Bioinformatics analysis of multi-omics data elucidates the primary function of Oct4 in gene regulation and pluripotency maintenance

The transcription factor (TF) Oct4 is essential for maintaining pluripotency in vitro as well as in vivo. The crucial function of Oct4 in control of pluripotency has been extensively studied during reprogramming of somatic cells to induced pluripotency cells (iPSCs). However, how Oct4 maintains pluripotency in native embryonic stem cells (ESCs) remains far from being fully understood. Moreover, Oct4 cooperates with the TF Sox2 at thousands of genomic sites in ESCs to fulfill its functional role, but the interplay between Oct4 and Sox2 in pluripotency control has not been completely elucidated. In this study, we aim to investigate the primary function of Oct4 in the control of pluripotency in mouse ESCs. To achieve this goal, we used a doxycycline inducible Oct4 loss-of-function mouse ES cell line. We then conducted a high-resolution time course study to monitor changes in the transcriptome by transient transcriptome sequencing (TT-seq), changes in chromatin accessibility by assay for transposase accessible chromatin with sequencing (ATAC-seq), and changes in Oct4 and Sox2 occupancy by chromatin immunoprecipitation and sequencing (ChIP-seq). We first used the TT-seq and ATAC-seq data to segment the genome into transcription units (TUs) and non-transcribed regions. We then classified TUs into different types including mRNAs and enhancer RNAs (eRNAs). Differential gene expression analysis revealed a rapid downregulation of the components of the pluripotency transcriptional network with super enhancer (SE)-controlled genes being immediately and strongly affected upon Oct4 depletion. We then derived a refined annotation of putative enhancers and identified Oct4-bound transcribed enhancers in mouse ESCs. We found that Oct4 depletion led to immediate loss of Oct4 binding and a rapid loss of eRNA synthesis at Oct4-bound transcribed enhancers, which correlated well with a decrease in mRNA synthesis from nearby putative target genes. In contrast, chromatin accessibility at Oct4-bound enhancers was generally decreased only later or remained unchanged at a subset of enhancers, arguing that Oct4 primarily functions in transcription activation, not chromatin opening. Furthermore, we found that binding of Sox2 at Oct4-bound enhancers remained unchanged during the first 6 hours of Oct4 depletion and it correlated well with the delayed chromatin accessibility changes. Our data provide new insights into the direct target genes of Oct4. We identified 446 early down-regulated genes upon Oct4 depletion and these genes are strongly enriched for stem cell population maintenance, suggesting that a key pluripotency gene network centered around SEs is the primary/direct downstream targets of Oct4. Importantly, our results suggest that the key role of Oct4 in pluripotency maintenance in ESCs is to act as a transcriptional activator that stimulates transcription of pluripotency enhancers and their target genes, whereas Sox2 may act as a factor that renders enhancers accessible. In summary, our work illustrates the primary function of Oct4 and its interplay with Sox2 in the controlling the enhancer landscape to regulate pluripotency gene expression in mouse ESCs