High resolution dissection of non-coding determinants of disease

While mutations at protein-coding sequences have been extensively characterized, the functional role of disease-associated non-coding variants remains largely elusive. In fact, germline non-coding variants are often associated with increased risk of childhood cancers. In B-cell derived acute lymphoblastic leukemia (B-ALL), the most common type of cancer in children, non-coding sequence variation at lineage-specific genes, such as IKZF1, GATA3 and CEBPE is associated with disease. Interestingly, the strongest risk factor maps at ARID5B, a DNA-binding protein described to promote the removal of repressive histone marks. Still, the role of ARID5B in leukemia and haematopoiesis remains largely uncharacterized. Thus, we hypothesize that non-coding variants associated with B-ALL affect the activity of distal regulatory elements, modulating the expression of ARID5B and other B-cell lineage-specific genes (objective 1). We further postulate that ARID5B promotes the de-repression of lineage-specific genes (objective 2), playing a crucial role in B-cell development and B-ALL initiation and progression (objective 3). To test these hypotheses, I will develop a novel CRISPR screen approach to dissect ARID5B enhancers at single nucleotide resolution (work package 1). I will use a combination of genomics and transcriptomics methods after acute and prolonged degradation of ARID5B to elucidate the molecular function of ARID5B in B-ALL (work package 2). Finally, I will use mouse models and focus on the physiological relevance of ARID5B in B-cell development and leukemia initiation and progression, in vivo. (work package 3). Together, I will dissect the role of non-coding variants at leukemia-associated loci, characterize the molecular function of ARID5B and elucidate the pathophysiological relevance of ARID5B.