c-JUN is a barrier in hESC to cardiomyocyte transition

The study reveals that knocking out c-JUN improves cardiomyocyte generation in humans by increasing chromatin accessibility and H3K4me3 modification on key genes regulating heart development. These findings shed new light on c-JUN's unexpected role in human cardiogenesis.

spreadsheets for the main figures of the manuscript. If you would like to add source data, we would welcome one PDF/Excel-file per figure for this information. These files will be linked online as supplementary "Source Data" files. ***IMPORTANT: It is Life Science Alliance policy that if requested, original data images must be made available. Failure to provide original images upon request will result in unavoidable delays in publication. Please ensure that you have access to all original microscopy and blot data images before submitting your revision.***
-The author mentioned many times in the manuscript that c-JUN "binds" to RBBP5/SETD1B and thus represses their expression. I agree RBBP5 and SETD1B are up-regulated upon c-JUN depletion. Then, experimental result such as ChIP, cut&tag (though there is cut & tag method in the M&M section and a heatmap only in Fig. 4D), or EMSA to show c-JUN really binds to or localizes at RBBP5/SETD1B loci.
-As the author claimed c-JUN is a barrier of cardiogenesis, I am wondering what would happen when overexpressing c-JUN during cardiomyocyte differentiation.
-Did the authors quantify the numbers of TNNT2 positive cells in WT and c-JUN scRNA-seq? I am wondering whether the authors also see more TNNT2 positive "cardiac cells" in KO compared to WT from scRNA-seq data.
-I think the authors should further discuss the contradictory phenomenon of c-JUN deletion observed between human and mouse early development.
-The whole manuscript suffers from grammar issues, mistakes, and typos. It is highly recommended to have the text proofread by someone more skilled with scientific writing.
Reviewer #2 (Comments to the Authors (Required)): The manuscript from Zhong, Zhang, and Li et al. present work describing a role for c-Jun in cardiomyocyte differentiation repression. The c-Jun knockout cells (or c-Jun inhibitor treated) produce more cardiomyocytes than the controls. The authors go on to show that c-Jun mutant cell lines have higher H3K4me3 signals than control lines at cardiomyocyte related genes. Using single cell RNA-seq, the authors describe that c-Jun mutant lines have more cardiomyocyte-related gene expression. The work presented here is describes a potential role for c-Jun in inhibiting some cells from undergoing cardiomyocyte differentiation but seems preliminary in its proposed mechanism.
Below are some comments and questions related to the work presented in no particular order: 1. In the section, "Chromatin dynamics during cardiogenesis," (lines 161-177), it is unclear from the text that the authors are describing chromatin accessibility near the example genes and not something else like motif analysis. 2. The ATAC-seq and RNA-seq analysis in Figure 3 is performed on bulk cells undergoing differentiation. Based on the results from Figure 2D and Figure 6D, it seems reasonable that the WT cell differentiation contains a mixture of cell fates and states. On the other hand, Figure 2D suggests that the c-Jun mutant cell lines represent a more homogenous cardiomyocyte population (or at least, 92-96.5% of cells are TNNT2+). The authors statement about c-Jun mutants having more/less accessibility at certain loci or increased/decreased gene expression of certain regulators seems more to reflect the nature of the pool of cells (more cardiomyocytes, less other cell fates or states) rather than the behavior or action of c-Jun in any given cell within the pool of cells. 3. A similar issue as my comment above (#2), the H3K4me3 ChIP-seq analysis is performed on bulk cells undergoing differentiation. While it is certainly intriguing that global levels of H3K4me3 were up in the cJun KO lines compared to controls, it seems premature to conclude that H3K4me3 levels were increased at specific loci since the WT differentiation is a mixture of cell types at each time point compared to a seemingly more homogenous pool in the cJun mutant differentiations. 4. In Figures 2G and 5C, the authors state that the cardiomyocytes have enhanced sarcomere structures. How was this determined or quantified? Likewise, in Movies EV3 and S4, how was beating strength measured or quantified? 5. From the single cell analysis, what were the distribution of the 13 clusters between WT and cJun mutant cells? It was reported in Figure 2 that almost all cells in the cJun mutant population at Day 15 were TNNT2+ cardiomyocytes, but it seems like in 6E many Day 15 mutant cells are epicardial, perhaps even more epicardial than WT differentiation ( Figure 6D). 6. Is there any explanation for why the KDM5 inhibitor would have the specific effect of activating cardiomyocyte genes and not other cell fates? 7. Lines 224 and 306, the authors state that they utilized ChIP-seq to demonstrate c-Jun binding, but the methods suggest they used a different method, namely, CUT&Tag. 8. Why were clones #2 and #10 chosen over any of the other clones? No details are provided in the text. 9. The manuscript and figures contained typos and grammatical errors that I would recommend fixing. Below are a few that jumped out to me during the review process: a. Figure 1A, "Cardiac Mesoderm" b. Figure 4C and 4G, it should say ATAC-seq along the lefthand side c. Line 125, it should read, "we analyzed the time course RNA-seq data" d. Line 138-139, it should read, "c-Jun was expressed throughout hESC" e. Line 140, it should read, "and used CRISPR/Cas9 to delete" f. Line 168, "MESP1" g. Line 180, "chromatin accessibility is driven" h. Line 189, "we compared the" Reviewer #3 (Comments to the Authors (Required)): c-JUN has been previously identified as a critical TF for mouse embryonic development. In this manuscript, the authors discovered that c-JUN plays an important role in cardiomyocyte development using in vitro differentiation of human embryonic stem cells (hESCs). Loos of c-JUN leads to increased expression of RBBP5 and SED1B, thereby improving chromatin accessibility and deposition of H3K4me3 on regulatory elements associated with cardiomyocyte development. This manuscript is somewhat innovative, as c-JUN is known to be essential for normal cardiac development in mice but acts as a barrier in hESCs to cardiomyocyte transition.
Major points: 1. Loss of c-JUN leads to early mouse embryonic death possibly due to failure to a normal cardiac system. develop. However, in this manuscript c-JUN is a barrier in hESCs to cardiomyocyte transition. There should be more explanation about the conflict function of c-JUN. 2. Are there any connections between the TFs predicted by SCENIN and H3K4me3? It would be valuable to supplement the discussion regarding the results shown in figure 6H.
Minor points： 1. The titles of video in the attachment do not align with the descriptions provided in the manuscript. 2. There are issues with the titles of Result1, Result2 and Reslut6, as they are not complete sentences. Furthermore, it is necessary to further standardize the language used in other sections of the manuscript 3. At line 158, the sentence "rather than a facilitator" lacks supporting results or references in the text. 4. How can we determine that "the differentiation trajectory of hESCs to cardiomyocytes transition was similar between WT and c-JUN KO conditions" from figure 3H. 5. The "h" in figure4 should be capitalized as "H". Additionally, are there RNA-seq data available for RBBP5 and SED1B as supplementary material? 6. There is some logical issue in lines 217-220. While chromatin accessibility ca be regulated by histone methylation, the WB detection includes histone acetylation. This discrepancy should be addressed. They further investigated how c-JUN inhibits the cardiomyocyte differentiation and showed that loss of c-JUN increases RBBP5 and SETD1B, two methyltransferases, and thus increases the enrichment of H3K4me3 at cardiac gene loci. The manuscript is of interest, but descriptive to some extent. A few key issues need to be fixed before the manuscript could be further considered for publication.
Response: We thank the reviewer for their comprehensive review of our paper, and the useful suggestions that we have utilized to improve our work.
1. The author mentioned many times in the manuscript that c-JUN "binds" to RBBP5/SETD1B and thus represses their expression. I agree RBBP5 and SETD1B are up-regulated upon c-JUN depletion. Then, experimental result such as ChIP, cut&tag (though there is cut & tag method in the M&M section and a heatmap only in Fig. 4D), or EMSA to show c-JUN really binds to or localizes at RBBP5/SETD1B loci.
Response: We apologize for not making this clearer. c-JUN is recruited to the RBBP5 and SETD1B transcription start sites as shown in the CUT&Tag data in Figure 4C (the last row). We emphasize this data in the revision, and have revised the text to make the evidence for the claim explicit. When c-JUN was knocked out, TNNT2 was highly activated in D7 and D15 (A), also the proportion of TNNT2+ cells increased at D7 (~80%), followed by a small decrease at D15 (~70%). But in WT cells, TNNT2+ cells were ~60% at D7, and increased to ~80% on D15 (B). Potentially, the discrepancy between the single cell RNA-seq data and FACS data may be caused by slower RNA transcription and reduced protein degradation at the cardiomyocyte stage. Those results suggest that the knock out of c-JUN promotes TNNT2+ cell generation. In the revised version, we have added this result in Figure S3A and B to make the conclusion more rigorous.  Figure 3 is performed on bulk cells undergoing differentiation. Based on the results from Figure 2D and Figure 6D, it seems reasonable that the WT cell differentiation contains a mixture of cell fates and states. On the other hand, Figure 2D suggests that the c-Jun mutant cell lines represent a more homogenous cardiomyocyte population (or at least, 92-96.5% of cells are TNNT2+). The authors statement about c-Jun mutants having more/less accessibility at certain loci or increased/decreased gene expression of certain regulators seems more to reflect the nature of the pool of cells (more cardiomyocytes, less other cell fates or states) rather than the behavior or action of c-Jun in any given cell within the pool of cells.

The ATAC-seq and RNA-seq analysis in
Response: We agree that this analysis can be difficult to interpret, However, as chromatin opening often precedes gene expression to reshape cell fate 2,3 . We think it is reasonable to identify the sequence of chromatin changes that occurs in the WT and KO cells. Here, we found that KO of c-JUN resulted in increased accessibility at the chromatin of key TFs. This consequently regulates cardiomyocyte generation as early as in D3 (Box 3, Fig. 3B), which precedes the expression of TNNT2 (Box 2). This indicates that the loss of c-JUN primes chromatin in advance of changes in cell composition. and also explains why knock-out c-JUN could increase the population of TNNT2+ cells to 92%. 3. A similar issue as my comment above (#2), the H3K4me3 ChIP-seq analysis is performed on bulk cells undergoing differentiation. While it is certainly intriguing that global levels of H3K4me3 were up in the cJun KO lines compared to controls, it seems premature to conclude that H3K4me3 levels were increased at specific loci since the WT differentiation is a mixture of cell types at each time point compared to a seemingly more homogenous pool in the cJun mutant differentiations.
Response: We agree that the signal of H3K4me3 in bulk ChIP-seq data was averaged among the mixture cell types at each day. Although this is true, it led to the use of the KDMi, which indeed supports the role of H3K4me3 in CM differentiation. Figures 2G and 5C, the authors state that the cardiomyocytes have enhanced sarcomere structures. How was this determined or quantified? Likewise, in Movies EV3 and S4, how was beating strength measured or quantified?

In
Response: We counted the number of sarcomere structures as white boxes marked in each image related to Fig 2G and Fig 5E, and compared WT/control with KO/KDMi treatment (Box 4). From the bar plot, the number of sarcomere structures in both knock out c-JUN or treated by KDMi was more than WT or control group. Those results also add in Fig 2H and Fig 5F. Beating strength was not measured, it was only a qualitative observation. Hence, we have changed the sentence to read: "Interestingly, we observed that spontaneous contractions of c-JUN KO cardiomyocytes Response: We speculate that the KDM5i is working to 'boost' the already established differentiation direction that is determined by the cell culture medium. As the KDMi treatment could highly activate the marker genes of mesoderm and cardiac mesoderm cells such as TBXT/DKK1/EOMES, MEST/HAND2/KDR (Fig.1C and Fig. 5G-H) Response: Picking and generating clones is a stressful process for the cells. In this case, clones #2 and #10 recovered first, and showed little evidence of spontaneous differentiation. Hence, we chose these two lines.
9. The manuscript and figures contained typos and grammatical errors that I would recommend fixing.
Below are a few that jumped out to me during the review process: a. Figure 1A, "Cardiac Mesoderm" b. Figure 4C and 4G, it should say ATAC-seq along the lefthand side c. Line 125, it should read, "we analyzed the time course RNA-seq data" d. Line 138-139, it should read, "c-Jun was expressed throughout hESC" e. Line 140, it should read, "and used CRISPR/Cas9 to delete" f. Line 168, "MESP1" g. Line 180, "chromatin accessibility is driven" h. Line 189, "we compared the" Response: We thank the reviewer for pointing out these errors, which we have corrected. We have also carefully proof read the manuscript and have worked to improve the clarity and flow of the writing.
Reviewer #3: c-JUN has been previously identified as a critical TF for mouse embryonic development.
In this manuscript, the authors discovered that c-JUN plays an important role in cardiomyocyte development using in vitro differentiation of human embryonic stem cells (hESCs). Loos of c-JUN leads to increased expression of RBBP5 and SED1B, thereby improving chromatin accessibility and deposition of H3K4me3 on regulatory elements associated with cardiomyocyte development. This manuscript is somewhat innovative, as c-JUN is known to be essential for normal cardiac development in mice but acts as a barrier in hESCs to cardiomyocyte transition.
Response: We thank the reviewer for their positive comments on our manuscript, and the points below which we have used to strengthen our work.
Major points: 1. Loss of c-JUN leads to early mouse embryonic death possibly due to failure to a normal cardiac system. develop. However, in this manuscript c-JUN is a barrier in hESCs to cardiomyocyte transition.
There should be more explanation about the conflict function of c-JUN.
Response: Reviewer #1 (Point 4) also raised this point. Previous study shows that c-Jun defect leads to mouse embryonic lethal around E13.5 due to cardiac problems 1 . Here, our work shows c-JUN is a barrier to cardiomyocyte differentiation. However, it is important to highlight critical differences between the in vivo knock-out of c-Jun and the in vitro differentiation system employed here. Potentially, an accelerated generation of cardiomyocytes may be deleterious for proper heart generation due to reduced numbers of support cells. Indeed, our single-cell RNA-seq data indicates that the cardiomyocyte population expands at the expense of epicardial cells. Potentially, in the even more complex in vivo development of the heart, other cell types may also be defective, leading to a failure to establish the correct balance of cells required for a functioning heart organ. We have added the discussion about the contradictory phenomenon of c-