Deciphering the membrane topology of the pestiviral non-structural protein 4B (NS4B)

Pestiviruses like bovine viral diarrhea virus and other members of the family Flaviviridae form replication complexes consisting of viral non-structural and cellular proteins at rearranged intracellular membranes. Despite the pivotal roles of non-structural protein 4B (NS4B) throughout the pestiviral life cycle, little is known about how this protein exerts its multiple functions. It is assumed that pestiviral NS4B promotes replication complex assembly and virion morphogenesis by interacting with defined sets of viral and host proteins. The membrane topology of the protein dictates the availability of individual protein structures and interfaces for such interactions. Thus, the knowledge of the NS4B membrane topology is required for a detailed functional understanding of this protein. Therefore, we experimentally determined the membrane topology for NS4B in cellulo by using the substituted cysteine accessibility method (SCAM) in combination with computational secondary structure and transmembrane domain (TMD) predictions. Our model indicates the formation of two TMDs in the N-terminal region of NS4B (TMD2-3) followed by nine putative membrane-associated α-helices. Furthermore, a dual topology of the N-terminal amphipathic α-helix AH1 was detected by applying a Split-GFP assay, exposing similarities to hepatitis C virus NS4B. The translocation of AH1 across the membrane and the luminal orientation of the proposed loop connecting TMD2-3 was further confirmed by glycosylation acceptor site recognition analysis. Together, our model will assist further studies on the diverse functions of pestiviral NS4B throughout the viral life cycle.IMPORTANCEMembrane proteins are of special importance for positive-strand RNA viruses due to their replication at remodeled intracellular membranes. Moreover, the multi-functionality of these proteins can rely on alternative topologies. Studying their membrane topologies is challenging since protein purification can induce misfolding. Similarly, random insertions of large N-glycosylation acceptor sites may disturb transmembrane domains and thus the topology, while minimal glycosylation motifs (NXT/S) are often inefficiently glycosylated. Therefore, we used the SCAM assay utilizing single cysteine substitutions to analyze the membrane topology of BVDV-1 NS4B. A dual topology of the N-terminal region was demonstrated by a Split-GFP assay. Glycosylation acceptor site insertions at pre-analyzed positions further corroborated the model. In sum, BVDV-1 NS4B topology shows similarities but also remarkable differences to the NS4B membrane topologies of other Flaviviridae orthologues. This new information will allow further studies to clarify the molecular basis of the multi-functionality of this critical viral component.