The LMO4-DEAF1 complex and its biological impacts

Abstract

LIM domains are non-classical zinc fingers that bind other proteins. They have diverse cellular functions that range from regulating gene expression to maintaining the structural integrity of the cell. LIM domain-only proteins (LMO1–4) contain only two LIM domains. They are involved in transcriptional regulation, embryonic development and are also implicated in numerous cancers. LMO4 is essential for embryonic development and it is overexpressed in 50% of sporadic breast cancers. LMO2 is known to nucleate haematopoietic transcriptional complexes by binding multiple protein co-factors. It has been generally assumed that LMO4 also nucleates multi-protein transcriptional complexes. This thesis describes how LMO4 binds Deformed Epidermal Autoregulatory Factor-1 (DEAF1). DEAF1 is a modular transcription factor that controls diverse processes such as embryonic development, mood, innate immunity and autoimmunity. LMO4 and DEAF1 are co expressed in a range of cell types, and exhibit similar knock-in or knock-out phenotypes. I show that the LIM interacting domain (LID) of DEAF1 (DEAF1LID), which spans residues 404–438, is unstructured in solution but becomes structured upon binding to LMO4. I show that DEAF1454–487, which harbours a nuclear export signal (NES), is a tetrameric coiled-coil domain. I also show that a dimerisation domain lies N terminal of the DNA-binding SAND domain. I show that although both LIM domains of LMO4 appear to be required for a high affinity interaction with DEAF1, the C-terminal LIM domain (LMO4LIM2) is the more important contributor to binding. I solved the solution structure of an engineered LMO4LIM2-DEAF1404–418 complex and confirm that DEAF1 binds the same face of LMO4 as LDB1 and CtIP, other well characterised protein partners of LMO4. This suggests that competition for binding LMO4 occurs when these proteins are co expressed. Thus, overexpression of LMO4 might disrupt the natural stoichiometries of LMO4 containing complexes and/or sequester binding partners that can perturb normal cell function to cause diseases, such as cancer. Finally, I propose a model of how a complex containing LMO4 and DEAF1 might regulate gene expression.