Introduction

Early mammalian development is a dynamic process that involves large-scale chromatin reorganization (1). Blastomeres acquire a defined cell identity through the activation of a subset of genes and specific epigenetic modifications. Among the latter, there is a tailored control over histone H3 lysine 9 trimethylation (H3K9me3), a hallmark of the transcriptionally repressed constitutive heterochromatin (2, 3). During the first cleavage stages, constitutive heterochromatin reorganizes in the nucleus to form highly compacted chromocenters (4, 5). Systematic identification of the underlying factors involved in de novo heterochromatin establishment and maintenance - thus chromocenter compaction-is still lacking, mostly because of the minuscule amount of material available during embryogenesis.

Embryonic stem cells (ESCs) can fluctuate back to a 2-cell embryo-like (2C-like) state under defined culture conditions (6). Although ESCs can spontaneously revert their fate to resemble early embryogenesis, this process happens at a very low frequency (6). Recently, early mouse embryo development has been modelled with high efficiency after downregulation of chromatin assembly factors (7) or modulation of key developmentally regulated genes in ESCs (8-11). ESCs can efficiently be converted into 2C-like cells by the overexpression of a single murine transcription factor, Dux (11). Interestingly, decondensation of HP1α foci and of chromocenters in 2C⁺ were previously reported, suggesting that 2C⁺ cells can be used to study the remodeling of heterochromatin foci (7, 12). Here, using the Dux-dependent reprogramming system, we show that 2C-like cells can be used as a model system to investigate de novo chromocenter formation and dynamics. Using chromatin proteomics, we profiled the dynamic changes occurring in the chromatin-bound proteome (chromatome) during 2C-like cell reprogramming and identified factors potentially involved in chromocenter reorganization. H3K9me3-marked heterochromatin foci in 2C-like cells generated via Dux overexpression became larger and decreased in number during the reprogramming of ESCs to 2C-like cells. The chromocenters re-formed upon transition of 2C-like cells into ESC-like cells. We identified the DNA TOPoisomerase II Binding Protein 1 (TOPBP1) and the chromatin remodeler SWI/SNF-Related, Matrix-Associated Actin-Dependent Regulator Of Chromatin, Subfamily A, Containing DEAD/H Box 1 (SMARCAD1) to be associated with H3K9me3 in heterochromatin foci of ESCs. The association of SMARCAD1 was reduced upon entry of ESCs in the 2C-like state, although SMARCAD1 nuclear localization was recovered after 2C-like state exit. Depletion of SMARCAD1 and of TOPBP1 induced mouse embryo developmental arrest, which was accompanied by a remodeling of the heterochromatin foci. Our results suggest a contributing role of SMARCAD1 and of TOPBP1 activity in the maintenance of heterochromatin formation during early development.

Results

Chromatin-bound proteome profiling allows the identification of dynamic chromatome changes during 2C-like cell reprogramming

To explore the molecular driving events for the establishment of constitutive heterochromatin during embryo development, we generated stable ESC lines carrying doxycycline-inducible cassettes that drive expression of either Dux (Dux-codon altered, CA) or luciferase (control) (Fig. 1A). These ESC lines also carry an EGFP reporter under the control of the endogenous retroviral element MERVL long terminal repeat (2C::EGFP) (7). The EGFP reporter allows the purification of 2C-like cells (hereinafter named 2C⁺) and low Dux expressing cells, which are negative for MERVL reporter expression (2C⁻) (Fig. 1A). 2C⁻ cells were reported to be an intermediate population generated during 2C-like reprogramming (13, 14). During the reprogramming process, in contrast to 2C⁻ cells, 2C⁺ cells do not show DAPI-dense chromocenters (7, 11). Therefore, we can study de novo chromocenter formation by following the transition of 2C⁺ cells toward an ESC-like state, thus modelling in culture the epigenetic reprogramming that occurs during mouse early development.

Fig. 1.

After culturing the Dux-CA line with doxycycline (Dox), the number of 2C-like cells increased to > 60 % as compared to luciferase control cells (fig. S1A-D). Dux overexpression resulted in the loss of DAPI-dense chromocenters and loss of the pluripotency transcription factor OCT4 (fig. S1E), in accordance with previous reports (6, 7). These changes were accompanied by an upregulation of specific genes of the 2-cell transcriptional program such as endogenous Dux, MERVL and major satellites (MajSat) (fig. S1F). Additionally, we looked at cell cycle progression in the heterogeneous population of cells generated after Dux overexpression since it has been previously shown that spontaneous 2C-like cells have an altered cell cycle (15). 2C⁻ cells displayed a cell cycle profile comparable to that of luciferase cells, whereas 2C⁺ cells accumulated in the G2/M cell cycle phase (fig. S1G) with a much-reduced S phase consistent in several clonal lines (fig. S1H). Overall, these data indicate that the 2C⁺ line we generated recapitulates known features of 2C-like cells.

Having characterized the Dux-CA line, we aimed to identify potential chromatin-associated factors involved in the de novo establishment of heterochromatin. For that, we performed DNA-mediated chromatin purification coupled to tandem mass spectrometry for the identification of proteins on total DNA (iPOTD) (16, 17). We captured the whole genome labelled with 5-ethynyl-2’-deoxyuridine (EdU) and identified candidate proteins differentially enriched in the 2C-like chromatin-bound (chromatome) fraction (Fig. 1A). We analyzed the chromatome of 2C⁺, 2C⁻ and luciferase (Luc) populations to characterize the chromatin-bound proteome profile of these distinct states. We first confirmed that we could enrich the iPOTD preparations for chromatin proteins, such as histone H3, and devoid them of cytoplasmic ones, such as vinculin (fig. S2A-C). We identified a total of 2396 proteins, suggesting an efficient pull-down of chromatin-associated factors (fig. S2D and Table S1). Chromatin-resident proteins, such as core histones and histone variants, were comparably enriched in all +EdU replicates (fig. S2E and Table S1). Pearson’s correlation coefficients (PCC) and principal component analysis (PCA) of independent replicates of 2C⁺, 2C⁻ and Luc samples showed consistent results regarding the abundance of the proteins detected (Fig. 1B-D and Table S1). Interestingly, Luc replicates clustered separately from 2C⁺ and 2C⁻ conditions, indicating significant changes in the chromatomes of these fractions (Fig. 1C, 1D).

Table 1

We then ranked the identified chromatin-associated factors according to their fold change to interrogate the differences in protein-chromatin interactions in the 2C⁺, 2C⁻ and Luc chromatomes (Fig. 1E-G). Members of the ZSCAN4 (Zinc finger and SCAN domain containing 4) family of proteins, that are well-characterized markers of the 2C stage (7, 11, 18, 19), were identified among the top enriched factors in the 2C⁺ chromatome (Fig. 1E, 1F, 1H). ZSCAN4 family members, such as ZSCAN4F and ZSCAN4C, were found associated with chromatin already in the 2C⁻ chromatome (Fig. 1F), supporting previous findings (14). However, we identified regulators of 2C-like cells such as TET1, the non-canonical Polycomb (PcG) Repressor Complex 1 (PRC1) member PCGF6, the TGF-β regulator SMAD7 and the heterochromatic H4K20me3 methyltransferase SUV420H2 depleted from the 2C⁺ chromatome when compared to the 2C⁻ (13, 20, 21) (Fig. 1E, 1G, 1H). The pluripotency transcription factors NANOG, OCT4, STAT3, SOX2 and ESRRB were, as expected, exclusively enriched in the 2C⁻ and Luc chromatomes (Fig. 1H). We also identified several transcriptional regulators and epigenetic enzymes differentially enriched in the 2C⁺, 2C⁻ and Luc chromatomes (Fig. 1H). Of note, we identified marked differences in the enrichment for H3K9 histone methyltransferases SUV39H1 and SUV39H2 (Fig. 1H). SUV39H2 gradually increased its abundance on chromatin as Luc cells converted to 2C⁻ and, ultimately, to 2C⁺ (Fig. 1H). Contrarily, SUV39H1 was overall less abundantly chromatin-bound, although with a slight enrichment in the 2C⁻ chromatome (Fig. 1H). Altogether, these data indicate that ESC reprogramming toward 2C-like state correlates with a major reorganization of the chromatin-bound proteome.

We used the Significance Analysis of INTeractome (SAINT) algorithm (22) to further interrogate protein-chromatin interactions in the iPOTD datasets. To identify molecular drivers of chromocenter reorganization, we compared the enriched proteins in the 2C⁺, 2C⁻ and Luc chromatomes (Fig. 1I and Table S1). We identified a total of 397 proteins shared by the 2C⁻ and Luc chromatomes that were not enriched in the 2C⁺ chromatome (Fig. 1I and Table S1). We focused on analyzing this cluster since chromocenters are present in 2C⁻ and Luc cells. This protein cluster included gene ontology (GO) terms associated with RNA and chromatin binding, active remodeling activity (e.g. ATPase activity), repressive chromatin (e.g. heterochromatin condensed chromosome, negative regulation of gene expression), and pluripotent stem cell identity (e.g. response to LIF, stem cell maintenance, blastocyst growth) (Fig. 1J). To identify putative factors responsible for chromocenter reorganization, we ranked the commonly identified proteins included in 2C⁻ and Luc chromatomes according to their fold change (Fig. 1K). Notably, this protein cluster included known transcriptional regulators such as the DNA methyltransferase DNMT3L, the bromodomain-containing protein BRD2, the core pluripotency factor OCT4 and the DNA topoisomerase 2-binding protein 1, TOPBP1 (Fig. 1K). We focused our attention on TOPBP1, since topoisomerases control genome structure and folding (23). We asked if the lack of topoisomerase activity could promote 2C⁺ cell induction. Thus, we treated ESCs with camptothecin (CPT) and ICRF-193, which are inhibitors of DNA topoisomerases I and II, respectively (24, 25). Inhibition of topoisomerase II alone increased the number of 2C⁺ cells 1.5-fold (fig. S3A) and triggered a prominent cell cycle arrest in the G2/M phase (25, 26) (fig. S3B). Simultaneous inhibition of topoisomerases I and II resulted in an enhanced effect, leading to a 2.4-fold increase in the fraction of 2C⁺ cells (fig. S3A, S3C). These results motivated us to further investigate TOPBP1 network. TOPBP1 has been shown to interact with chromatin remodelers such as the SWI/SNF-like remodeler SMARCAD1 in yeast and human cells (27, 28) (Fig. 1L). We then investigated TOPBP1 and SMARCAD1 as potential candidate factors controlling the remodeling of chromocenters.

Entry in the 2C-like state is characterized by the remodeling of H3K9me3 heterochromatic regions

To study remodeling of the chromocenters, we asked about the reorganization of heterochromatic regions upon reprogramming of ESCs into 2C⁺ cells. H3K9me3 is a well-known pericentric heterochromatin histone modification that prominently associates with constitutive heterochromatin (2, 3, 29, 30). H3K9me3 can therefore be used as marker for chromocenters. H3K9me3 foci in 2C⁺ cells were morphologically distinct from those of 2C⁻ cells (Fig. 2A). They were 2.3-fold fewer (3.89 ± 0.19 foci/nucleus) (Fig. 2B), and occupied 2.4-fold larger area (4.76 ± 0.33 µm²) in 2C⁺ as compared with both ESCs (8.88 ± 0.30 foci/nucleus; 1.99 ± 0.07 µm²) and 2C⁻ cells (8.72 ± 0.25 foci/nucleus; 1.93 ± 0.08 µm²) (Fig. 2C). These results suggest that H3K9me3 heterochromatin undergoes massive spatial reorganization, during the reprogramming of ESCs into 2C-like state. Importantly, the levels of H3K9me3 remain unchanged among ESCs, 2C⁻ and 2C⁺ cells, indicating that the remodeling of chromocenters was not due to loss of H3K9me3 (fig. S3D). The increased size of the H3K9me3 foci and the reduction in the number of H3K9me3 foci per nucleus might be due to the decompaction or fusion of several chromocenters.

Fig. 2.

We then imaged global DNA organization with Stochastic Optical Reconstruction super-resolution Microscopy (STORM). DNA was labelled using the nucleotide analogue 5-ethynyl-2’-deoxycytidine (EdC) (31, 32). DNA images were quantified by Voronoi tessellation analysis (33, 34) which can precisely determine the DNA density based on the number of localizations in each Voronoi tessel (see Methods). Voronoi analysis showed a marked decrease in the localization density of the chromatin in 2C⁺ cells (Fig. 2D). Furthermore, Voronoi analysis confirmed the decreased DNA density as a function of the GFP intensity in 2C⁺ cells (Fig. 2E). Interestingly, 2C⁻ cells were heterogeneous with respect to DNA density, with the majority of them showing low DNA density as compared with 2C⁺ cells, suggesting that DNA might undergo decompaction prior to GFP activation (Fig. 2E). Overall, the DNA decompaction of the chromatin fibers in 2C⁺ cells is in agreement with the chromatin landscape of early/late 2-cell embryos, which was reported to be more accessible by Assay for Transposase Accessible Chromatin with high-throughput sequencing (ATAC-seq) (11, 35).

H3K9me3 heterochromatin becomes rapidly formed following exit from the 2C-like state

We then asked whether 2C⁺ cells could undergo the reverse transition, exiting the 2C-like state and subsequently re-entering pluripotency, thereby becoming ESC-like cells. To answer this question about the kinetics of the reverse transition, we followed the expression of EGFP in FACS-sorted 2C⁺ cells 24 h, 48 h, 72 h and one week after sorting (Fig. 2F). Strikingly, over 60 % of the 2C⁺ cells in culture lost the expression of the MERVL reporter 24 h after sorting. Moreover, 48 h after sorting, only 6 % of the cells still expressed the reporter, suggesting rapid repression of the 2C program, and quick re-establishment of the pluripotency network (Fig. 2F). 72 h and 7 days (7 d) after sorting, EGFP expression levels were comparable to those derived from the endogenous fluctuation (“steady state”) of ESC cultures (Fig. 2F). The decay in EGFP levels was accompanied by a downregulation of MERVL, endogenous Dux, Zscan4, Nelfa, Zfp352 and Eif1a-like gene expression (Fig. 2G). These results indicate that 2C-like cells could revert their fate back to pluripotency after Dux overexpression, and that such transition occurs rapidly, as early as 24 h after sorting.

We then quantified the number and area of H3K9me3 foci during the 2C⁺ to ESC-like transition (Fig. 2H-J). Our results indicate that chromocenters underwent rapid re-formation and increased in number (24 h: 9.67 ± 0.50; 48 h: 7.07 ± 0.46; 72 h: 8.82 ± 0.44 foci/nucleus) as compared to 2C⁺ cells (3.89 ± 0.19 foci/nucleus), concomitantly to the loss of EGFP expression and to the exit from the 2C-like state (Fig. 2H, 2I). The areas of chromocenters in ESC-like cells were similar across the different time-points analyzed (24 h: 1.54 ± 0.15; 48 h: 1.77 ± 0.14; 72 h: 1.75 ± 0.10 µm²) and smaller of those of 2C⁺ cells (4.76 ± 0.33 µm²) (Fig. 2H, 2J). These results suggest that the in vitro transition of the 2C⁺ cells toward ESC-like state can be used as a model system to study chromocenter formation and chromatin reorganization occurring during early development.

SMARCAD1 and TOPBP1 associate with H3K9me3 in ESCs and can maintain heterochromatin foci

The results of the iPOTD revealed TOPBP1 as a potential regulator of chromocenter reorganization. SMARCAD1 has been shown to interact with TOPBP1 in yeast and human cells (27) (Fig. 1K, 1L). Interestingly, SMARCAD1, a SWI/SNF-like chromatin remodeler, is known to promote heterochromatin maintenance during DNA replication in terminally differentiated cells and silencing of endogenous retroviruses in ESCs (36, 37). Nonetheless, it is not known whether SMARCAD1 plays a role in 2C-like fate transition and early embryo development.

We, therefore, decided to investigate SMARCAD1 and TOPBP1 in 2C⁺ cells undergoing the transition to ESC-like cells where chromocenters are formed de novo. First, we evaluated whether SMARCAD1 was associated with TOPBP1 and/or H3K9me3 in ESCs, since TOPBP1 enriched in the 2C⁻ and Luc chromatomes. We observed that both SMARCAD1 and TOPBP1 were associated with H3K9me3 (Fig. 3A). However, we could not detect direct association between TOPBP1 and SMARCAD1, likely because of the suboptimal efficiency of available antibodies. Indeed, we cannot exclude their interaction as was shown previously in yeast and human cells (27). These data suggest SMARCAD1 and TOPBP1 to be potential regulators of H3K9me3 heterochromatin. Since SMARCAD1 is a chromatin remodeler, we then asked whether it localized at chromocenters in ESCs. We found that SMARCAD1 co-localized with H3K9me3 in heterochromatin foci of chromocenters in both ESCs and 2C⁻ cells (Fig. 3B-D). In contrast, the expression of SMARCAD1 decreased in 2C⁺ cells, where foci were much reduced in number (Fig. 3C, 3D). We then asked whether SMARCAD1 depletion would increase the fraction of 2C⁺ cells. We depleted Smarcad1 using two independent sgRNAs, as confirmed comparing to a control sgRNA targeting luciferase (fig. S3E). SMARCAD1 depletion resulted in no major impact in the 2C⁺ conversion either in the endogenous fluctuation or in Dux-induced cells when inspected at the steady state (Fig. 3E). We then investigated Smarcad1-depleted cells 24 h, 48 h, 72 h after the 2C⁺ exit. Control KO cells (sgLuc) followed comparable exit kinetics as compared to non-transfected (NT) 2C⁺ cells (fig. S3F). However, SMARCAD1 depletion resulted into a tendency to increased percentage of 2C⁺ cells at all time point after the exit (Fig. 3F). Accordingly, the nuclear distribution of SMARCAD1 during exit from the 2C-like state changed. We first observed a diminution in SMARCAD1 signal as ESCs started to express the MERVL reporter, attaining severe reduction of SMARCAD1 in 2C⁺ at the 24 h time point (Fig. 3H, 3I). SMARCAD1 nuclear signal was then gradually recovered in the heterochromatin foci as 2C⁺ cells were converted in ESC-like cells up to the 72 h from the exit, indicating reversibility of foci formation (Fig. 3H, 3I). Surprisingly, the fraction of cells that repressed retroelements within 24 h from the 2C⁺ exit (ESC-like at 24 h) already showed SMARCAD1 enriched foci (Fig. 3H). Altogether, these results suggest that SMARCAD1 was severely reduced from chromatin as ESCs progress to the 2C-like state and, later, SMARCAD1 nuclear distribution was reverted during the 2C⁺ exit.

Fig. 3.

We then used published single-cell RNA-seq (scRNA-seq) data (38) and found Smarcad1 expression starting at the 2-cell stage, but increasing at the 4-cell stage embryo, which is the time when chromocenters compact during mouse embryo development (Fig. 3J). Notably, Topbp1 showed a similar expression profile during preimplantation development (Fig. 3J). Similar to what observed for Smarcad1-depleted cells, Topbp1-depleted cells showed a tendency to increased percentage of 2C⁺ cells at 24 h, 48 h and 72 h after 2C⁺ exit (Fig. 3G). To further confirm these results and to investigate the role of TOPBP1 in the regulation of heterochromatin foci, we generated knocked-down ESC clones carrying shSmarcad1 or shTopbp1 (fig. S4A). We observed that the number of foci decreased and their area become larger after either knocking down Smarcad1 or Topbp1, with respect to scramble controls (fig. S4B, S4C). Moreover, larger and fewer chromatin foci were visible in 2C⁺ cells when compared to ESCs and 2C⁻ cells (fig. S4D). We confirmed these results investigating Topbp1-depleted cells 24 h, 48 h, 72 h after the 2C⁺ exit. We observed a decreased number of foci and their larger area at all time points analyzed after the exit (fig. S4E-G). These data suggest that heterochromatin foci are maintained in ESCs by SMARCAD1 and TOPBP1 and the depletion of both of these two proteins leads to a remodeling of the H3K9me3 foci.

Next, we asked about the functional interaction of SMARCAD1 and TOPBP1 and thus we evaluated the localization of SMARCAD1 after knocking down Topbp1. We found a significant reduction of SMARCAD1 co-localization with H3K9me3 in heterochromatin foci in Topbp1-depleted cells (Fig. 3K, 3L), suggesting that SMARCAD1 and TOPBP1 might work as complex in the maintenance of heterochromatin foci.

SMARCAD1 and TOPBP1 are necessary for early embryo development

Collectively, our findings suggested that both SMARCAD1 and TOPBP1 could be potential regulators of H3K9me3 heterochromatin in the 2C⁺ transition. With this in mind, we aimed at investigating their function in preimplantation embryos. We injected zygote-stage (E0.5) embryos with morpholino antisense oligos (MO) targeting Smarcad1 or Topbp1 along with a scrambled control morpholino (Ctrl MO) (Fig. 4A and fig. S5A). As expected from MO, which acts by blocking translation, SMARCAD1 was degraded from the 2-cell stage, and a reduction in its levels was observed up to the 8-cell stage, in Smarcad1 MO-injected embryos (fig. S5B, S5C). We could not image the degradation of TOPBP1 since available anti-TOPBP1 antibodies provide unspecific signal in immunofluorescence experiments. It is noteworthy that SMARCAD1 localizes exclusively in the nucleus of preimplantation embryos (fig. S5B). We observed that embryos developed slower than normal when Smarcad1 was silenced (Fig. 4A, 4B). Indeed, they did not show the formation nor expansion of a blastocoel cavity at the early blastocyst stage, indicating a severe developmental delay (Fig. 4A, 4B). Notably, 68 % of the embryos deficient for Smarcad1 arrested and did not develop until the late blastocyst stage (Fig. 4A, 4B). In the case of Topbp1 silencing, we observed an even a more severe phenotype. All the embryos, 100 % of the Topbp1 MO-injected ones, did not develop and arrest at 4-cell stage (Fig. 4A, 4B).

Fig. 4.

Since we observed that both SMARCAD1 and TOPBP1 were necessary for embryo developmental progression, we decided next to image H3K9me3 upon depletion of SMARCAD1 or of TOPBP1 (fig. S5A). H3K9me3 signal was significantly reduced in the embryos injected with Smarcad1 MO already at the 8-cell stage (E2.5), almost one day earlier than early blastocyst (E3.5), when the developmental delay was morphologically visible (Fig. 4C, 4D and fig. S5A). In Topbp1 MO embryos, we did not observe decreased intensity of the H3K9me3 signal since the developmental arrest was present already at 4-cell stage and variation in this histone mark might be clearly measurable only starting from morula stage (fig. S5A, S5D, S5E). On the other hand, we analyzed HP1b, a major component of constitutive heterochromatin which binds to both DNA and to H3K9me3 (39, 40). We observed a major remodeling of heterochromatin in both 2-cell and 4-cell Topbp1 MO arrested embryos, as indicated by the spreading and increased signal of HP1b (fig. S5F-S5I).

Finally, given that we observed SMARCAD1 reduction in heterochromatin foci in Topbp1-depleted cells (Fig. 3K, 3L), we investigated SMARCAD1 level in Topbp1 MO in 2-cell and 4-cell arrested embryos. We observed a severe reduction of SMARCAD1 that was even more pronounced when analyzing the pool of 2-cell arrested embryos (Fig 4E-4H and S5J-S5L).

Collectively, these results confirm the functional interaction between SMARCAD1 and TOPBP1showing that Smarcad1 or Topbp1 knockdown impair mouse embryo development and that their role in the maintenance of H3K9me3 heterochromatin foci might contribute to the developmental arrest. Overall, our results suggest that both SMARCAD1 and TOPBP1 contribute to proper early embryo development.

Discussion

Heterochromatin formation during early embryogenesis is a fundamental aspect of development (4). Here, we have reported that the transition from the 2C-like to the pluripotent state is a robust in vitro model system to study heterochromatin foci establishment and their reorganization in early embryo development. During the 2C-like to pluripotency transition, we found that heterochromatin foci are re-formed along with the DNA compaction of the chromatin fibers. Unlike previous reports that focused exclusively on transcriptional changes (13, 14, 41), our study exploited chromatin proteomics by genome capture to unravel an additional layer of information and complexity in the 2C-like system. Thus, we provided a detailed characterization of the stepwise chromatome dynamics occurring during the 2C-like state transition. Remarkably, we identified the chromatin remodeler factor SMARCAD1 and TOPBP1, a binding protein interacting with topoisomerase to contribute to embryo development. Depletion of SMARCAD1 or TOPBP1 in preimplantation embryos led to severe developmental arrest and to a substantial remodeling of H3K9me3 heterochromatin foci. These findings have important implications because the establishment and maintenance of heterochromatin foci during embryo development is a key step in the embryonic totipotent program of the 2-cell stage toward pluripotency (1, 42).

Endogenous retroviruses (ERVs) are transposable elements flanked by long terminal direct repeats (LTRs) (43, 44). Tight control of ERVs and their transposable activity is essential for genome integrity and play an important role in early development and pluripotency (43, 44). H3K9me3 has been associated with retrotransposons through the KRAB-associated protein 1, KAP1 (45). KAP1 led to the silencing of ERVs in ESCs by inducing H3K9me3 heterochromatin formation via the recruitment of the H3K9 histone methyltransferase SETDB1 (45-47). SMARCAD1 was discovered recently to directly interact with KAP1 and therefore be an important regulator of the KAP1-SETDB1 silencing complex in ESCs (36, 48). SMARCAD1 is also a key factor for ERV silencing in ESCs (36), where it remodels nucleosomes (49). Of note, although SMARCAD1 is highly expressed in ESCs, its depletion does not affect pluripotency (36, 49, 50).

SMARCAD1 has been described in ESCs, yet its function in 2C-like cells has not been explored. Our observation that SMARCAD1 enriches in H3K9me3 heterochromatin foci during the transition from the 2C-like state to pluripotency and that it contributes to early mouse embryo development is aligned with the observations previously reported in ESCs. It will be interesting in the future to study whether SMARCAD1 can tether the KAP1-SETDB1 to directly induce the formation of H3K9me3 heterochromatin foci at the exit of the 2-cell stage in the embryos. Recently, the H3K9 histone methyltransferase SUV39H2 has been reported to catalyze de novo H3K9me3 in the paternal pronucleus after fertilization (51). Yet, Suv39h2 downregulation in zygote-stage embryos did not translate on appreciable changes in H3K9me3 levels on the maternal chromatin. This opens up the possibility that different methyltransferases, and their regulators like SMARCAD1, could be responsible for H3K9me3 acquisition in this early developmental stage.

Topoisomerases likely cooperate with the chromatin remodeling factor SMARCAD1 in yeast (27). We observed that SMARCAD1 and TOPBP1 both interact with H3K9me3 suggesting they might work as a complex, this result is also confirmed by the observed reduction of SMARCAD1 in heterochromatin foci in both Topbp1-depleted cells and Topbp1-depleted embryos. We showed that topoisomerase inhibition led to an increase in the fraction of 2C-like cells and cell cycle arrest in the G2/M phase. Moreover, the knock-down of TOPBP1 leads to a severe developmental arrest. Thus, it is also tempting to speculate that cell cycle progression, especially since we observed that 2C⁺ cells might be arrested in the G2/M phase, has a role in regulating SMARCAD1 recruitment and/or function on chromatin during the 2C⁺ exit. In the in vivo scenario, this prolonged G2/M phase might be necessary to rewire specific epigenetic modifications in the 2-cell blastomeres to allow heterochromatin formation. TOPBP1, being a DNA topoisomerase 2-binding protein, might have a role in this process. This is a key step before the blastomeres can embark into the correct developmental process, as proposed for early Drosophila embryos (52).

By using chromatin proteomics, we have provided additional data that will help to elucidate the molecular intricacies of the 2C-like state and early mammalian development. In the current study, we focused on heterochromatin establishment and we identified SMARCAD1 and TOPBP1, which both interact with H3K9me3. SMARCAD1 might act in the complex as the remodeler factor that, by regulating methyltransferases, can facilitate H3K9me3 deposition at the exit of the totipotent 2-cell stage when heterochromatin is established de novo. Although we could not collect robust data on the alteration of the 2C program, we have indication of its prolonged activity when either SMARCAD1 or TOPBP1 are knocked down, in line with a role in regulating early development in the maintenance of heterochromatin and the regulation of the 2C program.

Materials and Methods

Cell lines and culture conditions

E14Tg2a mouse ESCs were cultured in gelatinized plates in high glucose DMEM supplemented with 15 % FBS (Sigma), GlutaMAX, sodium pyruvate, non-essential amino acids, penicillin/streptomycin, 100 µM 2-mercaptoethanol, 1000 U/ml mouse leukemia inhibitory factor (mLIF) (Millipore), 1 µM PD0325901 and 3 µM CHIR99021. After viral infection, ESCs were selected and maintained with ES medium containing the appropriate combination of selection drugs (250 µg/ml Geneticin (G418, Life Technologies), 0.5 µg/ml Puromycin (Life Technologies)). ESCs were treated with 2 µg/ml doxycycline (D9891, Sigma) for 24 h to induce Dux expression.

Lentivirus production and ESC infection

Lentiviral particles were produced following the RNA interference Consortium (TRC) instructions for viral production and cell infection (http://www.broadinstitute.org/rnai/public/). HEK293T cells were co-transfected with the lentiviral plasmid of interest (pCW57.1-Luciferase or pCW57.1-mDux-CA) and the viral packing vectors (pCMV-ΔR8.9 and pCMV-VSV-G) using the CalPhos mammalian transfection kit (631312, Clontech). pCW57.1-Luciferase and pCW57.1-mDux-CA were a gift from Stephen Tapscott (Addgene plasmids #99283 and #99284). Short hairpins targeting Smarcad1 (shSmarcad1), Topbp1 (shTopbp1 #1 and shTopbp1 #2) and a scramble control sequence (shScbl) were cloned into the pLKO.1-Hygro lentiviral vector (Addgene plasmid #24150). The lentiviral-containing medium was harvested from HEK293T cells at 48 h and 72 h after transfection, filtered and used for ESC infection. Two days after the last round of infection, ESCs were selected with the indicated concentration of the selection drug (see Cell culture).

Fluorescence-activated cell sorting (FACS)

Quantification of GFP positive cells and cell cycle analysis was performed with a LSR II Analyzer (BD Biosciences). For cell sorting, an Influx Cell Sorter (BD Biosciences) was used to sort the specified populations in each experiment.

Cell cycle analysis by flow cytometry

For cell cycle analysis of live cells, 5 ×10⁴ ESCs were plated per well in gelatin-coated 6-well plates one day before starting the experiment. At the moment of the assay, ESCs were trypsinized, collected and washed with PBS before incubation with ES medium supplemented with 10 µg/ml Hoechst 33342 (H1399, Thermo Fisher) for 30 min at 37 °C. Propidium iodide (PI) (1 µg/ml; P4864, Sigma) was added to stain dead cells. All flow cytometry data were processed and analyzed with FlowJo (v10).

Inhibition of DNA topoisomerases

To inhibit DNA topoisomerases, ESCs were treated with 500 nM of the topoisomerase I inhibitor camptothecin (CPT; ab120115, Abcam) and/or with 5 µM of the topoisomerase II inhibitor ICRF-193 (I4659, Sigma) for 12 h.

Immunostaining, image processing and quantification

Immunofluorescence staining of ESCs

ESCs were plated at a concentration of 56.000 cells/cm² in gelatin-coated borosilicate glass bottom Nunc Lab-Tek (155411, Thermo Fisher) or µ-Slide (80827, Ibidi) 8-well chambers. Cells were fixed with 4 % paraformaldehyde (PFA) for 10 min and were then washed three times with PBS. Cells were permeabilized and blocked (10 % GS, 2.5 % BSA, 0.4 % Triton X-100) for 30 min at room temperature (RT). Incubation with the corresponding primary antibodies at the indicated dilutions lasted 3 h at 37 °C. Cells were then washed and incubated with Alexa Fluor (Molecular Probes, Invitrogen) secondary antibodies for 1 h at RT. For H3K9me3 and SMARCAD1 co-staining, cells were washed three times with PBS after secondary antibody incubation. Then cells were incubated with second primary antibody and the corresponding secondary antibody as indicated above. Finally, cells were washed three times with PBS containing DAPI for nuclear counterstain. Images were acquired on a Leica TCS SP5 confocal microscope equipped with a 63x oil objective.

The following antibodies were used: chicken anti-GFP (1:500; ab13970, Abcam), mouse anti-Oct-3/4 (1:200; sc-5279, Santa Cruz), rabbit anti-histone H3K9me3 (1:500; ab8898, Abcam), mouse anti-SMARCAD1 (1:500; ab67548, Abcam), goat anti-chicken Alexa Fluor 488, goat anti-mouse Alexa Fluor 568, goat anti-rabbit Alexa Fluor 568, goat anti-mouse Alexa Fluor 647. All secondary antibodies were provided by Molecular Probes (Invitrogen).

EdC incorporation and DNA labelling

To label DNA, a 14 h incorporation pulse of 5-ethynyl-2’-deoxycytidine (EdC; T511307, Sigma) at 2.5 µM was performed in ESCs, in parallel to doxycycline treatment. Cells were plated in gelatin-coated borosilicate glass bottom chambers at a concentration of 56.000 cells/cm² in ES medium supplemented with EdC for 14 hours. At the end of EdC incorporation, ESCs were fixed with PFA 4 % (43368, Thermo Fisher Alfa Aesar) and permeabilized with 0.4 % Triton X-100. Click chemistry reaction was performed by incubating cells for 30 min at RT in click chemistry buffer: 100 mM Hepes pH 8.2, 50 mM Amino Guanidine (396494, Sigma), 25 mM Ascorbic Acid (A92902, Sigma), 1 mM CuSO₄, 2 % Glucose (G8270, Sigma), 0.1 % Glox solution [0.5 mg/ml glucose oxidase, 40 mg/ml catalase (G2133 and C100, Sigma)] and 10 mM Alexa Fluor 647 Azide (A-10277, Thermo Fisher) (31, 32, 53). After washing the samples three times with PBS, we directly proceeded to perform STORM imaging.

STORM imaging

Stochastic Optical Reconstruction Microscopy (STORM) imaging was performed on a N-STORM 4.0 microscope (Nikon) equipped with a CFI HP Apochromat TIRF 100x 1.49 oil objective and a iXon Ultra 897 camera (Andor) with a pixel size of 16 µm. This objective/camera combination provides an effective pixel size of 160 nm. STORM images were acquired with 10 msec exposure time for 60000 frames using highly inclined (HILO) illumination. An activator/reporter pair strategy was used with AF405 and AF647 fluorophores, respectively. Continuous imaging acquisition was performed with simultaneous 405 nm and 647 nm illumination. 647 nm laser was used at constant ∼2 kW/cm² power density. 405 nm laser was used at low laser power and gradually increased during the imaging to enhance fluorophore reactivation and to maintain the density of localizations per frame constant. Before STORM imaging, we acquired conventional fluorescence images of GFP for each nucleus to discriminate between 2C⁻ and 2C⁺ cells. Imaging buffer composition for STORM imaging was 100 mM Cysteamine MEA (30070, Sigma), 1 % Glox Solution and 5 % Glucose (G8270, Sigma) in PBS.

STORM images were analyzed and rendered in Insight3 as previously described (54, 55). Localizations were identified based on an intensity threshold and the intensity distribution of their corresponding Point Spread Functions (PSFs) fit with a 2D Gaussian to determine the x-y positions of their centers with high accuracy (∼20 nm).

Voronoi Tesselation analysis

For Voronoi Tesselation analysis, we used the list of localization from STORM (33, 34) and then we used a previously developed custom-made Matlab script (31). X-y coordinates of the localizations were used to generate the Voronoi polygons. Local densities were defined as the inverse value of the area of each Voronoi polygon. For visualization, we color-coded each Voronoi polygon based on their area, from yellow for the smallest polygons (density > 0.01 nm⁻²), to blue for larger polygons (density < 0.0001 nm⁻²). Finally, the largest 0.5 % of polygons were set to black. For each nucleus, we computed the mean Voronoi density (nm⁻²) as a measure of global DNA compaction.

For the GFP intensity score, we quantified the GFP conventional images (488 nm channel) with lower intensities in order to assign a GFP intensity score to each nucleus. We summed the fluorescence intensity ADU counts inside each nucleus and divided it by the total number of pixels to obtain the average GFP intensity. Then, we used the distribution of GFP intensities from the different nuclei to normalize the values, obtaining a GFP intensity score ranging from 0 (less bright) to 1 (most bright). We then performed a cell-by-cell analysis of the relation between GFP intensity score and global chromatin compaction obtained from Voronoi Tesselation analysis.

Immunofluorescence of preimplantation embryos

Preimplantation embryos at E1.5 and E2.5 stages were fixed with 2 % PFA for 10 min at RT, permeabilized (0.25 % Triton X-100) for 10 min, and then blocked (3 % BSA) for 1 h at 37 °C. Incubation with the corresponding primary antibodies at the indicated dilutions in 1 % BSA lasted one overnight at 4 °C. After washing, embryos were incubated with Alexa Fluor (Molecular Probes, Invitrogen) secondary antibodies diluted in 1 % BSA for 1 h at 37 °C. Finally, embryos were washed and transferred to an imaging buffer containing DR (1:500; 62251, Thermo Fisher) for DNA staining. Images were acquired on a Leica TCS SP8 STED3X confocal microscope equipped with a 63x oil objective.

The following antibodies were used: rabbit anti-histone H3K9me3 (1:500; ab8898, Abcam), mouse anti-SMARCAD1 (1:250; ab67548, Abcam), rabbit anti-HP1β (1:200; ab10478, Abcam), goat anti-rabbit Alexa Fluor 488 and goat anti-mouse Alexa Fluor 488. All secondary antibodies were provided by Molecular Probes (Invitrogen).

Image processing and quantification

Immunofluorescence images were processed and analyzed with the ImageJ software (https://imagej.net/download/). All immunofluorescence images were acquired with z-stacks. Z-stacks were projected using the maximum intensity z-projection type. For SMARCAD1 nuclear signal analysis, manual selection of nuclear area was performed and integrated intensity was measured. For SMARCAD1-H3K9me3 co-immunofluorescence images, a Gaussian blur filtering (σ = 0.5) was applied to the SMARCAD1 channel. Fluorescence intensities of H3K9me3 foci were analyzed using the 3D Object Counter function (https://imagej.net/3D_Objects_Counter, ImageJ). Co-localization analysis was done using the JACoP plugin (https://imagej.net/JaCoP, ImageJ). Manders’ coefficient was calculated with the JACoP plugin. Manders’ coefficient was used as a co-localization indicator because of its independence of the intensity of the overlapping pixels. For the quantification of H3K9me3 and SMARCAD1 fluorescence intensities in preimplantation embryos, manual selection of the nuclear area was performed for each blastomere. Fluorescent signals were measured and then normalized by the average cytoplasmic signal (background) in each condition. For the normalization step, the fluorescence intensity of a squared shape of equal size was taken for each individual blastomere.

RNA extraction and quantitative real-time PCR (qRT-PCR)

RNA was extracted from pelleted or sorted ESCs using the RNA isolation RNeasy Mini kit (QIAGEN), according to the manufacturer protocol. RNA was reverse-transcribed with iScript cDNA Synthesis kit (Bio-Rad). qRT-PCR reactions were performed using LightCycler 480 SYBR Green I Master (Roche) in a LightCycler 480 (Roche) instrument, according to the manufacturer recommendations. The oligos used are listed in Table 1. qRT-PCR data was normalized to Gapdh or β-actin expression. For each sample, we had at least a technical duplicate.

Chromatin-bound proteome profiling by genome capture (iPOTD)

ESCs were plated at a concentration of 34.000 cells/cm² in gelatin-coated 150-mm dishes. Then, ESCs were pulsed for 24 h with 0.1 µM 5-ethynyl-2’-deoxyuridine (EdU; T511285, Sigma), in parallel to doxycycline treatment. Sorted luciferase (± EdU), 2C⁻ +EdU and 2C⁺ +EdU cells were fixed with 1 % PFA, quenched with 0.125 mM glycine (pH 7) and harvested immediately after sorting. Of note, ∼ 10⁷ cells were sorted per replicate and condition. Cells were later processed as described previously to extract the chromatin-bound proteins (16, 17).

Mass spectrometry analysis

Sample preparation

Eluted proteins were reduced with dithiothreitol (37 °C, 60 min) and alkylated in the dark with iodoacetamide (25 °C, 20 min) prior to sequential digestion with endoproteinase LysC (1:10 w:w, 37 °C, overnight; 129-02541, Wako) and trypsin (1:10 w:w, 37 °C, 8 h) according to filter-aided sample preparation procedure (56). After digestion, the peptide mixtures were acidified with formic acid and desalted with a MicroSpin C18 column (The Nest Group, Inc) prior to LC-MS/MS analysis.

Chromatographic and mass spectrometric analysis

Samples were analyzed using a LTQ-Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific, San Jose, CA, USA) coupled to an EASY-nLC 1200 (Thermo Fisher Scientific (Proxeon), Odense, Denmark). Peptides were loaded directly onto the analytical column and were separated by reversed-phase chromatography using a 50-cm column with an inner diameter of 75 μm, packed with 2 μm C18 particles spectrometer (Thermo Scientific, San Jose, CA, USA).

Chromatographic gradients started at 95 % buffer A and 5 % buffer B with a flow rate of 300 nl/min for 5 minutes and gradually increased to 22 % buffer B and 78 % A in 79 min and then to 35 % buffer B and 65 % A in 11 min. After each analysis, the column was washed for 10 min with 10 % buffer A and 90 % buffer B. Buffer A was 0.1 % formic acid in water and buffer B was 0.1 % formic acid in acetonitrile.

The mass spectrometer was operated in positive ionization mode with nanospray voltage set at 1.9 kV and source temperature at 275 °C. Ultramark 1621 was used for external calibration of the FT mass analyzer prior the analyses, and an internal calibration was performed using the background polysiloxane ion signal at m/z 445.1200. The acquisition was performed in data-dependent acquisition (DDA) mode and full MS scans with 1 micro scans at resolution of 120,000 were used over a mass range of m/z 350-1500 with detection in the Orbitrap mass analyzer. Auto gain control (AGC) was set to 1E5 and charge state filtering disqualifying singly charged peptides was activated. In each cycle of data-dependent acquisition analysis, following each survey scan, the most intense ions above a threshold ion count of 10000 were selected for fragmentation. The number of selected precursor ions for fragmentation was determined by the “Top Speed” acquisition algorithm and a dynamic exclusion of 60 seconds. Fragment ion spectra were produced via high-energy collision dissociation (HCD) at normalized collision energy of 28 % and they were acquired in the ion trap mass analyzer. AGC was set to 1E4, and an isolation window of 1.6 m/z and maximum injection time of 200 ms were used. All data were acquired with Xcalibur software.

Digested bovine serum albumin (P8108S, NEB) was analyzed between each sample to avoid sample carryover and to assure stability of the instrument and QCloud has been used to control instrument longitudinal performance during the project (57).

Data analysis

Acquired spectra were analyzed using the Proteome Discoverer software suite (v2.3, Thermo Fisher Scientific) and the Mascot search engine (58) (v2.6, Matrix Science). The data were searched against a Swiss-Prot mouse database (as in October 2019) plus a list of common contaminants and all the corresponding decoy entries (22). For peptide identification a precursor ion mass tolerance of 7 ppm was used for MS1 level, trypsin was chosen as enzyme, and up to three missed cleavages were allowed. The fragment ion mass tolerance was set to 0.5 Da for MS2 spectra. Oxidation of methionine and N-terminal protein acetylation were used as variable modifications whereas carbamidomethylation on cysteines was set as a fixed modification. False discovery rate (FDR) in peptide identification was set to a maximum of 5 %. The analysis of specific chromatin interactors was carried out with SAINT (v2, Significance Analysis of INTeractome) as previously described (22, 58). Replicate 2 of the 2C⁺ condition was excluded for SAINT analysis due to abnormal lower peptide-spectrum matches (PSM) observed in this run. Hierarchical clustering of all the chromatome replicates was computed and visualized using Instant Clue (59) v0.5.2 (https://www.instantclue.uni-koeln.de/). Pearson’s correlation coefficients were calculated using the Prism software (v9.0, GraphPad, San Diego, CA). To identify proteins shared by the 2C⁻ and Luc chromatomes and not enriched in the 2C⁺ chromatome, an average enrichment value was computed from the respective pairwise comparisons (i.e., Luc vs 2C⁻; Luc vs 2C⁺; 2C⁻ vs Luc; 2C⁻ vs 2C⁺) and then selecting those hits that were more commonly enriched among the 2C⁻ and Luc chromatomes (FC ≥ 2). Gene ontology (GO) term enrichment was performed with GO Enrichment Analysis using the PANTHER tool (60, 61) (https://geneontology.org/). Protein interaction data were retrieved from the STRING database v11.0 (62) and visualized with Cytoscape v3.8.2 (63).

Western blot (WB) analysis

Protein extracts were boiled in Laemmli buffer, run in precast protein gel (Mini-PROTEAN TGX; 4561084, Bio-Rad) and then transferred to immuo-blot polyvinylidene difluoride membranes (162-0177, Bio-Rad). The membranes were blocked and incubated with the indicated primary antibodies overnight at 4 °C [rabbit anti-histone H3K9me3 (1:500; ab8898, Abcam), mouse anti-OCT4 (1:500; sc-5279, Santa Cruz), rabbit anti-TOPBP1 (1:1000; ab2402, Abcam) and mouse anti-SMARCAD1 (1:500; ab67548, Abcam)].

After washing, membranes were incubated with specific peroxidase-conjugated secondary antibodies [sheep anti-mouse IgG HRP-linked (1:1000; NA931, GE Healthcare) and donkey anti-rabbit IgG HRP-linked (1:2000; NA934, GE Healthcare)] and visualized on an Amersham Imager 600 (29083461, GE Healthcare Life Sciences).

Dot blot analysis

Samples were spotted in triplicates in 1 µl dots onto a nitrocellulose membrane (0.2 µM, Amersham Protan), air-dried, and detected following standard blotting procedures with the corresponding antibodies (rabbit anti-histone H3 (1:1000; ab1791, Abcam), mouse anti-vinculin (1:1000; V9131, Merck)). Quantification of dot blots was performed by Image Studio Lite software (v5.2, LI-COR, Biosciences). For quantification, each protein was normalized to its background signal.

Co-immunoprecipitation (co-IP)

ESCs were washed with PBS, trypsinized and pelleted. Freshly collected pellets were used for co-IP experiments. ESCs were lysed for 10 min at 4 °C (10 mM tris-HCl (pH 7.4), 10 mM KCl, 15 mM MgCl₂) in the presence of Protease Inhibitor Cocktail (PIC; 5056489001, Roche). Nuclei were pelleted and lysed for 1 h at 4 °C (300 mM NaCl, 50 mM Hepes (pH 7.5), 0.5 % NP-40, 2.5 mM MgCl₂) in the presence of benzonase endonuclease (100 U/ml; 71205, Novagen). Benzonase removes DNA and RNA, thereby releasing DNA-binding proteins. After nuclear lysis, samples were pelleted and supernatants containing protein extracts were harvested. Protein concentrations were determined using the Pierce BCA protein assay (23225, Thermo Fisher) according to the manufacturer’s protocol. For endogenous IPs, an aliquot of protein extract (0.8-1 mg) diluted in dilution buffer (300 mM NaCl, 50 mM Hepes (pH 7.5), 0.5 % NP-40, 5 mM EDTA + PIC) was incubated with 25 μl of pre-washed Dynabeads Protein G (10004D, Thermo Fisher) and with 4 µg of the antibody of interest overnight at 4 °C. For mock IP, ChromPure Mouse (015-000-003, Jackson Immunoresearch) or Rabbit (011-000-003, Jackson Immunoresearch) IgGs were used. Samples were then washed three times. Elution was performed by incubating the dried beads with 30 μl of 2x Laemmli buffer (1610747, Bio-Rad) supplemented with 50 mM of 2-mercaptoethanol for 10 min at 95 °C. After eluting the samples, we directly proceeded to perform western blotting.

The following antibodies were used: rabbit anti-histone H3K9me3 (ab8898, Abcam) and rabbit anti-TOPBP1 (ab2402, Abcam).

CRISPR-Cas9 plasmid generation and delivery

Single guide RNAs (sgRNAs) targeting each of the specific target genes were retrieved from the Mouse CRISPR Knockout Pooled Library (Addgene #73632). Two sgRNA sequences were selected per gene of interest (for sgRNAs sequences, see Table 2). The sgRNAs with the highest on-target activity score (Rule Set 2) were selected for assembly into the CRISPR-Cas9 vector. An sgRNA targeting the luciferase sequence was also included as control. Primers containing sequences for the sgRNAs were annealed in the presence of T4 ligation buffer (Thermo Fisher) and T4 PNK (NEB) in a heat block (30 °C for 30 min, 95 °C for 5 min and slow cool down to RT). Annealed primers were then cloned into the pU6-(BbsI)_CBh-Cas9-T2A-mCherry plasmid following a one-step cloning reaction. pU6-(BbsI)_CBh-Cas9-T2A-mCherry was a gift from Ralf Kuehn (Addgene plasmid #64324).

Table 2

Table 3

To generate CRISPR-Cas9-targeted ESCs, cells were nucleofected with 4 µg of the sgRNA-containing plasmid individually following the Amaxa Mouse ES cell Nucleofector kit recommendations (VPH-1001, Lonza). Later, ESCs were FACS-sorted 48 h after nucleofection to enrich for the modified cells.

Zygote collection and culture

Embryos were collected at E0.5 from 6 to 10 weeks BDF1 female mice (Charles River Laboratories) following 5 IU pregnant mare’s serum gonadotrophin (PMSG) and 5 IU human chorionic gonadotropin (hCG) injections at 48 hours intervals. Female mice were mated with BDF1 male mice immediately after hCG injection. Embryos were collected from the oviducts 24 hours post-hCG and were briefly cultured in M2 medium supplemented with 0.2 mg/ml hyaluronidase (H3506, Sigma) to remove cumulus cells. Cumulus-free embryos were washed and cultured with Advanced KSOM medium (MR-101-D, Millipore) at 37 °C until microinjection.

Microinjection of morpholino antisense oligos

Morpholino antisense oligos (MO) for Smarcad1 and Topbp1 and non-targeting control were designed and produced by Gene Tools (Gene Tools, LLC). MOs were microinjected into the cytoplasm of E0.5 embryos using a Narishige micromanipulator system mounted on an Olympus IX71 inverted microscope. Embryos were immobilized using a holding pipette and MOs were then microinjected using a Narishige pneumatic microinjector (IM-300, Narishige). After microinjection, embryos were cultured in Advanced KSOM medium in low oxygen conditions (5 % CO₂, 5 % O₂) at 37 °C for 5 days (until E5.5). Preimplantation development was examined every 24 hours using an AMG EVOS microscope.

The following MO sequences were used:

Control MO: TCCAGGTCCCCCGCATCCCGGATCC;

Smarcad1 MO: ATATTGGGAGGAACCACCACCCTGA;

Topbp1 MO: ACGGCTCTTGGTCATTTCTGGACAT;

All morpholino sequences are written from 5’ to 3’ and they are complementary to the translation-blocking target.

All animal experiments were approved and performed in accordance with institutional guidelines [Parc de Recerca Biomèdica de Barcelona (PRBB), Barcelona, Spain] and in accordance with the Ethical Committee for Animal Experimentation (CEEA) number PC-17-0019-PI, approved by La Comissió d’Experimentació Animal, Departament de Territori i Sostenibilitat, Direcció General de Polítiques Ambientals i Medi Natural, Generalitat de Catalunya.

Statistical analysis

As specified in the figure legends, data are presented either as scatter dot plots with line at mean ± SD or at median ± interquartile range, bar graphs showing mean ± SD, min to max boxplots with line at median, or as violin plots showing median and quartiles. All statistical tests and graphs were generated using the Prism software (v9.0, GraphPad, San Diego, CA), unless otherwise indicated. Depending on the experimental setup, we used unpaired two-tailed Student’s t-test, multiple t-test, Fisher’s exact test, Mann-Whitney test, one-way ANOVA or two-way ANOVA with the indicated post-comparison test. In all cases, a P value p ≤ 0.05 was considered significant (p ≤ 0.05*; p ≤ 0.01**; p ≤ 0.001***; p ≤ 0.0001****; p > 0.05^ns, not significant).

Acknowledgements

We would like to thank M.-E. Torres-Padilla (Helmholtz Zentrum München), B.R. Cairns (Huntsman Cancer Institute) and B. Huang (UCSF) for sharing E14 ESCs containing the 2C::EGFP reporter, an independent E14 ESC clone containing the Dux-CA cassette and Insight3 software. We are grateful to S. Sdelci for critical reading of the manuscript. We also thank the CRG/UPF Flow Cytometry Unit, the CRG Advanced Light Microscopy Unit, and the PRBB animal facility (PRBB, Barcelona).

Funding

This work was supported by the European Union’s Horizon 2020 Research and Innovation Programme (CellViewer No 686637 to M.P.C.); Ministerio de Ciencia e Innovación, grant BFU2017-86760-P (AEI/FEDER, UE); and an AGAUR grant from Secretaria d’Universitats i Recerca del Departament d’Empresa I Coneixement de la Generalitat de Catalunya (2017 SGR 689 to M.P.C.); National Natural Science Foundation of China (No 31971177 to M.P.C.); Spanish Ministry of Economy, Industry and Competitiveness (MEIC) (PID2019-108322GB-100 to L.D.C.), and from AGAUR (L.D.C.). We acknowledge the support of the Spanish Ministry of Science and Innovation to the EMBL partnership, the Centro de Excelencia Severo Ochoa and the CERCA Programme. The CRG/UPF Proteomics Unit is part of the Spanish Infrastructure for Omics Technologies (ICTS OmicsTech) and it is a member of the ProteoRed PRB3 consortium which is supported by grant PT17/0019 of the PE I+D+i 2013-2016 from the Instituto de Salud Carlos III (ISCIII) and ERDF. R.S.-P. was supported by a FI-AGAUR PhD fellowship from the Secretaria d’Universitats i Recerca del Departament d’Empresa i Coneixement de la Generalitat de Catalunya and the co-finance of Fondo Social Europeo (2018FI_B_00637 and FSE). X.T. is supported by a FPI PhD fellowship from the Ministerio de Ciencia e Innovación (PRE2018-085107). S.A. is funded by the Ramon y Cajal program of the Ministerio de Ciencia, Innovación y Universidades and the European Social Fund under the reference number RYC-2018-025002-I, and the Instituto de Salud Carlos III-FEDER (PI19/01814). L.M. is supported by a grant for the recruitment of early-stage research staff FI-2020 (Operational Program of Catalonia 2014-2020 CCI grant no. 2014ES05SFOP007 of the European Social Fund) and La Caixa Foundation fellowship (LCF/BQ/DR20/11790016). M.P. was supported by a Severo Ochoa PhD fellowship from the Subprograma Estatal de Formación del Ministerio de Economía y Competitividad (BES-2015-072802). M.V.N. is funded by FP7/2007–2013 under an REA grant (608959) and Juan de la Cierva-Incorporación 2017.

Author contributions

R.S.-P. and M.P.C. conceptualized this work. R.S.-P. designed and performed most of the experiments with contributions from S.N., X.T., S.A., M.P., M.A.-B., J.L.G.-V, E.B., M.N.-R. Data were primarily analyzed by R.S.-P. with contributions from S.N., X.T., S.A., P.A.G.-G., D.C., E.B., E.S., L.M., M.N.-R., E.M. R.S.-P. and M.P.C. wrote the manuscript with input from S.N., X.T., M.V.N., S.A. and L.D.C. M.P.C., M.V.N., and L.D.C. supervised the project.

Competing interests

Authors declare that they have no competing interests.

Data and materials availability

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE (64) partner repository with the dataset identifier PXD019703. All other data needed to evaluate the conclusions in this study are present in the paper and/or the Supplementary Materials. Additional materials generated in this study are available from the corresponding author upon request.