AI-Assisted Computed Structure Models for Pre-Ubiquitylation Complexes Assembled by Respiratory Syncytial Viral Suppressors of Cellular Interferon Response.

Multiple viruses suppress the antiviral defense system of the host for optimal growth and pathogenesis by co-opting the ubiquitin-mediated proteasomal system (UPS) that promotes the degradation of cellular substrates belonging to the interferon pathway. In the Orthopneumovirus genus, respiratory syncytial virus (RSV), a significant pathogen in human and other animals, employs a pair of viral nonstructural proteins (NS1, NS2) to assemble the UPS. The lack of experimental three-dimensional structures of the substrate proteins and the NS-assembled UPS has impeded progress in our understanding of the mechanism of this assembly process. In an effort to remedy this deficiency, I have taken advantage of the burgeoning field of AI (artificial intelligence) and machine learning programs, such as AlphaFold3, to model the pre-ubiquitylation cores in various combination of the subunits to construct three-dimensional structures, named 'computed structure models' (CSMs). The UPS core universally comprises an adapter protein connected to the "substrate" that is to be degraded by the "substrate receptor". The NS proteins are believed to act as receptors, and cellular Elongin BC as an adapter. These CSMs lend support to the biochemical results where known while also suggesting that the complete core of three proteins is energetically more stable than a complex of only the NS protein and the substrate. In the absence of experimental structures, these results offer, for the first time, a mechanistic insight into RSV-triggered assembly of the UPS, which should allow for a better design of future experiments, and eventually new antiviral regimens.