Embryonic development and organogenesis depend on the precise spatiotemporal expression of specific sets of genes. Precisely controlled gene expression ensures cell state transitions, especially in the early stages of development, as gastrulation. These complex multi-layered cellular processes are orchestrated by the interfacing of the epigenome, 3-dimensional (3D) nuclear organization, cis-regulatory elements (CREs) with transcription factors (TF), and long non-coding RNAs (lncRNAs). In the gastrulating embryo, definitive endoderm is specified from the pluripotent epiblast following a series of regulatory events, including the activation of SOX17, a key TF of that particular germ layer. Although SOX17 has been extensively studied in early embryonic development, the precise control of its activation, the locus, and the epigenetic rules governing its genetic regulatory network (GRN) remains poorly investigated. In my thesis, I in-depth characterized the human SOX17 locus, exploring the relevance and regulatory impact of 3D nuclear organization, its distal CREs, and their activity. I applied a series of loss of function (LOF) and transgenic experiments to dissect the locus at a satisfactory resolution. In particular, I showed SOX17 among a subset of developmental regulators topologically isolated within CTCF-CTCF loop domains and highlighted the importance of gene control in 3D within this type of domain. I pinpointed the relevance of SOX17’s distal CREs and their definitive endoderm-specific interaction and showed this interaction to be highly dependent on CTCF-CTCF loop-formation to guarantee proper gene control. I found CRE-dependent SOX17 gene deregulation associated with poor definitive endoderm differentiation outcome and a stalled “mesendodermal-like” phenotype. Assessing the genetic identity of different CREs, I divulged the presence of a novel lncRNA within the locus, namely LNCSOX17. I fully characterized LNCSOX17 and established its identity as a bona fide lncRNA through a series of genetic perturbations. I demonstrated the importance of LNCSOX17 for forming definitive endoderm and the lack of participation in SOX17 cis-acting gene control. I associated the loss of LNCSOX17 RNA but not its active transcription at the locus with an aberrant endodermal transcriptome, a lack of epithelial-to-mesenchymal transition (EMT), and the hyperactivity of the detrimental definitive endoderm JNK/JUN/AP1 signaling pathway. I found definitive endoderm lacking LNCSOX17 to be functionally impeded in the generation of pancreatic progenitor populations. The studies within this thesis serve as valuable examples to support the functional relevance of 3D nuclear organization and its importance for developmental gene control in cis via CTCF-CTCF loop domain-mediated CRE-promoter contact facilitation. They associate developmental gene expression levels with various phenotypes, identify a so far unknown developmental lncRNA molecule, and imply its relevance for the formation of definitive endoderm. The outlined results advance our knowledge of developmental TF gene-control and its importance for the development of human definitive endoderm.
