CANTAC-seq analysis reveals E2f1 and Otx1 as a coupled repressor-activator pair co-modulating zygotic genome activation in Xenopus tropicalis

doi:10.21203/rs.3.rs-4885809/v1

Download PDF

Article

CANTAC-seq analysis reveals E2f1 and Otx1 as a coupled repressor-activator pair co-modulating zygotic genome activation in Xenopus tropicalis

https://doi.org/10.21203/rs.3.rs-4885809/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Zygotic genome activation (ZGA) is tightly associated with the modulation of chromatin accessibility via maternal transcription factors. However, due to technical limitations, it remains elusive how the chromatin regulatory landscape is established during Xenopus tropicalis (X. tropicalis) ZGA and DNA binding transcription regulators involved in this process have therefore been underexplored. Here, by developing CANTAC-seq, we generated a first genome-wide map of accessible chromatin of early X. tropicalis embryos and found that the open chromatin landscape is progressively established at cis-regulatory elements during ZGA. Based on the motif analysis and perturbation experiments, we demonstrated E2f1, a well-known transcriptional activator, maintains a repressive chromatin environment independent of its negative effect on cell cycle progression before the MBT. Moreover, we identified another maternal factor Otx1 counteracts the inhibitory function of E2f1. The dynamic balance between the two factors determines the temporal regulation of a set of genes required for zygotic gene transcription and germ layer differentiation.

Biological sciences/Developmental biology/Embryogenesis

Biological sciences/Genetics/Epigenomics

Biological sciences/Molecular biology/Transcription

Following fertilization, the sperm and the egg fuse to form a zygote, which will ultimately give rise to a new organism. Initially, the zygote genome is transcriptionally quiescent, and it only gradually becomes activated through a period called “maternal to zygotic transition”¹. This onset of transcription is referred to as the zygotic genome activation (ZGA), which is a universal process tightly controlled during embryogenesis, though the timing of ZGA varies largely among different species^2,3. For example, slow-developing species such as mouse and human require one or more days to finish ZGA, while fast-developing species such as zebrafish and frog can complete this process in just a few hours⁴. In frog, embryos undergo 12 rapid synchronous divisions within five hours, followed by asynchronous cleavages, with the induction of gap phases during a critical time named mid-blastula transition (MBT)^5,6. While the transcription of most zygotic genes coincides with the MBT during the major wave of ZGA, some zygotic genes required for frog embryo patterning begin to express during the minor wave of ZGA that occurs before the MBT^7,8.

The process of ZGA is known to coincide with extensive chromatin remodeling that drives essential cellular processes, including gene expression, DNA replication, and repair³. Accessible chromatin typically marks cis-regulatory elements such as promoters and enhancers⁹, which modulate the binding of specific DNA-binding transcription factors thereby the transcription of key developmental regulators during ZGA. In turn, transcription factors also dynamically coordinate with other chromatin modifiers to modulate the local DNA access during this process^10,11. Conceptually, the accumulation of activators and the loss of repressors during ZGA regulate local chromatin targets, and the balance between these factors determines the time of the activation of zygotic genes². Therefore, characterization of chromatin accessibility dynamics and identification of responsible transcription activators and repressors during ZGA are vital for advancing our knowledge of early embryogenesis.

Recent studies utilizing ATAC-seq or DNase-seq have unveiled the dynamics of chromatin accessibility during ZGA in several model organisms^12–17. These studies have demonstrated that regulatory regions often display accessible chromatin prior to gene activation and have highlighted the pivotal role of specific transcription activators in driving changes in chromatin state. For instance, the pioneer factor OCT4 has been shown to establish chromatin patterns during human ZGA, but not in mice¹⁴. In zebrafish, maternal factors Pou5f3, Sox19b, and Nanog are critical in chromatin opening and primed for transcriptional activity^11,17. As to the repressors, histones have been identified as common repressors of gene transcription during ZGA in both zebrafish and frogs^18,19. Additionally, DNA methyltransferase Dnmt1 and the methyl-CpG repressor Kaiso, both with low DNA-binding sequence specificity, have been demonstrated with a general repressive role in ZGA of frogs^20,21. So far, the involvement of sequence- or target-specific transcription repressors in chromatin changes during this process remains elusive.

Xenopus tropicalis (X. tropicalis) is a classical model organism for exploring fundamental questions in developmental biology²². This is attributed to its abundant egg supply and easily manipulated embryos. Prior studies have examined chromatin accessibility changes via ATAC-seq and DNase-seq following the MBT during X. tropicalis development^23–25. However, compared to other model organisms, the global chromatin accessibility landscape and its molecular dynamics prior to the MBT remain poorly understood due to technical limitations^23,24. In this study, we developed CANTAC-seq, concanavalin A (ConA) beads-based nucleus capture followed by Tn5-mediated accessible chromatin assay with sequencing, to profile the chromatin accessibility landscape of early X. tropicalis embryos. For the first time, we generated a genome-wide map of chromatin accessible regions from stage 7 to stage 13 (blastula to neurula), revealing the dynamics of chromatin accessibility during X. tropicalis early embryogenesis. Bioinformatics analysis revealed that E2f binding motifs are highly enriched at promoter regions opened during X. tropicalis ZGA. Further perturbation experiments showed that E2f1 plays a repressive role in the establishment of open chromatin landscape independent of its negative effect on cell cycle progression prior to the MBT. Moreover, we found that the transcription factor Otx1 antagonizes the inhibitory function of E2f1. Mechanistically, E2f1 and Otx1 co-regulate a set of genes required for zygotic gene transcription before the MBT based on the dynamic balance between the two factors.

Development of CANTAC-seq

High yolk content interfered with transposase activity and made it almost impossible to apply ATAC-seq on whole X. tropicalis embryos of early developmental stages^23,24. Therefore, in order to investigate the chromatin regulatory landscape during early development, we have developed CANTAC-seq, which entails the use of ConA-coated magnetic beads with high glycoprotein affinity to capture nuclei, followed by yolk removal and optimized on-beads tagmentation to extract open chromatin regions for sequencing (see Methods; Fig. 1a).

To validate the CANTAC-seq method, we first compared it with standard ATAC-seq on mouse embryonic stem cells (mESCs) and human K562 cells. As shown in Fig. 1b and Fig. S1a, the chromatin accessibility profiles generated by CANTAC-seq correlated well with those obtained by standard ATAC-seq on mESCs (R = 0.99) and K562 cells (R = 0.99). Quality control metrics, such as insert size distribution and enrichment around transcriptional start sites (TSS), were almost identical between CANTAC-seq and ATAC-seq from the same cell line (Fig. 1c,d and Fig. S1b,c). Open chromatin regions detected by both methods showed similar profiles at the promoters of pluripotent marker genes, including Nanog, Klf4, Pou5f1, and Sox2 in mESCs (Fig. 1e), as well as at the well-characterized human β-globin locus control region (LCR) and its downstream hemoglobin genes in K562 cells Fig. S1d). Then, we performed both methods on X. tropicalis embryos collected at stage 13. As shown in Fig. 1f, whereas as reported before, standard ATAC did not produce DNA fragments of distinct sizes corresponding to accessible DNA, mono-nucleosome, and di-nucleosome, our CANTAC-seq did successfully and the distribution of insert size was similar to that obtained from mESCs and K562 cells. These results collectively demonstrated that while the performance of the CANTAC-seq method is on par with standard ATAC-seq approaches for cellular samples, only CANTAC-seq enables successful mapping of accessible chromatin of early X. tropicalis embryos with high yolk content.

Accessible chromatin landscape during X. tropicalis ZGA

After establishing CANTAC-seq, to investigate the chromatin regulatory landscape during X. tropicalis ZGA, we mapped the global chromatin accessibility using CANTAC-seq for five developmental stages, including early, middle, and late blastula (stage 7, 8, and 9, respectively), which encompass the first major developmental transition MBT, the onset of gastrula (stage 10), and the onset of neurula (stage 13) (Fig. 2a). These stages encompass both the minor wave (stage 7–8) and the major wave (stage 9–13) of ZGA processes. Each stage was analyzed in two independent biological replicates, which all showed a high degree of correlation (Pearson’s correlation coefficient, 0.93-1.00) (Fig. S2a). Therefore, to increase the specificity, we combined the data and use common peaks identified from both replicates for subsequent analysis. Overall, we could identify 339, 23,731, 28,331, 44,481, and 47,726 accessible regions at stage 7, 8, 9, 10, and 13, respectively (Fig. 2b, Table S1). This demonstrated that while the chromatin was almost inaccessible at stage 7, a drastic opening was started at stage 8 before the MBT, followed by a progressive increase of accessibility from stage 9 to stage 13 (Fig. 2b,c). A substantial proportion of accessible regions established during the early stage was largely maintained through later stages. Specifically, 61% (207/339) of stage 7 peaks, 44% (10,543/23,731) of stage 8 peaks, 53% (14,917/28,331) of stage 9 peaks, and 49% (21,897/44,481) of stage 10 peaks were maintained until stage 13 (Table S1). Furthermore, we analyzed the genomic distributions of newly emerged accessible regions at each stage and found a continuous increase of peaks at intergenic regions from stage 8 to 13 (Fig. 2d), suggesting a stronger acceleration of chromatin opening at distal regulatory regions during early development.

Then, we compared chromatin accessibility profiles generated by CANTAC-seq with published ATAC-seq data²⁵ as well as ChIP-seq data of H3K4me3 and H3K27me3 at stage 9²⁶. As shown in Fig. S2b, our CANTAC-seq captured many more open chromatin regions compared to published ATAC-seq data. Importantly, the open regions identified by CANTAC-seq, regardless of whether they were overlapped and non-overlapped with those detected by ATAC-seq, were marked with the active mark H3K4me3, but not with the repressive mark H3K27me3 (Fig. S2b).

Furthermore, we checked the changes of chromatin accessibility at cis-regulatory elements for some key developmental regulators, including miR-427 cluster and sox17 gene cluster (Fig. 2e). MiR-427, which represents the ortholog of zebrafish miR-430, was involved in the deadenylation and clearance of maternal mRNAs, and was highly expressed before the MBT and then switched off at the gastrula stage²⁷. In line with such expression pattern, we found that the chromatin region (chr3:146,260,000-146,294,000) of miR-427 cluster was already open before the MBT (stage 7, 8 and 9), but turned closed afterwards (stage 10 and 13). In contrast, we found that chromatin accessibility at sox17 cluster (chr6:115,150,000-115,191,000), including sox17a, sox17b.1, and sox17b.2, was gradually established, which is consistent with the expression and function of sox17 genes in germ layer differentiation²⁸.

Finally, to reveal the potential function of increased accessible regions, we carried out gene ontology (GO) analysis on genes that gain promoter accessibility at each stage. As shown in Fig. S2c,d, GO terms related to transcriptional regulation and GTPase activity as well as pathways related to P53 regulation and GTPase cycle were significantly enriched at stage 8, 9 and 10, which is consistent with the fact that these biological processes and pathways are important in regulating cell migration, adhesion, as well as cell cycles during early development. Interestingly, genes with gained promoter accessibility specifically at stage 13 were enriched for those functioning in the neuronal system and encoding stimuli-sensing channels, consistent with the differentiation of nervous tissue at this stage. These findings highlighted the important functional relevance of the regulation on chromatin landscape during X. tropicalis early development.

Taken together, these analyses demonstrated that CANTAC-seq is a highly efficient method for generating accessible chromatin profiles of X. tropicalis embryos during early development.

Proximal and distal accessible chromatin in X. tropicalis early development

To understand the association between chromatin accessibility and gene expression during ZGA, we conducted transcriptome analyses at the corresponding developmental stages (Table S2). Again, RNA-seq analyses performed on two independent biological replicates for each stage showed a high degree of correlation (Pearson’s correlation coefficient, 0.96-1.00, Fig. S3a). We then performed principal component analyses (PCA) on the RNA-seq and CANTAC-seq datasets, which revealed a consistent pattern of developmental progression in both gene expression and chromatin accessibility profiles, as depicted in Fig. S3b. Then, based on the promoter accessibility at each stage, we classified genes into three groups (i.e., low, medium, and high) and compared their expression levels. As illustrated in Fig. 3a, genes with more accessible promoters showed elevated expression levels at all stages examined. After the MBT, genes with high promoter accessibility showed higher expression level than before (stage 7 and 8), and the difference in the expression level between the different groups is becoming more obvious, which is consistent with a higher impact of transcriptional regulation on RNA abundance after ZGA. To further explore the temporal relationship between chromatin opening and transcription activity, we compared promoter accessibility of genes expressed only after the MBT along the developmental stages and found that promoters are accessible prior to the onset of gene transcription (Fig. 3b). Moreover, we analyzed publicly available nascent RNA-seq data generated in X. laevis²⁹, and identified 631 homologous zygotic transcripts that are induced at the MBT in X. tropicalis (FC > 2, FDR < 0.05). As shown in Fig. S3c, promoters of these genes are accessible prior to the MBT. In addition, we also identified 78 homologous genes in X. tropicalis based on the list of well-known ZGA genes highly induced during MBT²⁹. Again, promoters of these genes are accessible prior to the MBT. Taken together, these results suggested that chromatin opening precedes the transcription of these zygotic genes.

As promoters and enhancers are often binding sites of transcription factors (TFs), we investigated whether proximal and distal CANTAC-seq peaks contained the motifs of TFs that regulate ZGA in early development. Using HOMER, we observed that TFs with enriched binding motifs for both promoters and enhancers mainly belong to the SP, KLF, and NFY protein families (Fig. 3c,d). In addition, whereas proximal regions exhibit specific enrichment of USF, ETV, and E2F motifs (Fig. 3c), distal regions are highly enriched with SOX, POU, CTCF, LHX, TFAP, and SMAD motifs (Fig. 3d). Importantly, several of the TFs identified in our motif enrichment analysis have been shown to be crucial in early development^23,30–35. For instance, the maternal factor Sox3 with pioneering TF activity establishes pluripotency before the MBT by triggering chromatin remodeling²³, while Ctcf is required for the de novo establishment of topologically associating domains during X. tropicalis ZGA³⁵.

E2f1 is a repressor of minor zygotic genome activation

The enrichment of E2F binding motifs at promoters drew our attention, given that no function had been reported for E2F proteins during ZGA. In mammals, E2F family consists of eight members with similar structure domains, of which E2F1/2/3 are classified as transcription activators, while E2F4/5/6/7/8 are considered to be transcription repressors³⁶. All members of e2f genes, except for e2f2, were identified in X. tropicalis. During early development from stage 3 to stage 9, e2f1, e2f3 and e2f5 exhibit high expression levels, followed by a stark decrease between stage 9 and stage 10 (Fig. S4a), suggesting that these three members may play an important role in X. tropicalis ZGA.

To assess the role of E2fs in X. tropicalis early development, we designed morpholino antisense oligonucleotides (MOs) targeting E2f1 (E2f1 MO), E2f3 (E2f3 MO), and E2f5 (E2f5 MO), respectively. These MOs were separately microinjected at the one-cell stage to knockdown their respective targets. We then quantified genomic DNA content in both uninjected wild-type (WT) and MO-injected embryos to assess the progression of the cell division during early developmental stages. As seen in Fig. 4a, we observed an increase in total genomic DNA content at stage 7 and 8 prior to MBT only in E2f1 MO-treated embryos, but not in E2f3 MO or E2f5 MO-treated embryos (Fig. S4b). To delve deeper into the defect induced by E2f1 MO, we made time-lapse videos featuring WT and E2f1 MO embryos (Video S1). In comparison to WT embryos, the depletion of E2f1 led to accelerated synchronous divisions prior to the MBT (Fig. 4b, Video S1). We could observe more cells in E2f1 MO embryos than the WT embryos starting from stage 6. After MBT, cell divisions transitioned to an asynchronous pattern for both E2f1 MO and WT embryos (Fig. S4c), and there was no significant difference in DNA content between E2f1 MO and WT embryos at stage 9 (Fig. 4a). Furthermore, we noted delayed development and mortality by the tailbud stage in the E2f1 MO-treated embryos (Fig. 4c). To determine whether these outcomes resulted from E2f1 perturbation, we performed rescue experiments through co-injection of E2f1 MO and E2f1 mRNA containing mutated MO target sites. Interestingly, we found that both abnormal development and the increase of genomic DNA content could be partially rescued (Fig. 4b,c), suggesting these abnormalities are direct and specific effects of the E2f1 perturbation. Hereafter, we focused on E2f1 and elaborated on its role during X. tropicalis early development.

Then, we conducted CANTAC-seq to analyze chromatin accessibility in E2f1 MO-treated embryos at stages 7 and 8, and compared them to the uninjected WT embryos. As shown in Fig. 4d, the chromatin of E2f1 MO embryos displayed higher accessibility than that of WT embryos at stage 7, indicating that E2f1 has a repressive effect on the chromatin opening. Subsequently, we investigated how increased accessibility was accompanied by transcriptome alterations. To this end, we performed comparisons of transcriptomic profiles between E2f1 MO and WT embryos from stage 3 to stage 8 using RNA-seq. As shown in Fig. 4e and Fig. S4d, the gene expression profiles were almost identical between E2f1 MO and WT embryos before stage 7. In contrast, significant changes in gene expression were observed in E2f1 MO embryos starting from stage 7, with 743 and 610 genes up-regulated and 534 and 633 genes down-regulated at stage 7 and 8, respectively (FDR < 0.05 and |Log₂ FC|>1, Table S3). GO terms related to gene transcription and embryo development were enriched for up-regulated genes, and GO term related to translation was enriched for down-regulated genes at stage 7 (Fig. S4e). Importantly, co-injection of E2f1 mRNA and MO could significantly rescue the gene dysregulation observed in E2f1 MO stage 7 samples (Fig. 4f and Fig. S4f, Table S4). Indeed, as shown in the PCA plot, the transcriptome profiles of WT X. tropicalis from different stages aligned along the PC1 axis according to the developmental progress, E2f1 MO-treated embryos at stage 7 were closer to WT embryos at stage 8 at PC1 axis, whereas the rescued embryos moved back towards the WT stage 7 (Fig. 4f).

In order to test whether the genes affected by E2f1 MO were relevant for the early development, we compared the genes altered upon E2f1 MO at stage 7 with differentially expressed genes during normal development between stage 7 and 8, and found a significant overlap (p < 2.2e-16, Fig. 4h). Out of the 699 ZGA genes that are normally induced from stage 7 to stage 8, 268 (38%) genes are significantly upregulated upon E2f1 MO at stage 7 (Fig. 4f). These genes were enriched for those with GO terms related to transcription, cell differentiation, morphogenesis, gastrulation, and BMP signaling (Fig. 4f). Notably, E2f1 MO caused premature expression of a set of well-known zygotic TFs, including gsc, bix1.2, sox17a/b.1/b.2 as well as pou5f3.1. In consistence with this, we observed significant enrichment of several TF motifs in the open chromatin regions from E2f1-MO treated embryos at stage 7 (Fig. S4g), as well as from WT embryos at stage 8 (Fig. 3c,d), suggesting TFs repressed by E2f1 contribute to the chromatin opening during minor ZGA. Taken together, our analysis suggested that E2f1 functions as a maternal transcriptional repressor in X. tropicalis early development.

E2f1 represses zygotic gene transcription independent of its regulation of cell cycle progression

In a previous study, Jukam et al. reported that the DNA-to-cytoplasmic ratio regulates the timing of zygotic gene expression in hybrid frog embryos³⁷. To investigate whether the precocious initiation of zygotic gene transcription is a result of increased DNA content in E2f1 morphants with accelerated cell cycles, we decelerated cell division by introducing mRNA encoding the cyclin-dependent kinase inhibitor cdkn1a (Fig. S5a), which was known to regulate cell cycle and DNA replication. As evidenced by the genomic DNA content in Fig. 5a, upon injecting Cdkn1a mRNA, embryonic development was delayed for one cell cycle at stage 7 and 8 in WT embryos. Co-injection of Cdkn1a mRNA with E2f1 MO successfully rescued the accelerated cell division due to E2f1 MO alone (Fig. 5a and Fig. S5b). Intriguingly, the precocious expression of zygotic genes at stage 7 induced by E2f1 MO was not rescued with the co-injection of Cdkn1a mRNA (Fig. 5b). The observation that co-injection of Cdkn1a mRNA could rescue accelerated cell division, but not the precocious expression of zygotic genes induced by E2f1 MO demonstrated that the two effects of E2f1 were mechanistically independent. Given that E2f1 could directly regulate transcription by binding to target DNA via its well-characterized DNA binding domain (DBD, a.a. 119–184), to ascertain its direct repressive effect of zygotic gene activation, we performed rescue experiments using an E2f1 mutant lacking the DBD (∆DBD, Fig. S5c). As shown in Fig. 5c and Fig. S5d, same as the WT E2f1, the ∆DBD mutant successfully rescued the cell cycle acceleration defect. However, in contrast to WT E2f1, the mutant failed to rescue the precocious zygotic gene activation induced by E2f1 MO (Fig. 5b). As representative examples shown in Fig. 5d, zygotic genes, including sia1/2, mix1 and gsc were activated at stage 7 by E2f1 MO. The defect could be rescued upon co-injection of WT E2f1 mRNA, but not that of ∆DBD mutant. Similarly, co-injection of Cdkn1a mRNA could not rescue the precocious expression of these zygotic genes although it could normalize the cell division prior to MBT. These findings collectively suggest that the impact of E2f1 on these zygotic genes operates through transcriptional regulation rather than as a consequence of cell cycle acceleration.

Otx1 antagonizes the inhibitory effect of E2f1 during minor ZGA

To understand the molecular mechanisms underlying the E2f1-mediated repression of ZGA, we sought to identify E2f1 interaction partners. Due to the lack of specific antibodies that can recognize endogenous E2f1, we performed immunoprecipitation followed by mass spectrometry (IP-MS) analysis with the anti-HA antibody in stage 8 embryos injected with E2f1 MO and HA-tagged E2f1 mRNA (Fig. S6a). With a stringent cutoff (unique peptide > 2, coverage > 5% and Log₁₀ FC (LFQ) > 1), we identified a total of 22 potential interaction partners with E2f1 (Table 1), including the DP family protein Tfdpf1, which is known to form a heterodimeric complex with E2f1. Out of the remaining candidates, we focused on the orthodenticle homeobox transcription factor Otx1, as it was reported to be required for endoderm formation in X. tropicalis³⁸, and its paralogue OTX2 was recently suggested to be a potential maternal regulator of human ZGA³⁹. During X. tropicalis early development, otx1 is highly expressed while otx2 only starts to express after the MBT (Fig. S6b). To validate the interaction between E2f1 and Otx1, we co-injected HA-tagged E2f1 and Myc-tagged Otx1 mRNAs at one cell stage of X. tropicalis embryos (Fig. S6c). The interaction between E2f1 and Otx1 could be confirmed in Co-IP assays followed by western blotting with anti-HA and anti-Myc antibodies, respectively, at the blastula stage (Fig. 6a). Notably, this interaction was significantly reduced after treatment with benzonase, suggesting that the interaction between E2f1 and Otx1 is mediated largely by DNA.

Table 1

**E2f1 interaction partners were identified using immunoprecipitation followed by mass spectrometry (IP-MS).** A total of 22 proteins were identified in two independent biological replicates, with a minimum number of three peptides and 5% coverage, as well as the average Log₁₀ fold change (FC) of label-free quantity (LFQ) > 1.
Gene	Unique peptides		Coverage		Log₁₀ FC (LFQ)	Description
Gene	rep1	rep2	rep1	rep2	Log₁₀ FC (LFQ)	Description
Stk31	48	49	51	51.2	9.14	Serine/threonine kinase 31
Otx1	7	7	26.2	26.2	7.89	Orthodenticle homeobox 1
Dnaaf2	12	12	21.1	21.1	7.20	Dynein axonemal assembly factor 2
Acly	12	13	14	15.1	7.03	ATP citrate lyase
Tfdp1	5	5	20.7	20.7	6.80	Transcription factor Dp-1
Arrdc1	5	5	14.6	14.6	6.75	Arrestin domain containing 1
Park7	3	3	17.5	17.5	6.44	Parkinson protein 7
Acad9	6	6	12.2	12.2	6.42	Acyl-CoA dehydrogenase family member 9
Ywhae	6	3	30	14.2	6.34	Tyrosine 3/5-monooxygenase activation protein epsilon
Ddx6	4	4	10.2	10.2	6.33	DEAD-box helicase 6
Dbt	5	4	9.8	7.9	6.30	Dihydrolipoamide branched chain transacylase E2
Psmd2	4	4	7	7	6.23	Proteasome 26S subunit ubiquitin receptor, non-ATPase 2
Aldh7a1	4	4	12.1	12.1	6.22	Aldehyde dehydrogenase 7 family member A1
Ndufs3	4	4	20.8	20.8	6.17	NADH:ubiquinone oxidoreductase core subunit S3
Pspc1	3	3	9.4	9.4	6.08	Paraspeckle component 1
Ak2	6	3	30.3	14.1	6.07	Adenylate kinase 2
Cndp2	3	3	6.5	6.5	6.06	CNDP dipeptidase 2
Vdac3	3	3	13.1	13.1	6.03	Voltage-dependent anion channel 3
Ube2v2	4	3	25.9	19.7	5.98	Ubiquitin conjugating enzyme E2 V2
Acat1	3	3	9	9	5.95	Acetyl-CoA acetyltransferase 1
Amt	4	3	11.9	8.9	5.84	Aminomethyltransferase
Cct4	9	10	21.6	24.5	1.07	Chaperonin containing TCP1 subunit 4

To gain insight into the functional role of Otx1 during early development and, more importantly, to assess how Otx1 and E2f1 function together to regulate X. tropicalis ZGA, we blocked their translation by injecting MOs against E2f1 and Otx1 into embryos either alone or together. As shown in Fig. 6b, Otx1 MO-treated embryos were viable, whereas E2f1&Otx1 double MO-treated embryos had even more severe developmental defects than E2f1 MO-treated embryos, and all the double MO embryos died at the early tailbud stage. Next, to figure out the underlying mechanisms, we examined transcriptomic changes and chromatin accessibility alterations of Otx1 MO-treated embryos as well as E2f1&Otx1 double MO-treated embryos at stage 7 and 8 using RNA-seq and CANTAC-seq, respectively. Knockdown of Otx1 alone led to 520 and 718 genes to be up-regulated and 634 and 897 down-regulated at stage 7 and 8, respectively (FDR < 0.05 and |Log₂ FC|>1, Fig. S6d, Table S5). Then, we examined the developmental relevance of these genes and found a significant overlap (p < 2.2e-16, Fig. S6e) between the genes down-regulated upon Otx1 MO at stage 8 and the genes up-regulated between stage 7 and 8 during normal development. These overlapped genes, including several well-known zygotic TFs, were significantly enriched for the GO term related to transcription regulation (Fig. S6e). In line with this, a slight attenuation of global chromatin accessibility upon Otx1 MO was also observed at stage 8 (Fig. S6f). Interestingly, double knockdown of E2f1 and Otx1 did not affect a much larger group of genes (Fig. S6g, Table S6). The expression of up-regulated genes upon E2f1 MO was largely returned to WT level by double MO at both stage 7 and 8 (Fig. 6c). The prematurely opened chromatin induced by E2f1 MO at stage 7 was also partially rescued by the double MO (Fig. 6d). In contrast, the expression change induced by Otx1 MO remained the same in double MO (Fig. S6h). Further PCA analysis of RNA-seq and CANTAC-seq data showed that the transcriptome and the chromatin accessibility profiles of double MO embryos were indeed moving back and even lagged a bit behind the WT embryos along the developmental trajectory (Fig. 6e,f). The difference at the transcriptome level between WT embryos and double MO embryos at the PC1 axis was likely due to the effects of Otx1 independent of E2f1 (Fig. 6e). These results together suggested that Otx1 antagonizes the inhibitory effect of E2f1 during X. tropicalis minor ZGA.

Furthermore, to investigate the effect of different perturbations on chromatin accessibility of the 699 genes induced from stage 7 to stage 8, we analyzed the percent of these genes that gain, lose, or do not change accessibility upon E2f1 MO, Otx1 MO or double MO (Fig. 6g). At stage 7, 57% of these genes gain accessibility at genic regions upon E2f1 MO, with 17% genes gaining accessibility at promoter regions. These together demonstrated that E2f1 exerts a dominantly repressive effect on chromatin opening at stage 7. In contrast, 95% of these genes show no change of accessibility upon Otx1 MO at genic regions at stage 7. Similarly, these genes show no change of accessibility upon double MO, demonstrating that the chromatin opening induced by E2f1 MO requires Otx1 at stage 7. At stage 8, 49% and 43% genes gain and lose accessibility, respectively, at genic regions upon E2f1 MO, with 8% and 12% genes gaining and losing accessibility at promoter regions, respectively. In contrast, 63% and 64% genes loss accessibility at genic regions upon Otx1 MO or double MO, with 32% and 33% losing accessibility at promoter regions. This analysis together suggested that Otx1 exerts a dominantly positive effect on chromatin opening at stage 8.

E2f1 and Otx1 co-modulate zygotic gene transcription by a seesaw balance

Both E2f1 and Otx1 are known DNA binding transcription factors, we therefore asked how they bound to and localized on the genome before the MBT. Again, without a specific antibody, we performed chromatin immunoprecipitation followed by sequencing (ChIP-seq) of E2f1 and Otx1 with anti-HA and anti-Myc antibodies in stage 8 embryos injected with E2f1 MO and HA-tagged E2f1 mRNA or Otx1 MO and Myc-tagged Otx1 mRNA, respectively (Fig. 7a). In total, we identified 22,524 binding sites for E2f1, of which 30%, 42%, and 28% were located at the promoter, gene body, and intergenic region, respectively (Fig. 7b). For Otx1, we obtained 65,888 binding sites, with 6%, 52%, and 42% located at the promoter, gene body, and intergenic region, respectively (Fig. 7b). We observed a significant overlap of the DNA binding sites between E2f1 and Otx1, with almost half of E2f1 peaks overlapped with Otx1 (p < 2.2e-16, Fig. S7a), suggesting a genome-wide co-occupancy of these two transcription factors. We then categorized these binding regions into three subgroups, namely E2f1 only (regions exclusively bound by E2f1, n = 12,031), E2f1&Otx1 (regions co-bound by E2f1 and Otx1, n = 10,493), and Otx1 only (regions exclusively bound by Otx1, n = 55,395). Among the common binding sites, 23% and 32% were located at promoters and intergenic regions, respectively (Fig. S7b).

To further characterize the E2f1 and Otx1 binding sites, we performed sequence motif analysis in these three categories separately. As expected, E2F and OTX motifs were exclusively enriched in E2f1 only or Otx1 only regions, respectively, while E2f1&Otx1 regions were enriched for both E2F and OTX motifs (Fig. S7c). Moreover, FOXH, POU, SOX and bHLH motifs were commonly enriched in all three subgroups. In addition, we observed that motifs such as KLF and ETS tend to appear in E2f1 binding regions, while motifs like GSC and T-box motifs were enriched in Otx1 binding regions. Since the chromatin accessibility was gradually established with the most dramatic increase at stage 8, we then asked whether E2f1 and Otx1 binding was associated with chromatin opening. As shown in Fig. 7c, E2f1 bound regions appeared to be more accessible than Otx1 bound regions during early development.

To further check how the co-binding of E2f1 and Otx1 regulates the gene expression during minor ZGA, we next defined their common target genes which are also developmentally relevant by the following criteria: (1) up-regulated in WT embryos from stage 7 to stage 8, (2) up-regulated upon E2f1 MO at stage 7 and (3) rescued back to the WT level by double MO at stage7. A total of 220 genes fitting the three criteria were identified (Table S7). Compared to genes that were up-regulated between stage 7 and 8, but not affected by E2f1 and/or Otx1 perturbation, we observed a much higher enrichment of E2f1 and Otx1 binding at these 220 co-regulated target genes (Fig. S7d) with 38.2% and 20.5% of gene promoters harboring binding sites for E2f1 and Otx1, respectively (Fig. 7d). Among these, 11.8% of gene promoters contained both E2f1 and Otx1 binding sites. Interestingly, the expression of these 220 genes was largely inhibited in Otx1 MO-treated at stage 8, suggesting that Otx1 is the transcription activator of these genes also at stage 8 of WT embryos.

Finally, to further explore the 220 genes co-regulated by E2f1 and Otx1, we constructed an interaction gene network based on protein-protein interactions using the STRING database⁴⁰. As shown in Fig. S7e, the resulting network comprises a higher number of transcriptional regulators and components of signaling cascades. This is also reflected by enriched GO terms related to transcription regulation, embryo development and cell differentiation as well as TGF-beta and WNT signaling pathways. As representative examples shown in Fig. 7e, both E2f1 and Otx1 bound at proximal and/or distal regulatory regions of germ layer differentiation-related genes such as sia1/2, mixer, mix1 and gsc, and their binding sites were highly overlapping. Consistent with their binding pattern, the up-regulation of these genes upon E2f1 MO at stage 7 was rescued by E2f1&Otx1 double MO at stage 7, and Otx1 MO inhibits the expression of these genes at stage 8, indicating that E2f1 and Otx1 directly bind to and modulate the transcription of their targets in opposite directions.

Taken together, our findings suggested a seesaw model via which a coupled transcriptional repressor-activator controls the transcription of a subset of zygotic genes before the MBT in X. tropicalis early development (Fig. 8). For WT embryos, at stage 7 (early blastula), E2f1 plays a dominant role in maintaining an inactive chromatin state and repressing the zygotic gene transcription. As the embryo develops, the repression by E2f1 gradually fades, and the chromatin accessibility is remodeled from being closed to being opened. At stage 8 (middle blastula), the transcriptional activator Otx1 takes the leading role and starts to activate the transcription of zygotic genes important for zygotic genome activation and germ layer differentiation. Upon E2f1 MO, the chromatin opens earlier (at stage 7) and zygotic genes are activated by Otx1, which results in a premature ZGA. Upon Otx1 MO or E2f1&Otx1 double MO, the chromatin remains less open than WT and the transcription of those zygotic genes is blocked at stage 8, leading to a delayed ZGA.

Here, by developing CANTAC-seq, we systematically analyzed the dynamics of chromatin accessibility during X. tropicalis early development. We found that the chromatin accessibility was progressively established with the most dramatic increase before the MBT, and chromatin opening at cis-regulatory regions preceded zygotic gene transcription. Moreover, through analyzing these data and genetic perturbation experiments, we identified a coupled pair of transcriptional repressor and activator, E2f1 and Otx1, and the dynamic balance between the two factors determines the temporal expression pattern of their target genes during minor ZGA.

X. tropicalis has been widely used as a model organism in developmental biology, especially in studying the process of ZGA. However, high yolk content in X. tropicalis early embryos interfered with the application of existing methods (e.g., ATAC-seq, Dnase-seq) for analyzing chromatin accessibility^23,24. To overcome this issue, we established the CANTAC-seq, which is a sensitive tool for measuring the open chromatin of X. tropicalis embryos. Indeed, we observed a drastic opening of chromatins before the MBT at stage 8, followed by the gradual increase of accessibility at late stages. Moreover, the promoter accessibility of zygotic genes expressed after the MBT was already established before the MBT. A previous study based on the ATAC-seq to analyze chromatin accessibility demonstrated that regulatory elements became only accessible after the MBT in X. tropicalis (Bright et al., 2021). The inconsistency between the two studies is likely due to the superior capability of our CANTAC-seq in measuring chromatin accessibility of frog embryos at very early developmental stages. Furthermore, our CANTAC-seq approach could also be applied to low-input cell numbers with no need for centrifugation. By switching ConA to antibodies recognizing specific surface markers, this approach can be easily adapted to study chromatin profiles of specific cell types without the need for prior enrichment.

Bioinformatics analysis of chromatin landscape established during early development revealed a set of maternal TFs whose DNA binding motifs were enriched at proximal or distal cis-regulatory elements with gradually gained accessibility, including several several TFs shown to be crucial in early development. Focusing on E2F family, whose function during ZGA had not been investigated, we demonstrated that E2f1 knockdown not only sped up cell division before the MBT but also led to the precocious transcription of a subset of zygotic genes. To explore the relationship between these two phenomena, we suppressed cell division processes in E2f1 MO embryos by co-injecting Cdkn1a mRNA and performed rescue experiment using a mutant form of E2f1 lacking the DBD. It turned out that the repressive effect of E2f1 on zygotic gene transcription is independent of its role in cell division regulation.

During the normal cell cycle with gap phase, it is well-known that E2f1 regulates the G1/S transition via driving the expression of its targets necessary for S phase entry⁴¹. However, in the early development of Xenopus embryos, prior to the MBT, cell cycle progressed rapidly without gap phase and there is absence of transcriptional activity. Therefore, it is unclear how E2f1 regulates the cell cycle in this context. In this study, we focused on the transcriptional regulation of E2f1. To dissect the mechanism underlying the E2f1 mediated cell cycle regulation, the identification of protein interaction partners might provide a hint. Unfortunately, the list of proteins identified in our IP/MS experiment did not contain any one with obvious function related to cell cycle regulation. In the future study, the improved experimental design might improve the chance to identify the relevant interaction partners of E2f1 in regulating cell cycle: 1) instead of pulldown overexpressed tagged E2f1, use antibody against endogenous E2f1; 2) instead of stage 8, perform IP-MS analysis at earlier stages.

By manipulating the DNA content or mechanically separating the cytoplasm, previous works have suggested a classical nucleocytoplasmic (N/C) ratio model positing that changes in the N/C ratio trigger the MBT, which was defined by the near coincident onset of large-scale zygotic gene transcription and cell-cycle lengthening during early vertebrate development. It has been shown that an increase in DNA content or a decrease in cytoplasmic volume caused the MBT to occur earlier, while a decrease in DNA content delayed the MBT^5,37,42-44. However, the classic N/C ratio model falls short in elucidating the initiation of minor ZGA as these zygotic transcription initiates prior to cell-cycle lengthening and it remains unclear whether the timing of these early transcripts depends on an N/C ratio^8,29,45,46. Our study discovered that in embryos undergoing rapid cleavage prior to MBT, the precocious transcription of a subset of zygotic genes upon E2f1 MO does not depend on the accelerated cell division. Therefore, our findings underscore a previously unrecognized direct repressive function of the maternal factor E2f1 in modulating the onset of the transcription of a subset of zygotic genes during minor ZGA. Indeed, our data indicated that the cell cycle progression prior to MBT and minor ZGA could be uncoupled. On one hand, embryos with ∆DBD mutant or Cdkn1a mRNA co-injected with E2f1 MO divided “normally” prior to MBT, but started ZGA at stage 7 instead of stage 8. On the other hand, our observation that Otx1 MO at stage 8 failed to activate a large set of zygotic genes demonstrated that normal cell cycle progression prior to MBT is not sufficient for ZGA. The seesaw model proposed in our manuscript for E2f1 and Otx1 could serve a new paradigm for searching more TF pairs with antagonistic function, further our mechanistic understanding of ZGA regulation.

In mammals, E2f1 is a well-known transcriptional activator that regulates the expression of a number of genes important for the G1/S transition⁴⁷. A previous study using a tissue-specific E2f1 knockout mouse model has reported that E2f1-3 could form in complex with the repressor Rb to silence their targets in differentiating cells of the small intestine⁴⁸. However, Rb expression in early developmental stages is extremely low⁴⁹ and loss of Rb has no impact on cell cycling or differentiation of early X. laevis embryos⁵⁰. Indeed, we did not identify any known transcriptional repressors in our IP-MS analysis. How E2f1 inhibits chromatin opening and zygotic gene transcription during the early development of frog embryo await future elucidation.

Disruption of E2f1 and/or Otx1 led to alterations in both gene expression and chromatin accessibility. As the expression of many transcription factors was dysregulated upon E2f1 and/or Otx1 perturbation, the observed chromatin changes could be both direct and indirect effects of their perturbation. In line with a previous publication suggesting a pioneer activity of Otx1 in the establishment of functional enhancers and the specification of endoderm in the gastrula stage³⁸, we also discovered that Otx1 played an important role in activating zygotic gene transcription and chromatin states at an earlier developmental time point in our study. More importantly, we demonstrated that Otx1 could antagonize the inhibitory effect of E2f1 during minor ZGA. Via ectopic expression of tagged-protein, the extensive co-localization of E2f1 and Otx1 on the genome was observed using ChIP-seq, and the association between the two proteins was likely mediated by their independent DNA binding, as the effective Co-IP between the two was largely attenuated after DNA digestion. Importantly, it should be noted that the E2f1-Otx1 axis was not the only point via which E2f1-repressed genes get activated during ZGA, as not all E2f1 target genes got activated by Otx1 and half of E2f1-bound regions were not co-localized with Otx1. Indeed, we observed several TF motifs in regions bound by E2f1 but not Otx1, indicating that E2f1 may also pair with other TFs to modulate chromatin opening and zygotic gene transcription. The same might also hold true for Otx1, i.e., it could exert its transcriptional regulation also independent of E2f1. On the other hand, as shown above for E2f1, in addition to direct transcriptional regulation Otx1 may also play roles in other biological processes during early development.

In our seesaw model, we proposed that the titration of E2f1 and the accumulation of Otx1 with development determined the timing of their target gene activation. To further validate this model, one needs to measure the temporal protein abundance, ideally the DNA binding profile of both factors using specific antibodies against the endogenous E2f1 and Otx1, which we unfortunately could not generate after several failed attempts. Furthermore, it would be also interesting to explore how the abundance of E2f1 and Otx1 themselves are regulated during ZGA. Deciphering both upstream and downstream factors of E2f1/Otx1 regulatory network will largely facilitate our mechanistic understanding of ZGA during early development.

Finally, our findings could be evolutionally conserved. In X. laevis, e2f1 and otx1 transcripts have been shown to be enriched and co-localized in pre-MBT embryos⁵¹, and both factors are translated during ZGA^52,53, suggesting that E2f1 and Otx1 may regulate X. laevis ZGA in the same manner. More recently, OTX2, the paralogue of OTX1, was reported to be highly expressed upon oocyte meiotic resumption and began to decline after major ZGA in human, suggesting its functional role in human ZGA regulation³⁹. In the same study, several E2F factors were found also to be transcribed and translated during human ZGA (Table S8). It is therefore plausible that a similar regulatory mode might also exist in human, but may utilize different OTX and E2F family members.

X. tropicalis manipulations

X. tropicalis frogs were purchased from NASCO (USA) and housed in a room with a constant temperature at 25°C following a 12-hour light-dark cycle. Embryos were obtained at different developmental stages by artificial fertilization and cultured in 0.1× MBS medium (1xMBS: 80 mM NaCl, 10 mM HEPES, 2.4 mM NaHCO₃, 1 mM KCl, 0.82 mM MgSO₄, 0.33 mM Ca(NO₃)₂, 0.41 mM CaCl₂, pH 7.4) at 25 °C. The developmental stages were determined according to Nieuwkoop and Faber ⁵⁴. Briefly, embryos at developmental stages 3, 4, 5, 6, 7, 8, 9, 10 and 13 were collected at 1.5, 2, 2.5, 3, 3.5, 4, 5.5, 7 and 12 hours post fertilization (hpf), respectively. All animal experiments were carried out following the animal protocols approved by the Laboratory Animal Welfare and Ethics Committee of the Southern University of Science and Technology.

Plasmid construction and microinjection

Morpholino antisense oligonucleotides (MOs) targeting E2f1 (5’ GTGTTCATCTTTGCTTCAAGAGTTC 3’), E2f3 (5’ CCCTTTCTCATCTTCCTGACTAG 3’), E2f5 (5’ TACGGGCCGAGCAGAGCTACAGTC 3’) and Otx1 (5’ CATGCTCAAGGCTGGACAGAAACCC 3’) were obtained from Gene Tools. The open reading frames of X. tropicalis E2f1 (XM_012952820), Cdkn1a (XM_002935778) and Otx1 (NM_203885) containing mutated MO target sites were cloned into the pCS2+ vector. The E2f1 ∆DBD mutant was sub-cloned via depleting the sequence coding a.a. 119-184. The capped mRNAs were in vitro transcribed using the mMESSAGE mMACHINE SP6 Transcription Kit (Invitrogen, AM1340). Afterwards, the mRNAs were purified with the RNeasy Mini Kit (QIAGEN, NC9677589) and quantified by NanoDrop (Thermo).

For microinjection, MOs were injected into 1-cell stage embryos from the animal pole with a dose of 10ng per embryo using a pneumatic Pico Pump PV830 (WPI). The rescue experiments were carried out to confirm the specificity of MOs by co-injection of 300pg mRNA and 10ng corresponding MOs. 40pg Cdkn1a mRNA was injected into 1 cell stage embryos to inhibit cell division. Injected embryos were then collected for further experiments at desired time points. Images and videos of the whole embryos were acquired using a stereo microscope (SMZ18, Nikon).

Cell culture

K562 cells (ATCC, CCL-243) were cultured in RPMI 1640 medium supplemented with 10% FBS and 1% penicillin/streptomycin. MES cells were obtained from Dr. Qi Zhou and cultured in serum-free media. Briefly, 500mL of N2B27 media was generated by including: 240mL DMEM/F12 (Invitrogen, 11330-032), 240mL Neurobasal (Invitrogen, 21103049), 5mL N2 medium (Invitrogen, 17502048), 10mL B27 medium (Invitrogen, 17504044), 1% GlutaMAX (Life Technologies, 35050-061), 1% nonessential amino acids (Life Technologies, 1140-035), 0.1mM β-mercaptoethanol (Life Technologies, 21985-023),1% penicillin-streptomycin (Life Technologies, 15140-122), 1uM PD0325901 (Stemgent, 04-0006), 3uM CHIR99021 (Stemgent, 04-0004), 5ul mLIF (Millipore, LIF2050).

CANTAC-seq

In order to make tagmentation working on early embryos of X. tropicalis, we developed the CANTAC-seq method in which cells/nuclei were firstly captured and purified using ConA beads, followed by Tn5 transposition and library preparation. Briefly, X. tropicalis embryos were collected in a 1.5mL tube at blastula stages (stage 7, stage 8 and stage 9), at the onset of gastrulation (stage 10) and the onset of neurulation (stage 13). Embryos were pipetted up and down in 200μL binding buffer (20mM HEPES pH 7.5, 150mM NaCl, 0.5mM Spermidine) with a P20 pipette tip until a homogenized lysate is formed. ConA magnetic beads (Bangs Laboratories, BP531) were pre-washed once with washing buffer (20mM HEPES pH 7.5, 10mM KCl, 1mM CaCl₂, 1mM MnCl₂) and 20 µL of ConA beads were added per sample and incubated at RT for 20 min. The unbound supernatant containing yolk, pigment and lipids was removed and cells/nuclei-beads complex were washed once with 100µL resuspension buffer (10mM Tris-HCl pH7.4, 10mM NaCl, 3mM MgCl₂). A small aliquot was taken for DAPI staining and nuclei counting under the microscope. Afterwards, approximately 50,000 beads-bound cells/nuclei were incubated in 100µL lysis buffer (10mM Tris-HCl pH7.4, 10mM NaCl, 3mM MgCl₂, 0.1%NP40, 0.1% Tween-20, 0.01% Digitonin) for 10 min on ice followed by washing with 600µL lysis buffer without NP40 and Digitonin. Afterwards, an on beads tagmentation was carried out by resuspending bead-bound cells/nucleus in 100µL transposition mix (Vazyme, TD501) and incubated at 37˚C for 30 min with gentle shaking at 800 rpm. The tagmentation was stopped by adding 3.3 μL 0.5M EDTA, 1μL 10% SDS and 1.5μL 20 mg/mL Proteinase K to each sample and incubated at 55 ˚C for 1 hr.

Tagmented DNA was cleaned up using the DNA Clean & Concentrator kit (Zymo, D4013) and DNA was eluted in nuclease free water. Libraries were amplified under the following PCR conditions: 72 °C for 3 min; 98 °C for 30 sec; and thermocycling at 98 °C for 15 sec, 60 °C for 30 sec and 72 °C for 40 sec; followedby 72 °C for 5 min and hold at 4 °C. Libraries were cleaned up with VAHTS DNA Clean Beads (Vazyme, N411), followed by two times washing with 80% ethanol and eluted in nuclease free water. The libraries were sequenced in a 2x150nt manner on NovaSeq 6000 platform (Illumina).

ATAC-seq

For normal ATAC-seq library construction, cells were prepared as previously described with minor modifications. Briefly, 50000 fresh cells were lysed in lysis buffer for 10 minutes on ice to prepare the nuclei. Immediately after lysis, nuclei were spun at 500g for 5 min to remove the supernatant. Nuclei were then incubated with the Tn5 transposase (Vazyme, TD501) in tagmentation buffer at 37°C for 30 min. After tagmentation, PCR was performed to amplify the library for 12 cycles under the following PCR conditions: 72°C for 3 min; 98°C for 30 sec; and thermocycling at 98°C for 15 sec, 60°C for 30 sec and 72°C for 40 sec; following by 72°C for 5 min. After the PCR reaction, libraries were purified with the DNA purification beads (Vazyme, N411). The libraries were sequenced in a 2x150nt manner on NovaSeq 6000 platform (Illumina).

RNA extraction and mRNA sequencing library preparation

Total RNA was extracted from X. tropicalis embryos from different developmental stages using TRIzol™ Reagent (Invitrogen, 15596026). mRNA sequencing libraries were prepared using the standard protocol provided by VAHTS® mRNA-seq V2 Library Prep Kit for Illumina (Vazyme, NR601) with 1μg total RNA. The libraries were sequenced in a 2x150nt manner on NovaSeq 6000 platform (Illumina).

Chromatin immunoprecipitation (ChIP)

The ChIP assay was performed according to the standard protocol provided by SimpleChIP Plus Sonication Chromatin IP Kit (CST, 56383) with minor modifications. Briefly, X. tropicalis embryos were fixed with 1.5% formaldehyde for 30 min. Sonication was then carried out at the Bioruptor pico (Diagenode) by applying 20 cycles of 30 sec ON and 30 sec OFF to obtain chromatin fragments of approximately 100-500 bp. The HA antibody (Sigma, H6908) and Myc-Tag antibody (CST, 2276) were used to pulldown HA-tagged E2f1 and Myc-tagged Otx1, respectively. ChIP DNA was cleaned up using the ChIP DNA Clean& Concentrator kit (Zymo, D5205). ChIP-seq libraries were prepared using the standard protocol provided by VAHTSTM Universal DNA Library Prep Kit for Illumina® V3 (Vazyme, ND607). The libraries were sequenced in a 2x150nt manner on NovaSeq 6000 platform (Illumina).

DNA quantification

To compare the total amount of DNA in wild-type and MO embryos at different stages during embryogenesis, 15 embryos per stage per condition were collected. Genomic DNA was isolated together with 100,000 K562 cells as spike-in control according to the manual of DNeasy Blood & Tissue Kit (Qiagen, 69504). For DNA quantification, real-time PCR were carried out using primer pairs that specifically target X. tropicalis ACTB genomic region (forward primer: 5’ AGGCCAGGACAGCCCTGTAA 3’, reverse primer: 5’ CCCAGAGGAACACCCAGTGC 3’) and human ACTB genomic region (forward primer: 5’ GCCTTGTCACACGAGCCAGT 3’, reverse primer: 5’ GAGCTGCGCCCTTTCTCACT 3’), respectively. The relative DNA quantity was calculated by using the spiked in human DNA as an internal control. All the measurements were performed in at least triplicates.

Co-immunoprecipitation (Co-IP) and western blotting (WB)

X. tropicalis embryos at indicated stages were collected and homogenized in lysis buffer (20 mM Tris-HCl pH 7.4, 150 nM NaCl, 1 mM EDTA, 1% Triton, 1 mM DTT, 0.1 mM PMSF, protease inhibitor cocktail (Roche, 04693132001), 1 mM NaVO4). Protein lysates were then mixed with 2 volumes of 1,1,2-Trichlorotrifluoroethane. After centrifugation, the protein lysates at the upper layer were collected to measure concentrations using the BCA assay (Beyotime, P0011). For immunoprecipitation, the protein lysates were incubated with protein A/G magnetic beads (MCE, HY-K0202) coupled with indicated antibodies for 4 hr at 4°C. The beads were then washed three times with TBST and eluted in loading buffer by heating at 98°C for 5 minutes. The eluted protein complex was separated in 10% SDS gels and blotted on PVDF membranes by semi-dry blotting. The Spectra Multicolor Broad Range Protein Ladder (Thermo, 26634) was loaded for size estimation. The membranes were blocked in 5 % skim milk powder/TBST for 1h at room temperature and then incubated with the primary antibody at 4˚C overnight. After washing in TBST for three times, the membrane was incubated with the secondary antibody for 1 hr at room temperature. After washing for three times, the membranes were developed with Pierce ECL (Thermo, 32106) according to the manufacturer's instructions. Protein bands were recorded with the ChemiDoc MP Imaging System.

Immunoprecipitation followed by mass spectrometry analysis (IP-MS)

X. tropicalis embryos were homogenized in lysis buffer containing protease and phosphatase inhibitors. The E2f1-associated proteins were immunoprecipitated with Protein A/G magnetic beads (MCE, HY-K0202) coupled with HA antibody (Sigma, H6908) for 4 hr. After washing with TBST, an overnight on-beads digestion was carried out with sequencing-grade trypsin (Promega, V5111). Afterwards, the samples were analyzed on an LTQ-Orbitrap Elite mass spectrometer system (Thermo). IP-MS data was processed by MaxQuant for label-free quantification with “match between run” function activated and database searching was against the UniProt X. tropicalis database.

CANTAC-seq and ATAC-seq analysis

For ATAC-seq analysis, fastp (v0.23.2) ⁵⁵ were used to trim the reads with parameters -a CTGTCTCTTATA --detect_adapter_for_pe --length_required 20 -q 30. Alignment of these reads to the X. tropicalis v10.0 reference genome was performed with Bowtie2 (version 2.4.5)⁵⁶ with parameters -X 2000. Sambamba (v0.7.0)⁵⁷ was used to remove the duplicated reads. Peak calling was performed by MACS2 (v2.2.7.1)⁵⁸ with the parameters -g 1.435e9 --keep-dup all -q 0.05 --slocal 10000 --nomodel --nolambda -B --SPMR. The reads were counted using featrueCounts (version 2.0.1)⁵⁹ and the counts were converted to counts per million (CPM) for plotting. Normalized read coverage tracks were generated with bamCoverage from the deeptools package(version 3.5.1)⁶⁰ and are visualized using the Integrative Genomics Viewer (IGV, version 2.16.0)⁶¹. For motif enrichment analysis, the default motif libraries including JSAPAR, DMMPMM, AthaMap and etc., were used by the HOMER⁶² tool findMotifs.pl. Peaks located within the region of -1kb ~ +0.1kb from the transcription start site (TSS) were defined as proximal peaks, while other CANTAC peaks are classified as distal ones.

RNA-seq analysis

For RNA-sequencing data, quality control and adapter trimming were performed using fastp (v0.23.2)⁵⁵ with parameters -a AGATCGGAAGAGC --detect_adapter_for_pe -w 12 --length_required 30 -q 20. Reads were mapped to the X. tropicalis v10.0 reference genome with STAR (version 2.7.0e)⁶³. The counts for known genes were obtained using featureCounts (version 2.0.1)⁵⁹ with the parameters -s 2 -p -BC. Normalized gene expression level was further calculated as transcripts per million (TPM). Differentially expressed genes were identified with the DESeq2 package⁶⁴, with threshold FDR < 0.05 and Log₂|FC| > 1. The normalized read coverage tracks in bigwig format were generated with bamCoverage in the deepTools package (version 3.5.1)⁶⁰, with the parameters --normalizeUsing RPKM -bs 5. The tracks were further visualized using Integrative Genomics Viewer (IGV, version 2.16.0)⁶¹.

ChIP-seq analysis

To analyze ChIP-seq data, quality check was performed using fastp with parameters --length_required 20 -q 30. The sequence reads were mapped to the X. tropicalis v10.0 reference genome by Bowtie2 (version 2.4.5)⁵⁶ with parameters --very-sensitive. Duplicates were removed with Sambamba (version 0.7.0)⁵⁷. Peak calling was performed by MACS2 (version 2.2.7.1)⁵⁸ with parameters -g 1.435e9 --keep-dup all -q 0.01. The normalized read coverage tracks (bigwig files) were obtained by bamCoverage with parameters --normalizeUsing CPM -bs 5.

Peaks were overlapped with the Bedtools (v2.30.0)⁶⁵ intersect function. Peaks within the defined promoter region were acquired using the window function with parameters -l 2000 -r 500. The calculation of peak distribution was performed by the HOMER tool ⁶² annotatePeaks.pl. Heatmaps were plotted with normalized read coverage tracks using deepTools (version 3.5.1)⁶⁰ function computeMatrix and plotHeatmap. The normalized ChIP-seq sequence tracks were visualized using the Integrative Genomics Viewer (IGV, version 2.16.0)⁶¹.

Data availability

All the sequencing data generated from this study have been submitted to the NCBI under the accession number GSE232071. For review purposes, please go to https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE232071, and enter token mnylkkyulhmbdwd into the box. The proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD051084 (Username: [email protected]; Password: ZYkzFGek).

Acknowledgements

This work was supported by the National Key R&D Program of China (Grant No. 2021YFF1201000 and 2022YFC3400400), the Shenzhen Key Laboratory of Gene Regulation and Systems Biology (Grant No. ZDSYS20200811144002008), the Shenzhen Science and Technology Program (Grant No. KQTD20180411143432337). We thank Dr. Zilong Wen and Dr. Xi Chen from SUSTech for the helpful discussion on the project. We also thank the Center for Computational Science and Engineering of SUSTech for the support on computational resource.

Author contributions

W.C. and H.C. conceived the study. W.C., H.C. and W.L. designed the experiments. H.C. and W.L. developed CANTAC-seq and performed the experiments. Z.Y. and Y.C. prepared frog embryos and performed microinjections and imaging. L.Z., G.L. and C.T. performed bioinformatics analysis. R.C., D.G., X.S. and Z.S. assisted in performing experiments. W.C., H.C., Y.C., Y.H., H.H., L.F. and Q.Z. reviewed and discussed the results. W.C. and H.C. wrote the manuscript with the input from W.L. and L.Z..

Competing interests

The authors declare no competing interests.

Tadros, W. & Lipshitz, H. D. The maternal-to-zygotic transition: a play in two acts. Development 136, 3033-3042, doi:10.1242/dev.033183 (2009).
Lee, M. T., Bonneau, A. R. & Giraldez, A. J. Zygotic genome activation during the maternal-to-zygotic transition. Annu Rev Cell Dev Biol 30, 581-613, doi:10.1146/annurev-cellbio-100913-013027 (2014).
Schulz, K. N. & Harrison, M. M. Mechanisms regulating zygotic genome activation. Nat Rev Genet 20, 221-234, doi:10.1038/s41576-018-0087-x (2019).
Jukam, D., Shariati, S. A. M. & Skotheim, J. M. Zygotic Genome Activation in Vertebrates. Developmental Cell 42, 316-332, doi:10.1016/j.devcel.2017.07.026 (2017).
Newport, J. & Kirschner, M. A major developmental transition in early Xenopus embryos: II. Control of the onset of transcription. Cell 30, 687-696, doi:10.1016/0092-8674(82)90273-2 (1982).
Newport, J. & Kirschner, M. A major developmental transition in early Xenopus embryos: I. characterization and timing of cellular changes at the midblastula stage. Cell 30, 675-686, doi:10.1016/0092-8674(82)90272-0 (1982).
Tan, M. H. et al. RNA sequencing reveals a diverse and dynamic repertoire of the Xenopus tropicalis transcriptome over development. Genome Research 23, 201-216, doi:10.1101/gr.141424.112 (2013).
Owens, N. D. L. et al. Measuring Absolute RNA Copy Numbers at High Temporal Resolution Reveals Transcriptome Kinetics in Development. Cell Reports 14, 632-647, doi:10.1016/j.celrep.2015.12.050 (2016).
Klemm, S. L., Shipony, Z. & Greenleaf, W. J. Chromatin accessibility and the regulatory epigenome. Nature Reviews Genetics 20, 207-220, doi:10.1038/s41576-018-0089-8 (2019).
Zaret, K. S. & Carroll, J. S. Pioneer transcription factors: establishing competence for gene expression. Gene Dev 25, 2227-2241, doi:10.1101/gad.176826.111 (2011).
Lee, M. T. et al. Nanog, Pou5f1 and SoxB1 activate zygotic gene expression during the maternal-to-zygotic transition. Nature 503, 360-364, doi:10.1038/nature12632 (2013).
Lu, F. et al. Establishing Chromatin Regulatory Landscape during Mouse Preimplantation Development. Cell 165, 1375-1388, doi:10.1016/j.cell.2016.05.050 (2016).
Wu, J. et al. The landscape of accessible chromatin in mammalian preimplantation embryos. Nature 534, 652-657, doi:10.1038/nature18606 (2016).
Gao, L. et al. Chromatin Accessibility Landscape in Human Early Embryos and Its Association with Evolution. Cell 173, 248-259 e215, doi:10.1016/j.cell.2018.02.028 (2018).
Liu, G., Wang, W., Hu, S., Wang, X. & Zhang, Y. Inherited DNA methylation primes the establishment of accessible chromatin during genome activation. Genome Res 28, 998-1007, doi:10.1101/gr.228833.117 (2018).
Wu, J. et al. Chromatin analysis in human early development reveals epigenetic transition during ZGA. Nature 557, 256-260, doi:10.1038/s41586-018-0080-8 (2018).
Palfy, M., Schulze, G., Valen, E. & Vastenhouw, N. L. Chromatin accessibility established by Pou5f3, Sox19b and Nanog primes genes for activity during zebrafish genome activation. PLoS Genet 16, e1008546, doi:10.1371/journal.pgen.1008546 (2020).
Amodeo, A. A., Jukam, D., Straight, A. F. & Skotheim, J. M. Histone titration against the genome sets the DNA-to-cytoplasm threshold for the Xenopus midblastula transition. P Natl Acad Sci USA 112, E1086-E1095, doi:10.1073/pnas.1413990112 (2015).
Joseph, S. R. et al. Competition between histone and transcription factor binding regulates the onset of transcription in zebrafish embryos. Elife 6, doi:ARTN e2332610.7554/eLife.23326 (2017).
Stancheva, I. & Meehan, R. R. Transient depletion of xDnmt1 leads to premature gene activation in Xenopus embryos. Genes Dev 14, 313-327 (2000).
Ruzov, A. et al. Kaiso is a genome-wide repressor of transcription that is essential for amphibian development. Development 131, 6185-6194, doi:10.1242/dev.01549 (2004).
Hellsten, U. et al. The genome of the Western clawed frog Xenopus tropicalis. Science 328, 633-636, doi:10.1126/science.1183670 (2010).
Gentsch, G. E., Spruce, T., Owens, N. D. L. & Smith, J. C. Maternal pluripotency factors initiate extensive chromatin remodelling to predefine first response to inductive signals. Nat Commun 10, 4269, doi:10.1038/s41467-019-12263-w (2019).
Esmaeili, M. et al. Chromatin accessibility and histone acetylation in the regulation of competence in early development. Dev Biol 462, 20-35, doi:10.1016/j.ydbio.2020.02.013 (2020).
Bright, A. R. et al. Combinatorial transcription factor activities on open chromatin induce embryonic heterogeneity in vertebrates. EMBO J 40, e104913, doi:10.15252/embj.2020104913 (2021).
Hontelez, S. et al. Embryonic transcription is controlled by maternally defined chromatin state. Nat Commun 6, 10148, doi:10.1038/ncomms10148 (2015).
Lund, E., Liu, M. Z., Hartley, R. S., Sheets, M. D. & Dahlberg, J. E. Deadenylation of maternal mRNAs mediated by miR-427 in Xenopus laevis embryos. Rna 15, 2351-2363, doi:10.1261/rna.1882009 (2009).
Mukherjee, S. et al. Sox17 and beta-catenin co-occupy Wnt-responsive enhancers to govern the endoderm gene regulatory network. Elife 9, doi:ARTN e5802910.7554/eLife.58029 (2020).
Chen, H. & Good, M. C. Nascent transcriptome reveals orchestration of zygotic genome activation in early embryogenesis. Curr Biol 32, 4314-4324 e4317, doi:10.1016/j.cub.2022.07.078 (2022).
Zhang, C., Basta, T., Jensen, E. D. & Klymkowsky, M. W. The beta-catenin/VegT-regulated early zygotic gene Xnr5 is a direct target of SOX3 regulation. Development 130, 5609-5624, doi:10.1242/dev.00798 (2003).
Tao, J. et al. BMP4-dependent expression of Xenopus Grainyhead-like 1 is essential for epidermal differentiation. Development 132, 1021-1034, doi:10.1242/dev.01641 (2005).
Zhang, C., Basta, T., Fawcett, S. R. & Klymkowsky, M. W. SOX7 is an immediate-early target of VegT and regulates Nodal-related gene expression in Xenopus. Dev Biol 278, 526-541, doi:10.1016/j.ydbio.2004.11.008 (2005).
Cao, Y. et al. POU-V factors antagonize maternal VegT activity and beta-Catenin signaling in Xenopus embryos. EMBO J 26, 2942-2954, doi:10.1038/sj.emboj.7601736 (2007).
Chiu, W. T. et al. Genome-wide view of TGFbeta/Foxh1 regulation of the early mesendoderm program. Development 141, 4537-4547, doi:10.1242/dev.107227 (2014).
Niu, L. et al. Three-dimensional folding dynamics of the Xenopus tropicalis genome. Nat Genet 53, 1075-1087, doi:10.1038/s41588-021-00878-z (2021).
Kent, L. N. & Leone, G. The broken cycle: E2F dysfunction in cancer. Nat Rev Cancer 19, 326-338, doi:10.1038/s41568-019-0143-7 (2019).
Jukam, D., Kapoor, R. R., Straight, A. F. & Skotheim, J. M. The DNA-to-cytoplasm ratio broadly activates zygotic gene expression in Xenopus. Curr Biol 31, 4269-4281 e4268, doi:10.1016/j.cub.2021.07.035 (2021).
Paraiso, K. D. et al. Endodermal Maternal Transcription Factors Establish Super-Enhancers during Zygotic Genome Activation. Cell Rep 27, 2962-2977 e2965, doi:10.1016/j.celrep.2019.05.013 (2019).
Zou, Z. et al. Translatome and transcriptome co-profiling reveals a role of TPRXs in human zygotic genome activation. Science 378, abo7923, doi:10.1126/science.abo7923 (2022).
Szklarczyk, D. et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res, doi:10.1093/nar/gkac1000 (2022).
Chen, H. Z., Tsai, S. Y. & Leone, G. Emerging roles of E2Fs in cancer: an exit from cell cycle control. Nature Reviews Cancer 9, 785-797, doi:10.1038/nrc2696 (2009).
Edgar, B. A., Kiehle, C. P. & Schubiger, G. Cell cycle control by the nucleo-cytoplasmic ratio in early Drosophila development. Cell 44, 365-372, doi:10.1016/0092-8674(86)90771-3 (1986).
Kane, D. A. & Kimmel, C. B. The zebrafish midblastula transition. Development 119, 447-456, doi:10.1242/dev.119.2.447 (1993).
Syed, S., Wilky, H., Raimundo, J., Lim, B. & Amodeo, A. A. The nuclear to cytoplasmic ratio directly regulates zygotic transcription in Drosophila through multiple modalities. Proc Natl Acad Sci U S A 118, doi:10.1073/pnas.2010210118 (2021).
Strong, I. J. T., Lei, X., Chen, F., Yuan, K. & O'Farrell, P. H. Interphase-arrested Drosophila embryos activate zygotic gene expression and initiate mid-blastula transition events at a low nuclear-cytoplasmic ratio. PLoS Biol 18, e3000891, doi:10.1371/journal.pbio.3000891 (2020).
Chan, S. H. et al. Brd4 and P300 Confer Transcriptional Competency during Zygotic Genome Activation. Dev Cell 49, 867-881 e868, doi:10.1016/j.devcel.2019.05.037 (2019).
Attwooll, C., Lazzerini Denchi, E. & Helin, K. The E2F family: specific functions and overlapping interests. EMBO J 23, 4709-4716, doi:10.1038/sj.emboj.7600481 (2004).
Chong, J. L. et al. E2f1-3 switch from activators in progenitor cells to repressors in differentiating cells. Nature 462, 930-934, doi:10.1038/nature08677 (2009).
Destree, O. H. et al. Structure and expression of the Xenopus retinoblastoma gene. Dev Biol 153, 141-149, doi:10.1016/0012-1606(92)90098-2 (1992).
Cosgrove, R. A. & Philpott, A. Cell cycling and differentiation do not require the retinoblastoma protein during early Xenopus development. Developmental Biology 303, 311-324, doi:10.1016/j.ydbio.2006.11.015 (2007).
Owens, D. A. et al. High-throughput analysis reveals novel maternal germline RNAs crucial for primordial germ cell preservation and proper migration. Development 144, 292-304, doi:10.1242/dev.139220 (2017).
Subtelny, A. O., Eichhorn, S. W., Chen, G. R., Sive, H. & Bartel, D. P. Poly(A)-tail profiling reveals an embryonic switch in translational control. Nature 508, 66-71, doi:10.1038/nature13007 (2014).
Peshkin, L. et al. On the Relationship of Protein and mRNA Dynamics in Vertebrate Embryonic Development. Dev Cell 35, 383-394, doi:10.1016/j.devcel.2015.10.010 (2015).
Nieuwkoop & Faber. Normal Table of Xenopus laevis (Daudin). Garland Publishing Inc, New York (1994).
Chen, S., Zhou, Y., Chen, Y. & Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884-i890, doi:10.1093/bioinformatics/bty560 (2018).
Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat Methods 9, 357-359, doi:10.1038/nmeth.1923 (2012).
Tarasov, A., Vilella, A. J., Cuppen, E., Nijman, I. J. & Prins, P. Sambamba: fast processing of NGS alignment formats. Bioinformatics 31, 2032-2034, doi:10.1093/bioinformatics/btv098 (2015).
Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol 9, R137, doi:10.1186/gb-2008-9-9-r137 (2008).
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923-930, doi:10.1093/bioinformatics/btt656 (2014).
Ramirez, F. et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res 44, W160-165, doi:10.1093/nar/gkw257 (2016).
Robinson, J. T. et al. Integrative genomics viewer. Nat Biotechnol 29, 24-26, doi:10.1038/nbt.1754 (2011).
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell 38, 576-589, doi:10.1016/j.molcel.2010.05.004 (2010).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21, doi:10.1093/bioinformatics/bts635 (2013).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550, doi:10.1186/s13059-014-0550-8 (2014).
Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841-842, doi:10.1093/bioinformatics/btq033 (2010).

There is NO Competing Interest.

TableS1S8.xlsx
Table S1-S8
VideoS1.mp4
Video S1
SupplementaryFigures17.pdf
FIGURELEGENDSOFSUPPLEMENTARYFIGURES.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

CANTAC-seq analysis reveals E2f1 and Otx1 as a coupled repressor-activator pair co-modulating zygotic genome activation in Xenopus tropicalis

Status:

Version 1

Abstract

Figures

INTRODUCTION

RESULTS

Development of CANTAC-seq

E2f1 is a repressor of minor zygotic genome activation

E2f1 represses zygotic gene transcription independent of its regulation of cell cycle progression

Otx1 antagonizes the inhibitory effect of E2f1 during minor ZGA

E2f1 and Otx1 co-modulate zygotic gene transcription by a seesaw balance

DISCUSSION

METHODS

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1