Introduction

Embryonic Stem Cells (ESCs) pluripotency is characterized by the capacity to generate all somatic and germline lineages both in vitro and in vivo, depending on the culture conditions. When cultured in serum plus Leukemia Inhibitory Factor (LIF) (Serum/LIF), murine ESCs (mESCs) include cells interconverting to different metastable states ¹ that either resemble preimplantation blastocyst (naïve pluripotency) or post-implantation (primed pluripotency) embryonic stages ^2–5. The heterogeneity of mESCs in Serum/LIF is reduced by culturing them in the presence of GSK3β and MEK1/2 inhibitors (CHIR99021 and PD0325901, respectively) along with LIF (2i/LIF). Under these conditions, mESCs enter a ground state of naïve pluripotency characterized by self-renewal activity while suppressing any pro-differentiating signals ^6,7. This pluripotent state closely resembles the inner cell mass of the E4.0 mouse preimplantation blastocyst ⁸. After implantation, pluripotent capability of the mouse embryo (E5.5) becomes restricted to Epiblast Stem Cells (EpiSCs), which represent a primed pluripotent state, characterized by coexpression of pluripotency and lineage specification markers ^9,10. EpiSCs can give rise to cells from the three germ layers (endoderm, mesoderm and ectoderm), albeit their contribution to chimeras or to germ cells is greatly compromised ^4,11. EpiSCs can be derived in vitro from naïve pluripotent cells when cultured in a medium supplemented with Nodal and fibroblast growth factor (FGF) ¹². Different epigenetic features distinguish naïve from primed pluripotent states ¹³, such as changes in the X chromosome inactivation in female cells ^12,14, histone post-translational modifications ¹⁰ or a global DNA hypomethylation profile observed in naïve pluripotent stem cells ^15–17. These epigenetic changes are a prerequisite for the subsequent activation of gene circuits associated with pluripotency exit and germ layer specification ^9,10. Crucial regulatory elements for the transition from naïve to primed pluripotency state are located in distal DNA regions or enhancers to facilitate the resolution toward the primed state ^18,19. Therefore, identifying novel players to fine-tune the equilibrium between ground and primed states is crucial for preserving both pluripotency stability and differentiation capability in mESCs.

There is now strong evidence that inflammatory signals are critical for stem cell development ^20–22, with NF-κB being the main effector. Different strategies to inhibit or attenuate NF-κB in ESCs lead to increase pluripotency gene expression and impair cell differentiation ^23,24, or it facilitates reprogramming to induced pluripotent stem cells (iPSC) ²⁵. Canonical NF-κB signaling is triggered by the IKK kinase complex, which induce the phosphorylation and subsequent degradation of the NF-κB inhibitor, IκBα, leading to nuclear translocation of the NF-κB factors (e.g. p50/p65) ²⁶. However, we have previously identified a nuclear IκBα function, which is critical for tissue stem cell homeostasis and the proper differentiation of epidermal, intestinal and hematopoietic stem cells ^27–30. IκBα chromatin function is mediated by the interaction with histone deacetylases (HDACs), core elements of the Polycomb Repressor Complex 2 (PRC2) and histone H2A and H4 ^27,29,31.

We have now investigated the role of IκBα in the context of pluripotent stem cells. Whereas cytoplasmic IκBα is detected in the different states and its levels decrease after differentiation, in agreement with NF-κB activity as a promoter of differentiation, we specifically detected chromatin-bound IκBα in the naïve pluripotent stem cells. Importantly, IκBα depletion in ESCs did not affect NF-κB activity but it stabilizes the naïve (2i/LIF-like) state of pluripotency under Serum/LIF conditions, displaying an epigenetic rewiring, including changes in the PRC2-dependent H3K27me3 histone mark and impacting in proximal and distal regulation of pluripotent genes. Using a new Separation-Of-Function (SOF) IκBα mutants³², we have now definitively demonstrated that chromatin-bound IκBα is required for naïve pluripotency exit of mESCs in an NF-κB-independent manner. Overall, we identified IκBα as a novel key player in the regulation of the exit from naïve-to-primed pluripotency and essential for the activation of gene programs involved in germ layer specification.

Results

IκBα and NF-κB have opposite expression dynamics in naïve pluripotent cells

Due to the prominent role of NF-κB signaling in mESCs regulation, we analyzed the expression pattern of the different NF-κB members including the IκB inhibitors in the three defined pluripotent stages: ground state of naïve pluripotency (2i/LIF), naïve pluripotency (Serum/LIF), and primed pluripotency (Epiblast Stem Cells or EpiSCs) ⁵ (Figure 1A). In agreement with NF-κB being a pro-differentiation factor, canonical NF-κB inhibitors including IκBα (Nfkbia), IκBβ (Nfkbib) and IκBε (Nfkbie) were all expressed in the naïve pluripotency state (both 2i/LIF and Serum/LIF), with Nfkbia levels gradually reduced upon naïve pluripotency exit towards EpiSCs (Figure 1B). In contrast, Rela, Nfkb1 and Nfkb2 genes, codifying for the NF-κB factors p65/RelA, p105 and p100, respectively, showed opposite expression dynamics increasing the levels in EpiSCs (Figure 1C). We further characterized the distribution of IκBα in the cytoplasm, nucleus and chromatin fractions of the three states of pluripotency (Figure 1D). Total IκBα protein levels correlated with the detected RNA levels in the three different cellular states being highest in the 2i/LIF condition (Figure 1B), with a large accumulation in the cytoplasm. A fraction of IκBα protein was detected in the chromatin of naïve pluripotent cells, being highest in cells cultured in Serum/LIF conditions compared to 2i/LIF and absent in the EpiSCs. (Figure 1D). The dynamics of IκBα RNA and protein levels from naïve pluripotent ESCs to EpiSCs was further corroborated by analysis of an additional database that includes both naïve and primed pluripotent states³³ (Figure 1E).

Figure 1.

Next, we studied IκBα expression in differentiated embryoid bodies (EBs) from mESCs (Serum/LIF) (see materials and methods) and analyzed the expression levels of different pluripotency, differentiation-linked, and NF-κB genes by RNA sequencing. Nfkbia (IκBα gene) was expressed at higher levels specifically in mESCs (Serum/LIF) whereas cells undergoing differentiation (48h and 96h) displayed a reduction in its expression levels (Figure 1F). The expression pattern of Nfkbia was similar to the pluripotency genes whereas most of the NF-κB genes were upregulated upon differentiation, clustering together with developmental genes, as previously described ²⁴. Notably, several polycomb elements were highly expressed in mESCs, consistent with the possibility that chromatin-bound IκBα modulates PRC2 activity ²⁷ in the pluripotent state.

Taken together, these data show that, unlike the other NF-kB inhibitors and target genes, the expression of Nfkbia is high in the naive pluripotent state, with IκBα protein present in the cytoplasm and chromatin of naive pluripotent cells. Moreover, whereas canonical NF-κB subunits are primarily absent, we detected several Polycomb elements expressed in mESCs.

IκBα is required in mESCs to exit the naïve pluripotency state

To investigate the functional implication of IκBα in naïve pluripotency, we knocked-out (KO) IκBα protein in mESCs by CRISPR-Cas9 (IκBα-KO mESCs) (Figure S1A and materials and methods). Notably, IκBα KO mESCs have an impaired ability to undergo differentiation and cannot properly switch off the pluripotency program after 216h (9 days) of pro-differentiation signals, as it is demonstrated by high levels of pluripotency markers OCT3/4 and NANOG (Fig 1A and Fig S1C), higher number of alkaline phosphatase cells (Fig 1B and Fig S1) and higher percentage of SSEA-1⁺ cells (Fig S1B) within the 216h EBs. These results were also supported by the detection of increased expression level of the naïve pluripotency genes Pou5f1 (Oct3/4 gene), Gbx2, Klf2, Sox2, Zfp42 (Rex1) and Nanog in IκBα KO EBs after 216h of differentiation (Figure S1B).

To shed light into the molecular mechanisms governing the differentiation impairment in IκBα KO EBs, we performed RNA-seq of IκBα WT and IκBα KO mESCs in the first stages of the embryoid body differentiation. Analysis of the data indicated that IκBα KO EBs cannot successfully evolve towards differentiation trajectories, as shown by principal component analysis (PCA) of IκBα-WT and IκBα-KO EBs at 48h and 96h in differentiation media (Figure 2C). Analysis of naïve pluripotency markers at 96h (Gab1, Sox2, Esrrb, Zfp42, Dppa4, Klf2, Prdm14, Klf4, Tfcp2l1) also showed higher expression levels in IκBα KO compared to their IκBα WT counterparts (Figures 2D and S1F). Defective differentiation capacity of IκBα KO EBs involved programs associated with specification of all 3 germ layers (endoderm, mesoderm, and ectoderm) (Figures 2E-G). Moreover, IκBα KO EBs were smaller in size (Figure S1E), in agreement with the reduced developmental potential of IκBα-depleted mESCs ³⁴. To further study the requirement of IκBα in pluripotency exit in vivo, we established teratomas by subcutaneous injection of IκBα-WT and IκBα-KO mESCs (Figure 2H and materials and methods) into NSG (NOD.Cg-Prkdcˢᶜⁱᵈ Il2rgᵗᵐ¹ᵂʲˡ/SzJ) mice. IκBα-KO teratomas showed a decreased differentiation potential, as indicated by higher number of OCT4⁺ cells 6 weeks after injection (Figures 2H-J). These observations further confirm the essential role of IκBα in the exit from the pluripotency state not only in vitro but also under in vivo differentiation signals. To understand the conditions in which IκBα is required for pluripotency exit, we differentiated IκBα-WT and IκBα-KO mESCs cultured in Serum/LIF (naïve pluripotency) towards EpiSCs (primed pluripotency) (Figure 3A, upper panel). IκBα-KO cells were not committed towards the primed pluripotency stage and maintained an elevated expression levels of naïve pluripotency genes (Dppa3, Nanog, Sox2, Rex1, Klf2, Klf4, Gbx2, Tbx3) (Figure 3A, bottom panel). Additionally, IκBα-KO cells retained a naïve pluripotent morphology, forming tight clusters that resemble cells cultured in 2i/LIF medium (Figure S2A).

Figure 2.

Figure 3.

mESCs cultured in serum/LIF are highly heterogeneous, comprising a mixture of cell states that resemble developmental transitions from preimplantation (naïve pluripotency) to postimplantation (primed pluripotency) embryo ^2,3. To investigate how IκBα deficiency was affecting the exit of mESCs from naïve pluripotency, we performed a Gene Set Enrichment Analysis (GSEA) using RNA-seq data from IκBα-KO and IκBα-WT mESCs cultured in Serum/LIF against gene signatures for 2i/LIF and Serum/LIF pluripotency ³⁵. GSEA results revealed a significant enrichment in Serum/LIF IκBα-KO mESCs transcriptome for genes specifically expressed in the naïve ground state (2i/LIF), while they negatively correlate with mESCs cultured in Serum/LIF (Figure 3B) ³⁵. These results agree with the fact that IκBα-KO mESCs cultured in Serum/LIF resemble 2i/LIF morphology, with homogeneous tight clusters of cells, whereas IκBα-WT mESCs retain the colony heterogeneity (Figure S2B). Interestingly, the higher levels of ground-state-related genes Zfp42 (Rex1), Klf2 and Tbx3 in IκBα-KO mESCs were partially obtained when IκBα-WT mESCs were cultured in 2i/LIF for two consecutive passages (Figure 3C).

One of the key features of the naïve pluripotency state is its global DNA hypomethylation pattern ^15,16. We assessed the DNA methylation status of IκBα-WT and IκBα-KO mESCs cultured in Serum/LIF conditions by analysis of 5-methyl-cytosine (5mC) mark and using DNA methylation arrays ³⁶. We found that IκBα-KO mESCs contained much lower levels of 5mC compared to IκBα-WT cells (Figure 3D). Moreover, DNA methylation arrays confirmed a global pattern of DNA hypomethylation in IκBα-KO mESCs, which was maintained in 96h EBs (Figure 3E). Interestingly, the DNA hypomethylation status occurred in a genome-wide fashion affecting all chromosomes (Figure S2C), in accordance with a higher pluripotent state of IκBα-KO mESCs.

Overall, these data demonstrate that IκBα is required for the exit of the naïve state of pluripotency and its deficiency results in impaired differentiation of mESCs in vivo and in vitro. Accordingly, IκBα-KO mESCs cultured in Serum/LIF showed genome-wide pattern of DNA hypomethylation.

Lack of IκBα causes an epigenetic rewiring in pluripotent stem cells to resemble 2i/LIF naïve pluripotency in Serum/LIF culture

ChIP Enrichment Analysis (ChEA) of genes identified in our RNA-seq data of IκBα-KO and WT mESCs revealed that differentially expressed genes were putative targets of chromatin regulators including the PRC2 subunit SUZ12 and MTF2 (Figure 4A). We previously described that IκBα is important for the proper deposition of the H3K27me3 mark in different types of tissue stem cells by associating with elements of the PRC2 complex ^27,28. Since 2i/LIF and Serum/LIF states are characterized by important differences in epigenetic marks, we carried out an epigenetic profiling of IκBα-WT and IκBα-KO mESCs cultured in Serum/LIF. We performed chromatin immunoprecipitation followed by sequencing (ChIP-seq) of histone modifications associated with gene activation (H3K4me3) and gene repression (H3K27me3) (Figures 4B and S3A-B). Overall, Serum/LIF IκBα-KO mESCs exhibited a general increased H3K4me3 status and redistribution of the PRC2-catalyzed H3K27me3 mark (Figure 4B). Notably, gain of the H3K4me3 mark was associated with a reduction of H3K27me3 levels in genes associated with naïve pluripotency function (Figures 4B-C and S3A-B). In particular, 41 genes with concomitant gain of H3K4me3 and loss of H3K27me3 were found to be enriched in mechanisms associated with pluripotency function (Figure 4C), including Tbx3 and Gbx2 (Figure 4D-4E). A second set of pluripotency-related genes showed an increase in H3K4me3 without changes in H3K27me3 (Zfp42 or Klf2). Both sets of pluripotency genes showed increased expression in the absence of IκBα (Figure 4H).

Figure 4.

We next aimed to study whether the lack of IκBα was also affecting distal regulatory regions by analyzing enhancer-associated histone marks in IκBα-WT and IκBα-KO mESCs. We quantified the total amount of H3K27Ac (active enhancer) and H3K4me1 (poised enhancer) by ChIP-seq experiments conducted in both IκBα-WT and IκBα-KO mESCs. Overall, IκBα-KO mESCs showed a reduction in poised enhancers at genes (or related genomic regions) associated with differentiation processes (Figures 4F and S3C) and an increase in active enhancers at genes associated with pluripotency (Figure 4F and S3D), such as Tbx3, Tfcp2l1 and Zfp42 (Rex1) (Figure 4G), which positively favors their higher expression in IκBα-KO mESCs (Figure 4H). These results further support that mESCs lacking IκBα are epigenetically remodeled to favor the ground state of naive pluripotency, which may negatively impact on their differentiation potential, similar to that described for other models with stabilized naïve pluripotent state^37,38.

The chromatin function of IκBα in pluripotency exit is independent of classical NF-κB activity

Although previous results from C. elegans ³⁹ and Drosophila ²⁷ support the functional relevance of chromatin-related IκBα function, the investigation of this alternative IκBα activity remains challenging due to the predominant role of this protein in NF-κB regulation. However, and similar to that found in other models, we did not detect major changes in canonical NF-κB target genes upon IκBα depletion neither in Serum/LIF mESCs, nor 48h and 96h EB-differentiating conditions (Figure 5A). Thus, either NF-κB regulation does not require IκBα at this stage or it might be compensated by the other IκBs (IκBβ, IκBε or p100), as previously demonstrated ^40,41. This observation suggests that either NF-κB signaling is not active in naïve pluripotency, which is supported by the low expression levels of NF-κB subunits (see Figures 1B-C and 1E-F), or IκBα function on NF-κB is compensated by the other IκBs (IκBβ, IκBε or p100), as previously demonstrated^40,41.

Figure 5.

To further investigate the relative impact of both IκBα functions in mESCs, we took advantage of recent results from our group that led to the identification of the specific IκBα protein residues that define NF-κB- or chromatin-binding through H2A/H4. Mutation of these residues allowed the generation of doxycycline-inducible (i) separation-of-function IκBα mutants that are specifically deficient in one or the other function, at least in the intestinal epithelial cells³². We called i-IκBα^ΔChromatin the i-SOF IκBα mutant that interacts with NF-κB but is deficient for H2A and H4 binding and i-IκBα^ΔNF-κB the one that interact with histones but is deficient for NF-κB subunits binding (Figure 5B and ³²). The three different IκBα forms (i-IκBα^WT, i-IκBα^ΔNF-κB and i-IκBα^ΔChromatin) were stably transfected in the IκBα KO mESCs and induced by 16h of doxycycline treatment (Figure S4A). We used these cells to investigate the biochemical properties and functional impact of SOF IκBα mutants in mESCs. By pull-down experiments, we found that i-IκBα^ΔNF-κB expressed in ESCs was able to interact with histone H2A but not with the NF-κB member p50. In contrast, i-IκBα^ΔChromatin interacts with p50 but not with H2A, while i-IκBα^WT preserves both binding capacities (Figure 5C). In concordance with these results, i-IκBα^WT and i-IκBα^ΔNF-κB were found in both the cytoplasmic and chromatin fractions of mESCs, whereas IκBα^ΔCromatin was retained in the cytoplasm (Figures 5D-E). In multiple experiments, we noticed a significant reduction in protein levels of IκBα^Δchromatin despite the mRNA expression levels of the different IκBα forms being comparable (Figures 5C-D and S4A).

Then, we studied whether i-IκBα^WT and/or the i-SOF IκBα mutants (IκBα^ΔNF-κB or IκBα^ΔChromatin) were able to reverse any of the phenotypes observed in the IκBα-KO mESCs (Figure S4B). Induction of i-IκBα^WT and i-IκBα^ΔNF-κB reduced the expression levels of the naïve pluripotent genes Zfp42, Klf2, Sox2 and Tbx3, which were increased by i-IκBα^ΔChromatin (Figure 5F). In addition, the ground state-associated DNA hypomethylation was specifically reverted in i-IκBα^WT and i-IκBα^ΔNF-κB mESCs but not in i-IκBα^ΔChromatin mESCs (Figure 5G). Finally, we addressed whether i-IκBα^WT, i-IκBα^ΔNF-κB and i-IκBα^ΔChromatin mESCs were capable to revert the differentiation blockage of IκBα-KO ESCs (Figure S4B). Induction of IκBα^WT and IκBα^ΔNF-κB was enough to restore the differentiation potential of IκBα-KO mESCs into EBs, as indicated the increase in the number of differentiated cells (based on percentage of SSEA-1^neg cells) (Figures 5H and S4C) and the reduced number of AP+ colonies (in the 216h IκBα-KO reconstituted EB assays), which was similar to the number of colonies obtained from IκBαWT EBs (Figures 5I). Lastly, the number of remaining pluripotent-like cells (defined by OCT3/4 NANOG staining) was also reduced in i-IκBα^WT and i-IκBα^ΔNF-κB 216h EBs, and only small and few regions of undifferentiated cells was observed. In contrast, induction of i-IκBα^ΔChromatin did not restore the differentiation potential of IκBα-KO mESCs (Figures 5J and S4D).

These results demonstrate that the chromatin-dependent IκBα function, but not its NF-κB-related activity, is necessary for the proper regulation of chromatin marks at specific genomic regions that are linked to transcriptional changes that are likely at the base of the capacity of ESCs to exit the ground state of naïve pluripotency and differentiate.

Discussion

Here we have identified a new and unexpected function for the IκBα protein in the regulation of the naïve pluripotency exit. Using mESCs to investigate the role of IκBα in pluripotency maintenance and differentiation, we discovered that the absence of IκBα severely impairs the exit of pluripotency. Furthermore, IκBα-KO mESCs stabilizes the ground state of naïve pluripotency under Serum/LIF conditions, shedding light on its function as a critical regulator of the transition from naïve to primed pluripotency. While IκBα is commonly known for its role as an inhibitor of the NF-κB signaling pathway ²⁶, our laboratory has previously furnished compelling evidence of its alternative role through its interaction with histones and other chromatin components ^27,29,31. We have demonstrated that chromatin-bound IκBα plays a crucial role in regulating skin and intestinal stem cells ^27,28, hematopoietic stem cell development ⁴², and it influences the regenerative capacity of these tissues, and in some cases, their susceptibility to neoplastic transformation. Disentangling the chromatin-related function of IκBα from its canonical role as an inhibitor of NF-κB has been challenging, making it difficult to determine its truly biological significance. To tackle this challenge, we engineered separation-of-function (SOF) mutants of IκBα, as detailed in ⁴³, by identifying specific residues that are essential for binding either NF-κB elements or histones. Interestingly, we have shown that only the histone-binding proficient IκBα mutant (IκBα^ΔNF-κB), which lacks the ability to bind to NF-κB, has the capability to reverse the stabilization of the ground state observed in IκBα-KO mESCs. This is sufficient to facilitate their exit from the state of naïve pluripotency, restoring their full differentiation potential (Figure 5). The involvement of NF-κB in pluripotency has been a subject of prior investigation. Studies have revealed that pluripotent cells display a dampened inflammatory signaling pathway, which is induced upon pluripotency dissolution and activation of differentiation ^24,44. However, the precise molecular mechanisms underlying the attenuated status of NF-κB during this stage remain incompletely understood. A recent work has shed light on a potential regulatory axis involving ATG5-β-TrCP1-NF-κB, proposing it as a mechanism that diminishes NF-κB activity by stabilizing the IκBα protein in mESCs ²³. In any case, since modification of canonical NF-κB activity directly impacts on the levels IκBα expression, the previous NF-κB-associated phenotypes should be reevaluated considering this chromatin role of IκBα. Furthermore, the expression of IκBα^ΔChromatin, which retains the inhibitory function of NF-κB, failed to mitigate the ground state stabilization observed in IκBα-KO mESCs (Figure 5). This finding suggests that the observed role of IκBα in regulating the exit from naïve pluripotency is not due to NF-κB inhibition. In fact, there is evidence that the chromatin function of IκBα is an ancestral function preceding its role as NF-κB inhibitor. In Caenorhabditis Elegans (C.elegans), which lacks NF-κB orthologs but possesses two homologs of IκBα, the absence of IκBα orthologs has severe multi-organ differentiation defects ³⁹. This observation further supports the notion that the chromatin-related function of IκBα plays a fundamental role in developmental processes, independent of its canonical role in NF-κB signaling.

We have observed that IκBα-KO mESCs exhibit epigenetic and transcriptomic profiles resembling the ground state of naïve pluripotency under Serum/LIF culture (Figure 3). In this regard, the fact that IκBα is preferentially located in the chromatin of mESCs cultured in Serum/LIF, when they exhibit a high degree of heterogeneity, suggests that IκBα might act as a regulator of this heterogeneity. One potential explanation might be that IκBα could be affecting the stability of fluctuating transcription factors, which is crucial to mediate the pluripotent-to-differentiation balance ³. Thus, the lack of IκBα might stabilize those transcription factors that would favor the ground-state observed in IκBα-KO mESCs cultured in Serum/LIF. Further experiments are required to demonstrate this hypothesis. Moreover, the differences in murine and human pluripotency ¹¹ point out the necessity to address the real impact of IκBα modulation in the stabilization of the ground state in human pluripotent stem cells.

In summary, our work establishes that IκBα mediates the exit from naïve pluripotency by modulating the activation of naïve pluripotency genes. Importantly, this newly identified role of IκBα operates independently of its canonical function in inhibiting the NF-κB pathway. Our findings underscore the intricate interplay between inflammation and pluripotency status, revealing a previously unrecognized complexity in their interaction.

Methods

Cell culture

mESC line ES-E14TG2a (ATCC; Cat #CRL-1821) was cultured on plastic dishes precoated with 0.1% (w/v) Gelatin (Sigma; Cat #G2500-100G) in Serum/LIF medium. Serum/LIF is composed by DMEM basal medium (Sigma-Aldrich; Cat #D5796) supplemented with 15% FBS (ESC-qualified; Gibco Cat #26140079), 1X Glutamax (Gibco; Cat #35050061), 1X NEAA (Gibco; Cat#11140050), 1mM Sodium Pyruvate (Gibco; Cat#11360070), 1000U/mL Leukemia Inhibitory Factor (LIF) (Millipore, Cat #ESG1107) and 0.125mM 2-mercaptoethanol (Gibco; Cat#31350010). Medium was changed every day, and cells were splitted every other day using TrypLE Express (Gibco; Cat #12605010) for harvesting. Cells were maintained in a 5% CO2 incubator at 37°C.

For inducing the ground-state of naïve pluripotency, mESCs were cultured in 2i/LIF medium for 2 consecutive passages. 2i/LIF medium is composed by NDiff 227 medium (TAKARA, Cat #Y40002) supplemented with 1000U/mL LIF, 0.4 µM PD032591 (Selleck Chemicals, Cat #S1036) and 3 µM CHIR99021 (Merck, Cat #SML1046).

Epiblast stem cell differentiation

Differentiation from mESCs towards Epiblast Stem Cells (EpiSCs) was performed as previously described ⁴⁵. 3x10⁴ mESCs were seeded in 6-well plastic dishes, and they were cultured in Serum/LIF medium for 24h. Medium was then switched to NDiff 227 medium (TAKARA, Cat #Y40002) supplemented with 20 ng/ml activin A (Cat. #120-14E; PeproTech) and 10 ng/ml Fgf basic (R&D Systems, Cat #233-FB-025/CF). Cells were submitted to daily media changes till day 5 (120h), when they were further analyzed.

Embryoid bodies differentiation from mESCs

Embryoid bodies differentiation was established as described in ⁴⁶. Briefly, mESCs were splitted twice in Serum/LIF medium before inducing differentiation. Once mESCs are 80% confluent, cells were collected using TrypLE Express. One well of 6-well plate was splitted in an entire 6-well plate in IMDM-ES medium for 48h. IMDM-ES medium is composed of Iscove’s Modified Dulbecco’s Medium (IMDM) (Cytiva, Cat #16SH30259.01) supplemented with 20% FBS (ESC-qualified; Gibco Cat #26140079), 1X Glutamax (Gibco; Cat #35050061), 1X NEAA (Gibco; Cat#11140050), 1mM Sodium Pyruvate (Gibco; Cat#11360070), 10ng/mL Leukemia Inhibitory Factor (LIF) (Millipore, Cat #ESG1107) and 0.125mM 2-mercaptoethanol (Gibco; Cat#31350010). For embryoid body induction, mESCs were harvested, and they were rinsed twice with DPBS (Gibco; Cat #14190144). Cells were very well disaggregated into single-cells, and 1.2 x 10⁴ cells/mL embryoid body differentiation (EB^diff) medium were resuspended to a total volume of 25mL of EB^diff medium. EB^diff medium is composed by IMDM supplemented with 15% FBS (ESC-qualified; Gibco Cat #26140079), 1X Glutamax (Gibco; Cat #35050061), 50 μg/mL ascorbic acid (Sigma; Cat #A-4544), 180 μg/mL Transferrin (Roche; Cat #10652202001) and 0.45 mM alpha-monothioglycerol (MTG) (Sigma; Cat #M6145). EBs were formed in suspension for 5 days (120h). At day 5, in order to elongate the differentiation up to 216h (day 9), EB^diff medium was refreshed by harvesting EBs with 10mL serological pipette, centrifuge at 200 g for 3 min and EBs were resuspended in fresh EB^diff medium for 4 more days.

CRISPR/Cas9 gene editing and cell lines generation in mESCs

The two guide RNAs (gRNAs) targeting the Nfkbia locus were designed using the CRISPR design tool from MIT (http://crispr.mit.edu). The best 2 gRNAs (based on on-target and off-target scores) that were targeting the exon 1 of Nfkbia (IκBα gene) were selected. After annealing, one of the gRNAs was cloned into SpCas9(BB)-2A-GFP (px458) plasmid (Addgene; Cat #48138), and the other gRNA was cloned into the px330-mCherry plasmid (modified from Addgene; Cat #98750 to incorporate mCherry reporter). mESCs were co-transfected with the two plasmids (px458-gRNA1 and px330-mCherry-gRNA2). 3x10⁵ cells were seeded a day before per well of 6-well plate. At day of transfection, cells were washed once with 1X DPBS (Gibco; Cat #14190144), and 2mL Opti-MEM medium (Gibco; Cat #31985070) was added into cells. 1.25 μg of each plasmid (2.5 μg of total DNA) was incubated with 10 μL of Lipofectamine 2000 (Invitrogen; Cat #11668019) in 240 μL of Opti-MEM for 20 min. OptiMEM::Lipofectamine::DNA mixture was added into mESCs, and cells were incubated for 5 hours. Cells were then washed with DPBS, and medium was replaced by Serum/LIF. 48h after transfection, GFP⁺ mCherry⁺ single cells were sorted by FACS using BD FACSAria II Cell Sorter (BD Bioscience). Mutant clones were screened through PCR and western blotting to identify single clones with no IκBα protein expression. Non-targeting scrambled gRNAs were cloned into px458 and px330-mCherry plasmids as wildtype clones. Three independent clones from each genotype (IκBα-WT and IκBα-KO) were selected for further experiments.

Doxycycline-inducible IκBα (IκBα^WT, IκBα^ΔNF-κB and i-IκBα^ΔH2A/H4) mESCs were generated by cloning the three different IκBα versions into a PiggyBac transposon system ⁴⁷. IκBα protein-coding cDNAs were PCR-amplified, and NheI and SalI restriction sites were placed at the 5’ and 3’ ends, respectively. T2A-EGFP fragment was amplified from PX458 plasmid, and SalI and AgeI restriction sites were introduced at its 5’ and 3’ ends. PB-TRE backbone (Addgene, Cat# 63800) was digested with NheI and AgeI enzymes, and mNfkbia and T2A-EGFP fragments were ligated into the PB-TRE digested vector. To generate mESCs containing the inducible vector, IκBα-KO mESCs were co-transfected with PB-TRE-mNfkbia-T2A-EGFP and PiggyBac transposase plasmids, and pool of transfected cells was selected by Hygromycin (100µg/mL) for 7 days. Cells were doubly screened by FACS sorting green fluorescent protein after 24h of doxycycline (1µg/mL) treatment. Pool of cells were used for further experiments. Oligos sequences are included in supplementary table 1.

Teratoma formation assay

IκBα-WT and IκBα-KO mESCs grown in Serum/LIF (5 × 10⁵) were injected intramuscularly into severe combined immunodeficient mice (NSG). In order to favor 3D aggregation of cells and teratoma formation, cells were resuspended in Matrigel Matrix (BD; Cat #356234) prior to injection. Six weeks later, mice with tumors were euthanized, and tumors were fixed in formaldehyde, embedded in paraffin, sectioned and stained with hematoxylin and eosin for histological analysis.

Immunofluorescence staining

For immunostaining of mESCs, cells were seeded on 0.1% gelatin-coated coverslips on 6-well plates; for EB immunofluorescence, they were collected directly from the plate. Samples were washed twice with 1X PBS, and they were fixed at 4°C for 30 minutes with 4% PFA (Electron Microscopy Sciences; Cat #15713S). For immunofluorescence of 5-Methylcytosine (5mC), fixed cells were incubated with 2M Hydrochloric Acid (Sigma-Aldrich, Cat #H1758) for 20 minutes at room temperature. Samples were then washed twice with 1X Tris-buffered saline (TBS; 50 mM Tris-Cl, pH 8) and permeabilized and blocked with 1X TBS supplemented with 1% Triton-X100 (MERCK; Cat #9036-19-5) and 6% FBS (Biological Industries; Cat #04-001-1A) for 2h at 4°C. Samples were washed twice with 1X TBS supplemented with 6% FBS. Primary antibody incubation was performed in 1X TBS plus 6% FBS and 0.3% or 0.5% Triton-X100 (0.3% for 2D culture and 0.5% for EBs) overnight at 4°C. The following primary antibodies were used: OCT3/4 (1:250; Santa Cruz; Cat #sc-5279), NANOG (1:250; Novus Biologicals; Cat #NB100-588), 5-Methylcytosine (1:500, Invitrogen, Cat #MA5-24694), H3K27Ac (1:2000, Abcam, Cat #AB4729). Samples were then washed 4 times with 1X TBS, 5 min each washing. Secondary antibody incubation was performed in TBS 1X plus 1% BSA (Sigma-Aldrich; Cat #9048-46-8) for 2h at room temperature. The following antibodies were used: Alexa Fluor 488 donkey anti-mouse antibody (1:1000; Invitrogen; Cat #A-21202), Alexa Fluor 647 donkey anti-mouse antibody (1:1000; Invitrogen; Cat #A-31571), Alexa Fluor 594 donkey anti-rabbit antibody (1:1000; Life Technologies; Cat #A-21207), Alexa Fluor 647 donkey anti-rabbit antibody(1:1000, Invitrogen; Cat #A-31573). Samples were washed three times with 1X TBS for 5 min at room temperature each rinse. Samples were mounted using DAPI Fluoromount-G (Southern Biotech; Cat #0100-20).

Microscopy and image acquisition

Fluorescence images were acquired using Confocal Leica TCS SP5 (Leica Microsystems), and Leica application software LAS AF (Leica Microsystems) was used to visualize the images. Images of teratoma haematoxylin and eosin and immunohistochemistry stainings were acquired using BX61 Olympus Microscope (Olympus), and PRECiV™ 2D Image and Measurement Software (Olympus) was used to visualize the images. The Image J (version 1.15) was used for further analysis ⁴⁸.

Flow cytometry sample preparation and analysis

Cells were collected and they were purified based on FACS using BD FACSAria II Cell Sorter (BD Bioscience). Cells were sorted at 3,500 events/second, and at maximum flow rate of 4 and 85 μm nozzle. 5 μg/ml of DAPI (Biotium; Cat #BT-40043) was used as a viability dye. For flow cytometry analysis of pluripotency exit, cells were disaggregated using TrypLE Express, and they were incubated at 4°C for 20 minutes with SSEA1-eFluor 660 Monoclonal Antibody (1:200, Invitrogene, Cat #50-8813-42). BD LSRFortessa Cell Analyzer or BD LSR II Flow Cytometer (BD Bioscience) were used for Flow Cytometry Analysis experiments. Flow cytometry data were analyzed using FlowJo X v10.0.7 (BD Biosciences).

RNA isolation, cDNA synthesis and quantitative RT-PCR

Total RNA isolation from cells was performed using the RNeasy Plus Mini Kit (Qiagen; Cat #74136) or RNeasy Micro Kit (Qiagen; Cat #74004) following manufacturer’s instructions. Amount of RNA was quantified with Nanodrop (Thermo Fisher; Cat #ND2000CLAPTOP), and 2 μg of total RNA was retro-transcribed using Transcriptor First Strand cDNA Synthesis Kit (Roche; Cat #04897030001) following the manufacturer’s instructions.

Quantitative RT-PCR was performed in triplicates for each sample, and SYBR Green I Master Kit (Roche; Cat #04887352001) was used to carry out the reaction. qRT-PCR was performed using the LightCycler 480 system (Roche). Relative expression levels were calculated as 2^-ΔCT normalized with the average CT of the housekeeping gene Tbp or Gapdh. Oligos sequences are found in supplementary table 1.

Chromatin Immunoprecipitation (ChIP)

4-6x10⁷ mESCs were cross-linked by incubating them in DPBS (Gibco, Cat #14190094) supplemented with 1% formaldehyde (Sigma-Aldrich; Cat #252549) for 10 min rocking at room temperature. Cross-linking was stopped using 125 mM Glycine (Sigma-Aldrich, Cat #G8790) rocking for 5 minutes at room temperature. Fixed cells were then washed twice with ice-cold DPBS, and scrapped and collected using DPBS supplemented with protease/phosphatase inhibitor cocktail composed by 1X cOmplete™, EDTA-free Protease Inhibitor Cocktail (Roche, Cat #11873580001), 1 mM PMSF, 1mM Sodium Ortovanadate and 20 mM β-Glycerol phosphate. Cells were centrifuged at 3200g for 5 minutes at 4°C. Cells were lysed by resuspending cell pellet in ice-cold ChIP buffer (1 volume of SDS buffer [100mM NaCl, 50mM Tris-HCl pH 8.1, 5mM EDTA pH 8, 0.5% SDS] and 0.5 volume of triton dilution buffer [100mM Tris-HCl pH 8.6, 100mM NaCl, 5mM EDTA pH 8, 5% Triton X-100]) supplemented with protease/phosphatase inhibitor cocktail, and samples were sonicated using sonication beads (Diagenode; Cat #C01020031) and the Bioruptor Pico Sonicator (Diagenode; Cat #B01060010) for 20 cycles (each cycle 30 seconds on/30 seconds off) or till DNA fragments have 100-300bp size. Samples were centrifuged at 16000g for 20 minutes at 4°C. Chromatin was quantified, and 30 μg of chromatin was incubated with every 5μg of antibody rotating for 16 hours (or overnight) at 4°C. The following antibodies were used: H3K4me3 (Abcam, Cat #ab8580), H3K27me3 (Millipore, Cat ##07-449), H3K27Ac (Abcam, Cat #ab4729), H3K4me1 (Abcam, Cat #ab8895). Samples were pulled down by incubation with Protein A-Sepharose™ CL-4B (previously hydrated and blocked with 0.05% BSA) for 3h at 4°C in rotation. Sample was washed with a low salt washing buffer (50mM HEPES pH 7.5, 140mM NaCl and 1% Triton) three times, one wash with a high salt wash buffer (50mM HEPES pH 7.5, 500mM NaCl and 1% Triton) and one wash with TE pH 8 (10mM Tris-HCl pH8, 1mM EDTA). Samples were then eluted in 1% SDS and 100mM NaHCO3. Samples were descrosslinked by incubating for 16 hours (or overnight) at 65°C shaking at 450 rpm, and were treated with Proteinase K for 1h at 45°C. DNA was purified using QIAquick PCR Purification Kit (Qiagen, Cat #28106), following the manufacturer’s instructions.

For ChIP followed by sequencing (ChIP-seq) Purified DNA concentration and integrity were determined using Agilent Bioanalyzer (Agilent Technologies; Cat #G2939BA). Libraries were prepared using standard protocols. Chromatin was sequenced using Illumina HiSeq platform (Illumina, Inc.) (50bp single-end reads). Samples sequencing depth range was the following: (i) H3K4me3: 2.56-2.74x10⁷ reads, (ii) H3K27me3: 6.19-7.2x10⁷ reads, (iii) H3K27Ac: 9.11-10.65x10⁷ reads and (iv) H3K4me1: 9.76-10.58x10⁷ reads.

ChIP-seq data analysis

Quality control was performed on raw data with FASTQC tool. Raw reads were trimmed to remove adapters presence with Trimgalore (v0.6.6) ⁴⁹. Default parameters were used except for a minimum quality of 15 (Phred score) and an adapter removal stringency of 3bp overlap. For H3K4me3, H3K27me3 and H3K4me1 data, trimmed reads were aligned to the reference genome with Bowtie2 (v2.4.4) which was executed with default parameters ⁵⁰. Required genome index was built with corresponding GRCm38 fasta file retrieved from Ensembl (http://ftp.ensembl.org/pub/release-102/). Multimapped reads and those exhibiting MAPQ < 20 were removed. Randomly placed multi mappers were removed from the mapped reads. For all cases, duplicated reads were marked with SAMtools (v1.15) ⁵¹. Drosophila Melanogaster spike-ins, present in H3K27me3, were discarded. NSC and RSC quality metrics were computed with PhantomPeakQualTools (v1.2) ⁵². ENCODE blacklisted regions (mm10 v2) were removed prior to peak calling. BigWig files were individually generated using deepTools (v3.5.1) bamCoverage with -ignoreDuplicates -binSize 10 -smoothLength 30 - effectiveGenomeSize 2308125349 -normalizeUsing RPGC and -extendReads Fragment_Length options ⁵³. The effective genome size was particularized to a read length of 50bp and directly retrieved from deepTools web site (https://deeptools.readthedocs.io/en/develop/content/feature/effectiveGenomeSize.html). Fragment_Length was retrieved from PhantomPeakQualTools results. For all histone marks except for H3K4me3, peak calling was conducted by means of epic2 (v0.0.52) with -effective-genome-fraction 0.8452 -fragment-size Fragment_Length options and chromosome sizes only referring to canonical chromosomes ⁵⁴. For H3K4me3, MACS2 (v2.2.7.1) was used to identify peaks with -nomodel -extsize Fragment_Length -g 2308125349 options ⁵⁵. The corresponding input sample was used in all peak calling computations. For the histone marks, peaks were called with adjusted p-values (FDR) < 0.05.

Peak annotation was performed with ChIPseeker R (v4.2.1) package (v1.34.1) considering a TSS region range between 5000 bp upstream and 100 bp downstream (Yu et al., 2015). Required TxDb object was generated from the Ensembl GRCm38 gtf file retrieved (http://ftp.ensembl.org/pub/release-102/). A consensus peakset was obtained from the three biological replicates per histone mark and condition (IκBα-WT or IκBα-KO) by identifying overlapping peaks in at least two out of the three replicates.

Differential binding analysis (DBA) was conducted with DiffBind (v3.8.4) per histone mark ⁵⁶. Default parameters were used except for the summits parameter which was set to 500bp for all histone marks except for H3K4me3 (150bp) in order to consider, for testing, intervals of 1000bp or 300bp respectively. Summit values were selected to have interval widths between the minimum and first quartile peak width values for each histone mark. EdgeR was the statistical method used for all the three analyses. Differentially bound regions (DBRs) between IκBα-KO and IκBα-WT samples were called with adjusted p-values (FDR) < 0.05. Identified DBRs with any annotated gene were plotted with EnrichedHeatmap R package (v1.18.1) ⁵⁷. To summarize replicates per condition, normalizeToMatrix function was used based on corresponding BigWig files in w0 mean mode in 50bp windows. Tracks visualization were obtained by means of Integrative Genomics Browser (IGV) tool ⁵⁸.

Enhancers activity identification

Consensus peaksets derived from the three H3K27ac and H3K4me1 IκBα-WT mESCs replicates were used to identify poised and active enhancers in an IκBα-WT scenario. Putative poised enhancers were defined as those regions with H3K4me1 but no H3K27ac peaks. Putative active enhancers were defined as those regions with H3K27ac and presence/absence H3K4me1 peaks. To assess any differential enhancers activity in IκBα-KO vs IκBα-WT mESCs cultured in Serum/LIF, differentially bound regions and consensus peaksets (IκBα-KO) identified in H3K4me1 and H3K27ac were required. Putative gained or lost poised/active enhancers in IκBα ^-/- were obtained with following criteria: (i) for gained/lost poised enhancers in IκBα ^-/-: a differential H3K4me1 increase/decrease and absence of H3K27ac peaks in IκBα ^-/- consensus peakset, (ii) for gained/lost active enhancers (without H3K4me1) in IκBα-KO: a differential H3K27ac increase/decrease and absence of H3K4me1 in IκBα-KO consensus peakset and (iii) for gained/lost active enhancers (with H3K4me1) in IκBα-KO: a differential H3K27ac increase/decrease and presence of H3K4me1 in IκBα ^-/- consensus peakset.

A maximum gap of 1000bp was allowed for checking the overlap between two different histone marks. For this purpose, the ‘subsetByOverlaps’ function from the IRanges R package was used (v.2.34.1).

RNA-seq experiments

Total RNA from three independent clones from mESCs, 48h EBs and 96h EBs was isolated using the RNeasy Plus Mini Kit (Qiagen; Cat #74136) following manufacturer’s instructions. Amount of RNA was quantified with Nanodrop (Thermo Fisher; Cat #ND2000CLAPTOP), and RNA integrity was addressed by agarose gel and Agilent Bioanalyzer (Agilent Technologies; Cat #G2939BA). Libraries sequenced using Illumina HiSeq 2500 (Illumina, Inc.) (125bp paired-end reads). Samples sequencing depth ranged between 35M and 52M reads (average 41M reads) per sample.

RNA-seq data analysis

Quality control was performed on raw data with the FASTQC tool (v0.11.9). Raw reads were trimmed to remove adapter presence with Trimgalore (v0.6.6) ⁴⁹. Default parameters were used except for a minimum quality of 15 (Phred score) and an adapter removal stringency of 3bp overlap. Trimmed reads were aligned to reference the genome with STAR aligner tool (v2.7.8). STAR was executed with default parameters except for the number of allowed mismatches which was set to 1. Required genome index was built with corresponding GRCm38 gtf and fasta files retrieved from Ensembl (http://ftp.ensembl.org/pub/release-102/). Obtained BAM files with uniquely mapped reads were considered for further analysis. Raw gene expression was quantified using featureCounts tool from subRead software (v2.0.1) with exon as feature ⁵⁹. The raw counts matrix was imported into the R Statistical Software environment (v4.2.1) for downstream analysis. Raw expression matrix included 55,487 genes per 18 samples in total. Experimental design considered three timepoints: mouse ESCs, EBs at 48h and EBs at 96h. Each time point included 6 samples distributed in two conditions: 3 IκBα-WT and 3 IκBα-KO. Prior to statistical analysis, those genes with less than 10 raw counts across the 6 samples under test were removed. After pre-filtering, 21,843 genes (mESCs), 21,424 genes (EBs 48h) or 21,664 genes (EBs 96h) were available for testing. For visualization purposes, counts were normalized by the variance-stabilizing transformation method as implemented in DESeq2 R package ⁶⁰ (v1.38.3). Differential expression analysis (DEA) was conducted with DESeq2. Each time point was independently analyzed. Fitted statistical model included sample conditions as covariable with IκBα-WT as the reference. Obtained log2 fold change values were shrunken with apeglm shrinkage estimator R package (v1.20.0) ⁶¹. Raw p-values were adjusted for multiple testing using the Benjamini-Hochberg False Discovery Rate (FDR) ⁶². Differentially Expressed Genes (DEGs) between IκBα-KO and IκBα-WT samples were called with adjusted p-values (FDR) < 0.05 and absolute shrunken log2 Fold change > 1. Data visualization was performed with the ggplot2 (v3.4.1).

Genome-wide DNA methylation samples preparation

DNA from frozen mESCs pellets was extracted using DNeasy Blood and Tissue Kit (Qiagen GmbH, Hilden, Germany). Purified genomic DNA was quantified with Qubit (Invitrogen, Carlsbad, CA, USA) according to manufacturer’s instructions. Infinium Mouse Methylation BeadChip (Illumina, Inc., San Diego, CA, USA) arrays were used to profile DNA methylation genome wide. This platform allows over 285,000 methylation sites per sample to be interrogated at single-nucleotide resolution. The samples were bisulfite converted using EZ DNA Methylation-Gold™ Kit (Zymo Research, CA, USA) and were hybridized in the array following the manufacturer’s instructions.

DNA methylation data analysis

The DNA methylation profile of the studied samples was assessed using the Infinium Mouse Methylation BeadChip Array (∼285,000 methylation sites) as previously described ⁶³. Briefly, raw signal intensities were obtained with GenomeStudio Software 2011.1 (Illumina) and DNA methylation beta values were computed from raw IDAT files using GenomeStudio default normalization with control probes and background subtraction. Quality control steps to remove erratic probe signals were performed within the R statistical environment (v4.0.3). We removed probes with detection p value > 0.01, genotyping probes and manufacturing flagged (MFG) probes described in the Illumina manifest file (https://support.illumina.com/downloads/infinium-mouse-methylation-manifest-file.html). The differentially methylated probes (DMPs) between IκBα-WT and IκBα-KO samples were computed separately for each time point (mESCs and EBs at 96h) by deriving a linear model with the limma R package (v3.46.0). Each condition included 3 samples per time point. DMPs with adjusted p-value (FDR) <0.05 and absolute mean methylation beta value difference between conditions >0.3 were considered significant. DNA methylation analysis was performed using the mm10 mouse genome reference build and the complete annotation was downloaded from the annotated manifest file (http://zwdzwd.github.io/InfiniumAnnotation#mouse) ³⁶.

Functional analysis

Overrepresentation analysis was applied over lists of selected genes derived from RNA-seq data (DEGs) or from ChIP-seq data (differentially bound regions). The Gene Ontology (Biological Process ontology, GO BP terms), KEGG PATHWAY and WikiPathways databases for Mus Musculus ^64–66 were interrogated by means of clusterProfiler R package (v4.6.2) (Wu et al., 2021). Corresponding Entrez identifiers were used. Benjamini-Hochberg procedure was used to obtain adjusted p-values. Obtained GO BP terms were simplified using the simplify function from clusterProfiler with default parameters. Overrepresented terms or processes were called with adjusted p-values (FDR) < 0.05.

Gene Set Enrichment Analysis (GSEA) was performed for mESCs against the three cell states signatures defined by other authors ⁶⁷. Briefly, each signature was composed of 200 genes uniquely expressed in each state obtained from bulk RNA-seq performed per state. For this purpose, the complete list of genes from mESCs samples (21,843 genes) was ranked based on the shrunken log2 Fold Change obtained from DEA. GSEA was conducted through the fgseaMultilevel function from fgsea R package (v1.24.0) ⁶⁸ with default parameters. Enrichment plots were generated with the same package.

Additionally, the testEnrichment function from SeSAMe R package (v1.14.2) ⁶⁹ was used for conducting functional analysis with default parameters over the list of DMPs. Probe design, transcription factor binding site and histone modifications consensus database sets, included in the same package, were interrogated.

Gene Set Variation Analysis (GSVA) R package (v1.46.0) with default parameters were used to obtain Z-score values for genes annotated to endoderm (GO:0001706), mesoderm (GO:0001707) and ectoderm (GO:0001705) GO BP terms ⁷⁰.

Statistical analysis

Statistical analysis was performed with GraphPad Prism v.8.0.1. (GraphPad Software, Inc.). Unless specified, the comparison between two groups was performed with unpaired two-sided t-test. A p-value < 0.05 was considered significant.

Data availability

All sequencing data is deposited at GEO under a SuperSeries with the accession number GSE239565)[https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE239565]. Individual SubSeries can be found at: GSE239563 (RNA-seq data), GSE239564 (ChIP-seq data) and GSE239562 (Methylation array data).

Acknowledgements

We would like to acknowledge all members of the Bigas/Espinosa labs for helpful discussions. We are grateful for technical support to CRG/UPF Flow Cytometry Unit, CRG Genomics and Advanced Light Microscopy Units. This work has been supported by Spanish Ministry of Science and Innovation (PID2019-104695RB-I00, PLEC2021-007518, PDC2021-120817-I00), Generalitat de Catalunya (2021SGR 39) and Departament de Salut (SLT002/16/00299) to AB; the Spanish Ministry of Science and Innovation (PID2019-108322GB-100 and PID2022-142679NB-I00) to LDC; the Spanish Ministry of Science and Innovation (PID2021-123383NB-I00.) and the Agencia de Gestió d’Ajuts Universitaris i de Recerca (2021 SGR 01222) to BP. L.G.P has been a recipient of FI AGAUR fellowship (2019 FI-B 00151/2020 FI_B1 00130) from Generalitat de Catalunya. D.A-V was funded by the FIS fellowship (FI20/00130) from Instituto Carlos III. M.B. received funding from the Ramón Areces Foundation. M.M. is a recipient of a grant from the Instituto Carlos III, grant number CA22/00011 (co-funded by the European Social Fund Plus, ESF+ and by the European Union).

Additional information

Code availability

Scripts used to process the generated bulk RNA-seq and ChIP-seq data are available in Github repository: https://github.com/BigaSpinosaLab/PAPER_IkBa_Naive_Pluripotency

Author contribution

L.G.P, L.E and A.B designed the study. L.G.P, D.A-V, M.B, A.I, J.G, J.B, D.A.E, C.A.G.P, C.B, V.R.C and A.V performed experiments. L.G.P, M.M, M.B and C.A.G.P performed bioinformatics analysis. L.G and A.B wrote the manuscript with input from all authors. C.B, A.V, A.V, P.M, L.D.C, G.S, B.P, M.E and L.E provided input on experimental design. L.E and A.B provided resources and supervised the research.