RNP interfaces as regulatory "active sites": a repurposing strategy for small molecules targeting viral...

RNA-protein (RNP) complexes regulate nearly every stage of gene expression and play central roles in viral infection and human disease. Despite decades of research, relatively few small molecules (SMs) have been shown to modulate RNP assemblies with mechanistic understanding. A major challenge in targeting RNA arises from the absence of well-defined SM binding pockets analogous to the catalytic active sites that guide conventional protein-directed drug discovery. However, in many biological contexts, RNP interfaces function as the effective "active sites" of regulatory RNA, where RNA structure and protein recognition surfaces converge to regulate gene expression. In this Perspective, we argue that the major obstacle to therapeutic progress is the difficulty of identifying functionally and structurally characterized RNP interfaces within highly dynamic regulatory networks. To address this challenge, we propose an integrative discovery framework centered on RNP interfaces that integrates state-of-the-art structure prediction, molecular dynamics simulations, ensemble-based virtual screening, and orthogonal biophysical validation to enable rational repurposing of FDA approved SMs. Viral $5^{\text{'}}$ -UTR untranslated regions ( $5^{\text{'}}$ -UTRs) provide a compelling context for this strategy, as they function as structural scaffolds that present conserved RNP interfaces essential for translation and replication. By focusing on minimal RNP fragments that consist of recurrent structural motifs such as bulge loops, ensemble sampling can reveal transient pockets suitable for SM virtual docking. Using Enterovirus A-71 $5^{\text{'}}$ -UTR as an illustrative example, we outline how interface-guided modeling can prioritize SMs capable of modulating specific RNP interactions. This ensemble-guided framework offers a generalizable strategy for accelerating the development of RNP-targeted therapies in viral and disease contexts.