FUNCTIONAL DISSECTION OF TRANSCRIPTION FACTOR NETWORKS GOVERNING HAEMATOPOIETIC PROGENITOR STATES

Maintenance of the adult haematopoietic system requires carefully balanced cell fate decisions at the level of multipotent progenitor cells. These decisions are controlled by complex transcription factor (TF) networks. While some TF-target connections have been experimentally established, they are mostly inferred from correlations of either gene expression (RNA-Seq) or genome-wide binding events (ChIP-Seq). This limits our understanding and prevents informed manipulation of cell states. To address these issues, we used CRISPR/Cas9 system and small scale RNA-Seq to experimentally assay gene regulatory functions of 38 TFs in the Hoxb8-FL cellular system, which captures in vitro a lymphoid-primed multipotent progenitor (LMPP) like state. Our RNA-Seq analysis following each TF knockout revealed >15,000 functional activating/inhibitory relationships across >7000 genes. Network analysis augmented with genome-wide TF binding information predicted direct interactions, regulatory elements and putative new circuits maintaining the multipotent state. Among many others, this included a novel interplay between co-existing Ebf1 and Gata3 regulating expression of lymphoid and B-cell specific genes. To gain further insights into how TFs regulate cell state transitions, we developed a new computational method that visualises perturbation effects in the context of single cell expression landscapes. Importantly, this avoids using externally curated ontology databases and instead directly employs human or mouse expression data. In the current study, the analysis revealed how each TF is linked with diverse differentiation programs. In addition to revealing known functions, we also identified previously unrecognized associations, such as Meis1 suppression of the monocytic programme. Studies are underway to validate new circuits and their influence on differentiation in primary cells.