3018 – FUNCTIONAL DISSECTION OF TRANSCRIPTION FACTOR NETWORKS GOVERNING MULTIPOTENT PROGENITOR STATES

Abstract

Recent breakthroughs in single cell technology have delivered maps of cell states together with their differentiation journeys at unprecedented resolution and scale. However, despite great progress in network inference methods, our understanding of molecular networks governing differentiation is limited due to the complexity and lack of systematic, functional data. In this study, we demonstrate a scalable approach to functionally define gene networks using genetic perturbations (CRISPR/Cas9) followed by transcriptional profiling. We utilise a lympho-myeloid progenitor model (Hoxb8-FL) to investigate mechanisms maintaining the propensity to differentiate into multiple lineages. In these cells, we built a network spanning ∼17,000 interactions between 39 transcriptions factors (TFs) and their target genes. As functional interpretation of gene perturbations is challenging, we provide a new method (DoT score) which visualises state transitions using scRNA-Seq landscapes as references, offering an alternative to Gene Ontology analysis. Furthermore, in-depth analysis of the network revealed how several lineage-associated TFs cooperate to establish lymphoid and myeloid transcription programmes. We find that myeloid gene expression is activated by Cebpa, but is kept in check by enforced expression of Hoxb8 and its downstream regulators Meis1 and Hoxa9. In parallel, Ebf1 is essential for a proper balance of cell growth and proliferation and, consistently with its established role, promotes the B cell expression programme. Conversely, B-cell differentiation is counteracted by Gata3, which efficiently blocks Pax5 expression. Altogether, our work unravels a core TF circuit, simultaneously activating several lineage programmes but wired specifically to prevent any lineage progression. Molecularly defined regulatory networks such as the one reported here may be utilised for targeted modulation of cell states, with broad applicability for the production of desired cells through targeted differentiation, as well as resetting abnormal cellular states in haematopoietic malignancies.