Elucidating the cis-regulatory logic of Runx1 during developmental haematopoiesis

The study of dynamic gene expression during development utilises model differentiation systems. The haematopoietic system is one of the archetypal differentiation systems and many model gene loci used to shed light on basic mechanisms of transcriptional regulation are important haematopoietic genes. In this project I set out to uncover transcriptional regulatory mechanisms of the gene Runx1. Runx1 is a transcription factor important for the development and function of the haematopoietic system. In particular, Runx1 is required for haematopoietic stem cell (HSC) generation during embryonic development. HSCs are born during an endothelial-to-haematopoietic transition (EHT). Runx1 is expressed during, and is absolutely required for, EHT and the birth of HSCs. Previous work in the lab established that Runx1 exhibits complex transcriptional regulation during EHT; Runx1 has two alternative promoters and several haematopoietic enhancers have been identified. Using cutting-edge chromatin assays I interrogated cis-interactions within the transcriptional regulatory domain of Runx1 when it was expressed at low and high levels. Capture-C showed that interactions between Runx1 promoters and the haematopoietic enhancers previously identified were more frequent in haematopoietic cells compared to undifferentiated mouse embryonic stem cells (mESCs). Increased Runx1 transcription was associated with an apparent increase in the hallmarks of loop extrusion, suggesting that the two might be causally related. Next, I set out to investigate whether functional redundancy exists between the haematopoietic Runx1 enhancers. Distinct upstream regulators of the enhancers that were identified suggested that they might not have entirely overlapping functions. While using CRISPR/Cas9 in mESCs to delete endogenous Runx1 enhancers and test their functional requirements, I uncovered interesting caveats about common genome engineering practices. Functional characterisation of enhancer-deleted mESCs differentiated in vitro to haematopoietic cells indicated that partial enhancer redundancy may exist. Finally, I investigated a potential mechanism of Runx1 alternative promoter choice involving CTCF/cohesin. I identified tissue-specific and ubiquitous CTCF binding sites in the Runx1 domain that were capable of acting as insulators in vitro. Cell type-specific binding of CTCF upstream of each of the Runx1 promoters was correlated to DNA demethylation and promoter activity. CTCF binding was seen upstream of a signifcant proportion of active promoters genome-wide, suggesting that modulating CTCF binding close to promoters, possibly via DNA methylation, may be a widespread transcriptional regulatory mechanism.