Eight groups of rotaviruses (A-H) have been described. The group A rotaviruses are the most widely studied, as they represent the primary cause of rotavirus-associated morbidity and mortality in many animals (including humans). All rotaviruses face the similar challenge of replicating in cells with innate immune systems capable of sensing rotavirus RNAs, activating interferon (IFN) production, and synthesizing antiviral IFN-stimulated gene (ISG) products that can prevent virus progeny formation. Remarkably, although all rotavirus groups seem to have similar capsid structures and employ analogous genome replication cycles, the number and types of proteins that the viruses use to antagonize IFN and ISG expression differ significantly. This variation in protein antagonists reveals that the rotavirus groups have evolved multiple and distinct species-specific approaches to counter highly conserved cellular antiviral pathways. In the past year, the efforts of my research group have focused on defining the structure and function of three such antagonists: (i) the putative E3 ubiquitin ligase (NSP1) of group A, C, D, and F rotaviruses that block the activation of NFkB and induce the degradation of IFN-transcription factors (e.g., IRF3, IRF7), (ii) the VP3 phosphodiesterase (PDE) domain of group A, B, and G rotaviruses that degrade the RNase L activator made by oligoadenylate synthetase (OAS), and (iii) the dsRNA-binding protein (dsRBP) of group C and H rotaviruses that prevent activation of host dsRNA sensors (e.g., MDA5). Our results indicate that these antagonists all derive from cellular proteins that have been stripped down and repurposed, creating viral proteins that effectively impede host antiviral pathways. Remarkably, other classes of viruses encode proteins that appear structurally and functionally similar to the rotavirus antagonists suggesting an evolutionary sharing of genetic information. Results in specific research areas: (1) Putative viral E3 ubiquitin ligase (NSP1). The NSP1 proteins of clade A rotaviruses (group A, C, D and F) have features characteristic of simple single-chain E3 ligases, including an N-terminal RING domain and a C-terminal targeting domain. The RING domain of NSP1 E3 ligases is proposed to recruit a ubiquitin-charged E2 conjugating enzyme, positioning it in proximity to a target protein bound to the targeting domain. In this complex, the ubiquitin moiety is transferred onto the target protein, marking it for subsequent proteasomal degradation. We have performed experiments directed at identifying the E2 subtypes that support the function of the NSP1 E3 ligase. The approach that we used takes advantage of the fact that bacterial growth in chloramphenicol (CAM) is inhibited, unless overcome by the activity of soluble chloramphenicol acetyltransferase. Because of the insolubility properties of NSP1, fusion proteins consisting of CAT and NSP1 lack CAT activity, preventing bacterial growth in the presence of CAM. However, the solubility of CAT-NSP1 fusions is restored by interaction with E2 proteins that recognize the NSP1 RING domain. Using this assay system, we determined that the E2 subtypes UBCH5A and UBCH7 supported bacterial growth in the presence of CAM, suggesting that these E2s interact with the NSP1 RING. Interestingly, UBCH5A was previously found to support the activity of an HPV-associated E3 ligase. (2) Double-stranded RNA binding protein (dsRBP). Among the eight groups of rotaviruses, the group C viruses are unique, as its gene 7 RNA not only encodes the expected NSP3 protein, but also the smaller dsRBP. Although NSP3 and dsRBP derive from a single ORF within the gene 7 RNA, the presence of an intervening 2A-like stop-go translation element between the two cause NSP3 and dsRBP to be produced as separate proteins. Based on structure modeling and sequence homology, the group C dsRBP appears to represent the structural and functional equivalent of the dsRNA-binding domain (or dsRNA-binding motif, dsRBM) found in many cellular proteins that recognize and interact with dsRNA (e.g., ADAR, PKR, RNaseIII, Dicer). In the past year, we developed protocols for expressing and, then, purifying the group C dsRBP to homogeneity. Afterwards, the protein was crystallized and its structure determined at Argonne National Lab. The analysis indicates that the group C dsRBP is structurally similar to the dsRBM, indicating that the virus has acquired genetic information for its protein from a host gene encoding a protein with a dsRBM. Functional studies with the group C dsRBP confirm that the protein has affinity for dsRNA. Functional studies also indicate that the protein interferes with activity of cellular dsRNA sensors, perhaps by sequestering viral RNA molecules that serve as activators of such sensors.
Showing the most recent 10 out of 22 publications
