Rotaviruses (RVs), members of the Reoviridae family, have genomes consisting of eleven segments of double-stranded (ds) RNA. The genome of the RV virion is contained in a non-enveloped icosahedral capsid composed of three concentric protein layers. The outer layer consists of VP7 trimers, organized with T=13 symmetry, and VP4 spikes. The intermediate layered is formed by VP6 trimers, also with T=13 symmetry. The innermost protein layer is a smooth, thin, pseudo T=1 assembly formed from 12 decamers of the core lattice protein, VP2. Tethered to the underside of the VP2 layer are complexes comprised of the viral RNA-dependent RNA polymerase (RdRP), VP1, and the RNA-capping enzyme, VP3. Together, VP1, VP2, VP3, and the dsRNA genome form the core of the virion. The core proteins function together to transcribe the segmented dsRNA genome, producing eleven capped plus-sense (+)RNAs. The viral RdRP uses the (+)RNAs as templates for the synthesis of the dsRNA genome. Although the RdRP alone can recognize viral (+)RNAs, the polymerase is only active when VP2 is present. The VP2-dependent activity of VP1 provides a means by which genome replication (dsRNA synthesis) can be linked with genome packaging and core assembly. Newly made (+)RNAs pass from the RdRP to VP3, an enzyme which introduces m7G caps to the 5'-end of the transcripts through associated guanylyltransferase and methyltransferase activities. Genome replication and core assembly take place in cytoplasmic inclusion bodies of infected cells;these structures are referred to as viroplasms. Two viral nonstructural proteins, the octamer, NSP2, and the phosphoprotein, NSP5, direct the formation of viroplasms. The interactions of NSP2 and NSP5 with VP1, VP2, and VP3 are believed to coordinate genome replication and core assembly. The overriding goal of this project is to characterize the structure and function of the core proteins VP1, VP2, and VP3 and the viroplasm building-blocks NSP2 and NSP5. This includes defining the structural interfaces between the proteins and establishing how these interactions affect and regulate the activities of the proteins, and the assembly of virions. Progress toward this goal in 2013-2014 is summarized below. (1) A FLEXIBLE LOOP NEAR THE NTP/PPi EXCHANGE TUNNEL CONTRIBUTES TO RNA SYNTHESIS BY THE RV RNA POLYMERASE. The RV RdRP, VP1 (1089 amino acids), directs the synthesis of the viral dsRNA genome and mRNA transcripts during infection. The activities of VP1 take place within the viral core and hinge on indeterminate interactions between VP1 and the innermost capsid shell protein, VP2. Structural studies have identified four tunnels that lead to the large cavernous active site of VP1;these tunnels mediate template entry, negative-strand RNA/dsRNA exit, mRNA exit and NTP/PPi substrate exchange. The substrate exchange tunnel of VP1 is bordered by a highly conserved, solvent-exposed flexible loop consisting of 12 amino acids;unresolved in VP1 crystal structures analyzed to date. The role that this flexible loop plays in the activity of VP1 was investigated by structure- and chemical-based mutagenesis. Several attempts to shorten the loop rendered the polymerase insoluble and unsuitable for functional assays. Alanine mutations altering the chemical nature of the loop had little to no effect on the expression and purification of the polymerase, although such mutations decreased the activity of the polymerase relative to wild type. Several lysine residues were identified in the substrate exchange tunnel of VP1, including on the flexible loop, that likely participate in the interchange of negatively charged NTP/PPi moieties. While the NTP/PPi exchange tunnel is a common structural feature of RNA and DNA polymerases, the relationship between the function of the tunnel and polymerase fidelity is poorly understood. We propose that the tunnel contains several charged amino acids in strategic locations that facilitate the NTP/PPi exchange during RNA synthesis. Because the NTP/PPi tunnel affects the rate of nucleotide movement, the tunnel can be expected to also influence the rate of RNA synthesis;an activity likely mirrored by other DNA/RNA polymerases. Therefore, the NTP/PPi exchange tunnel probably not only moderates enzyme pace, but also may have a gatekeeping function that is connected to polymerase fidelity. (2) STRUCTURE AND FUNCTION OF THE RV CAPPING ENZYME, VP3. The RV capping enzyme, VP3, is responsible for capping viral RNA transcripts and has associated guanylyltransferase (GTase), N7 and 2'-O methyltransferase (MTase), and possibly RNA triphosphatase (RTPase) activities. Based on homology modeling, the organization and structure of VP3 domains related to capping are probably similar to those of the bluetongue virus (BTV) capping enzyme, VP4. At the C-terminus of RV VP3 is an additional domain with phosphodiesterase (PDE) activity. This activity hydrolyzes 2'-5'oligoadenylate (2'-5'OA), a product of 2'-5'oligoadenylate synthetase (OAS) that activates the antiviral host enzyme, RNase L. Thus VP3, through its PDE domain, can function as an antagonist of cellular innate immune responses. In the past year, we have successfully expressed, purified, and crystallized a fragment of VP3 that exhibits PDE activity. From this, we have determined the atomic structure of the PDE domain. Through mutagenesis, we have also identified residues critical to the function of PDE domain in 2'-5'OA hydrolysis. (3) STRUCTURE OF GROUP C RV (RVC). RVs are classified into 8 groups (RVA-RVH) based on the antigenic and/or sequence properties of the intermediate capsid protein, VP6. Viruses belonging to the RVA, RVB, and RVC groups are important pathogens in humans and many animal species, including pigs and cows. While the atomic structure of the RVA virion (a triple-layered particle, TLP) has been determined, there is little to no structural information available for the RVB or RVC TLP, or their individual proteins. Such lack of information limits our ability to define antigenic epitopes on RVB and RVC virions that could be useful in identifying distinct virus serotypes and developing vaccines. To gain information for RVC, we used Phyre2-based homology-modeling to build a predicted structure for the RVC TLP (VP2, VP4, VP6, VP7) of the human Bristol strain, relying on the known atomic structure of the RVA TLP. The model, along with sequence alignments, indicates that the exposed surface of the RVC VP7 outer capsid likely includes 1 or 2 putative antigenic loops not present on the RVA VP7 outer capsid. Thus, the location and number of neutralizing epitopes on the RVC TLP may be distinct from those of the RVA TLP. In addition, RVC VP7 proteins appear to have 1 or 2 more surface-exposed N-linked glycosylation sites than RVA VP7 proteins;additional glycosylation may help RVC mask their more elaborate antigenic virion surface. RVC strains grow poorly in cell culture, hindering the generation of high resolution RVC TLP structures by cryo-electron microscopy and crystallography. Our modeled structure may serve as a useful alternative for understanding the architecture of the RVC virion.
Showing the most recent 10 out of 20 publications
