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 that interact with other proteins. Many LMO and LIM-HD co-factors contain intrinsically disordered LIM interacting domains (LIDs), with competition between different LIM and LID transcription factors playing an important role in neural development and several cancers. Estimating affinities for these interactions would help define the molecular basis of LMO and LIM-HD function; however LIM domain aggregation prevents the use of standard binding methods.
We have designed Förster Resonance Energy Transfer (FRET)-based approaches to study LIM:LID interactions. We have expressed LID and LIM domains fused with a flexible linker to prevent aggregation, with each domain in turn fused to fluorescent proteins optimised for FRET. Proteolytic cleavage of this linker has allowed us to study binding using homologous competition and dilution-based approaches. These have revealed LMO and LIM-HD proteins can bind their major LID-containing co-factor with association rates that differ by 3-orders of magnitude, and that these differences are not due to electrostatic enhancement. Perturbation experiments suggest related LMO and LIM-HD proteins can interact with the same LID through different binding mechanisms. We have performed modeling to show the importance of these highly differential association rates in the regulation of LMO and LIM-HD co-factor competition during motor neuron development. Overall, our suite of FRET techniques provide simple yet informative methods for studying the biophysical basis of LMO and LIM-HD activity, as well as providing a tool for studying other aggregation-prone proteins.