Quiescence enables unrestricted cell fate in naive embryonic stem cells

Khoa, Le Tran Phuc; Yang, Wentao; Shan, Mengrou; Zhang, Li; Mao, Fengbiao; Zhou, Bo; Li, Qiang; Malcore, Rebecca; Harris, Clair; Zhao, Lili; Rao, Rajesh C.; Iwase, Shigeki; Kalantry, Sundeep; Bielas, Stephanie L.; Lyssiotis, Costas A.; Dou, Yali

doi:10.1038/s41467-024-46121-1

Download PDF

Article
Open access
Published: 26 February 2024

Quiescence enables unrestricted cell fate in naive embryonic stem cells

Nature Communications volume 15, Article number: 1721 (2024) Cite this article

4297 Accesses
13 Altmetric
Metrics details

Subjects

An Author Correction to this article was published on 12 March 2024

This article has been updated

Abstract

Quiescence in stem cells is traditionally considered as a state of inactive dormancy or with poised potential. Naive mouse embryonic stem cells (ESCs) can enter quiescence spontaneously or upon inhibition of MYC or fatty acid oxidation, mimicking embryonic diapause in vivo. The molecular underpinning and developmental potential of quiescent ESCs (qESCs) are relatively unexplored. Here we show that qESCs possess an expanded or unrestricted cell fate, capable of generating both embryonic and extraembryonic cell types (e.g., trophoblast stem cells). These cells have a divergent metabolic landscape comparing to the cycling ESCs, with a notable decrease of the one-carbon metabolite S-adenosylmethionine. The metabolic changes are accompanied by a global reduction of H3K27me3, an increase of chromatin accessibility, as well as the de-repression of endogenous retrovirus MERVL and trophoblast master regulators. Depletion of methionine adenosyltransferase Mat2a or deletion of Eed in the polycomb repressive complex 2 results in removal of the developmental constraints towards the extraembryonic lineages. Our findings suggest that quiescent ESCs are not dormant but rather undergo an active transition towards an unrestricted cell fate.

Spatially organized cellular communities form the developing human heart

Article Open access 13 March 2024

Single-mitosis dissection of acute and chronic DNA mutagenesis and repair

Article Open access 16 April 2024

N6-methyladenine in DNA antagonizes SATB1 in early development

Article 15 July 2020

Introduction

Cellular quiescence refers to a state of reversible cell cycle arrest at G₀ phase, distinct from irreversible G₀ arrest observed in terminally differentiated and senescent cells¹. Quiescence is generally considered as a protective mechanism to evade unfavorable external conditions or serve as a reservoir to prevent stem cell exhaustion and aging^2,3. Many adult stem cells, including hematopoietic stem cells, hair follicle stem cells, muscle stem cells, and neural stem cells, reside in the quiescent state in vivo¹. They play essential roles in tissue homeostasis and regeneration. The quiescent stem cells are mostly considered “dormant”, capable of rapid exit upon receiving activation or differentiation signals^1,2. Recent studies including ours have shown that naive mouse embryonic stem cells (ESCs) cultured with leukemia inhibitory factor (LIF) and GSK3β/MEK inhibitors (2i)⁴ can enter a quiescent state in vitro either spontaneously or in response to depletion or inhibition of the histone acetyltransferase MOF (also called KAT8 or MYST1), MYC, or fatty acid oxidation (FAO) pathway^5,6. In contrast, inhibiting the mechanistic target of rapamycin (mTOR) pathway in both naive and serum/LIF ESCs leads to a diapause-like growth arrest without inducing quiescent G₀ phase⁷. The quiescent naive ESCs (qESCs) mimic embryonic diapause in vivo, retaining self-renewal ability and developmental potential for three embryonic germ layers^5,6. Interestingly, cancer cells can also enter an embryonic diapause-like quiescent state to evade cell death from drug treatments^8,9. These drug-induced cancer persister cells contribute to most cases of relapse in patients and possess a transcriptional signature similar to that of diapausing blastocysts or quiescent ESCs^8,9. Despite these studies, functional significance of ESC quiescence and how it can be leveraged in regenerative medicine remain unclear.

Preimplantation embryo development is concomitant with remarkable changes in mitochondria. Nuclear distribution of mitochondrial tricarboxylic acid cycle (TCA) enzymes in 2-cell (2C) embryo coincides with global re-setting of chromatin structure and zygotic genome activation¹⁰. Furthermore, metabolic switch between oxidative phosphorylation and glycolysis coincides with the epiblast development and lineage specification¹¹. Naive ESCs derived from the blastocyst are metabolically heterogenous^5,12. We have shown that naive ESCs with inherently low mitochondrial membrane potential (MMP or ΔΨ_m) are quiescent⁵. These spontaneous qESCs reside in a reversible G₀ phase with low metabolic and biosynthetic activity^2,5. Notably, reduced MMP and adenosine triphosphate (ATP) production are common features of the induced qESCs and adult stem cells as well^5,13,14. However, whether they have any impacts on the chromatin milieu and the transcriptional landscape in ESCs is largely unknown.

Here we show that the qESCs are functionally distinct from cycling ESCs. The qESCs are able to differentiate into the extraembryonic lineages with high efficiency in vitro. These cells have molecular characteristics akin to 2C embryos, with increased expression of MERVL, 2C-specific genes and trophoblast master regulators. We further shed light on the active and coordinated changes of the mitochondrial activity, the one-carbon metabolite S-adenosylmethionine (SAM) homeostasis and polycomb repressive complex 2 (PRC2)-mediated H3K27me3 in qESCs during the transition from a restricted to an unrestricted cell fate. Our study suggests that quiescent stem cells are not simply “dormant” or “idle” with poised potential². Instead, they actively undergo the metabolic and epigenetic changes in acquisition of the expanded cell fate plasticity.

Results

The spontaneous qESCs have a unique transcriptome similar to that of 2C embryo

The spontaneous qESCs can be isolated by live-cell sorting after staining with the lipophilic cationic dye Tetramethylrhodamine Methyl Ester (TMRM)^5,15 (Supplementary Fig. 1), which measures the MMP across the inner mitochondrial membrane. The MMP is a direct indicator of mitochondrial function and integrity^16,17. Naive ESCs display intrinsic heterogeneity in MMP, demonstrated by both TMRM staining (Supplementary Fig. 2a, top) and the mitochondrial redox-sensing Grx1-roGFP2 reporter¹⁸ (Supplementary Fig. 2b). Naive ESCs with low MMP signals (bottom 5%, Supplementary Fig. 2a, bottom) exhibited typical features of quiescence^2,5, including the low level of total RNAs per cell, an extremely slow proliferation rate and an increase of G₀ arrest (Fig. 1a, Supplementary Fig. 2c–f). Like cycling ESCs, the spontaneous qESCs (low ΔΨ_m/MMP) retained their ability to differentiate into three embryonic germ layers (Supplementary Fig. 2g, h).

**Fig. 1: The spontaneous qESCs possess a unique transcriptional landscape.**

Total RNA sequencing (RNA-seq) identified 1641 genes with altered expression in the spontaneous qESCs as compared to the ESCs with top 5% MMP (high ΔΨ_m/MMP) after normalization against the ERCC spike-in control (false discovery rate [FDR] < 0.05) (Fig. 1b; Supplementary Fig. 3a, b and Supplementary Data 1). They included 1238 and 403 up- or down-regulated genes in the spontaneous qESCs, respectively. Expression of cell cycle genes (e.g., Cdk1, Ccnb1, Ccnd3, Pdk1) was significantly downregulated in the qESCs as expected (Fig. 1b, left and Supplementary Fig. 3c). While the pluripotency genes (e.g., Oct4, Nanog, Sox2) were expressed at similar levels in the quiescent and cycling ESCs (Supplementary Fig. 3d), endogenous retroviruses (e.g., MERVL, RLTR20C, RLTR6B_Mm, ETnERV3-int) and 2C-specific genes (e.g., Zscan4, Dux, Gm4340)¹⁹ were expressed at significantly elevated levels in the qESCs (Fig. 1b, c, Supplementary Fig. 3e and Supplementary Data 1). The RNA-seq results for the 2C gene signatures¹⁹ were validated by reverse-transcription and real-time PCR (RT-qPCR) (Fig. 1d). Importantly, increased expression of MERVL and 2C genes was also observed in naive ESCs treated with Carbonyl Cyanide 3-chlorophenylhydrazone (CCCP) or complex I electron transport chain inhibitor Phenformin, both of which reduce the MMP^20,21 (Fig. 1e, f and Supplementary Fig. 3f, g). Elevated expression of 2C genes in the spontaneous qESCs was confirmed in two independent naive ESC lines, Oct4-GiP²² and V6.5 2i/LIF ESCs (Supplementary Fig. 3h). Consistently, the spontaneous qESCs isolated from the MERVL-2C::EGFP²³ and Oct4-GiP naive ESC lines had higher levels of 2C::EGFP reporter signals and lower levels of OCT4-GFP signals, respectively (Supplementary Fig. 4a, b). ESCs cultured under the serum/LIF condition do not enter quiescence when mitochondrial function is blocked^5,24 and therefore, did not show changes in 2C gene expression (Supplementary Fig. 3h).

Hierarchical clustering analysis showed that transcriptome of the spontaneous qESCs closely correlated with that of early-stage blastomeres (e.g., zygote, early 2C, 2C)²⁵ as well as transient or genetically induced 2-cell-like cells (2CLCs) with expanded pluripotency^{19,26,27,28,29,30,31,32,33} (Fig. 1g, green-shaded area). Interestingly, the spontaneous qESCs were more closely associated with zygote and early 2C embryos as compared to the transient MERVL-tdTomato⁺ 2CLCs¹⁹ (Fig. 1g). Notably, the extraembryonic placenta development pathways were among the top ranked gene ontology (GO) terms upregulated in the spontaneous qESCs (Fig. 1h and Supplementary Data 1). As expected, transcriptome of the cycling ESCs (high ΔΨ_m/MMP) resembled that of the inner cell mass and 2C-negative cells (e.g., Zscan4^low, MERVL-tdTomato^-) (Fig. 1g, unshaded area).

The spontaneous qESCs exhibit unrestricted developmental potential

The transcriptomic similarity between the spontaneous qESCs and 2C embryos suggests that the spontaneous qESCs may have unrestricted developmental potential. To test this, we subjected the cycling ESCs and qESCs to trophoblast stem cell (TSC) differentiation in vitro. The E14 qESCs exhibited a very high TSC induction efficiency, as denoted by around 40% of CDX2⁺/OCT4⁻ cells at day 7 post TSC induction, whereas the cycling ESCs still maintained robust OCT4 expression with little to no expression of trophoblast marker CDX2 (Fig. 2a). The qESCs expressing the iCdx2 Elf5-2A-mCherry reporter³⁴ gave rise to stable TSC lines with ~77% efficiency (Fig. 2b, c) (referred to as iTSCs herein). The iTSCs demonstrated robust proliferation and exhibited expression levels of the canonical TSC markers Cdx2, Elf5, Eomes, and Tfap2c comparable to those derived directly from blastocysts (Fig. 2d). Similar to TSCs, the iTSCs also lacked expression of the pluripotent gene Oct4 (Fig. 2d). Immunofluorescence analysis confirmed elevated levels of CDX2 and EOMES proteins in these cells, while OCT4 or the primitive endoderm marker GATA6 were not detected (Fig. 2e and Supplementary Fig. 4c). Furthermore, the CpG sites within the Elf5 promoter were hypomethylated in iTSCs when compared to ESCs³⁵ (Fig. 2f). Upon differentiation, the iTSCs underwent morphological alterations, characterized by enlarged and flattened cells, along with the emergence of trophoblast giant cells (TGCs) (Fig. 2g). Furthermore, iTSCs demonstrated the capability to differentiate into all subtypes of trophoblast lineages. Gene markers for the terminally differentiated TGCs, spongiotrophoblasts and labyrinthine trophoblasts, such as Ctsq, Prl3b1, Prl3d1, Prl2c2, Tpbpa, Cdh3, and Itm2a, were significantly upregulated (Fig. 2h). These data strongly support that the spontaneous qESCs have unrestricted cell fate in vitro.

**Fig. 2: The spontaneous qESCs give rise to extraembryonic trophoblast stem cells.**

The spontaneous qESCs have distinct cellular metabolism

We performed untargeted metabolomics analysis for naive ESCs with bottom or top 5% MMP (Fig. 3a). A total of 68 metabolites consistently displayed significant reductions in the qESCs across four independent experiments, with cystine being the sole metabolite showing a notable increase (Fig. 3a and Supplementary Data 2). Specifically, the qESCs demonstrated decreased levels of ATP and metabolites associated with the glycolysis, TCA cycle, pentose phosphate pathway, purine and pyrimidine metabolism (Fig. 3b, Supplementary Fig. 5a–c and Supplementary Data 2). Particularly striking was the marked downregulation of metabolites linked to one-carbon (1C) metabolism in the qESCs (Fig. 3a–c). The 1C metabolic pathway (encompassing folate and methionine cycles) recycles 1C units from glucose, vitamins and amino acid catabolism or through de novo synthesis via serine synthesis pathway (SSP) to fuel important cellular processes³⁶ (Fig. 3c). As shown in Fig. 3b, the glycine, serine, threonine and methionine metabolic pathways were all significantly downregulated in the qESCs. Consistently, the levels of 1C metabolites such as L-homocysteine, SAM, and SAH, as well as SPP metabolites L-Serine and O-Phospho-L-Serine were notably reduced in the qESCs compared to the control cells (Fig. 3d). The metabolic changes in the qESCs were consistent with diminished expression of key genes in the 1C metabolic pathway (Fig. 3e), including Gldc, Phgdh, Psat1, Shmt2, Mthfd1l, and Mthfd2³⁶.

**Fig. 3: Low one-carbon metabolism confers unrestricted cell fate of the spontaneous qESCs.**

SAM supplement is widely used to reverse low levels of intracellular SAM^37,38. Upon SAM supplementation (Fig. 3f), the efficiency of iTSC induction decreased sixfold, as evidenced by the percentage of CDX2⁺/GATA6⁻ cells at day 7 post-induction (Fig. 3g). SAM supplementation also resulted in a global shift of the transcriptome from that of qESCs/2CLCs to conventional ESCs (see below). These results suggest that lower level of 1C metabolite SAM may contribute to increased cell fate plasticity in the qESCs.

The spontaneous qESCs have global reduction of H3K27me3 and higher chromatin accessibility at genes important for the extraembryonic development

The 1C metabolite SAM is a universal methyl donor for histone/DNA methylases³⁹. To examine whether the qESCs have a distinct epigenetic landscape, we assessed chromatin accessibility in qESCs by performing the Assay for Transposase Accessible Chromatin with high-throughput sequencing (ATAC-seq) (Supplementary Data 3). Chromatin accessibility was modestly reduced at 8768 gene loci in the qESCs, many of which were involved in development of three embryonic germ layers (Supplementary Fig. 6a, b, bottom). Interestingly, increase of chromatin accessibility was found at 2183 chromatin loci in the qESCs, which were enriched for genes important for extraembryonic development (Supplementary Fig. 6a, b, top).

We performed Cleavage Under Targets and Release Using Nuclease (CUT&RUN) analysis for H3K4me3 and H3K27me3 marks, given their prominent functions in naive ESCs^40,41. While expression of the core subunits (Suz12, Eed, Ezh1, Ezh2) of PRC2 complex⁴² remained unchanged (Supplementary Fig. 8a), there was a drastic global reduction of H3K27me3 in the qESCs (Fig. 4a and Supplementary Data 4). This was further confirmed by the immunofluorescence analysis (Fig. 4b). In contrast, the H3K4me3 level had no discernable difference in the qESCs (Supplementary Fig. 6c, d, and Supplementary Data 5). Reduction of H3K27me3 in the qESCs mostly occurred at intergenic or intron regions in the genome (Supplementary Fig. 7a). These regions were enriched for the consensus sequences for CDX2 and ELF5⁴³ by Motif analysis (Supplementary Fig. 7b) and were associated with multi-tissue development, including the placenta development pathways (Fig. 4c). Representative H3K27me3 occupancy at Cdx2 and Tfap2c was shown in Supplementary Fig. 7c.

**Fig. 4: Global reduction of H3K27me3 enables unrestricted cell fate of the spontaneous qESCs.**

We next examined chromatin changes at the long terminal repeat (LTR) sequences of endogenous retrovirus MERVL elements. The LTRs evolve as functional promoters to regulate transcription of nearby genes (e.g., 2C genes), leading to generation of chimeric transcripts in ESCs and early embryos^19,44. Consistent with elevated expression of MERVL (Fig. 1c, d), H3K27me3 was significantly downregulated in 662 MERVL elements in the qESCs, with concomitant increase of H3K4me3 and chromatin accessibility (Fig. 4d). Epigenomic and transcriptomic features at a representative MERVL locus were shown in Supplementary Fig. 7d. In contrast, only modest changes in H3K27me3, H3K4me3 and ATAC-seq signals were found at other 2C gene loci in the qESCs (Supplementary Fig. 7e–g), suggesting a specific and non-redundant function of H3K27me3 in de-repressing MERVL expression in qESCs.

CCCP treatment and MAT2A depletion lead to unrestricted cell fate in naive ESCs

To examine whether reducing MMP alters the metabolic and epigenetic states (Fig. 3a, d), we treated naive ESCs with CCCP for 9 h (Fig. 1e). Metabolomics analyses showed a decrease in the SAM level (Fig. 4e) while almost all TCA cycle metabolites, except alpha-ketoglutaric acid, remained unchanged in ESCs after CCCP treatment (Supplementary Fig. 8b and Supplementary Data 6). The CCCP treatment led to elevated expression of 2C gene signature, MERVL and the trophoblast master regulators (e.g., Cdx2, Tfap2c, Eomes) (Fig. 1e, f and Supplementary Fig. 8c). We also observed global reduction of H3K27me3, which was fully restored by SAM supplementation (Fig. 4f). H3K9me3 and DNA methylation marks were not affected by the CCCP treatment (Supplementary Fig. 8d, e).

We next examined whether blocking SAM biosynthesis in naive ESCs affects H3K27me3 and de-repression of 2C gene signature. The methionine adenosyltransferase (MAT) family proteins, consisting of three isoforms MAT1A, MAT2A and MAT2B⁴⁵, are the rate-limiting enzymes that catalyze SAM synthesis from methionine and ATP in mammals^45,46. Among them, Mat2a was expressed at the highest level in naive ESCs (Fig. 5a). Knockdown of Mat2a by siRNA in naive ESCs resulted in global reduction of H3K27me3 levels (Fig. 5b, c). Notably, Mat2a knockdown significantly upregulated the expression of MERVL and 2C genes, as revealed by RNA-seq analysis and validated by RT-qPCR (Fig. 5d, e and Supplementary Data 7). Similarly, MERVL-2C::EGFP naive ESCs had over 20-fold increase in GFP signals after Mat2a knockdown (Fig. 5f). Additionally, the MERVL⁺ cells exhibited a loss of chromocenters, indicated by the absence of DAPI-dense regions, and reduced levels of OCT4 (Fig. 5g). RNA-seq analyses revealed a close correlation between the transcriptome of Mat2a-depleted naive ESCs and that of spontaneous qESCs and early-stage blastomeres (Fig. 5h), distinct from qESCs with SAM supplementation and conventional ESCs (Fig. 5h). These findings underscore the importance of low MMP and SAM homeostasis in regulating the fate plasticity of qESCs.

**Fig. 5: Knockdown of *Mat2a* induces 2C-like features in naive ESCs.**

Inducing quiescence in naive ESCs is sufficient to confer unrestricted cell fate

To examine whether inducing quiescence is sufficient to confer unrestricted cell fate, we performed RNA-seq analysis for the qESCs induced by FAO⁵ or MYC⁶ inhibitor (referred to as FAOi or MYCi herein). Hierarchical clustering analysis showed that the transcriptome of FAOi and MYCi-induced qESCs was closely clustered with that of early stage blastomeres (Fig. 5h), akin to the clustering pattern observed in the spontaneous qESCs and Mat2a-depleted naive ESCs. RT-qPCR analysis confirmed a significant increase in MERVL expression in the FAOi and MYCi-induced qESCs (Fig. 6a). In contrast, mTOR inhibition-mediated ESC growth arrest⁷ (Supplementary Fig. 9a, b), which does not induce quiescence, did not upregulate the MERVL expression (Fig. 6a). Both FAOi- or MYCi-induced qESCs exhibited efficient differentiation into TSCs, yielding a high percentage of CDX2⁺/GATA6⁻ cells by day 7 post-induction (Fig. 6b, c).

**Fig. 6: Induction of quiescence in naive ESCs is sufficient to unlock the unrestricted developmental potential via downregulating H3K27me3 and metabolic modulation.**

The untargeted metabolomics analyses conducted on the spontaneous and induced qESCs showed that FAOi or MYCi-induced qESCs had significantly lower levels of TCA metabolites, such as alpha-ketoglutaric acid, as well as metabolites within the purine and pyrimidine metabolism pathways (Fig. 6d, left, Supplementary Figs. 5c,9c, d and Supplementary Data 6), mirroring the metabolic profiles observed in spontaneous qESCs. However, FAOi or MYCi-induced qESCs also displayed significant upregulation in glycerolipid, pentose phosphate and glutathione metabolism pathways (Fig. 6d, right and Supplementary Data 6). Importantly, MYCi and to a lesser degree, FAOi, led to a decrease in SAM levels (Fig. 6e), downregulation of the PRC2 subunits (Fig. 6f) and reduction of global H3K27me3 (Fig. 6g). Together, these results suggest that inducing quiescence in naive ESCs results in a convergent change on H3K27me3.

Genetic deletion of Eed promotes unrestricted ESC fate

Given the observed reduction of H3K27me3 in both spontaneous and induced qESCs, we investigated whether the genetic deletion of Eed (embryonic ectoderm development), the core subunit of the PRC2 complex⁴⁷, could remove the developmental constraints in naive ESCs. To this end, we derived naive ESCs from the Eed⁻^/− (null) and Eed^fl/− (heterozygous) mouse embryos and maintained them continuously in 2i/LIF medium (Fig. 7a, b and Supplementary Fig. 10a, b, see method). The Eed^−/− naive ESCs exhibited global loss of H3K27me3 (Fig. 7b). RNA-seq analysis identified 2153 upregulated and 1166 downregulated genes ([FDR] < 0.05) in the Eed^−/− naive ESCs (Fig. 7c). Similar to spontaneous qESCs and Mat2a-depleted naive ESCs, the Eed^−/− naive ESCs had elevated expression of MERVL and 2C genes (such as Zscan4, Tcstv3 Gm2022) (Fig. 7c, d and Supplementary Data 7). Genes from both embryonic and extraembryonic differentiation pathways were upregulated in the Eed^−/− naive ESCs (Fig. 7e–g). Strikingly, the Eed^−/− naive ESCs had a significant increase in expression of trophectoderm gene markers (e.g., Cdx2, Elf5, Eomes), even without TSC induction (Fig. 7a, b, e). These results imply that the Eed^−/− naive ESCs are different from the spontaneous qESCs, which had minimal changes in expression of ectoderm, mesoderm, and endoderm genes (Fig. 7g).

**Fig. 7: Deletion of *Eed* promotes unrestricted fate potential in naive ESCs.**

Quiescence endows a dynamic reservoir of naive ESCs with the potential for unrestricted cell fate

Unlike adult stem cells, which predominantly remain quiescent within their niche in vivo, naive ESCs continuously proliferate in vitro. We postulated that quiescence enables naive ESCs to sustain a dynamic reservoir of cells with unrestricted cell fate. Mathematical modeling through simulation supported the notion that stochastic mitochondrial flux in daughter cells during cell division can sustain metabolic heterogeneity in naive ESCs (Supplementary Fig. 11a, b). In line with the modeling predictions, isolated single spontaneous qESCs were capable of generating ESC lines with similar heterogeneity in MMP as the parental cell line (Fig. 8a). Importantly, the low MMP qESCs derived from each newly established naive ESC line displayed unrestricted cell fate. They expressed high levels of MERVL and 2C genes (Zscan4, Dux, Tcstv3) (Fig. 8b, left) as well as pluripotency markers Oct4, Nanog, and Prdm14 (Fig. 8b, right). These cells efficiently differentiated into TSCs, as evidenced by a high percentage of CDX2⁺/OCT4⁻ cells at day 7 post-induction (Fig. 8c, d). These results suggest that quiescence enables maintenance of the intrinsic unrestricted cell fate in naive ESCs.

**Fig. 8: Quiescence preserves innate unrestricted cell fate in naive ESCs.**

Discussion

Quiescent stem cells are widely considered to reside in a “dormant” or “idle” state^1,2. Here we report that spontaneous or induced quiescent ESCs represent a dynamic reservoir of cells endowed with unrestricted differentiation potential. These qESCs exhibit convergent alterations in mitochondrial activity, decreased levels of SAM, and global reduction of H3K27me3. These metabolic and epigenetic alterations accompany the drastic transcriptomic and functional changes in qESCs, including de-repression of the 2C and extraembryonic gene programs, and expanded differentiation potential towards the extraembryonic lineages. Our results strongly argue that stem cell quiescence transcends a mere dormant or inactive state. Instead, the qESCs actively acquire new distinctive metabolic and epigenetic states, enabling unrestricted cell fate determination. Our findings underscore a previously overlooked paradigm wherein quiescence serves as an active strategy employed by the stem cells to maximize cell fate plasticity.

We show that qESCs have convergent changes in PRC2-mediated H3K27me3 along with removal of developmental constraints. Unlike serum/LIF ESCs^40,48, naive mouse ESCs are devoid of DNA methylation and mostly rely on H3K27me3 to repress lineage-specific gene expression^41,49. Our study extends the function of PRC2 to direct regulation of endogenous retroviruses, 2C and TSC genes (e.g., Cdx2, Elf5, Eomes). Our results contrast to a recent report⁴¹ indicating that Eed deletion in mouse ESCs leads to minimal changes in the expression of TSC genes (Supplementary Data 7). The discrepancy might stem from the method employed in the prior study^41,50, where Eed deletion occurred in serum/LIF ESCs prior to their transfer to 2i/LIF medium. Given that inhibition of PRC2 in naive human ESCs (hESCs) sporadically induces the trophoblast marker GATA3 in about 10% of cells^51,52, it is plausible that PRC2 plays a conserved role in suppressing extraembryonic cell fate. Future investigations are warranted to explore whether quiescence in hESCs⁵ can expand cell fate and if so, whether it is through modulating PRC2-mediated H3K27me3.

Our study suggests importance of SAM homeostasis in naive ESCs. Cellular metabolites such as SAM, acetyl-CoA and α-ketoglutarate are substrates or cofactors for DNA- and histone-modifying enzymes and may contribute to regulation of the epigenetic landscapes in cells^16,53. We find that both spontaneous and induced qESCs have lower levels of intracellular SAM. Acute reduction of MMP by CCCP also lead to reduced SAM levels in naive ESCs. Previous studies have indicated that the mitochondrial polarity is essential for transporting charged molecules in and out of mitochondria¹⁷. How MMP modulates SAM homeostasis in naive ESCs remains to be determined. Nonetheless, we find that blocking SAM biosynthesis by depletion of Mat2a is sufficient to induce the unrestricted cell fate. Conversely, SAM supplementation in low MMP ESCs can restrict extraembryonic differentiation. Collectively, these results suggest a potential connection between MMP state and SAM biosynthesis in regulating developmental potential of naive ESCs, which awaits further investigations.

One surprising finding is that low levels of SAM in qESCs correspond to a prominent reduction of global H3K27me3, along with minimal changes in H3K4me3, H3K9me3 or DNA methylation. The mechanism for such specific, rather than general, changes in chromatin modifications remains unclear. Specific regulation of H3K27me3 by intracellular SAM has been reported in naive hESCs⁵⁴. The nonlinear correlation between the cellular SAM level and H3K27me3 could be influenced by various factors, including the overall abundance of enzymes and their antagonists (e.g., demethylases), as well as adaptive epigenetic mechanisms in response to SAM depletion⁵⁵. These aspects merit further investigations in future studies. In contrast to H3K27me3, which emerges as a convergent change in both spontaneous and induced qESCs, the metabolic alterations in these cell types are largely distinct. Induced qESCs exhibit significantly elevated metabolites associated with pyrimidine metabolism and pentose phosphate pathway (Fig. 6d, right and Supplementary Data 6) that were not observed in the spontaneous qESCs. The functional significance of these pathways in cell fate determination awaits further studies.

Finally, our study shows that ESC quiescence can serve as a valuable platform for comprehensive metabolomics and epigenomic characterizations associated with cell fate alteration, which are somewhat limited for the in vivo quiescent adult or cancer stem cells. Bona fide quiescent ESCs can be isolated by FACS after TMRM live stain or stably induced using a single chemical agent (i.e., FAOi or MYCi), offering a more accessible and well-controlled experimental setting. Leveraging qESCs enables the establishment of fundamental principles related to cell metabolism, cell cycle regulation, and epigenetic modifications in stem cell fate determination.

Methods

Ethical regulation compliance

The 3–8-week-old female and 2–6-month-old male mice were used for all experiments. Mice were generated and maintained on the 129/S1 background on a 12 h light/dark cycle in the Transgenic Animal Model Core, University of Michigan. All animal procedures were conducted according to the protocols approved by the University Committee on Use and Care of Animals at the University of Michigan (protocol #PRO00004007 and PRO00006455).

Cell lines and cell culture conditions

Mouse ES cell lines (ESCs) used in this study include: E14tg2a (E14) (ATCC, #CRL-1821TM), Oct4-GiP ESCs²², male V6.5 ESCs (Novus Biologicals, #NBP1-41162) and iCdx2 Elf5-2A-mCherry reporter ESCs³⁴. All ESCs were routinely cultured on tissue culture plates coated with Geltrex (Thermo Fisher Scientific, #A1413202, 1:100 dilution from stock) at 37 ^oC and 5% CO₂. Naive ESCs were maintained under 2i/LIF culture condition, which contains N2B27 basal medium supplemented with PD0325901 (SIGMA ALDRICH, #PZ0162, 1 µM), CHIR99021 (SIGMA ALDRICH, #SML1046, 3 µM) and LIF (Millipore, #ESG1106, 1000 U/ml). N2B27 consists of a 1:1 mix of DMEM/F12 (Thermo Fisher Scientific, #21331-046) and Neurobasal (Thermo Fisher Scientific, #21103-049), N2 (Thermo Fisher Scientific, #17502001, 1:100), B27 (Thermo Fisher Scientific, #17504-044, 1:100), 2 mM Glutamine (Thermo Fisher Scientific, #25030024), Penicillin-Streptomycin (Thermo Fisher Scientific, #15140122, 1:100) and 0.055 mM 2-mercaptoethanol (Thermo Fisher Scientific, #21985023). Cells were routinely plated at the density of 1 × 10⁵ cells per well on the six-well tissue culture plates. Cells were passaged every 2 days by dissociation with StemPro Accutase (Gibco, #A1110501). Where indicated, ESCs were grown in serum/LIF culture condition containing the KnockOut™ DMEM medium (Thermo Fisher Scientific, #10829018) supplemented with 15% fetal bovine serum (FBS) (SIGMA ALDRICH, #F0926), 2 mM Glutamine, 1X Non-essential amino acids (Thermo Fisher Scientific, #11140050), 0.055 mM 2-mercaptoethanol, Penicillin-Streptomycin (1:100) and LIF (1000 U/ml).

Mouse embryonic fibroblasts (MEFs) were isolated from E13.5 embryos of FVB triple transgenic mice at the University of Michigan Transgenic Animal Model Core. Mitomycin-C-treated MEFs were maintained in KnockOut™ DMEM medium containing 15% FBS, 2 mM Glutamine, 1X Non-essential amino acids, Penicillin-Streptomycin (1:100) and 0.055 mM 2-mercaptoethanol. MEFs were prepared at the appropriate density (see below) and used within one week after plating.

TSCs were derived directly from Ẻ.3.5 blastocyst as previously described⁵⁶. TSCs were routinely maintained on top of MEFs (30,000 cells per cm²) at 37 ^oC and 5% CO₂ in the TSC medium containing 70% MEF-conditioned media (R&D systems, #AR005), 30% base medium (RPMI 1640, Thermo Fisher Scientific, #11875093; sodium pyruvate, Thermo Fisher Scientific, #11360070, 1:100; Penicillin-Streptomycin, 1:100; 0.1 mM 2-mercaptoethanol and 20% FBS), 25 ng/ml FGF4 (Reprotech, #100-31) and 1 µg/ml Heparin (SIGMA ALDRICH, #H3149-10KU). Medium change was performed daily, and cells were passaged by dissociation with Trypsin-EDTA 0.05% (Thermo Fisher Scientific, #25300054).

Derivation of Eed-null naive ESCs

Mouse strain JF1 carries a conditional mutation of Eed gene in which its exon 7 was flanked by loxP sequences. The Eed^fl/fl mice were generated and maintained on the 129/S1 background as described⁴⁷. E3.5 embryos were obtained from a cross between Eed heterozygous females and Eed heterozygous males carrying Protamine CRE recombinase. Each E3.5 blastocyst was placed on top of MEF feeder in 2i/LIF culture condition containing the KnockOut™ DMEM base medium, 15% KnockOut™ Serum Replacement (Thermo Fisher #10828028), 5% FBS, 2 mM Glutamine, 1X Non-essential amino acids, 0.055 mM 2-mercaptoethanol and Penicillin-Streptomycin (1:100), PD0325901 (1 µM), CHIR99021 (3 µM) and LIF. At 4–5 days post plating, each inner cell mass was picked, dissociated into single cells and expanded on MEF feeders in the 2i/LIF ESC medium.

Derivation of ESC lines from single low ΔΨ_m qESC

Each single naive ESC with bottom 5% TMRM fluorescence intensity was FACS-sorted directly onto each well of 96-well tissue culture plate. After 5–7 days, each colony was dissociated into single cell by StemPro Accutase, resuspended into 2i/LIF medium and transferred to 48-well tissue culture plates. Cell lines were routinely cultured in 2i/LIF medium.

Generation of MERVL-2C::EGFP naive ESC lines

About 3 × 10⁴ E14 naive ESCs were seeded on 24-well plates 24 h before transfection. Cells were transfected with the 2C-EGFP plasmid (addgene, #69071)²³ for 2 h by Lipofectamine^TM 2000 (Thermo Fisher Scientific, #11668027). After transfection, fresh medium was added overnight. Transfected cells were dissociated and replated at clonal density in 6-well tissue culture plates. Selection was performed with 250 µg/ml G418 (Thermo Fisher Scientific, #10131035) for one week before individual colonies were manually picked and expanded in 2i/LIF condition.

Generation of mito-Grx1-roGFP2 naive ESC lines and live-cell imaging

The pPBmitoCOX8-Grx1-roGFP2 was generated by PCR amplification of mitoCOX8-Grx1-roGFP2 cassette using the Forward-GGAAGATCTATGTCCGTCCTGACGCCGC; Reverse-CGACGCGTTAACGTTCGAGGTCGACTC primers. PCR products were cloned into BglII/MluI sites in pPBhCMV1cHApA. E14 naive ESCs (2i/LIF) were transfected with 0.5 µg of pPBCAG-rtTAM2-IN, 0.5 µg of pPBmitoCOX8-Grx1-roGFP2 and 1 µg pCAG-hyPBase transposase by Lipofectamine^TM 2000. Individual colonies were manually picked and expanded under 2i/LIF condition after one-week selection with 250 µg/ml G418. Expression of mitoCOX8-Grx1-roGFP2 in the mitochondria (i.e., mito-Grx1-roGFP2 ESCs) was induced by Dox (1 µg/ml) and analyzed on a Nikon X1 Yokogawa Spinning Disk Confocal (Nikon Instruments Inc.). The 60× objective (water immersion) was used for live-cell imaging. The mito-Grx1-roGFP2 sensor was excited by the 405 nm and 488 nm laser line and detected the single emission (~520) through 500–554 nm bandpass filter.

Generation of induced TSC lines and in vitro differentiation

The naive ESC line with the iCdx2 Elf5-2A-mCherry reporter³⁴ was used to generate induced TSC lines (iTSC). About 400 FACS-sorted cells were plated directly onto MEF feeder (10,000 cells per cm²) in TSC medium. At day 7, mCherry+ cells were FACS-sorted and plated at clonal density for 7 days before individual TSC-like colonies were picked and expanded. The iTSC lines were routinely cultured at high density on MEF feeders (30,000 cells per cm²). To induce differentiation, the iTSCs (after MEF removal) were cultured on gelatin-coated dishes in the absence of FGF4 and Heparin for 6 days before analyses.

CCCP and phenformin treatment of naive ESCs

To inhibit MMP, carbonyl cyanide 3-chlorophenylhydrazone (CCCP, SIGMA ALDRICH, #C2759, 3 µM) was added 24 h after cell plating. All analyses were performed 9 h after CCCP treatment. To inhibit complex I electron transport chain, naive ESCs were treated with phenformin (Selleck Chemical, #S2542, 1 mM) for 2 days. For S-(5′-Adenosyl)-L-methionine chloride dihydrochloride (SIGMA ALDRICH, #A7007, 0.5 mM) supplement, naive ESCs were plated on Lab-Tek II Chamber Coverglass (Thermo Fisher Scientific, #155382). At 24 h post plating, cells were pre-treated with 0.5 mM of SAM or vehicle for 2 h before treatment with CCCP for 6 h. SAM at a final concetration of 25 or 50 µM was used for culturing the spontaneous qESCs for 24 h.

Induction of quiescence

Naive ESCs were induced to achieve a stable quiescent state by inhibition FAO or MYC (FAOi or MYCi). To inhibit FAO, Etomoxir (SIGMA ALDRICH, #E1905, 100 µM) was added at the time of plating for 4 days⁵. 10058-F4 (SIGMA ALDRICH, #475956) was used at a final concentration of 64 µM for 3 days to inhibit MYC⁶. As a control, naive ESCs were treated with mTOR inhibitor (mTORi) INK128 (Medchem Express, HY13328, 200 nM) for 3 days as described⁷. H₂O and DMSO were added at the same final volume as FAOi, MYCi or mTORi. The induced qESCs and control cells were then harvested for RT-qPCR, metabolomics analysis and TSC differentiation.

Mitochondrial membrane potential (ΔΨ_m) staining and FACS sorting

Fractionation of ESCs with high and low ΔΨ_m levels was performed as previously described^5,15. 25 nM of TMRM (Thermo Fisher Scientific, #T668), was used for staining. To visualize ΔΨ_m profile, naive ESCs were stained with 25 nM of TMRM directly on the tissue culture plate at 37 ^oC, 5% CO₂ for 15 min. Nuclei was stained with Hoechst 33342 (Thermo Fisher Scientific, #H3570, 1:2000). Live-cell imaging was performed using a Nikon X1 Yokogawa Spinning Disk Confocal (Nikon Instruments Inc.) (see below).

Embryoid body (EB) differentiation

The EB medium consists of the KnockOut™ DMEM medium, 15% FBS, 2 mM Glutamine, 1X Non-essential amino acids, 0.055 mM 2-mercaptoethanol and Penicillin-Streptomycin (1:100). 2000 FACS-sorted ESCs with high and low ΔΨ_m were resuspended into 200 µl EB medium and plated onto each well of the Nunclon^TM Sphera^TM 96U plate (day 0). EBs were cultured in suspension at 37 ^oC and 5% CO₂. Half of medium were replaced on day 1, followed by full medium change every other day from day 2. EBs in suspension at day 3 and 6 were collected for RT-qPCR analysis. At day 7, EBs were transferred to 24-well tissue culture plates coated with Gelatin for additional 7 days. Differentiating cells were then fixed by 4% PFA/PBS for immunostaining to detect ectoderm (TUJ1), mesoderm (SMA) and endoderm (GATA6) markers as described below.

Mathematical simulation modeling for mitochondrial activity

Flow cytometry data of TMRM staining for bulk naive ESCs were used for curve fitting. The distribution of TMRM fluorescence intensity exhibited multimodal features that are captured by a combination of n gaussian distributions using the R package mixtools⁵⁷. The probability of high ΔΨ_m and low ΔΨ_m ESCs were obtained using the Gaussian model. To simulate the dynamics of mitochondrial activity through cell divisions, we constructed a mathematical model based on following assumptions: 1) the mitochondrial activity for the parental cells was generated from a normal distribution with the same mean and standard deviation from the bulk ESCs; 2) the range of mitochondrial activities for the progenies was calculated using the following formula, where a is the activity for each daughter: a = (max(1, a*runif(1,m,1)), a, a*runif(1,1,n)); 3) the activity for each daughter cell had three possible outcomes (I, J, K):

I: the activity was decreased by multiplying a by a value between m and 1 and the lowest value was set to be 1; runif(1, m, 1) generated a value from a uniform distribution between m and 1;
J: the activity remained at a;
K: the activity was increased by multiplying a by a value between 1 and n; runif(1, 1, n) generated a value from a uniform distribution between 1 and n.

Using this model, the experimental data were applied to estimate the activity fluctuation coefficient, t. Most low ΔΨ_m daughter cells had increased (Pk = 0.8) activity and these cells regained the mitochondrial heterogeneity within 2–3 cell divisions. By comparing the difference in absolute value between the mean activity of daughter and parental cell populations, the activity fluctuation coefficient was estimated as 1.25, which means daughter cells regain similar parental ΔΨ_m distribution within 3 cell divisions if mitochondrial activity alters ~25% per division.

Gene knockdown by siRNAs

To ensure the complete knockdown, we used the Dharmacon ON-TARGETplus siRNA SMARTPool that contains a mixture of four different siRNAs to target all transcript/splice variants of Mat2a. siRNA transfections were performed with the DharmaFECT 1 Transfection Reagent (Dharmacon) using reverse transfection protocol according to manufacturer’s instructions. Briefly, 2 × 10⁴ naive ESCs per 24-well culture area were transfected with either Dharmacon ON-TARGETplus Mouse Mat2a siRNA SMARTPool or ON-TARGETplus Non-targeting Control Pool at a final concentration of 40 nM. 18 h post transfection, medium change was performed, and cells were maintained for another 2 days prior to sample collection for subsequent analyses.

Cell proliferation and cell cycle analysis

About 5000 FACS-sorted ESCs were replated onto Geltrex-coated 24-well tissue culture plates. Cells were dissociated into single cell by StemPro Accutase, stained with Trypan Blue 0.4% (Thermo Fisher Scientific, #15250061) and counted by hemocytometer every day for 3 consecutive days. For cell cycle analysis, cells were stained with 5 µg of Hoeschst 33342 (Thermo Fisher Scientific, #62249) and 1 µg of Pyronin Y (SIGMA ALDRICH, #213519) as previously described⁵. Stained cells were centrifuged, resuspended in ice-cold staining solution and subjected to flow cytometry analysis using the Propel Bigfoot Cell Sorter. Data were further analyzed by the FlowJo 10.8.1 (FlowJo LLC).

Elf5 promoter methylation analysis

Genomic DNA was extracted and 0.5 µg of genomic DNA per sample was subjected to bisulfite conversion using the Methylation-Gold^TM Kit (ZYMO RESEARCH, #D5005). About 4 µl cDNA was PCR amplified with primers³⁵: Forward-TTTGTAGTTTGAGTATTTTGGTG; Reverse-ACCTTTCCACTCTAAACACCCAAA. Next-generation sequencing was performed at Center for Computational and Integrative Biology (CCIB) DNA Core at Massachusetts General Hospital. The raw paired-end bisulfite sequencing reads were processed using Trim Galore. The refined reads were aligned to the mouse genome (mm10) using Bismark⁵⁸ with default settings. To determine the methylation levels of individual CpG sites within the regions of interest, the “bismark_methylation_extractor” function was employed to calculate the percentage of methylated CpG over the total calls for each specific site. A robust coverage exceeding 100× was attained for each examined CpG sites. The methylation levels were visualized using a violin plot in the ggplot2 package and a nonparametric Mann-Whitney test was employed for statistical analysis.

Immunofluorescence

Cells were fixed with 4% paraformaldehyde (PFA, Ted Pella Inc. #18505)/PBS for 15 min at room temperature for histone antibodies or 30 min at 4 ^oC for other antibodies. Cells were then washed three times with PBS, permeabilized in 0.25% Triton X-100/PBS (PBST) for 30 min at 4 ^oC and blocked in blocking solution (10% Goat serum (Cell Signaling Technology, #5425) or 10% Donkey serum (SIGMA ALDRICH, #D9663), 0.1% BSA and 0.01% Tween20 in PBS) for 1 h at 4 ^oC. Primary antibody incubation was performed at 4 ^oC overnight. For detection of 5 mC, fixed cells were permeabilized with 0.4% TritonX-100/PBS for 30 min at 4 ^oC, denatured with 2 N HCl for 30 min at room temperature, and neutralized with 100 mM Tris-HCl (pH 8.0) for 10 min at room temperature. After incubation with primary antibodies, cells were washed with 0.1% PBST three times prior to secondary antibody incubation at room temperature for 1 h, followed by 0.1% PBST wash for three times. Nuclei were stained with DAPI (Thermo Fisher Scientific, #62248, 1:1000) diluted in 0.1% PBST at room temperature for 10 min. Images were captured by the Leica Stellaris 5 Inverted confocal microscope (Leica Microsystems) with a 40× dry, 60× oil or 100× oil objectives across 0.5 µm stacks. Fluorescence was excited with a 405 nm UV laser (DAPI), a 488 nm laser (Alexa Fluor488) and a 561 nm laser (Alexa Fluor555). Images were also captured by the IX73 microscope system (Olympus). Image data were further processed by either Leica software or ImageJ software⁵⁹.

Antibodies for IF

The following primary antibodies were used: Mouse anti-OCT3/4 (Santa Cruz Biotechnology, Clone C-10, #sc-5279, 1:50), Rabbit anti-CDX2 (BioGenex, Clone EP25, #NU777-5UC, 1:100), Mouse anti-CDX2 (BioGenex, Clone CDX2-88, #MU392A-5UC, 1:100), Rabbit anti-EOMES (Abcam, #ab23345, 1:100), Mouse anti-SMA (SIGMA ALDRICH, Clone 1A4, #A2547, 1:200), Mouse anti-TUJ1 (Biolegend, Clone TUJ1, #801201, 1:200), Goat anti-GATA6 (R&D Systems, #AF1700, 1:100), Mouse anti-5mC (Millipore, Clone 33D3, #MABE146, 1:200), Rabbit anti-H3K27me3 (Millipore, #07-449, 1:100) and Rabbit anti-H3K9me3 (Abcam, #ab8898, 1:100). The secondary antibodies used were: Goat anti-Mouse IgG FITC-conjugated (Thermo Fisher Scientific, #62-6511, 1:50), Goat anti-Mouse IgG Alexa Fluor^R 555 (Abcam, #ab150114, 1:1000), Donkey anti-Goat IgG Alexa Fluor^TM 488 (Thermo Fisher Scientific, #A11055, 1:1000), Donkey anti-Mouse IgG Alexa Fluor^TM 555 (Thermo Fisher Scientific, #A31570, 1:1000), Donkey anti-Rabbit IgG Alexa Fluor^TM 555 (Thermo Fisher Scientific, #A31572, 1:1000), and Donkey anti-Rabbit IgG Alexa Fluor^TM 488 (Thermo Fisher Scientific, #A21206, 1:1000).

Quantitative reverse transcription PCR (RT-qPCR) analysis

Total RNAs were extracted by the RNeasy mini kit. cDNA was synthesized using the SuperScriptTM III First-Strand Synthesis System with oligoT primer (Thermo Fisher Scientific, #18080051). qPCR was performed with the Radiant Green Lo-ROX qPCR Kit (Alkali Scientific Inc., #QS1020) on the 7500 Real Time PCR System (Applied Biosystems). The comparative cycling threshold (ΔΔCt) method was used to analyze the qPCR data. List of primers was shown in the Supplementary Data 8.

Total RNA sequencing

Total RNAs were extracted by the RNeasy mini kit (QIAGEN). Genomic DNA was removed on-column by RNA-free DNase I for 15 min at room temperature. Total RNA amounts were determined by NANODROP 2000/2000c (Thermo Fisher Scientific) and assessed for quality using the TapeStation (Agilent, #5067-5576). Samples were prepared using the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (NEB, #E7760L), NEBNext rRNA Depletion Dit (Human/Mouse/Rat) (NEB, #E6310X), NEBNext Multiplex Oligos for Illumina (NEB, #E6440L) and ERCC RNA Spike-In control Mix (Thermo Fisher Scientific, #4456740). About 400 ng of total RNA was ribosomal depleted and fragmented and reverse transcribed. Samples underwent end repair and dA-Tailing step followed by ligation of NEBNext adapters. The libraries were pooled and sequenced on the Illumina NovaSeq 6000 with paired-end 150 bp at the University of Michigan core facility or AZENTA.

Untargeted metabolomics analysis

The untargeted metabolomics analyses were performed as previously described⁶⁰ with modifications. All samples were subjected to FACS sorting to isolate 5 × 10⁵ live cells. They were lysed using 1 ml of 80% cold methanol. Since the qESCs have lower biomass than cycling ESCs^5,6, cell numbers were used to normalize metabolite fractions across samples before metabolite extraction. The extracts were incubated at −80 °C for 10 min and centrifugated at 14,000 rpm for 10 min at 4 °C. The supernatants were speed dried and re-suspended in 35 μl 50:50 MeOH:H2O mixture for LC-MS analysis. LC-MS analyses were performed using an Agilent Technologies Triple Quad 6470 LC-MS/MS system with a 1290 Infinity II LC Flexible Pump (Quaternary Pump), 1290 Infinity II Multisampler, 1290 Infinity II Multicolumn Thermostat with 6 port valve and 6470 triple quad mass spectrometers in both positive and negative acquisition mode as previously described^5,60. The Agilent MassHunter Workstation Software was used for data acquisition. Agilent MassHunter Quantitative Analysis for QQQ (Version B.10.1.733.0) was used for initial raw data extraction. Results were then exported as CVS file. Metabolite counts were normalized by the total intensity of all metabolites to scale equal sample loading⁵. Finally, abundance of each metabolite in the sample was normalized divided by the median abundance levels across all samples for comparisons, statistical analyses and visualizations among metabolites. The statistical analysis was performed by the two-tailed Student’s t test with a significant threshold level of 0.05. Pathway enrichment analysis was performed using the MetaboAnalyst 5.0 (https://www.metaboanalyst.ca)⁶¹.

Cleavage under targets and release using nuclease (CUT&RUN)

CUT&RUN analysis was performed with approximately 2 × 10⁵ naive ESCs for each library. Bead-bound cells were incubated overnight at 4 ^oC with Rabbit anti-H3K4me3 (Millipore, #07-473) or Rabbit anti-H3K27me3 (Millipore, #07-449) antibody diluted in the antibody buffer (EDTA/Digitonin/wash buffer, EMD Millipore, #300410) (1 µg/500 µl). The beads were washed three times and incubated with 200 µl pA-MNase (0.7 ng/µl) for 1 h at 4 ^oC. 2 mM of CaCl₂ was added to activate MNase for 30 min on ice. The reaction was quenched by 2× stop buffer containing 340 mM NaCl, 20 mM EDTA, 4 mM EGTA, 0.02% Digitonin, 50 µg/ml RNase A (QIAGEN, #19101), 50 µg/ml glycogen (Roche, #10901393001) and 2 pg/ml Drosophila spike-in DNA at 37 ^oC for 10 min. The soluble chromatins were converted to libraries using the NEBNext Ultra^TM II DNA library Prep kit (NEB, #E7645L). Libraries were subject to pair-ed sequencing on the Illumina NovaSeq 6000 instrument at the University of Michigan core facility.

Assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq)

Approximately 2 × 10⁵ naive ESCs were used for ATAC-seq analysis. Briefly, FACS-sorted cells were washed in cold D-PBS and resuspended in 50 µl of cold lysis buffer (10 mM Tris-HCl pH 7.4, 10 mM NaCl, 3 mM MgCl₂, and 0.1% IGEPAL CA-630). Samples were spun at 500 g for 10 min, 4 ^oC. Each cell pellet was resuspended in 50 µl of transposition reaction mix containing 25 µl of 2X TD buffer (Illumina, #FC-121-1030), 2.5 µl Tn5 transposase (Illumina, #FC-121-1030) and 22.5 µl nuclear free H₂O. Tagmentation by Tn5 was performed at 37 ^oC, 30 min. Libraries were sequenced on the Illumina NovaSeq 6000 instrument with paired end reads at the University of Michigan core facility.

Bioinformatic analysis

RNA-seq. To perform data cleaning, the cutadapt (https://cutadapt.readthedocs.org/en/stable/) and FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) tools wrapped in Trim Galore (http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) were used to trim low quality bases (Q < 20) and adapter for raw sequences. RNA-seq data were mapped to the mm10 (GRCm38) genome of mouse and ERCC reference using Tophat2 (2.1.1)⁶². Duplicated reads for pair-end data were removed by SAMtools (v1.9)⁶³. The bigwig files for visualization in Integrative Genomics Viewer (IGV)⁶⁴ were generated from BAM files by using “bamCoverage” from deepTools3 (v3.2.1)⁶⁵ with parameters “--outFileFormat bigwig --ignoreForNormalization chrM chrX chrY --skipNAs --bam input.accepted_hits.bam --outFileName output.bigwig --normalizeUsing RPKM --minMappingQuality 30 --binSize 20 --smoothLength 60 --ignoreDuplicates”.

Annotation of mRNA (GRCm38 v99) was downloaded from Ensembl database⁶⁶. LncRNA annotation was downloaded from NONCODE database⁶⁷ and repeat annotation was downloaded from Homer (v4.10)⁶⁸. All three annotations were merged for gene annotation. Fragment (or template) count for each feature was performed by using featureCounts with parameters “-g ‘gene_id’ -p -M -O --fraction -C -F GTF -T 64 -s 0”⁶⁹. Each reported alignment from a multi-mapping read (identified via “NH” tag) carried a fractional count of 1/x, instead of 1 (one), where x is the total number of alignments reported for the same read. Each overlapping feature received a fractional count of 1/y, where y is the total number of features overlapping with the read. Briefly, each alignment carried a fractional count of 1/(x*y). Fragment counts were transformed to fragments per kilobase of transcript per million (FPKM) by using edgeR⁷⁰. Genes (or features) have more than three fragment counts in all samples were included. We employed RUVSeq⁷¹ to perform differential expression analysis by using ERCC as control genes with a likelihood ratio test from DESeq2⁷². Genes (or features) with the adjusted P value < 0.05 were reported. The following public datasets were used: GSM2711863 (ESC_2i)⁴¹, GSM2711865 (ESC_2i_EedKO)⁴¹, GSM838738 (MERVL-tdtomato⁺ and MERVL-tdtomato^-)¹⁹, GSM1966767 (MERVL⁺_Zscan4⁺ and MERVL⁺_Zscan4⁺)²⁷, GSM3384433 (CRISPRa_EV and CRISPRa_MERVL)²⁹, GSM3110917 (NELFA^high and NELFA^low)³⁰, GSM1415501 (Zscan4^high and Zscan4^low)²⁶, GSM2279983 (Dux-GFP⁺, Dux-GFP^-, Plus Dox-Dux, and Minus Dox-Dux)²⁸, and GSE66582 (distinct developmental stages of preimplantation embryos)²⁵.

To facilitate the comparisons between our and public RNA-seq data, we followed the “batch mean-centering” approach for batch effect removal⁷³. Briefly, we have separately mean-centered the log₂(FPKM + 1) value of each gene by subtracting the mean log₂(FPKM + 1) across all samples. We used “1 - Spearman” correlation coefficient as the distance in the hierarchical clustering⁷. We computed the dissimilarity values with “as.dist” function in R, which were fed into hclust with “average” algorithm for agglomeration. The dendrogram were plotted by R.

CUT&RUN and ATAC-seq

Paired-end raw sequencing reads were trimmed with Trim Galore⁷⁴ to remove adaptors and low-quality reads. Processed read (>20 bp) were mapped against mouse mm10 reference genome using bowtie2 (v2.4.2)⁷⁵ with the following parameters “--local --very-sensitive-local --no-unal --no-mixed --no-discordant -q --phred33 -I 10 -X 700 --threads 12”. PCR duplicated reads were removed using samtools (v1.9)⁶³. For H3K4me3, filtered BAM files for each sample were submitted to the callpeak function of MACS2 (v2.2.7.1)⁷⁶ for peak regions calling with parameters “-p 1e-5 -g mm”. For H3K27me3 domains, peak calling was performed using SICER (v1.1)⁷⁷ with default parameters for calling broad peaks. ATAC-seq paired-end raw sequencing reads were trimmed with Trim Galore⁷⁴, then aligned to reference genome (mouse mm10) using Bowtie2 (v2.4.2)⁷⁵ with parameters: “--local --very-sensitive-local --no-unal --no-mixed --no-discordant -q --phred33 -I 10 -X 2000 --threads 12”. The PCR duplicates from each dataset were removed using the rmdup function of samtools (v1.9)⁶³. Peak calling was done using MACS2 (v2.2.7.1)⁷⁶ with the parameters: “-B –nomodel –shif 100 –extsize 200”.

Differential binding analyses were performed using the edgeR package DiffBind (v3.4.11)^70,78. Peaks derived from individual samples were loaded into dba.count function with minOverlap=1 and summits= FALSE.. The annotatePeak.pl script from Homer (v4.11.1)⁶⁸ with default settings was employed for annotation. BAM files were merged for the biological replicates and converted to bigwig files for visualization in Integrative Genomics Viewer (IGV)⁶⁴ by using the command “bamCoververage” of deeptools (v3.5.1)⁷⁹ with parameters “-ignoreDuplicates -normalUsing RPKM -skipNonCoveredRegions -binSize 20”. Heatmaps were plotted using “computeMatrix” and “plotHeatmap” in deepTools (v3.5.1)⁷⁹. Violin plots were performed by R package ggplot2.

De novo motifs were identified by the EPIGRAM⁸⁰ pipeline. For H3K27me3, peaks with higher signals in high ΔΨ_m group were extracted as the target set and all the other H3K27me3 peaks were treated as the background set. Motifs were identified with following command line “perl epiplay-P.pl h3k27me3.target.bed h3k27me3.backgroup.bed 4 mm10 both 4G both”. TOMTOM (v5.4.1) (https://meme-suite.org/meme/tools/tomtom), an online motif comparison tool from the MEME suite⁸¹ was employed to identify the most matched motifs.

To determine the epigenome profiles near MERVL, deepTools computeMatrix was executed in the scale-region mode to calculate the signal density at each MERVL element using following parameters: “computeMatrix scale-regions -m 5000 -a 5000 -b 5000 -S bigwig.files -R bed.files”. For 2C genes, the signal density was also computed with computeMatrix in the scale-region mode in a normalized region 25 kb plus the 5 kb upstream to 5 kb downstream. DeepTools profiler was employed to generate the line plot to display the average values across the region. The violin plot was generated to show the intensity distribution by R package ggplot2, and a nonparametric Mann-Whitney test was used.

Statistical analysis

Statistical analyses were performed by the two-tailed Student’s t test using GraphPad Prism 7.0 and 9.0 software. Data were presented as standard error of the mean (SEM). P values of less than 0.05 were considered statistically significant. GO term enrichment analysis for RNA-seq data was conducted using DAVID (https://david.ncifcrf.gov)⁸². GO terms with fold enrichment > 2 and the adjusted P value < 0.05 (Benjamini-Hochberg method) were considered as significant enrichment. GO analysis for CUT&RUN peaks was conducted using Panther Classification System Version 17.0 (http://pantherdb.org/). Statistical overrepresentation test was performed with “default settings”. Fisher’s Exact with false discovery rate (FDR < 0.05) multiple test correction was employed. GO terms (FDR < 0.05) for ATAC-seq used GREAT software with the Mouse Phenotype Single KO annotation⁸³. Rules for the region-gene association included 5 kb upstream and 1 kb downstream of each gene.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The raw data from RNA-seq, ATAC-seq, CUT&RUN for H3K4me3, and H3K27me3, and bisulfite sequencing for the Elf5 promoter have been deposited in the Gene Expression Omnibus (GEO) under the accession number GSE210915. The previous published data used in this study are available in the GEO under the accession numbers GSM2711863 (ESC_2i), GSM2711865 (ESC_2i_EedKO), GSM838738 (MERVL-tdtomato⁺ and MERVL-tdtomato^-), GSM1966767 (MERVL⁺_Zscan4⁺ and MERVL⁺_Zscan4⁺), GSM3384433 (CRISPRa_EV and CRISPRa_MERVL), GSM3110917 (NELFA^high and NELFA^low), GSM1415501 (Zscan4^high and Zscan4^low), GSM2279983 (Dux-GFP⁺, Dux-GFP^-, Plus Dox-Dux and Minus Dox-Dux), and GSE66582 (distinct developmental stages of preimplantation embryos). All source and processed data generated in this study, along with individual replicate values, are provided in the Supplemental Information and Source Data file. Source data are provided with this paper.

Code availability

The source codes for mathematical simulation modeling are available at Zenodo [https://zenodo.org/records/10633128].

Change history

12 March 2024
A Correction to this paper has been published: https://doi.org/10.1038/s41467-024-46566-4

References

Cheung, T. H. & Rando, T. A. Molecular regulation of stem cell quiescence. Nat. Rev. Mol. Cell Biol. 14, 329–340 (2013).
Article CAS PubMed Google Scholar
van Velthoven, C. T. J. & Rando, T. A. Stem cell quiescence: dynamism, restraint, and cellular idling. Cell Stem Cell 24, 213–225 (2019).
Article PubMed PubMed Central Google Scholar
Marescal, O. & Cheeseman, I. M. Cellular mechanisms and regulation of quiescence. Dev. Cell 55, 259–271 (2020).
Article CAS PubMed PubMed Central Google Scholar
Ying, Q. L. et al. The ground state of embryonic stem cell self-renewal. Nature 453, 519–523 (2008).
Article CAS PubMed PubMed Central ADS Google Scholar
Khoa, L. T. P. et al. Histone acetyltransferase MOF blocks acquisition of quiescence in ground-state escs through activating fatty acid oxidation. Cell Stem Cell 27, 441–458.e410 (2020).
Article CAS PubMed PubMed Central Google Scholar
Scognamiglio, R. et al. Myc depletion induces a pluripotent dormant state mimicking diapause. Cell 164, 668–680 (2016).
Article CAS PubMed PubMed Central Google Scholar
Bulut-Karslioglu, A. et al. Inhibition of mTOR induces a paused pluripotent state. Nature 540, 119–123 (2016).
Article CAS PubMed PubMed Central ADS Google Scholar
Dhimolea, E. et al. An embryonic diapause-like adaptation with suppressed Myc activity enables tumor treatment persistence. Cancer Cell 39, 240–256.e211 (2021).
Article CAS PubMed PubMed Central Google Scholar
Rehman, S. K. et al. Colorectal cancer cells enter a diapause-like DTP state to survive chemotherapy. Cell 184, 226–242 e221 (2021).
Article CAS PubMed Google Scholar
Nagaraj, R. et al. Nuclear localization of mitochondrial TCA cycle enzymes as a critical step in mammalian zygotic genome activation. Cell 168, 210–223.e211 (2017).
Article CAS PubMed PubMed Central Google Scholar
Zhang, J. et al. Metabolism in pluripotent stem cells and early mammalian development. Cell Metab. 27, 332–338 (2018).
Article CAS PubMed Google Scholar
Dey, S. K. et al. Repurposing an adenine riboswitch into a fluorogenic imaging and sensing tag. Nat. Chem. Biol. 18, 180–190 (2022).
Article CAS PubMed Google Scholar
Liang, R. et al. Restraining lysosomal activity preserves hematopoietic stem cell quiescence and potency. Cell Stem Cell 26, 359–376.e357 (2020).
Article CAS PubMed PubMed Central Google Scholar
Rodgers, J. T. et al. mTORC1 controls the adaptive transition of quiescent stem cells from G0 to G(Alert). Nature 510, 393–396 (2014).
Article CAS PubMed PubMed Central ADS Google Scholar
Khoa, L. T. P. & Dou, Y. In vitro derivation of quiescent mouse embryonic stem cells based on distinct mitochondrial activity. STAR Protoc. 1, 100136, https://doi.org/10.1016/j.xpro.2020.100136 (2020).
Article PubMed PubMed Central Google Scholar
Martinez-Reyes, I. & Chandel, N. S. Mitochondrial TCA cycle metabolites control physiology and disease. Nat. Commun. 11, 102 (2020).
Article CAS PubMed PubMed Central ADS Google Scholar
Zorova, L. D. et al. Mitochondrial membrane potential. Anal. Biochem. 552, 50–59 (2018).
Article CAS PubMed Google Scholar
Gutscher, M. et al. Real-time imaging of the intracellular glutathione redox potential. Nat. Methods 5, 553–559 (2008).
Article CAS PubMed Google Scholar
Macfarlan, T. S. et al. Embryonic stem cell potency fluctuates with endogenous retrovirus activity. Nature 487, 57–63 (2012).
Article CAS PubMed PubMed Central ADS Google Scholar
Fessler, E. et al. A pathway coordinated by DELE1 relays mitochondrial stress to the cytosol. Nature 579, 433–437 (2020).
Article CAS PubMed PubMed Central ADS Google Scholar
Dykens, J. A. et al. Biguanide-induced mitochondrial dysfunction yields increased lactate production and cytotoxicity of aerobically-poised HepG2 cells and human hepatocytes in vitro. Toxicol. Appl. Pharmacol. 233, 203–210 (2008).
Article CAS PubMed Google Scholar
Ying, Q. L., Nichols, J., Evans, E. P. & Smith, A. G. Changing potency by spontaneous fusion. Nature 416, 545–548 (2002).
Article CAS PubMed ADS Google Scholar
Ishiuchi, T. et al. Early embryonic-like cells are induced by downregulating replication-dependent chromatin assembly. Nat. Struct. Mol. Biol. 22, 662–671 (2015).
Article CAS PubMed Google Scholar
Mandal, S., Lindgren, A. G., Srivastava, A. S., Clark, A. T. & Banerjee, U. Mitochondrial function controls proliferation and early differentiation potential of embryonic stem cells. Stem Cells 29, 486–495 (2011).
Article CAS PubMed Google Scholar
Wu, J. et al. The landscape of accessible chromatin in mammalian preimplantation embryos. Nature 534, 652–657 (2016).
Article CAS PubMed ADS Google Scholar
Akiyama, T. et al. Transient bursts of Zscan4 expression are accompanied by the rapid derepression of heterochromatin in mouse embryonic stem cells. DNA Res. 22, 307–318 (2015).
Article CAS PubMed PubMed Central Google Scholar
Eckersley-Maslin, M. A. et al. MERVL/Zscan4 network activation results in transient genome-wide DNA demethylation of mESCs. Cell Rep. 17, 179–192 (2016).
Article CAS PubMed PubMed Central Google Scholar
Hendrickson, P. G. et al. Conserved roles of mouse DUX and human DUX4 in activating cleavage-stage genes and MERVL/HERVL retrotransposons. Nat. Genet. 49, 925–934 (2017).
Article CAS PubMed PubMed Central Google Scholar
Yang, F. et al. DUX-miR-344-ZMYM2-mediated activation of MERVL LTRs induces a totipotent 2C-like state. Cell Stem Cell 26, 234–250.e237 (2020).
Article CAS PubMed PubMed Central Google Scholar
Hu, Z. et al. Maternal factor NELFA drives a 2C-like state in mouse embryonic stem cells. Nat. Cell Biol. 22, 175–186 (2020).
Article CAS PubMed Google Scholar
Genet, M. & Torres-Padilla, M. E. The molecular and cellular features of 2-cell-like cells: a reference guide. Development 147, dev189688 (2020).
Article CAS PubMed Google Scholar
Riveiro, A. R. & Brickman, J. M. From pluripotency to totipotency: an experimentalist’s guide to cellular potency. Development 147, dev189845 (2020).
Article CAS PubMed Google Scholar
Baker, C. L. & Pera, M. F. Capturing totipotent stem cells. Cell Stem Cell 22, 25–34 (2018).
Article CAS PubMed Google Scholar
Posfai, E. et al. Evaluating totipotency using criteria of increasing stringency. Nat. Cell Biol. 23, 49–60 (2021).
Article CAS PubMed Google Scholar
Ng, R. K. et al. Epigenetic restriction of embryonic cell lineage fate by methylation of Elf5. Nat. Cell Biol. 10, 1280–1290 (2008).
Article CAS PubMed PubMed Central Google Scholar
Ducker, G. S. & Rabinowitz, J. D. One-carbon metabolism in health and disease. Cell Metab. 25, 27–42 (2017).
Article CAS PubMed Google Scholar
Bian, Y. et al. Cancer SLC43A2 alters T cell methionine metabolism and histone methylation. Nature 585, 277–282 (2020).
Article CAS PubMed PubMed Central Google Scholar
Sun, Y. & Locasale, J. W. Rethinking the bioavailability and cellular transport properties of S-adenosylmethionine. Cell Stress 6, 1–5 (2022).
Article PubMed Google Scholar
Dai, Z., Ramesh, V. & Locasale, J. W. The evolving metabolic landscape of chromatin biology and epigenetics. Nat. Rev. Genet. 21, 737–753 (2020).
Article CAS PubMed PubMed Central Google Scholar
Marks, H. et al. The transcriptional and epigenomic foundations of ground state pluripotency. Cell 149, 590–604 (2012).
Article CAS PubMed PubMed Central Google Scholar
van Mierlo, G. et al. Integrative proteomic profiling reveals PRC2-dependent epigenetic crosstalk maintains ground-state pluripotency. Cell Stem Cell 24, 123–137 e128 (2019).
Article PubMed Google Scholar
Piunti, A. & Shilatifard, A. The roles of Polycomb repressive complexes in mammalian development and cancer. Nat. Rev. Mol. Cell Biol. 22, 326–345 (2021).
Article CAS PubMed Google Scholar
Latos, P. A. et al. Elf5-centered transcription factor hub controls trophoblast stem cell self-renewal and differentiation through stoichiometry-sensitive shifts in target gene networks. Genes Dev. 29, 2435–2448 (2015).
Article CAS PubMed PubMed Central Google Scholar
Fueyo, R., Judd, J., Feschotte, C. & Wysocka, J. Roles of transposable elements in the regulation of mammalian transcription. Nat. Rev. Mol. Cell Biol. 23, 481–497 (2022).
Article CAS PubMed PubMed Central Google Scholar
Li, C. et al. Overview of methionine adenosyltransferase 2A (MAT2A) as an anticancer target: structure, function, and inhibitors. J. Med. Chem. 65, 9531–9547 (2022).
Article CAS PubMed Google Scholar
Sanderson, S. M., Gao, X., Dai, Z. & Locasale, J. W. Methionine metabolism in health and cancer: a nexus of diet and precision medicine. Nat. Rev. Cancer 19, 625–637 (2019).
Article CAS PubMed Google Scholar
Maclary, E. et al. PRC2 represses transcribed genes on the imprinted inactive X chromosome in mice. Genome Biol. 18, 82 (2017).
Article PubMed PubMed Central Google Scholar
Bernstein, B. E. et al. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315–326 (2006).
Article CAS PubMed Google Scholar
Habibi, E. et al. Whole-genome bisulfite sequencing of two distinct interconvertible DNA methylomes of mouse embryonic stem cells. Cell Stem Cell 13, 360–369 (2013).
Article CAS PubMed Google Scholar
Schoeftner, S. et al. Recruitment of PRC1 function at the initiation of X inactivation independent of PRC2 and silencing. EMBO J. 25, 3110–3122 (2006).
Article CAS PubMed PubMed Central Google Scholar
Kumar, B. et al. Polycomb repressive complex 2 shields naive human pluripotent cells from trophectoderm differentiation. Nat. Cell Biol. 24, 845–857 (2022).
Article CAS PubMed PubMed Central Google Scholar
Zijlmans, D. W. et al. Integrated multi-omics reveal polycomb repressive complex 2 restricts human trophoblast induction. Nat. Cell Biol. 24, 858–871 (2022).
Article CAS PubMed PubMed Central Google Scholar
Intlekofer, A. M. & Finley, L. W. S. Metabolic signatures of cancer cells and stem cells. Nat. Metab. 1, 177–188 (2019).
Article PubMed PubMed Central Google Scholar
Sperber, H. et al. The metabolome regulates the epigenetic landscape during naive-to-primed human embryonic stem cell transition. Nat. Cell Biol. 17, 1523–1535 (2015).
Article CAS PubMed PubMed Central Google Scholar
Haws, S. A. et al. Methyl-metabolite depletion elicits adaptive responses to support heterochromatin stability and epigenetic persistence. Mol. Cell 78, 210–223 e218 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kalantry, S. et al. The Polycomb group protein Eed protects the inactive X-chromosome from differentiation-induced reactivation. Nat. Cell Biol. 8, 195–202 (2006).
Article CAS PubMed PubMed Central Google Scholar
Benaglia, T., Chauveau, D., Hunter, D. R. & Young, D. S. mixtools: an R package for analyzing mixture models. J. Stat. Softw. 32, 1–29 (2010).
Google Scholar
Krueger, F. & Andrews, S. R. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics 27, 1571–1572 (2011).
Article CAS PubMed PubMed Central Google Scholar
Schneider, C. A., Rasband, W. S. & Eliceiri, K. W. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675 (2012).
Article CAS PubMed PubMed Central Google Scholar
Lee, H. J., Kremer, D. M., Sajjakulnukit, P., Zhang, L. & Lyssiotis, C. A. A large-scale analysis of targeted metabolomics data from heterogeneous biological samples provides insights into metabolite dynamics. Metabolomics 15, 103 (2019).
Article PubMed PubMed Central Google Scholar
Pang, Z. et al. MetaboAnalyst 5.0: narrowing the gap between raw spectra and functional insights. Nucleic Acids Res. 49, W388–W396 (2021).
Article CAS PubMed PubMed Central Google Scholar
Kim, D. et al. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 14, R36 (2013).
Article PubMed PubMed Central Google Scholar
Li, H. et al. The sequence alignment/map format And SAMtools. Bioinformatics 25, 2078–2079 (2009).
Article PubMed PubMed Central Google Scholar
Thorvaldsdottir, H., Robinson, J. T. & Mesirov, J. P. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief. Bioinform. 14, 178–192 (2013).
Article CAS PubMed Google Scholar
Ramirez, F. et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 44, W160–W165 (2016).
Article CAS PubMed PubMed Central Google Scholar
Yates, A. D. et al. Ensembl 2020. Nucleic Acids Res. 48, D682–D688 (2020).
CAS PubMed Google Scholar
Zhao, Y. et al. NONCODE 2016: an informative and valuable data source of long non-coding RNAs. Nucleic Acids Res. 44, D203–208, (2016).
Article CAS PubMed Google Scholar
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).
Article CAS PubMed PubMed Central Google Scholar
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014).
Article CAS PubMed Google Scholar
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2010).
Article CAS PubMed Google Scholar
Risso, D., Ngai, J., Speed, T. P. & Dudoit, S. Normalization of RNA-seq data using factor analysis of control genes or samples. Nat. Biotechnol. 32, 896–902 (2014).
Article CAS PubMed PubMed Central Google Scholar
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Article PubMed PubMed Central Google Scholar
Lazar, C. et al. Batch effect removal methods for microarray gene expression data integration: a survey. Brief. Bioinform. 14, 469–490 (2013).
Article CAS PubMed Google Scholar
Krueger, F. Trim Galore: a wrapper tool around Cutadapt and FastQC to consistently apply quality and adapter trimming to FastQ files, with some extra functionality for MspI-digested RRBS-type (Reduced Representation Bisufite-Seq) libraries. accessed 28 April 2016 http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/ (2012).
Langmead, B. Aligning short sequencing reads with Bowtie. Curr. Protoc. Bioinform. 32, 11.17. 11–11.17. 14 (2010).
Article Google Scholar
Gaspar, J. M. Improved peak-calling with MACS2. BioRxiv, 496521 (2018).
Zang, C. et al. A clustering approach for identification of enriched domains from histone modification ChIP-Seq data. Bioinformatics 25, 1952–1958 (2009).
Article CAS PubMed PubMed Central Google Scholar
Stark, R. & Brown, G. DiffBind: differential binding analysis of ChIP-Seq peak data. R package version 100 (2011).
Ramírez, F., Dündar, F., Diehl, S., Grüning, B. A. & Manke, T. deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res. 42, W187–W191 (2014).
Article PubMed PubMed Central Google Scholar
Whitaker, J. W., Chen, Z. & Wang, W. Predicting the human epigenome from DNA motifs. Nat. Methods 12, 265–272 (2015).
Article CAS PubMed Google Scholar
Bailey, T. L. et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 37, W202–W208 (2009).
Article CAS PubMed PubMed Central Google Scholar
Sherman, B. T. et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res. 50, W216–W221 (2022).
Article CAS PubMed PubMed Central ADS Google Scholar
McLean, C. Y. et al. GREAT improves functional interpretation of cis-regulatory regions. Nat. Biotechnol. 28, 495–501 (2010).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank members of Drs. Yali Dou and Stephanie L. Bielas labs, Dr. Sriram Chandrasekaran and Scott Campit for helpful discussions. We thank Dr. Sally A. Camper for sharing key equipment. We are grateful to Dr. Austin Smith for providing the Oct4-GiP ESCs and to Dr. Janet Rossant for providing the iCdx2 Elf5-2A-mCherry ESCs. Drs. Ursula Jakob, Hitoshi Niwa, and Tobias P. Dick kindly provided the mitoCOX8-Grx1-roGFP2, doxycycline (Dox) inducible PiggyBac pPBhCMV1cHApA, pPBCAG-rtTAM2-IN, and pCAG-hyPBase, and the pCaSper4mitoCOX8-Grx1-roGFP2 plasmids, respectively. Y.D. was supported by NIH grants (R01 NS100156 and GM082856-15) as well as Norris Comprehensive Cancer Center Grant (P30CA014089). C.A.L. was supported by the NCI grants (R01CA248160, R01CA244931) and UMCCC Core Grant (P30CA046592). R.M. thanks the University of Michigan Genetics Training Program (NIH T32GM007544) and the University of Michigan Career Training in Reproductive Biology (NIH T32HD079342). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Author information

Le Tran Phuc Khoa
Present address: Department of Molecular and Integrative Physiology, University of Michigan Medical School, Ann Arbor, MI, 48109, USA

Authors and Affiliations

Department of Medicine, Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, CA, 90033, USA
Le Tran Phuc Khoa, Wentao Yang & Yali Dou
Department of Molecular and Integrative Physiology, University of Michigan Medical School, Ann Arbor, MI, 48109, USA
Mengrou Shan, Li Zhang & Costas A. Lyssiotis
Institute of Medical Innovation and Research, Peking University Third Hospital, Beijing, China
Fengbiao Mao
Department of Human Genetics, University of Michigan Medical School, Ann Arbor, MI, 48109, USA
Bo Zhou, Rebecca Malcore, Clair Harris, Shigeki Iwase, Sundeep Kalantry & Stephanie L. Bielas
Department of Ophthalmology & Visual Sciences, W.K. Kellogg Eye Center, University of Michigan, 1000 Wall St., Ann Arbor, MI, 48105, USA
Qiang Li & Rajesh C. Rao
Beaumont Hospital, Wayne, 33155 Annapolis St., Wayne, MI, 48184, USA
Lili Zhao

Authors

Le Tran Phuc Khoa
View author publications
You can also search for this author in PubMed Google Scholar
Wentao Yang
View author publications
You can also search for this author in PubMed Google Scholar
Mengrou Shan
View author publications
You can also search for this author in PubMed Google Scholar
Li Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Fengbiao Mao
View author publications
You can also search for this author in PubMed Google Scholar
Bo Zhou
View author publications
You can also search for this author in PubMed Google Scholar
Qiang Li
View author publications
You can also search for this author in PubMed Google Scholar
Rebecca Malcore
View author publications
You can also search for this author in PubMed Google Scholar
Clair Harris
View author publications
You can also search for this author in PubMed Google Scholar
Lili Zhao
View author publications
You can also search for this author in PubMed Google Scholar
Rajesh C. Rao
View author publications
You can also search for this author in PubMed Google Scholar
Shigeki Iwase
View author publications
You can also search for this author in PubMed Google Scholar
Sundeep Kalantry
View author publications
You can also search for this author in PubMed Google Scholar
Stephanie L. Bielas
View author publications
You can also search for this author in PubMed Google Scholar
Costas A. Lyssiotis
View author publications
You can also search for this author in PubMed Google Scholar
Yali Dou
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

L.T.P.K. and Y.D., conceived and designed the original study. L.T.P.K., performed most experiments with the help from B.Z. and Q.L. W.Y. and F.M., performed bioinformatics analyses and W.Y., wrote bioinformatics Method and deposited the data to the GEO and Zenodo; M.S. and L. Zhang., under the supervision of C.A.L., performed LC-MS/MS-based metabolomics. R.M., under the supervision of S.K., derived the Eedfl/- and Eed−/− ESC lines. C.H., under the supervision of S.K., derived the TSCs. L. Zhao performed mathematical modeling. S.I. and R.R., contributed to project development and supervised B.Z. and Q.L., respectively. L.T.P.K. and Y.D., wrote the manuscript with input from other authors. Y.D. and S.L.B., supervised the overall study.

Corresponding author

Correspondence to Yali Dou.

Ethics declarations

Competing interests

C.A.L. is a paid consultant for Astellas Pharmaceuticals, Odyssey Therapeutics, and T-Knife Therapeutics. C.A.L. received no specific funding for this work. The remaining authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Peer Review File

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Supplementary Data 7

Supplementary Data 8

Reporting Summary

Source data

Source Data

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 licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence 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 licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Khoa, L.T.P., Yang, W., Shan, M. et al. Quiescence enables unrestricted cell fate in naive embryonic stem cells. Nat Commun 15, 1721 (2024). https://doi.org/10.1038/s41467-024-46121-1

Download citation

Received: 08 December 2023
Accepted: 14 February 2024
Published: 26 February 2024
DOI: https://doi.org/10.1038/s41467-024-46121-1

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.