Hyperactivation of mTORC1 blocks stem cell fate transitions through TFE3-NuRD association
Peizhi Li, Shuhui Xu, Xinyu Wu, Yin Gao, Tanveer Ahmed, Yinghua Huang, Dajiang Qin, Baoming Qin, Lulu Wang, Xueting Xu
TL;DR
Excessive mTORC1 activity blocks stem cell development by working with TFE3 and NuRD to silence key genes.
Contribution
Identifies a shared mechanism involving TFE3 and NuRD in mTORC1-mediated repression of stem cell fate transitions.
Findings
Hyperactivated mTORC1 causes TFE3 to move to the nucleus and recruit the NuRD complex.
The TFE3-NuRD complex represses genes essential for stem cell fate transitions.
This mechanism applies to both pluripotency exit and somatic cell reprogramming.
Abstract
Mechanistic target of rapamycin complex 1 (mTORC1) integrates signals from nutrients, growth factors, and cellular stress to regulate biosynthesis and maintain homeostasis. Dysregulated mTORC1 disrupts stem cell homeostasis and impairs cell fate transitions in vivo and in vitro. Previous studies have shown that mTORC1 hyperactivation promotes nuclear translocation of TFE3, blocking pluripotency exit in both mouse and human naïve embryonic stem cells. Similarly, our earlier work has demonstrated that sustained mTORC1 activation impedes somatic cell reprogramming via the transcriptional coactivator PGC1α. This raises the question of how mTORC1 coordinates gene transcription across distinct transitions in pluripotent cells. Here, we show that TFE3 mediates the transcriptional blockade induced by mTORC1 hyperactivation during reprogramming. Notably, during both pluripotency exit and…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 10
Figure 11
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 12
Figure 13- —the National Key Research and Development Program of China
- —the National Natural Science Foundation of China
- —the China Postdoctoral Science Foundation
- —the Science and Technology Planning Project of Guangdong Province, China
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsPluripotent Stem Cells Research · PI3K/AKT/mTOR signaling in cancer · FOXO transcription factor regulation
Introduction
The mechanistic target of rapamycin complex 1 (mTORC1) is a central regulator of cellular metabolism in eukaryotes. It integrates upstream signals to coordinate a broad network of anabolic and catabolic processes (Liu and Sabatini 2020). Multiple checkpoints monitor the availability of nutrients, energy, and growth factors, preventing full mTORC1 activation until these conditions are met. Once activated, mTORC1 promotes cell growth and proliferation by stimulating a range of biosynthetic pathways-including the synthesis of proteins, ribosomes, and even organelles such as mitochondria. Conversely, various forms of cellular stress-including reactive oxygen species (ROS), organelle dysfunction, and DNA damage-can inhibit mTORC1 activity via these upstream checkpoints. This tightly regulated system ensures proper development, growth, and homeostasis. Unsurprisingly, disruption of these regulatory checkpoints can profoundly affect cellular function and identity, with chronic mTORC1 dysregulation contributing to metabolic disorders, cancer, and aging.
Pluripotent stem cells (PSCs), including embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs), possess the unique ability to self-renew indefinitely while retaining the potential to differentiate into all somatic lineages (Hackett and Surani 2014). They are indispensable for regenerative medicine and for studying development and disease. While basal mTORC1 activity is essential for PSC self-renewal (Bulut-Karslioglu et al, 2016; Xu et al, 2022), its activity is lower in mouse ESCs compared to somatic or differentiated cells. Activation of mTORC1 is required for differentiation (Sampath et al, 2008), whereas its inhibition is necessary for iPSC generation (Wu et al, 2015). Dysregulation of mTORC1, particularly hyperactivation, blocks both pluripotency exit/differentiation and somatic cell reprogramming (Betschinger et al, 2013; Mathieu et al, 2019; Wu et al, 2015). Notably, the mechanisms by which hyperactive mTORC1 impairs pluripotent cell fate transitions converge on transcriptional regulation, involving direct activation of Esrrb by TFE3 during pluripotency exit and sustained induction of mitochondrial biogenesis via PGC1α during reprogramming (Betschinger et al, 2013; Wang et al, 2020). These observations led us to hypothesize that a shared transcriptional regulatory mechanism underlies mTORC1-induced blockade of pluripotent cell fate transitions.
In this study, we first demonstrate that nuclear translocation and activation of TFE3 mediates the blockade of somatic cell reprogramming caused by mTORC1 hyperactivation. We further show that, beyond its canonical role in activating lysosome and autophagy-related genes, TFE3 physically associates with the NuRD transcriptional repressor complex to directly inhibit the expression of pluripotency genes during reprogramming and differentiation-related genes during pluripotency exit. Our findings uncover a novel and direct mechanism by which the TFE3-NuRD complex governs pluripotent cell fate transitions, highlighting the critical role of metabolic control in stem cell fate regulation.
Results
TFE3 mediates the blockade of somatic cell reprogramming induced by mTORC1 hyperactivation
To investigate how hyperactivated mTORC1 impairs reprogramming at the transcriptional level, we utilized mouse embryonic fibroblasts (MEFs) derived from an inducible Tsc1 knockout (KO) mouse model (Fig. 1A). Introduction of Cre recombinase via adeno-associated virus (AAV), as described by Wang et al, 2020, efficiently ablated TSC1 protein expression and activated mTORC1 during reprogramming (Fig. 1B). By day 12, alkaline phosphatase staining showed a moderate decrease in the total number of colonies, whereas immunofluorescence staining for NANOG revealed a more substantial inhibition (Fig. 1C), consistent with our previous findings using a Tsc2 knockdown approach (Wu et al, 2015; Wang et al, 2020).Figure 1. Hyperactivation of mTORC1 promotes the nuclear translocation of TFE3, which in turn inhibits somatic cell reprogramming.(A) Schematic depicting how Tsc1 WT and KO MEFs were acquired and the OSKM reprogramming procedure. (B) Western blot showing TSC1, TFE3 protein levels, and mTORC1 activity on day 5 in Tsc1 WT and KO MEFs transduced with OSKM, and either Ctrl or Tfe3 shRNAs. D day, hereafter in all similar experiments. (C) Images and quantification of AP^+^ colonies (left panel) and phase contrast or NANOG-GFP^+^ colonies (right panel) at day 12 in Tsc1 WT and KO MEFs transduced with OSKM (n = 3 each with biological replicates). Scale bar: 100 µm. (D) The number of differentially expressed genes (DEGs) in Tsc1 WT and KO MEFs transduced with OSKM at day 5 (D5) and day 10 (D10). Fold change (FC) >1.5 and P value <0.05. (E) Venn diagram showing the overlap of upregulated and downregulated genes from OSKM day 5 and day 10 in Tsc1 KO cells compared with WT cells. (F) Gene ontology (GO) analysis of upregulated (UP) and downregulated (DOWN) genes in Tsc1 KO cells compared with WT cells at OSKM day 5 and day 10. (G) Discovery of known motifs from differentially expressed genes (DEGs) at day 5 and day 10 in Tsc1 KO compared to WT MEFs transduced with OSKM as in (D). (H) Fluorescence images showing DAPI and TFE3 localization (left) and quantification of nuclear (N)/cytoplasmic (C) TFE3 ratios (right) at day 5 in Tsc1 WT and KO MEFs transduced with OSKM (n = 2 each with biological replicates, each replicate of data was from more than 100 cells). Scale bar: 10 µm. (I) Western blot showing H3 (nuclear marker), TUBULIN (cytoplasmic marker) and TFE3 protein levels on day 5 in nuclear and cytoplasmic extracts of Tsc1 WT, KO, and KOR (KO + rapamycin) MEFs transduced with OSKM. (J) Western blot showing TSC1, TFE3 protein levels, and mTORC1 activity on day 5 in Tsc1 WT and KO MEFs transduced with OSKM, and either Ctrl or Tfe3 shRNAs. (K) Phase contrast or NANOG immunofluorescent images and quantification of NANOG-GFP^+^ colonies at day 12 in Tsc1 WT and KO MEFs transduced with OSKM and either Ctrl or Tfe3 shRNAs (n = 3 each with biological replicates). Scale bar: 100 µm. Data information: In (C, K), data are presented as mean values ± SD. ***P < 0.001, **P < 0.01, *P < 0.05, n.s., not significant (Student’s t-test). In (F), Gene Ontology (GO) analysis for RNA-seq data was performed at the Gene Ontology Resource website. Source data are available online for this figure.
We next performed RNA sequencing (RNA-seq) on both wild-type (WT) and Tsc1 knockout (KO) cells at early (day 5) and late (day 10) stages of reprogramming. Differential expression analysis identified a total of 2450 differentially expressed genes (DEGs) at the early stage (KO vs. WT, fold change >1.5, P value <0.05), with 1539 upregulated and 911 downregulated genes (Fig. 1D). At the late stage, 3383 DEGs were detected, including 1776 upregulated and 1607 downregulated genes (Fig. 1D). While the upregulated DEGs showed substantial overlap between the two stages, overlap among the downregulated DEGs was relatively limited, with many more genes suppressed at the late stage (Fig. 1E). Gene ontology (GO) analysis revealed that the upregulated DEGs were mainly enriched in metabolic pathways related to autophagy/lysosome, oxidative stress, mitochondria, and apoptosis, reminiscent of increased ROS production due to hyperactive mitochondrial metabolism, potentially linked to autophagy and apoptosis activation (Wu et al, 2015; Wang et al, 2020) (Fig. 1F, left). In contrast, the downregulated genes were enriched in pathways associated with cell proliferation and the cell cycle at the early stage, and with stem cell differentiation and epigenetic remodeling at the late stage (Fig. 1F, right).
HOMER motif analysis of all DEGs revealed significant enrichment of transcription factor binding sites belonging to the MITF/TFE3 basic helix-loop-helix (bHLH) family and the SP/KLF zinc finger (ZF) family at both early and late stages of reprogramming (Fig. 1G). Notably, TFE3 emerged as a key candidate, consistent with previous reports showing its nuclear translocation upon mTORC1 hyperactivation and its role in blocking pluripotency exit (Betschinger et al, 2013; Villegas et al, 2019). To assess whether TFE3 mediates reprogramming inhibition downstream of mTORC1 hyperactivation, we first examined its subcellular localization in Tsc1 KO versus WT cells at the early reprogramming stage. We observed a pronounced shift of TFE3 from the cytoplasm to the nucleus in KO cells (Fig. 1H). Moreover, nuclear TFE3 levels were significantly elevated in Tsc1 KO cells and could be reversed by rapamycin treatment, indicating mTORC1 dependency (Fig. 1I). Next, we knocked down Tfe3 using two effective small hairpin RNAs (shRNAs) in Tsc1 KO reprogramming cells and observed a marked rescue of reprogramming efficiency (Fig. 1J,K). This rescue effect was even more pronounced in our earlier reprogramming model using Oct4-GFP (OG2) MEFs with Tsc2 knockdown-induced mTORC1 hyperactivation (Wu et al, 2015; Wang et al, 2020) (Fig. EV1A–C). Furthermore, nuclear overexpression of ectopic TFE3, but not its cytoplasmic form, effectively blocked reprogramming (Fig. EV1D–F).
Together, these findings establish TFE3 as a central effector of mTORC1-mediated inhibition of somatic cell reprogramming.
TFE3 recruits the NuRD complex to directly bind and repress genes required for pluripotency exit
Villegas et al (2019) reported that hyperactivated TFE3 represses the expression of differentiation-associated genes both in mouse embryonic stem cells (mESCs) and during pluripotency exit. However, their detailed analysis did not support the notion that this repression was mediated by direct chromatin binding. To further investigate the potentially distinct mechanisms through which TFE3 represses gene transcription in these two contexts of pluripotent cell fate transition, we re-analyzed their RNA-seq data (Villegas et al, 2019). Principal component analysis (PCA) showed that cells undergoing pluripotency exit for 34 h were distinct from control mESCs along PC2. In contrast, the two groups with nuclear TFE3 overexpression clustered closely together but were separated from control mESCs along both PC1 and PC2 (Fig. 2A). These patterns suggest that TFE3 induces substantial and persistent transcriptomic changes that extend into the pluripotency exit phase. We regrouped the differentially expressed genes (DEGs) across the four conditions and identified both activated (cluster 4) and repressed (clusters 1–3) gene sets regulated by TFE3, either in mESCs or during pluripotency exit (Fig. 2B). Gene ontology (GO) analysis revealed that activated genes (cluster 4) were enriched in pathways related to lysosomal function, mitochondrial activity, autophagy, and oxidative stress responses (Fig. 2B). In contrast, the repressed genes (clusters 1–3) were enriched in pathways associated with stem cell differentiation, transcriptional regulation, and LIF signaling (Fig. 2B).Figure 2TFE3 directly recruits and interacts with the NuRD complex to repress exit of pluripotency.(A) PCA analysis of the indicated samples. RNA-seq data from Florian Villegas et al, 2019. (B) Hierarchical clustering of RNA-seq data for DEGs, corresponding GO analysis at the indicated samples as in (A). Fold change (FC) >1.5 and P value <0.05. (C) TFE3 interactome overlap was observed between proteins upregulated with TFE3 nuclear localization (Ectopic TFE3 binding) and those downregulated upon Tfe3 knockout (endogenous TFE3 binding). Fold change (FC) >2, P value <0.05. (D) Co-immunoprecipitation using IgG or TFE3 antibody in extracts from TFE3-ERT (+Tam) 2iL mESCs, followed by Western blotting with the indicated antibodies. (E) Schematic depicting how Tsc1 WT and KO OG2 mESCs acquire and pluripotency exit procedure. (F) Western blot showing H3, TUBULIN and TFE3 protein levels in nuclear and cytoplasmic extracts of WT, KO, and KOR (KO + rapamycin, hereafter in all similar experiments) mESCs. (G) Co-immunoprecipitation using TFE3 antibody in nuclear extracts of WT, KO, and KOR cells after N2B27 40 h exit, followed by Western blotting with the indicated antibodies. (H) Knockdown efficiency of Gatad2a and Mbd3 in 2iL mESCs (n = 3 each with biological replicates). (I) Images and quantification of OCT4-GFP^+^ colonies at 72 h in Tsc1 WT and KO cells transduced with Ctrl, Gatad2a or Mbd3 shRNAs in 2iL medium, after initial culture in N2B27 medium for 96 h (n = 3 each with biological replicates). Scale bar: 100 µm. (J) Images and quantification of OCT4-GFP^+^ colonies at 72 h in mESCs transduced with FLAG or TFE3-ERT in combination with Ctrl, Gatad2a, or Mbd3 shRNAs in 2iL medium, after initial culture in N2B27 medium for 96 h (n = 3 each with biological replicates). Scale bar: 100 µm. Data information: In (H–J), data were presented as mean values ± SD. ***P < 0.001, **P < 0.01, and *P < 0.05, n.s. not significant (Student’s t-test). In (B), Gene Ontology (GO) analysis for RNA-seq data were performed at the Gene Ontology Resource website. Source data are available online for this figure.
We hypothesized that nuclear TFE3 interacts with other transcription factors or co-repressors to inhibit the activation of genes required for transitions in pluripotent cell fate. To identify such factors, we leveraged published TFE3 interactome datasets in mESCs (Villegas et al, 2019). Specifically, we examined two datasets: one derived from nuclear-localized TFE3 under mTORC1 hyperactivation, and the other from endogenous TFE3 in wild-type (WT) cells. These analyses revealed a shared set of 49 interacting proteins, among which components of the nucleosome remodeling and deacetylation (NuRD) complex were the most significantly enriched based on GO analysis (Fig. 2C). The NuRD complex is a highly conserved and abundant transcriptional co-regulator in higher eukaryotes, known for its diverse roles in development and disease (Basta and Rauchman 2015). It consists of two functional submodules: an ATP-dependent nucleosome remodeling unit (comprising CHD4, GATAD2A/B, and CDK2AP1) and a histone deacetylation unit (including HDAC1/2, MTA1/2/3, and RBBP4/7), linked by either MBD2 or MBD3 (Tong et al, 1998; Zhang et al, 1998; Xue et al, 1998; Lai and Wade, 2011). Our analysis identified most paralogs across all seven core NuRD subunits (Fig. 2C). The interaction between TFE3 and the NuRD complex was further validated by co-immunoprecipitation (co-IP) in E14 mESCs overexpressing nuclear TFE3 (Fig. 2D).
Researches, primarily from the Hendrich lab, has established that NuRD is essential for proper pluripotency exit and lineage differentiation through finely tuning transcriptional programs (Kaji et al, 2006; Reynolds et al, 2012; Burgold et al, 2019). To investigate whether the TFE3-NuRD interaction mediates the blockade of pluripotency exit in mTORC1-hyperactivated mESCs, we generated Tsc1 KO OG2 mESCs by introducing AAV-Cre into Tsc1^Loxp/Loxp^ mice-derived mESCs (Figs. 2E and EV2A). Consistent with somatic cell reprogramming, TFE3 underwent nuclear translocation in Tsc1 KO cells, which was reversed by rapamycin treatment (Fig. EV2B), nuclear TFE3 levels were significantly elevated in Tsc1 KO OG2 mESCs, and this increase was also reversed by rapamycin treatment (Fig. 2F), knockdown of Tfe3 effectively rescued the block in pluripotency exit in Tsc1 KO OG2 mESCs, as demonstrated both at 96 h after exit and after re-addition of 2iL for 72 h (Figs. EV2C,D and EV3A). Moreover, nuclear overexpression of ectopic TFE3, but not cytoplasmic TFE3, impaired pluripotency exit under the same conditions (Figs. EV2E,F and EV3B).
Importantly, co-immunoprecipitation following nuclear-cytoplasmic fractionation revealed an increased nuclear interaction between TFE3 and the NuRD complex in Tsc1 KO OG2 mESCs compared to WT cells, which was abolished by rapamycin treatment (Fig. 2G). Previous studies have reported that ESCs lacking the NuRD component Mbd3 fail to commit to developmental lineages (Kaji et al, 2006; Reynolds et al, 2012), and those deficient in all three MTA proteins (MTA1/2/3) fail to exit pluripotency (Burgold et al, 2019). However, other reports have shown that ESCs lacking Gatad2b maintain normal differentiation potential (Wang et al, 2021), and that Chd4 silencing leads to upregulation of differentiation-associated genes during ESCs differentiation (Zhao et al, 2017). To rule out the possibility that the observed effects were solely due to NuRD complex disruption, we knocked down Gatad2a and Mbd3, two specific NuRD components, in both Tsc1 WT and KO OG2 mESCs (Fig. 2H). Knockdown of these components did not significantly affect pluripotency exit in Tsc1 WT OG2 mESCs (Figs. EV2G and EV3C), but notably rescued the blocked pluripotency exit phenotype in Tsc1 KO OG2 mESCs (Figs. 2I and EV3D). Similar rescue effects were observed in the nuclear TFE3 overexpression model upon knockdown of these NuRD components (Figs. 2J and EV3E), as well as following chemical inhibition of HDAC1/2 with sodium butyrate (NaB) (Figs. EV2H,I and EV3F,G).
Naïve pluripotency gene expression declined after 34 h of pluripotency exit, whereas expression of primed pluripotency genes increased (Fig. EV4A). To investigate whether these changes are regulated by the nuclear association between TFE3 and the NuRD complex, we examined the expression of three naïve pluripotency genes (Nanog, Esrrb, Klf4) and two primed pluripotency genes (Fgf5, Zic3) after 96 h of pluripotency exit (Fig. EV4B,C). Unexpectedly, although naïve pluripotency gene expression remained higher in Tsc1 KO cells compared to WT cells at 96 h, their expression levels in both groups were markedly reduced compared to those in 2iL mESCs, and this reduction was not reversed by knockdown of NuRD components (Fig. EV4B). In contrast, primed pluripotency gene expression was markedly upregulated in Tsc1 WT mESCs after 96 h of exit, but not in Tsc1 KO cells, this impaired induction was rescued by knockdown of NuRD components (Fig. EV4C). These findings suggest that the nuclear TFE3-NuRD interaction primarily represses the expression of primed pluripotency genes, which may play a key role in driving the exit from the naïve pluripotent state in mESCs.
To investigate how the TFE3-NuRD interaction regulates gene expression during pluripotency exit, we performed CUT&Tag analysis for TFE3 and RBBP4 (a core component of the NuRD complex) in Tsc1 WT and KO mESCs following nuclear-cytoplasmic fractionation under 2iL conditions. We identified 10,735 TFE3 peaks associated with 6627 genes, and 21,260 RBBP4 peaks associated with 11,372 genes, exhibiting either novel or increased binding in Tsc1 KO cells compared to WT cells (Fig. 3A). By overlapping these genes with the downregulated DEGs from TFE3-ERT overexpression during pluripotency exit (Fig. 2B, cluster 1 and 2), we identified 667 genes that were both downregulated and showed increased binding of TFE3 and RBBP4 under 2iL conditions (Fig. 3B, top). These genes were enriched in pathways related to stem cell differentiation, development, and cell fate commitment (Fig. 3B, bottom). Strikingly, primed pluripotency genes, including Lin28a, Zic3, and Fgf5 were among these 667 genes (Fig. 3C), suggesting that primed pluripotency genes are already repressed by the TFE3-NuRD interaction prior to exit. Given previous studies showing that Lin28a is critical for the transition from naïve to primed pluripotency in mESCs (Zhang et al, 2016; Tan et al, 2024), we assessed LIN28A protein levels in our models. As expected, LIN28A expression was reduced in both Tsc1 KO mESCs and in mESCs overexpressing nuclear TFE3 under 2iL conditions (Fig. 3D,E). Moreover, overexpression of Lin28a in TFE3-ERT mESCs rescued the blocked pluripotency exit phenotype (Figs. 3F,G and EV3H), further supporting the conclusion that repression of primed pluripotency genes by the nuclear TFE3-NuRD complex plays a key role in controlling exit from the pluripotent state.Figure 3. The mTORC1-TFE3-NuRD axis inhibits genes essential for pluripotency exit.(A) Heatmap of TFE3 (left), RBBP4 (right) and IgG binding in Tsc1 WT and KO mESCs under 2iL conditions. (B) Venn diagram showing the overlap of genes with increased binding of TFE3 and RBBP4 in Tsc1 KO 2iL mESCs and downregulated DEGs in TFE3-ERT 34 h exit mESCs (Fig. 2B, clusters 1 and 2) (top) and corresponding GO analysis (bottom). (C) Selected genomic views of TFE3, RBBP4 and IgG binding sites in 2iL condition for the indicated genes in Tsc1 WT and KO mESCs. (D) Western blot showing TSC1 and LIN28A protein levels in Tsc1 WT and KO mESCs. (E) Western blot showing TFE3 and LIN28A protein levels in mESCs transduced with FLAG or TFE3-ERT in 2iL medium with or without Tam. (F) RT-qPCR detection of exogenous Lin28a expression levels in 2iL mESCs transduced with FLAG or TFE3-ERT with or without Tam (n = 3 each with biological replicates). (G) Images and quantification of OCT4-GFP^+^ colonies at 72 h in mESCs transduced with FLAG or TFE3-ERT in combination with Ctrl or LIN28A overexpression in 2iL medium, after initial culture in N2B27 medium for 96 h (n = 3 each with biological replicates). Scale bar: 100 µm. Data information: In (F, G), data were presented as mean values ± SD. ***P < 0.001, **P < 0.01, *P < 0.05, n.s., not significant (Student’s t-test). In (B), Gene Ontology (GO) analysis for RNA-seq data was performed at the Gene Ontology Resource website. Source data are available online for this figure.
We also performed CUT&Tag analysis for TFE3 and RBBP4 in Tsc1 WT and KO mESCs following nuclear-cytoplasmic fractionation under pluripotency exit conditions (N2B27, 40 h). Under these conditions, we identified 9643 TFE3 peaks associated with 6702 genes and 14,664 RBBP4 peaks associated with 8667 genes that showed either novel or increased binding in KO cells compared to WT (Fig. EV5A). Overlap with the downregulated DEGs from TFE3-ERT overexpression (Fig. 2B, clusters 1 and 2) revealed 537 genes with increased TFE3 and RBBP4 binding under exit conditions (Fig. EV5B, top). These genes were enriched in pathways related to stem cell development, proliferation, and maintenance of the stem cell population (Fig. EV5B, bottom), and included differentiation-associated genes such as Gata6 (Fig. EV5C). Notably, over 50% of these genes overlapped with the set identified under 2iL conditions (Fig. EV5D). Furthermore, comparison of TFE3 and RBBP4 co-bound genes between Tsc1 WT and KO mESCs under both 2iL and exit (N2B27, 40 h) conditions showed that the co-bound targets in Tsc1 KO cells were consistently enriched in developmental and differentiation-related pathways across both conditions (Fig. EV5E,F).
These findings suggest that the nuclear TFE3-NuRD interaction functions as a transcriptional repressor of genes essential for pluripotency exit-including primed pluripotency genes under 2iL conditions and differentiation-related genes under exit conditions.
TFE3 recruits the NuRD complex to directly bind and repress genes required for the re-establishment of pluripotency
To investigate whether TFE3 also associates with the NuRD complex to block reprogramming, we performed co-immunoprecipitation of TFE3 in cells overexpressing nuclear TFE3 at day 5 of OSKM-induced reprogramming. We detected a strong interaction between TFE3 and NuRD components (Fig. 4A). Moreover, nuclear TFE3 exhibited enhanced binding to NuRD components in Tsc1 KO cells, which was reversed by rapamycin treatment (Fig. 4B). Knockdown of Gatad2a and Mbd3 rescued the reprogramming blockade induced either by mTORC1 hyperactivation (via Tsc2 depletion) or by nuclear TFE3 overexpression (Fig. 4C–E). Similarly, inhibition of HDAC1/2 activity using sodium butyrate (NaB) restored reprogramming efficiency in both models (Fig. EV6A,B).Figure 4. The mTORC1-TFE3-NuRD axis plays a critical role in regulating somatic cell reprogramming.(A) Co-immunoprecipitation using IgG or TFE3 antibody in extracts from OSKM-induced reprogramming TFE3-ERT (+Tam) MEFs at day 5, followed by Western blotting with the indicated antibodies. (B) Co-immunoprecipitation using TFE3 antibody in nuclear extracts from WT, KO and KOR OSKM-induced reprogramming cells, followed by Western blotting with the indicated antibodies. (C) Knockdown efficiency of Gatad2a and Mbd3 shRNAs at day 5 in MEFs transduced with OSKM and either Ctrl or Tsc2 shRNAs (n = 3 each with biological replicates). (D) Images and quantification of phase contrast or OCT4-GFP immunofluorescent images at day 12 in MEFs transduced with OSKM and Ctrl, or Tsc2 in combination with Ctrl, Gatad2a or Mbd3 shRNAs (n = 3 each with biological replicates). Scale bar: 100 µm. (E) Images and quantification of phase contrast or OCT4-GFP immunofluorescent images at day 12 in MEFs transduced with OSKM and either FLAG, or TFE3-ERT in combination with Ctrl, Gatad2a and Mbd3 shRNAs in medium with Tam (n = 3 each with biological replicates). Scale bar: 100 µm. (F) PCA of the specified samples at day 5/10 in Tsc1 WT and KO MEFs transduced with OSKM and either Ctrl or Tfe3 shRNAs. MEF data from GSE93027, iPSCs data from GSE93027, ESCs data from GSE93027. (G) Hierarchical clustering of RNA-seq data for DEGs and corresponding Gene Ontology (GO) analysis at day 5/10 in Tsc1 WT and KO MEFs transduced with OSKM and either Ctrl or Tfe3 shRNAs. Fold change (FC) >2, P value <0.05. (H) RT-qPCR detection of Nanog and Esrrb expression levels at day 12 in MEFs transduced with OSKM and either Ctrl, Tsc2 or Tfe3 shRNAs (n = 3 each with biological replicates). (I) RT-qPCR detection of Nanog and Esrrb expression levels at day 12 in MEFs transduced with OSKM and Ctrl, or Tsc2 in combination with Ctrl, Gatad2a or Mbd3 shRNAs (n = 3 each with biological replicates). Data information: In (C–E, H, I), data were presented as mean values ± SD. ***P < 0.001, **P < 0.01, *P < 0.05, n.s., not significant (Student’s t-test). In (G), Gene Ontology (GO) analysis for RNA-seq data were performed at the Gene Ontology Resource website. Source data are available online for this figure.
Given the context-dependent roles of the NuRD complex in reprogramming (Mor et al, 2018; dos Santos et al, 2014; Wang et al, 2023) and the well-established reprogramming-enhancing effects of HDAC inhibition (Huangfu et al, 2008; Karantzali et al, 2008; Zhang et al, 2014), we conducted additional experiments to clarify the specific role of NuRD in our system. Treatment of control cells with varying doses of NaB revealed that low-dose NaB enhanced iPSC generation, whereas high-dose NaB did not (Fig. EV6A). Importantly, both low and high doses of NaB rescued iPSC generation in Tsc2-depleted cells (Fig. EV6A), suggesting that the observed rescue under mTORC1 hyperactivation is specifically mediated through TFE3-NuRD nuclear binding.
TFE3 is known to induce autophagy and lysosomal functions in somatic cells by activating its canonical CLEAR targets (Sardiello et al, 2009). However, during pluripotency exit in mESCs, TFE3 instead recruits the NuRD complex to repress genes required for pluripotency exit (Figs. 3B and EV5B), indicating a context-dependent chromatin targeting and transcriptional function. To determine whether the downregulated DEGs in Tsc1 KO cells during reprogramming were mediated by TFE3, we performed RNA-seq on KO cells following Tfe3 knockdown. Principal component analysis (PCA) confirmed that mTORC1 hyperactivation impedes reprogramming, whereas Tfe3 knockdown restores the transcriptome to a state resembling that of wild-type cells (Fig. 4F). K-means clustering of Tsc1 KO-induced and Tfe3 knockdown-rescued DEGs revealed four distinct clusters (Fig. 4G). In clusters 3 and 4, canonical TFE3 targets-including metabolic and lysosome-related genes were upregulated under mTORC1 hyperactivation and reversed by Tfe3 depletion, validating our data analysis (Fig. 4G). In contrast, genes in cluster 1 were downregulated in Tsc1 KO cells and rescued upon Tfe3 knockdown. These genes were enriched in biological processes such as in utero embryonic development, embryonic morphogenesis, and chromosome segregation (Fig. 3G). Notably, most core pluripotency genes, including Nanog and Esrrb, were grouped into cluster 1 (Fig. EV6C), and their expression was restored by knockdown of either Tfe3 or NuRD components (Fig. 4H,I).
To further explore how TFE3 regulates gene expression during reprogramming, and whether this regulation is context-dependent as observed during pluripotency exit, we performed CUT&Tag analysis for TFE3 and RBBP4 in Tsc1 WT, KO, and rapamycin-rescued KO reprogramming cells following nuclear-cytoplasmic fractionation at the late stage of reprogramming, when gene deregulation in Tsc1 KO cells was most pronounced (Fig. 1F). We identified 56,266 TFE3 peaks associated with 11,307 genes and 14,770 RBBP4 peaks associated with 5771 genes that showed either novel or increased binding in Tsc1 KO cells, which was reversed by rapamycin treatment (Fig. 5A). By overlapping these TFE3- and RBBP4-bound genes with the downregulated DEGs in Tsc1 KO cells that were rescued by Tfe3 knockdown (Fig. 4G, cluster 1), we identified 386 genes enriched in pathways related to stem cell development, differentiation, and cell fate commitment (Fig. 5B,C). This group included several key pluripotency-related genes, such as Esrrb, Dnmt3a, and Nr5a2, which are essential for maintaining or re-establishing pluripotency (Betschinger et al, 2013; Charlton et al, 2020; Heng et al, 2010) (Fig. 5D).Figure 5. The mTORC1-TFE3-NuRD axis inhibits genes essential for somatic cell reprogramming.(A) Heatmap illustrating the density of TFE3(left), RBBP4(right) and IgG CUT&Tag reads on day 10 in Tsc1 WT, KO, and KOR MEFs transduced with OSKM. (B) Venn diagram showing the overlap of genes with increased binding of TFE3 and RBBP4 and downregulated DEGs in Tsc1 KO reprogrammed cells (Fig. 4G, cluster 1). (C) GO analysis of 386 genes from (B). (D) Selected genomic views of TFE3, RBBP4 and IgG binding sites for the indicated genes at day 10 in Tsc1 WT, KO and KOR MEFs transduced with OSKM. (E) Schematic diagram illustrating how the mTORC1-TFE3-NuRD axis controls pluripotent stem cells homeostasis maintenance cell fate transitions. Data information: In (G), Gene Ontology (GO) analysis for RNA-seq data was performed at the Gene Ontology Resource website. Source data are available online for this figure.
Furthermore, comparison of TFE3 and RBBP4 co-bound genes among Tsc1 WT, KO, and rapamycin-rescued cells revealed that genes involved in stem cell development, differentiation, cell fate commitment, and chromatin remodeling were specifically enriched in the TFE3-RBBP4 co-binding cluster in Tsc1 KO cells, but not in WT or rapamycin-rescued cells (Fig. EV6D). These findings provide strong evidence that mTORC1 hyperactivation promotes nuclear TFE3-NuRD association, which directly binds to and represses core pluripotency and epigenetic regulators during reprogramming.
Taken together, our results highlight the central role of the TFE3-NuRD complex in mediating the blockade of pluripotent cell fate transitions induced by mTORC1 hyperactivation.
Discussion
This study reveals that under conditions of mTORC1 hyperactivation, TFE3 translocates to the nucleus and recruits the NuRD complex to repress genes critical for pluripotent cell fate transitions. This shared mechanism, observed in both the re-establishment and exit of pluripotency, underscores a previously unrecognized role for TFE3 as a direct transcriptional repressor that disrupts normal transcriptional remodeling (Fig. 5E).
We previously demonstrated that during somatic cell reprogramming, suppression of the mTORC1-PGC1α pathway is essential for reducing mitochondrial metabolism and cell size (Wu et al, 2015; Wang et al, 2020). Earlier studies reported that nuclear TFE3 directly activates PGC1 coactivators, thereby promoting mitochondrial biogenesis in adipose tissue (Wada et al, 2016). In the present study, transcriptomic analysis identifies TFE3 as a key mediator of the reprogramming blockade induced by mTORC1 hyperactivation. Together with prior findings that nuclear TFE3 accumulation caused by mTORC1 activation hinders pluripotency exit (Betschinger et al, 2013; Villegas et al, 2019), our results suggest that a shared mTORC1-TFE3 signaling axis regulates both the acquisition and loss of pluripotency. Analysis of a published IP-MS dataset revealed that multiple NuRD complex components are among the most prominent interacting partners of TFE3 in mESCs. Functional assays confirmed that the TFE3-NuRD interaction mediates the inhibitory effects of mTORC1 hyperactivation in both somatic cell reprogramming and differentiation contexts. Mechanistically, the TFE3-NuRD complex directly binds to and represses pluripotency genes necessary for reprogramming, while targeting differentiation-related genes critical for pluripotency exit. Although the precise mechanisms by which TFE3-NuRD engages these targets remain to be clarified, our reprogramming data suggest direct associations between TFE3 and key factors such as KLF4 and/or somatic AP-1 transcription factors (Fig. 1G). Furthermore, TFE3 co-occupies a broad set of genomic loci with RBBP4 across various cell types during both directions of pluripotent fate transitions (Figs. EV5E,F and EV6D), suggesting a widespread but context-dependent regulatory role. However, under mTORC1 hyperactivation, this basal TFE3-NuRD co-binding network is disrupted and replaced with an alternative regulatory network that impairs pluripotent cell fate transitions. Whether this basal TFE3-NuRD interaction contributes to gene repression under physiological conditions remains to be explored.
TFE3 is a well-characterized transcription factor known for regulating genes involved in autophagy, lysosomal biogenesis, and cellular metabolism, and has been implicated in numerous diseases and cancers (Li et al, 2023). mTORC1 normally phosphorylates TFE3, sequestering it in the cytoplasm and thereby inhibiting its transcriptional activity, including the activation of autophagy and lysosomal genes (Martina et al, 2014). However, recent studies have shown that Tsc1/2 knockout leads to mTORC1 activation and nuclear accumulation of TFE3, which blocks pluripotency exit by disrupting RagC/D-mediated lysosomal signaling (Villegas et al, 2019). This disruption causes mTORC1 to dissociate from lysosomes, leading to TFE3 hypophosphorylation and nuclear translocation (Zwakenberg et al, 2024). Consistent with these findings, our study demonstrates that mTORC1 hyperactivation drives ectopic nuclear localization of TFE3 and induces expression of lysosome- and autophagy-related genes. Paradoxically, however, autophagy is ultimately suppressed during reprogramming (Wu et al, 2015), suggesting the presence of a yet-unidentified barrier between transcriptional activation and functional autophagy under conditions of mTORC1 hyperactivation.
Several limitations of our current model must be acknowledged. The N2B27-based pluripotency exit system is limited by its short timeframe and substantial cell death, which may obscure the full mechanism of the TFE3-NuRD axis. In particular, this axis may also contribute to promoting cell survival, further complicating efforts to comprehensively model the complexity of differentiation and development. This underscores the need to investigate the TFE3-NuRD interaction in more physiologically relevant models of differentiation and embryogenesis. Nonetheless, our findings may have implications for preventing the generation of abnormal progeny during stem cell differentiation. This is partly supported by observations that mTORC1 dysregulation due to upstream germline mutations or loss-of-function alterations often results in benign tumors, rather than malignancies, during early development or tissue rejuvenation. Future studies are needed to fully elucidate how the mTORC1-TFE3-NuRD network maintains stem cell homeostasis under physiological conditions, while contributing to disease, cancer, and aging in pathological states.
Methods
Reagents and tools tableReagent/resourceReference or sourceIdentifier or catalog number Experimental models Plat-EATCCCVCL_B488HEK293T cellsATCCCVCL_0063E14 mouse ESCsATCCCRL-1821Tsc1^Loxp/Loxp^ miceThe Jackson Laboratory005680129 miceBeijing Vital River Laboratory Animal Technology Co., Ltd.J002448OG2 transgenic miceThe Jackson Laboratory004654 Recombinant DNA pMXs-Oct3/4Addgene13366pMXs-Sox2Addgene13367pMXs-Klf4Addgene13370pMXs-c-MycAddgene13375pLKO.1-TRC cloning vectorAddgene10878pPB-CAG-Tfe3::ERT2-pgk-hphAddgene48756 Antibodies rabbit anti-ACTINSigma-AldrichA2066rabbit anti-TFE3Sigma-AldrichHPA023881rabbit anti-TSC1Cell Signaling Technology6935rabbit anti-TSC2Cell Signaling Technology4308rabbit anti-p-S6Cell Signaling Technology4858rabbit anti-p-S6KCell Signaling Technology9205rabbit anti-S6KCell Signaling Technology9202rabbit anti-S6Cell Signaling Technology2217rabbit anti-p-4EBP1 T37/46Cell Signaling Technology2855rabbit anti-4EBP1Cell Signaling Technology9452rabbit anti-EIF4ECell Signaling Technology2067rabbit anti-CHD4Cell Signaling Technology12011rabbit anti-MBD3Cell Signaling Technology14540 Trabbit anti-RBBP7Cell Signaling Technology6882 Trabbit anti-HDAC1Cell Signaling Technology5356 Trabbit anti-HDAC2Cell Signaling Technology5113 Trabbit anti-GATAD2AAbcamab188472rabbit anti-GATAD2BAbcamab224391rabbit anti-RBBP4Novus BiologicalsNB500-123rabbit anti-MTA1SelleckF0788Rabbit anti-NANOGBethylA300-397AHRP-labeled goat anti-rabbit IgG (H + L)BeyotimeA0208Anti-rabbit IgGBeyotimeA0208Alexa Fluor 488 goat anti-rabbit IgGInvitrogenA11008 Chemicals, enzymes and other reagents DMEM high glucoseThermo Fisher41965062Fetal bovine serum (FBS)NTCSFBEFetal bovine serum (FBS)BiowestS1580GlutaMaxGibco35050079NEAAGibco11140050Penicillin/streptomycinHycloneSV30010Knockout serum replacementInvitrogen10828028β-mercaptoethanolGibco2198503LIFMilliporeLIF2010Neurobasal mediumGibco21103049DMEM/F12HycloneSH30023.01B27Gibco17504044N2Gibco17502048PD0325901StemRD391210-10-9CHIR99021StemRD252917-06-9PolyethylenimineSigma764965polybreneSigmaH9268DMSOSigmaD1435RapamycinSelleckS1039NaBSigmaB5887TamoxifenSigmaH6278TRIzol reagentThermo Fisher15596026 Reagent/resource Nuclear/Cytoplasmic Protein Extraction KitBeyotime BiotechnologyBeyoKit® P0028NovoNGS® CUT&Tag 3.0 High-Sensitivity KitNovoproteinN259-YH01 Others Carl Zeiss microscopeAxio Vert.A1ZEN software v2.0ZeissThe FUSION SOLO 4 M machineVilber LourmatABI 7500 machineRocheThe MGIEasy RNA Library Prep SetMGI1000006384 Target sequences for shRNAs in this study
Gatad2a 5′- CAGACCTCCTCCACTCGAATA -3′5′- GCCTCTTCATGGCCATCTATA -3′ Mbd3 5′- GCGCTATGATTCTTCCAACCA -3′5′- AAGTCACTTTCCTTCAATAAA -3′ Tfe3 5′- AGCCAGAACAGCTGGACATTG-3′5′- GAATGGTGGCAAAGGTATAAT -3′ Tsc2 5′- GGATAGAGCCTCAGAGAAG-3′ Primers for qPCR assays in this study Gatad2aForward: 5′- CGGGGTACCCAGAACATTCC -3′Reverse: 5′- TCTGATGTTGTGAATCTGGGC -3′ Mbd3Forward: 5′- GGCTATTACAGGCCAGGGTC -3′Reverse: 5′- AAAGTGACTTCCTGGTGGGC -3′ Lin28aForward: 5′- GGCATCTGTAAGTGGTTCAACG -3′Reverse: 5′- CCCTCCTTGAGGCTTCGGA -3′ NanogForward: 5′- CTCAAGTCCTGAGGCTGACA -3′Reverse: 5′- TGAAACCTGTCCTTGAGTGC -3′ EsrrbForward: 5′- TTCTCATCTTGGGCATCGTG -3′Reverse: 5′- AATCTGAGTTGGCGAGGGC -3′ Klf4Forward: 5′- CCAGCAAGTCAGCTTGTGAA -3′Reverse: 5′- GGGCATGTTCAAGTTGGATT -3′ Fgf5Forward: 5′- GATCGCGGACGCATAGGTATTA -3′Reverse: 5′- CTGTGTCTCAGGGGATTGTAGG -3′ Zic3Forward: 5′- TCCTTCAAGGCGAAGTACAAACTG -3′Reverse: 5′- GGTTTCTCACCTGTATGGGTCCT -3′
Animal experiments
Animal experiments were compliant with all relevant ethical regulations regarding animal research, and were conducted under the approval of the Animal Care and Use Committee of the Guangzhou Institutes of Biomedicine and Health under license numbers 2021008 and 2024018. Tsc1^Loxp/Loxp^ mice were bought from The Jackson Laboratory (005680), and Tsc1^Loxp/Loxp^ mice were crossed with Oct4-GFP (OG2) mice to generate Tsc1^Loxp/+^-OG2^−/+^ mice, which were subsequently bred to obtain Tsc1^Loxp/Loxp^-OG2 mice. The 129 and OG2 mice were generously provided by Professor Jing Liu (Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou, China).
Plasmids, molecular cloning, and virus packaging
pMXs retroviral vectors separately expressing OSKM were transduced into Plat-E cells using a modified polyethylenimine method (Longo et al, 2013). The medium was changed 8–12 h after transduction. Retrovirus-containing medium was collected at 48 and 72 h post-transfection, filtered, and diluted 20–25% with fresh medium before use. The retrovirus mixture was supplemented with 8 µg/mL polybrene to enhance infection efficiency.
For shRNA vectors, oligonucleotides containing the target sequences were annealed and cloned into pLKO.1 vector (Addgene). Target sequences for shRNAs in this study have been deposited in the Reagents and tools table. Lentiviruses carrying shRNAs were packaged in HEK293T cells using packaging vectors (Warlich et al, 2011) and a modified polyethylenimine method. Fresh medium was added 10–14 h after transfection. Lentivirus-containing medium was harvested 48 h post-transfection, filtered, aliquoted, and stored at −80 °C. Prior to infection, shRNA lentiviruses were diluted 1:40 (for MEFs) or 1:1 (for mESCs) with fresh medium, and 8 µg/mL polybrene was included in the mixture to aid in infection.
The TFE3-ERT mutation plasmid was packaged in HEK293T cells using packaging vectors (Warlich et al, 2011) and a modified polyethylenimine method. Lentivirus-containing medium was collected 48 h after transfection, filtered, aliquoted, and stored at −80 °C. Before infection, TFE3-ERT lentiviruses were diluted 1:15 (for MEFs) or 1:1 (for mESCs) with fresh medium, and 8 µg/mL polybrene was added to the mixture to facilitate infection.
MEF preparation and culture
Mouse embryonic fibroblasts (MEFs), including Tsc1^Loxp/Loxp^ and OG2 (Esteban et al, 2010) MEFs, were derived from E13.5 embryos and cultured in Dulbecco’s modified Eagle’s medium (DMEM) high glucose supplemented with 10% fetal bovine serum, 1% GlutaMax, 1% nonessential amino acids (NEAA) and 0.5% penicillin/streptomycin.
iPSC generation
OG2 MEFs at the second passage (P2) were seeded at 2 × 10^4^ cells/well of a 12-well plate and transduced with 2 mL of retrovirus-containing medium for a standard OSKM reprogramming experiment (Esteban et al, 2010). Day 0 was designated as the time point for the first infection. After 12 h of OSKM virus infection, 1 mL of shRNA lentivirus-containing medium was added per well for 8–10 h. A second round of 12 h OSKM virus infection followed. After completing the retroviral infections, the medium was changed to KSR reprogramming medium (high-glucose DMEM supplemented with 10% FBS, 10% knockout serum replacement (KSR), 1% GlutaMax, 1% NEAA, 1% sodium pyruvate, 0.5% penicillin/streptomycin, 0.1 mM β-mercaptoethanol, and 1000 U/mL LIF). GFP^+^ or NANOG^+^ iPSC colonies were counted at the indicated time points. Images were captured using ZEN software v2.0 as indicated. For alkaline phosphatase (AP) staining, cells were washed twice with PBS, fixed with 4% paraformaldehyde for 2 min, washed with TBST (50 mM Tris-HCl at pH 8.0, 150 mM NaCl, and 0.1% Tween 20), and then incubated with freshly prepared AP staining solution (4.5 μL 50 mg/mL nitro blue tetrazolium, 3.5 μL 50 mg/mL 5-bromo-4-chloro-3-indolyl phosphate in 100 mM Tris–HCl, pH 9.5, 100 mM NaCl, 50 mM MgCl_2_) in the dark for 15 min at room temperature. Cells were washed twice with PBS, and data were collected using Image Pro Plus 6.0 after scanning.
For Cre or control (empty) adenovirus treatments, Tsc1^Loxp/Loxp^ MEFs at the first passage (P1) were transduced with adenoviruses purchased from Vigene Biosciences (Jinan, China) for 12 h, followed by a change to MEF medium. Cre-transduced P2 Tsc1^Loxp/Loxp^ MEFs were used for subsequent reprogramming.
For inhibitor treatments, DMSO, rapamycin, NaB, or Tamoxifen were added post-infection.
mESCs preparation and culture
Tsc1^Loxp/Loxp^-OG2 mESCs were derived from E3.5 embryos of Tsc1^Loxp/Loxp^-OG2 mice and subsequently infected with either AAV-empty (WT OG2 mESCs) or AAV-Cre (KO OG2 mESCs). E14 mESCs were provided by Dr. I. Samokhvalov (Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, China). Cultured in 2iL condition, mESCs were seeded on 0.2% gelatin-coated plates and cultured in 2iL medium, consisting of a 1:1 mix of Neurobasal medium and DMEM/F12, supplemented with 1% GlutaMax, 1% NEAA, 1% sodium pyruvate, 0.5% penicillin/streptomycin, 0.1 mM β-mercaptoethanol, 1% B27, 0.5% N2, 1 µM PD0325901, 3 µM CHIR99021, and 1000 U/mL LIF.
Pluripotency exit experiment
WT/KO OG2 mESCs infected with shRNA or TFE3-ERT virus were seeded at 4 × 10^4^ cells/well of a 12-well plate coated with 0.2% gelatin in 2iL medium. After overnight attachment, the medium was changed to N2B27 medium (2iL medium withdrawal 2iL) and maintained for 96 h. Subsequently, cells were passaged at 4 × 10^4^ cells/well in a 12-well plate coated with 0.2% gelatin in 2iL medium, and RNA samples were collected. The mESCs were cultured in 2iL conditions for an additional 72 h, followed by counting GFP^+^ colonies. Photos were captured using ZEN software v2.0, where indicated.
For inhibitor treatments, DMSO, rapamycin (0.3 nM), NaB (0.5 mM), or Tamoxifen (0.15 µM, Tam) was were added after infection.
RNA isolation and RT-qPCR
Cells were lysed using TRIzol reagent for RNA isolation, and 2 µg of RNA were used for reverse transcription. Quantitative PCR (qPCR) reactions were performed in triplicate using an ABI 7500 machine with SYBR green. Data were analyzed and normalized relative to Actb expression values. Primers for qPCR assays in this study have been deposited in the Reagents and tools table.
Immunofluorescence
For NANOG protein (iPSCs) and TFE3 protein (MEFs, mESCs and cells from reprogramming day 5) immunofluorescence, cells on plates were fixed with 4% paraformaldehyde for 30 min, washed three times with PBS, and then incubated in 0.1 M glycine for 10 min. They were subsequently blocked and permeabilized in 3% BSA and 0.2% Triton-X100 in PBS for 2 h. The cells were then incubated overnight at 4 °C with primary NANOG antibody, followed by incubation with Alexa Fluor 488 goat anti-rabbit IgG secondary antibodies for 2 h after washing three times with PBS. Finally, cells were stained with DAPI (1:5000 in PBS) for 2 min, washed three times with PBS, and observed using a Carl Zeiss microscope. Images were captured with ZEN software v2.0. Both primary and secondary antibodies were diluted 1:400.
Nuclear-cytoplasmic fractionation
Nuclear and cytoplasmic proteins were isolated from mESCs or iPSCs, using the Nuclear/Cytoplasmic Protein Extraction Kit according to the manufacturer’s instructions with minor adjustments. Briefly, 1 × 10⁷ cells were harvested, washed twice with ice-cold PBS, and lysed in cytoplasmic extraction buffer (containing 1 mM PMSF) on ice for 2 min. The lysate was centrifuged at 12,000 × g for 10 min at 4 °C to collect the cytoplasmic supernatant. The nuclear pellet was then resuspended in nuclear extraction buffer (supplemented with 1 mM PMSF), vigorously vortexed at maximum speed (2500 rpm) for 15–30 s to disperse nuclei, and incubated on ice for 30 min with intermittent vortexing every 2 min. Finally, the nuclear suspension was centrifuged at 12,000 × g for 10 min at 4 °C, and the supernatant containing nuclear proteins was aliquoted and stored at −80 °C.
Western blotting and co-immunoprecipitation
For Western blotting, cells were washed twice with DPBS and lysed on ice in lysis buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 1% bromophenol blue, and 10% glycerol). Samples were boiled for 8 min, subjected to SDS/PAGE and transferred to a 0.45 µm PVDF membrane (Millipore). Membranes were blocked with blocking buffer (5% nonfat milk in TBST (50 mM Tris-HCl pH 8.0, 150 mM NaCl, and 0.1% Tween 20)) and incubated with primary antibodies (diluted in blocking buffer) for 2 h at room temperature or overnight at 4 °C, After washing five times with TBST, membranes were incubated with the secondary antibody for 1 h at room temperature, then washing five times with TBST. Signals were detected using the Amersham ECL Prime Western Blotting Detection Kit (GE Healthcare, Boston, MA, USA), visualized with the FUSION SOLO 4 M machine and analysed with FusionCapt Advance Solo4.16.15. All primary antibodies were diluted 1:1000. The secondary antibody was diluted 1:2000.
For co-immunoprecipitation, 1 × 10^7^ TFE3-ERT (+Tam) mESCs and cells from reprogramming day 6 were lysed in 600 μL of TNE buffer (50 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, 0.5% Nonidet P-40, 1 mmol/L EDTA plus 6 μL cocktail(100x), and 6ul PMSF (100x)) for 1 h on ice. Cell lysates were cleared by centrifugation at 14000 × g for 5 min at 4 °C, take 40 μL of the supernatant cell lysate was saved as input, 250 μL of the supernatant was separately incubated overnight at 4 °C with anti-TFE3 antibody (1:500) or anti-rabbit IgG as a negative control. Protein A/G beads were washed three times with TNE buffer and then mixed with the cell lysate for 3–4 h or overnight. The bead mixture were washed five times with washing buffer (50 mM Tris-HCl, pH 7.9, 100 mM KCl, 5 mM MgCl2, 0.2 mM EDTA, 10% Glycerol, 0.1% NP-40, and 3 mM β-mercaptoethanol). The beads were transferred to a new tube and eluted in lysis buffer, boiled for 8 min, and prepared for Western blotting.
RNA-seq and analysis
RNA from reprogramming days 5 and 10 were extracted from all experimental groups using a standard protocol. A total of 1 µg RNA per sample was utilized for library preparation at China National GeneBank (CNGB). The sequencing libraries were generated using the MGIEasy RNA Library Prep Set (MGI, 1000006384) following the manufacturer’s protocol. Subsequently, the libraries were sequenced on an MGISEQ-2000 sequencer in paired-end 100 mode.
For RNA-seq data analysis, the reads were initially processed by removing adapter sequences, reads containing poly-N sequences, and low-quality reads. All cleaned reads were then mapped to the mouse reference genome using the HISAT tool in an RESM-based pipeline, P value is corrected for multi-testing. Pairwise differential expression and principal component analysis (PCA) plots were generated using DESeq2 (v.1.24) in R (v.3.6.0). Gene Ontology (GO) analysis for RNA-seq data were performed at the Gene Ontology Resource website. The RNA-Seq expression matrices have been deposited in Dataset EV1.
CUT&Tag and data analysis
For CUT&Tag library construction, NovoNGS® CUT&Tag 3.0 High-Sensitivity Kit for Illumina was used. Around 1 × 10^5^ living cells were collected for each sample, and the CUT&Tag library was sequenced on an Illumina Nova6000, carried out by CNGB.
Sequenced reads were aligned to the mouse reference genome (mm10) using Bowtie2 with the parameters: --end-to-end--very-sensitive--local--no-mixed--no-discordant--no-unal--phred33-I10-X700. Then, Sambamba was used to remove duplicate reads and Ssamtools was used to sort the data. Peaks were called using MACS2 (v.2.2.7.1) with the following parameters: --SPMR--nomodel--extsize 200--keep-dup all--q 0.001. Reads that mapped to mitochondrial DNA or unassigned sequences were discarded. For paired-end sequencing data, only transformed into read coverage files (bigwig format) using deepTools with RPKM normalization. For the sample without repeats, differential binding region was analysis by manorm (v.1.3.0) with P value ≤0.05, and P value is corrected for multi-testing. The CUT&Tag peak location matrices have been deposited in Dataset EV1.
Statistical analysis
We used a two-tailed t-test (*P < 0.05; **P < 0.01; ***P < 0.001) to analyze multiple independent experiments, with biological but not technical replicates. Values are shown as the mean ± standard deviation (s.d.). Details of the experimental replicates are displayed in the corresponding legends.
Supplementary information
Peer Review File Dataset EV1 Source data Fig. 1 Source data Fig. 2 Source data Fig. 3 Source data Fig. 4 Source data Fig. 5 Figure EV1 Source Data Figure EV2 Source Data Figure EV3 Source Data Figure EV3 Source Data Figure EV4 Source Data Figure EV5 Source Data Figure EV6 Source Data Expanded View Figures