Drosophila Groucho (Gro) is the founding member of a conserved metazoan family of global corepressors, with orthologs in worms and vertebrates including humans. Related corepressors are also found in fungi and plants. Gro family members are characterized by conserved N- and C-terminal domains, termed the Q and WD-repeat domains, respectively, and a less conserved internal region that can be divided into subregions termed the GP, CcN, and SP domains. In contrast to the Q and WD-repeat domains, the internal domains have not been extensively studied, and indeed their lack of conservation between phyla leads to speculation that they may be expendable for repression.

Gro family proteins mediate combinatorial regulation of gene expression by allowing repressors to dominate over activators under certain conditions. They facilitate a myriad of developmental processes that are essential for normal growth and development, and their misregulation can lead to disease. As a corepressor, Gro is recruited to silencers by interactions with DNA-bound transcription factors. Once Gro is recruited to the template, it represses in a manner that is at least partly dependent upon Gro self-association and Gro interactions with histone deacetylases. Interactions between Gro and histone tails may also contribute to repression. The exact nature of the various protein-protein interactions that contribute to Gro-mediated repression, and the composition and stoichiometry of the repression complexes formed are poorly understood.

In this thesis, I seek to elucidate the mechanism of Gro-mediated repression and the role of Gro in pattern formation by analyzing multiple Gro deletion mutants lacking one or more of the central domains. Using both gain-of-function and loss-of-function approaches, I find that the GP and CcN domains are essential for Gro function. Conversely, I find that the SP domain negatively regulates Gro function, and this negative regulation is required for full viability. Furthermore, the SP domain is required for the specificity of repression, and in its absence Gro promiscuously represses transcription of genes that are not normal targets of Gro. My studies indicate that development requires the correct balance between these positively and negatively acting domains as excess Gro activity results in the inappropriate resetting of transcription factor concentration thresholds sufficient to mediate target gene repression. They also show that the unconserved central regions are critical for Gro-mediated repression challenging the notion that functionality leads to evolutionary conservation. I speculate that these domains are intrinsically disordered, thus allowing them to bind a wide variety of targets and thereby serve as regulatory hubs.

To further facilitate the mechanistic understanding of the diverse roles of Gro in development and the contribution of the unconserved central region to this process, recombinant proteins consisting of glutathione-S-transferase (GST) fused to individual Gro domains were used as affinity reagents to purify Gro-interacting proteins from Drosophila embryo nuclear extracts. Affinity purified proteins were then identified by Multidimensional Protein Identification Technology (MudPIT). This approach led to the identification over 140 Gro interacting partners involved in such processes as nucleosome assembly, pattern formation, DNA repair, RNA processing, and signal transduction. The list of interacting proteins includes expected target proteins known from previous genetic and biochemical studies, as well as novel targets, and is consistent with the role of Gro as a pleiotropic global regulator. The finding that the Gro central regions engage in such a diverse array of interactions is consistent with the idea that the central region may be a regulatory hub and may account for Gro's position as the "champion corepressor".