Non-canonical mechanisms of gene regulation by the histone demethylase KDM5

Abstract The lysine demethylase 5 (KDM5) family of transcriptional regulators are important for normal development and their dysregulation is associated with intellectual disability and with several forms of cancer. However, a lack of understanding of the normal physiological roles of KDM5 proteins has hindered our understanding of how their loss or gain leads to disease states. Thus, the long-term goal of these studies is to define the molecular mechanisms by which KDM5 regulates essential gene regulatory programs using the genetically elegant model organism Drosophila. Specifically, we focus on defining mechanisms of transcriptional regulation by KDM5 that are independent of its well- established histone demethylase activity. The importance of illuminating demethylase-independent gene regulatory mechanisms is highlighted by our observation that KDM5 carries out its essential developmental activities separately from its enzymatic activity. However, the molecular mechanisms underlying these non-canonical functions of KDM5 remain unknown. Our new data show that a region within the C-terminus of KDM5 that has no previously known function is required for viability. Moreover, interactome studies link the C-terminus of KDM5 to the non-specific lethal (NSL) transcriptional activation complex. These and other data lead us to propose the central hypothesis that a KDM5 interacts with the NSL complex to orchestrate gene expression programs needed for animal survival. To test this, we will use a range of genetic, molecular and cell biological approaches to define the target genes and pathways regulated by KDM5 that are critical for development, and to determine the molecular links between KDM5, NSL and transcriptional regulation. This study is innovative in our focus on defining demethylase-independent activities of KDM5 and our use of state- of-the-art techniques. This work is significant because we will provide fundamental insights into KDM5 function that will be broadly relevant for our understanding of gene regulation by multi-domain proteins. Our work is also expected to highlight potential mechanisms by which KDM5 dysregulation could contribute to intellectual disability and/or cancer.