The deubiquitinase USP21 maintains the stemness of mouse embryonic stem cells via stabilization of Nanog

Jin, Jiali; Liu, Jian; Chen, Cong; Liu, Zhenping; Jiang, Cong; Chu, Hongshang; Pan, Weijuan; Wang, Xinbo; Zhang, Lingqiang; Li, Bin; Jiang, Cizhong; Ge, Xin; Xie, Xin; Wang, Ping

doi:10.1038/ncomms13594

Download PDF

Article
Open access
Published: 25 November 2016

The deubiquitinase USP21 maintains the stemness of mouse embryonic stem cells via stabilization of Nanog

Jiali Jin¹^na1,
Jian Liu^2,3^na1,
Cong Chen¹,
Zhenping Liu⁴,
Cong Jiang¹,
Hongshang Chu¹,
Weijuan Pan¹,
Xinbo Wang¹,
Lingqiang Zhang⁵,
Bin Li⁶,
Cizhong Jiang ORCID: orcid.org/0000-0001-5406-2360⁴,
Xin Ge⁷,
Xin Xie^2,3 &
…
Ping Wang^1,4

Nature Communications volume 7, Article number: 13594 (2016) Cite this article

8732 Accesses
66 Citations
23 Altmetric
Metrics details

Subjects

Abstract

Nanog is a master pluripotency factor of embryonic stem cells (ESCs). Stable expression of Nanog is essential to maintain the stemness of ESCs. However, Nanog is a short-lived protein and quickly degraded by the ubiquitin-dependent proteasome system. Here we report that the deubiquitinase USP21 interacts with, deubiquitinates and stabilizes Nanog, and therefore maintains the protein level of Nanog in mouse ESCs (mESCs). Loss of USP21 results in Nanog degradation, mESCs differentiation and reduces somatic cell reprogramming efficiency. USP21 is a transcriptional target of the LIF/STAT3 pathway and is downregulated upon differentiation. Moreover, differentiation cues promote ERK-mediated phosphorylation and dissociation of USP21 from Nanog, thus leading to Nanog degradation. In addition, USP21 is recruited to gene promoters by Nanog to deubiquitinate histone H2A at K119 and thus facilitates Nanog-mediated gene expression. Together, our findings provide a regulatory mechanism by which extrinsic signals regulate mESC fate via deubiquitinating Nanog.

Targeting DCAF5 suppresses SMARCB1-mutant cancer by stabilizing SWI/SNF

Article 27 March 2024

Sandi Radko-Juettner, Hong Yue, … Charles W. M. Roberts

Spatially organized cellular communities form the developing human heart

Article Open access 13 March 2024

Elie N. Farah, Robert K. Hu, … Neil C. Chi

Transcription bodies regulate gene expression by sequestering CDK9

Article Open access 08 April 2024

Martino Ugolini, Maciej A. Kerlin, … Nadine L. Vastenhouw

Introduction

The pluripotency of embryonic stem cells (ESCs) is regulated by a network of transcription factors (TFs), including Oct4, Sox2 and Nanog¹. Among them, Nanog plays an essential role in the transcriptional network of pluripotency and early embryonic development^2,3, controlling the epiblast versus primitive endoderm decision in the blastocyst⁴. Nanog-depleted blastocysts fail to generate epiblasts and produce only parietal endoderm-like cells³, whereas the ectopic expression of Nanog is sufficient to induce leukaemia inhibitory factor (LIF)-independent self-renewal of cultured mouse ESCs (mESCs)². Downregulation of Nanog in mESCs results in differentiation into a broader repertoire of cell lineages, with a primary contribution to the trophectoderm and primitive endoderm⁵. Due to its essential roles in pluripotency, Nanog is among the four factors (Oct4, Sox2, Nanog and Lin28) that were initially used to reprogram human somatic cells into pluripotent stem cells⁶.

The amounts of the pluripotency factors in ESCs are precisely controlled to maintain ESCs self-renewal^5,7,8. A number of TFs have been reported to activate and/or repress Nanog expression in mESCs, with Oct4 and Sox2 representing the major regulators⁹. Fine-tuning of Nanog expression is achieved via several signal transduction cascades. For instance, Nanog expression is promoted by LIF through two parallel pathways: the JAK/STAT3 pathway, which activates Klf4, and the PI3K/AKT pathway, which activates Tbx3 (ref. 10). A STAT3-binding site has been identified in an enhancer region upstream of the Nanog promoter¹¹. Other signalling pathways, such as FGF/MEK¹², GSK3β¹³ and TGFβ¹⁴, also participate in the regulation of Nanog expression. Interestingly, Nanog expression is primarily monoallelic in mESCs cultured in standard LIF/serum conditions. However, culturing mESCs in 2i (MEK inhibitor and GSK3 inhibitor)/LIF conditions significantly increases the level of biallelic Nanog expression, indicating that Nanog is a marker of ground-state pluripotency^15,16.

In addition to the regulation of their expression, the degradation of these pluripotency factors is also tightly controlled by the ubiquitin–proteasome system (UPS). The UPS is one of the key systems that regulate cellular protein levels under various conditions. Protein ubiquitination is a reversible and balanced process catalysed by Ub-activating enzyme (E1), Ub-conjugating enzyme (E2), Ub-protein ligase (E3) and deubiquitinases (DUBs). Several studies have investigated the mechanism by which the UPS controls the protein levels of key pluripotency factors. As a key pluripotency factor, Nanog is a short-lived protein that is rapidly degraded by E3 ligase FBXW8-mediated ubiquitination¹⁷. However, the DUBs that regulate Nanog stability are unknown.

In this study, we sought to identify DUBs that specifically regulate Nanog stability. We screened 46 DUBs and identified USP21 as an efficient deubiquitinating enzyme that governs Nanog stability in ESCs. Our study thus reveals a dynamic regulatory mechanism underlying Nanog stability and transcriptional activity through external signals.

Results

The DUB USP21 regulates Nanog stability

To search for DUBs that could stabilize Nanog, we fused firefly luciferase to the C terminus of Nanog and used this fusion protein (Nanog-Luc) as a reporter of Nanog stability (Supplementary Fig. 1a). As the fused luciferase is degraded together with the Nanog, the degradation of Nanog can be easily quantified by measuring the luciferase activity. Forty-six mammalian DUBs were screened with this reporter. We found that co-expression with USP21, but not the other DUBs, significantly increased the luciferase activity of Nanog-Luc (Fig. 1a).

**Figure 1: USP21 directly stabilizes and interacts with Nanog.**

Two USP21 isoforms (USP21-LV and USP21-SV) with different subcellular locations have been reported (Supplementary Fig. 1b)¹⁸. Co-expression of either USP21 isoform significantly prolonged the half-life of Nanog (Fig. 1b). In contrast, USP21 had almost no effect on the half-life of Oct4, Klf4 and Sox2 protein (Supplementary Fig. 1c). The stabilization of Nanog by USP21 was dependent on its deubiquitinating enzyme activity because the catalytic-inactive mutant C221A (CA-USP21) neither stabilized Nanog (Fig. 1c) nor increased the Nanog transcriptional activity measured by a luciferase reporter assay (Supplementary Fig. 1d). Depletion of endogenous USP21 by short hairpin RNA (shRNA) in mESC E14 cells reduced both the protein level (Fig. 1d) and half-life of Nanog, but not the half-life of other pluripotent factors such as Sox2, Oct4 and Klf4 (Fig. 1e). It is interesting to note that knockdown of USP21 led to reduced protein levels of Nanog, Sox2, Oct4 and Klf4, indicating a loss of pluripotency. To ensure the starting protein levels of these factors were similar in the protein half-life assay, we increased the loading amount of the USP21 knockdown group by twofold (which is why you see an increase in the GAPDH level). The half-life of Nanog was also shortened in USP21 knockout (USP21-KO) mouse embryonic fibroblasts (MEFs; Supplementary Fig. 1e, see below). In addition of the loss of Nanog protein, knockdown USP21 also led to the loss of the tight compact morphology of mESC (Fig. 1f).

To examine whether Nanog is specifically regulated by USP21, we examined the regulation of Nanog by USP2, the most closely related DUB to USP21. Our data showed the overexpression of USP2 slightly prolonged the half-life of Nanog (Supplementary Fig. 1f). However, knockdown USP2 in E14 cells had little effect on the half-life of Nanog (Fig. 1e), indicating a less significant role of USP2 on Nanog stability in native condition. Knockdown of USP2 and USP21 was monitored by PCR with reverse transcription (Supplementary Fig. 1g). Together, these data indicate that USP21, but not USP2, is a specific DUB that regulates Nanog stability.

USP21 directly interacts with and deubiquitinates Nanog

Since USP21 is a DUB, we examined whether USP21 could directly interact with and deubiquitinate Nanog. Both USP21 isoforms were found to bind to Nanog and vice versa when examined by co-immunoprecipitation (co-IP) assay (Fig. 1g,h). In contrast, the binding between Nanog and USP2 was much weaker than that between Nanog and USP21 (Supplementary Fig. 1h). The glutathione S-transferase (GST) pull-down assay using purified Nanog and USP21 proteins demonstrated that USP21 directly interacted with Nanog (Fig. 1i). In E14 cells, the endogenous USP21 was also found to interact with endogenous Nanog (Fig. 1j), but not other stem cell factors (Oct4, Sox2 and Klf4; Fig. 1k). The results were confirmed using an overexpression system in HEK293T cells (Supplementary Fig. 1i). Together, these data indicate that USP21 is a direct binding partner for Nanog.

To determine the region(s) of USP21 and Nanog responsible for their interaction, several deletion mutants of USP21 and Nanog were generated, and their interactions were examined using co-IP assays. The C-terminal UCH (catalytic domain), but not the N-terminal domain of USP21 was found to be responsible for its interaction with Nanog (Supplementary Fig. 1j). Mouse Nanog contains serine-rich N-terminal (N), NK-2-type homeobox (H) and C-terminal (C1WC2) domains. We found that the C-terminal domain of Nanog was required for its binding to USP21 (Supplementary Fig. 1k).

Next, we examined whether USP21 could deubiquitinate Nanog. Our data showed that co-expression of the wild type (WT)-USP21, but not the CA-USP21, almost completely removed the ubiquitin chains from Nanog (Fig. 2a; Supplementary Fig. 2a). In addition, the deubiquitination ability of USP21 was stronger than USP2 (Supplementary Fig. 2a). To examine whether USP21 deubiquitinated Nanog directly, both WT-USP21 and CA-USP21 were purified and their ability to deubiquitinate Nanog was tested in an in vitro deubiquitination assay. We found that WT-USP21, but not CA-USP21, efficiently deubiquitinated Nanog (Fig. 2b), indicating that USP21 was a direct DUB for Nanog. Moreover, depletion of USP21 by shRNAs in mESCs significantly increased the ubiquitination of endogenous Nanog (Fig. 2c), indicating that endogenous Nanog was also a target of USP21.

**Figure 2: USP21 deubiquitinates Nanog.**

Mouse Nanog contains 16 lysines. Mutation of all of these lysine residues (16KR) completely abolished Nanog ubiquitination (Fig. 2e), indicating that Nanog was ubiquitinated at lysine residues. Nanog has been reported to be ubiquitinated at K138 and K169 residues (http://www.phosphosite.org/proteinAction.do?id=15240190&showAllSites=true), which are conserved among various species including humans, rats and rabbits (Fig. 2d). Our data confirmed that the simultaneous replacement of K138 and K169 with Arg (K138/K169R) markedly reduced Nanog ubiquitination (Fig. 2e). A Nanog double mutant (K138/K169R) also showed a longer half-life than WT Nanog (Fig. 2f). The K to R mutations did not affect the interaction between USP21 and Nanog (Supplementary Fig. 2b), suggesting that binding of USP21 was independent of Nanog ubiquitination. Consistently with its stability, the Nanog K138/K169R mutant showed slightly stronger transcriptional activity than WT Nanog (Supplementary Fig. 2c). Together, these data indicate that USP21 is a bona fide DUB of Nanog.

USP21 regulates mESC function

Nanog is essential for the maintenance of mESC self-renewal and pluripotency¹⁹. Thus, we examined whether USP21 plays a role in the self-renewal of mESCs. We derived Usp21^Loxp/Loxp ESCs by intercrossing Usp21^Loxp/Loxp mice; Usp21 was knocked out by infection with a lentivirus carrying Cre recombinase (Supplementary Fig. 3a,b). For mESCs, the colonies are typically tightly packed, bright, dome-shaped with smooth edge. In alkaline phosphatase (AP) staining, the pluripotent mESC colonies will be stained deeply in a reddish brown colour with smooth edge, while differentiating ESC colonies (much less compact compared with the pluripotent ones) will stain in less saturated colour or completely lose AP staining²⁰. We found that deletion of USP21 caused the loss of mESC morphology and exhibited reduced AP staining (Fig. 3a). Moreover, USP21 knockdown with specific shRNAs in E14 cells also led to the loss mESC morphology and reduced AP staining (Fig. 3b; Supplementary Fig. 3c). Consistently, USP21 knockdown significantly reduced the expression of the pluripotency genes including Nanog, Oct4 and Rex1, but increased the expression of lineage-specific genes, including the ectoderm markers Fgf5 and Nestin, mesoderm marker Desmin, and endoderm markers Gata4 and Sox17, in a pattern similar to Nanog depletion (Supplementary Fig. 3d), indicating that USP21 is required for the self-renewal of mESCs. Furthermore, Usp21 depletion reduced the ability of mESCs to form teratomas (Fig. 3c). Conversely, overexpression of either USP21-LV or USP21-SV promoted mESC self-renewal under less optimal culture conditions. LIF withdrawal led to the significant differentiation of the control mESCs in 3 days. However, mESCs overexpressing USP21 were still able to maintain the typical mESC morphology and remain AP-positive in culture without LIF for 3 days (Supplementary Fig. 3e). Western blotting and PCR analysis also revealed the upregulation of pluripotency genes in mESCs overexpressing USP21 (Supplementary Fig. 3f). Although overexpression of USP21 promoted mESC self-renewal in vitro, it did not affect the developmental potential of these cells in vivo. Tissues derived from all three germ layers were found in teratomas formed with these USP21-overexpressing mESCs (Supplementary Fig. 3g).

**Figure 3: USP21 is required for mESC self-renewal.**

Next, we investigated whether USP21 maintained mESC self-renewal by regulating Nanog. To this end, USP21 was knocked down with a specific shRNA targeting the 3′-untranslated region (UTR) of USP21; then, a shRNA-resistant USP21 was reintroduced. Reintroduction of USP21 efficiently rescued the mESC differentiation caused by USP21 depletion, but failed to rescue mESC differentiation caused by Nanog knockdown (Fig. 3d,e). Next, we examined whether the ubiquitination of Nanog contributed to ESC self-renewal. We found that the ubiquitination-resistant K138/K169R mutant had stronger ability to maintain the self-renewal of mESCs than WT Nanog. Moreover, the K138/169R mutant showed a stronger ability to rescue the effect of USP21 knockdown on mESC self-renewal (Fig. 3f,g). These data indicate that USP21 is required for the self-renewal of mESCs and Nanog is downstream of USP21.

Nanog is required to attain the ground state of authentic pluripotency in the final phase of somatic cell reprogramming²¹. Therefore, we examined whether USP21 was required for the generation of induced pluripotent stem cells (iPSCs) from MEFs with Oct4, Sox2 and Klf4. USP21-KO MEFs displayed a significantly lower reprogramming efficiency than WT MEFs, as indicated by the reduced number of AP-positive colonies on day 12 of induction (Supplementary Fig. 3h). On the other hand, overexpression of USP21-LV or USP21-SV in OG2 MEFs carrying an Oct4-GFP reporter significantly increased the number of GFP⁺ iPSC colonies induced by Oct4, Sox2, Klf4 and Nanog on day 16. In contrast, CA-USP21 had no effect (Supplementary Fig. 3i). Collectively, our data indicate that USP21 is a specific DUB for Nanog and plays important roles in mESC self-renewal and somatic cell reprogramming by regulating Nanog stability.

USP21 is a transcriptional target of LIF/STAT3

Because USP21 is required for the self-renewal of ESCs, we investigated whether its expression was regulated by signals involved in ESC maintenance. Interestingly, we found that withdrawal of LIF induced spontaneous differentiation of mESCs that was accompanied by the marked decrease of both the messenger RNA and protein levels of USP21 and the pluripotency markers Nanog, Oct4 and Sox2 (Fig. 4a,b). Similar results were obtained for the retinoic acid (RA)-induced differentiation of mESCs (Supplementary Fig. 4a,b).

To examine the mechanism by which USP21 expression was downregulated during mESC differentiation, we analysed the potential binding sites of transcriptional factors on the USP21 promoter using the Vista 2.0 software. We did not find potential binding sites of pluripotency factors, including Oct4, Nanog and Sox2. However, two potential STAT3-binding sites were identified within the −836 to −828 and −3,054 to −3,033 regions (Fig. 4c). STAT3 is a TF that is required to maintain mESCs in their undifferentiated state and is activated by LIF/JAK signalling²². Therefore, we examined whether the expression of USP21 was regulated by STAT3. Chromatin immunoprecipitation (ChIP) analysis revealed that STAT3 was preferentially enriched at the USP21 promoter in mESCs cultured in LIF-containing medium, and this enrichment was significantly reduced when the mESCs were cultured in medium without LIF for three days (Fig. 4c). Moreover, the expression of exogenous STAT3 effectively activated USP21 promoter-driven luciferase expression in a dose-dependent manner (Fig. 4d), whereas mutating the core STAT3-binding sequence (TGCTTCCCC to TGCCCACCC) within the USP21 promoter abolished the effect of STAT3 (Fig. 4e). Knockdown of STAT3 in mESCs also significantly reduced the messenger RNA and protein levels of USP21 (Fig. 4f). Phosphorylation at Tyr 705 (pY705) has been reported to be essential for STAT3 activation²². In accordance with these results, treatment with Cryptotanshinone, a specific inhibitor of STAT3 by inhibiting STAT3 Y705 phosphorylation, reduced the expression of USP21 (Supplementary Fig. 4c). Our data indicated that STAT3 (Y705D, which mimics pY705), but not STAT3 (Y705F, which cannot be phosphorylated), stimulated USP21 expression in E14 cells (Fig. 4g). The level of STAT3 pY705 was reduced upon ESCs differentiation induced by LIF withdrawal or RA treatment (Supplementary Fig. 4d,e). Together, these data indicate that USP21 is a direct transcriptional target of the LIF/STAT3 pathway in mESCs.

USP21 phosphorylation blocks its effect on Nanog

Extracellular-signal regulated kinase (ERK) activation leads to mESC differentiation²³. Thus, we were interested in whether the regulation of Nanog by USP21 was influenced by ERK signalling. Our data revealed that the interaction between USP21 and Nanog was decreased upon epidermal growth factor stimulation in HEK293T cells, and that this effect was partially blocked by ERK1/2 inhibitor (SCH772984) (Supplementary Fig. 5a), suggesting that the interaction between USP21 and Nanog is regulated by the ERK pathway. Furthermore, treatment of mESCs with FGF4 (a growth factor that induces ERK activation and mESC differentiation¹⁵) reduced the interaction between Nanog and USP21 in mESCs in a manner that could be reversed by the MEK inhibitor PD0325901 (Fig. 5a). Consistently, Nanog ubiquitination was increased in FGF4-stimulated mESCs, which was blocked by PD0325901 (Fig. 5b). Moreover, activation of ERK1 by co-expression of a constitutively active MEK1 (MEK1^CA) in HEK293T cells significantly impaired the interaction between USP21 and Nanog (Supplementary Fig. 5b). Interestingly, we also noticed that co-expression of MEK1^CA and ERK resulted in the band shift of both Nanog and USP21 (Supplementary Fig. 5b), which is possibly caused by the phosphorylation of Nanog and USP21 by the activated ERK. Furthermore, co-expression of MEK1^CA and ERK1 inhibited the USP21-mediated deubiquitination of Nanog (Supplementary Fig. 5c). These data indicate that the USP21-mediated deubiquitination of Nanog is regulated by ERK activation.

**Figure 5: USP21 is phosphorylated by ERK1.**

Next, we investigated the mechanism by which ERK1 regulates USP21-mediated Nanog deubiquitination. We first examined whether ERK1 was a binding partner of USP21 and found that exogenously expressed USP21 interacted with ERK1 in HEK293T cells, as analysed by a co-IP assay (Supplementary Fig. 5d). Next, we evaluated whether USP21 was a phospho-substrate of ERK1. We analysed USP21 phosphorylation by using a phospho-Tag gel and found that both endogenous and exogenous USP21 were significantly phosphorylated upon the activation of ERK (Fig. 5c; Supplementary Fig. 5e). By using an antibody that specifically recognizes the ERK1/2 phosphorylation motif, USP21 could be found in the immunoprecipitation (IP) products from E14 cell stimulated with FGF4. In PD0325901-treated cells, however, much less USP21 was IPed by the ERK1/2 phosphorylation motif antibody (Fig. 5d). These results confirmed that USP21 is a phospho-substrate of ERKs.

Then, we analysed potential ERK phosphorylation motifs in USP21 by using Scansite (www.scansite.mit.edu) and found that USP21 contains three putative sites with the consensus ERK phosphorylation motifs XXpS/pTP (S93, S335 and S539), with S539 conserved in mice, rats and humans (Fig. 5e). Interestingly, mutation of S539, but not the other two sites, to Ala, markedly reduced the ERK1-mediated phosphorylation of USP21 (Fig. 5f), suggesting that USP21 is phosphorylated by ERK1 at S539. To further confirm this finding, we generated an antibody that specifically recognized phospho-S539 (p-S539) in USP21 (Supplementary Fig. 5f). Treating mESCs with FGF4 significantly induced S539 phosphorylation of the endogenous USP21 or exogenously expressed WT-USP21, but not the S539A mutant (Fig. 5g,h). In an in vitro phosphorylation assay, WT-USP21, but not the S539A mutant, was found to be phosphorylated by ERK1 at S539, confirming that USP21 is a direct substrate of ERK1 (Fig. 5i). Collectively, these data clearly demonstrate that USP21 is directly phosphorylated by ERK1 at Ser539.

Next, we asked whether the phosphorylation of USP21 at S539 affected its ability to stabilize Nanog. Interestingly, we found that the WT and dephosphomimetic mutant S539A, but not the phosphomimetic mutant S539D, significantly stabilized Nanog (Fig. 6a; Supplementary Fig. 6a). Concordantly, the USP21 S539D mutant also partially lost the ability to deubiquitinate Nanog (Fig. 6b). We found that S539D, but not S539A, S93D or S335D, showed a reduced ability to interact with Nanog (Fig. 6c; Supplementary Fig. 6b). In mESCs, both the MEK inhibitor PD0325901 and ERK1/2 inhibitor SCH772984 significantly stabilized Nanog (Supplementary Fig. 6c). Overexpression of both the WT-USP21 and S539A mutant, but not the S539D mutant, stabilized Nanog (Fig. 6d) and promoted self-renewal of mESCs in normal culture condition (Supplementary Fig. 6d). More significantly, S539A mutant almost completely prevented mESC differentiation induced by FGF4 (Fig. 6e,f). Furthermore, compared with the WT and S539A mutant USP21, mESCs with overexpression of S539D displayed a reduced ability to form teratomas in vivo (Fig. 6g; Supplementary Fig. 6e). Collectively, these data demonstrate that ERK1-induced phosphorylation of USP21 at S539 prevents its interaction with Nanog and thus negatively regulates USP21-mediated Nanog deubiquitination and stabilization.

**Figure 6: Phosphorylation of USP21 at S539 blocks its effect on Nanog function.**

USP21 regulates H2A ubiquitination via Nanog

USP21-LV is a shuttling protein that is primarily localized in the cytosol because it contains a nuclear export signal at its N terminus²⁴. Interestingly, we found that both isoforms of USP21 were re-localized into the nucleus when co-expressed with Nanog in HeLa cells (Supplementary Fig. 7a). These results intrigued us to examine whether USP21 has other functions in the nucleus by interacting with Nanog. USP21 has been reported to deubiquitinate the monoubiquitination of H2A at K119 (H2AK119Ub1)²⁵, this deubiquitination is required for polycomb repressive complex 1-mediated gene repression and plays critical roles in maintaining the stemness of ESCs^26,27,28. Thus, we hypothesized that Nanog might recruit USP21 to gene promoters where it deubiquitinates H2A and regulates gene expression.

To test this hypothesis, we examined whether the level of H2AK119Ub1 was affected during mESC differentiation. H2AK119Ub1 was significantly increased in mESCs cultured under LIF-free conditions and was accompanied by the downregulation of USP21 and Nanog. The monoubiquitination of H2A at K119 has been reported to reduce the level of H3K4me3 (ref. 25). In agreement with this observation, we found that the level of H3K4me3 was decreased, and that of H3K27me3 increased during in mESCs cultured in LIF-free medium (Fig. 7a). Similar results were obtained in RA-induced differentiation of mESCs (Supplementary Fig. 7b). We also found that H2AK119ub1 was increased with the reduction of H3K4me3 upon USP21 or Nanog knockdown in mESCs (Fig. 7b).

**Figure 7: USP21 is recruited to gene promoters and modulates H2A ubiquitination by interaction with Nanog.**

Next, we tested whether the monoubiquitination of H2A at K119 was regulated by USP21 in mESCs. E14 cell lines expressing Flag-USP21 (WT, S539A or S539D) were generated by insertion of USP21 WT or mutants at the Rosa26 locus with CRISPR/Cas9 and homologous recombination (Supplementary Fig. 7c). Cells expressing WT or S539A USP21, but not S539D USP21, showed significantly reduced levels of H2AK119ub1 (Supplementary Fig. 7d). Genome-wide USP21 and Nanog ChIP sequencing (ChIP-Seq) revealed that ∼60% of USP21 target genes were co-bound by Nanog (Fig. 7c). These data suggest that Nanog is an important and major cofactor of USP21 in mESCs. These data also suggested that USP21 and Nanog might independently regulate a number of other genes. ChIP–quantitative PCR (qPCR) analysis using an anti-Flag antibody indicated that Flag-USP21 was significantly enriched at Nanog-regulated genes, such as Sall4, Sox2, Lefty1, Cdx2 and Hoxa10 (refs 29, 30) (Fig. 7d). Furthermore, we observed decreased enrichment of ubH2A and increased enrichment of H3K4me3 in the Flag-USP21 expressing mESCs compared with the parental control cells (Fig. 7e; Supplementary Fig. 7e). Conversely, depletion of USP21 or Nanog decreased the enrichment of H3K4me3 to gene promoters (Fig. 7f). Together, these results indicate that in addition to stabilizing Nanog, USP21 is recruited to chromatin by Nanog to deubiquitinate H2A and facilitate Nanog-mediated gene expression in mESCs.

Discussion

Post-translational protein modifications (PTMs) play important roles in controlling the activity, interactions, subcellular localization and stability of their target proteins. Nanog is a critical TF that controls the pluripotency of ESCs and early embryonic development. However, its regulations by PTMs are not well understood. The most well-studied Nanog PTM is phosphorylation³¹. Several Nanog phosphorylation sites have been revealed, and several kinases, including focal adhesion kinase, protein kinase C, ERKs and CDK1, have been reported to phosphorylate Nanog. The phosphorylation of Nanog has been reported to be involved in ESC self-renewal and carcinogenesis^{17,32,33,34,35}.

Ubiquitination is another important form of PTM mediated by ubiquitin ligases and DUBs that regulate protein functions and stability. The protein levels of key pluripotency factors are controlled by the UPS. For example, Oct4 is regulated by Itch, which is a C2-WW-HECT domain E3 ligase, in undifferentiated mESCs³⁶. Oct4 and Sox2 are regulated by the E3 ligase WWP2 during differentiation^37,38. Klf4 is degraded by E3 ubiquitin ligase β-TrCP1 and β-TrCP2 (refs 28, 39). c-Myc is degraded by E3 ligase Fbxw7 during ESC differentiation⁴⁰ and Rex1 is degraded by E3 ligase RNF12 (ref. 41). Although several ubiquitin E3 ligases of stem cell transcription factors (SCTFs) have been identified, little is known about the mechanism and function of the deubiquitinating enzymes that controls the protein levels of these key pluripotency factors in ESC differentiation and function. In this study, we identified USP21 as a DUB that specifically regulates Nanog stability and function in mESCs.

On the basis of our studies and previous studies by other groups, we propose that each SCTF is regulated by both ubiquitination and deubiquitination at the post-translational level. The net balance of the ubiquitination and deubiquitination of SCTFs could have a significant impact on the cell fate determination of stem cells. Increased ubiquitination of SCTFs leads to degradation and induces cell differentiation, whereas dominant deubiquitination of SCTFs stabilizes these TFs, thus promoting the maintenance of stem cells. Our studies also suggest that the protein stability of each SCTF might be controlled by at least a pair of specific DUB and ubiquitin E3 ligase. Therefore, dissecting this paradigm of reciprocal post-translational control, especially ubiquitination and deubiquitination, in stem cell regulatory networks not only advances stem cell biology but also promotes understanding of cancer stem cells.

USP21 can deubiquitinate both the nuclear and cytoplasmic proteins, such as GATA3, RIPK1 and RIG-1 (refs 42, 43, 44). USP21 functions as a negative regulator in antiviral responses through binding and deubiquitinating RIG-1 in the cytosol⁴⁴. USP21 regulates centrosome- and microtubule-associated functions^18,45. Moreover, USP21 affects the transcription of NF-κB p65 through deubiquitinating and stabilizing interleukin-33 in the nucleus⁴⁶. In this study, we provided evidence showing that USP21 specifically deubiquitinated and stabilized Nanog, but not other pluripotency factors (for example, including Sox2 and Oct4) in mESCs. Consequently, USP21 was required for mESC maintenance. Depletion of USP21 promoted the differentiation of mESCs and reduced the efficiency of somatic cell reprogramming. Thus, our study uncovers an essential function of USP21 in mESC stemness maintenance.

We found that USP21 is regulated at both transcriptional and post-translational levels in mESC to regulate Nanog function. At the transcriptional level, the expression of USP21 in mESCs was activated by the LIF/STAT3 pathway, which was critical for the maintenance of mESC and the self-renewal of mESCs. Upon mESC differentiation, the expression of USP21 was significantly downregulated. At the post-translational level, USP21 was phosphorylated by ERKs induced by differentiation cues. This phosphorylation reduced the binding of USP21 to Nanog and led to Nanog degradation. These data suggest that regulation of Nanog by USP21 is a precise and important event to determine the mESC fate.

The ERK signalling pathway has been shown to negatively regulate the stability of several pluripotent factors. For example, ERKs phosphorylate Klf4 and induce Klf4 degradation by promoting its binding to β-TrCP1 or β-TrCP2 (ref. 47), ERK2 also negatively regulates the stability of Klf2 (ref. 48). In addition, Nanog is a target of ERK1 and phosphorylation by ERK1 induces Nanog degradation via the E3 ligase FBXW8 (ref. 17). Although it is well established that the LIF–STAT3 pathway is essential for the maintenance of mESC, it is also reported that LIF can stimulate the ERK pathway and promote mESCs differentiation. Inhibition of ERK activation enhances mESCs response to LIF and maintains mESCs self-renewal⁴⁹. Here we demonstrated that ERK1 regulated Nanog stability by phosphorylating its DUB USP21 at S539, which impaired the interaction between USP21 and Nanog, and led to increased ubiquitination and degradation of Nanog. All these studies strongly support that ERK signalling is an essential negative regulatory mechanism that controls mESC stemness at multiple levels.

USP21 has been reported to deubiquitinate histone H2A and activate transcriptional initiation via trans-histone cross-talk with H3K4 di- and trimethylation during hepatocyte regeneration²⁵. H2A ubiquitination and chromatin compaction combine to mediate the polycomb repressive complex 1-dependent repression of genes that are crucial for the maintenance of ESC identity⁵⁰. However, limited information is available concerning the role of USP21 in the regulation of histone function and gene expression. We demonstrated that USP21 not only deubiquitinated Nanog by direct interaction but also regulated H2A ubiquitination through being recruited to histones by Nanog.

Above all, our data demonstrate that USP21 is necessary to maintain the self-renewal of mESCs, and that depletion of Usp21 leads to enhanced differentiation. These results indicate that Usp21 depletion may result in abnormality in development. It is most surprising to note that a recent report showed that Usp21-KO mice developed normally and did do not show obvious phenotypes⁵¹. The discrepancy is probably caused by the compensation of other DUBs for Nanog during early development in the Usp21-KO mice. A recent study has shown that Egf17 knockdown leads to the abnormal development of blood vessels, whereas Egf17-KO does not affects blood vessel growth; these results indicate knockdown or KO of the same gene might lead to different consequences⁵². It has been reported that transient downregulation of Nanog in mESCs appears to direct cells towards differentiation, but mESCs with genetic deletion of Nanog can self-renew indefinitely, although they are more prone to differentiation⁵³. This is consistent with the report showing that Stat3-null mESCs can also self-renewal under 2i culture conditions¹⁵. However, these Nanog-KO mESCs failed to form mature germ cells⁵³. Another report has found that, although contributing greatly to reprogramming efficiency, Nanog is not absolutely required for the generation of iPSCs, and the Nanog-null iPSCs can contribute to mature germ cells⁵⁴. Another possibility is that Nanog is only transiently required and is quickly degraded around the time of implantation². These somewhat conflicting results indicating more studies are necessary to fully elucidate the functions of Nanog in pluripotency control and development. Thus, although we have shown that the stabilization of Nanog by USP21 is required for the long-term maintenance of mESCs in vitro, their in vivo regulations remain to be discovered.

In summary, we identified USP21 as a specific deubiquitinating enzyme for the pluripotency factor Nanog. USP21 plays a critical role in maintaining the self-renewal of mESCs by stabilizing Nanog and facilitate the chromatin remodelling that favors Nanog-mediated gene transcription. Depletion of USP21 leads to Nanog degradation and mESC differentiation. The regulation of Nanog by USP21 is controlled by both LIF/STAT3 and FGF/ERK signalling pathway: the expression of USP21 is regulated by LIF/STAT3 signalling and the function of USP21 is regulated by ERKs. Our study suggests a regulatory mechanism by which extrinsic signals regulates the fate of mESCs via the deubiquitination of pluripotency factor (Fig. 7g). The dynamic reversibility of the ubiquitin modification and the recent development of general proteasome inhibitors⁵⁵ and specific blockers and activators of ligase functions^56,57 pave the way for the future manipulation of the UPS in ESCs. In particular, this approach may lead to more efficient reprogramming of somatic cells or provide new therapies targeting cancer stem cells.

Methods

Antibodies and reagent

Antibodies were obtained from commercial sources: the anti-HA, anti-Flag and secondary antibodies were obtained from Sigma. The anti-ubiquitin (P4D1 sc-8017, 1:1,000), anti-H2A (L2313, 1:1,000) and anti-H2B (C2312, 1:1,000) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The antibodies against STAT3 (9139S, 1:1,000), P44/42 MAPK (9194S, 1:2,000), p-P44/42 MAPK (4370P, 1:500), p-MAPK/CDK substrates (2325S, 1:500), H3 (4620S, 1:5,000), H3K4me3 (9751S. 1:1,000) and H3K27me3 (9733S, 1:1,000) were obtained from CST. The anti-Oct4 (ab19857, 1:5,000) and anti-USP21 (ab171028, 1:500) antibodies were from Abcam. The antibody against Nanog was from Bethyl (A300-397A, 1:2,000) and CST (8822, 1:1,000). The anti-tubulin (1:3,000) antibody was kindly provided by Dr Dianqing Wu. The antibody against GAPDH (2118s, 1:10,000) was from CST. The anti-p-STAT3 (Y705; 2236-1, 1:1,000) and anti-Sox2 (2683-1, 1:3,000) were from Epitomics. The antibodies against ubH2A (K119; 05-678, 1:500) and ubH2B (K120; 05-1312, 1:500) were purchased from Millipore. Glutathione SepharoseTM 4B (17-0756-01) was obtained from GE Healthcare. The NI-NTA Agarose (30210) was purchased from Qiagen. The rabbit polyclonal antibody against phospho-USP21 was generated using a CNDSRVPp[Ser]PVSEN peptides (Abmart, 1:500). The rabbit polyclonal antibody against USP21 was generated using a CNDSRVSPVSEN peptides by Abgent (1:500). DMEM (amino acid free) was purchased from Genetimes Technology, and fetal bovine serum (FBS; 10099), GlutaMAX (100 × ; 1293468), non-essential amino acids (NEAA; 100 × ; 11140), β-mercaptoethanol (741953), penicillin and streptomycin (15140) were purchased from Gibco. The LIF (ESG1107) and RA (554723) were from Millipore. MG132 (10 μΜ) and cycloheximide (10 μΜ) were from Sigma. MEK inhibitor PD0325901 (1 μΜ) was from Sigma, ERK1/2 inhibitor SCH772984 (10 μΜ) and STAT3 inhibitor xryptotanshinone (10 μΜ) were from Selleck. Recombinant mouse fibroblast growth factor-4 (25 ng ml⁻¹) was from R&D and epidermal growth factor (100 ng ml⁻¹) was from Sigma.

Cell culture and derivation of MEFs

Human embryonic kidney 293T cells (American Type Culture Collection), Henrietta Lacks strain of cancer cells (HeLa cells, American Type Culture Collection) and MEF cells were cultured in DMEM supplemented with 10% heat-inactivated fetal bovine serum at 37 °C in 5% CO₂. OG2 mice, which carry the Rosa26-lacZ allele and a transgenic Oct4 promoter-driven GFP expression, were mated with C57 mice. MEFs were isolated from E12.5 embryos. Gonads and internal organs were removed before MEF isolation. MEFs were grown in DMEM supplemented with 10% (vol/vol) FBS, 2 mM L-glutamax, 0.1 mM NEAAs, 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin. Isolated MEFs in passage 1 were used for further experiments. E14 cells were maintained feeder free on gelatin-coated plastics in mES media (DMEM with 15% FBS, 2 mM GlutaMAX, 0.1 mM NEAAs, 0.1 mM β-mercaptoethanol, 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin supplemented with 1,000 U ml⁻¹ LIF and passaged every 3 days.

Cell transfection

Transfections were carried out by SunbioTrans-EZ for HeLa cells, using calcium phosphate–DNA coprecipitation for HEK293T cells, using X-tremeGENE HP and Lipofectamine 2000 for E14 cells according to the manufacturer’s instructions.

Animal procedures

Male 6-week-old BALB/cA nude mice used in the present study were purchased from National Rodent Laboratory Animal Resources (Shanghai, China). Usp21^Floxp/Floxp mice in a C57BL/6N background were purchased from the Jackson Laboratory (MECN EPD0716_10_H09; Taconic, USA). Animals were caged in the groups of five in a laminar airflow cabinet under specific pathogen-free conditions, fed with sterilized food and water, and kept on a 12-h light/dark cycle. All treatments were administered according to the guidelines of Institution Animal Care and Use Committee and all the protocols were approved by East China Normal University.

Teratoma formation

Approximately 1 × 10⁶ E14 cells were suspended in 200 μl of mES medium and injected into the thigh muscle of male 6-week-old BALB/cA nude mice. The animals were checked two to three times per week. Five to six weeks after injection, the teratomas were permitted to grow to 15 mm in any direction by the Institutional Animal Care and Use Committee at Massachusetts General Hospital and collected, fixed overnight with 4% paraformaldehyde, embedded in paraffin and sectioned. Sections of the teratomas were stained with haematoxylin and eosin, and analysed histologically.

Derivation and characterization of ESCs

Blastocysts were flushed out from the uterus of Usp21^Floxp/Floxp and WT mice E3.5 female mice with a 27.5-G needle connected to 1-ml syringe filled with 1 ml M2 medium. Blastocysts were then transferred to matrigel-coated plates with 200-ml sterile tip and cultured with M2 medium or regular ESC medium for 3–5 days. ESC-like colonies formed from inner cell mass (ICM) outgrowths were handpicked using glass needle and transferred to plates coated with a feeder layer. Several days later, ESC-like colonies were dissociated by 0.25% trypsin and expanded in ESC medium. Genotyping to detect the presence of the USP21^Loxp used the following primer sequences: USP21_55571_F: 5′-GGTGAGGACTTGACAGCACA-3′; USP21-AS1: 5′-CCTAAGGACCCAAGGAAGAA-3′; and USP21 Mutant-AS1: 5′-AACTACAAGGGGATCCCAAG-3′.

Plasmids and virus

DUBs plasmids were obtained from Addgene. Other constructs were generated by standard molecular cloning method. The MEK1^CA was kindly provided by Dr MienChie Hung. The NANOG reporter plasmids were kindly provided by Dr Duanqing Pei⁵⁸. For the NANOG reporter plasmids, oligonucleotides containing the NANOG-binding site (5′-cccttcgccgattaagtacttaag-3′) and a SalI restriction site were chemically synthesized (sense:5′-tcgacacccttcgccgattaagtacttaag-3′; antisense: 5′-tcgacttaagtacttaatcggcgaagggtg-3′). After being 5′-end phosphorylated by T4 kinase and ATP, these oligonucleotides were annealed and ligated to the SalI site of p37tk-luciferase. Positive clones were randomly picked, and the copy numbers of the inserted NANOG-binding site were determined by sequencing. Nanog, OCT4, Sox2, Klf4, ERK1, USP21 and its mutants, USP2, STAT3 and its mutants were cloned into the pCDNA3.1 vector that a Flag or HA tag was fused to the N terminus using standard cloning methods. USP21 basic promoter luciferase reporter and its mutants were cloned into the PGL3 basic vector. GST-tagged ERK1, GST-tagged USP21 and its mutants were cloned into pGEX-4T-1. His-tagged Nanog was cloned into pET28a. Nanog-luciferase-IRES-Renilla was cloned into pCDNA3.1 vector. To generate lentivirus, USP21 and its mutants, and Cre were cloned into the pWPI lentiviral vector. shRNAs were cloned into the FG12 vector. All the vectors were confirmed using DNA sequencing.

Short hairpin RNA design

Target sequences: USP21 shRNA1: 5′-CCCACTTTGAGACGTAGTA-3′; USP21 shRNA2: 5′-CTCCTGTGAAGCTGTGAAT-3′; USP21 shRNA3: 5′-GACCGAGCCAACTTAATGT-3′; USP21 shRNA4 (target USP21 3′UTR): 5′-GTCCCATGTACAGAAACCA-3′; USP2 shRNA: 5′-GCGAATGGCACACTTTCAA-3′; Nanog shRNA: 5′-GCCAACCTGTACTATGTTTAA-3′; Scramble shRNA: 5′-TTCTCCGAACGTGTCACGTTT-3′; STAT3 siRNA1: 5′-GACCAGCAATATAGCCGAT-3′; STAT3 siRNA2: 5′-GGGTCTCGGAAATTTAACA-3′; STAT3 siRNA3 (target STAT3 3′UTR): 5′-GCACTTTCAACCTTGCTAA-3′.

Virus preparation and viral infection

To prepare lentivirus for the knockdown experiments, HEK293T cells were transduced with the FG12 vector and the lentivirus packaging vectors PHCMVG, PMDLg/PRRE and PRSV-Rev using the calcium phosphate–DNA co-precipitation method. To prepare lentivirus for protein expression, the HEK293T cells were transduced with pWPI vectors and the packaging vectors PSPAX2 and pMD2G using the calcium phosphate–DNA co-precipitation method. Medium containing the virus was collected 48 h after transfection. The E14 cells were cultured in the collected viral supernatant in the presence of polybrene (8 μg ml⁻¹).

Luciferase reporter assay

HEK293T cells were transiently co-transfected with firefly luciferase reporter vectors, effector vectors and the renilla luciferase vector. After 36 h, cells were collected in lysis buffer (25 mM Tris-Cl (pH 7.8), 25 mM dithiothreitol (DTT), 2 mM 1,2-diaminocyclo-hoxane N,N,N,N′-tetracetic acid, 10% glycerol and 1% Triton X-100), and luciferase assays were performed using the dual-luciferase reporter assay system (Promega).

Immunoprecipitation and western blotting

Transfected HEK293T cells were lysed in lysis buffer (50 mM Tris-Cl, pH 7.4, 0.5% NP-40, 150 mM NaCl, 1 mM EDTA, 10% glycerophosphate and a cocktail of proteinase inhibitors). After lysis for 30 min, the soluble fraction of the cell lysates was isolated via centrifugation at 12,000 r.p.m. in a microcentrifuge for 15 min at 4 °C. For IP, the cell lysates were centrifuged to remove the cell debris and were incubated in HA-conjugated beads (Abmart) or M2 beads (Sigma) for 2–3 h. Endogenous Nanog was immunoprecipitated using an anti-Nanog polyclonal antibody. The beads were boiled after extensive washing; the proteins were boiled, resolved via SDS–polyacrylamide gel electrophoresis (SDS–PAGE) gel electrophoreses and analysed via immunoblotting. The proteins were detected using the Odyssey system (LI-COR Biosciences). All uncropped western blots can be found in Supplementary Figs 8–10.

Ubiquitination assay

For the in vivo ubiquitination assay using Ni-NTA beads, the cells were transfected with His-ubiquitin. Then, the transfected cells were lysed using denaturing buffer A (6 M guanidine-HCl, 0.1 M Na₂HPO₄/NaH₂PO₄, 10 mM imidazole, pH 8.0), and the ubiquitinated proteins were purified using Ni-NTA beads. The beads were then washed sequentially with buffer A, buffer B (8 M urea, 0.1 M Na₂HPO₄/NaH₂PO₄, pH 8.0, 0.01 M Tris-HCl, pH 8.0, 10 mM β-mercaptoethanol) and buffer C (8 M urea, 0.1 M Na₂HPO₄/NaH₂PO₄, pH 6.3, 0.01 M Tris-HCl, pH 6.3, 10 mM β-mercaptoethanol) containing 0.2% Triton X-100, and buffer C. The washed beads were incubated in 40 μl of elution buffer (200 mM imidazole, 0.15 M Tris-HCl, pH 6.7, 30% glycerol, 5% SDS, 0.72 M β-mercaptoethanol) at room temperature for 30 min. The input fractions and eluates were analysed via western blotting. For endogenous ubiquitinated Nanog, E14 cells were lysed using RIPA lysis buffer (50 mM Tris-HCl, PH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deocycholate, 1% SDS, 1% Triton X-100). After lysis for 30 min, the soluble fraction of the cell lysates was isolated via centrifugation at 12,000 r.p.m. in a microcentrifuge for 15 min at 4 °C. For IP, the cell lysates were centrifuged to remove the cell debris and were incubated with anti-Nanog polyclonal antibody or 2–3 h. The beads were boiled after extensive washing; the proteins were boiled, resolved via SDS–PAGE gel electrophoreses, and analysed via immunoblotting.

To detect the in vitro deubiquitination of Nanog, His-USP21 was purified from Escherichia coli. The in vitro deubiquitination reaction was performed using Flag-Nanog, His-USP21, 50 mM Tris-Cl (PH 8.0), 50 mM NaCl, 1 mM EDTA, 10 mM DTT and 5% glycerol in a final volume of 25 μl at 37 °C for 4 h. The ubiquitinated Nanog proteins were detected via western blotting using an anti-Flag antibody.

GST pull-down assay

The His-Nanog protein was purified from E. coli and incubated in 10 μg of purified GST or GST-USP21 protein. The GST proteins were purified using glutathione sepharose 4B, and the bound Nanog was detected via western blotting.

In vitro phosphorylation assay

Flag-USP21 and Flag-USP21 S539A were expressed in HEK293T cells and purified using M2 beads for immunoprecipitation. After washing twice with kinase buffer (25 mM Tris-HCl (pH 7.5), 5 mM glycerophosphate, 10 mM MgCl₂, 2 mM DTT and 0.1 mM Na₃VO₄), purified Flag-USP21 or Flag-USP21SA were incubated with 2 μg GST-ERK1 in 25 μl of kinase buffer containing 200 mM ATP at 37 °C for 1 h. The reaction was terminated by the addition of 8 μl of 4XLaemmli SDS sample buffer. The reaction mixtures were boiled and separated by SDS–PAGE and phosphorylation USP21 was immunobloted by anti-phosphor-USP21 antibody.

Immunofluorescence

Cells were seeded on fibronectin-coated glass coverslips in 24-well tissue culture plates. After transfection or infection, the cells were rinsed once with PBS and fixed in 4% paraformaldehyde for 15 min at room temperature. The fixed cells were permeabilized using 0.1% Triton X-100 and rinsed twice with PBS. The coverslips were blocked with blocking buffer for 1 h (0.3% bovine serum albumin in PBS) and incubated in a primary antibody in blocking buffer overnight at 4 °C. Next, the coverslips were rinsed twice with blocking buffer and incubated in secondary antibodies for 1 h at room temperature in the dark, followed by tyramide signal amplification. The glass coverslips were mounted using Mowiol and were examined using a Zeiss LSM 510 Meta confocal system.

Mouse iPSC generation

Retrovirus were produced by transfection of plat-E cells with pMXs retroviral vectors containing the coding sequences of mouse Oct4, Sox2, Klf4, Nanog, Nanog K138/K169R mutant, USP21-SV, USP21-LV and USP21-C221A. MEFs were seeded at a density of 150,000 cells per well in a six-well plate 18 h before infection. Virus containing supernatants, supplemented with 4–8 μg ml⁻¹ polybrene, were added onto the plates of MEF cultures and spun at 2,500 r.p.m. for 90 min to ensure their infection. Medium was changed immediately after virus transduction. Two days post virus infection, MEFs were digested into single cells and reseeded at a density of 5,000 or 10,000 cells per well on 24-well plates pre-seeded with irradiated MEF feeders, supplemented with mES medium (DMEM supplemented with 15% FBS, 2 mM L-glutamax, 0.1 mM NEAA, 0.1 mM β-mercaptoethanol 1,000 U ml⁻¹ LIF, 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin). At day 6, culture medium was replaced with KSR medium (KO DMEM supplemented with 15% KO serum replacement, 2 mM L-glutamax, 0.1 mM NEAAs, 0.1 mM β-mercaptoethanol, 1,000 U ml⁻¹ LIF, 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin). Images of representative wells were taken, and AP-positive colonies or GFP-positive colonies were counted at day 12 or day 16 post infection.

Real-time qPCR

Total RNA was Trizol-extracted, column-purified and reverse-transcribed using PrimeScript 1st Strand Cdna Synthesis kit (Takara). Primer sequences used to test gene expression are available in Supplementary Table 1. ChIP was performed according to the manufacturer’s instructions (CST, #9004). For ChIP–qPCR analysis, 1 ng ChIP DNA was used for each PCR. All qPCR analyses were performed using Fast SYBR Green (Takara). Primer sequences are available in Supplementary Table 2.

ChIP sequencing

Library preparation. An amount of 5–10 ng of ChIP DNA was used for the generation of libraries for deep-sequencing using the NEXTflex ChIP-Seq kit. In brief, the DNA was end-repaired following adding an A-base to the end-repaired DNA fragments. Illumina adaptors (regular or multiplex) were ligated to the ChIP DNA fragments and 100–300 bp of size fractions were excised from 2% agarose gel. Adaptor-modified fragments library preparation was validated in bioanalyser for quantity and size before putting into sequencing (The Beijing Genomics Institute).

ChIP-Seq data analysis

All fastq files including previously published samples were aligned to the Mouse reference genome (UCSC, mm10 assembly) using Bowtie (settings: -p 20 -S -m 1). Only uniquely mapped reads were retained for analysis in this study. Peak detection was performed using the MACS version 2 software⁵⁹ with default settings (except: ‘--broad --format=BAM -g mm’). The ‘—broad’ parameters causes MACS to output both ‘normal’ and ‘broad’ peaks, where normal peaks are merged into broad peaks if they are closely located on the same chromosome. All the analyses were based on broad peaks. IgG samples were used as peak calling controls. Chromosomal positions were annotated to the RefSeq database (mm9) using the UCSC refFlat table⁶⁰. To avoid redundancy, only the longest transcript of each gene was used. All chromosomal interval overlaps were performed using ‘intersectBed’ script from UCSC. To create a bigwig (that is, wiggle format) file, each mapped read was first extended in the 3′-direction to a total length of 250 bases (our estimated fragment length). The modified alignment file was then transformed to a bedGraph file, where densities were scaled according to millions of uniquely mapped reads, before finally being converted to a bigwig file. All the target genes were analysed by the BETA-minus of cistrome. Peaks considered to contribute to the genes: 10,000 bp; the distance from gene transcription start site within which peaks will be selected: 100,000 bp. The USP21 peaks were used as input to Cistrome BETA-minus tool with default settings to predict the target genes. The target genes of Nanog were predicted the same way. The ChIP-Seq data have been submitted to the Gene Expression Omnibus (GEO) database (GSE79890). The Nanog-binding site used to test in our study are showed in Supplementary Table 3. Nanog ChIP-Seq data refer to GEO DataSets under accession code GSE44764 (ref. 61).

Generation of the Flag-USP21 stable cell line

Flag-USP21 WT/SD/SA mESC stable cell line was generating using CRISPR-CAS methods⁶². In brief, guide RNA (gRNA) expression vectors were constructed for pGS3-T7-gRNA. The sequence of gRNA (target to Rosa26) is 5′-TCGTGATCTGCAACTCCAG-3′. Flag-USP21 was ligated to the Sall and MluI site of Rosa26-puro-CAG donor vector. Then, we inserted the Flag-USP21 into the mouse Rosa26 safe harbour site via homologous recombination. An amount of 40 μg donor vector was electroporated into 0.3–0.4 million E14 cells along with 5 μg each of the Rosa26 gRNA and pCas9-GFP plasmid (Amaxa kit R, program A-24; Lonza). One week after electroporation, cells were replated into 150-mm dishes at clonal density. Individual colonies were subsequently selected under the fluorescence microscope and expanded in 24-well dishes. Analysis of Flag-USP21 stable cell line by western blotting.

ESC differentiation assays

The differentiation of E14 cells was induced by the withdrawal of LIF or treatment with 10 μM RA.

MEK–ERK pathway stimulation in mESCs

E14 cells were cultured in N2B27 medium overnight and then treated with 25 ng ml⁻¹ mFGF4 for 12 h before collecting for immunoblotting analysis.

Alkaline phosphatase

For AP staining, the cells were fixed with 4% paraformaldehyde in PBS for 30 s and rinsed once with PBS. Detection was performed using a leukocyte AP kit (Sigma, catalogue no 85L3R) according to the manufacturer’s protocol.

Statistical analysis

Statistical analyses were performed with a two-tailed unpaired Student’s t-test and experiments that contain multiple samples were performed with analysis of variance followed by a multiple comparisons test. The data are presented as the means±s.d. The mean is calculated from truly independent experiments. The values of P<0.05 were considered statistically significant.

Data availability

The authors declare that all data supporting the findings of this study are available within the article and its Supplementary Information files or from the corresponding author on reasonable request. Usp21 ChIP-Seq data have been deposited in the NCBI GEO database under accession code GSE79890 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE79890).

Additional information

How to cite this article: Jin, J. et al. The deubiquitinase USP21 maintains the stemness of mouse embryonic stem cells via stabilization of Nanog. Nat. Commun. 7, 13594 doi: 10.1038/ncomms13594 (2016).

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

References

Do, J. T. & Scholer, H. R. Regulatory circuits underlying pluripotency and reprogramming. Trends Pharmacological Sci. 30, 296–302 (2009).
Article CAS Google Scholar
Chambers, I. et al. Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643–655 (2003).
Article CAS Google Scholar
Mitsui, K. et al. The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 113, 631–642 (2003).
Article CAS Google Scholar
Chazaud, C., Yamanaka, Y., Pawson, T. & Rossant, J. Early lineage segregation between epiblast and primitive endoderm in mouse blastocysts through the Grb2-MAPK pathway. Dev. Cell 10, 615–624 (2006).
Article CAS Google Scholar
Ivanova, N. et al. Dissecting self-renewal in stem cells with RNA interference. Nature 442, 533–538 (2006).
Article ADS CAS Google Scholar
Yu, J. et al. Induced pluripotent stem cell lines derived from human somatic cells. Science 318, 1917–1920 (2007).
Article ADS CAS Google Scholar
Niwa, H., Miyazaki, J. & Smith, A. G. Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nat. Genet. 24, 372–376 (2000).
Article CAS Google Scholar
Adachi, K., Suemori, H., Yasuda, S. Y., Nakatsuji, N. & Kawase, E. Role of SOX2 in maintaining pluripotency of human embryonic stem cells. Genes Cells 15, 455–470 (2010).
CAS PubMed Google Scholar
Pan, G. & Thomson, J. A. Nanog and transcriptional networks in embryonic stem cell pluripotency. Cell Res. 17, 42–49 (2007).
Article CAS Google Scholar
Niwa, H., Ogawa, K., Shimosato, D. & Adachi, K. A parallel circuit of LIF signalling pathways maintains pluripotency of mouse ES cells. Nature 460, 118–122 (2009).
Article ADS CAS Google Scholar
Suzuki, A. et al. Nanog binds to Smad1 and blocks bone morphogenetic protein-induced differentiation of embryonic stem cells. Proc. Natl Acad. Sci. USA 103, 10294–10299 (2006).
Article ADS CAS Google Scholar
Santostefano, K. E., Hamazaki, T., Pardo, C. E., Kladde, M. P. & Terada, N. Fibroblast growth factor receptor 2 homodimerization rapidly reduces transcription of the pluripotency gene Nanog without dissociation of activating transcription factors. J. Biol. Chem. 287, 30507–30517 (2012).
Article CAS Google Scholar
Luo, Y., Lim, C. L., Nichols, J., Martinez-Arias, A. & Wernisch, L. Cell signalling regulates dynamics of Nanog distribution in embryonic stem cell populations. J. R. Soc. Interface 10, 20120525 (2012).
Article Google Scholar
Xu, R. H. et al. NANOG is a direct target of TGFbeta/activin-mediated SMAD signaling in human ESCs. Cell Stem Cell 3, 196–206 (2008).
Article CAS Google Scholar
Ying, Q. L. et al. The ground state of embryonic stem cell self-renewal. Nature 453, 519–523 (2008).
Article ADS CAS Google Scholar
Miyanari, Y. & Torres-Padilla, M. E. Control of ground-state pluripotency by allelic regulation of Nanog. Nature 483, 470–473 (2012).
Article ADS CAS Google Scholar
Kim, S. H. et al. ERK1 phosphorylates Nanog to regulate protein stability and stem cell self-renewal. Stem Cell Res. 13, 1–11 (2014).
Article CAS Google Scholar
Okuda, H. et al. The USP21 short variant (USP21SV) lacking NES, located mostly in the nucleus in vivo, activates transcription by deubiquitylating ubH2A in vitro. PLoS ONE 8, e79813 (2013).
Article ADS Google Scholar
Silva, J. et al. Nanog is the gateway to the pluripotent ground state. Cell 138, 722–737 (2009).
Article CAS Google Scholar
Evans, M. J. & Kaufman, M. H. Establishment in culture of pluripotential cells from mouse embryos. Nature 292, 154–156 (1981).
Article ADS CAS Google Scholar
Schwarz, B. A., Bar-Nur, O., Silva, J. C. & Hochedlinger, K. Nanog is dispensable for the generation of induced pluripotent stem cells. Curr. Biol. 24, 347–350 (2014).
Article CAS Google Scholar
Huang, G., Yan, H., Ye, S., Tong, C. & Ying, Q. L. STAT3 phosphorylation at tyrosine 705 and serine 727 differentially regulates mouse ESC fates. Stem Cells 32, 1149–1160 (2014).
Article CAS Google Scholar
Li, Z. et al. BMP4 signaling acts via dual-specificity phosphatase 9 to control ERK activity in mouse embryonic stem cells. Cell Stem Cell 10, 171–182 (2012).
Article CAS Google Scholar
Garcia-Santisteban, I., Banuelos, S. & Rodriguez, J. A. A global survey of CRM1-dependent nuclear export sequences in the human deubiquitinase family. Biochem. J. 441, 209–217 (2012).
Article CAS Google Scholar
Nakagawa, T. et al. Deubiquitylation of histone H2A activates transcriptional initiation via trans-histone cross-talk with H3K4 di- and trimethylation. Genes Dev. 22, 37–49 (2008).
Article CAS Google Scholar
Kallin, E. M. et al. Genome-wide uH2A localization analysis highlights Bmi1-dependent deposition of the mark at repressed genes. PLoS Genet. 5, e1000506 (2009).
Article Google Scholar
Boyer, L. A. et al. Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 441, 349–353 (2006).
Article ADS CAS Google Scholar
Kim, J., Chu, J., Shen, X., Wang, J. & Orkin, S. H. An extended transcriptional network for pluripotency of embryonic stem cells. Cell 132, 1049–1061 (2008).
Article CAS Google Scholar
Loh, Y. H. et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat. Genet. 38, 431–440 (2006).
Article CAS Google Scholar
Wu, X., Johansen, J. V. & Helin, K. Fbxl10/Kdm2b recruits polycomb repressive complex 1 to CpG islands and regulates H2A ubiquitylation. Mol. Cell 49, 1134–1146 (2013).
Article CAS Google Scholar
Yates, A. & Chambers, I. The homeodomain protein Nanog and pluripotency in mouse embryonic stem cells. Biochem. Soc. Trans. 33, 1518–1521 (2005).
Article CAS Google Scholar
Moretto-Zita, M. et al. Phosphorylation stabilizes Nanog by promoting its interaction with Pin1. Proc. Natl Acad. Sci. USA 107, 13312–13317 (2010).
Article ADS CAS Google Scholar
Ho, B. et al. Nanog increases focal adhesion kinase (FAK) promoter activity and expression and directly binds to FAK protein to be phosphorylated. J. Biol. Chem. 287, 18656–18673 (2012).
Article CAS Google Scholar
Xie, X. et al. Phosphorylation of Nanog is essential to regulate Bmi1 and promote tumorigenesis. Oncogene 33, 2040–2052 (2014).
Article CAS Google Scholar
Brumbaugh, J. et al. NANOG is multiply phosphorylated and directly modified by ERK2 and CDK1 in vitro. Stem Cell Rep. 2, 18–25 (2014).
Article CAS Google Scholar
Liao, B. et al. Itch, an E3 ligase of Oct4, is required for embryonic stem cell self-renewal and pluripotency induction. J. Cell Physiol. 228, 1443–1451 (2013).
Article CAS Google Scholar
Xu, H. et al. WWP2 promotes degradation of transcription factor OCT4 in human embryonic stem cells. Cell Res. 19, 561–573 (2009).
Article CAS Google Scholar
Fang, L. et al. A methylation-phosphorylation switch determines Sox2 stability and function in ESC maintenance or differentiation. Mol. Cell 55, 537–551 (2014).
Article CAS Google Scholar
Ye, Y. et al. Polyubiquitin binding and cross-reactivity in the USP domain deubiquitinase USP21. EMBO Rep. 12, 350–357 (2011).
Article CAS Google Scholar
Buckley, S. M. et al. Regulation of pluripotency and cellular reprogramming by the ubiquitin-proteasome system. Cell Stem Cell 11, 783–798 (2012).
Article CAS Google Scholar
Gontan, C. et al. RNF12 initiates X-chromosome inactivation by targeting REX1 for degradation. Nature 485, 386–390 (2012).
Article ADS CAS Google Scholar
Zhang, J. et al. Identification of the E3 deubiquitinase ubiquitin-specific peptidase 21 (USP21) as a positive regulator of the transcription factor GATA3. J. Biol. Chem. 288, 9373–9382 (2013).
Article CAS Google Scholar
Xu, G. et al. Ubiquitin-specific peptidase 21 inhibits tumor necrosis factor alpha-induced nuclear factor kappaB activation via binding to and deubiquitinating receptor-interacting protein 1. J. Biol. Chem. 285, 969–978 (2010).
Article CAS Google Scholar
Fan, Y. et al. USP21 negatively regulates antiviral response by acting as a RIG-I deubiquitinase. J. Exp. Med. 211, 313–328 (2014).
Article CAS Google Scholar
Urbe, S. et al. Systematic survey of deubiquitinase localization identifies USP21 as a regulator of centrosome- and microtubule-associated functions. Mol. Biol. Cell 23, 1095–1103 (2012).
Article CAS Google Scholar
Tao, L. et al. Deubiquitination and stabilization of IL-33 by USP21. Int. J. Clin. Exp. Pathol. 7, 4930–4937 (2014).
CAS PubMed PubMed Central Google Scholar
Kim, M. O. et al. ERK1 and ERK2 regulate embryonic stem cell self-renewal through phosphorylation of Klf4. Nat. Struct. Mol. Biol. 19, 283–290 (2012).
Article CAS Google Scholar
Yeo, J. C. et al. Klf2 is an essential factor that sustains ground state pluripotency. Cell Stem Cell 14, 864–872 (2014).
Article CAS Google Scholar
Burdon, T., Stracey, C., Chambers, I., Nichols, J. & Smith, A. Suppression of SHP-2 and ERK signalling promotes self-renewal of mouse embryonic stem cells. Dev. Biol. 210, 30–43 (1999).
Article CAS Google Scholar
Endoh, M. et al. Histone H2A mono-ubiquitination is a crucial step to mediate PRC1-dependent repression of developmental genes to maintain ES cell identity. PLoS Genet. 8, e1002774 (2012).
Article CAS Google Scholar
Pannu, J. et al. Ubiquitin specific protease 21 is dispensable for normal development, hematopoiesis and lymphocyte differentiation. PLoS ONE 10, e0117304 (2015).
Article Google Scholar
Rossi, A. et al. Genetic compensation induced by deleterious mutations but not gene knockdowns. Nature 524, 230–233 (2015).
Article ADS CAS Google Scholar
Chambers, I. et al. Nanog safeguards pluripotency and mediates germline development. Nature 450, 1230–1234 (2007).
Article ADS CAS Google Scholar
Carter, A. C., Davis-Dusenbery, B. N., Koszka, K., Ichida, J. K. & Eggan, K. Nanog-independent reprogramming to iPSCs with canonical factors. Stem Cell Rep. 2, 119–126 (2014).
Article CAS Google Scholar
Adams, J. & Kauffman, M. Development of the proteasome inhibitor Velcade (Bortezomib). Cancer Invest. 22, 304–311 (2004).
Article CAS Google Scholar
Aghajan, M. et al. Chemical genetics screen for enhancers of rapamycin identifies a specific inhibitor of an SCF family E3 ubiquitin ligase. Nat. Biotechnol. 28, 738–742 (2010).
Article CAS Google Scholar
Orlicky, S. et al. An allosteric inhibitor of substrate recognition by the SCF(Cdc4) ubiquitin ligase. Nat. Biotechnol. 28, 733–737 (2010).
Article CAS Google Scholar
Pan, G. & Pei, D. The stem cell pluripotency factor NANOG activates transcription with two unusually potent subdomains at its C terminus. J. Biol. Chem. 280, 1401–1407 (2005).
Article CAS Google Scholar
Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008).
Article Google Scholar
Rhead, B. et al. The UCSC Genome Browser database: update 2010. Nucleic Acids Res. 38, D613–D619 (2010).
Article CAS Google Scholar
Kartikasari, A. E. et al. The histone demethylase Jmjd3 sequentially associates with the transcription factors Tbx3 and Eomes to drive endoderm differentiation. EMBO J. 32, 1393–1408 (2013).
Article CAS Google Scholar
Yang, H. et al. One-step generation of mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineering. Cell 154, 1370–1379 (2013).
Article CAS Google Scholar

Download references

Acknowledgements

We thank Drs Dianqing Wu, Xiaoqing Zhang, Duanqing Pei, Yin Jing and Mienchie Hung for kindly providing reagents. This work was supported by grants from the Ministry of Science and Technology of China (973 program 2012CB910404, 2015CB964503), the Chinese Academy of Sciences (XDA01040301), the National Natural Science Foundation of China (91519322, 91440104, 31222037, 31171338, 81502559, 81425024 and 31371511) and the Doctoral Fund of Ministry of Education of China (20130076110022).

Author information

Jiali Jin and Jian Liu: These authors contributed equally to this work

Authors and Affiliations

Shanghai Key Laboratory of Regulatory Biology, Institute of Biomedical Sciences and School of Life Sciences, East China Normal University, 500 Dongchuan Road, Shanghai, 200241, China
Jiali Jin, Cong Chen, Cong Jiang, Hongshang Chu, Weijuan Pan, Xinbo Wang & Ping Wang
Chinese Academy of Sciences Key Laboratory of Receptor Research, National Center for Drug Screening, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, 201203, China
Jian Liu & Xin Xie
University of Chinese Academy of Sciences, No. 19A Yuquan Road, Beijing, 100049, China
Jian Liu & Xin Xie
Department of Central Laboratory, Shanghai Tenth People’s Hospital of Tongji University, School of Life Science and Technology, Tongji University, Shanghai, 200072, China
Zhenping Liu, Cizhong Jiang & Ping Wang
State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Collaborative Innovation Center for Cancer Medicine, Beijing, 100850, China
Lingqiang Zhang
Key Laboratory of Molecular Virology and Immunology, Unit of Molecular Immunology, Institute Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, 200031, China
Bin Li
Department of Clinical Medicine, Shanghai Tenth People's Hospital of Tongji University, Tongji University, Shanghai, 200072, China
Xin Ge

Authors

Jiali Jin
View author publications
You can also search for this author in PubMed Google Scholar
Jian Liu
View author publications
You can also search for this author in PubMed Google Scholar
Cong Chen
View author publications
You can also search for this author in PubMed Google Scholar
Zhenping Liu
View author publications
You can also search for this author in PubMed Google Scholar
Cong Jiang
View author publications
You can also search for this author in PubMed Google Scholar
Hongshang Chu
View author publications
You can also search for this author in PubMed Google Scholar
Weijuan Pan
View author publications
You can also search for this author in PubMed Google Scholar
Xinbo Wang
View author publications
You can also search for this author in PubMed Google Scholar
Lingqiang Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Bin Li
View author publications
You can also search for this author in PubMed Google Scholar
Cizhong Jiang
View author publications
You can also search for this author in PubMed Google Scholar
Xin Ge
View author publications
You can also search for this author in PubMed Google Scholar
Xin Xie
View author publications
You can also search for this author in PubMed Google Scholar
Ping Wang
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

J.J., J.L., X.X. and P.W. conceived the project and designed the experiments. J.J., J.L., C.C., C.J, H.C., W.P. and X.W. performed the experiments. Z.L. and C.J. performed the bioinformatics analyses of ChIP-Seq data. B.L. generated the conditional KO Usp21^Floxp/Floxp mice. L.Z. generated Usp2 constructs. J.J., J.L., H.C., C.J., B.L., L.Z., X.G., X.X. and P.W. analysed the data. J.J., J.L., X.G., X.X. and P.W. wrote the manuscript.

Corresponding authors

Correspondence to Xin Xie or Ping Wang.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Supplementary information

Supplementary Information

Supplementary Figures 1-10, Supplementary Table 1-3. (PDF 1934 kb)

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Reprints and permissions

About this article

Cite this article

Jin, J., Liu, J., Chen, C. et al. The deubiquitinase USP21 maintains the stemness of mouse embryonic stem cells via stabilization of Nanog. Nat Commun 7, 13594 (2016). https://doi.org/10.1038/ncomms13594

Download citation

Received: 26 December 2015
Accepted: 18 October 2016
Published: 25 November 2016
DOI: https://doi.org/10.1038/ncomms13594

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.