Fatty acid oxidation is required for embryonic stem cell survival during metabolic stress
2021; Springer Nature; Volume: 22; Issue: 6 Linguagem: Inglês
10.15252/embr.202052122
ISSN1469-3178
AutoresHualong Yan, Navdeep Malik, Young‐Im Kim, Yunlong He, Mangmang Li, Wendy Dubois, Huaitian Liu, Tyler J. Peat, Joe Nguyen, Yu‐Chou Tseng, Gamze Ayaz, Waseem AlZamzami, King C. Chan, Þorkell Andrésson, Lino Tessarollo, Beverly A. Mock, Maxwell P. Lee, Jing Huang,
Tópico(s)Autophagy in Disease and Therapy
ResumoArticle5 May 2021free access Source DataTransparent process Fatty acid oxidation is required for embryonic stem cell survival during metabolic stress Hualong Yan Hualong Yan Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Navdeep Malik Navdeep Malik Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Young-Im Kim Young-Im Kim Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Yunlong He Yunlong He Sequencing Facility, Cancer Research Technology Program, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA Search for more papers by this author Mangmang Li Mangmang Li Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Department of Cell Biology, School of Basic Medical Sciences, Southern Medical University, Guangzhou, China Search for more papers by this author Wendy Dubois Wendy Dubois Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Huaitian Liu Huaitian Liu Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Tyler J Peat Tyler J Peat Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Joe T Nguyen Joe T Nguyen Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Yu-Chou Tseng Yu-Chou Tseng Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Gamze Ayaz Gamze Ayaz Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Waseem Alzamzami Waseem Alzamzami Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author King Chan King Chan Cancer Research Technology Program, Protein Characterization Laboratory, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA Search for more papers by this author Thorkell Andresson Thorkell Andresson Cancer Research Technology Program, Protein Characterization Laboratory, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA Search for more papers by this author Lino Tessarollo Lino Tessarollo Mouse Cancer Genetics Program, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Frederick, MD, USA Search for more papers by this author Beverly A Mock Beverly A Mock Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Maxwell P Lee Maxwell P Lee Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Jing Huang Corresponding Author Jing Huang [email protected] orcid.org/0000-0002-7163-5156 Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Hualong Yan Hualong Yan Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Navdeep Malik Navdeep Malik Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Young-Im Kim Young-Im Kim Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Yunlong He Yunlong He Sequencing Facility, Cancer Research Technology Program, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA Search for more papers by this author Mangmang Li Mangmang Li Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Department of Cell Biology, School of Basic Medical Sciences, Southern Medical University, Guangzhou, China Search for more papers by this author Wendy Dubois Wendy Dubois Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Huaitian Liu Huaitian Liu Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Tyler J Peat Tyler J Peat Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Joe T Nguyen Joe T Nguyen Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Yu-Chou Tseng Yu-Chou Tseng Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Gamze Ayaz Gamze Ayaz Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Waseem Alzamzami Waseem Alzamzami Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author King Chan King Chan Cancer Research Technology Program, Protein Characterization Laboratory, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA Search for more papers by this author Thorkell Andresson Thorkell Andresson Cancer Research Technology Program, Protein Characterization Laboratory, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA Search for more papers by this author Lino Tessarollo Lino Tessarollo Mouse Cancer Genetics Program, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Frederick, MD, USA Search for more papers by this author Beverly A Mock Beverly A Mock Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Maxwell P Lee Maxwell P Lee Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Jing Huang Corresponding Author Jing Huang [email protected] orcid.org/0000-0002-7163-5156 Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Search for more papers by this author Author Information Hualong Yan1, Navdeep Malik1, Young-Im Kim1, Yunlong He2, Mangmang Li1,3, Wendy Dubois1, Huaitian Liu1, Tyler J Peat1, Joe T Nguyen1, Yu-Chou Tseng1, Gamze Ayaz1, Waseem Alzamzami1, King Chan4, Thorkell Andresson4, Lino Tessarollo5, Beverly A Mock1, Maxwell P Lee1 and Jing Huang *,1 1Laboratory of Cancer Biology and Genetics, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA 2Sequencing Facility, Cancer Research Technology Program, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA 3Department of Cell Biology, School of Basic Medical Sciences, Southern Medical University, Guangzhou, China 4Cancer Research Technology Program, Protein Characterization Laboratory, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, USA 5Mouse Cancer Genetics Program, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Frederick, MD, USA *Corresponding author. Tel: +1 240 760 6796; E-mail: [email protected] EMBO Reports (2021)22:e52122https://doi.org/10.15252/embr.202052122 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Abstract Metabolic regulation is critical for the maintenance of pluripotency and the survival of embryonic stem cells (ESCs). The transcription factor Tfcp2l1 has emerged as a key factor for the naïve pluripotency of ESCs. Here, we report an unexpected role of Tfcp2l1 in metabolic regulation in ESCs—promoting the survival of ESCs through regulating fatty acid oxidation (FAO) under metabolic stress. Tfcp2l1 directly activates many metabolic genes in ESCs. Deletion of Tfcp2l1 leads to an FAO defect associated with upregulation of glucose uptake, the TCA cycle, and glutamine catabolism. Mechanistically, Tfcp2l1 activates FAO by inducing Cpt1a, a rate-limiting enzyme transporting free fatty acids into the mitochondria. ESCs with defective FAO are sensitive to cell death induced by glycolysis inhibition and glutamine deprivation. Moreover, the Tfcp2l1-Cpt1a-FAO axis promotes the survival of quiescent ESCs and diapause-like blastocysts induced by mTOR inhibition. Thus, our results reveal how ESCs orchestrate pluripotent and metabolic programs to ensure their survival in response to metabolic stress. SYNOPSIS Tfcp2l1-induced fatty acid oxidation is vital for the survival of embryonic stem cells under metabolic stress. Diapause-like blastocysts use a similar mechanism to survive in a "dormant" state. The transcription factor Tfcp2l1 is dispensable for self-renewal and pluripotency of ESCs. Tfcp2l1 induces fatty acid oxidation (FAO) through the free fatty acid transporter Cpt1a. Tfcp2l1-induced FAO is essential for the survival of embryonic stem cells under metabolic stresses. Introduction Epiblasts in preimplantation blastocysts exist in a postulated naïve pluripotent state (also called ground-state) and give rise to all cell types in embryos (Nichols & Smith, 2009). Due to the transient nature of this naïve pluripotent state, it remains technically challenging to study how epiblasts maintain pluripotency and respond to environmental cues. Mouse embryonic stem cells (ESCs) are a widely used model to study the biology of epiblasts, and they are immortalized stem cells derived from the inner cell mass (ICM) of preimplantation blastocysts (Evans & Kaufman, 1981; Martin, 1981). The conventional culturing medium of ESCs contains fetal bovine serum (FBS) and leukemia inhibitory factor (LIF). ESCs cultured in FBS and LIF (FBS-ESCs) are heterogeneous in morphology and expression of several transcription factors, such as Nanog and Rex1 (Hayashi et al, 2008). FBS-ESCs exist in metastable states containing both pluripotent stem cells, which are similar to naïve (or ground state) epiblasts in preimplantation blastocysts, and cells expressing various lineage markers (Graf & Stadtfeld, 2008; Hayashi et al, 2008). This heterogeneity of FBS-ESCs can complicate experimental interpretations. Recently, a serum-free, chemically defined medium formula containing an ERK inhibitor (PD0325901) and a GSK-3b inhibitor (CHIR99021), termed 2i medium, has been developed (Ying et al, 2008). In contrast to FBS-ESCs, ESCs grown in 2i (2i-ESCs) are more homogenous and less susceptible to spontaneous differentiation (Ying et al, 2008). Transcriptional and epigenomic analyses have demonstrated that 2i-ESCs mimic the naïve pluripotent epiblasts within the ICM of preimplantation blastocysts and therefore are a better model than FBS-ESCs to investigate the cellular and molecular mechanisms underlying naïve pluripotency and stress responses (Nichols & Smith, 2009). Through studying the transcriptional and epigenetic programs of 2i-ESCs (Marks et al, 2012), several transcription factors, such as Nanog (Silva et al, 2009), Klf4 (Guo et al, 2009), Klf2 (Qiu et al, 2015), Nr5a (Guo & Smith, 2010), and Tfcp2l1 (Martello et al, 2013; Ye et al, 2013; Qiu et al, 2015), have been reported to be critical for establishing and maintaining the naïve pluripotency. Tfcp2l1 (transcription factor cellular promoter 2-like 1) is a member of the Grainyhead-like (GRHL) transcription factor family and is highly expressed in ESCs as well as several ductal somatic tissues (Yamaguchi et al, 2006). Its functions in ESCs are context-dependent: Tfcp2l1 is critical for self-renewal in FBS-ESCs but it is dispensable in 2i-ESCs (Martello et al, 2013; Ye et al, 2013). Tfcp2l1 cooperates with another transcription factor, Klf2, to maintain naïve pluripotency of 2i-ESCs (Qiu et al, 2015) and co-inhibition of Tfcp2l1 and Klf2 leads to differentiation of 2i-ESCs. Mouse embryos with homozygous deletion of Tfcp2l1 can survive to birth (Yamaguchi et al, 2006), indicating that Tfcp2l1 alone is not absolutely necessary for naïve pluripotency of ICM under normal physiological conditions. These previous studies together suggest that Tfcp2l1 either has a yet to be discovered role or has redundant functions with other factors in 2i-ESCs and blastocysts during embryonic development. The metabolic regulation of ESCs has recently gained increasing attention (Folmes et al, 2012; Sperber et al, 2015; Shyh-Chang & Ng, 2017; Zhang et al, 2018; Intlekofer & Finley, 2019). It has been reported that metabolic genes are upregulated in 2i-ESCs compared with FBS-ESCs (Marks et al, 2012), but little is known regarding the resultant biological consequence of this observation. Recently, FBS-ESCs, but not 2i-ESCs, were shown to be dependent on glutamine for survival, a differential dependence that is mediated by alpha-ketoglutarate (α-KG) (Carey et al, 2015; Vardhana et al, 2019). Although these results demonstrate that 2i-ESCs have a unique metabolic program, the factors in metabolic regulation and the mechanisms underlying the crosstalk between metabolic regulation and transcriptional regulation in 2i-ESCs remain largely unclear. Metabolic regulation of pluripotency in development is evident in diapause, a reversible state in which mammalian preimplantation blastocysts are arrested under unfavorable conditions, such as metabolic stress (Fenelon et al, 2014). Diapausing blastocysts are metabolically quiescent with low transcription and translation activity, high autophagic flux, and low levels of histone H4 lysine 16 acetylation (H4K16ac). When the environment becomes favorable, "dormant" blastocysts re-enter the developmental process. Evolutionarily, embryonic diapause can be regarded as a survival mechanism for mammalian embryos to cope with adverse environmental conditions, including metabolic stress (Murphy, 2012; Renfree & Fenelon, 2017). The diapause phenomenon can be largely recapitulated in cultured ESCs. In response to mTOR inhibition or Myc deletion, 2i-ESCs and FBS-ESCs enter a diapause-like state with features similar to natural diapause (Bulut-Karslioglu et al, 2016; Scognamiglio et al, 2016). Pharmacological inhibition of mTOR by INK-128 also prolongs survival of ex vivo blastocysts (Bulut-Karslioglu et al, 2016), and the LKB1-AMPK-mTOR axis is critical for the induction of a diapause-like state in cultured ESCs (Hussein et al, 2020). It is unclear whether other factors and signaling pathways are involved in diapause. In this study, we set out to identify key transcription factors critical for the naïve pluripotency of 2i-ESCs and found that Tfcp2l1 is enriched in the 2i-ESCs. To our surprise, Tfcp2l1 is dispensable for self-renewal and pluripotency of 2i-ESCs. Genomics analyses revealed that Tfcp2l1 activates metabolic genes and represses transcription/chromatin-related genes. Tfcp2l1 deletion in 2i-ESCs leads to a major metabolic reprogramming, including decreased fatty acid oxidation (FAO), increased glycolysis, and enhanced glutamine catabolism. In addition, we demonstrated that Tfcp2l1 induces FAO through activating a rate-limiting enzyme Cpt1a. Furthermore, the Tfcp2l1-Cpt1a-FAO axis promotes the survival of 2i-ESCs and blastocysts in a diapause-like state upon mTOR inhibition. Taken together, we establish Tfcp2l1 as an important transcription factor in the metabolic regulation in ESCs and uncover FAO as a key pathway for pluripotent stem cells and blastocysts to survive during metabolic stress. Results Tfcp2l1 expression is higher in 2i-ESCs than in FBS-ESCs To identify the key transcription factors that are more highly expressed in ESCs grown in 2i medium (2i-ESCs) than in ESCs in FBS-containing medium (FBS-ESCs), we applied an algorithm that considers both absolute expression levels and fold-change of expression in 2i-ESCs versus FBS-ESCs (Fig 1A). Using public RNA-seq datasets (Data ref: Marks et al, 2012; Marks et al, 2012), we calculated and rank-ordered the Z-scores for all 981 transcription factors annotated in the gene ontology (GO) term "DNA-binding transcription factor activity" (GO:0003700). Nanog, Oct4, Cdkn2a, Klf2, and Tfcp2l1 were among the top-ranked transcription factors. Of these, we chose to pursue Tfcp2l1, because its functions in ESC biology have been understudied compared with well-known transcription factors. Immunoblotting and RNA-seq confirmed that Tfcp2l1 expression levels were higher in 2i-ESCs than in FBS-ESCs (Figs 1B and EV1A). Figure 1. Tfcp2l1 shields 2i-ESCs from spontaneous cell death but is dispensable for renewal and pluripotency A. Z-score shows transcription factors enriched in 2i-ESCs and FBS-ESCs. B. Immunoblotting shows protein levels of Tfcp2l1, Nanog, and H3K27me3 in 2i-ESCs and FBS-ESCs. C, D. Real-time PCR (C) and immunoblotting (D) showing Tfcp2l1 deletion by CRISPR-Cas9. n = 5 clones; error bars indicate standard error of the mean (SEM); P-values, t-test; five clones were pooled for immunoblotting. E. Alkaline phosphatase staining. The scale bar equals 100 µm. Left, representative stained colonies; right, a bar chart showing the percentage of differentiated and undifferentiated colonies of WT and Tfcp2l1_KO 2i-ESCs. F. Hematoxylin and eosin staining of teratoma formed by WT and Tfcp2l1_KO 2i-ESCs in NSG mice. Scale bars = 100 µm. G, H. Propidium iodide staining to measure spontaneous cell death of WT and Tfcp2l1_KO 2i-ESCs (G) and ESC-FBS (H). A representative histogram is shown on the left and the bar chart on the right is the percentage of spontaneous dead cells. Error bars indicate SEM; n = 3 biological repeats; P-values were calculated using t-test. I, J. Cumulative cell numbers of WT and Tfcp2l1_KO 2i-ESCs (I) and FBS-ESCs (J). Results are from three independent clones of WT and Tfcp2l1_KO ESCs. Source data are available online for this figure. Source Data for Figure 1 [embr202052122-sup-0006-SDataFig1.zip] Download figure Download PowerPoint Click here to expand this figure. Figure EV1. Tfcp2l1 is dispensable for self-renewal and pluripotency of ESCs mRNA levels (FPKM, fragment per kilobase per million, by RNA-seq) of Tfcp2l1 and cMyc in FBS and 2i. n = 4 biological repeats; error bars indicate SEM; P-values were calculated using t-test. The morphology of WT and Tfcp2l1_KO 2i-ESC colonies. Scale bars = 400 µm. Macroscopic views of alkaline phosphatase staining of WT and Tfcp2l1_KO 2i-ESCs. Scale bars = 1 cm. Relative diameter of colonies (arbitrary unit) in alkaline phosphatase staining; n (WT) = 33 clones; n (KO) = 37 clones; error bars indicate SEM; P-value was calculated using Mann–Whitney test. Weight of teratomas formed from WT and Tfcp2l1_KO 2i-ESCs. n = 4 tumors; error bars indicate SEM; P-value was calculated using two-sample t-test. Dotplots of Click-iT EdU incorporation assay showing proliferation of WT and Tfcp2l1_KO 2i-ESCs. Quantitative analysis G0-G1, S, and G2 phases. n = 3 biological repeats; error bars indicate SEM; P-values were calculated using two-sample t-test. Source data are available online for this figure. Download figure Download PowerPoint Tfcp2l1 is dispensable for self-renewal, pluripotency, and proliferation of 2i-ESCs To investigate the function of Tfcp2l1 in 2i-ESCs, we generated Tfcp2l1 knockout (Tfcp2l1_KO) 2i-ESCs using the CRISPR-Cas9 technology (Fig 1C and D). The morphology of WT and Tfcp2l1_KO 2i-ESC colonies was grossly similar (Fig EV1B), as was the self-renewal capacity based on the alkaline phosphatase staining of colonies formed from single cells (Fig 1E). Tfcp2l1_KO 2i-ESCs formed fewer and smaller colonies than WT 2i-ESCs (Fig EV1C and D), observations that may be due to poor adhesion, cell cycle defect, and/or enhanced spontaneous cell death in the KO cells. To test whether Tfcp2l1 plays a role in pluripotency, we transplanted WT and Tfcp2l1_KO 2i-ESCs into immunocompromised NSG (NOD-SCID-gamma) mice to form teratomas. Histological analyses showed that teratomas from both WT and KO 2i-ESCs formed all three germ layers (Fig 1F). In addition, teratomas generated from WT and KO 2i-ESCs were of similar weight (Fig EV1E). Together, these results show that Tfcp2l1 is dispensable for the pluripotency of 2i-ESCs. Furthermore, WT and Tfcp2l1_KO 2i-ESCs had very similar cell cycle profiles (Fig EV1F and G), indicating that Tfcp2l1 is not involved in the proliferation of naïve pluripotent ESCs. Tfcp2l1 shields 2i-ESCs from spontaneous cell death In the course of culturing 2i-ESCs, we observed a higher percentage of spontaneous death in Tfcp2l1_KO 2i-ESCs than WT 2i-ESCs (8% vs. 2%, respectively) (Fig 1G). This protective role of Tfcp2l1 was specific to 2i-ESCs, because WT and Tfcp2l1_KO FBS-ESCs had similar levels of spontaneous cell death (Fig 1H). After 3 passages, the number of WT 2i-ESCs was about 3 times that of Tfcp2l1_KO 2i-ESCs (Fig 1I), while no difference was observed between WT and Tfcp2l1_KO FBS-ESCs (Fig 1J). This difference in cumulative cell number is likely due to the difference in spontaneous cell death because cell proliferation was similar in both WT and Tfcp2l1_KO 2i-ESCs (Fig EV1F and G). Tfcp2l1 activates metabolic genes and represses transcription-related genes in 2i-ESCs To search for the mechanism underlying the protective role of Tfcp2l1 against spontaneous cell death, we set off to determine genes that are downstream targets of Tfcp2l1. To this end, we identified Tfcp2l1 bound genes by ChIP-seq (Fig 2A and Dataset EV1) and Tfcp2l1 regulated genes by RNA-seq (Fig 2B and Dataset EV2). After integrating both bound and regulated gene datasets, we obtained 1,368 genes whose expression was dependent on Tfcp2l1 and that were bound by Tfcp2l1 (Fig 2C and Dataset EV3). These 1,368 genes were designated as Tfcp2l1 direct targets. Among them, 852 were activated and 516 repressed by Tfcp2l1. Figure 2. Integrated genomic approaches are showing Tfcp2l1 regulates metabolic genes in 2i-ESCs A. Heatmaps demonstrating a global view of Tfcp2l1 binding on 30641 genes. Shown are two independent repeats of Tfcp2l1 ChIP-seq. The heatmap scale is enrichment (ChIP versus input, log2). B. Hierarchical clustering based on 1,582 differentially expressed (WT versus Tfcp2l1_KO 2i-ESCs FDR adjusted P-value < 0.001, ANOVA) genes identified by RNA-seq. Shift genes to mean of zero and scale to standard deviation of one. The heatmap scale is fold change (log2). C. Identify direct targets of Tfcp2l1 by integrating Tfcp2l1 bound genes (ChIP-seq) and dependent genes (RNA-seq). Pathway analysis is done by using Tfcp2l1-activated and Tfcp2l1-repressed genes. D, E. Gene Ontology (GO) terms (biological function) of Tfcp2l1-activated genes (D) and Tfcp2l1-repressed genes (E). GO terms related to metabolism are highlighted in purple, and those to transcription are orange. F, G. Gene set enrichment analyses for assessing the enrichment of Tfcp2l1-activated genes (F) and Tfcp2l1-repressed genes (G) in 2i-ESCs and FBS-ESCs. The number of permutations is 1,000 times of phenotypes. Download figure Download PowerPoint We reasoned that Tfcp2l1-activated and Tfcp2l1-repressed direct targets might have different functions since they moved in different directions upon Tfcp2l1 deletion. Therefore, we applied Gene Ontology (GO) analysis separately to Tfcp2l1-activated and Tfcp2l1-repressed direct targets. GO analysis showed that six of the ten top GO terms enriched in Tfcp2l1-activated genes were related to metabolism (Fig 2D), while the six top enriched GO terms in Tfcp2l1-repressed genes were associated with transcription and chromatin binding (Fig 2E), suggesting that Tfcp2l1 mainly activates metabolic genes and represses transcription/chromatin-related genes. We then tested whether Tfcp2l1-activated genes and Tfcp2l1-repressed genes were associated with the phenotypes of 2i-ESCs and FBS-ESCs. To this end, we applied the gene set enrichment analysis (GSEA) by using Tfcp2l1-activated genes and Tfcp2l1-repressed genes as two separate gene sets. Tfcp2l1-activated genes were enriched in features (genes) associated with 2i-ESCs, while Tfcp2l1-repressed genes were enriched in features associated with FBS-ESCs (Fig 2F and G). The Leading Edge Analysis revealed that lipid metabolic genes contributed to the enrichment of the Tfcp2l1-activated gene set in 2i-ESCs (Dataset EV4). Therefore, Tfcp2l1-activated lipid metabolic genes connect Tfcp2l1 to the unique transcriptional phenotype of naïve pluripotent stem cells. Tfcp2l1 maintains homeostasis of FAO in 2i-ESCs The Tfcp2l1-mediated activation of metabolic genes in 2i-ESCs prompted us to determine the major metabolic pathway(s) that is (are) altered in Tfcp2l1_KO 2i-ESCs. To this end, we initially focused on measuring fatty acid oxidation (FAO), glycolysis, and mitochondrial function in WT and Tfcp2l1_KO ESCs using real-time metabolic analyses based on the Seahorse platform. By utilizing the Seahorse FAO Stress Test, we found that FAO decreased in Tfcp2l1_KO 2i-ESCs compared with WT cells (Fig 3A). Basal, maximal, and spare respiration capacity of mitochondria resulting from FAO were all higher in WT 2i-ESCs than in Tfcp2l1_KO cells (Fig 3B), indicating that Tfcp2l1 induces FAO in 2i-ESCs. The regulatory role of Tfcp2l1 in FAO is specific to 2i-ESCs, because there was no obvious difference between WT and Tfcp2l1_KO FBS-ESCs (Fig EV2A and B). The isotopic tracing experiment confirmed that FAO levels were reduced in Tfcp2l1_KO 2i-ESCs compared with WT cells (Fig EV2C). Figure 3. Tfcp2l1 deletion impairs fatty acid oxidation in 2i-ESCs A. Real-time fatty acid oxidation (FAO) profile of WT and Tfcp2l1_KO 2i-ESCs using the FAO stress test kit in Seahorse XF. Oxygen consumption rate (OCR) was measured. Error bars indicate SEM; n = 5 biological repeats. Palm-BSA, palmitate-conjugated BSA, a substrate of FAO; Oligomycin, an inhibitor of ATP synthase (complex V); FCCP, Carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone, a proton gradient uncoupler; Rotenone and Antimycin, a complex I inhibitor and a complex III inhibitor, respectively. B. Basal respiration, maximal respiration, and spare respiration capacity were calculated from (A). n = 5 biological repeats; error bars indicate SEM; P-values were calculated using Student's t-test. C, D. Mito Stress Assay (C) and Glycolysis Stress Assay (D) in Seahorse XF. Error bars indicate SEM; n = 3 biological repeats; no time point shows a statistically significant difference between WT and KO 2i-ESCs. E. Real-time OCR profile using FAO substrate (Palm-BSA) or BSA control followed by FAO inhibitor, etomoxir. n = 5 biological repeats; error bars indicate SEM. F. Quantification of OCR changes after etomoxir addition. n = 5 biological repeats; error bars indicate SEM; P-values were calculated using two-sample t-test. G. Mitochondrial membrane potential (left) and amount (right) were measured by TMRM and MitoTracker Green, respectively. Representative histograms were shown. The numbers shown are mean fluorescence intensity; P-values were calculated using two-sample t-test. Source data are available online for this figure. Source Data for Figure 3 [embr202052122-sup-0007-SDataFig3.xlsx] Download figure Download PowerPoint Click here to expand this figure. Figure EV2. Tfcpl21 regulates FAO in 2i-ESCs but not FBS-ESCs The FAO Stress Test of WT and Tfcp2l1_KO FBS-ESCs using Seahorse XF. n = 3 biological repeats; error bars indicate SEM. Basal, maximal, and spare respiration. E
Referência(s)