Development and application of simple FRET-based methods for aggregation-prone LIM domain interactions

Abstract

LIM-only (LMO) and LIM-homeodomain (LIM-HD) proteins are important mediators of cell specification, proliferation and differentiation. These transcription factors all contain two tandem LIM domains (LIM1+2), which are non-classical zinc finger motifs that mediate protein-protein interactions. Many co-factors of these proteins contain LIM interacting domains (LIDs). The LID is a ~30-residue intrinsically disordered region (IDR) that folds upon binding to LIM1+2 domains. LID:LIM1+2 interactions and the competition established through different combinations of different binding partners play an important role in neural development and breast cancer. The ability to estimate affinities for these interactions would help provide mechanistic insight into LMO and LIM-HD complex formation and regulation. However, the propensity of LIM1+2 domains from LMO/LIM-HD proteins to aggregate and precipitate during recombinant protein production have made it difficult to measure binding affinities for LID:LIM1+2 interactions. This thesis outlines the design, optimisation and application of a series of Förster Resonance Energy Transfer (FRET)-based approaches to study LID:LIM1+2 interactions. LIM1+2 aggregation is prevented by tethering the domains to a LID using a flexible polypeptide linker. The interacting domains are in turn fused to fluorescent proteins that are optimised for FRET. Specific proteolytic cleavage of the linker allows equilibrium binding constants and dissociation rates to be determined using homologous competition and dilution-based approaches. Through the application of these simple FRET-based binding methods, this thesis reveals previously unappreciated and unknown properties of LMO and LIM-HD proteins. This work provides tools for studying other aggregation-prone proteins, as well as general implications for the activity of transcription factors and IDR interactions.