Mechanistic comparison of structurally divergent transcriptional coactivators through covalent activator‐coactivator complexes

Abstract

Transcriptional coactivators act as essential linchpins between DNA‐bound transcriptional activators and the rest of the transcriptional machinery. The coactivator Med25, of the human Mediator complex, contains a unique coactivation domain, the Activator Interacting Domain (AcID), which links several disease‐associated activators and the transcriptional machinery. The AcID motif is composed of both alpha helices and beta sheets whereas the most well characterized coactivation domains (KIX and TAZ1) are entirely alpha helical in structure. The molecular basis of activator‐coactivator interactions are most thoroughly understood in those between activators and the KIX domain of CBP/p300. Intrinsic disorder and allostery have been identified as critical KIX features for mediating binary and ternary complex formation between transcriptional activators and KIX at two distinct activator binding sites. We sought to probe the molecular basis of AcID‐activator interactions specifically examining the roles of intrinsic disorder and allostery in mediating AcID‐transcriptional activator complexes. The transcriptional activator, VP16 (Herpes simplex virus), contains two tandem, yet functionally independent, transcriptional activation domains (TAD) that have been shown to interact with AcID at two distinct binding surfaces. These protein‐protein interactions (PPI) were probed via the construction of cysteine‐containing VP16‐derived peptides that can be covalently tethered to AcID through the formation of activator‐AcID disulfides. Native, or mutationally introduced, cysteines at several positions on AcID enable specific formation of VP16‐AcID binary complexes at either AcID binding surface. This selective control enables us to probe activator binding at each binding surface individually, vastly simplifying our analysis of these binding events and allowing observation of essential differences in the mechanism of binary and ternary complex formation of the AcID and KIX domains. A suite of direct binding, kinetics, and NMR performed using these covalent complexes have revealed that Med25 AcID, with its unique beta‐sheet rich architecture, diverges from KIX in its mechanism of transcriptional activator complex formation. AcID, like KIX, possesses an allosteric network linking both activator binding surfaces but these AcID binding sites differ in their ability to accommodate substrates bound in many different conformations forming so‐called “fuzzy” interfaces. This refined understanding of Med25 AcID has guided efforts to develop small‐molecule inhibitors of AcID capable of inhibiting specific activator‐AcID interactions.