Research project Establishment of epigenetic landscapes

The early Drosophila embryo is patterned along the anterior-posterior and dorsal-ventral axes by transcription of developmental control genes in different parts of the embryo. Dorsal-ventral patterning is controlled by an intra-nuclear concentration gradient of Dorsal, a Rel-family transcription factor related to NF-kappaB. Over 50 Dorsal target genes are known, and this gene regulatory network constitutes one of the best understood in the development of any animal. Dorsal enters ventral nuclei at high levels in response to signaling by the transmembrane receptor Toll. Target genes such as twist (twi) and snail (sna) with low-affinity bindning sites are turned on in ventral, presumptive mesodermal cells where Dorsal concentration is highest.  In lateral regions of the embryo, cells destined to become neuroectoderm activate short-gastrulation (sog), brinker (brk) and rhomboid (rho) in response to low levels of Dorsal. The Snail repressor prevents expression of these neuroectodermal genes in the mesoderm. Finally, when Dorsal sites are positioned in proximity to AT-rich binding sites, Dorsal is converted to a repressor and thereby prevents expression of dorsal ectoderm targets such as tolloid (tld), decapentaplegic (dpp) and zerknüllt (zen) in lateral and ventral parts of the embryo.

The Dorsal protein forms a gradient along the ventral-dorsal axis to specify mesoderm, neurogenic ectoderm, and dorsal ectoderm. High Dorsal concentration activates genes in the mesoderm, whereas intermediate Dorsal concentration activates neurogenic ectoderm genes that are repressed in the mesoderm by Snail. Dorsal can be converted to a repressor to prevent expression of dorsal ectoderm targets in mesoderm and neurogenic ectoderm. In Toll10B mutants, the entire embryo is converted to mesoderm, Tollrm9/rm10 mutant embryos become neurogenic ectoderm, and in gd7 embryos all cells are turned into dorsal ectoderm.

Investigation of dorsal-ventral patterning in Drosophila provides a unique opportunity to study cell specification, since mutants in the Toll signaling pathway can be used to generate embryos with a homogenous population of cell types (see Fig. 1). In embryos derived from gd⁷ mutant mothers, Dorsal fails to enter the nucleus and the entire embryo is transformed into dorsal ectoderm. In embryos derived from Toll^rm9/rm10 mothers, all cells become neurogenic ectoderm since Dorsal uniformly enters nuclei at an intermediate concentration. Toll^10B mothers produce embryos where Dorsal enters all nuclei at high concentration thereby transforming them into mesoderm. Chromatin extracts from these mutant embryos circumvent the problem of a mixed population of cell types, but enable investigation of histone modifications and regulator binding to endogenous loci under conditions that the different cell types normally experience during development. These tools allow us to study mechanisms of tissue-specific gene expression in vivo, in a manner that is exclusive to the early Drosophila embryo.

We have found that the epigenetic landscape at two neuroectoderm genes, brk and sog, differ between the three gemlayers (Boija and Mannervik 2016). These loci are decorated with Polycomb-mediated H3K27me3 in the dorsal ectoderm where they are off, have H3K27ac in the neuroectoderm where they are on, and lack both modifications in the mesoderm where they are repressed by Snail (Boija and Mannervik 2016). Using PRO-seq, ChIP-seq, CUT&Tag, and ATAC-seq we have characterized how the chromatin gradually gains accessibility and histone modifications at all dorso-ventral genes as the cells become specified. This results in release of a promoter-proximal paused Pol II that is established already in naïve cells in the cell type where the gene is expressed (Hunt et al, in prep.) Interestingly, despite differences in chromatin state and transcription between tissues, the three-dimensional genome architecture does not differ between the three germlayers (Ing-Simmons et al. 2021). This shows that differential gene expression is independent of 3D chromatin conformation in the early embryo. Current work investigates how transcriptional enhancers communicate with promoters without changes in chromatin conformation, and how the chromatin state influences this communication and transcriptional bursting.