Abstract
Mouse somatic cells can be chemically reprogrammed into pluripotent stem cells (CiPSCs) through an intermediate extraembryonic endoderm (XEN)-like state. However, it is elusive how the chemicals orchestrate the cell fate alteration. In this study, we analyze molecular dynamics in chemical reprogramming from fibroblasts to a XEN-like state. We find that Sox17 is initially activated by the chemical cocktails, and XEN cell fate specialization is subsequently mediated by Sox17 activated expression of other XEN master genes, such as Sall4 and Gata4. Furthermore, this stepwise process is differentially regulated. The core reprogramming chemicals CHIR99021, 616452 and Forskolin are all necessary for Sox17 activation, while differently required for Gata4 and Sall4 expression. The addition of chemical boosters in different phases further improves the generation efficiency of XEN-like cells. Taken together, our work demonstrates that chemical reprogramming is regulated in 3 distinct “prime–specify–transit” phases initiated with endogenous Sox17 activation, providing a new framework to understand cell fate determination.
Similar content being viewed by others
Introduction
Somatic cells can be reprogrammed to become pluripotent by nuclear transfer into oocytes, by delivery of transcription factors or by treatment with a cocktail of chemicals1,2,3. These somatic reprogramming techniques hold great promise in regenerative medicine for providing an unlimited source for functional cells. In comparison with the other two strategies, chemical reprogramming is attractive for future applications due to its non-integrative nature, ease to be standardized and temporally controlled, and lower tumorigenicity3,4.
In recent years, the understanding of the cell dynamics and the molecular mechanisms of chemical reprogramming has gone deeper and broader. For instance, an extraembryonic endoderm (XEN)-like state bridges the chemical reprogramming towards chemically reprogrammed into pluripotent stem cells (CiPSCs) from different somatic cell types4,5. Dynamic early-embryonic-like programs are found critical for the transition of XEN-like state into a pluripotent state6. In addition, the chemical reprogramming efficiency has been found greatly improved by additional chemical boosters, such as bromodeoxyuridine, retinoic acid agonists, Dolt1L inhibitors, and glycogen synthase kinase 3 inhibitors, and CiPSC can even be induced with a chemically defined medium4,7,8,9.
Furthermore, chemical reprogramming strategies have been extended to inducing direct cell lineage conversion into functional cell types without an intermediate pluripotent state. For instance, neural progenitors10, functional neurons11,12, cardiomyocytes13,14, skeletal muscles15, brown adipocytes16,17, astrocytes18, endoderm progenitor-like cells19, and photoreceptor-like cells20 are reported to be induced from fibroblasts with chemicals alone. Besides, endoderm progenitor cells are induced from gut epithelium with pure chemicals21, and human fetal astrocytes are converted into functional neurons by chemical combinations22.
Intriguingly, the small molecules essential for XEN induction, CHIR99021 (a GSK3 inhibitor), 616452 (Repsox, an ALK5 inhibitor), and Forskolin (a cAMP agonist) have frequently been used for the direct induction of the many of different cell types noted above. Unlike the master genes used in transgenic reprogramming, which are associated with the target cell type, these chemicals always target signaling pathways that play roles in different cell types and are not associated with any specific cell lineage. Therefore, it is still unclear how the chemical cocktails determine the target cell type, and the molecular dynamics during chemically induced cell fate transition are still elusive23.
Here, to better understand how chemically induced cell fate alteration and determination are orchestrated, we studied the chemical reprogramming process from fibroblasts to XEN-like cells in terms of the time-course and at the single-cell resolution. We revealed that cell fate transition was primed by endogenously expressed Sox17, which mediated further hierarchical activation of master transcription factors in chemical reprogramming. We further investigated the role of small molecules in various stages throughout the process.
Results
Chemically induced Sox17 expression initiates XEN-like cell fate reprogramming
To investigate how the chemical cocktail determines XEN-like cell fate during C6FAE-mediated reprogramming (C, CHIR99021; 6, 616452; F, Forskolin; A, AM580; E, EPZ004777) (Fig. 1a), we analyzed the reprogramming process at 10 time points over a course of 20 days with single-cell RNA-sequencing (Fig. 1b). In comparison with the existing dataset6, our data detected more UMIs and genes, and the expression pattern of XEN and fibroblast master genes in various periods were comparable (Supplementary Fig. 1a–d). Importantly, MEF cells and XEN cells (day 20) merged perfectly with those in the existing dataset (Supplementary Fig. 1e), indicating the fidelity of our single-cell RNA-seq data.
By principal component analysis (PCA) of single-cell RNA-seq data from days 0 to day 20, we found that cells in the earlier stage were quite close to each other and then separated gradually, and ultimately divided into two branches (Fig. 1c). This bifurcated reprogramming trajectory was also confirmed by pseudo-time analysis (Supplementary Fig. 1f). These two branches mainly consisted of cells from day 12 and day 20 and they belonged to different clusters in which cells were grouped using Louvain clustering with a resolution 0.9524,25,26 (Supplementary Fig. 1g). We evaluated cell identities by using the analytical technique based on quadratic programming with 100 genes representing for MEF cell identity and 100 genes representing for XEN cell identity27. The left branch had established the major XEN identity without MEF identity (Fig. 1d and Supplementary Fig. 1h), indicating successful reprogramming into XEN-like cells and was further referred to as “the proceeding branch”. While, the right branch reserved the major profiles of MEF identity, without the establishment of XEN identity, which was further termed as “the trapped branch”. The cells located at the proceeding branch had a remarkable higher expression of the XEN master genes Sox17, Sall4, and Gata4, which were reported to promote the differentiation into XEN cells from mouse embryonic stem cells by forming a self-activation loop28,29,30,31 (Fig. 1e). The trapped branch included cells with a lower expression of Sall4 and Gata4 while still retained the high expression of fibroblast master genes, such as Osr1, Prrx1, and Twist2 (Fig. 1e). Interestingly, very few cells had low scores of MEF identity before XEN-like cells were induced. This indicates no distinct de-differentiated, or other kinds of intermediate, cells during chemical reprogramming from MEFs to XEN-like cells (model 1–4 in Fig. 1a).
We noticed that the major differences between the proceeding and the trapped branches were the differential expression of XEN and fibroblast master genes. The expression of XEN master transcriptional factors (TFs), especially Sox17, were enriched in the proceeding branch (Fig. 1e and Supplementary Fig. 1i). Besides, the order of the activated expression of XEN master TFs was Sox17, Sall4, Gata4, and Foxa2, suggesting that Sox17 was upstream of the other XEN TFs (Supplementary Fig. 1j–l).
In line with the above, we found that Sox17 knockdown impaired the activation of most of the XEN TFs, Gata4, Sall4, and Foxa2, as well as XEN-like colony formation (Fig. 1f). Meanwhile, Sox17 overexpression promoted the upregulation of Sall4, Gata4, and Foxa2 (Fig. 1g). Then, we analyzed the co-expression of XEN master gene expression every day throughout the reprogramming process by immunofluorescence. We found that the expression of Sox17 was detected as early as day 4. Sall4, Gata4, and Foxa2-expressing cells were all subpopulations of Sox17-expressing cells that emerged in day 5-8 (Fig. 1h), indicating Sall4, Gata4, and Foxa2 were only activated in Sox17-positive cells.
These findings suggest that chemical-mediated XEN-like cell reprogramming is mediated by the endogenously activated Sox17 in fibroblasts (Fig. 1i).
XEN-like cell fate specification and transition with the accumulated master TFs downstream of Sox17
To further investigate how the cell fate reached a XEN-like state after Sox17 activation, we focused on the activation of Gata4, Sall4, and Foxa2. We found that Gata4-positive cells were a subset of Sall4-positive cells and no Gata4-positive/Sall4-negative cells appeared before day 6 by analyzing the co-expression of Gata4 and Sall4 using immunostaining. This suggests Gata4 activation is only in Sall4-expressing cells (Fig. 2a, b). Afterward from day 7 to day 12, the number of Gata4-positive colonies greatly increased while the number of Sall4-positive colonies declined (Fig. 2c). Finally, at day 12, Sall4-positive cells turned out to be a subpopulation of Gata4-positive cells (Fig. 2a). This was probably due to the self-repression function of Sall4 expression as reported32 or resulted from another wave of Gata4 activation without Sall4. Staining for Foxa2 revealed a subpopulation of Gata4 expressing cells, leading us to believe that Gata4 might be upstream of Foxa2 in cell fate specification (Fig. 2d).
In a knockdown experiment of Sall4 the expression of Gata4 and Foxa2 was severely disrupted, suggesting that Sall4 is the upstream regulator of Gata4 and Foxa2, which is consistent with the immunostaining data (Fig. 2e). Knockdown of Gata4 decreased the expression of Foxa2 but had no influence on the transcription of Sall4 (Fig. 2f), which further supported that Foxa2, but not Sall4, is a downstream factor of Gata4. In summary, we found the activation of Gata4 and Sall4 was regulated differently, and the mutual regulation between them was dynamic. Figure 2g shows the hierarchical regulatory network of XEN master TFs.
The regulatory network for the cell fate specification and the transition was established after the core network of XEN master TFs was constructed (Fig. 2h). The cell fates of MEF and XEN-like were seen mutual antagonizing from the regulatory network. In the transition process, the up-regulation of XEN master TFs promoted the down-regulation of fibroblast master TFs, and vice versa (Fig. 2i, j and Supplementary Fig. 2). Such a positive feedback loop could account for the fast transition from fibroblasts to XEN-like cell fates in the final stage of reprogramming, which was exhibited in the single-cell analysis (Fig. 1f and Supplementary Fig. 1h).
Taken together, the core XEN transcriptional network, including Sox17, Sall4, Gata4, and Foxa2, was established consecutively and hierarchically, and thus completing the cell fate specification and transition process in the end. These findings supported the “prime, specify and transit” model that previously speculated (model 5 in Fig. 1a).
Chemicals are essential for Sox17 expression while play different roles in the specification and transition process
Since CHIR99021, 616452 and Forskolin are pivotal to chemical reprogramming to XEN-like cells (Supplementary Fig. 3a), we further explored whether they were necessary to the entire process of reprogramming. We found that the chemical compounds worked with a stepwise approach.
Subtracting any one of CHIR99021, 616452 and Forskolin, respectively, from C6FAE from day 0 hampered the expression of XEN master TFs, especially the major TF, Sox17 (Supplementary Fig. 3b). We then withdrew CHIR99021, 616452 and Forskolin after 4-day induction when Sox17 was already activated albeit at a lower level (Fig. 3a). We found that the presence of CHIR99021 and Forskolin was essential for Gata4 activation, while the addition of 616452 was critical for the up-regulation of Sall4 (Fig. 3a). The expression of Foxa2 was also greatly impaired when CHIR99021 or Forskolin was removed (Fig. 3a), which is consistent with our previous finding that Foxa2 activation might be downstream of Gata4 activation.
We further studied the requirement of CHIR99021, 616452, and Forskolin for the protein expression of Sall4 and Gata4, by subtracting CHIR99021, 616452, and Forskolin after day 6, when Sox17-positive cell number was greatly increased. Similar to the transcriptional level, 616452 was essential for the expression of Sall4 protein, and chemical cocktails containing 616452 after 6-day treatment of C6FAE were sufficient to induce the expression of Sall4 protein (Fig. 3b, c). Moreover, we detected the expression of Gata4 in Sall4-positive colonies when CHIR99021 or Forskolin was subtracted from the cocktail after day 6 in the presence of 616452 (Fig. 3b, c). It was consistent with our previous findings that Sall4 activated Gata4 expression. Interestingly, when 616452 was removed from the cocktail after day 6, in the presence of CHIR99021 and Forskolin, Gata4 expression was still detected at a high level, and Sall4 expression was substantially impaired. This indicates CHIR99021 and Forskolin were sufficient to induce Gata4 expression after the activation of Sox17, which is independent of Sall4 expression (Fig. 3d, e). This was also consistent with another wave of Gata4 expression that was found after 6 days of C6FAE treatment (Fig. 2a–c).
In summary, it was the cooperation of CHIR99021, 616452 and Forskolin that activated Sox17; thereafter, CHIR99021/Forskolin and 616452 activated the expression of Gata4 and Sall4, respectively, in the specification stage, which further established the entire core regulatory network of XEN (Fig. 3f). After day 8, CHIR99021, 616452, and Forskolin were not essential for XEN gene expression (Supplementary Fig. 3c), suggesting that the transition phase was a self-organizing process by XEN master genes. These were also in line with the findings that the transduction of Sall4 and Gata4 was able to replace the function of CHIR99021, 616452 and Forskolin in inducing XEN-like colonies4. Overall, CHIR99021, 616452, and Forskolin played different roles in the reprogramming processes before and after Sox17 expression although they were required for both of the two phases.
Endogenously activated BMP signaling is critical for Sox17 activation and XEN induction
We further explored the upstream factors of Sox17 after chemical treatment. Using bulk RNA-sequencing in the very early stage of XEN reprogramming, we found that Bmp2 was one of the factors that were activated before Sox17 expression (Fig. 4a and Supplementary Fig. 4a).
To investigate the roles of Bmp2 in the activation of Sox17 and the subsequent reprogramming into XEN-like cells, we inhibited Bmp signaling with small molecule inhibitors Dorsomorphin and DMH1. We found that both the transcription of Sox17 and the number of Sox17-positive colonies remarkably decreased (Fig. 4b–d). Meanwhile, the overexpression of Bmp2 promoted the activation of Sox17 drastically (Supplementary Fig. 4b–e). Adding recombinant BMP2 or BMP4 in the reprogramming medium also improved the messenger RNA (mRNA) level of Sox17 and the number of Sox17-positive colonies (Fig. 4e–g). Consistently, Dorsomorphin and DMH1 compromised the upregulation of Sox17 expression by BMP2 or BMP4 (Supplementary Fig. 4f–i).
Importantly, the mRNA level of other XEN master genes (Sall4, Gata4, Foxa2) and the efficiency of XEN-like cell induction were hampered by Dorsomorphin and DMH1 (Fig. 4h, i), and were promoted by BMP2 and BMP4 (Fig. 4j, k). Also, we found that BMP4 notably promoted the up-regulation of Sox17 in the iCD1 serum-free medium used in CiPSC induction9 (Supplementary Fig. 4j). However, the effects of BMP4 on the activation of Sox17 relied on the presence of C6F. BMP4 could not replace the role of C6F on the activation of Sox17 (Supplementary Fig. 4k).
These results indicate that the early activation of endogenous Bmp signaling by chemical cocktails promoted the expression of Sox17 and thus facilitated the stepwise induction of XEN-like cells (Fig. 4l and Supplementary Fig. 4l).
The chemical boosters, CH55 and VPA, benefit Sox17 activation and XEN specification differently
The two phases before and after Sox17 expression revealed in our study, raised the possibility that the chemical boosters played different roles in the stepwise process from fibroblast to XEN-like cells. Thus, we compared the gene expression profiles induced with and without the previously reported chemical boosters, CH55 and valproic acid (VPA), in the presence of C6FAE3,5. Interestingly, CH55 promoted the expression of Sox17 notably in the first 4 days, even on the basis of exogenously provided Bmp4 (Supplementary Fig. 5a), while had nearly no function in further activation of other XEN genes from day 4 to 12 (Fig. 5a, c). VPA was found to promote the up-regulation of most XEN genes and XEN identity from day 4 to day 12 (Fig. 5b, c). However, VPA has no beneficial effect on Sox17 expression in the first 4 days, suggesting that VPA improved XEN reprogramming efficiency by supporting the up-regulation of the XEN network after the activation of endogenous Sox17.
We also found that the cocktail mainly induced “smoothed” colonies co-expressing Sox17, Gata4, and Sall4 in the presence of VPA (VC6FAE). Without VPA, it induced many “fuzzy” colonies with robust Sox17 expression and very low expression of Sall4 and Gata4 (Supplementary Fig. 5b–d). The fuzzy colonies had higher mRNA levels of the fibroblast master genes, Osr1, Prrx1 and Twist2 (Supplementary Fig. 5e) and could rarely be induced into XEN-like cells (Supplementary Fig. 5f, g). Importantly, smoothed colonies, but not fuzzy colonies, could be induced into pOct4-GFP-positive CiPSCs (Supplementary Fig. 5h, i). VPA promoted the induction of pOct4-GFP-positive CiPSCs (Supplementary Fig. 5j). These results support that VPA improves the C6FAE-mediated XEN reprogramming by promoting the XEN specification process, which was previously reported to bridge chemical reprogramming from fibroblasts to pluripotency4,5.
We further determined whether using chemical boosters, CH55 and VPA, in a stepwise manner could promote the XEN reprogramming efficiency. We found that treating the cells with CH55 only in the first 4 days was more efficient than treating for the entire process in promoting the expression of Sox17 and Sall4 (Fig. 5d). Also, VPA induced a higher level of Gata4 and Foxa2 mRNA when using in the last 8 days rather than in the entire process (Fig. 5e). Collectively, the sequential use of CH55 and VPA in different steps reached the highest efficiency of XEN-like colony generation (Fig. 5f, g). Taken together, these findings not only suggested the “prime, specify and transit” model in chemical reprogramming but also revealed the roles of the chemicals on the stepwise processes (Fig. 1a, h).
Discussion
A major question in chemical reprogramming is “how does a set of chemicals, which bear no obvious relation to any genes or molecules that are directly associated with a specific cell type, enable the determination of a specific cell fate”. In this study, we made a significant conceptual leap towards an answer to this question. We demonstrated that the chemical reprogramming was a stepwise process by studying the molecular roadmap from fibroblasts to XEN-like cells. First, the chemicals orchestrated a priming state with the activated expression of Sox17, a master gene of XEN, without substantial cell fate alteration or determination. Afterward, the chemicals further guided hierarchical accumulation of endogenous master transcription factors for cell fate specification. Finally, cell fate was transited with the combination of those activated master transcription factors. In brief, chemicals used in reprogramming guided the hierarchical activation of master genes in the cell-fate-associated regulatory network in a stepwise manner.
Therefore, we indicate that the chemicals previously used in the entire process from fibroblasts to XEN-like cells had different functions in different phases and played different roles in activating different genes. The core chemicals CHIR99021, 616452, and Forskolin (C6F) were all essential to stimulate Sox17 expression in the priming phase. Afterward, they supported the activation of other master genes, such as Sall4 and Gata4 for the XEN cell fate specification in the Sox17 expressing cells differently. CHIR99021 and Forskolin facilitated Gata4 expression, while 616452 enabled the expression of Sall4. CH55 and BMP signaling functioned through elevating Sox17 activation, while VPA functioned through activating the other XEN master genes in the Sox17 expressing cells.
Importantly, the “prime–specify–transit” model may be extended to other chemical reprogramming systems according to the gene expression profiling data during the reprogramming processes. For instance, in the chemical reprogramming process from fibroblast to neural stem cells, Sox2 was activated very early by small molecules in the first 4 days and might initiate a priming state to neural stem cell33. After that, neural stem cell core regulators network was built up, which was reminiscent of a specification process. Besides, the molecular dynamics during the chemical reprogramming from fibroblast to photoreceptor-like cells (CiPCs) was probably initiated by the activation of photoreceptor-specifying transcription factors, such as Rorb, Ascl1 and Pias3, reminiscent of a priming phase20. The stepwise manner in the activation of the master transcription factors suggested identifying the key molecular events in a chemical reprogramming process could help to optimize the protocol in a stepwise manner to achieve higher efficiency.
Unexpectedly, although cells in the priming phase had already expressed some of the XEN master TFs, they were still fibroblast-like with a high level of the fibroblasts program, as well as the high expression of fibroblast master genes. This was rather different from an intermediate multipotent cell type that was presumed in previous reports34,35 and from the other possible intermediates that were speculated before this study (Fig. 1a). The “Disc model” for cell fate reprogramming matches these findings since cell fate priming helps the cell to escape the attractor of an initial cell type without cell fate determination, while the hierarchically accumulated expression of endogenous transcription factors provides the “guide rail” to determine a cell fate progressively, without entering into a multipotent attractor36.
Our findings also highlight the importance of activating some or even one master transcription factor of the target cell type in developing a chemical reprogramming system. In the reprogramming process to XEN-like cells, the activation of Sox17 was a molecular event that was not easy to be triggered. It required most chemicals in the cocktail like C, 6, F and CH55, and even took advantage of the endogenously activated expression of BMP2 or other BMP signaling stimuli from serum or KSR. Thereafter, Sox17 expression made the subsequent molecular events possible and easier. Thus, the activation of one or more transcription factors in a cell type may represent the major molecular basis for cell plasticity. The cells initially express one or more transcription factors of another cell type may have superiority in cell fate transition in chemical reprogramming.
Moreover, since the chemicals C, 6, F, and their combinations have been widely used in different chemical reprogramming systems and generate many different cell fates8,12,13,15,16,17,19,20, it is still unclear whether Sox17 activation is a specific outcome of the chemical treatment and whether these chemicals can prime the cells and facilitate cell fate conversion into other lineages simultaneously. These are some of the questions we intend to address in our future study.
Methods
MEF isolation
MEFs were isolated from E13.5 embryos of ICR mouse. After the removal of head, limbs, and viscera, embryos were minced with scissors and dissociated in trypsin-EDTA at 37 °C for 10 min. After centrifugation, MEF cells were collected and cultured in MEF medium, which included: high-glucose Dulbecco’s modified Eagle’s medium (DMEM) supplemented, 10% fetal bovine serum (FBS), 1% GlutaMAX, 1% nonessential amino-acids (NEAAs), and 1% penicillin-streptomycin. Oct4-EGFP mice were obtained from The Jackson Laboratory (004654). This study was performed under in accordance with protocols by Peking University laboratory animal research center.
Generation of XEN-like cells from fibroblasts
Twenty-thousand MEF cells were seeded into a well of 12-well plate. Twenty-four hours later later, the medium was changed to XEN reprogramming medium, which included: KnockOut DMEM supplemented, 10% KnockOut Serum Replacement (KSR), 10% FBS, 1% GlutaMAX, 1% NEAAs, 0.055 mM 2-mercaptoethanol, 1% penicillin–streptomycin (Invitrogen), 50 ng/ml basic fibroblast growth factor (bFGF), and the small-molecule cocktail VC6FAE (0.5 mM valproic acid, 20 μM CHIR99021, 10 μM 616452, 50 μM Forskolin, 0.05 μM AM580, and 5 μM EPZ004777). XEN reprogramming medium was changed every 4 days for 12 to 20 days.
Immunofluorescence
Primary antibodies were those specific to rabbit anti-SALL4 (Abcam, 1:500), goat anti-SOX17 (R&D, 1:500), goat anti-GATA4 (Santa Cruz, 1:300), goat anti-GATA6 (R&D, 1:200), rabbit anti-Nanog (Sigma Aldrich, 1:200). The secondary antibodies used were FITC-conjugated secondary antibodies and TRITC-conjugated secondary antibodies (Jackson ImmunoResearch, 1:200).
Cells were fixed in 4% paraformaldehyde for 15 min at room temperature. Then, removing 4% paraformaldehyde and washing cells with PBS for two times. cells were permeabilized and blocked in PBS containing 0.2% Triton X-100 and 3% donkey serum for 1 h at room temperature. Then the cells were incubated with primary antibodies at 4 °C overnight. After washing three times with PBS, secondary antibodies (Jackson ImmunoResearch) were incubated at 37 °C for 1 h. The nuclei were stained with DAPI (Roche Life Science) for 5 min.
Quantitative reverse transcription PCR (RT-qPCR)
RT-qPCR was performed according to protocols. Briefly, total RNA samples were extracted by using the EasyPure RNA Kit (TransGen Biotech) and were reverse transcribed into complementary DNA (cDNA) using TransScript One-step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech); Real-time PCR was performed on a Quantagene q225 System (KUBO technology) using 2 × T5 Fast qPCR Mix (TSINGKE Biological Technology).
Single-cell RNA-seq
Individual cell at different time points was manually picked after digestion, lysed and subjected to cDNA synthesis37,38. Single-cell cDNA was then amplified and fragmented as published steps37,38. The sequencing library was constructed (New England Biolabs) and sequenced with paired-end 150-bp reads on an Illumina HiSeq X-Ten platform (Novogene). Raw reads were firstly separated by cell barcodes, then trimmed, and aligned to the mm9 mouse transcriptome and de-duplicated by UMIs information as described previously39.
Pseudotime analysis
Monocle (v2.6.4) were adopted to perform the pseudotime analysis. Differentially expressed genes (DEGs) identified from each cell type were used as ordering genes. The whole workflow followed the recommended pipeline with default parameters.
Integration analysis of gene expression between data in this study and in Zhao et al.6
Cells from day 0 to day 20 in this study, and cells belong to MEFs, Stage I day 5, Stage I day 12 and XEN-like cells in Zhao et al.6 were used to perform the integration analysis. To integrate different data sets, CCA algorithm from Seurat was used.
Statistics and reproducibility
All experiments contain at least three independent biological replicates. No randomization or blinding was used. The statistical analysis in this paper uses Student’s t-test. p-value of <0.05 is considered a significant difference.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Data availability
The accession number for the RNA-seq and single-cell RNA-seq data reported in this paper is NCBI GEO: GSE144097. Source data underlying plots shown in figures are provided in Supplementary Data 1. Full blots are shown in Supplementary Information. All other data, if any, are available upon reasonable request.
References
Gurdon, J. B. The developmental capacity of nuclei taken from intestinal epithelium cells of feeding tadpoles. J. Embryol. Exp. Morphol. 10, 622–640 (1962).
Takahashi, K. & Yamanaka, S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676 (2006).
Hou, P. et al. Pluripotent stem cells induced from mouse somatic cells by small-molecule compounds. Science 341, 651–654 (2013).
Zhao, Y. et al. A XEN-like State Bridges Somatic Cells to Pluripotency during Chemical Reprogramming. Cell 163, 1678–1691 (2015).
Ye, J. et al. Pluripotent stem cells induced from mouse neural stem cells and small intestinal epithelial cells by small molecule compounds. Cell Res. 26, 34–45 (2016).
Zhao, T. et al. Single-cell RNA-seq reveals dynamic early embryonic-like programs during chemical reprogramming. Cell Stem Cell 23, 31–45 e37 (2018).
Long, Y., Wang, M., Gu, H. & Xie, X. Bromodeoxyuridine promotes full-chemical induction of mouse pluripotent stem cells. Cell Res. 25, 1171–1174 (2015).
Li, X. et al. Small-molecule-driven direct reprogramming of mouse fibroblasts into functional neurons. Cell Stem Cell 17, 195–203 (2015).
Cao, S. et al. Chromatin accessibility dynamics during chemical induction of pluripotency. Cell Stem Cell 22, 529–542 e525 (2018).
Cheng, L. et al. Generation of neural progenitor cells by chemical cocktails and hypoxia. Cell Res. 24, 665–679 (2014).
Li, X. et al. Direct reprogramming of fibroblasts via a chemically induced XEN-like state. Cell Stem Cell 21, 264–273 e267 (2017).
Hu, W. et al. Direct conversion of normal and Alzheimer’s disease human fibroblasts into neuronal cells by small molecules. Cell Stem Cell 17, 204–212 (2015).
Fu, Y. et al. Direct reprogramming of mouse fibroblasts into cardiomyocytes with chemical cocktails. Cell Res. 25, 1013–1024 (2015).
Cao, N. et al. Conversion of human fibroblasts into functional cardiomyocytes by small molecules. Science 352, 1216–1220 (2016).
Bansal, V. et al. Chemical induced conversion of mouse fibroblasts and human adipose-derived stem cells into skeletal muscle-like cells. Biomaterials 193, 30–46 (2019).
Nie, B. et al. Brown adipogenic reprogramming induced by a small molecule. Cell Rep. 18, 624–635 (2017).
Takeda, Y., Harada, Y., Yoshikawa, T. & Dai, P. Direct conversion of human fibroblasts to brown adipocytes by small chemical compounds. Sci. Rep. 7, 4304 (2017).
Tian, E. et al. Small-molecule-based lineage reprogramming creates functional astrocytes. Cell Rep. 16, 781–792 (2016).
Cao, S. et al. Chemical reprogramming of mouse embryonic and adult fibroblast into endoderm lineage. J. Biol. Chem. 292, 19122–19132 (2017).
Mahato, B. et al. Pharmacologic fibroblast reprogramming into photoreceptors restores vision. Nature 581, 83–88 (2020).
Wang, Y. et al. Conversion of human gastric epithelial cells to multipotent endodermal progenitors using defined small molecules. Cell Stem Cell 19, 449–461 (2016).
Yin, J. C. et al. Chemical conversion of human fetal astrocytes into neurons through modulation of multiple signaling pathways. Stem Cell Rep. 12, 488–501 (2019).
Zhao, Y. Chemically induced cell fate reprogramming and the acquisition of plasticity in somatic cells. Curr. Opin. Chem. Biol. 51, 146–153 (2019).
Blondel, V. D., Guillaume, J.-L., Lambiotte, R. & Lefebvre, E. Fast unfolding of communities in large networks. J. Stat. Mech.: Theory Exp. 2008, P10008 (2008).
Levine, J. H. et al. Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis. Cell 162, 184–197 (2015).
Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: large-scale single-cell gene expression data analysis. Genome Biol. 19, 15 (2018).
Treutlein, B. et al. Dissecting direct reprogramming from fibroblast to neuron using single-cell RNA-seq. Nature 534, 391–395 (2016).
Lim, C. Y. et al. Sall4 regulates distinct transcription circuitries in different blastocyst-derived stem cell lineages. Cell Stem Cell 3, 543–554 (2008).
Niakan, K. K. et al. Sox17 promotes differentiation in mouse embryonic stem cells by directly regulating extraembryonic gene expression and indirectly antagonizing self-renewal. Genes Dev. 24, 312–326 (2016).
Shimosato, D. et al. Extra-embryonic endoderm cells derived from ES cells induced by GATA factors acquire the character of XEN cells. BMC Dev. Biol. 7, 80 (2007).
Hwang, J. T. et al. GATA6 and FOXA2 regulate Wnt6 expression during extraembryonic endoderm formation. Stem Cells Dev. 21, 3220–3232 (2012).
Yang, J., Gao, C., Chai, L. & Ma, Y. A novel SALL4/OCT4 transcriptional feedback network for pluripotency of embryonic stem cells. PLoS ONE 5, e10766 (2010).
Zhang, M. et al. Pharmacological reprogramming of fibroblasts into neural stem cells by signaling-directed transcriptional activation. Cell Stem Cell 18, 653–667 (2016).
Han, X. et al. A molecular roadmap for induced multi-lineage trans-differentiation of fibroblasts by chemical combinations. Cell Res. 27, 386–401 (2017).
Xie, X., Fu, Y. & Liu, J. Chemical reprogramming and transdifferentiation. Curr. Opin. Genet. Dev. 46, 104–113 (2017).
Maamar, H. et al. Noise in gene expression determines cell fate in Bacillus subtilis. Science 317, 526–529 (2007).
Li, L. et al. Single-cell RNA-seq analysis maps development of human germline cells and gonadal niche interactions. Cell Stem Cell 20, 858–873 e854 (2017).
Gu, C., Liu, S., Wu, Q., Zhang, L. & Guo, F. Integrative single-cell analysis of transcriptome, DNA methylome and chromatin accessibility in mouse oocytes. Cell Res. 29, 110–123 (2019).
Trapnell, C. et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat. Biotechnol. 28, 511–515 (2010).
Acknowledgements
We thank Chunyan Yang, Yang Liu, and Siyuan Zhang for technical assistance, and Jiayu Chen for providing pOct4-GFP mice. We thank Iain C. Bruce for editing the manuscript. This work was supported by the National Key Research and Development Program of China (2018YFA0800504), the National Natural Science Foundation of China (31771645, 31922020, 31821091 and 31771590), the Science and Technology Department of Sichuan Province (2018JZ0025).
Author information
Authors and Affiliations
Contributions
Z.Y., J.L., Q.W., C.Y., L.M., J.Y., and K.B. performed experiments. X.X., C.G., and A.N. conducted the bioinformatics analyses. Y.Z., C.T., and F.G. supervised this project. Y.Z., Z.Y., and X.X. wrote the paper.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yang, Z., Xu, X., Gu, C. et al. Chemicals orchestrate reprogramming with hierarchical activation of master transcription factors primed by endogenous Sox17 activation. Commun Biol 3, 629 (2020). https://doi.org/10.1038/s42003-020-01346-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s42003-020-01346-w
This article is cited by
-
Developmental progression continues during embryonic diapause in the roe deer
Communications Biology (2024)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.