Defining Binding Principles to Target Coactivator Med25-Activator Interaction

Abstract

Transcription is the process by which the information encoded in our DNA is converted to RNA. This important event is guided by an assembly of proteins brought together at specific gene promoter sites. A key part of the assembly is a coactivator complex called Mediator that interacts with transcriptional activator proteins. Mediator is therapeutically relevant because these interactions are misregulated in many diseases. Unfortunately, the coactivator class of proteins are notoriously difficult to target with druglike molecules. Coactivators are usually dynamic in nature and poorly understood mechanistically, making it quite challenging to design or discover small molecule modulators. Because of this, there is a lack of useful small molecule probes that target coactivators such as those within the Mediator complex. The coactivator Med25 is a component of Mediator that binds transcriptional activators using its Activator Interaction Domain (AcID) and there is considerable evidence that its network of protein-protein interactions (PPIs) are dysregulated in certain cancers. Although Med25 is conformationally dynamic like other coactivators, Med25-AcID has a structurally unique fold. It contains a central β-barrel that is flanked by dynamic loop and helix substructures and the prevailing structural model is that the featureless surfaces of the β-barrel are the key binding surfaces for activators. This dissertation defines important binding mechanism principles for Med25-AcID and activator protein interactions that allow for small molecule probes. It was found using a series of binding and mutagenesis structures that the activators ATF6 and ERM, despite sequence similarities, bind to opposite faces of the AcID β-barrel. Further computational and binding experiments revealed that charge-containing dynamic loop structures adjacent to the binding surfaces are important in these interactions. In Chapter 3, I report the successful targeting of one of the dynamic substructures in Med25 with a covalent disulfide fragment that influences the conformation and binding of Med25-AcID. This was accomplished through a mass spectrometry based Tethering screen used because of our lab’s previous success in targeting dynamic coactivator proteins. A disulfide fragment library was used to identify a lead fragment that changes the kinetic koff signature of two Med25-binding activators. Furthermore, the fragment stabilizes the AcID domain and uses a chiral surface that allows for sufficient disulfide exchange that may prove beneficial in changing activator binding effects. Taken together, results defined in this thesis will provide a framework to learning about coactivator-activator interaction’s mechanistic features to discover probes for this class of proteins.