.5 by flow cytometry. (E) Heatmap showing transcriptional changes from day 1.5 to day 5.5 associated with CM formation and maintenance of major cardiomyocyte genes, including cardiac mesoderm markers. VST: Dispersion-stabilized converted count. (F) Results of organoid RT-qPCR at day 14, where activin and CHIR99021 concentrations were varied during the induction phase, and retinoid acid concentrations were varied in the presence or absence of SB 43154 during the cardiac mesoderm stage. The scale bar is normalized for the magnification changes from housekeeping genes, PBGD, and pluripotent stem cells. (G) Immunostaining results of CM in the ventricle (IRX4) and atrium (NR2F2) of a cardioid (d10) showing optimized left venous conditions (A4CH4 RA50) compared to unoptimized conditions (A50CH4 RA500). A, activin; CHIR, CHIR99021, RA, retinoic acid; SB, SB 43154. Scale bar: 100 μm. (H) ScRNA-seq analysis of the proportion of CMs under intermediate concentration WNT(CHIR6) / high activin(A50) / RA500 conditions (N=2, n=1717) versus low concentration WNT(CH4) / low activin(A4) / RA50 conditions (N=2, n=5097) expressing more atrial markers versus more ventricular markers. TNNT2 is expressed in all CMs. Cell line used in this figure: WTC. [Figure 15-1]Figure 15 shows how cardiac mesoderm self-organizes to form cavities (in vitro and in vitro). (A) Time course of cardioid formation. Top image. Quantification of size changes during differentiation. Bottom image. Immunostaining of full-thickness bright-field images and frozen sections shows that the size increases over time and cavities are formed. Cavities can be confirmed for the first time at day 2.5 and are not formed by differences in apoptosis (Casp3 negative) or local proliferation (KI67+). Scale bar: 500 μm (bright-field image), 50 μm (section). (B) Cardiac mesoderm excisions from chicken embryos under human cardiac mesoderm conditions form chamber-like CM structures with cavities. Scale bar: 200 μm. (C) Detailed images of organoids (from peripheral to central) from day 1.5 to day 2.5. Differences in the expression of F-actin (phalloidin), membrane-bound β-catenin (PY-654-β-catenin), and N-cadherin indicate the formation of loose and dense mesodermal compartments at the start of the cardiac mesoderm stage. (D) Representative images of 65 μm² squares of the loose (central) and dense (peripheral) layers of cardiac mesoderm, with quantification of the number of nuclei (N=3, n=14). Cell line used in this figure: WTC. [Figure 15-2]Figure 15 shows how cardiac mesoderm self-organizes to form cavities (in vitro and in vitro). (A) Time course of cardioid formation. Top image. Quantification of size changes during differentiation. Bottom image. Immunostaining of full-thickness bright-field images and frozen sections shows that the size increases over time and cavities are formed. Cavities can be confirmed for the first time at day 2.5 and are not formed by differences in apoptosis (Casp3 negative) or local proliferation (KI67+). Scale bar: 500 μm (bright-field image), 50 μm (section). (B) Cardiac mesoderm excisions from chicken embryos under human cardiac mesoderm conditions form chamber-like CM structures with cavities. Scale bar: 200 μm. (C) Detailed images of organoids (from peripheral to central) from day 1.5 to day 2.5. Differences in the expression of F-actin (phalloidin), membrane-bound β-catenin (PY-654-β-catenin), and N-cadherin indicate the formation of loose and dense mesodermal compartments at the start of the cardiac mesoderm stage. (D) Representative images of 65 μm² squares of the loose (central) and dense (peripheral) layers of cardiac mesoderm, with quantification of the number of nuclei (N=3, n=14). Cell line used in this figure: WTC. [Figure 15-3]Figure 15 shows how cardiac mesoderm self-organizes to form cavities (in vitro and in vitro). (A) Time course of cardioid formation. Top image. Quantification of size changes during differentiation. Bottom image. Immunostaining of full-thickness bright-field images and frozen sections shows that the size increases over time and cavities are formed. Cavities can be confirmed for the first time at day 2.5 and are not formed by differences in apoptosis (Casp3 negative) or local proliferation (KI67+). Scale bar: 500 μm (bright-field image), 50 μm (section). (B) Cardiac mesoderm excisions from chicken embryos under human cardiac mesoderm conditions form chamber-like CM structures with cavities. Scale bar: 200 μm. (C) Detailed images of organoids (from peripheral to central) from day 1.5 to day 2.5. Differences in the expression of F-actin (phalloidin), membrane-bound β-catenin (PY-654-β-catenin), and N-cadherin indicate the formation of loose and dense mesodermal compartments at the start of the cardiac mesoderm stage. (D) Representative images of 65 μm² squares of the loose (central) and dense (peripheral) layers of cardiac mesoderm, with quantification of the number of nuclei (N=3, n=14). Cell line used in this figure: WTC. [Figure 15-4]Figure 15 shows how cardiac mesoderm self-organizes to form cavities (in vitro and in vitro). (A) Time course of cardioid formation. Top image. Quantification of size changes during differentiation. Bottom image. Immunostaining of full-thickness bright-field images and frozen sections shows that the size increases over time and cavities are formed. Cavities can be confirmed for the first time at day 2.5 and are not formed by differences in apoptosis (Casp3 negative) or local proliferation (KI67+). Scale bar: 500 μm (bright-field image), 50 μm (section). (B) Cardiac mesoderm excisions from chicken embryos under human cardiac mesoderm conditions form chamber-like CM structures with cavities. Scale bar: 200 μm. (C) Detailed images of organoids (from peripheral to central) from day 1.5 to day 2.5. Differences in the expression of F-actin (phalloidin), membrane-bound β-catenin (PY-654-β-catenin), and N-cadherin indicate the formation of loose and dense mesodermal compartments at the start of the cardiac mesoderm stage. (D) Representative images of 65 μm² squares of the loose (central) and dense (peripheral) layers of cardiac mesoderm, with quantification of the number of nuclei (N=3, n=14). Cell line used in this figure: WTC. [Figure 16-1]Figure 16 shows that WNT and BMP control cardioid self-organization and normalization. (A) CHIR99021 concentration range during mesothelial induction has a dramatic effect on cardioid diameter (day 3.5) and CM identification (day 7.5). Protocol starting with 2500 hPSCs. Scale bar: 2500 μm. (B) Quantification of cardioid diameter at day 3.5 of cardiac mesoderm differentiation. (C) BMP target genes (HAND1, IRX3, BMP4, BMP2, BMPR2) are upexpressed under cavity formation conditions (8 μM CHIR99201), and EC genes (PECAM1, CDH5, VEGFA) are upregulated when using 4 μM CHIR99021. (D) BMP inhibition with either noggin (100 ng / ml) or LDN193189 (0.2 μM) reduces cardioid diameter. Protocol starting with 7500 hPSCs. Scale bar: 1958 μm. (E) Quantification of cardioid diameter at day 3.5 with and without BMP inhibition. (F) Cell / organoid cytometry revealed that the decrease in diameter was not due to a function that reduces cell number. (G) Noggin and LDN193189 treatment inhibits epigastric dilation. Scale bar: 200 μm. (H) Wnt activation at the cardiac mesoderm stage inhibits CM differentiation but not cavity enlargement. Scale bar: 200 μm. All bar graphs are shown. Mean + / - SD. Cell lines used in this figure: H9, WTC. [Figure 16-2]Figure 16 shows that WNT and BMP control cardioid self-organization and normalization. (A) CHIR99021 concentration range during mesothelial induction has a dramatic effect on cardioid diameter (day 3.5) and CM identification (day 7.5). Protocol starting with 2500 hPSCs. Scale bar: 2500 μm. (B) Quantification of cardioid diameter at day 3.5 of cardiac mesoderm differentiation. (C) BMP target genes (HAND1, IRX3, BMP4, BMP2, BMPR2) are upexpressed under cavity formation conditions (8 μM CHIR99201), and EC genes (PECAM1, CDH5, VEGFA) are upregulated when using 4 μM CHIR99021. (D) BMP inhibition with either noggin (100 ng / ml) or LDN193189 (0.2 μM) reduces cardioid diameter. Protocol starting with 7500 hPSCs. Scale bar: 1958 μm. (E) Quantification of cardioid diameter at day 3.5 with and without BMP inhibition. (F) Cell / organoid cytometry revealed that the decrease in diameter was not due to a function that reduces cell number. (G) Noggin and LDN193189 treatment inhibits epigastric dilation. Scale bar: 200 μm. (H) Wnt activation at the cardiac mesoderm stage inhibits CM differentiation but not cavity enlargement. Scale bar: 200 μm. All bar graphs are shown. Mean + / - SD. Cell lines used in this figure: H9, WTC. [Figure 16-3]Figure 16 shows that WNT and BMP control cardioid self-organization and normalization. (A) CHIR99021 concentration range during mesothelial induction has a dramatic effect on cardioid diameter (day 3.5) and CM identification (day 7.5). Protocol starting with 2500 hPSCs. Scale bar: 2500 μm. (B) Quantification of cardioid diameter at day 3.5 of cardiac mesoderm differentiation. (C) BMP target genes (HAND1, IRX3, BMP4, BMP2, BMPR2) are upexpressed under cavity formation conditions (8 μM CHIR99201), and EC genes (PECAM1, CDH5, VEGFA) are upregulated when using 4 μM CHIR99021. (D) BMP inhibition with either noggin (100 ng / ml) or LDN193189 (0.2 μM) reduces cardioid diameter. Protocol starting with 7500 hPSCs. Scale bar: 1958 μm. (E) Quantification of cardioid diameter at day 3.5 with and without BMP inhibition. (F) Cell / organoid cytometry revealed that the decrease in diameter was not due to a function that reduces cell number. (G) Noggin and LDN193189 treatment inhibits epigastric dilation. Scale bar: 200 μm. (H) Wnt activation at the cardiac mesoderm stage inhibits CM differentiation but not cavity enlargement. Scale bar: 200 μm. All bar graphs are shown. Mean + / - SD. Cell lines used in this figure: H9, WTC. [Figure 16-4]Figure 16 shows that WNT and BMP control cardioid self-organization and normalization. (A) CHIR99021 concentration range during mesothelial induction has a dramatic effect on cardioid diameter (day 3.5) and CM identification (day 7.5). Protocol starting with 2500 hPSCs. Scale bar: 2500 μm. (B) Quantification of cardioid diameter at day 3.5 of cardiac mesoderm differentiation. (C) BMP target genes (HAND1, IRX3, BMP4, BMP2, BMPR2) are upexpressed under cavity formation conditions (8 μM CHIR99201), and EC genes (PECAM1, CDH5, VEGFA) are upregulated when using 4 μM CHIR99021. (D) BMP inhibition with either noggin (100 ng / ml) or LDN193189 (0.2 μM) reduces cardioid diameter. Protocol starting with 7500 hPSCs. Scale bar: 1958 μm. (E) Quantification of cardioid diameter at day 3.5 with and without BMP inhibition. (F) Cell / organoid cytometry revealed that the decrease in diameter was not due to a function that reduces cell number. (G) Noggin and LDN193189 treatment inhibits epigastric dilation. Scale bar: 200 μm. (H) Wnt activation at the cardiac mesoderm stage inhibits CM differentiation but not cavity enlargement. Scale bar: 200 μm. All bar graphs are shown. Mean + / - SD. Cell lines used in this figure: H9, WTC. [Figure 17-1]Figure 17 shows that HAND1 is necessary for the self-organization of cardiac mesoderm (d3.5). (A) HAND1 KO cardioid showing downregulation of NKX2-5. Scale bar: 200 μm. (B) WT cardioid has more cavities (arrows) and a larger diameter compared to HAND1 KO cells. Scale bar: 2000 μm. (C) HAND1 KO cardioid shows smaller cavities and a reduced number of cavities. Scale bar: 2000 μm. (C') and (C''), details of C and HAND1 KO confirmation by HAND1 staining. Scale bar: 200 μm. (D) Quantification of cardioid cavity expansion as a function of diameter in HAND1 KO and WT cardioids. (E) Percentage of cardioid area covered by cavities in WT and KO cardioids. (F) Timeline of organoid formation up to day 3.5 (cardiac mesoderm stage), the point of analysis. Increased WNT signaling (CHIR99021) during mesoderm induction rescues cavity defects in HAND1 KO organoids (arrow). Scale bar: 2500 μm. (G) Quantification of cardioid diameter suggests that increased WNT activation rescues cavity defects in HAND1 KO. All bar graphs are shown. Mean + / - SD. All data in this figure are from cardioids on day 3.5. Cell lines used in this figure: HAND1 KO (H9), NKX2-5 KO (H9). [Figure 17-2]Figure 17 shows that HAND1 is necessary for the self-organization of cardiac mesoderm (d3.5). (A) HAND1 KO cardioid showing downregulation of NKX2-5. Scale bar: 200 μm. (B) WT cardioid has more cavities (arrows) and a larger diameter compared to HAND1 KO cells. Scale bar: 2000 μm. (C) HAND1 KO cardioid shows smaller cavities and a reduced number of cavities. Scale bar: 2000 μm. (C') and (C''), details of C and HAND1 KO confirmation by HAND1 staining. Scale bar: 200 μm. (D) Quantification of cardioid cavity expansion as a function of diameter in HAND1 KO and WT cardioids. (E) Percentage of cardioid area covered by cavities in WT and KO cardioids. (F) Timeline of organoid formation up to day 3.5 (cardiac mesoderm stage), the point of analysis. Increased WNT signaling (CHIR99021) during mesoderm induction rescues cavity defects in HAND1 KO organoids (arrow). Scale bar: 2500 μm. (G) Quantification of cardioid diameter suggests that increased WNT activation rescues cavity defects in HAND1 KO. All bar graphs are shown. Mean + / - SD. All data in this figure are from cardioids on day 3.5. Cell lines used in this figure: HAND1 KO (H9), NKX2-5 KO (H9). [Figure 17-3]Figure 17 shows that HAND1 is necessary for the self-organization of cardiac mesoderm (d3.5). (A) HAND1 KO cardioid showing downregulation of NKX2-5. Scale bar: 200 μm. (B) WT cardioid has more cavities (arrows) and a larger diameter compared to HAND1 KO cells. Scale bar: 2000 μm. (C) HAND1 KO cardioid shows smaller cavities and a reduced number of cavities. Scale bar: 2000 μm. (C') and (C''), details of C and HAND1 KO confirmation by HAND1 staining. Scale bar: 200 μm. (D) Quantification of cardioid cavity expansion as a function of diameter in HAND1 KO and WT cardioids. (E) Percentage of cardioid area covered by cavities in WT and KO cardioids. (F) Timeline of organoid formation up to day 3.5 (cardiac mesoderm stage), the point of analysis. Increased WNT signaling (CHIR99021) during mesoderm induction rescues cavity defects in HAND1 KO organoids (arrow). Scale bar: 2500 μm. (G) Quantification of cardioid diameter suggests that increased WNT activation rescues cavity defects in HAND1 KO. All bar graphs are shown. Mean + / - SD. All data in this figure are from cardioids on day 3.5. Cell lines used in this figure: HAND1 KO (H9), NKX2-5 KO (H9). [Figure 17-4]Figure 17 shows that HAND1 is necessary for the self-organization of cardiac mesoderm (d3.5). (A) HAND1 KO cardioid showing downregulation of NKX2-5. Scale bar: 200 μm. (B) WT cardioid has more cavities (arrows) and a larger diameter compared to HAND1 KO cells. Scale bar: 2000 μm. (C) HAND1 KO cardioid shows smaller cavities and a reduced number of cavities. Scale bar: 2000 μm. (C') and (C''), details of C and HAND1 KO confirmation by HAND1 staining. Scale bar: 200 μm. (D) Quantification of cardioid cavity expansion as a function of diameter in HAND1 KO and WT cardioids. (E) Percentage of cardioid area covered by cavities in WT and KO cardioids. (F) Timeline of organoid formation up to day 3.5 (cardiac mesoderm stage), the point of analysis. Increased WNT signaling (CHIR99021) during mesoderm induction rescues cavity defects in HAND1 KO organoids (arrow). Scale bar: 2500 μm. (G) Quantification of cardioid diameter suggests that increased WNT activation rescues cavity defects in HAND1 KO. All bar graphs are shown. Mean + / - SD. All data in this figure are from cardioids on day 3.5. Cell lines used in this figure: HAND1 KO (H9), NKX2-5 KO (H9). [Figure 18-1]Figure 18 shows that WNT, activin, and VEGF regulate the self-organization of endoderm and cardiomyocytes. (A) Low WNT doses (CHIR99021, 4 μM) during mesoderm induction result in higher expression of EC-specific genes (VEGFA, TAL1, LMO2, ETV2, PECAM1) compared to high WNT (CHIR99021, 8 μM). (A') Time series diagrams per differentiation formed under no VEGF supplementation, low or high WNT, and activin doses. (B) The proportion of VEGF-expressing CM (unicellular sequences) increases in the optimized ventricular-like CM (low WNT (CHIR: 4 μM) / low activin (4 ng / ml), N=2, n=1717) compared to the intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) condition (N=2, n=5097). (C) Image example of partial EC-lining appearance under the "no activin" condition. Scale bar: 200 μm. (D) Quantification of different EC formation and lining development. N=4, n=29. (E) Image examples of categories 1 and 2. Scale bar: 200 μm. (F) Time series diagram of differentiation performed with added VEGF and intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) doses and WNT inhibition. (F') Quantification of FACS data showing the ratio of CM to EC rigidity in cardioids. Mean + / - SD. (F'') Cardioids show separation of CM and EC layers and appearance of a fibroblast-like cell (COL1A1+) layer. Scale bar: 200 μm. (G) Time series diagram of differentiation performed with VEGF addition, intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) administration, and lack of WNT inhibition. (G') Frozen sections of cardioids containing only EC cells (which do not contain CM) and fibroblast-like cells, when WNT was not inhibited during the cardiac mesoderm stage and VEGF was added. Cell line used in this figure: WTC. [Figure 18-2]Figure 18 shows that WNT, activin, and VEGF regulate the self-organization of endoderm and cardiomyocytes. (A) Low WNT doses (CHIR99021, 4 μM) during mesoderm induction result in higher expression of EC-specific genes (VEGFA, TAL1, LMO2, ETV2, PECAM1) compared to high WNT (CHIR99021, 8 μM). (A') Time series diagrams per differentiation formed under no VEGF supplementation, low or high WNT, and activin doses. (B) The proportion of VEGF-expressing CM (unicellular sequences) increases in the optimized ventricular-like CM (low WNT (CHIR: 4 μM) / low activin (4 ng / ml), N=2, n=1717) compared to the intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) condition (N=2, n=5097). (C) Image example of partial EC-lining appearance under the "no activin" condition. Scale bar: 200 μm. (D) Quantification of different EC formation and lining development. N=4, n=29. (E) Image examples of categories 1 and 2. Scale bar: 200 μm. (F) Time series diagram of differentiation performed with added VEGF and intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) doses and WNT inhibition. (F') Quantification of FACS data showing the ratio of CM to EC rigidity in cardioids. Mean + / - SD. (F'') Cardioids show separation of CM and EC layers and appearance of a fibroblast-like cell (COL1A1+) layer. Scale bar: 200 μm. (G) Time series diagram of differentiation performed with VEGF addition, intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) administration, and lack of WNT inhibition. (G') Frozen sections of cardioids containing only EC cells (which do not contain CM) and fibroblast-like cells, when WNT was not inhibited during the cardiac mesoderm stage and VEGF was added. Cell line used in this figure: WTC. [Figure 18-3]Figure 18 shows that WNT, activin, and VEGF regulate the self-organization of endoderm and cardiomyocytes. (A) Low WNT doses (CHIR99021, 4 μM) during mesoderm induction result in higher expression of EC-specific genes (VEGFA, TAL1, LMO2, ETV2, PECAM1) compared to high WNT (CHIR99021, 8 μM). (A') Time series diagrams per differentiation formed under no VEGF supplementation, low or high WNT, and activin doses. (B) The proportion of VEGF-expressing CM (unicellular sequences) increases in the optimized ventricular-like CM (low WNT (CHIR: 4 μM) / low activin (4 ng / ml), N=2, n=1717) compared to the intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) condition (N=2, n=5097). (C) Image example of partial EC-lining appearance under the "no activin" condition. Scale bar: 200 μm. (D) Quantification of different EC formation and lining development. N=4, n=29. (E) Image examples of categories 1 and 2. Scale bar: 200 μm. (F) Time series diagram of differentiation performed with added VEGF and intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) doses and WNT inhibition. (F') Quantification of FACS data showing the ratio of CM to EC rigidity in cardioids. Mean + / - SD. (F'') Cardioids show separation of CM and EC layers and appearance of a fibroblast-like cell (COL1A1+) layer. Scale bar: 200 μm. (G) Time series diagram of differentiation performed with VEGF addition, intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) administration, and lack of WNT inhibition. (G') Frozen sections of cardioids containing only EC cells (which do not contain CM) and fibroblast-like cells, when WNT was not inhibited during the cardiac mesoderm stage and VEGF was added. Cell line used in this figure: WTC. [Figure 18-4]Figure 18 shows that WNT, activin, and VEGF regulate the self-organization of endoderm and cardiomyocytes. (A) Low WNT doses (CHIR99021, 4 μM) during mesoderm induction result in higher expression of EC-specific genes (VEGFA, TAL1, LMO2, ETV2, PECAM1) compared to high WNT (CHIR99021, 8 μM). (A') Time series diagrams per differentiation formed under no VEGF supplementation, low or high WNT, and activin doses. (B) The proportion of VEGF-expressing CM (unicellular sequences) increases in the optimized ventricular-like CM (low WNT (CHIR: 4 μM) / low activin (4 ng / ml), N=2, n=1717) compared to the intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) condition (N=2, n=5097). (C) Image example of partial EC-lining appearance under the "no activin" condition. Scale bar: 200 μm. (D) Quantification of different EC formation and lining development. N=4, n=29. (E) Image examples of categories 1 and 2. Scale bar: 200 μm. (F) Time series diagram of differentiation performed with added VEGF and intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) doses and WNT inhibition. (F') Quantification of FACS data showing the ratio of CM to EC rigidity in cardioids. Mean + / - SD. (F'') Cardioids show separation of CM and EC layers and appearance of a fibroblast-like cell (COL1A1+) layer. Scale bar: 200 μm. (G) Time series diagram of differentiation performed with VEGF addition, intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) administration, and lack of WNT inhibition. (G') Frozen sections of cardioids containing only EC cells (which do not contain CM) and fibroblast-like cells, when WNT was not inhibited during the cardiac mesoderm stage and VEGF was added. Cell line used in this figure: WTC. [Figure 18-5]Figure 18 shows that WNT, activin, and VEGF regulate the self-organization of endoderm and cardiomyocytes. (A) Low WNT doses (CHIR99021, 4 μM) during mesoderm induction result in higher expression of EC-specific genes (VEGFA, TAL1, LMO2, ETV2, PECAM1) compared to high WNT (CHIR99021, 8 μM). (A') Time series diagrams per differentiation formed under no VEGF supplementation, low or high WNT, and activin doses. (B) The proportion of VEGF-expressing CM (unicellular sequences) increases in the optimized ventricular-like CM (low WNT (CHIR: 4 μM) / low activin (4 ng / ml), N=2, n=1717) compared to the intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) condition (N=2, n=5097). (C) Image example of partial EC-lining appearance under the "no activin" condition. Scale bar: 200 μm. (D) Quantification of different EC formation and lining development. N=4, n=29. (E) Image examples of categories 1 and 2. Scale bar: 200 μm. (F) Time series diagram of differentiation performed with added VEGF and intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) doses and WNT inhibition. (F') Quantification of FACS data showing the ratio of CM to EC rigidity in cardioids. Mean + / - SD. (F'') Cardioids show separation of CM and EC layers and appearance of a fibroblast-like cell (COL1A1+) layer. Scale bar: 200 μm. (G) Time series diagram of differentiation performed with VEGF addition, intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) administration, and lack of WNT inhibition. (G') Frozen sections of cardioids containing only EC cells (which do not contain CM) and fibroblast-like cells, when WNT was not inhibited during the cardiac mesoderm stage and VEGF was added. Cell line used in this figure: WTC. [Figure 18-6]Figure 18 shows that WNT, activin, and VEGF regulate the self-organization of endoderm and cardiomyocytes. (A) Low WNT doses (CHIR99021, 4 μM) during mesoderm induction result in higher expression of EC-specific genes (VEGFA, TAL1, LMO2, ETV2, PECAM1) compared to high WNT (CHIR99021, 8 μM). (A') Time series diagrams per differentiation formed under no VEGF supplementation, low or high WNT, and activin doses. (B) The proportion of VEGF-expressing CM (unicellular sequences) increases in the optimized ventricular-like CM (low WNT (CHIR: 4 μM) / low activin (4 ng / ml), N=2, n=1717) compared to the intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) condition (N=2, n=5097). (C) Image example of partial EC-lining appearance under the "no activin" condition. Scale bar: 200 μm. (D) Quantification of different EC formation and lining development. N=4, n=29. (E) Image examples of categories 1 and 2. Scale bar: 200 μm. (F) Time series diagram of differentiation performed with added VEGF and intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) doses and WNT inhibition. (F') Quantification of FACS data showing the ratio of CM to EC rigidity in cardioids. Mean + / - SD. (F'') Cardioids show separation of CM and EC layers and appearance of a fibroblast-like cell (COL1A1+) layer. Scale bar: 200 μm. (G) Time series diagram of differentiation performed with VEGF addition, intermediate WNT (CHIR: 6 μM) / high activin (50 ng / ml) administration, and lack of WNT inhibition. (G') Frozen sections of cardioids containing only EC cells (which do not contain CM) and fibroblast-like cells, when WNT was not inhibited during the cardiac mesoderm stage and VEGF was added. Cell line used in this figure: WTC. [Figure 19-1]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 19-2]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 19-3]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 19-4]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 19-5]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 19-6]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 19-7]Figure 19 shows the characteristics of cardioid cells after VEGF treatment. (A) Classification of VEGF-treated cardioid cells (human fetal heart) based on markers determined by Cui et al. (2019). CM: cardiomyocyte, EC: endothelial cell, EP: epicardial cell. Selected markers for endothelial cells (PECAM1, CDH5, NPR3), cardiomyocytes (TNNT2, MYL7), and fibroblast-like cells (COL1A1, COL3A1) are shown. (B, B', B'') Cardioid endothelial cells express mechanosensing genes (e.g., KLF2, SOX18). Scale bar: 200 μm. (C) PCA of 2D-derived ECs (Anterior, Patsch et al., Human Cardiac Microvascular (HCMEC) Human Umbilical Vein (HUVEC)), hPSCs, and 3D-derived ECs (cardioids (Day 7.5, intermediate WNT dose) and vascular organoids (Wimmer et al., 2019) > 18). Anterior ECs: H9 line, Patsch et al., vascular organoid ECs as shown in (Wimmer et al., 2019). (D) Differences in the expression of mechanosensing and maturation genes, cardiac transcription factors, EC transcription factors, endocardial-like and general EC genes in FACS-sorted ECs. Cell lines used in this figure (unless otherwise specified): WTC. [Figure 20-1]Figure 20 shows that the epicardium interacts with cardioids, and fibroblasts derived from the epicardium or endocardium respond to cryo-injury. (A) Schematic diagram of co-culture of the epicardium and cryo-injury. (B) Confocal image of fluorescently labeled epicardium derivatives after 7 days of co-culture with cardioids (TNNT2+) containing EC(PECAM1+) in the presence of 100 ng / ml VEGF-A. Scale bar: 200 μm. (CD) Confocal images of COL1A1+ (C) and ACTA2+ (D) epicardium derivatives after 7 days of co-culture with cardioids containing EC(PECAM1+). Scale bar: 50 μm. (EF) Confocal images of cryo-injury responses (1 hour and 3 days post-injury) in cardioids stained with fibronectin (FN1) and COL1A1+ cells in the presence (E) or absence (F) of the epicardium. Scale bar: 200 μm (E, F) and 50 μm (details of e'-E). (G) Masson trichrome staining of freeze-damaged cardioids with and without epicardium (blue: connective tissue, red: muscle) 1 hour and 3 days after injury. Scale bar: 200 μm. (H) Confocal image of CM+EC+fibroblast codifferentiated cardioid, 3 days after injury, with FN1 and COL1A1+ cells marked. Scale bar: 200 μm. (I) Quantification of COL1A1+ cells in the damaged site (purple) versus healthy site (blue) of CM+EC+fibroblast codifferentiated cardioid 3 days after injury. The number of COL1A1+ cells was normalized by damaged site / total area. In the figure, the dashed line indicates the damaged site. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC cell line as well. Cell line used for CM / EC / fibroblast codifferentiation: WTC. [Figure 20-2]Figure 20 shows that the epicardium interacts with cardioids, and fibroblasts derived from the epicardium or endocardium respond to cryo-injury. (A) Schematic diagram of co-culture of the epicardium and cryo-injury. (B) Confocal image of fluorescently labeled epicardium derivatives after 7 days of co-culture with cardioids (TNNT2+) containing EC(PECAM1+) in the presence of 100 ng / ml VEGF-A. Scale bar: 200 μm. (CD) Confocal images of COL1A1+ (C) and ACTA2+ (D) epicardium derivatives after 7 days of co-culture with cardioids containing EC(PECAM1+). Scale bar: 50 μm. (EF) Confocal images of cryo-injury responses (1 hour and 3 days post-injury) in cardioids stained with fibronectin (FN1) and COL1A1+ cells in the presence (E) or absence (F) of the epicardium. Scale bar: 200 μm (E, F) and 50 μm (details of e'-E). (G) Masson trichrome staining of freeze-damaged cardioids with and without epicardium (blue: connective tissue, red: muscle) 1 hour and 3 days after injury. Scale bar: 200 μm. (H) Confocal image of CM+EC+fibroblast codifferentiated cardioid, 3 days after injury, with FN1 and COL1A1+ cells marked. Scale bar: 200 μm. (I) Quantification of COL1A1+ cells in the damaged site (purple) versus healthy site (blue) of CM+EC+fibroblast codifferentiated cardioid 3 days after injury. The number of COL1A1+ cells was normalized by damaged site / total area. In the figure, the dashed line indicates the damaged site. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC cell line as well. Cell line used for CM / EC / fibroblast codifferentiation: WTC. [Figure 21-1]Figure 21 shows the generation and functional characterization of ventricular-like structures. Related to Figure 1. A) An example of a pulsatile ventricular-like structure containing a large cavity (day 8.5) obtained by adding laminin LN511 (0.1 μg / ml) to the cell suspension during seeding prior to differentiation in a two-dimensional planar bottom well. Scale bar: 100 μm. (B) Pulsatile three-dimensional structures containing cavities were formed by CM differentiation in an ultra-low adhesion plate, even in the absence of exogenous ECM molecules ("no ECM" column). Measurement time: day 8.5. Scale bar: 500 μm. (C) Expression of several CM-specific markers in cardioid myocardium (day 7.5). Scale bar: 200 μm. (C') Shows sarcomerial tissue within cardioid CM. (C) Details of (TTN-GFP / ACTN2). Scale bar: 6 μm. (D) Electron micrograph of cardioid at day 7.5. S is sarcomere, ID is intercalated disk, Z is Z-line, and D is desmosome. Scale bar: 1 μm. (E) Immunostaining with CM markers showing robust cardioid formation using various hPSC lines (hESC H7)(hiPSC176 / 177 / 178). Scale bar: 200 μm. (E') Whole-mount bright-field images of several technical replicas of the cardioids shown in (E). Scale bar: 200 μm. (F) Quantification of overlap between DAPI and TNNI1± regions in optical sections capturing 500 μm of organoids. N=3, n=8. Cell line used in all experiments in this figure: WTC (unless otherwise noted). [Figure 21-2]Figure 21 shows the generation and functional characterization of ventricular-like structures. Related to Figure 1. A) An example of a pulsatile ventricular-like structure containing a large cavity (day 8.5) obtained by adding laminin LN511 (0.1 μg / ml) to the cell suspension during seeding prior to differentiation in a two-dimensional planar bottom well. Scale bar: 100 μm. (B) Pulsatile three-dimensional structures containing cavities were formed by CM differentiation in an ultra-low adhesion plate, even in the absence of exogenous ECM molecules ("no ECM" column). Measurement time: day 8.5. Scale bar: 500 μm. (C) Expression of several CM-specific markers in cardioid myocardium (day 7.5). Scale bar: 200 μm. (C') Shows sarcomerial tissue within cardioid CM. (C) Details of (TTN-GFP / ACTN2). Scale bar: 6 μm. (D) Electron micrograph of cardioid at day 7.5. S is sarcomere, ID is intercalated disk, Z is Z-line, and D is desmosome. Scale bar: 1 μm. (E) Immunostaining with CM markers showing robust cardioid formation using various hPSC lines (hESC H7)(hiPSC176 / 177 / 178). Scale bar: 200 μm. (E') Whole-mount bright-field images of several technical replicas of the cardioids shown in (E). Scale bar: 200 μm. (F) Quantification of overlap between DAPI and TNNI1± regions in optical sections capturing 500 μm of organoids. N=3, n=8. Cell line used in all experiments in this figure: WTC (unless otherwise noted). [Figure 22-1]Figure 22 shows the generation and functional characterization of ventricular-like structures. Related to Figure 1. (A) Heatmap showing gene expression of cardiomyocyte structure genes / ion channels in cardioids on day 7 / d10 / d27. (B) Comparison of gene expression signatures at each time point between cardioids, 3D assembled CMs, 2D CMs, and hPSCs. Cell lines used: H9, WTC. (C) GO terms whose expression increased in cardioids on day 7 compared to 2D CMs on day 7. Cell lines used. (D) GO terms whose expression increased between cardioids on day 7 and 3D aggregated CMs on day 27. Cell lines used: H9, WTC. E) Ventricular-like structures and 2D CMs show similar calcium transients. Calcium transients were identified by loading cells with Fluo-4-AM and analyzing fluorescence intensity over time. F / F0: Fluorescence intensity relative to background. ISI: Interspike interval. (F) Analysis of pulse parameters using a published algorithm (Huebsch et al., 2015, Tissue Engineering Part C: Methods 21, 467-479). (G) Representative mean action potential traces (d10) of ventricular cardioids, labeled APD30, APD50, and APD90. The organoids contracted at a mean frequency (dT) of 7.33 seconds. The mean was taken over 14 action potentials from one organoid using the FluoVolt system. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 22-2]Figure 22 shows the generation and functional characterization of ventricular-like structures. Related to Figure 1. (A) Heatmap showing gene expression of cardiomyocyte structure genes / ion channels in cardioids on day 7 / d10 / d27. (B) Comparison of gene expression signatures at each time point between cardioids, 3D assembled CMs, 2D CMs, and hPSCs. Cell lines used: H9, WTC. (C) GO terms whose expression increased in cardioids on day 7 compared to 2D CMs on day 7. Cell lines used. (D) GO terms whose expression increased between cardioids on day 7 and 3D aggregated CMs on day 27. Cell lines used: H9, WTC. E) Ventricular-like structures and 2D CMs show similar calcium transients. Calcium transients were identified by loading cells with Fluo-4-AM and analyzing fluorescence intensity over time. F / F0: Fluorescence intensity relative to background. ISI: Interspike interval. (F) Analysis of pulse parameters using a published algorithm (Huebsch et al., 2015, Tissue Engineering Part C: Methods 21, 467-479). (G) Representative mean action potential traces (d10) of ventricular cardioids, labeled APD30, APD50, and APD90. The organoids contracted at a mean frequency (dT) of 7.33 seconds. The mean was taken over 14 action potentials from one organoid using the FluoVolt system. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 22-3]Figure 22 shows the generation and functional characterization of ventricular-like structures. Related to Figure 1. (A) Heatmap showing gene expression of cardiomyocyte structure genes / ion channels in cardioids on day 7 / d10 / d27. (B) Comparison of gene expression signatures at each time point between cardioids, 3D assembled CMs, 2D CMs, and hPSCs. Cell lines used: H9, WTC. (C) GO terms whose expression increased in cardioids on day 7 compared to 2D CMs on day 7. Cell lines used. (D) GO terms whose expression increased between cardioids on day 7 and 3D aggregated CMs on day 27. Cell lines used: H9, WTC. E) Ventricular-like structures and 2D CMs show similar calcium transients. Calcium transients were identified by loading cells with Fluo-4-AM and analyzing fluorescence intensity over time. F / F0: Fluorescence intensity relative to background. ISI: Interspike interval. (F) Analysis of pulse parameters using a published algorithm (Huebsch et al., 2015, Tissue Engineering Part C: Methods 21, 467-479). (G) Representative mean action potential traces (d10) of ventricular cardioids, labeled APD30, APD50, and APD90. The organoids contracted at a mean frequency (dT) of 7.33 seconds. The mean was taken over 14 action potentials from one organoid using the FluoVolt system. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 22-4]Figure 22 shows the generation and functional characterization of ventricular-like structures. Related to Figure 1. (A) Heatmap showing gene expression of cardiomyocyte structure genes / ion channels in cardioids on day 7 / d10 / d27. (B) Comparison of gene expression signatures at each time point between cardioids, 3D assembled CMs, 2D CMs, and hPSCs. Cell lines used: H9, WTC. (C) GO terms whose expression increased in cardioids on day 7 compared to 2D CMs on day 7. Cell lines used. (D) GO terms whose expression increased between cardioids on day 7 and 3D aggregated CMs on day 27. Cell lines used: H9, WTC. E) Ventricular-like structures and 2D CMs show similar calcium transients. Calcium transients were identified by loading cells with Fluo-4-AM and analyzing fluorescence intensity over time. F / F0: Fluorescence intensity relative to background. ISI: Interspike interval. (F) Analysis of pulse parameters using a published algorithm (Huebsch et al., 2015, Tissue Engineering Part C: Methods 21, 467-479). (G) Representative mean action potential traces (d10) of ventricular cardioids, labeled APD30, APD50, and APD90. The organoids contracted at a mean frequency (dT) of 7.33 seconds. The mean was taken over 14 action potentials from one organoid using the FluoVolt system. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 23]Figure 23 shows the characterization of cardioids and chicken cardiac mesoderm. Related to Figure 2. (A) Self-organizing cardiac mesoderm at day 2.5 showing the absence of SOX17+ and EOMES+ endoderm. Scale bar: 200 μm. (B) Treatment with sunitinib (100 nM) leads to the formation of cavities in cardioids (day 7.5) in the absence of endothelial cells (PECAM1+). Scale bar: 200 μm. (C) Chicken cardiac mesoderm exoplants form a spherical structure with cavities in vitro when cultured in cardiac mesoderm medium. Scale bar: 200 μm. (D) Control chicken cardiac mesoderm grafts (same structure as Figure 2B) show almost no foregut marker SOX2. Scale bar: 200 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 24-1] Figure 24 shows the characterization of NKX2-5 and HAND1 KO hPSCs and cardioids. Related to Figure 4. (A) NKX2-5 KO cardioids do not show dilation of damaged cavities. Scale bar: 200 μm. (B) Quantification of cardioid diameter confirms no decrease in NKX2-5 KO cardioids at day 3.5. (C) Immunostaining results for HAND1, NKX2-5, and TNNT2 in WT and NKX2-5 KO cardioids at days 3.5 and 7.5. (D) Immunostaining results for NKX2-5 in WT and HAND1 KO organoids at day 3.5. (E) Close-up of immunostaining results of cardioids at day 7.5 showing the presence of CM (NKX2-5+ / TNNT2+) in HAND1 KO cardioids. (F) Quantification of cell number / cardioid, with no significant difference between WT (N=3) and HAND1 KO (N=3). (G) Immunostaining results (day 3.5) of cardioid sections (48 WT / 47 HAND1 KO). Shows impaired cavity formation, as well as lack of HAND1 protein expression in HAND1 KO cardioids. Scale bar: 2000 μm. Cell lines used in all experiments in this figure: HAND1 KO (H9), WT (H9), NKX2-5 KO (H9). [Figure 24-2]Figure 24 shows the characterization of NKX2-5 and HAND1 KO hPSCs and cardioids. Related to Figure 4. (A) NKX2-5 KO cardioids do not show dilation of damaged cavities. Scale bar: 200 μm. (B) Quantification of cardioid diameter confirms no decrease in NKX2-5 KO cardioids at day 3.5. (C) Immunostaining results for HAND1, NKX2-5, and TNNT2 in WT and NKX2-5 KO cardioids at days 3.5 and 7.5. (D) Immunostaining results for NKX2-5 in WT and HAND1 KO organoids at day 3.5. (E) Close-up of immunostaining results of cardioids at day 7.5 showing the presence of CM (NKX2-5+ / TNNT2+) in HAND1 KO cardioids. (F) Quantification of cell number / cardioid, with no significant difference between WT (N=3) and HAND1 KO (N=3). (G) Immunostaining results (day 3.5) of cardioid sections (48 WT / 47 HAND1 KO). Shows impaired cavity formation, as well as lack of HAND1 protein expression in HAND1 KO cardioids. Scale bar: 2000 μm. Cell lines used in all experiments in this figure: HAND1 KO (H9), WT (H9), NKX2-5 KO (H9). [Figure 25-1] Figure 25 shows further examples of the extent of EC self-organization. (A) Time series of differentiation under VEGF-free, low WNT, activin-free, or low activin conditions. (A') Image examples of categories 1 and 2 under low WNT / low activin conditions. Scale bar: 200 μm. (B) Time series of differentiation when VEGF was added after CM designation under low WNT / high activin conditions. (C), (D) When VEGF was added after CM formation, ECs sometimes partially covered the cavities, but in many cases, ECs also partially formed the outer layer. Scale bar: 200 μm. Cell line used in (D): H9. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 25-2]Figure 25 shows further examples of the extent of EC self-organization. (A) Time series of differentiation under VEGF-free, low WNT, activin-free, or low activin conditions. (A') Image examples of categories 1 and 2 under low WNT / low activin conditions. Scale bar: 200 μm. (B) Time series of differentiation when VEGF was added after CM designation under low WNT / high activin conditions. (C), (D) When VEGF was added after CM formation, ECs sometimes partially covered the cavities, but in many cases, ECs also partially formed the outer layer. Scale bar: 200 μm. Cell line used in (D): H9. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-1]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-2]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-3]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-4]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-5]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-6]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-7]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 26-8]Figure 26 shows the characterization of cardioids with endothelial and fibroblast-like cell layers. (A) Three types of biological replication that reliably generate cardioids with separate CM and EC layers using an intermediate WNT dose (CHIR99021, 6 μM). Scale bar: 2000 μm. (B) The EC in aggregated CM microtissue forms a rudimentary network around the cardiomyocytes, rather than being a separate layer. Scale bar: 400 μm. Cell line used: H9. (C) The EC forms a network layer surrounding the cardioid. Maximum Int. projection. Frozen sections show that FN1 is expressed in all three cell types, and VIM is mainly expressed in EC and fibroblast-like cells. Scale bar: 200 μm. (D) An example of a FACS plot showing the distribution of cardiomyocytes (CM) and (EC) in codifferentiated cardioids at day 7.5 using the MYL7-GFP / CDH5-Tomato iPSC (WTC) strain. Control: WT WTC line. (E) Heatmap of Smart-Seq2 data from sorted hPSC, CM, EC, and Non-EC / Non-CM (fibroblast-like) cells, showing the expression of major genes in each. Neither fibroblast-like cells express genes associated with putative EC-derived fibroblast-like cells, as can be confirmed by the staining results in (C). (F) Three biological replicas demonstrate reproducibility of cardioid formation without CM contribution (MYL7-, CDH5+). Scale bar: 500 μm. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 27-1]Figure 27 shows further characterization of cardioid cell identity. (A) MuSiC deconvolution of bulk RNA-seq data of cardiac organoids (Wang et al., 2019, Nature Communications 10, 380-389) shows CM / EC contributions similar to FACS quantification and the absence of EC in cardioids with only CM. Reference dataset: Single-cell RNA-seq of human cardiac developmental trajectory (Cui et al., 2019, Cell Rep 26, 1934-1950.e1935). Cell line used: H9, WTC. B) Proteomic analysis of cardioids shows that major markers of CM and EC are detected at the protein level. (C) HOX gene expression analysis of cardioid ECs, HCMECs, and anterior ECs shows expression of the anterior HOX gene, and that the posterior HOX gene is also expressed in other ECs. VST: Dispersion-stabilized conversion. Anterior ECs. Vascular organoid ECs shown in the H9 line, Wimmer et al., 2019, Nature 565, 505-510. (D) GO terms of genes upregulated in CM indicate terms related to cardiomyocyte development. (E) GO terms of genes upregulated in cardiomyocyte ECs are shown. (F) GO terms of genes expressed in fibroblast-like cells (GFP- / Tomato-) indicate terms related to connective tissue formation and ECM composition. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 27-2]Figure 27 shows further characterization of cardioid cell identity. (A) MuSiC deconvolution of bulk RNA-seq data of cardiac organoids (Wang et al., 2019, Nature Communications 10, 380-389) shows CM / EC contributions similar to FACS quantification and the absence of EC in cardioids with only CM. Reference dataset: Single-cell RNA-seq of human cardiac developmental trajectory (Cui et al., 2019, Cell Rep 26, 1934-1950.e1935). Cell line used: H9, WTC. B) Proteomic analysis of cardioids shows that major markers of CM and EC are detected at the protein level. (C) HOX gene expression analysis of cardioid ECs, HCMECs, and anterior ECs shows expression of the anterior HOX gene, and that the posterior HOX gene is also expressed in other ECs. VST: Dispersion-stabilized conversion. Anterior ECs. Vascular organoid ECs shown in the H9 line, Wimmer et al., 2019, Nature 565, 505-510. (D) GO terms of genes upregulated in CM indicate terms related to cardiomyocyte development. (E) GO terms of genes upregulated in cardiomyocyte ECs are shown. (F) GO terms of genes expressed in fibroblast-like cells (GFP- / Tomato-) indicate terms related to connective tissue formation and ECM composition. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 27-3]Figure 27 shows further characterization of cardioid cell identity. (A) MuSiC deconvolution of bulk RNA-seq data of cardiac organoids (Wang et al., 2019, Nature Communications 10, 380-389) shows CM / EC contributions similar to FACS quantification and the absence of EC in cardioids with only CM. Reference dataset: Single-cell RNA-seq of human cardiac developmental trajectory (Cui et al., 2019, Cell Rep 26, 1934-1950.e1935). Cell line used: H9, WTC. B) Proteomic analysis of cardioids shows that major markers of CM and EC are detected at the protein level. (C) HOX gene expression analysis of cardioid ECs, HCMECs, and anterior ECs shows expression of the anterior HOX gene, and that the posterior HOX gene is also expressed in other ECs. VST: Dispersion-stabilized conversion. Anterior ECs. Vascular organoid ECs shown in the H9 line, Wimmer et al., 2019, Nature 565, 505-510. (D) GO terms of genes upregulated in CM indicate terms related to cardiomyocyte development. (E) GO terms of genes upregulated in cardiomyocyte ECs are shown. (F) GO terms of genes expressed in fibroblast-like cells (GFP- / Tomato-) indicate terms related to connective tissue formation and ECM composition. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 27-4]Figure 27 shows further characterization of cardioid cell identity. (A) MuSiC deconvolution of bulk RNA-seq data of cardiac organoids (Wang et al., 2019, Nature Communications 10, 380-389) shows CM / EC contributions similar to FACS quantification and the absence of EC in cardioids with only CM. Reference dataset: Single-cell RNA-seq of human cardiac developmental trajectory (Cui et al., 2019, Cell Rep 26, 1934-1950.e1935). Cell line used: H9, WTC. B) Proteomic analysis of cardioids shows that major markers of CM and EC are detected at the protein level. (C) HOX gene expression analysis of cardioid ECs, HCMECs, and anterior ECs shows expression of the anterior HOX gene, and that the posterior HOX gene is also expressed in other ECs. VST: Dispersion-stabilized conversion. Anterior ECs. Vascular organoid ECs shown in the H9 line, Wimmer et al., 2019, Nature 565, 505-510. (D) GO terms of genes upregulated in CM indicate terms related to cardiomyocyte development. (E) GO terms of genes upregulated in cardiomyocyte ECs are shown. (F) GO terms of genes expressed in fibroblast-like cells (GFP- / Tomato-) indicate terms related to connective tissue formation and ECM composition. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 27-5]Figure 27 shows further characterization of cardioid cell identity. (A) MuSiC deconvolution of bulk RNA-seq data of cardiac organoids (Wang et al., 2019, Nature Communications 10, 380-389) shows CM / EC contributions similar to FACS quantification and the absence of EC in cardioids with only CM. Reference dataset: Single-cell RNA-seq of human cardiac developmental trajectory (Cui et al., 2019, Cell Rep 26, 1934-1950.e1935). Cell line used: H9, WTC. B) Proteomic analysis of cardioids shows that major markers of CM and EC are detected at the protein level. (C) HOX gene expression analysis of cardioid ECs, HCMECs, and anterior ECs shows expression of the anterior HOX gene, and that the posterior HOX gene is also expressed in other ECs. VST: Dispersion-stabilized conversion. Anterior ECs. Vascular organoid ECs shown in the H9 line, Wimmer et al., 2019, Nature 565, 505-510. (D) GO terms of genes upregulated in CM indicate terms related to cardiomyocyte development. (E) GO terms of genes upregulated in cardiomyocyte ECs are shown. (F) GO terms of genes expressed in fibroblast-like cells (GFP- / Tomato-) indicate terms related to connective tissue formation and ECM composition. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 27-6]Figure 27 shows further characterization of cardioid cell identity. (A) MuSiC deconvolution of bulk RNA-seq data of cardiac organoids (Wang et al., 2019, Nature Communications 10, 380-389) shows CM / EC contributions similar to FACS quantification and the absence of EC in cardioids with only CM. Reference dataset: Single-cell RNA-seq of human cardiac developmental trajectory (Cui et al., 2019, Cell Rep 26, 1934-1950.e1935). Cell line used: H9, WTC. B) Proteomic analysis of cardioids shows that major markers of CM and EC are detected at the protein level. (C) HOX gene expression analysis of cardioid ECs, HCMECs, and anterior ECs shows expression of the anterior HOX gene, and that the posterior HOX gene is also expressed in other ECs. VST: Dispersion-stabilized conversion. Anterior ECs. Vascular organoid ECs shown in the H9 line, Wimmer et al., 2019, Nature 565, 505-510. (D) GO terms of genes upregulated in CM indicate terms related to cardiomyocyte development. (E) GO terms of genes upregulated in cardiomyocyte ECs are shown. (F) GO terms of genes expressed in fibroblast-like cells (GFP- / Tomato-) indicate terms related to connective tissue formation and ECM composition. Cell line used in all experiments in this figure: WTC (unless otherwise specified). [Figure 28-1]Figure 28 shows that the epicardium interacts with cardioids, and epicardium or endocardium-derived fibroblasts respond to freeze damage. (A) Schematic diagram of the embryologically aligned protocol. (B) Differences in pluripotency, mesoderm-specific, cardiac mesoderm, and epicardium-specific gene expression comparing 2D replating and 3D aggregated pericardium at day 8.5 (complete differentiation). Cell line used: H9. (C) Whowell WT1 immunostaining results of 2D replated epicardium. Scale bar: 1 mm. 1×1 mm detailed view of C'C. Cell line used: H9. (D) Differentiation expression of growth factors involved in epicardium sublineage identification in pluripotency in cardioid cells at days 7 / 10 / 27 without co-culture. Cell line used: WTC. (E) Fluorescently labeled 3D aggregates of 2D epicardium differentiation co-cultured with time-matched cardioids. Arrows indicate morphological changes and expansion over time. Scale bar: 200 μm. (FG) Confocal images of TNNT2 (CM) and WT1 (epidimal) co-cultures of epicardium and cardioid at 2 and 7 days. Scale bar: 200 μm. (HI) Detailed overall view of cardioid in Figures 7C-D, scale bar: 200 μm. (JK) Confocal image of CM+EC+fibroblast co-differentiated cardioid, marking cell death (TUNEL, Cleaved Caspase 3) and lack of cell proliferation (MKI67) 3 days after freeze injury. Scale bar: 200 μm. (L) An example of the region used for normalization between healthy and injured individuals in the quantification shown in Figure 7I. The dashed line indicates the injured area in the figure. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC. Cell line used for CM+EC+fibroblast co-differentiation: WTC. [Figure 28-2]Figure 28 shows that the epicardium interacts with cardioids, and epicardium or endocardium-derived fibroblasts respond to freeze damage. (A) Schematic diagram of the embryologically aligned protocol. (B) Differences in pluripotency, mesoderm-specific, cardiac mesoderm, and epicardium-specific gene expression comparing 2D replating and 3D aggregated pericardium at day 8.5 (complete differentiation). Cell line used: H9. (C) Whowell WT1 immunostaining results of 2D replated epicardium. Scale bar: 1 mm. 1×1 mm detailed view of C'C. Cell line used: H9. (D) Differentiation expression of growth factors involved in epicardium sublineage identification in pluripotency in cardioid cells at days 7 / 10 / 27 without co-culture. Cell line used: WTC. (E) Fluorescently labeled 3D aggregates of 2D epicardium differentiation co-cultured with time-matched cardioids. Arrows indicate morphological changes and expansion over time. Scale bar: 200 μm. (FG) Confocal images of TNNT2 (CM) and WT1 (epidimal) co-cultures of epicardium and cardioid at 2 and 7 days. Scale bar: 200 μm. (HI) Detailed overall view of cardioid in Figures 7C-D, scale bar: 200 μm. (JK) Confocal image of CM+EC+fibroblast co-differentiated cardioid, marking cell death (TUNEL, Cleaved Caspase 3) and lack of cell proliferation (MKI67) 3 days after freeze injury. Scale bar: 200 μm. (L) An example of the region used for normalization between healthy and injured individuals in the quantification shown in Figure 7I. The dashed line indicates the injured area in the figure. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC. Cell line used for CM+EC+fibroblast co-differentiation: WTC. [Figure 28-3]Figure 28 shows that the epicardium interacts with cardioids, and epicardium or endocardium-derived fibroblasts respond to freeze damage. (A) Schematic diagram of the embryologically aligned protocol. (B) Differences in pluripotency, mesoderm-specific, cardiac mesoderm, and epicardium-specific gene expression comparing 2D replating and 3D aggregated pericardium at day 8.5 (complete differentiation). Cell line used: H9. (C) Whowell WT1 immunostaining results of 2D replated epicardium. Scale bar: 1 mm. 1×1 mm detailed view of C'C. Cell line used: H9. (D) Differentiation expression of growth factors involved in epicardium sublineage identification in pluripotency in cardioid cells at days 7 / 10 / 27 without co-culture. Cell line used: WTC. (E) Fluorescently labeled 3D aggregates of 2D epicardium differentiation co-cultured with time-matched cardioids. Arrows indicate morphological changes and expansion over time. Scale bar: 200 μm. (FG) Confocal images of TNNT2 (CM) and WT1 (epidimal) co-cultures of epicardium and cardioid at 2 and 7 days. Scale bar: 200 μm. (HI) Detailed overall view of cardioid in Figures 7C-D, scale bar: 200 μm. (JK) Confocal image of CM+EC+fibroblast co-differentiated cardioid, marking cell death (TUNEL, Cleaved Caspase 3) and lack of cell proliferation (MKI67) 3 days after freeze injury. Scale bar: 200 μm. (L) An example of the region used for normalization between healthy and injured individuals in the quantification shown in Figure 7I. The dashed line indicates the injured area in the figure. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC. Cell line used for CM+EC+fibroblast co-differentiation: WTC. [Figure 28-4]Figure 28 shows that the epicardium interacts with cardioids, and epicardium or endocardium-derived fibroblasts respond to freeze damage. (A) Schematic diagram of the embryologically aligned protocol. (B) Differences in pluripotency, mesoderm-specific, cardiac mesoderm, and epicardium-specific gene expression comparing 2D replating and 3D aggregated pericardium at day 8.5 (complete differentiation). Cell line used: H9. (C) Whowell WT1 immunostaining results of 2D replated epicardium. Scale bar: 1 mm. 1×1 mm detailed view of C'C. Cell line used: H9. (D) Differentiation expression of growth factors involved in epicardium sublineage identification in pluripotency in cardioid cells at days 7 / 10 / 27 without co-culture. Cell line used: WTC. (E) Fluorescently labeled 3D aggregates of 2D epicardium differentiation co-cultured with time-matched cardioids. Arrows indicate morphological changes and expansion over time. Scale bar: 200 μm. (FG) Confocal images of TNNT2 (CM) and WT1 (epidimal) co-cultures of epicardium and cardioid at 2 and 7 days. Scale bar: 200 μm. (HI) Detailed overall view of cardioid in Figures 7C-D, scale bar: 200 μm. (JK) Confocal image of CM+EC+fibroblast co-differentiated cardioid, marking cell death (TUNEL, Cleaved Caspase 3) and lack of cell proliferation (MKI67) 3 days after freeze injury. Scale bar: 200 μm. (L) An example of the region used for normalization between healthy and injured individuals in the quantification shown in Figure 7I. The dashed line indicates the injured area in the figure. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC. Cell line used for CM+EC+fibroblast co-differentiation: WTC. [Figure 28-5]Figure 28 shows that the epicardium interacts with cardioids, and epicardium or endocardium-derived fibroblasts respond to freeze damage. (A) Schematic diagram of the embryologically aligned protocol. (B) Differences in pluripotency, mesoderm-specific, cardiac mesoderm, and epicardium-specific gene expression comparing 2D replating and 3D aggregated pericardium at day 8.5 (complete differentiation). Cell line used: H9. (C) Whowell WT1 immunostaining results of 2D replated epicardium. Scale bar: 1 mm. 1×1 mm detailed view of C'C. Cell line used: H9. (D) Differentiation expression of growth factors involved in epicardium sublineage identification in pluripotency in cardioid cells at days 7 / 10 / 27 without co-culture. Cell line used: WTC. (E) Fluorescently labeled 3D aggregates of 2D epicardium differentiation co-cultured with time-matched cardioids. Arrows indicate morphological changes and expansion over time. Scale bar: 200 μm. (FG) Confocal images of TNNT2 (CM) and WT1 (epidimal) co-cultures of epicardium and cardioid at 2 and 7 days. Scale bar: 200 μm. (HI) Detailed overall view of cardioid in Figures 7C-D, scale bar: 200 μm. (JK) Confocal image of CM+EC+fibroblast co-differentiated cardioid, marking cell death (TUNEL, Cleaved Caspase 3) and lack of cell proliferation (MKI67) 3 days after freeze injury. Scale bar: 200 μm. (L) An example of the region used for normalization between healthy and injured individuals in the quantification shown in Figure 7I. The dashed line indicates the injured area in the figure. Cell line used for the epicardium figure: H9. Experiments were repeated with WTC. Cell line used for CM+EC+fibroblast co-differentiation: WTC. [Figure 29] Figure 29 shows that a damaged cardioid reproduces the response at the time of injury. [Figure 30] Figure 30 shows that necrosis of freeze-damaged cardioids, as indicated by TUNEL staining, reproduces aspects of the injury response. [Figure 31] Figure 31 shows no increase in proliferation at the site of injury - endothelial cells outside / part of the site of injury. [Figure 32] Figure 32 shows the increase in the number of fibroblasts in the injured area. [Figure 33] Figure 33 shows that FN1 is accumulating at the site of injury. [Figure 34] Figure 34 shows the quantification of the average fibronectin intensity value. [Modes for carrying out the invention]

[0017] The present invention provides a self-organizing myocardial tissue or organoid model capable of generating a single large lumen reminiscent of the development of an early left heart in animals and humans, such as the ventricles and atrioventricular regions. Such tissues are shown, for example, in Figures 1, 2, and 14. The present invention's method for generating the tissue model utilizes native growth factors and signaling factors. This generation does not require an exogenous matrix scaffold.

[0018] Accordingly, the present invention provides a cardiac tissue model having at least 60%, preferably at least 80%, cardiac cells, independently of any additional cells of the vascular tissue system of the tissue model, where cardiac cells surround the lumen and are selected from cardiomyocytes, endocardial cells (also called cardiac endothelial cells) and epicardial cells (e.g., Figure 6, 11G). "Selected from" means that the selected species can be any one of the members grouped as selectable species. The cardiac tissue model may have at least 60%, preferably at least 80%, cardiac cells, particularly when confirming step c) of the method of the present invention. It is possible to add additional cells infiltrating the tissue model, particularly cells of the vascular tissue system, which may not be counted in the figure of at least 60%, preferably at least 80%, cardiac cells. The features of the present invention remain unchanged regardless of these infiltrating cells, as long as the infiltrating cells do not replace the cardiac cells and impair or render the functions of the cardiac model, such as beating, impaired. In a preferred embodiment, the cardiac tissue model includes at least 50%, preferably at least 60%, cardiac cells selected from cardiac cells, cardiomyocytes, endocardial cells, and epicardial cells (absolute cell count, further infiltrating cells are also counted).

[0019] In this specification, the term “contains” is open to the presence of any further components unless otherwise specified. With respect to compositions such as the tissues of the present invention that are specified to contain a cell-like component in an amount defined by a numerical range, a proviso is necessary to exclude the presence of that component in amounts outside that range.

[0020] In particular, the tissue model of the present invention contains almost all cells of the cardiomyocyte lineage, for example, at least 60%, preferably at least 70%, particularly preferably at least 80%, and even more preferably at least 90% of the cells of the cardiomyocyte lineage. The tissue model may also contain lateral plate mesoderm (HAND1+) cells, which are precursors to all three major cardiomyocyte lineages. For example, the tissue model may contain at least 0.5% lateral plate mesoderm cells and / or up to 30% lateral plate mesoderm cells, for example, 1% to 20% lateral plate mesoderm cells. In particular, the tissue model may contain no foregut endoderm cells or up to 5% or less, for example, up to 3%, up to 1%, or up to 0.1% foregut endoderm cells. Preferably, the cardiac tissue model of the present invention contains up to 3% or none of foregut endoderm-derived cells (e.g., expressing SOX17+ and / or EOMES+), and / or up to 3% or none of hematopoietic cells, preferably up to 1% and / or up to 3% of epithelial-derived cells (SOX2+) (e.g., Figures 2 and 3) or none.

[0021] For example, a characteristic of the cardiac tissue model of the present invention is that it possesses rhythmic pulsation activity, similar to that of a naturally growing heart in its early stages of development.

[0022] Cardiac cells can be selected from several cardiomyocyte types, including cardiomyocytes, endocardial cells, and epicardial cells. The composition of these cell types may vary based on any different treatments as disclosed herein. According to most options, the cardiac tissue model consists of at least 40% cardiomyocytes. In a particularly preferred embodiment, the cardiac tissue model contains at least 50%, more preferably at least 60%, and especially preferably at least 80% cardiomyocytes. In special embodiments, the number may even reach at least 90% cardiomyocytes (e.g., Figures 4, 6, and 10).

[0023] In a more preferred embodiment of the present invention, the cardiac tissue model comprises at least 2%, preferably at least 5%, and more preferably at least 8% endocardial cells. Endocardial cells are cardiac endothelial cells that form the endocardium in vivo. They are endothelial cells of the cardiomyocyte. Endocardial cells form compartments / liners on the inside (facing the lumen), and possibly the outside, of the tissue model of the present invention. Such cells can be identified by the expression markers NPR3, NFATC1, HOX1, HOX2, HOX3, HOX4, and HOX5. Importantly, these endothelial cells (ECs) most closely resemble those in the cardiomyocytes in vivo (i.e., ECs of a real heart) (e.g., Figure 4). They express higher levels of NPR3 and NFATC1 compared to non-cardiac control ECs. In this specification, all cell comparisons are made under standard cell culture conditions for maintaining cells that express relevant characteristic markers or signature markers, in particular, so that the expression of markers or markers can be determined as cell-specific properties; such conditions typically include a culture medium containing all the nutrients necessary to maintain cells under ambient pressure and at physiological temperature. Furthermore, the endocardial cells (ECs) of the present invention express the correct HOX genes (HOX1-5) that are typical for the heart (e.g., Figure 4). Activation of mechanosensing genes (SOX18, KLF2, CDH5, FOS, TEK, FOXO1) is also observed (Figures 4, 19, etc.). This is considered to be a difference from other non-self-organizing organoids, where ECs are simply mixed with cardiomyocytes. The expression of mechanosensing genes is an important biological feature of more functional ECs. Therefore, the endocardial cells of the present invention are more functional than ECs in conventional two-dimensional and three-dimensional models.

[0024] According to the present invention, the amount of endocardial cells can be controlled during the proliferation of a tissue model. The tissue model can have various amounts of endocardial cells, such as at least 2%, preferably at least 4%, or at least 8%, more preferably at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, or at least 40%. One option for controlling the formation of endocardial cells is to use VEGF (vascular endothelial growth factor) when creating the tissue model. Without VEGF, endocardial cells are usually present in small amounts, less than 10%, and are particularly formed as the inner lining. These autologous endocardial cells express endothelial markers such as CD31 and CDH5 (e.g., Figures 8 and 19).

[0025] The tissue model of the present invention is essentially hollow in shape, preferably with a continuous layer of endocardial cells surrounding the lumen, and preferably facing the lumen (e.g., Figures 2, 8, 11G). The continuous layer preferably has no pores and completely surrounds the lumen in all directions when viewed from the center of the lumen.

[0026] The tissue model of the present invention has a structured composition, and certain types of cells are typically found in specific compartments. Such compartments are usually layers. Preferably, the cardiac tissue model of the present invention includes cardiomyocytes and endocardial cells in the kit. Endocardial cells may be located in different tissue layers, preferably the inner layer and / or outer layer (e.g., Figures 4, 7, and 8).

[0027] In a preferred embodiment of the present invention, a gap exists between the endothelial cell compartment, preferably a layer, and the cardiomyocyte compartment, preferably a layer, as can be observed in vivo (cardiac jelly) at this stage of development (e.g., Figure 7).

[0028] In a further preferred embodiment, the cardiac tissue model of the present invention comprises at least 50%, more preferably at least 60%, at least 70%, at least 80%, at least 85%, at least 90%, or at least 95% cardiomyocytes (e.g., Figures 4 and 6).

[0029] Cardiac muscle cells are responsible for the majority of the beating in the tissue model, and are the main component of the tissue model of this invention, forming the layer surrounding the lumen in the tissue model. In early tissue, cardiomyocytes directly face the lumen, but in later, more developed tissue, the inner layer is formed by endocardial cells (for example, Figures 2 and 8).

[0030] The tissue model of the present invention is essentially hollow, and preferably, a continuous layer of cardiomyocytes surrounds the lumen. The continuous layer preferably has no pores and completely surrounds the lumen in all directions when viewed from the center of the lumen.

[0031] One important aspect of the tissue model of the present invention is that it develops with highly purified cardiac cells. Non-cardiac cells such as ectoderm cells, endoderm cells, and hematopoietic endothelial cells are either absent (e.g., Figures 2 and 3) or, if added later to the tissue model after development according to the present invention, up to the step of differentiating mesoderm cells into cardiac cells, are preferably present in very small amounts. Therefore, in a preferred embodiment, the tissue model of the present invention contains no endoderm cells at all, or contains a maximum of 5%, preferably a maximum of 2%, or a maximum of 1%. Furthermore, or in combination therewith, in a preferred embodiment, the tissue model of the present invention contains no hematopoietic endothelial cells, or contains a maximum of 5%, preferably a maximum of 2%, or a maximum of 1%.

[0032] One cell type that is preferable to add to this tissue model later is epicardial cells. These cells can form a layer surrounding the tissue model of the present invention. Epicardial cells preferably emit the cell markers WT1 and TCF21 (e.g., Figure 10). The tissue model of the present invention may have at least 0.5%, preferably at least 5%, or at least 10%, epicardial cells, for example, 0.5% to 30% epicardial cells (e.g., Figure 11).

[0033] In preferred embodiments of the present invention, the cardiac tissue model includes epicardial-derived smooth muscle cells (also known as cardiac smooth muscle cells) and / or epicardial-derived cardiac fibroblasts (also known as "endocardial-derived fibroblasts" or simply cardiac fibroblasts). These types of smooth muscle cells (SMCs) and fibroblasts develop from the epicardium and are not cells from non-cardiac lineages. Because these cells belong to the cardiac lineage, their distribution within organoids is unique, and their expression patterns for smooth muscle cell markers SM22 and calponin, and cardiac fibroblast markers DDR2 and vimentin are also different. These cells (both SMCs and fibroblasts are cardiac lineage) migrate to cardiomyocyte tissue or layers (e.g., Figures 10, 11).

[0034] In certain embodiments, the cardiac tissue model contains up to 3%, preferably up to 1%, or none of non-epidinal smooth muscle cells, and / or up to 3%, preferably up to 1%, or none of non-cardiac fibroblasts, i.e., non-cardiac SMCs and fibroblasts that do not originate from the epicardium. Non-cardiac cells may interfere with the natural behavior of the cardiac model and are therefore preferably absent or present in low amounts so as not to interfere with cardiac studies.

[0035] The tissue model of the present invention is artificial and grown by culture using the principles of natural development, but it does not represent any developmental stage of a heart in a living organism, nor is it a heart grown in a living organism. Its size is usually limited depending on the culture process. In a preferred embodiment of the present invention, a heart tissue model having a maximum dimension of 0.3 mm to 15 mm, for example, 0.5 mm to 1 mm, is used. This tissue model is usually hollow with a substantially spherical shape, but it may have irregularities. Therefore, for reference, its size is given by the maximum dimension of its shape (for example, Figures 1 and 2).

[0036] This size is typically applied to organoids obtained by the method of the present invention after steps a) to c) of developing into a single cavity. Since the cardiac tissue model of the present invention is fused, it is possible to generate larger structures having two or more lumens. The preferred size is then applied to one ventricle of the fused tissue model, i.e., one lumen with surrounding tissue, and is not extended to the tissue layer surrounding the other lumen.

[0037] The lumen is considerably larger (compared to previous cardiac tissue models) and occupies a large portion of the tissue model's volume (for both cavity organoids and fused organoids). Preferably, the size of the lumen at its maximum dimension is at least 60% of the size of the cardiac tissue model at its maximum dimension. As described above, for fused organoids with two or more lumens, this lumen applies to one cavity, including the surrounding tissue, and does not extend to the tissue layers surrounding another lumen (e.g., Figures 1 and 2).

[0038] Preferably, the tissue model is a mammalian cell, preferably a human cell or a non-human primate cell, or a cell model of a rodent, mouse, hamster, pig, cat, dog, horse, or cattle.

[0039] The present invention further provides a method for developing a cardiac tissue model, comprising the steps of: a) providing pluripotent stem cells; b) inducing mesoderm differentiation in a low-adhesion three-dimensional culture in the presence of a WNT activator and / or a GSK3-β inhibitor, and further in the presence of a PI3 kinase inhibitor, thereby generating an aggregate of mesoderm cells by the cells binding to each other and forming cell aggregates instead of a culture vessel; and c) differentiating the mesoderm cells from step b) into cardiac cells in a low-adhesion three-dimensional culture, and in the presence of cardiomyocyte differentiation factors, and in the absence of a WNT activator and / or a WNT antagonist, for at least 3 days, preferably 3 to 7 days, to form cardiac mesoderm and lumen (e.g., Figure 1).

[0040] Step b) can be modified using a similar effect, for example, by inducing mesoderm differentiation in the presence of a WNT activator and / or a GSK3-β inhibitor, the WNT activator and / or GSK3-β inhibitor and / or any PI3 kinase inhibitor generate aggregates of mesoderm cells, so that at least 90% of the pluripotent stem cells are in sufficient quantity to differentiate into pluripotent stem cells by eliminating pluripotency within 40 hours of the start of induction, and the cells are treated with activin A, osteomorphonimbus, fibroblast growth factor and / or albumin. In this case, the PI3 kinase inhibitor is not necessary but is still preferred.

[0041] In a particularly preferred embodiment, the treatment with a WNT activator and / or a GSK3-β inhibitor in step b) is combined with the treatment with an osteomorphic protein, particularly BMP4, in step c). Preferably, the osteomorphic protein, particularly BMP4, may have already been used in the cell treatment in step b). The use of osteomorphic proteins is particularly advantageous in constructing a heartbeat model because it forms a solid cavity.

[0042] In the process of embryonic development in vivo, the heart is the first functional organ to form in a developing human embryo. Whether or not an embryo survives depends on whether the cardiomyocyte self-organizes to form a four-chambered heart and subsequently matures. Self-organization in biology is the ability of cells to self-assemble when differentiating into organized tissue / organ-like structures under acceptable in vivo-like conditions in vivo or in vitro. This ability of cells to self-organize under appropriate conditions is crucial for the function of tissues and organs. The method of the present invention provides conditions that enable self-organization into organoids, including the formation of a large central cavity self-organization that does not require support from any artificial scaffold. A key feature of the present invention is that a scaffold, such as a polymer scaffold, is unnecessary and is not present in the final tissue model of the present invention. Such scaffolds, which can be avoided according to the present invention, do not need to be used according to the present invention, but include, for example, scaffolds described by Ma et al. (in the background art section above), which include polymers that are larger than most of the tissue model, for example, whose maximum dimensions extend across at least 75% of the tissue model size. Preferably, polymer scaffolds having a molecular weight of at least 1 MiODa are to be avoided.

[0043] According to the method of the present invention, self-assembly is driven by growth factor signaling and does not require an exogenous extracellular matrix such as a polymer scaffold (Ma et al., op. cit.) or Matrigel (International Publication No. 2019 / 174879 (A1) pamphlet). However, Matrigel is less disruptive than a polymer scaffold and can still be used. The addition of an extracellular matrix leads to significant variability in the degree of organoid formation and is therefore preferable to avoid.

[0044] The present invention's method is based on the use of pluripotent cells differentiated into mesoderm and cardiac cells. Initially, pluripotent cells could be cultured in a general two-dimensional culture, but at least from step c) onward, three-dimensional culture in a low-adhesion culture vessel is desirable. Step b) may be two-dimensional, but three-dimensional is preferred (similar to step c).

[0045] The cells provided in step a) are pluripotent. They are not human totipotent cells.

[0046] Pluripotent cells are preferably derived from mammals, preferably human cells or non-human primate cells, or from rodents, mice, hamsters, pigs, cats, dogs, horses, or cattle.

[0047] Pluripotent cells may be derived from cell lines or cell cultures. They may be pluripotent stem cells or induced pluripotent cells, i.e., cells derived from differentiated cells that have been re-made pluripotent, such as by treating them with Yamanaka factor. By using induced pluripotent cells, cardiac development in a specific individual can be investigated, regardless of whether or not they have genetic abnormalities that may affect cardiac development. In such embodiments, the cells may be induced pluripotent cells derived from patients with cardiac disease, particularly genetic cardiac disease.

[0048] In a preferred embodiment, pluripotent stem cells are cultured in a culture medium before being subjected to additional steps in the method of the present invention. This allows the cells to be primed and activated for further differentiation in the method of the present invention. The pluripotent cells may preferably be passaged and / or cultured, grown or maintained in a culture medium containing albumin and / or fibroblast growth factor. These components, preferably both, have been shown to significantly prime the cells and develop as desired in the present invention. Albumin may be, for example, serum albumin, e.g., bovine serum albumin (BSA). Fibroblast growth factor is preferably FGF2, particularly human FGF2.

[0049] It is preferable that pluripotent stem cells are cultured in a medium containing at least 1.5% (w / v) (at least 1.5 g / l) of albumin, preferably BSA, and / or at least 100 ng / ml of fibroblast growth factor, preferably FGF2.

[0050] Preferably, a cell culture medium is used, such as E8 medium. This medium preferably contains amino acids necessary for cell proliferation and an energy source such as carbohydrates, particularly preferably glucose. Furthermore, the medium contains salts and ions necessary for cell proliferation, such as Ca, Fe, Mg, K, Na, Zn, Cl, SO4, NO3, and PO4 ions. Even more preferred components of the medium are vitamins, such as vitamin B-12, biotin, choline, folic acid, inositol, niacinamide, pantothenic acid, pyridoxine, riboflavin, and thiamine.

[0051] Preferred amino acids include essential amino acids, and preferably one of the following amino acids: alanine, arginine, asparagine, aspartic acid, cystine, cystine, glutamic acid, glutamine, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, and valine.

[0052] Two-dimensional culture may involve growing or maintaining cells on a surface such as a tissue culture plate. Such a surface may be coated with a suitable culture substrate such as vitronectin. The cells may then be dissociated or used as aggregates in step b). If one or more passage steps are performed before supplying the cells to step b), the cells may be re-passaged.

[0053] In a preferred embodiment, the cells are those that have been cultured or treated in such a medium, particularly a medium containing albumin and / or fibroblast growth factor, for at least one day, preferably at least two days, for example, 1 to 30 days, for example, 2 to 20 days—while of course maintaining pluripotency.

[0054] In step b), preferably 1 to 1 million pluripotent cells (depending on the culture vessel) are provided and cultured in a single culture vessel. The culture vessel may be a two-dimensional culture vessel, such as in step a), or a three-dimensional culture vessel, such as those further described below for step c), where three-dimensional culture is essential.

[0055] In step b), the cells no longer divide or dissociate, but form aggregates that maintain their overall state. Pluripotent cells form aggregates by aggregating with each other, either on their own or during the initial differentiation process. The number of cells involved in this process affects the morphology of the subsequent tissue model. Generally, tissue models grown from fewer cells are more homogeneous, so it is recommended to use 100 to 10,000, for example, 200 to 6,000, pluripotent cells. Pluripotency can be controlled, for example, by determining SOX2, OCT4, and / or NANOG markers.

[0056] The culture medium used in step b) for culturing cells (forming aggregates, usually one main aggregate that persists in each culture vessel for the formation of the final tissue model) may have the same components as described above for step a), for example, it preferably includes energy sources such as amino acids and carbohydrates necessary for cell proliferation, particularly preferably glucose. The medium of step b) further includes signaling factors or modulators and any supporting factors that regulate signaling.

[0057] One such signaling regulator is a PI3 kinase inhibitor. An example of a PI3 kinase inhibitor is LY294002, but others are also effective. Using a PI3 kinase inhibitor is particularly preferred because it results in the cleanest and most homogeneous tissue model. However, instead of using a PI3 kinase inhibitor, one can use large amounts of a WNT activator. PI3 kinase inhibitors have the effect of causing many pluripotent cells to lose their pluripotency and differentiate into mesodermal cells, which then determines a highly homologous cardiac tissue model (with little to no non-cardiac cells such as endoderm and foregut). LY294002 is preferably used at concentrations of 3 μM to 15 μM.

[0058] In step b), the important signaling factor is a WNT activator. The WNT activator in step b) may be WNT-3a or a WNT ligand such as CHIR99021. Further WNT activators include the human WNT1, WNT2, WNT2B, WNT3, WNT3A, WNT4, WNT5A, WNT5B, WNT6, WNT7A, WNT7B, WNT8A, WNT8B, WNT9A, WNT9B, WNT10A, WNT10B, WNT11, and WNT16. These are highly conserved, and homologous forms exist in other organisms. Another possible WNT activator is R-spondin 1, which acts synergistically with WNT4. WNT4 may be endogenous in cells or administered externally.

[0059] The activation of WNT in this invention causes cells to undergo mesodermal differentiation. This is particularly efficient when combined with fibroblast growth factor (FGF), bone morphogenetic protein (BMP, preferably BMP4), activin A and / or albumin, preferably all four. In such cases, especially when using high levels of WNT activation, it may not even be necessary to use a PI3 kinase inhibitor. In fact, of these factors, the combination of WNT activation and albumin is sufficient to generate the robust and reproducible tissue model of this invention. Thus, as an alternative to step b), WNT activation and albumin can be utilized to cause a desired amount of pluripotent stem cells to lose their pluripotency (and differentiate into mesodermal cells) within 24-40 hours of culture according to step b). This is the most effective method for obtaining higher levels of WNT activation than usual. As an alternative to albumin (of course, it can also be used in combination), FGF is used in combination with BMP and / or activin A.

[0060] Preferably, if a sufficient amount of WNT activator is present and at least 90% of the pluripotent stem cells are differentiated into pluripotent stem cells by eliminating pluripotency within 40 hours after the start of induction, the amount of CHIR99021 as the WNT activator is at least 6 μM, preferably at least 9 μM, for example, at a concentration of 12 μM.

[0061] When a PI3 kinase inhibitor is used, the level of WNT activation may be at the usual low level; for example, the concentration of CHIR99021 as a WNT activator is at least 0.5 μM, preferably 0.5 μM to 12 μM.

[0062] Preferably, mesoderm differentiation (step b)) is induced in a medium containing activin A and / or osteomorphic protein, preferably further fibroblast growth factor. The preferred osteomorphic protein (BMP) is BMP4, particularly human BMP4. Preferably, mesoderm differentiation is induced in a medium containing at least 8 ng / ml of osteomorphic protein, preferably BMP4. Alternatively, or in combination, mesoderm differentiation is induced in a medium containing fibroblast growth factor (FGF) and / or albumin, preferably BSA. The preferred concentration of FGF is at least 10 ng / ml, preferably at least 50 ng / ml, for example 150 ng / ml or more, depending on the batch activity; the preferred concentration of albumin is preferably 0.2% (w / v) or more of BSA, for example 0.4% (w / v) or more. In particular, albumins such as serum albumin and / or FGFs are preferred culture medium additives because these compounds help promote the uniform and vigorous differentiation of the forming cell aggregates and their arrival at the mesoderm (and later the heart) using different pluripotent cell lines.

[0063] Activin, such as activin A, can be used to direct tissue models toward either the ventricles or atria. Generally, high activin levels (e.g., ≥10 ng / ml) direct tissue models toward the atria, while low activin levels (e.g., ≤10 ng / ml) direct them toward the ventricles. These exemplary concentrations may vary depending on the cell type used.

[0064] The culture in step b) in a specific medium is preferably 18 to 40 hours, preferably 24 to 39 hours, particularly 30 to 38 hours, for example 36 hours. Mesodermal differentiation can be controlled by identifying mesodermal cells that express the markers BRA / T (e.g., Figure 2), MIXL1, TBX6 and / or GSC. At the end of step b), it is desirable that the majority of cells, for example at least 90% of the desired cells, express these markers.

[0065] Steps b) (optional) and c) (and optional further steps, e.g., d) and maturation) involve culturing in a three-dimensional culture, meaning that aggregates are free-floating so that they do not adhere to the surface and can expand uniformly in all directions in three dimensions. Such culturing is performed in a low-adhesion culture so that aggregates do not adhere to the culture vessel walls. Preferably, low-adhesion culture involves culturing cells in a container having a low-adhesion surface that inhibits cell adhesion. The low-adhesion surface is preferably hydrophilic, neutrally charged, or nonionic. Alternatively, a hydrogel layer or coating can be applied to inhibit cell adhesion and keep the cells suspended. Such low-adhesion culture vessels are well known in the art and are described, for example, in International Publication No. 2019 / 014635 or International Publication No. 2019 / 014636. Preferably, the culture vessel is circular, especially concave. Alternatively, it may be flat or have a V-shaped bottom.

[0066] Step c), the step of differentiating mesoderm cells into cardiac cells, includes differentiating the mesoderm cells from step b) into cardiac cells in a three-dimensional culture in a low-adhesion culture, wherein the cells do not bind to the culture vessel but bind to each other to form cell aggregates, and cardiac mesoderm formation and lumen formation occur for at least two days, preferably three to seven days, for example five days, in the presence of cardiomyocyte differentiation factors, in the absence of a WNT activator and / or in the presence of a WNT antagonist. The WNT antagonist is preferably not used for the entire duration of step c). Preferably, the WNT antagonist is used for 12 to 72 hours, preferably 24 to 60 hours. Step c) is preferably carried out until at least 80% of the cells in the tissue culture have left the pluripotent and non-cardiac mesoderm stage. The "non-cardiac mesoderm stage" is the pluripotent stage, a precursor to the cardiomyocyte differentiation stage in which the cells have not yet differentiated into the cardiomyocyte system. Such processing steps are preferably for at least two, three, four, or five days, preferably seven or five days.

[0067] By identifying cardiac cells expressing the markers NKX2-5, HAND1, HAND2, TBX5, MYH6, MYH7, MYL7, Troponin T, Troponin I, and / or α-Actinin (e.g., Figures 2, 4, 5, 6, 7), it is possible to control cardiac mesoderm formation and myocardial progenitor differentiation.

[0068] Preferably, step c) includes culturing in a medium containing fibroblast growth factor, osteomorphogenetic protein, or a combination thereof. These compounds further differentiate the mesodermal cells obtained from step b) into cardiac cells, i.e., act as cardiomyocyte differentiation factors. The medium preferably further contains insulin, which aids in cell proliferation, diffusion, and survival, particularly when the medium contains glucose as an energy source.

[0069] In preferred embodiments of all the embodiments of the present invention, cells are treated in step c) with activin A, bone morphogenetic protein (BMP, preferably BMP4), and / or fibroblast growth factor (FGF, preferably FGF2), and particularly preferably all of these compounds. These compounds and factors improve the robustness and cavity formation of the tissue model of the present invention. Preferably, and / or albumin is used in combination with Activin A, BMP, and / or FGF, at the option of choice. Albumin is highly recommended to be used not only in step c) but also in step b), or in place of step c), as it protects cells from trace amounts of toxic substances such as WNT activators / inhibitors.

[0070] WNT antagonists are known in the art. They are also called WNT inhibitors and inhibit the activity of the WNT pathway. WNT antagonists are preferably selected from Wnt-C59, IWR-1, XAV939, IWP-2, IWP-4, DKK1, or combinations thereof (e.g., Figure 9). Particularly preferred WNT antagonists are IWP2 and XAV, which are used, for example, at a concentration of at least 1 μM, preferably at least 5 μM. WNT antagonists do not have to be used for the entire duration of step c) (at least 3 days), but may be used, for example, only at the beginning of step c) from 12h to 48h, preferably 18h to 36h, or at the start of the process within 12h of the start of step c). Preferably, WNT antagonists are used for at least 3 days or throughout step c). The culture medium used in step c) may be the same as in step b), and preferably contains amino acids necessary for cell proliferation and an energy source such as carbohydrates, particularly preferably glucose. As mentioned above, vitamins and / or salts are even more preferred.

[0071] The culture medium preferably contains fibroblast growth factor, preferably FGF2, at a preferred concentration of at least 6 ng / ml. The culture medium preferably contains osteomorphonomastic protein, preferably BMP4, at a preferred concentration of at least 8 ng / ml.

[0072] After step c), the cells that have formed aggregates with cavities at this point are left to differentiate further in a suitable culture medium. This is performed in the optional step d). Therefore, this method includes step d), which is a step of differentiating the aggregates having cardiac mesoderm into tissue having a cardiomyocyte layer using cardiomyocyte differentiation factors for a further 1 day or more, for example 1 to 3 days, preferably 2 days. The culture medium, culture vessel, and differentiation factors can be the same as in step c). WNT inhibition is no longer used at this point. The culture medium in step d) preferably contains fibroblast growth factor, preferably FGF2, at a preferred concentration of at least 6 ng / ml. Preferably, the medium contains osteomorphonidate, preferably BMP4, at a preferred concentration of at least 8 ng / ml. Insulin is also preferably used. Retinoic acid may be used further during this step, which further promotes differentiation toward complete differentiation.

[0073] Preferably, the culture medium in each step a), b), and / or c) and / or d) is changed daily, every two days, or at any interval between these days.

[0074] The method of the present invention may further include maturing the cells in a nutrient-rich culture medium, preferably until at least the lumen of the endothelial cells is lined. No further differentiation induction is required in such a maturation step. After step c) or d), the tissue model is stable and can be maintained. While being maintained in this way, the cells can be further matured while maintaining the large lumen that is characteristic of the present invention. This maturation can lead to the formation of a lining of cardiomyocytes that resemble the endocardium in vivo (e.g., Figure 8).

[0075] The differentiation of cardiomyocyte endothelial cells can be prevented by using VEGF signaling (vascular endothelial growth factor) inhibitors. According to one option of the present invention, in steps b) and / or c), and / or optional step d), the differentiation of cardiomyocyte endothelial cells can be prevented by applying a VEGF inhibitor. Exemplary VEGF inhibitors are the VEGFR inhibitor sunitinib or anti-VEGF antibodies (e.g., Figures 5 and 6).

[0076] On the other hand, it is also possible to culture the tissue with the addition of VEGF to promote the formation of cardiomyocyte endothelial cells. An example of VEGF is VEGF-A. This yields a tissue model with a high endothelial cell content, for example, 40% or more of cardiac endothelial cells. VEGF is preferably applied in step b), preferably optionally in step c), or not applied in step c) (for example, Figures 4 and 7).

[0077] Surprisingly, when using VEGF, it has been found that the location of endothelial cell formation can be controlled using WNT activation (step b), i.e., to be controlled to be on the inside (facing the lumen) or outside of the tissue model. Such a method may involve adding a sufficient amount of WNT activator, preferably CHIR99021, to the outside of the cardiac tissue model during step b), to induce the formation of cardiomyocyte endothelial cells, preferably 1 μM to 10 μM of CHIR99021. Lower concentrations of WNT can be used to guide the formation of the endothelial cell layer inward, depending on the cell line, from no CHIR to 5 μM of CHIR99021 (e.g., Figures 4, 7, 8). The addition of VEGF helps to develop a tissue model having three separate layers in the following orientations, for example, from inside to outside: cavity-cardiac cardiomyocytes, endocardial-derived fibroblasts, and cardiomyocyte endothelial cells (Figures 18F, F', F'', and 19A). The tissue model of the present invention may have such orientation. Furthermore, it is possible to stimulate the development of the inner endothelium under conditions that do not involve VEGF, for example, by using low concentrations of activin such as 5 ng / ml or less, or low concentrations of activin A. These values ​​may vary depending on the cell type used. This tissue model may also have an inner endothelial layer.

[0078] The method of the present invention may optionally further include the step of adding epicardial cells to the tissue model (e.g., Figure 11). This step may be performed directly after steps c) and d), or at any time during maturation. The addition of epicardial cells forms an epicardial layer on the outside of the tissue model. Further maturation can cause this epicardial layer to give rise to myocardial SMCs and / or fibroblasts, as described above. The epicardial cells may be simply added to the culture of the cardiac tissue model, or they may be allowed to adhere to the model and extend / migrate onto it to form an outer layer.

[0079] The cardiac tissue of the present invention can be used to represent a specific, normal heart or an abnormal condition such as disease, illness, or injury, and to restore and regenerate from such a condition. The cardiac tissue model of the present invention may include injured or healed injury. The method of the present invention may include inflicting injury on the tissue model and optionally further restoring it from such injury. Such conditions can be induced using agents such as test compounds, or the effects of the compounds can be tested using agents during injury and during the recovery and / or healing process.

[0080] The damaged or healed portion of the tissue model may contain higher concentrations of fibroblasts and / or collagen or fibronectin compared to the undamaged portion of the cardiac tissue model. For example, a higher concentration of fibroblasts in the damaged or healed portion is at least twice, preferably at least three times, and particularly preferably three to 25 times, for example, 3.3 to 21.7 times, compared to the undamaged portion of the cardiac tissue model. A higher concentration of collagen and / or fibronectin in the damaged or healed portion may be at least 1.1 times, preferably at least 1.25 times, and particularly preferably 1.3 to 3 times, for example, 1.33 to 2.13 times, compared to the undamaged portion of the cardiac tissue model.

[0081] The damage or healed damage may be caused by freezing, for example, freeze damage. Freeze damage is a reproducible damage model and is preferably used for standardization studies using the above-mentioned test compounds.

[0082] The damaged or healed portion of the cardiac tissue model is preferably 1% to 30%, preferably 4% to 20%, of the volume of the cardiac tissue model. Since cells, such as fibroblasts, migrate from healthy tissue to damaged tissue, it is preferable that sufficient healthy tissue remains. In other embodiments, fibroblasts may also migrate from adjacent tissue, and this migration is independent of the volume of healthy cells. The damaged or healed portion may also be larger in size, for example, 1% to 80%, preferably 10% to 60%, of the volume of the tissue model.

[0083] Damaged or healed damaged areas may contain a higher amount of collagen compared to the undamaged areas or collagen deposits in the tissue model. This higher amount may be, for example, at least 1.5 times or at least 2 times.

[0084] The present invention further provides a cardiac tissue model obtained by any method of the present invention. The obtained cardiac tissue model may have any of the structural elements described above.

[0085] The present invention further includes a kit for carrying out the method of the present invention. Such a kit may include i) a WNT activator and / or a GSK3-β inhibitor, ii) a PI3 kinase inhibitor, and iii) a low-adhesion cell culture vessel. All of these (particularly preferred examples) are described above. The kit may also include any further components used in the culture medium as described above. For example, it may further include a WNT inhibitor, BMP, FGF, insulin, albumin, etc. Particularly preferred, for the reasons described above, is BMP, or a combination with other further or alternative compounds. The kit of the present invention can be used in a method using the method of the present invention as described above.

[0086] The kit may further include instructions for carrying out the method of the present invention. Such instructions may be in printed form or in a computer-readable form on a suitable data carrier.

[0087] The present invention further provides a container plate comprising at least 10 compartments. The method of the present invention provides homogeneous and reproducible results such that the tissue models according to the present invention exist in each compartment, preferably at substantially the same developmental stage. The high reproducibility and robustness of the method are particularly advantageous when performing parallel comparative tests such as gene screening or compound testing (e.g., Figures 9 and 12).

[0088] The method of the present invention can be used for screening or testing candidate compounds with respect to their effects on cardiac development and function (e.g., Figure 9). Such a method may involve generating a cardiac tissue model according to the present invention by treating cells (at any stage of development, e.g., steps a), b), and / or c), and / or d), with the candidate compound. Alternatively, the final tissue model may be used to test the candidate compound. The method may involve comparing the development or functionality of the cardiac tissue model with that of a cardiac tissue model that was not treated with the candidate compound. For comparison, all treatment steps except treatment with the candidate compound should be the same. Examples of functionality include, for example, any cardiac function such as heart rate behavior (e.g., intensity and / or rhythm of arrhythmias), metabolic turnover, or gene expression of target genes, which may be affected by the candidate compound. Functional development may also be toxic development or function that may be caused by the candidate compound. Thus, the method or tissue model of the present invention can be used for toxicity testing or screening.

[0089] Similar to candidate compounds, gene modification tests can be performed. For example, the present invention provides a method for observing the effects of mutated (e.g., disease-related), repressed, or overexpressed genes during cardiac development, which includes generating a cardiac tissue model according to the present invention by having cells possess or overexpressing a candidate repressor gene, and comparing the development of the cardiac tissue model according to the present invention with the development of a cardiac tissue model not generated using a repressor or overexpressed gene (e.g., Figure 12). Mutation, overexpression, or repression can be performed by any method known in the art, for example, by gene knockout, siRNA inhibition, or CRISPR / Cas-based inactivation. Overexpression can be performed by introducing a transgene or by applying a gene activator that enhances gene expression. It is also possible to use mutations in such mutated tissue or cells to generate the cardiomyocyte of the present invention using these cells for screening or testing of the candidate compounds described above. Thus, the present invention provides for using the cardiomyocyte of the present invention or the method thereof as a disease model, for example, a cardiac disease model, to study cardiac functionality in particular. Therefore, it is also possible to combine a method for screening or testing candidate compounds with a method for observing the effects of repressor or overexpressed genes. In the combined method, comparisons can be made between tissue variants, between candidate treatment methods (with or without drug administration) (both involving mutations), between mutant and non-mutated variants in both drug-administered configurations, or between all four states (mutated + drug administration, mutant + no drug administration, non-mutated + drug administration, non-mutated + no drug administration).

[0090] An experimental overview is provided below, highlighting several preferred elements that can be selected according to the present invention. Further preferred elements are given in parentheses. The first pluripotency medium for growing pluripotent stem cells, which may be provided in step a), is preferably based on E8 medium. This medium contains BSA (2.5 μg / ml) and is modified by increasing the dose of human FGF2 to 200 ng / ml. These cells are passaged in 12-well plates as single cells by TrypLE or as small aggregates by EDTA, which does not affect organoid formation. When pluripotent cells were cultured under standard commercial conditions without the above combination, the protocol did not function well. The pluripotency stage is characterized by the expression of SOX2, OCT4, and NANOG as major cellular markers.

[0091] The next step in the protocol is mesoderm induction in differentiation (step b). The base differentiation medium CDM (Johansson & Wiles Molecular and Cellular Biolog, 1995, pp. 141-151) is used. Notable additives include BSA. The CDM differentiation medium contains the signaling ligands FGF2 (200 ng / ml), activin A (50 ng / ml), BMP4 (10 ng / ml), and CHIR99021 (an activator of WNT signaling) at different concentrations. Uniquely, the PI3-kinase inhibitor LY294002 (5-10 μM) is used. Its purpose is to ensure that all cells lose pluripotency within 24 hours, thereby making subsequent differentiation more homogeneous. However, the addition of activin, FGF2, and BMP4 is not essential, and mesoderm induction is possible with CHIR99021 (also abbreviated as "CHIR") alone in CDM, although in this case the results may be more variable. The dosage of CHIR has been optimized for other strains and varies from approximately 1 μM to 10 μM. The representative mesoderm markers expressed at this stage are as follows: BRA, EOMES, MIXL1, TBX6, and GSC.

[0092] The key to mesoderm induction is the timing of differentiation induction. Three options have been tested and are working: 1) Separating pluripotent cells from a 2D 12-well plate and seeding them into a 96-well ULA (ultra-low adhesion) plate, then inducing mesoderm after a further 18-36 hours in E8 pluripotent medium. 2) Induction of mesoderm immediately with CDM after seeding pluripotent cells. 3) Induction of mesoderm in 2D culture, then after the initial 36-hour step, separating into a 96-well low-density culture plate for 3D culture. Options 2 and 3 tend to yield cleaner and more homogeneous differentiation with larger cell counts (e.g., 2000+ cells / aggregate; option 1 yields homogeneous differentiation even with 2500+ cells).

[0093] The next stage is cardiac mesoderm differentiation (step c), which is continued for 4 days in a 96-well ULA plate, during which the lumen emerges. This is performed in CDM medium with IWR-1, XAV939, and IWP-2 inhibitors, along with FGF2, insulin, BMP4, and WNT inhibitors (as has been done in other studies). The results showed that cavity formation in CDM medium could be induced by external WNT inhibition alone. However, the endogenous activity of other pathways downstream of WNT, such as BMP, is also beneficial. This stage is where a cardiac-specific program is activated, involving several structural markers that continue to rise until day 7 of the protocol. The structural markers observed are MYH6, 7, MYL7, troponin T, and α-actinin; the transcription factors driving this program upstream are NKX2-5, HAND1, TBX5, ISL1, and GATA4 / 6.

[0094] In the final maturation stage, which involves further differentiation, the use of FGF2 and BMP4 for two days strongly promotes terminal differentiation, resulting in cardiomyocytes that beat in vivo. This stage is again not essential, but unstable, as cells can be left alone with only CDM and insulin. Cells at this stage continue to express the structures and transcription factor markers described in the cardiac mesoderm stage. If the final stage is maturation with only CDM + insulin, the cardiomyocyte organoids tend to mature further after about one month, as seen by the upregulation of MYL2 ventricular markers. In general, cardiomyocytes in these three-dimensional organoids are known to have higher expression levels of cardiomyocyte genes compared to two-dimensional cardiomyocytes. In this protocol, cardiac endothelial cells (EC, CD31+, CDH5+) appear over time and are already detectable by days 4-5, but increase in number after one month, forming to neatly cover the internal lumen. However, the majority of cells in this basic protocol are still cardiomyocytes (about 80-90%), with cardiac endothelial cells making up the remainder. No other cell types were detected. No endoderm or ectoderm derivatives were detected.

[0095] If organoids with a higher proportion of cardiomyocytes (over 95%) must be produced without even endothelial cells, then a VEGF inhibitor such as sunitinib (100 nM, VEGF pathway inhibitor) can be used during the first 10 days of differentiation or thereafter (e.g., Figure 6). Advantageously, both protocols (with and without VEGF) allow these structures to be maintained for several months, which is again very important for applications and screening studies.

[0096] The ratio of these cells can be controlled. For example, the cardiac mesoderm stage can be modified by adding VEGF (200 ng / ml) to induce the migration of endothelial cells. With 200 ng / ml of VEGF, the ratio of cardiomyocytes to endothelial cells is typically about 53% endothelial cells and 41% cardiomyocytes (with a variation of about 5%) (other ratios are possible by changing the VEGF concentration) (see, for example, Figure 4). Under high WNT activity conditions (5-8 μM CHIR) during mesoderm induction, endothelial cells form compartments / liners around cardiomyocytes, while in organoids induced using low WNT activity (1-4 μM CHIR), they form liners facing the lumen. In other words, not only the ratio of cells but also whether the endothelial liner is on the inside or outside can be controlled. This control is an important aspect because it may differ depending on the application. Furthermore, a space exists between the endothelium and cardiomyocyte compartments, similar to the space observed in vivo at the cardiac duct stage. Fibronectin is expressed in the cardiomyocytes facing that space, confirming that this is consistent with the conditions in vivo. This is a major component of the cardiac jelly that forms between endothelial cells and cardiomyocytes.

[0097] The important point here is that these endothelial cells are the most similar to the endocardium (and therefore true cardiac myocardial endothelial cells) because they express higher levels of NPR3 and NFATC1 compared to control non-cardiac endothelial cells (e.g., Figure 4E). Furthermore, they typically express the correct HOX genes (ranging from HOX1 to HOX5) for cardiac function.

[0098] Another important point is that activation of mechanosensing genes (SOX18, KLF2, CDH5, FOS, TEK, FOXO1) in endothelial cells from organoids is observed in contrast to when endothelial cells are simply cultured in two dimensions. This is a very important biological feature of more functional, true cardiomyocytes (e.g., Figure 4).

[0099] A third preferred addition is the epicardium (WT1+, TCF21+, TBX18+) applied to the additional step. Here again, CDM is preferably used in three-dimensional culture (e.g., Figures 10, 11). When epicardium cells are added to cardiac organoids, an epicardium enveloping the organoids from the outside is formed within a few days (similar to in vivo), and then the epicardium cells differentiate into smooth muscle cells (SM22+, Calponin+) and cardiac fibroblasts (DDR2+, Vimentin+) and migrate to the cardiomyocyte tissue (e.g., Figure 11). This also occurs in vivo. It is important that smooth muscle cells and fibroblasts do not simply molt, but rather undergo spontaneous transformation, which is their natural function. Thus, spontaneous development is crucial for the functionality of the final product.

[0100] In summary, there are several possible end products. Preferred tissue models or organoids have: a) 95% or more pure cardiomyocyte organoids, b) 90% cardiomyocytes and 10% endothelial cell liners, c) 45% cardiomyocytes and 55% endothelial cells (or other ratios that can be manipulated by VEGF), d) cardiomyocytes and epicardium only, f) all three major lineages combined (e.g., Figure 11G), and g) if the epicardium differentiates into cardiac smooth muscle cells and cardiac fibroblasts, there will be five cell types present, all essential for cardiac function and all types of cardiac cells as in vivo (e.g., Figure 11). Cell types not belonging to the heart (such as endoderm, hematopoietic endothelium, non-cardiac smooth muscle cells and non-cardiac fibroblasts not derived from the epicardium) are not detected. Preferably, the tissue model contains 0.1% to 40%, preferably 1% to 35%, and particularly preferably 10% to 30% endocardium-derived fibroblasts. Preferably, the tissue model contains at least 1%, preferably at least 5%, of endocardial-derived fibroblasts.

[0101] The defining feature of this invention is its architecture—namely, cardiomyocytes or endothelial cells facing a single lumen, epicardial cells surrounding the structure, and cardiac smooth muscle cells and cardiac myofibroblasts migrating into the myocardium. This is the same way that cardiac chambers / ventricles are formed in vivo, and this particular architecture, with all its variations, is crucial for its function and application in models and studies of development and physiological behavior that closely resembles that in vivo.

[0102] The present invention is further described by the following numbered embodiments.

[0103] 1. A cardiac tissue model comprising at least 60% cardiac cells or at least 50% cardiac cells, independently of any additional cells in the vascular tissue of the cardiac tissue model, wherein the cardiac cells surround a lumen and the cardiac cells are selected from the group consisting of cardiomyocytes, endocardial cells and epicardial cells.

[0104] 2. A cardiac tissue model as described in 1., consisting of at least 30% cardiomyocytes.

[0105] 3. The cardiac tissue model described in 2., further comprising at least 2% endocardial cells.

[0106] 4. A cardiac tissue model described in any one of 1-3, which includes cardiomyocytes and endocardial cells in different tissue layers.

[0107] 5. A cardiac tissue model according to any one of 1 to 4, wherein epicardial cells are included in another tissue, preferably the outer layer.

[0108] 6. A cardiac tissue model according to any one of 1 to 4, comprising at least 40%, preferably at least 60%, and more preferably at least 80% cardiomyocytes.

[0109] 7. A cardiac tissue model according to any one of 1 to 6, wherein foregut endoderm cells are present in a maximum of 5%, preferably a maximum of 3%, or none at all, and / or hematopoietic cells are present in a maximum of 3%, or none at all.

[0110] 8. A cardiac tissue model described in any one of 1. to 7., comprising epicardial smooth muscle cells and / or epicardial myocardial fibroblasts.

[0111] 9. A cardiac tissue model according to any one of 1. to 8., which contains no more than 3% or no non-epidial smooth muscle cells, and / or contains no more than 3% or no non-epidial or endocardial fibroblasts.

[0112] 10. A cardiac tissue model described in any one of items 1-9, with a maximum dimension of 0.3mm to 15mm.

[0113] 11. A cardiac tissue model according to any one of 1 to 10, wherein the maximum dimension of the lumen is at least 20% of the size of the cardiac tissue model in its maximum dimension.

[0114] 12. A cardiac tissue model according to any one of 1 to 11, wherein cardiomyocytes or endocardial cells are directly facing the lumen.

[0115] 13. A cardiac tissue model described in any one of 1. to 12., including the injured or healed injured portion.

[0116] 14. The cardiac tissue model according to 13, wherein the damaged portion or the healed damaged portion contains higher concentrations of fibroblasts, collagen and / or fibronectin compared to the undamaged portion of the cardiac tissue model.

[0117] 15. The cardiac tissue model according to 14, wherein the high concentration of fibroblasts in the damaged portion or the healed damaged portion is at least twice, preferably at least three times, particularly preferably three to 25 times, for example, 3.3 to 21.7 times, compared to the undamaged portion of the cardiac tissue model.

[0118] 16. The cardiac tissue model according to 14. or 15., wherein the high concentration of fibronectin and / or collagen in the damaged portion or the healed damaged portion is at least 1.1 times, preferably at least 1.25 times, particularly preferably 1.3 to 3 times, for example 1.33 to 2.13 times, compared to the undamaged portion of the cardiac tissue model. 17. A cardiac tissue model according to any one of 13-16, wherein the damaged portion or the healed damaged portion is damaged by freezing, for example, freezing.

[0119] 18. The cardiac tissue model according to any one of 13 to 17, wherein the damaged portion or the healed damaged portion of the cardiac tissue model accounts for 1% to 30%, preferably 4% to 20%, of the volume of the cardiac tissue model.

[0120] 19. The following steps: a) To provide pluripotent stem cells, b) Inducing mesodermal differentiation in three-dimensional culture in low-adhesion culture in the presence of a WNT activator and / or a GSK3-β inhibitor, and further in the presence of a PI3 kinase inhibitor, thereby generating aggregates of mesodermal cells by having the cells bind to each other instead of the culture vessel to form aggregates of the cells. c) Differentiate the mesoderm cells from step b) into cardiac cells in a low-adhesion three-dimensional culture, and in the presence of cardiomyocyte differentiation factors, and in the absence of WNT activators and / or in the presence of WNT antagonists, for at least 3 days, preferably 3 to 7 days, the cells aggregate with each other instead of binding to the culture vessel to form aggregates of the cells, thereby forming cardiac mesoderm and lumen. A method for generating a cardiac tissue model, including the development of the cardiac tissue model.

[0121] This method may further include the case where the endothelial and cardiomyocyte compartments are separated by space, for example, in vivo.

[0122] 20. The following steps: a) To provide pluripotent stem cells, b) Induce mesoderm differentiation in the presence of a WNT activator and / or a GSK3-β inhibitor, wherein the WNT activator and / or GSK3-β inhibitor, along with an optional PI3 kinase inhibitor, cause at least 90% of the pluripotent stem cells to form aggregates of mesoderm cells within 40 hours of the start of induction, in an amount sufficient to deplete pluripotency and differentiate the pluripotent stem cells, and then treat the cells with fibroblast growth factor and / or albumin. c) Differentiate the mesoderm cells from step b) into cardiac cells in a three-dimensional culture using low-adhesion culture, and perform cardiac mesoderm and lumen formation for at least 3 days, preferably 3 to 7 days, in the presence of cardiomyocyte differentiation factors, in the absence of WNT activators and / or in the presence of WNT antagonists. A method for generating a cardiac tissue model, including the development of the cardiac tissue model.

[0123] The method may further include separating the endothelial and cardiomyocyte compartments by space, for example, in vivo. 21. The method according to 19. or 20., wherein the WNT activator in step b) is a WNT ligand such as WNT-3a or CHIR99021.

[0124] 22. The method according to 20, wherein the amount of the WNT activator is sufficient to de-pluripotently differentiate at least 90% of the pluripotent stem cells within 40 hours after the start of induction, the amount of CHIR99021 as the WNT activator is at least 1 μM, preferably at least 6 μM, and 12 μM if a PI3 kinase inhibitor is present, and CHIR99021 as the WNT activator is present at a concentration of at least 0.5 μM, preferably 0.5 μM to 12 μM.

[0125] 23. The method according to any one of 19 to 22, wherein the pluripotent stem cells are induced pluripotent stem cells or cells from a cell line, and / or the pluripotent stem cells provided in step a) are subcultured in a medium preferably comprising albumin and / or fibroblast growth factor, more preferably further comprising BMP and / or insulin.

[0126] 24. The method according to any one of 19-23, wherein the pluripotent stem cells provided in step a) are grown in a medium further comprising at least 1.5% (w / v) albumin, preferably BSA, and / or at least 100 ng / ml of fibroblast growth factor, preferably FGF2, more preferably BMP, and / or insulin.

[0127] 25. The method according to any one of claims 19 to 24, wherein the PI3 kinase inhibitor is LY294002 or any other PI3 kinase inhibitor.

[0128] 26. The method according to any one of claims 19-25, wherein mesoderm differentiation is induced in a culture medium containing activin A and / or osteomorphogenetic protein, preferably further containing fibroblast growth factor.

[0129] 27. The method according to 26, wherein mesoderm differentiation is induced in a medium containing at least 1 ng / ml of osteomorphonectomy protein, preferably BMP4.

[0130] 28. The method according to any one of claims 19-27, wherein mesoderm differentiation is induced in a medium containing fibroblast growth factor and / or albumin, preferably BSA.

[0131] 29. The method according to any one of claims 19 to 28, wherein the low-adhesion culture includes culturing the cells in a container having a low-adhesion surface that inhibits adhesion to the surface of the cells.

[0132] 30. The method according to any one of claims 19 to 28, wherein differentiating the mesodermal cells into cardiac cells involves culturing them in a medium containing fibroblast growth factor, insulin, osteomorphonectomy protein, or a combination thereof.

[0133] 31. The method according to any one of the methods described in 19. to 30., wherein the WNT antagonist is selected from Wnt-C59, IWR-1, XAV939, IWP-2, and IWP-4 DKK1.

[0134] 32.d) The method according to any one of 19. to 31., further comprising differentiating the aggregate having cardiac mesoderm into tissue having a cardiomyocyte layer with a cardiomyocyte differentiation factor for one day or more.

[0135] 33. The method according to any one of claims 19 to 32, further comprising, preferably, maturing the cells in a nutrient-rich culture medium until at least an inner or outer layer of endothelial cells is formed, preferably lining the lumen or surrounding the cardiomyocyte layer, and optionally separated by space.

[0136] 34. The method according to any one of 19 to 33, further comprising applying a VEGF inhibitor in step b) and / or c) and / or optional step d) of 32. to prevent the formation of the inner layer of cardiac endothelial cells.

[0137] 35. The method according to any one of 19-33, further comprising culturing cells with VEGF to increase the formation of cardiac endothelial cells and / or cardiac fibroblasts.

[0138] 36. The method according to 35, comprising adding a sufficient amount of WNT activator, preferably CHIR99021, preferably 1 to 12 μM CHIR99021, to induce the formation of cardiac endothelial cells on the outside of the cardiac tissue model during step b).

[0139] 37. The method according to any one of claims 19 to 36, further comprising the step of adding epicardial cells to the tissue model.

[0140] 38. The method according to any one of 19. to 37., wherein in step b), the cells are treated with albumin.

[0141] 39. The method according to any one of items 19 to 38, wherein step a) is a two-dimensional culture, and preferably step b) is also a two-dimensional culture.

[0142] 40. The method according to any one of 19 to 39, comprising the step of causing damage, preferably freeze damage, to a portion of the tissue model in a portion of the cardiac tissue model that is 1% to 30%, preferably 4% to 20%, of the volume of the cardiac tissue model.

[0143] 41. A kit for carrying out any one of the methods described in 19. to 40., comprising i) a WNT activator and / or a GSK3-β inhibitor, ii) a PI3 kinase inhibitor, and iii) a low-adhesion cell culture vessel.

[0144] 42. The kit described in 41, further comprising a WNT inhibitor.

[0145] 43. A cardiac tissue model that can be obtained by any one of the methods described in any one of 19 to 40, preferably further defined in any one of 1 to 18.

[0146] 44. A container plate comprising at least 10 compartments, each compartment containing a tissue model described in any one of 1. to 18. and 43. at substantially the same developmental stage.

[0147] 45. A method according to any one of 19 to 40, for screening or testing a candidate compound for its effect on cardiac development and / or functionality, comprising treating the cells with the candidate compound while generating a cardiac tissue model according to any one of 19 to 40, and comparing the development of the cardiac tissue model with the development and / or functionality of a cardiac tissue model not treated with the candidate compound.

[0148] 46. ​​A method for observing the effect on cardiac development, including the development of a cardiac tissue model according to any one of 19 to 40, comprising having a candidate repressor gene in the cells or overexpressing a candidate gene, and comparing the development of the cardiac tissue model with the development of a cardiac tissue model that did not develop with the repressor or overexpression gene.

[0149] A method that combines screening or testing of candidate compounds for their effects on cardiac development and / or function as described in 47.45. with observation of the effects of inhibitory or overexpressed genes on cardiac development as described in 46.

[0150] The present invention is not limited to these embodiments and will be further illustrated by the following figures and examples. [Examples]

[0151] Example 1: Materials and Method General Human Pluripotent Stem Cell Culture - Human pluripotent stem cell lines (H9, WiCell;WT and modified WTC, Allen Institute for Cell Science) were cultured in a homemade culture medium modified from the E8 culture system. 0.5% BSA (Europa Biosciences) and in-house produced FGF2 were added to the E8 recipe. Cells were cultured on Corning or Eppendorf tissue culture plates coated with vitronectin (Stem Cell Technologies) and subcultured every 3-4 days using TrypLE Express Enzyme (Gibco) or PBS-EDTA.

[0152] HPSC differentiation into cardiomyocytes in 2D and 3D aggregates—hPSCs—was seeded in E8+ROCKi (Y-27632, Tocris) at 160-175,000 cells / 24-well plates for 24 hours. Subsequently, cells were induced for 36-40 hours in CDM medium containing FGF2 (30 ng / ml, University of Cambridge), LY294002 (5 μM, Tocris), activin A (50 ng / ml, University of Cambridge), BMP4 (10 ng / ml, R&D System), and CHIR99021 (1-1.5 μM for H9, 3-4 μM for WTC, Tocris) (Mendjan et al., Cell Stem Cell (2014) Vol.15, pp.310-325). To enhance cell viability at this stage, 1 microg / ml of insulin (Roche) was optionally added. This culture medium was named FLyABCH(Ins). After 36-40 hours, the cells were cultured for 4 days in CDM medium containing BMP4 (10 ng / ml), FGF2 (8 ng / ml), insulin (10 μg / ml), IWP2 (5 μM, Tocris), and Retinoic Acid (0.5 μM, Sigma Aldrich), with daily medium changes. This medium was named BFIIWPRa. Subsequently, the medium was changed to CDM medium containing BMP4 (10 ng / ml), FGF2 (8 ng / ml), and insulin (10 μg / ml), and the medium was changed daily for 2 days. This medium was named BFI. To maintain the obtained cardiomyocytes, the medium was changed to CDM medium containing I insulin (10 μg / ml), and half of the medium was changed daily. This medium was named CDM-I. Aggregated cardiomyocytes were developed by dissociating cardiomyocytes maintained until day 21 using the StemDiff cardiomyocyte dissociation kit (Stem Cell Technologies). 1000 cells / well were then re-seeded as aggregates in CDM-I and 5% FBS (PAA) on AggreWell 400 plates (Stem Cell Technologies). Two days later, the formed aggregates were transferred to ultra-low cluster 96-well plates (Corning) in CDM-I and shaken at 58 rpm, 37°C, and 5% CO2. After two days, the culture medium was changed, and four days later, the aggregates were used for analysis.

[0153] Cardiac organoid generation - hPSCs were harvested at approximately 70% confluence. Subsequently, 5000 cells / well were seeded into an ultra-low adhesion 96-well plate (Corning) containing E8+ROCKi (Y-27632, Tocris) and harvested by spinning at 200G for 5 minutes. After 24 hours, the aggregates formed were induced with FLyAB(Ins) containing 4-8 μM CHIR99021. Cardiac differentiation was carried out in the same manner as in 2D culture. To maintain the obtained cardiomyocytes, the culture medium was replaced with CDM medium containing insulin (10 μg / ml) every two days.

[0154] Endothelial cardiomyocyte development - the pluripotency maintenance medium was refreshed 6 hours before cell seeding. Next, 2500-3000 hPSCs were seeded over 36-40 hours in ultra-low cluster 96-well plates (Corning) in FLYAB(Ins) medium containing CHIR (5-6 μM) and ROCKi (5 μM). Then, the medium was changed to BFIIWPRa supplemented with VEGF-A (200 ng / ml, Peprotech), and the cells were cultured for 4 days with daily medium changes. After that, the medium was changed to BFI+VEGF-A (100 ng / ml), and the medium was changed again after 1 day, and the cells were cultured for 2 days. For maintenance, CDM medium supplemented with 100 ng / ml of VEGF-A was used, and the medium was changed every two days.

[0155] Extracardiac involvement of cardiomyocyte organoids—for extracardiac differentiation, ~70% confluent hPSCs were seeded in E8+ROCKi (Y-27632, Tocris) at 55,000 cells / 24 well plates 24 hours prior to differentiation. Cells were induced in CDM medium (Mendjan et al., 2014, above) containing FGF2 (30 ng / ml, University of Cambridge), LY294002 (7.5 μM, Tocris), BMP4 (10 ng / ml, R&D Systems), and CHIR99021 (1.5 μM, Tocris). After 36-40 hours, the differentiation medium was replaced for 2 days with CDM medium containing BMP4 (10 ng / ml), FGF2 (8 ng / ml), insulin (10 μg / ml), IWR-1 (1 μM, Tocris), and retinoic acid (1 μM, Sigma Aldrich), with the medium changed once daily. Subsequently, the medium was changed again for 5 days with CDM medium containing BMP4 (10 ng / ml), insulin (10 μg / ml), and retinoic acid (1 μM), with one medium change during this period. To maintain the resulting epicardium, cells were seeded at the end of differentiation on bovine plasma fibronectin (2 μg / ml, Sigma) coated plates in CDM medium containing insulin (10 μg / ml) and SB431542 (10 μM, Tocris), supplemented with ROCKi on the first day of seeding. This medium was named CDM-SBI. The replicated epicardium was routinely passaged at a 1:3 ratio every 3-5 days at 80-90% confluence, or subsequently differentiated into cardiac fibroblasts (CF) or smooth muscle cells (SMC) for 12 days using CDM medium containing FGF2 (30 ng / ml, University of Cambridge), TGFb2 (2 ng / ml, R&D Systems), L-ascorbic acid (100 μm / ml, Sigma), and insulin (10 μg / ml), or CDM medium supplemented with PDGF-BB (10 ng / ml, R&D Systems), L-ascorbic acid (100 μg / ml, Sigma), and insulin (10 μg / ml) supplemented with TGFb2 (2 ng / ml, R&D Systems).

[0156] To generate epicardial aggregates for the engagement assay, 8.5-day-old epicardial cells were dissociated using TrypLE Express Enzyme (Gibco) and re-seeded as aggregates in CDM-SBI and 5% FBS (PAA) at 1000 cells / well in Ag-greWell400 plates (Stem Cell Technologies). After 2 days, an average of 8-12 aggregates / well were transferred to ultra-low cluster 96-well plates (Corning) containing myocardial organoids differentiated in CDM-I, shaken at 58 rpm, 37°C, and 5% CO2, and refreshed with CDM-I medium every 2 days for 1 week of co-culture.

[0157] Example 2: Cardiac mesoderm with a cavity formed in a test tube The central morphological feature of the heart is the presence of a cavity surrounded by cardiomyocytes (CMs), with the potential to develop into beating CMs, and lined with endothelium (endocardium). In vivo, this cavity is formed through a complex process involving the migration of cardiac mesoderm and the fusion of the endocardial ducts due to the contraction of the foregut endoderm. As a result, the cardiac ducts develop into a multi-ventricled heart. However, cardiac ducts and ventricles can be formed in vivo even without the contraction of the endocardium or foregut endoderm.

[0158] The study investigated whether cardiac mesoderm was sufficient to form cavities under tolerant conditions in vitro. For this purpose, the effects of activin, BMP, FGF, retinoic acid, and WNT on differentiation were tested, along with a high-throughput differentiation approach in adherent 96-well plates based on a time protocol of cardiogenic signaling pathways, while differentiating hPSCs into mesoderm, cardiac mesoderm, and (pulsating) cardiomyocyte precursors (Figure 1). To screen for factors involved in cavity formation in vitro, extracellular matrix (ECM) proteins involved in early cardiomyogenesis were selectively supplemented with culture media. For example, the addition of laminin 521 / 511 before mesoderm induction resulted in rapid, dose-dependent autoaggregation at different stages of differentiation, transforming from a two-dimensional layer to a pulsating three-dimensional spherical structure by day 7 of differentiation (Figure 1). Surprisingly, these structures were hollow, and the presence of cavities was confirmed by confocal microscopy and histological sectioning (Figures 1 and 2). Analysis of the MYL7-GFP reporter and staining results for the CM marker TROPO-T / MYL7 confirmed that this structure is mostly constructed from cardiomyocytes.

[0159] Next, we investigated whether exogenous ECM is necessary for mesoderm autoaggregation or directly involved in cavity formation. Cardiac differentiation in a 3D non-adherent 96-well plate revealed that exogenous ECM is not necessary for rapid autoaggregation, robust self-organization, and differentiation into pulsating TROPO-T+ / MYL7+ structures containing cavities (Figure 2, HT image, live image & section). Therefore, this high-throughput approach allowed for rapid optimization of timing, cell count, and culture medium conditions for a highly reproducible in vitro self-organizing cardiomyocyte model.

[0160] To determine the timing of cardiac cavity formation, live imaging and time-course analysis of frozen sections revealed that cardiac cavities appear during the cardiac mesoderm stage, before the expression of major cardiac structural markers (Figures 1 and 2). These cavities typically merge into a single major cavity. Since cavity formation occurred in the absence of VEGF signaling and the undetectable endothelial markers CD31 and VE-Cad, lumen formation was not mediated by the endothelium (Figure 1, ICC section (Figure 5)). Importantly, the absence of SOX17+ and EOMES+ endoderm during differentiation indicates that these cardiac mesoderm cavities did not arise from the contraction of the foregut endoderm (Figure 3). In contrast, during systole, the cardiac mesoderm strongly expressed the cardiac mesoderm marker HAND1, and specific structural markers (MYL7, TROPO-T) were upregulated. Inhibition of VEGF signaling resulted in over 90% of these structures becoming positive for MYL7 by day 7.5 (Figure 6). In conclusion, the combination of signal transduction and high-throughput differentiation in three dimensions revealed that cardiac mesoderm can specialize and self-organize into CM-based cavities, which are key features of the heart.

[0161] The current state of CM differentiation technology includes either two-dimensional or three-dimensional approaches. Therefore, an attempt was made to compare three-dimensional cavity-containing structures with CM differentiated in two dimensions. Typically, on day 7 of differentiation, these structures began to pulsate at a similar rate and frequency of Ca2+ transient currents as CM differentiated in two dimensions. These structures were able to maintain pulsation for several months without the appearance of non-cardiac cells. Molecular-level comparisons using RNA-seq time-series analysis revealed that these structures exhibited the closest expression signature to the first cardiac cortex subsystem (HAND1+, TBX5+, NKX2-5+, TBX1-, HOXB1-) of cardiac mesoderm, which in vivo gives rise to the cardiovascular system and later to parts of the left ventricle and both atria. Overall, compared to two-dimensional CM (PCA and heatmap), higher expression of cardiomyocyte genes was observed in cavity-forming structures and three-dimensional aggregates. Genes encoding ion channels, structural proteins, cardiomyocyte transcription factors, and sarcoplasmic reticulum proteins showed significantly higher expression levels. This effect was also confirmed at the protein level, as seen in whole-proteome analysis. Thus, hPSC-derived cardiac mesoderm is sufficient to firmly form functional, reproducible, and sustainably maintained cavity-containing CM structures in long-term culture.

[0162] Example 3: Endothelial cells line the cavities of cardiac organoids. In vivo, before the formation of the cardiac duct, the cardiac mesoderm co-develops with endocardial progenitor cells to form bilateral endocardial canals (independent compartments with lumens). The co-differentiation of cardiac mesoderm into endocardial-like ECs and CMs was investigated to determine if these compartments could be reconstituted in vitro (Figure 4). To promote co-differentiation, VEGF-A, an inducer of embryonic endothelial cell identity, was added at the cardiac mesoderm stage, and optimal conditions were screened using a double-reporter hPSC strain for the CM marker (MYL7-GFP+) and the endothelial cell (EC) marker (Figure 4). Surprisingly, in the presence of VEGF, when CMs and ECs co-differentiated from cardiac mesoderm into structures with cavities, they separated into CM and EC layers, creating a space between them that resembled the in vivo environment (Figure 4). The degree of WNT signaling activation during mesoderm induction allowed control over whether the EC layer surrounded the CM or faced the cavity (Figures 2, 7, 8). The ratio of endothelium (CM) to endothelium (EC) was extremely stable, ranging from 41% (MYL7+) to 53% (CHD5+). When VEGF was not administered, myocardial organoids had a lower proportion of EC and, over time (from day 10 to day 27), formed endothelium facing only the cardiac chambers, similar to the in vivo environment (Figure 8).

[0163] ECs within the layer have the potential to form rapidly extending CD31+ networks within their compartments, which also extend into the CM layer (Figures 4, 7). To investigate whether early co-differentiation is necessary for compartment formation, control 3D cardiac microtissues were developed by first differentiating CM and EC separately from the cardiac mesoderm and then condensing them three-dimensionally. In these cardiac microtissues, ECs only formed networks and were not compartmentalized to form independent layers. In conclusion, it was found that, as in vivo, only CM and EC co-differentiated three-dimensionally from the cardiac mesoderm self-organize, forming distinct layers containing cavities, endothelial linings, and extended endothelial networks. These structures were named cardiac organoids because they arise from a homogeneous population of hPSCs and self-organize into cavities, compartments, and EC linings while giving rise to the first two types of cardiac cells.

[0164] Generally, cardiac endocardium (EC) is characterized by the expression of markers such as CD31 and CDH5, but early cardiac EC (endocardium) has even more specific signatures. To compare these molecular features of myocardial organoid EC, SMART-seq2 analysis was performed on selected CDH5+ cells (Figure 4). Compared to ECs derived from established 2D hPSC EC differentiation protocols (Patsch et al., Nature Cell Biology 17, 994-1003 (2015)), ECs derived from 3D vascular organoids, human umbilical vein endothelial cells (HUVECs), human cardiac microvascular endothelial cells (HCMECs), and CHD5+ cardiac organoids were most similar to those derived from 3D vascular organoids (Wimmer et al., Nature 565, 505-510 (2019) (Figure 4E)). Importantly, ECs from cardiac organoids upregulated cardiomyocyte transcription factors such as GATA4 / 5 endocardial-related genes (NFATC1, NPR3). Their anterior HOX gene expression profiles were consistent with HCMECs (Figure 4E) and consistent with 2D cardiac mesoderm-derived ECs, but not with other EC subtypes. On the other hand, MYL7+ sorted from cardiac organoids... SMART-seq2 analysis of cardiomyocytes (CMs) showed that sorted cells that were neither MYL7+ nor CDH5+ upgraded genes related to lateral plate mesoderm and ECM genes. Therefore, the molecular signature of ECs derived from cardiomyocytes is consistent with an endocardial-like identity.

[0165] The endothelium's ability to sense fluid flow, pressure, and mechanical stretching plays a crucial developmental and physiological role. As predicted from a true model of cardiomyogenesis, RNA-seq analysis of bulk cardiomyocyte organoids revealed upregulation of major mechanosensing genes (SOX18, KLF2, CHD5) compared to three-dimensional cardiomyocyte microtissue and two-dimensional cardiomyocytes (CM) and ECs (Figure 4E). This was further supported by SMART-seq2 analysis of sorted ECs from cardiomyocytes, which showed induction of mechanosensing stress genes (KLF2, FOXO1, TEK, FOS), and immunofluorescence staining of cardiomyocyte organoids for CDH5, SOX18, and KLF2 (Figure 4E). These results suggest that co-differentiation, self-organization, and compartmentalization of CM and EC are important for achieving major cardiomyocyte identity and activation of mechanosensing genes (both essential aspects of cardiac physiology).

[0166] Example 4: The epicardium encloses the myocardial organoid. Following cavity formation and the construction of the endothelium and lining, the involvement of the epicardium with the early myocardial ventricles is the third major cardiac self-organization event. In vivo, the epicardium involves the myocardium from a small cluster of cells called the pre-epidimal. After the involvement of the myocardium, signals from the cardiomyocyte differentiate the epicardinal cells into smooth muscle cells (SMCs) and cardiac fibroblasts (CFs), which become important cell types for later cardiac development and maturation. To mimic this crucial early feature of myocardial formation, two-dimensional and three-dimensional epicardial differentiation protocols were developed that can predict the developmental stages of the human epicardium and the estimated timing of signaling (known to be able to identify the pre-epidimal in vertebrates). Importantly, these protocols are compatible with the myocardial organoid approach in that they use basic culture medium conditions (Figure 10, summary). Time-course RNA-seq and flow cytometry analysis confirmed efficient epicardial differentiation by broad expression of markers WT1, TCF21, and TBX18 (%) (Figure 10). Similar to in vivo conditions, activation of TGF-β, FGF, and PDGF signaling pathways had the potential to differentiate epicardial cells into SMCs expressing calponin, α-SMA, and SM22 in two dimensions, and CFs expressing DDR2 and vimentin (Figure 10).

[0167] To mimic the process of epicardial involvement, cardiac organoids and epicardial aggregates were co-cultured under basal medium conditions without the addition of growth factors (Figure 11). The results showed that epicardial cells efficiently spread onto the cardiac organoids within 4 days (Figure 11). Importantly, when these structures were further cultured under the same conditions without the addition of growth factors, epicardial cells were observed to migrate to the cardiomyocyte compartment and differentiate into SMC (Calponin+, SM22+) and CF (DDR2, vimentin, Decorin) (Figure 11). We conclude that epicardial involvement of cardiac organoids, co-culture without external signaling, is sufficient to stimulate epicardial involvement of cardiac organoids, epicardial migration to the CM compartment, and differentiation into SMC and CF. Overall, our high-throughput 3D differentiation approach can establish a self-organizing, systematically controlled cardiac organoid platform that mimics key aspects of human cardiomyogenesis.

[0168] Example 5: Mechanism of cardiac chamber formation While reductionist in vitro models of molecular and cellular systems cannot fully replicate the complexity of in vivo models, they have proven complementary and useful in resolving mechanistic questions. Leveraging our organoid platform, we were able to develop a bridge for semi-automated image analysis, enabling phenotypic quantification with high statistical power. Using this platform, we explored several possibilities regarding how signaling pathways control cardiomegaly, namely, that signaling at the mesoderm stage may influence lumen formation at the cardiac mesoderm stage, that cardiac mesoderm signaling may be decisive, or a combination of both. Taking advantage of the controlled stage / lineup of our high-throughput myocardial organoid platform, we systematically investigated the effects of doses of major mesoderm and cardiac mesoderm signaling pathways (e.g., WNT, BMP) on cardiomegaly. Surprisingly, doses of WNT signaling activation (e.g., CHIR99021) at mesoderm induction significantly affected lumen expansion at the cardiac mesoderm stage (Figure 4F). This dramatic effect was independent of cell proliferation. Furthermore, an optimal intermediate WNT activation level was found that promoted both cavity formation and CM differentiation. WNT-dependent cavity formation was observed in both organoids (with EC lining) and organoids (without EC) differentiated in the presence of VEGF (Figures 4 and 5). The optimal WNT activation level differs depending on the hPSC strain, but it can be easily determined using the cell culture system described in this book, for example, by using a parallel method on a well plate.

[0169] To identify downstream mediators of WNT that control cavity formation, RNA-seq analysis was performed to compare mesodermal gene expression profiles induced by high (large cavity) or low (small cavity) WNT signaling doses. Multiple components and targets of the BMP signaling pathway (BMP4, BMP2, BMPR) were found to be differently expressed genes at the start of cardiac mesoderm development. While BMPs are known as drivers that define myocardium at multiple stages, their direct role in cavity formation has not been proven. Therefore, it was investigated whether differences in BMP signaling levels at the cardiac mesoderm stage promote cavity formation. Inhibiting BMP signaling with the natural inhibitor noggin during the first two days of cardiac mesoderm development significantly inhibited cavity expansion (Figure 12). Furthermore, a clear dose-effect relationship of BMP signaling was observed, with cavity expansion increasing with higher BMP4 concentrations (Figure 12). These results clearly demonstrate that the WNT-BMP signaling pathway controls cavity expansion in myocardial organoids.

[0170] In vivo, several well-known cell biological mechanisms related to embryonic lumen formation exist downstream of signal regulation. However, the driving forces of cardiac chamber enlargement are not well understood, particularly in mammals and humans; therefore, we used a cardiac organoid platform to explore the fundamental mechanisms of cardiac chamber formation. Cardiac chamber dilation at optimal doses of WNT and BMP was not caused by differences in apoptosis or local proliferation, as seen from Caspase3 and Ki67 staining results (Figure 2). Instead, compression of the peripheral cardiac mesodermal layer was observed, characterized by high density of N-cadherin / actin / MYH10 / DAPI and lack of E-cadherin, while the inner layer where the cavity first formed appeared mesenchymal (Figure 2). Importantly, the compression of the outer layer was observed only under high WNT and BMP conditions that promoted cavity formation, and not under low WNT and BMP activation without cavity formation. These observations are consistent with the localization pattern of N-cadherins in compressed regions of cardiomyocytes and splenic mesoderm in vertebrate embryos, with mesenchymal regions facing the endocardial canal. As supporting a compact, epithelial barrier, the cavity structure of cardiac mesoderm was impermeable to low molecular weight (4 kDa) dextran. Since epithelial lumen expansion generally depends on osmotic gradients mediated by ion pumps, the involvement of chloride ion and sodium / potassium pumps in cardiac organoids was then investigated. The results revealed that WNT-BMP-driven compression of cardiac mesoderm is involved in cavity formation in vitro.

[0171] Example 6: Modeling of cardiac chamber defects Mutations in transcription factors (TFs) affect the development of the cardiovascular and ventricular systems and are the most well-known underlying cause of cardiac chamber defects, which lead to severe birth defects in humans. For example, disruption of NKX2-5 and HAND1, which are downstream of BMP, causes severe cardiac chamber defects in vertebrates and also causes the most severe cardiac malformations in humans. However, because these factors are present throughout the entire process of cardiomyogenesis and in multiple cell types, it has been difficult to identify their underlying mechanisms. Here, we hypothesized that one or more of these TFs downstream of the WNT-BMP axis are important for cavity formation in cardiomyocyte organoids. Therefore, we generated hPSC strains lacking heterozygotes and homozygotes of the HAND1 and NKX2-5 genes. Surprisingly, no defects in cavity formation during the cardiac mesoderm stage were detected in the NKX2-5 mutant strain. On the other hand, clear defects in cardiac chamber formation were observed in the HAND1 homozygote knockout strain at the cardiac mesoderm and CM stages (Figure 12B). The localization of N-cadherin in these KO organoids was also affected. Importantly, increasing the dose of WNT signaling during mesoderm induction partially maintained this effect, confirming the role of WNT in cavity enlargement. Furthermore, HAND1 expression levels were found to be higher under WNT signaling dose conditions that promoted cavity formation. Based on these findings, a high-throughput and systematically controlled cardiac organoid platform appears to be useful for quantitatively deciphering the mechanisms of cardiac development and genetic defects.

[0172] Example 7: Extended Materials and Methods General human pluripotent stem cell cultures—human pluripotent stem cell lines (WT H9, WiCell and constitutively fluorescent H9 clones (Wimmer et al., 2019, Nature 29, 40) WT and modified WTC, Allen Institute for Cell Science)—were cultured in a modified autologous medium based on the E8 culture system (Chen et al., 2011, Nature Methods 8, 424-429). The original E8 recipe was supplemented with 0.5% BSA (Europa Biosciences, #EQBAH70), autologous FGF2, and 1.8 ng / ml TGFβ1 (R&D RD-240-B-010). Cells were cultured on Corning or Eppendorf tissue culture plates coated with Vitronectin XF (Stem Cell Technologies, #7180) and subcultured every 2-4 days using TrypLE Express Enzyme (Gibco, #12605010) or PBS-EDTA (Biological Industries, 01-862-1B) until they reached approximately 70% confluence. Cells were regularly tested for mycoplasma.

[0173] hPSC differentiation into cardiomyocytes in 2D and 3D aggregates was performed by seeding 160-175,000 cells / 24-well plates in E8+ROCKi (Y-27632, Tocris #1254) for 24 hours. Subsequently, the cells were induced for 36-40 hours in CDM medium (Mendjan et al., 2014, Cell Stem Cell 15, 310-325) containing FGF2 (30 ng / ml, University of Cambridge), LY294002 (5 μM, Tocris, #1130), activin A (50 ng / ml, University of Cambridge), BMP4 (10 ng / ml, R&D Systems RD-314-HP-050), and CHIR99021 (R&D Systems RD-4423 / 50). To increase cell viability at this stage, 1 μg / ml insulin (Roche, #11376497001) was optionally added. This medium was named FLYABCH(Ins). After 36-40 hours, cells were induced for 4 days with daily medium changes in CDM medium containing BMP4 (10 ng / ml), FGF2 (8 ng / ml), insulin (10 μg / ml), IWP2 (5 μM, Tocris, #3533) (optionally, IWR-1 (1 μM, Tocris, #3532 / 10)) or XAV-939 (5 ​​μM, SelleckChem, #S1180)) and retinoic acid (0.5 μM, Sigma Aldrich, #R2625). This medium was named BFIIWPRa. Subsequently, the culture medium was changed to CDM medium containing BMP4 (10 ng / ml), FGF2 (8 ng / ml), and insulin (10 μg / ml), and the medium was changed daily for two days. This medium was referred to as BFI. To maintain the obtained cardiomyocytes, the culture medium was changed again to CDM medium containing insulin (10 μg / ml), and half of the medium was changed daily. This medium was referred to as CDM-I.Aggregated cardiomyocytes were maintained until day 21. Cardiomyocytes were dissociated using the STEMdiff cardiomyocyte dissociation kit (Stem Cell Technologies, #05025), and re-seeded as aggregates at 1000 cells / well in CDM-I and 5% FBS (PAA, #A15-108) on AggreWell 400 plates (Stem Cell Technologies, #34425). Two days later, the formed aggregates were transferred to an ultra-low-adhesion 96-well plate (Corning, #7007) in CDM-I and shaken at 58 rpm, 37°C, and 5% CO2. The culture medium was changed after two days, and the aggregates were used for analysis after four days.

[0174] The ECM molecules used—vitronectin (10 μg / ml), laminin-511 E8 fragment (Takara Bio, #T303, 0.05-2 μg / ml), and laminin-521 (Bio-Olaminin, #LN521-02, 0.1-5 μg / ml)—were used either as a pre-coat for the wells or added to the cell suspension before seeding. Further differentiation of cardiomyocytes was performed as described above.

[0175] Cardioid Generation - hPSCs were collected at approximately 70% confluence. 7500 cells / well were seeded for KO and BMP inhibition experiments, and 5000 cells / well for the remaining experiments to generate cardioids. Cells were seeded in 200 μl volumes into ultra-low-adhesion 96-well plates (Corning) containing E8+ROCKi and collected by centrifugation at 200 g for 5 minutes. After 24 hours, the formed aggregates were induced with FLyAB(Ins) containing the WNT activator CHIR99021 (see below for cell lineage-dependent concentrations). Cardiomyocyte differentiation was performed in the same manner as in 2D culture. To maintain the obtained cardioid cardiomyocytes, the medium was changed to CDM-I and replaced every two days. To stop endothelial cell differentiation, 100 nM sunitinib malate (Biovision, #1611) was added from the cardiac mesoderm stage (BFIIWPRa) onward. For BMP inhibition experiments, we used either 100 ng / m² of noggin (R&D System, #RD-6057-NG-025) or 0.2 μM of LDN-193189 (Stemgent, #04-0074).

[0176] To optimize the development of cardioids (more ventricular / medial EC lining), hPSCs were induced in two dimensions using F, Ly, B, low activin (4 ng / ml), Ins, and CHIR99021 (see below for cell line specific concentrations). Subsequently, 15,000 cells were seeded in 200 μl volumes into ultra-low adhesion 96-well plates (Corning) and differentiated according to the CM differentiation protocol described above.

[0177] Cardioid-generating and pluripotency-maintaining medium containing CM, EC, and fibroblast-like cells in the standard layer was refreshed 6 hours before cell seeding. Next, 2500 hPSCs were directly seeded for 36–40 hours in ultra-low adhesion 96-well plates (Corning) in FLYAB(Ins) medium containing CHIR99021 (see below for cell line-specific CHIR99021 concentrations) and ROCKi (5 μM). Next, the medium was changed to BFIIWPRa medium supplemented with VEGF-A (200 ng / ml, Peprotech, #AF-100-20), and the cells were cultured for 4 days, changing the medium daily. After that, the medium was changed to BFI+VEGF-A (100 ng / ml), and the medium was changed after 1 day, followed by 2 days of culture. For maintenance, CDM medium supplemented with VEGF-A (100 ng / ml) was used, and the medium was changed every two days. The protocol was followed except that IWP2 was not added during mesoderm induction and low WNT (CHIR99021: 4 μM) / low activin (4 ng / ml) was used to generate cardioids containing only EC and fibroblast-like cells. For Smart-Seq2 analysis, 7.5-day cardioids containing cardiomyocytes and endothelial cells were dissociated using the STEMdiff cardiomyocyte dissociation kit (STEMCELL Technologies, #05025), and GFP+CM, Tomato+EC, and GFP- / Tomato- cells were sorted using a homemade lysis buffer via FACS.

[0178] Freeze-damage of cardioids - Cardioids were temporarily transferred to a 10 cm container without culture medium and observed under an EVOS microscope (Thermo Fisher) placed in a laminar flow hood. Then, the cardioids were touched with an N2 iron rod cooled with liquid nitrogen until the wavefront of the frozen tissue / culture medium became clearly visible within the cardioid. The cardioids were then transferred back to wells containing maintenance medium for further cultivation.

[0179] Cell line-dependent CHIR99021 concentration - Different hPSC lines were found to respond to different concentrations of CHIR99021 (Wnt-activated) for optimal differentiation. This is consistent with a previous report (Strano et al., 2020, Cell Reports 31, 107732), where the optimal "high" (large lumen) and "low" (small lumen) CHIR99021 concentrations for different hPSC lines were empirically determined. In two-dimensional differentiation, H9 cells were induced with 1-2 μM CHIR99021, while WTC cells were induced with 3-4 μM CHIR99021. In three-dimensional cardioid differentiation, CM-only cardioids were generated according to the protocol described in the "Cardioid Development" section, with "high" and "low" CHIR99021 concentrations being 8 μM and 4 μM for WTC cells. 1.5-3 μM CHIR99021 was used for H9 cells. Following the section on "Cardioid Development," which includes CM, EC, and fibroblast-like cells, a "low" concentration of 4 μM of CHIR99021 was used for co-differentiation of CM / EC / fibroblast-like cells, while an "intermediate" concentration of 5-6 μM (optimal concentration) and a "high" concentration of 9 μM were used for CM / EC differentiation and cavity dilation.

[0180] Epicardial co-culture with cardioids: hPSCs were seeded in E8+ROCKi (5-10 μM) at 55,000 cells / 24 well plate 24 hours prior to differentiation. Cells were induced in CDM medium containing FGF2 (30 ng / ml, University of Cambridge), LY294002 (7.5 μM), BMP4 (10 ng / ml), and CHIR99021 (1.5 μM) (Mendjan et al., 2014, Cell Stem Cell 15, 310-325) (Iyer et al., 2015, Development 142, 1528-1541). After 36 to 40 hours, the differentiation medium was changed to CDM medium containing BMP4 (10 ng / ml), FGF2 (8 ng / ml), insulin (10 μg / ml), IWR-1 (1 μM), and retinoic acid (1 μM), and the culture was performed daily for 2 days. Subsequently, the medium was changed to CDM medium containing BMP4 (10 ng / ml), insulin (10 μg / ml), and retinoic acid (1 μM), with one medium change in between, and the culture was performed for 5 days (Guadix et al., 2017, Stem Cell Reports 9, 1754-1764). To maintain the obtained epicardium, cells were seeded at the end of differentiation on bovine plasma fibronectin (2 μg / ml, Sigma, #F1141) coated plates in CDM medium containing insulin (10 μg / ml) and SB431542 (10 μM, Tocris, #1614) supplemented with ROCKi at the start of seeding. The replicated epicardium was routinely passaged at a ratio of 1:3 every 3-5 days at 80-90% confluence.

[0181] To generate aggregated epicardium for use in the phagocytic assay, 8.5-day-old epicardium cells were dissociated using TrypLE Express Enzyme and re-seed as aggregates in AggreWell400 plates at 1000 cells / well in CDM-SBI and 5% FBS. After 2 days, an average of 8-12 aggregates / well were transferred to ultra-low adhesion 96-well plates (Corning) containing differentiated cardioids in CDM-I, and co-cultured with CDM-I medium refreshed every 2 days while shaking at 58 rpm, 37°C, and 5% CO2. Control (epidamicum only) aggregates were stored in CDM-SBI medium in ultra-low adhesion 96-well plates (Corning).

[0182] Two-dimensional anterior endothelial cell differentiation—pluripotent stem cells—were seeded at 100,000 cells / 24 wells (coated with Vitron-Echitin in E8 medium supplemented with 10 μM ROCK inhibitor). The following day, cells were induced with FLYABCH(Ins), 1-3 μM (for H9), and 3-6 μM (for WTC) CHIR99021 and cultured for 36-40 hours. Subsequently, the medium was changed to BFIIWPRa for two days. Then, differentiation medium containing CDM with 200 ng / ml VEGF and 2 μM forskolin (Sigma-Aldrich, #F3917) was provided for two days, followed by culture in CDM + 100 ng / ml VEGF for one day. EC was maintained in CDM supplemented with 100 ng / ml VEGF.

[0183] Culture of Human Cardiac Microvascular Endothelial Cells - Human cardiac microvascular endothelial cells (HCMECs) were obtained from PromoCell (PC-c-12285 HCMEC-c) and cultured using Endothelial Cell Growth Medium MV (PromoCell, #PC-C-22020) according to the manufacturer's instructions. For Smart-Seq2 analysis, HCMECs were dissociated with TrypLE Express Enzyme and FACS sorted into a homemade lysis buffer.

[0184] Chicken Cardiac Mesoderm Extraction and Culture: Explants were isolated from the cardiogenic region of developing chicken (Gallus gallus) embryos at Hamburger and Hamilton stages 7-8 and cultured in cardiac mesoderm (BFIIWPRa) medium at 37°C for 24 hours. The explants were then embedded, frozen, and fractured, and immunostained for further analysis as described below.

[0185] Freeze-fracturing was performed based on {Bag-ley:2017ga}. Briefly, 4% PFA-fixed tissue was cryoprotected overnight at 4°C with 30% sucrose in PBS, and the following day embedded using OCT cryoembedding medium (Seigen, #4586K1). The embedded tissue was frozen using a metal surface immersed in liquid nitrogen and stored in a -80°C freezer until sectioning was performed in a Leica cryostat. Sections were collected on Ultra Plus slides and stored at -20°C or -80°C until immunostaining. OCT was removed by washing with PBS before immunostaining.

[0186] Immunostaining - After fixation with 4% PFA (Sigma-Aldrich, #16005), the specimens were washed twice with 1×PBS, and the 3D structures were washed once more with PBS / Tween20 (0.1%, Sigma-Aldrich, #P1379) for at least 15 minutes each. The tissues were incubated for at least 15 minutes in a blocking solution of PBS (Gibco, #14190094) with 4% goat (Bio-Rad Laboratories, #C07SA) or donkey serum (Bio-Rad Laboratories, #C06SB) and 0.2% Triton X-100 (Sigma-Aldrich, #T8787). Subsequently, the primary antibody was applied in the blocking buffer as described above, for 1-3 hours at room temperature and overnight at 4°C for 2D samples, and for 2 days at 4°C on a shaker for 3D samples. After two washes with PBS / Tween20, the 3D tissues were incubated with secondary antibody solution at 4°C on a shaker for a further 2 days, while the 2D samples were incubated at room temperature for up to 2 hours. Following these washing steps and additional PBS washes, the tissues were prepared for analysis or stored in PBS at 4°C, and the slides were mounted using fluorescent mounting medium (Dako Agilent Pathology Solutions, #S3023). The 3D tissues were cleared with FocusClear (CellExplorer Labs, #FC-101) before imaging.

[0187] Trichrome staining—Masson trichrome staining (Bio Optica, #04-010802) was performed on 20 μM frozen sections as recommended by the manufacturer.

[0188] Electron Microscopy - The sample was fixed overnight at 4°C using a mixture of 2% glutaraldehyde (EM grade; Agar Scientific, Essex, UK) and 2% paraformaldehyde (EM grade; Electron Microscopy Services, Hatfield, US) in 0.1 mol / L sodium cacodylate buffer, pH 7.2. The organoid was then rinsed with the same buffer and fixed on ice with 1% osmium tetroxide (Electron Microscopy Services, Hatfield, US) in the buffer for 40 minutes. After three rinsing steps, the organoid was dehydrated on ice using high-grade acetone and embedded in Agar 100 resin (Agar Scientific, Essex, UK). 70 nm sections were selected using a 100-mesh Cu / Pd grid (Agar Scientific, Essex, UK) pre-coated with a holbaum support film, and then post-stained with 2% uranyl acetate (Merck) and Reynolds lead citrate.

[0189] Dextran conjugated with the dextran and Fluo-4 uptake assay-TAMRA (Sigma Aldrich, T1037) was added to cardioid cultures for 64 to 90 hours. The cardioids were live-imaged when cavity formation began. Fluo-4 AM (Thermo Fisher Scientific, #F14217) was mounted on 3D cardioids and 2D CMs to image and analyze calcium transients. After incubation for 15 minutes, the cardioids were incubated for another 15 minutes in Tyrode solution (Sigma Aldrich, #T2397). Subsequently, the cardioids were live-imaged and video-analyzed using FIJI software (Schindelin et al., 2012, Nature Methods 9, 676-682) to obtain signal intensity (F) and background (F0) of the region of interest.

[0190] Contraction characteristics were measured - 2D CM and 3D cardioids were imaged live, and the video was analyzed using a publicly available algorithm (Huebsch et al., 2015, Tissue Engineering Part c:Methods 21, 467-479) to determine contraction velocity and pulsation rate.

[0191] Optical action potential-cardioids were incubated in CDM medium at 37°C and 5% CO2. Before the experiment, the organoids were transiently exposed to the voltage-sensitive dye (VSD) FluoVolt (×0.5, room temperature for 30 minutes). Subsequently, the medium containing the VSD was replaced with fresh serum-free medium (DMEM, Sigma-Aldrich). The multiwell plates were placed in an environmentally controlled stage incubator (37°C, 5% CO2, water-saturated air atmosphere, Okolab Inc, Burlingame, CA, USA). The FluoVolt fluorescence signal was recorded from a 0.2 × 0.2 mm area of ​​the organoid. The excitation wavelength was set to 470 ± 10 nm using a light-emitting diode (LED), and the emitted light was collected using a photomultiplier tube (PMT, Cairn Research Ltd., Kent, UK). The fluorescence signal was digitized at 10 kHz. Subsequently, the 120-second recordings were analyzed offline using the pClamp software package v.10.0 (Molecular Devices, Inc., Sunnyvale, CA, USA). APD was measured at 30%, 50%, and 90% repolarization.

[0192] Image Acquisition and Analysis: Fixed hole mounts and sections were imaged using point scanning (upright Zeiss LSM800 Axio Imager with a 20× Apochromatic objective lens at 1× magnification), spinning disk confocal microscope (Olympus spinning disk system based on the IX3 series (IX83) inverted microscope, equipped with Yokogawa W1 rotating disk), or wide-field microscope (Zeiss Axio Imager 2, Axio Vert A1, Panoramic FLASH 250 II System). Live imaging experiments were performed using the Zeiss Celldiscoverer 7 or the spinning disk microscopes mentioned above. For high-throughput imaging and analysis, images were acquired using a Celigo Imaging Cytometer microscope (Nexcelom Biosciences, LLC) and analyzed using a custom-made manual written for FIJI software. A transmission electron microscope, the FEI Morgagni 268D TEM (FEI, Eindhoven, The Netherlands), was used for observation at 80kV. Images were acquired using an 11-megapixel Morada CCD camera (Olympus-SIS).

[0193] Statistical Data - Data are shown as mean + / - SD. To calculate statistical significance, the data were analyzed for normality and log-normality using the D'Agostino-Pearson and Shapiro-Wilk tests in Prism 8 software (GraphPad Software Inc.). If the data was normally distributed, parametric tests (two-sided t-test, one-way ANOVA) were performed to determine significance. If the data was not normally distributed, non-parametric tests (two-sided Mann-Whitney, Kruskal-Wallis) were performed using Prism 8 software. Correction for multiple comparisons using statistical hypothesis testing (Tukey's test for parametric, Dunn's test for non-parametric) was performed using Prism 8 software. The p-values ​​for significant differences are visualized as follows: *: p<0.05, **: p<0.01, ***: p<0.001, ****: p<0.0001.

[0194] Cells were dissociated using a flow cytometry-CM dissociation kit (Stem Cell Technologies, #05025). After centrifugation at 130g for 3 minutes, the cells were resuspended in 300 μl of PBS with 0.5 mM EDTA (Biological Industries, #01-862-1B) and 10% FBS (PAA Laboratories, #A15-108). Cells were acquired using a FACS LSR Fortessa II (BD) and analyzed using FlowJo V10 (FlowJo, LLC) software. FACS sorting was performed using a Sony SH800 Cell Sorter (Sony Biotechnology).

[0195] RNA isolation and RNA-seq / Smart-Seq2 / unicellular (sc)RNA-seq preparation were performed using the RNeasy Mini Kit (Qiagen, #74104). Bulk RNA-seq library preparation was performed using the QuantSeq 3'mRNA-Seq Library Prep Kit FWD (Lexogen GmbH, #015) according to the manufacturer's instructions. After library preparation, samples were verified for appropriate size distribution using a fragment analyzer (Advanced Analytical Technologies, Inc.) and submitted to the Vienna Biocenter Core Facility (VBCF), a next-generation sequencing (NGS) facility, for sequencing. For Smart-Seq2 analysis, 400 cells were divided into lysis buffers and stored at -80°C until further processing. The samples were QC'd / libraries prepared and sequenced at the VBCF NGS facility using a homemade Smart-Seq2 kit. For scRNA-seq, cardioids (two biological replicas, 3 cardioids each) were dissociated at day 7.5 of differentiation, and the cells were submitted to the VBCF NGS institution for library preparation using a 10× Genomics Chromium platform (10× Genomics, California, USA).

[0196] Bioinformatics analysis and trimming were performed on Smart-Seq2 experiments using trim-galore v0.5.0 and on QuantSeq 3'mRNA-Seq experiments using BBDuk v38.06 (ref=polyA.fa.gz, truseq.fa.gz k=13 ktrim=r useshortkmers=t mink=5 qtrim=r trimq=10 minlength=20). Reads mapping to extra sequences contained in the iGenomes UCSC hg38 reference (human rDNA, human mitochondrial chromosome, phiX174 genome, adapter) were removed using bowtie2 v2.3.4.1 alignment. The remaining reads were analyzed using genome and UCSC gene annotations provided in the iGenomes UCSC hg38 bundle (support.illumina.com / sequencing / sequencing_software / igenome.html). Reads were aligned to the hg38 genome using star v2.6.0c, and intragenetic reads were counted using featureCounts (subread v1.6.2) with strand-specific read counting for QuantSeq experiments (-s 1). Differential gene expression analysis of raw counts and principal component analysis of variance-stabilized count data were performed using DESeq2 v1.18.1. In addition, functional annotation enrichment analysis of differentially expressed genes was performed using clusterprofiler v3.6.0 in R v3.4.1.

[0197] Bulk tissue cell type deconvolution was performed using MuSiC v0.1.1 (Wang et al., 2019, Nature Communications 10, 1-9). Cell type-specific marker genes and cell type-specific single-cell expression criteria for developing human hearts (Cui et al., 2019, CellReports 26, 1934-1950.e5) were used. Bulk cardioid RNA-seq samples were processed using a previously reported pipeline with the hg19 UCSC iGenomes reference to ensure consistency with previously reported data. The proportion of cell types from developing hearts in bulk cardioid RNA-seq samples was estimated, and the MuSiC estimated proportions were visualized in a heatmap.

[0198] Single-cell RNA-seq reads were processed using cellranger count v4.0.0 with a pre-built 10× GRCh38 reference, refdata-gex-GRCh38-2020-A. Count data were further analyzed using Seurat v3.2.2. Over 500 genes were detected, and cells with mitochondrial content less than 15% were retained. Doublets detected by scDblFinder v1.4.0 were removed. As a result, 9632 cells were further analyzed for cell type analysis within the cardioids. To compare ventricular-like CMs, 1717 CMs under standard conditions (intermediate CHIR / high activin) and 5097 CMs under optimized low CHIR / low activin conditions were analyzed. Log-normalized expression values ​​were derived using the log-normalization method with the default scale factor of 10000. Replication was integrated in 15 dimensions and with default settings using the FindIntegrationAnchors and IntegrateData functions. Using singleR v1.4.0, I annotated single cells using the Cui et al. reference (supra). The 2D representation was generated using uniform manifold approximation and projection with uwot v0.1.9.

[0199] Proteomics - Cells were lysed in 100 mM 8M urea 4-(2-hydroxyethyl)piperazine-1-ylethanesulfonate buffer (HEPES), reduced with 10 mM 1,4-dithioerythritol and 1 U benzonase (MERCK KGaA, #1.01654.0001), and alkylated with 20 mM 2-iodoacetamide. The cells were then immersed in 100 mM 4M urea HEPES with LysC (Wako, #121-05063, 1 / 100 (w / w) protease / substrate) at 37°C for 3 hours, followed by trypsin digestion (Promega, #V5280, 1 / 100 (w / w) protease / substrate) and retested overnight at 37°C. The peptides were desalted using a reversed-phase solid-phase extraction cartridge (Sep-Pak c-18, Waters, #186000308), dried under vacuum, reconstituted to neutral pH with HEPES, and labeled with TMT10-plex (Thermo Fisher Scientific, #90110) as instructed by the manufacturer. Equal amounts of the TMT-labeled peptides were pooled and fractionated by high-pH reversed-phase chromatography (UPLC Peptide CSH c18 column, 130 Å, 1.7 μm, 1 mm × 150 mm, ACQUITY) to obtain 10 fractionations.

[0200] The samples were separated by reverse-phase chromatography (75 μm × 250 mm PepMap C18, particle size 5 μm, Thermo Fisher Scientific), and within 60 minutes, they were spread to a linear gradient from 2% to 80% acetonitrile in 0.1% formic acid (RSLC nano, Dionex-Thermo Fisher Scientific). The samples were then analyzed by electrospray ionization tandem mass spectrometry (Orbitrap QExactive HFX, Thermo Fisher Scientific) via MS / MS. The instruments were operated with the following parameters. MS1 resolution 120,000; MS1 AGC target 3e6; MS1 maximum injection time 50ms; MS1 scan range 380~1650m / z; MS2 resolution 45,000; MS2 AGC target 1e5; Maximum injection time 250; TopN 10; Separation window 0.7m / z; Fixed first mass 110m / z; Normalized collision energy 35; Minimum AGC target 1e4; Peptide matching pre-selection; Isotope exclusion ON; Dynamic exclusion 30s.

[0201] All MS / MS data were processed and analyzed using Proteome Discoverer 2.3 (PD 2.3.0.484, Thermo Scientific), and a search was performed against the Homo sapiens database (SwissProt TaxID=9606) (v2017-10-25) using MSAmanda v2.0.0.14114. Maximum missed cleavage: 2. Tests were performed on iodoacetamide derivatives on cystine and peptide N-terminal 10-plex tandem mass tags (fixed mod.), oxidized methionine, and 10-plex tandem mass tags on lysine (variable mod.). Peptide mass tolerance: ±5 ppm; Fragment mass tolerance: ±15 ppm. The proteins and peptides were filtered to 1% FDR using a percolator, and the reporter ion was quantified using IMP Hyperplex (Doblmann et al., 2019, Journal of Proteome Research 18, 535-541) (ms.imp.ac.at / index.php?action=hyperplex).

[0202] Generation of a MYL7-GFP / CDH5-Tomato double reporter strain - An endogenously tagged WTC MYL7-GFP hPSC strain was obtained from the Allen Institute for Cell Science (cell line number: AICS-0052). A gBlock of the CDH5 promoter sequence (-1135~-5 relative to TSS) (Prandini et al., 2005) was ordered from Integrated DNA Technologies and cloned into a modified backbone of a vector integrated into the AAVS1 locus using TALEN technology (Hockemeyer et al., 2009, Nature Biotechnology 27, 851-857) according to Bagley et al. The modified backbone contained a flanking tandem repeat of the co-chicken HS4 insulator (2×CHS4). Thus, the following reporter expression cassette was inserted into the AAVS1 locus: 2×CHS4-CDH5promoter-dTomato-WPRE-SV40-2×CHS4. Nucleofection and clone picking / verification were performed as described in (Bagley et al., 2017, Nature Methods 14, 743-751).

[0203] Generation of HAND1 and NKX2-5 knockout cell lines - HAND1 and NKX2.5 were knocked out in H9 cells using CRISPR / Cas9. sgRNAs targeting the specified sites were identified using the Sanger Institute Genome Editing (WGE) website and the Benchling sgRNA designing tool (HAND1_sgRNA1:GAGCATTAACAGCGCATTCG (SEQ ID NO: 1); NKX2.5_sgRNA1:GACGCACACTTGGCCGGTGA (SEQ ID NO: 2); NKX2.5_sgRNA2:ACTTGGCCGGTGAAGGCGCG (SEQ ID NO: 3)). sgRNA was cloned into pSpCas9(BB)-2A-Puro(PX459) V2.0 (Feng Zhang Lab; Addgene plasmid #62988; n2t.net / addgene:62988; RRID: Addgene_62988) according to the Zhang Lab General Cloning Protocol (Ran et al., 2013, Nature Protocols 8, 2281-2308). Cells were transfected using P3 Primary Cell 4D-Nucleofector™ X Kit S (Lonza-BioResearch, Cat #: V4XP-3032) and Amaxa™ 4D-Nucleofector™ (Lonza-BioResearch). After nucleofection, cells were incubated for 24 hours in E8 with 10 μM Y-27632 (Cat #72302). After that period, cells were selected for 48 hours with puromycin (0.2 ng / μL; Sigma-Aldrich, Cat #P8833). Following this treatment, the cell culture medium was returned to E8 with 10 μM Y-27632 (Cat #72302) added to promote regrowth. Once the cells formed colonies, they were extracted and transferred to a 96w-plate (Corning, Cat #CLS3370). The success of the editing was first evaluated at the pool level. Subsequently, single colonies were independently genotyped twice to confirm successful knockout.Genome editing at the pool and clone levels was evaluated using Synthego's online tool ICE (ice.synthego.com / ♯ / ). Primer: HAND1_G1_forward 5'-CACCGAGCATTAACAGCGCATTCG-3'(Sequence ID 4) HAND1_G1_reverse 5'-AAACCGAATGCGCTGTTAATGCTCC-3'(Sequence ID: 5) NKX2.5_G1_forward 5'-CACCGGACGCACTTGGCCGGTGA-3'(Sequence ID 6) NKX2.5_G1_reverse 5'-AAACTCACCGGCCAAGTGTGCGTCC-3'(Sequence ID: 7) NKX2.5_G2_forward 5'-CACCGACTTGGCCGGTGAAGGCGCG-3'(Sequence ID 8) NKX2.5_G2_reverse 5'-AAACCGCGCCTTCACCGGCCAAGT-3'(Sequence ID: 9)

[0204] Example 8: Formation of ventricular-like structures in a test tube To investigate whether three-dimensional ventricular-like structures can be intrinsically formed in vitro, we developed a differentiation approach based on the temporal control of key signaling pathways for cardiomyogenesis: activin, BMP, FGF, retinoic acid, and WNT. By replicating in vivo developmental stages, we sequentially identified hPSCs into mesoderm, cardiac mesoderm, and beating cardiomyocyte progenitor cells with over 90% efficiency in two-dimensional culture (Mendjan et al., 2014, Cell Stem Cell 15, 310-325) (Figure 14A). To screen for factors sufficient to stimulate intrinsic three-dimensional cardiomyocyte structure formation in two-dimensional culture, we supplemented the culture medium with selected ECM proteins involved in mesoderm development (Yap et al., 2019, Trends in Cell Biology 29, 987-1000). When Laminins 521 / 511 was added before mesoderm induction, cells self-assembled intrinsically, and after 7 days of differentiation, hollow, pulsating three-dimensional structures expressing the CM marker TNNT2 were prominently formed (Figure 21A). When complete cardiomyocyte differentiation was performed using three-dimensional non-adherent high-throughput culture, it was found that exogenous ECM was not required for rapid and reproducible self-assembly into pulsating cavity-containing structures positive for CM markers ACTN2, TNNT2, TNNI1, MYL7, TTN, NPPA, and ATP2A2 (Figures 14B, 14c, 21B, 21c). At the ultrastructural level, these CMs contained organized sarcomeres and were interconnected via intercalating disks (Figures 21C', 21D). Self-assembly was robust in the WTC hiPSC line, which includes a vast collection of live fluorescence reporter resources (Roberts et al., 2019, Stem Cell Reports 12, 1145-1158), the widely used hESC lines H9 and H7, and three hiPSC lines routinely used for cerebral organoid generation (Figures 21E, 21E'). CM differentiation efficiencies, measured by flow cytometry (N=3, n=5) and 3D Z-stack image analysis (N=3, n=8), averaged approximately 90% (Figures 14D, 21F). Hereafter, these cavity-containing myocardial structures will be referred to as cardioids.

[0205] Next, we aimed to characterize the cardioid at the molecular level. RNA-seq time-course analysis of the cardioid revealed an expression signature most similar to that of the first cardiac area (FHF) lineage of cardiac mesoderm (HAND1+, TBX5+, NKX2-5+, TBX1-) (Figure 14E). This is responsible for the development of the cardiovascular system (the major precursor of the left ventricle and smaller ventricles) in vivo. During the identification and maturation of the cardioid, the expression of structural and ion channel genes, as well as the expression of β-adrenergic receptors 1 and 2, increased (Figure 22A). Comparing the expression profiles of cardioids with those of 2D differentiated CMs, genes encoding ion channels (e.g., HERG channel KCNH2), structural proteins (TNNI1, TTN, MYH6), cardiac transcription factors (TBX5, MEF2c), and sarcoplasmic reticulum proteins (RYR2, ATP2A2) showed higher expression levels in the 3D cavity-forming structure, suggesting improved function (Figure 22B). GO-term analysis revealed that cardioids exhibited gene expression patterns related to cardiac morphogenesis and development, which were significantly upregulated compared to 2D CMs (Figure 22C) and aggregated 3D CM microtissues (Figure 22D). In both models, heartbeats began between days 5 and 7 of differentiation and continued at similar rates and frequencies (Ca2+ transients, beating frequency) (Figures 22E, 22F), and cardioids could be maintained in culture for at least 3 months. In this way, we succeeded in reproducing self-assembly and generating functional hPSC-derived cardioids while maintaining molecular identity.

[0206] Furthermore, we explored the possibility of cardioid CM subtypes. Initial RNA-seq analysis of cardioids revealed mixed ventricle (IRX4+, MYL2+) and atrial (NR2F2+, KCNJ3+) profiles consistent with their FHF origin (Figures 22A, 22B). Considering that mesoderm induction and low doses of RA signaling are important for immature ventricular formation, we optimized the doses of WNT and activin during mesoderm induction and reduced the RA dose during the cardiac mesoderm stage. These optimizations resulted in increased expression of ventricle-specific markers (IRX3, IRX4, HEY2, MYL2) and almost complete elimination of atrial-specific markers (NR2F2, HEY1), as revealed by RTqPCR, immunocytochemistry, and single-cell RNA-seq (scRNA-seq) analysis of CMs (Figures 14F, G, H). Consistent with these data, potential-of-action measurements using the FluoVolt dye assay showed a predominantly ventricular-like profile (Figure 22G). In conclusion, the cardioid shape may have been directed towards an early left ventricle-like identity.

[0207] Example 9: Cardiac mesoderm self-organizes and forms cavities in vitro and in vitro. Next, this system was used to determine whether the cavities within the cardioid were formed by intrinsic morphogenesis. Analysis of the time course of cavity formation revealed that the cavities initiated in the cardiac mesoderm (HAND1+) stage, preceding the expression of major cardiac structural markers such as MYL7, and expanded robustly (Figure 14E; Figure 15A). The cardioid expanded rapidly and reproducibly through expansion and the formation of multiple cavities. Most of the small cavities eventually merged into one large cavity. Cavity expansion was not caused by apoptosis or differences in localized expansion, as evidenced by the results of cleavage-type CASP3 and MKI67 staining (Figure 15A). Importantly, SOX17 + / EOMES +The endoderm disappeared during differentiation, resulting in the failure of cardiac mesoderm cavity formation as a result of endoderm direction (Figure 23A). This is consistent with in vivo findings, as both hearts are formed even when foregut endoderm morphogenesis is disrupted. Lumen formation also occurred when VEGF-driven endothelial cell (EC) differentiation was inhibited using the potent VEGFR inhibitor sunitinib (Figure 23B), suggesting an endothelium-independent mechanism. It was concluded that cardiac mesoderm possesses an inherent ability to self-organize into CM-based ventricle-like structures.

[0208] In vivo, cardiac mesoderm does not require foregut endoderm for the basic morphogenesis of the heart in mice (Li et al., 2004, Science 305, 1619-1622) and chickens (DeHaan and DeHaan, 1959, developmental Biology 1, 586-602). Therefore, we investigated whether mesoderm dissected in vitro from developing chicken embryos could form ventricle-like structures under the conditions developed for the self-organization of the human heart. Surprisingly, in the absence of SOX2+ foregut, excised chicken mesoderm developed into ventricular-like structures that pulsated in vitro, similar to human cardiac mesoderm, demonstrating that in vitro myocardial auto-formation is well preserved under acceptable conditions (Figures 15B, 23C, 23D).

[0209] In addition to self-morphogenesis during the identification and determination phase, the intrinsic self-pattern formation of a homogeneous starting cell population is an important feature of self-organization. For this purpose, detailed analysis of cardiac mesoderm was performed to determine when the initial self-pattern formation phenomenon occurs. Although the mesoderm appears homogeneous during the induction phase, it was confirmed that peripheral signals of F-actin and membrane-bound β-catenin increase at the start of the cardiac mesoderm phase. Subsequent cavitation was consistent with the increase in mesoderm density in the periphery, reflected by the accumulation of F-actin and N-cadherin and high nuclear density (Figures 15c, 15D). Since the cavity structure was impermeable to low molecular weight (4kDa) dextran, it is thought that this high-density cardiac mesoderm layer functions as a permeable barrier. On the other hand, in the central part of the developing structure where the cavity first appears, the signals of N-cadherin and β-catenin decreased, exhibiting a looser appearance (Figure 15C). These observations are consistent with the in vivo patterns of N-cadherin and β-catenin in the high-density dorsal regions of the cardiac mesoderm and the low-density regions facing the endocardial canal and foregut endoderm (Linask, 2003, Birth Defects Research Part c: Embryo Today: Reviews 69, 14-24). Thus, we conclude that human cardioids possess important features such as self-organization, ongoing designation, intrinsic self-patterning into the mesodermal layers, and self-morphogenesis that forms cavities.

[0210] Example 10: WNT and BMP control cardioid self-organization. Next, we used cardioids to elucidate how signaling controls morphogenesis and patterning during cardioid identification. A high-throughput cardioid platform was combined with a custom-made semi-automated imaging / analysis FIJI pipeline to quantify phenotypes with high statistical power. Using this setup, we investigated which signals control cardioid self-organization and at what stage of mesoderm identification they act. First, we systematically examined the effects of key mesoderm and cardiac mesoderm signaling levels (e.g., WNT, BMP) on cardiac chamber autoformation. Surprisingly, high doses of WNT signaling during mesoderm induction were found to promote cavity enlargement in late cardiac mesoderm (Figures 16A, 16B), a previously unreported finding. Intermediate WNT doses promoted both cavity morphogenesis and CM identification. The optimal WNT activation range was consistent across hPSC strains but differed between lines, consistent with many studies showing line-specific signaling responses. Importantly, at the highest WNT dose, cavity formation was promoted without CM identification (Figure 16A), highlighting a significant difference in signaling regulation between cell destination identification and morphogenesis.

[0211] To identify WNT downstream mediators controlling cardiac chamber morphogenesis, RNA-seq analysis was performed, comparing mesoderm gene expression profiles induced by higher (large cavity) and lower (small cavity) WNT signaling doses. Genes with differing expression in late cardiac mesoderm included known BMP signaling cardiomyocyte mediators (BMP4, BMP2, BMPR2) and their mesoderm targets (HAND1, IRX3) (Figure 16C). BMP drives cardiomyocyte identification at multiple stages. Therefore, we investigated whether BMP can direct patterning and morphogenesis for cardiac chamber formation. To answer this question, BMP signaling was inhibited during the first two days of cardiac mesoderm using the natural inhibitor noggin or the compound LDN193189. BMP inhibition resulted in impaired epigastric morphogenesis, densification of cardiac mesoderm, and reduced cardioid size, although the cell count per cardioid remained stable (Figures 16D, 16E, 16F, 16G). On the other hand, WNT inhibition at the cardiac mesoderm stage was not necessary for cavity formation (Figure 16H). These findings highlight that the control of identification and the control of morphogenesis may be fundamentally different processes, and that the mesoderm WNT-BMP signaling axis controls both self-pattern formation and self-morphogenesis—both important self-organizing processes.

[0212] Example 11: HAND1 acts on the self-assembly of cardinoids. Signal transduction and mutations in downstream transcription factors affect cardiovascular and ventricular development and cause severe congenital heart malformations in humans. For example, in left ventricular heart failure syndrome, the most severe congenital defect in humans, disruption of the levels of the BMP regulatory genes NKX2-5 and HAND1 is associated with a significantly reduced cardiac chamber in the left ventricle. The earliest phenotypes in mutant Nkx2-5 and HAND1 mice appear as defects in cardiovascular and left ventricular early morphogenesis, respectively, but the disease etiology and underlying morphogenetic mechanisms in humans are less clear. Therefore, we generated either HAND1 or NKX2-5 knockout (KO) hPSC lines and evaluated whether these genes are required for endogenous self-organization in the absence of non-cardiac tissues. In NKX2-5 KO lines, no defects in cavity formation were detected at the cardiac mesoderm stage, and they ultimately formed TNNT2+ cardioids (Figures 24A, 24B, 24C). HAND1 expression in NKX2-5 KO cardiac mesoderm remained unaffected, which is consistent with the delayed onset of NKX2-5 relative to HAND1 expression in cardioids (Figures 24C, 14E), similar to human and mouse FHF cardiac mesoderm. However, HAND1 expression appeared to be reduced at the CM stage in NKX2-5 KO cardioids (Figure 24C). This is consistent with findings in mouse and human CMs where NKX2-5 acts upstream of HAND1.

[0213] On the other hand, in HAND1 KO, a decrease in NKX2-5 protein levels was observed in the cardiac mesoderm but not in the pericardial chamber (CM) (Figures 17A, 17D, 24E). These results suggest that HAND1 functions upstream of NKX2-5 in the cardiac mesoderm, and that NKX2-5 functions upstream of HAND1 in late CM, highlighting the importance of stage-specific analysis. Consistent with this hypothesis, HAND1 KO cardioids, unlike NKX2-5-deficient cardioids, showed a clear defect in cardiac chamber self-organization and size at the cardiac mesoderm stage. This phenotype manifested as cardioids forming smaller cardiac chambers (N=3, n=130) (N=9, n=246) (Figures 17B-17E, 24G). These defects were not caused by differences in cell number per pericardium, as both KO and WT showed similar cell numbers (Figure 24F). Importantly, despite these defects in cardiac mesoderm patterning and morphogenesis, subsequent CM identification (TNNT2+) was still functioning in HAND1 KO cardioids (Figure 24E). These observations further highlight the crucial distinction between the control of cell identification and tissue patterning and organ morphogenesis, which manifest as myocardial malformations in cardioids.

[0214] Next, we investigated whether the HAND1 KO phenotype could be rescued by exogenous signaling factors. Increasing the dose of WNT signaling during mesoderm induction rescued the HAND1 KO phenotype, confirming the involvement of WNT in cavity morphogenesis (Figure 17F, Figure 17G). Consistently, human-specific FHF and HAND1, an early ventricular marker, were upregulated under high WNT conditions, which also promoted cavity enlargement (Figure 16C). Furthermore, HAND1 protein levels decreased with BMP inhibition, confirming that the WNT-BMP-HAND1 axis drives cardioid cavity self-organization (Figure 16G). Taken together, these data demonstrate that self-organization and genetic cardiac defects can be quantitatively modeled on our high-throughput cardioid platform.

[0215] Example 12: WNT, activin, and VEGF regulate endothelial and myocardial self-organization. Next, to elucidate the signaling pathways that instruct the patterning and separation of the myocardium and endocardium to form the endocardium, which is a characteristic of the ventricular chamber, we explored whether cardioids could be used. To examine these relationships, we compared the RNA-seq time courses of cardioids generated with high and low amounts of WNT activation during mesoderm induction. As a result, it was found that under low-concentration WNT activation, the expression of VEGF-A and other EC specification factors (ETV2, TAL1, LMO2, PECAM1) increased during the cardiac mesoderm stage (Figures 18A, 18A'). At the same time, single-cell RNA-seq analysis of ventricular-like cardioids induced under low-concentration WNT / low-concentration activin conditions revealed that the proportion of CMs expressing VEGF-A was higher compared to cardioids induced under medium-concentration WNT / high-concentration activin (Figure 18B). Therefore, we hypothesized that low amounts of WNT and activin signals coordinate the co-specification of CMs and ECs by inducing VEGF-A in the cardiac mesoderm and CM and stimulating EC differentiation. Indeed, it was found that the combination of low-concentration levels of WNT signal and low-dose activin signal during mesoderm induction promoted the subsequent self-organization of ECs within the cardioids (Figures 18C, Figure 18D). These ECs often (55.2%, N = 4, n = 29) formed a partial lining inside the cardioid cavity (Figures 18c, 18D, 18E, 25A, 25A'), but not outside the cardioid, and resembled the architecture of in vivo tissues. On the other hand, under conditions of high doses of WNT and activin, no EC self-organization was observed without exogenous VEGF (Figure 18C). In conclusion, the optimal doses of WNT and activin signals during mesoderm induction control the subsequent EC self-organization that results in a partial lining in the cardioid cavity.

[0216] Next, we investigated the effect of exogenous VEGF on the co-specification of CM and EC in cardioids (Figure 18F). When VEGF-A was added after the CM specification stage, the formation of an EC layer lining the cardiac chambers was occasionally observed, and some EC specification was also seen on the surface of the cardioid (Figures 25B, 25c, 25D). To further clarify whether the early self-organization of CM and EC layers in cardiac mesoderm could be controlled, VEGF-A was added at this stage. In the presence of VEGF-A, when CM and EC co-differentiated from cardiac mesoderm into structures containing cavities, they separated into CM and EC layers (Figures 18F, 18F', 18F''). Here, when optimal (intermediate) WNT activation was used during mesoderm induction, the EC layer consistently surrounded the CM layer (Figure 18F'', Figures 26A, 26c, 26D). Furthermore, a third layer of COL1A1+ cells appeared next to the EC layer (Figure 18F''). In contrast to cardioids, the aggregates of EC and CM that initially differentiated in two dimensions formed a mixed network, but no pattern was formed into layers (Figure 26B). This suggests that VEGF stimulates the initial separation of the two layers, a crucial aspect observed in embryonic cardiac mesoderm and cardiac tubule stages. However, exogenous VEGF was insufficient to control the correct lateral and medial orientation of the EC lining. Thus, the dosage of WNT / activin and the timing of VEGF signaling modulate the identification of CM and EC lineages, luminal morphogenesis, and in vivo-like patterninization.

[0217] To further elucidate the WNT and VEGF signaling regulation of ciliary identification and morphogenesis in cardioids, we investigated whether the EC cell layer can be formed without CM codifferentiation. We found that without WNT inhibition, the presence of VEGF during the cardiac mesoderm stage reproducibly formed cardioid-like structures mainly consisting of CM-free ECs and COL1A1+ cells (Figure 18G, G', Figure 26F-). This observation again highlights how morphogenesis signaling regulation is decoupled from cell identification signaling regulation in cardioids. Taken together, these data suggest that cardioids self-organize and give rise to the first two lineages of the heart, and therefore can be used to elucidate aspects and stages of myocardial and endocardial codevelopment.

[0218] Example 13: Formation of cardioids by endocardial cells and fibroblast-like cells All self-organizing organoids share related tissue-like identification, patterning, and morphogenesis processes, but are distinguished by their organ-specific cell types. Therefore, the heterogeneity of cells within cardioids was investigated. Using WNT intermediate activation and VEGF, the ratio of CM to EC remained stable at 41% (MYL7+) and 53% (CDH5+) (Figure 18F'), facilitating quantitative analysis using this model. This observation was further confirmed by deconvolution of bulk RNA-seq data (Wang et al., 2019, Nature Communications 10, 1-9) and reference to cardiac cell type-specific single-cell expression in developing human hearts (Cui et al., 2019, CellReports 26, 1934-1950.e5) (Figure 27A). Smart-seq2 analysis of sorted cells (Figure 26E) and scRNA-seq analysis used to classify cells based on in vivo human cardiac data (Cui et al., 2019) further confirmed major CM and EC marker profiles and GO terms (Figures 19A, 27D, 27E), while the remaining cells (MYL7- / CDH5-) expressed genes associated with putative EC-derived fibroblast-like cells (e.g., SOX9, MSX1 / 2, COL1A1, COL3A1) (Figures 19A, 26E, 27F). Bulk proteomics analysis of cardioids confirmed CM and EC proteome expression signatures (Figure 27B).

[0219] All vascular tissues and organs contain specific EC subtypes, and therefore, the endocardium has an identity-specific EC gene expression signature. To determine the identity of ECs in cardioids, Smart-seq2 analysis was performed on sorted CDH5+ cardioid ECs. Those, We compared cardioid-derived ECs (endothelial cells) with those generated using our 2D differentiation protocol, which employs well-established 2D differentiation protocols (Patsch et al., 2015, Nature Cell Biology 17, 994-1003), vascular organoids (Wimmer et al., 2019, Nature 29, 40), human umbilical vein endothelial cells (HUVECs), and human cardiac microvascular endothelial cells (HCMECs) (Figure 19C, 19D). Cardioid-derived ECs were found to be the most similar to ECs from vascular organoids, despite the age difference (7.5 days vs. 18 days). Importantly, cardioid-derived ECs showed increased transcription levels of cardiac transcription factors such as GATA4 / 5 and genes related to endocardial-like identity (NFATC1, NPR3) (Figure 19D). Importantly, NFATC1 was found in the proteomic analysis data (Figure 27B), and NPR3 was found in the scRNA-seq data (Figure 19A). The anterior HOX gene expression profiles were consistent with those of HCMECs derived from adult human hearts and ECs derived from two-dimensional cardiac mesoderm (anterior ECs), but not with those of other analyzed, more posterior EC subtypes (Figure 27C). Therefore, the signature of cardioid-derived ECs is consistent with an endocardial-like identity.

[0220] The endothelium's ability to sense fluid flow, pressure, and mechanical expansion and contraction is an instrumental requirement for its developmental and physiological roles, particularly in the heart. As expected from a benevolent model of cardiac development, Smart-seq2 analysis of cardioid organoid endocardium (ECs) sorted from cardioids showed upregulation of mechanosensory genes (SOX18, KLF2, FOXO1, FOS) compared to 2D ECs (Figures 19B, 19B', 19B'', 19D), similar to more mature 3D vascular organoid ECs. These observations were confirmed by scRNA-seq datasets and SOX18 staining results. Markers of EC maturation (VWF, TEK, TIE1) were also upregulated in cardioid ECs compared to 2D EC differentiation (Figure 19D). These results suggest that the self-organization of CMs and ECs in 3D triggers essential aspects of endocardial identity and endothelial physiology.

[0221] Example 14: A three-system cardioid platform as a developmental injury model Following endocardial formation, the epicardium envelops the initial myocardial ventricles, thereby adding a third major cardiac system to the heart. The epicardium develops from small cell clusters called prepericardial organelles and eventually covers the outer surface of the heart. The epicardium differentiates into smooth muscle cells (SMCs) and cardiac fibroblasts (CFs) in response to signals from the CM layer (TGF-β, PDGF-β, FGFs), which are involved in further development, maturation, and regeneration of the heart in the event of injury. To mimic this self-organizing process in cardioids, a cardioid-compatible epicardial differentiation protocol was developed based on signaling sequences known to specify the precardia in vertebrates and hPSCs (Figures 20A, 28A, 28B, 28C). Next, to investigate whether the endogenous expression of these signaling factors by cardioids is sufficient to stimulate epicardial / cardioid interactions, cardioids and epicardial aggregates were co-cultured in a medium without exogenous TGF-β, FGF, and PDGF (Figure 28D). Within 2–7 days, epicardial cells were observed spreading over the cardioid (Figures 28E, 28F, 28G). After 7 days, without the addition of growth factors, epicardial cells interacted with the CM layer, migrated to the CM layer, and underwent differentiation (Figures 20B, 20c, 20D, 28H, 28I). The migrated cells decreased epicardial WT1 and increased the SMC and CF markers ACTA2 (Figures 20D, 28I) and COL1A1 (Figures 20C, 28H) to levels that would occur in vivo. Surprisingly, some of the migrating epicardial-derived cells began interacting with cardioid EC (Figures 20B, 20D). In conclusion, co-culture without external signals induces intrinsic spreading to the epithelium on the cardioid, its inward migration, differentiation, and interaction with CM and EC.

[0222] The hypothetical advantage of self-organizing development or cardioid models lies in their pathophysiological responses. However, current myocardial in vitro models fail to replicate either the myocardial regeneration seen in fetal, early postnatal, and adult in vivo injury models, or the fibrosis seen in disease models and patients. For example, previous tissues have shown only limited proliferation after cryo-injury of bioengineered cardiac organoids, without the early extracellular matrix (ECM) accumulation typically seen in the early stages of both regeneration and fibrosis. Because cardioids encompass all three major cardiac lineages, depend solely on developmental mechanisms, and do not require an external ECM scaffold, it was thought that cardioids might evoke a more physiological response to cryo-injury. To investigate the potential of the cardioid platform as a model for injury, cryo-injury was performed on monophyletic (CM only) and triphyletic (CM (Figure 20G, 20F), EC (Figure 20H, 20I, 28J, 28K, 28L) or epicardium (Figure 20E, 20G) and associated fibroblast-like cells) cardioids (Figures 29-34). Injury sites were characterized by strongly compressed DAPI signaling, severe necrosis detected by TUNEL staining (Figure 28J), limited apoptosis, and a clear CM proliferation response (Figure 28K). In time-course experiments, trichrome staining was observed at the injury site, indicating localized ECM accumulation (Figure 20G). Monophyletic cardioids showed low fibronectin accumulation and the absence of COL1A1+ expressing fibroblast-like cells at the injury site (Figures 20G, 20F). On the other hand, in the three cardioid strains, significant tropism of COL1A1+ fibroblasts to the injury site and strong fibronectin accumulation were observed (Figures 20E, 20G). This rapid increase in EC or epicardial-associated fibroblast-like cells (Figures 20E, 20H, 20I) is consistent with in vivo observations during the injury response. Therefore, in conclusion, we have elucidated cell type-specific processes during cardiac injury and concluded that cardioids can mimic important early stages of regeneration and fibrous responses.

[0223] conclusion In conclusion, a high-throughput human cardioid platform with patterned layers and an inherent self-organizing capability into a three-dimensional structure reminiscent of the early human left ventricle was established. Furthermore, this resource is shown to be usable for modeling the underlying mechanisms of development in the three major cardiac systems, including cavity formation and injury response.

[0224] While organoids mimic the self-organization of organoids in vitro, their diversity and complexity still hinder accurate modeling of morphogenesis defects. This diversity challenge was addressed by eliminating exogenous ECM and employing a high-throughput approach to reach optimal conditions from a range of optimized parameters. By incorporating and controlling three major cardiac lineages into the organoid platform, and leveraging its high reproducibility, we can statistically determine when and where gene mutations cause defects. This system could potentially incorporate other important cardiac sublineages, such as the second cardiac area lineage and the conduction system. Using this platform, we demonstrated that in vitro cardiac mesoderm can adequately form cavities via the WNT-BMP-driven mechanism, even in the absence of endothelium and endoderm. Thus, controlling myocardial formation in vitro holds broad potential for exploring developmental mechanisms and heart failure, as well as for developing more mature and complex human myocardial models suitable for drug discovery and regenerative medicine. However, the diversity and complexity of self-organizing organ-small animal systems hinder the quantitative modeling of morphogenesis defects. Cardioids address this challenge by eliminating exogenous extracellular media (ECM) and employing high-throughput techniques to obtain optimal signaling conditions. Furthermore, high reproducibility is achieved by strictly controlling signal-mediated self-organization and sequentially incorporating three major cardiac systems into the cardioid. This approach allows for the analysis of when and where the function of specific factors is required with high statistical power. Moreover, the simplicity of the system, which can include one, two, or three cardiac systems and is unaffected by non-cardiac system interference, allows for the reduction of self-organization and the underlying molecular and cellular biological mechanisms to the bare minimum. Therefore, the complexity of cardioids corresponds to biological questions. This is an important advantage for organoid models, as complex biological systems often contain redundant mechanisms that are difficult to elucidate.

[0225] Cardioids, like all other self-organizing organoid systems, replicate some aspects of development but differ from embryogenesis in other respects. Self-organization encompasses only a portion of the intrinsic developmental mechanisms, which is sufficient to replicate aspects of architecture similar to those found in vivo. As a result, it has been shown that cardioids, using only signal-mediated cardiac mesoderm, are sufficient to form a ventricle-like cavity in vitro. It is proposed that this cavity may resemble the cardiac duct or the cavity of the early left ventricle. In vivo, the foregut endoderm causes the bilateral cardiac mesoderm and endocardial duct to migrate and fuse to form a single cardiac duct, creating the first cavity. However, the mechanism by which bilateral cardiac ducts and ventricles can form without the contraction of the endocardium or foregut endoderm was unknown. This suggests that cardiac mesoderm inherently possesses the ability to form cavities and ventricles in vivo, which is consistent with the self-organization observed in cardioids in vitro and in chicken embryo blastodiscs. The lateral plate mesoderm, one of the subspecies that make up the cardiac mesoderm, has a similar potential to form a cavity called the pericardial coelom. Therefore, cavitation appears to be a more common mesodermal characteristic required in foregut-deficient embryos. Furthermore, the phenotype of the HAND1 KO cavity in cardioids is consistent with the phenotype of left ventricular hypoplasia in HAND1 KO mice and the phenotype of the cavity in human left heart hypoplastic syndrome, demonstrating the potential for modeling cardioids.

[0226] What we demonstrated using the cardioid platform is that WNT and BMP drive ventricular self-organization. While these pathways are known to control cardiomyocyte identification in vivo and in vitro, it was unclear whether they control cardiomyocyte patterning and morphogenesis, and at what stage. The surprising finding that early mesoderm WNT controls later cardiomyocyte self-organization is consistent with the diversification of early cardiomyocyte lineages during mesoderm induction in vivo. Patterning and morphogenesis occur in parallel with identification, but not necessarily in conjunction. Consistent with this idea, cavities can self-organize even without cardiomyocyte identification, and HAND1 KO cardiomyocytes have defects in self-organization but not in CM identification. Conversely, inhibition of WNT signaling at the cardiac mesoderm stage is essential for CM identification but does not affect cardiomyocyte self-organization. Therefore, cardioids are a powerful system for essentially unraveling the control of differentiation and morphogenesis. At the same time, the cardioid is simple enough to determine one of these processes to a satisfactory level, and therefore complements more complex systems.

[0227] It was found that WNT, activin, and VEGF control the self-organization of CM and EC in cardioids. In vivo, the myocardial EC first forms the incarcerate, is later separated from the outer CM duct by a gap filled with ECM (myocardial jelly), and then forms the inner wall of the ventricle. It was unclear how signaling identifies and coordinates these patterns and morphogenesis processes. In cardioids, the patterning and morphogenesis of the CM and EC layers are controlled by the doses of WNT and activin administered at the earliest stages of mesoderm differentiation, and by VEGF, which directs both the identification and patterning of the EC layer in cardiac mesoderm. Interestingly, it was found that the same signaling levels of WNT and activin promoted ventricular formation and EC layer formation, suggesting a coordinated relationship between these processes. When EC and CM are aggregated in microtissue, they do not form separate layers or linings. Formation of EC layers and linings is crucial for activating mechanosensing within the ventricle. The mechanobiology of the ventricle is essential for physiological EC-CM crosstalk and drives subsequent stages of cardiac development, such as cavernous formation, cardioid compression, and interaction with the epicardium. Therefore, the cardioid is a promising system for studying the fundamental mechanisms of CM and EC patterning and crosstalk.

[0228] During development, the (pre)epidermis comes into contact with and envelops the early ventricles, simultaneously differentiating and migrating into the cardioid. Co-culturing the cardioid and epicardium allowed for the observation of epicardium expansion, migration, and differentiation, confirming that these processes mirror those in vivo. While co-culturing of the epicardium and cardioid mesentericin (CM) has been studied using microtissues, its relationship with ventricular-like models had not been investigated. This is important because the crosstalk between derivatives of the epicardium, EC, and CM lineages depends on the mechanobiology of the ventricle. Therefore, it is proposed that self-organization of the three myocardial lineages in ventricular-like cardioids is crucial to reproduce the developmental and regenerative crosstalk that drives cardiac proliferation, maturation, and pathophysiology in a manner similar to that in vivo. The significant difference in the response to freeze damage between bioengineered organoids and self-organized cardioids supports the claim that developmental mechanisms influence subsequent pathophysiology.

Claims

1. A cardiac organoid comprising at least 60% cardiac cells or at least 50% cardiac cells, independently of any additional cells of the vascular system of the cardiac organoid, wherein the cardiac cells surround a lumen, and the cardiac cells are selected from the group consisting of cardiomyocytes, endocardial cells and epicardial cells, and the cardiac organoid comprises cardiomyocytes or endocardial cells directly facing the lumen.

2. The cardiac organoid according to claim 1, comprising at least 30% cardiomyocytes.

3. The cardiac organoid according to claim 2, further comprising at least 2% endocardial cells.

4. The cardiac organoid according to claim 1 or 2, comprising cardiomyocytes and endocardial cells in different tissue layers.

5. A cardiac organoid according to any one of claims 1 to 4, comprising 0% to a maximum of 5% foregut endoderm cells and / or 0% to a maximum of 3% hematopoietic cells.

6. The cardiac organoid according to claim 5, comprising up to 3% foregut endoderm cells.

7. The cardiac organoid according to claim 5, which contains no foregut endoderm cells and / or no hematopoietic cells.

8. A cardiac organoid according to any one of claims 1 to 7, comprising epicardial smooth muscle cells and / or epicardial cardiac fibroblasts, and / or comprising 0% to a maximum of 3% non-epidial smooth muscle cells, and / or comprising 0% to a maximum of 3% non-epidial fibroblasts.

9. The cardiac organoid according to claim 8, which contains no non-epidial smooth muscle cells and / or no non-epidial fibroblasts.

10. A cardiac organoid according to any one of claims 1 to 9, wherein the maximum dimension is 0.3 mm to 15 mm, and / or the size of the lumen at the maximum dimension is at least 30% of the size of the cardiac organoid at the maximum dimension.

11. A method for generating cardiac organoids according to any one of claims 1 to 10, The following steps: a) Inducing mesodermal differentiation from pluripotent stem cells in a low-adhesion three-dimensional culture in the presence of a WNT activator and / or a GSK3-β inhibitor, and further in the presence of a PI3 kinase inhibitor, thereby generating an aggregate of mesodermal cells by having the cells bind to each other in place of the culture vessel to form aggregates of the cells, or Inducing mesoderm differentiation in the presence of a WNT activator and / or a GSK3-β inhibitor, the WNT activator and / or the GSK3-β inhibitor and / or any PI3 kinase inhibitor, within 40 hours of the start of induction, cause at least 90% of the pluripotent stem cells to lose their pluripotency and form aggregates of mesoderm cells in an amount sufficient to differentiate the pluripotent stem cells, and then treating the cells with fibroblast growth factor and / or albumin. b) Differentiate the mesoderm cells from step a) into cardiac cells in a low-adhesion three-dimensional culture, and in the presence of cardiomyocyte differentiation factors, and in the absence of a WNT activator and / or in the presence of a WNT antagonist, for at least three days, the cells will aggregate with each other instead of binding to the culture vessel to form aggregates of the cells, thereby forming cardiac mesoderm and lumen. A method for generating cardiac organoids, including the development of cardiac organoids.

12. The method according to claim 11, wherein step b) is performed for 3 to 7 days.

13. The method according to claim 11 or 12, wherein the WNT activator in step a) is a WNT ligand such as WNT-3a or CHIR99021.

14. If the amount of the WNT activator is sufficient to de-pluripotently differentiate at least 90% of the pluripotent stem cells within 40 hours after the start of induction, then the amount of CHIR99021 as the WNT activator is at least 1 μM; The method according to claim 13, wherein, in the presence of the PI3 kinase inhibitor, CHIR99021 as a WNT activator is present at a concentration of at least 0.5 μM.

15. The method according to claim 14, wherein the amount of the WNT activator is sufficient to deactivate and differentiate at least 90% of the pluripotent stem cells within 40 hours after the start of induction, and the amount of CHIR99021 as the WNT activator is at least 6 μM.

16. The method according to claim 15, wherein the amount of the WNT activator is sufficient to deactivate and differentiate at least 90% of the pluripotent stem cells within 40 hours after the start of induction, and the amount of CHIR99021 as the WNT activator is at least 12 μM.

17. The method according to claim 14, wherein, in the presence of the PI3 kinase inhibitor, CHIR99021 as a WNT activator is present at a concentration of 0.5 μM to 12 μM.

18. The method according to any one of claims 11 to 17, wherein the pluripotent stem cells are induced pluripotent stem cells or cells from a cell line, and / or the pluripotent stem cells of step a) are passaged, and / or the pluripotent stem cells provided in step a) are grown in a medium containing at least 1.5% (w / v) albumin and / or at least 100 ng / ml fibroblast growth factor.

19. The method according to claim 18, wherein the pluripotent stem cells of step a) are subcultured in a culture medium containing albumin and / or fibroblast growth factor.

20. The method according to claim 19, wherein the pluripotent stem cells of step a) are subcultured in a medium further comprising albumin and / or fibroblast growth factor, and BMP and / or insulin.

21. The method according to claim 19 or 20, wherein albumin is BSA.

22. The method according to claim 19, 20, or 21, wherein the fibroblast growth factor is FCF2.

23. The method according to any one of claims 19 to 22, wherein the pluripotent stem cells of step a) are grown in a medium further comprising at least 1.5% (w / v), albumin and / or at least 100 ng / ml of fibroblast growth factor, and BMP and / or insulin.

24. The method according to any one of claims 11 to 23, wherein mesoderm differentiation is induced in a culture medium containing activin A and / or bone morphogenetic protein.

25. The method according to claim 24, wherein mesoderm differentiation is induced in a culture medium comprising activin A and / or osteomorphogenetic protein and further fibroblast growth factor.

26. The method according to claim 24 or 25, wherein mesoderm differentiation is induced in a culture medium containing at least 1 ng / ml of osteomorphonoprotein.

27. The method according to claim 26, wherein the bone morphogenetic protein is BMP4.

28. d) The method according to any one of claims 11 to 27, further comprising differentiating the aggregate having cardiac mesoderm into a tissue having a cardiomyocyte layer with a cardiomyocyte differentiation factor for one day or more.

29. The method according to any one of claims 11 to 28, using a kit comprising i) a WNT activator and / or a GSK3-β inhibitor, ii) a PI3 kinase inhibitor, iii) a low-adhesion cell culture vessel, and iv) albumin.

30. The method according to claim 29, wherein the kit further comprises a WNT inhibitor.

31. A container plate comprising at least 10 compartments, wherein each compartment contains organoids according to any one of claims 1 to 10 at substantially the same developmental stage.

32. The method according to any one of claims 11 to 28 for screening or testing a candidate compound for its effect on cardiac development or function, comprising: treating any of the cells in step a) and / or b) of claim 11 with the candidate compound while generating a cardiac organoid according to any one of claims 1 to 10, and comparing the development or function of the cardiac organoid with the development of a cardiac organoid that was not treated with the candidate compound.

33. Either of the cells in step a) and / or b) of claim 11 has a candidate repression gene or overexpresses the candidate gene, The development of the aforementioned cardiac organoids is compared with the development of cardiac organoids that did not develop due to suppressed or overexpressed genes. A method for observing the effect of the suppressed or overexpressed gene on the cardiac development process, comprising generating a cardiac organoid according to any one of claims 1 to 10, including the above.

Images