Structural And Functional Dissection Of The Nidovirus Proteins By Reverse Genetics

Nidoviruses are a group of enveloped positive-stranded RNA viruses, which can infect mammals,avian, or invertebrate hosts. The taxon derives its name from the common nidovirus strategy to expressall genes located downstream of the replicase gene from a3’ coterminal nested set of subgenomic (sg)mRNAs. The most important families of order Nidovirales which were also studied extensively containthe Arteriviridae and Coronaviridae.Thus, the representatives from these two families, the PorcineReproductive and Respiratory Syndrome Virus (PRRSV) and Mouse Hepatitis Virus (MHV), werechosen for my study. The structural and functional relationship of the PRRSV nucleocapsid (N) andMHV accessory protein ns2were dissected, and also, the study was extended to a disparate group Arotavirus-encoded VP3protein.PRRSV is belonged to the family Arteriviridae, and its strains are classified into two distinctgenotypes, North American (type II) and European (type I). These two genotypes, which were identifiedalmost simultaneously as the causative agents of what was then known as “blue-ear disease” on twodifferent continents, share only approximately60%nucleotide sequence identity of the overall genome.The nucleocapsid (N) protein, encoded by ORF7, is the most abundant component of the virion. Thenucleotide and predicted amino acid sequences of the N protein are extremely conserved within eachgenotype but show much low identity between genotypes. To dissect the functionality of the N proteinand elucidate the mechanism of virion assembly, inter-genotypic swapping, site-directed mutagenesis, invitro biochemical and in vivo bimolecular fluorescence complementation (BiFC) assay were conductedbased on the infectious cDNA clones of both type II and I PRRSVs that have been constructed in ourlab. The study in details is as follows:Given the low similarity of N proteins from two genotypes, we speculated that the conserveddomans or amino acids may share the same functions. To this end, the possibility of N swappingbetween the genotypes were investigated. Results showed that the ORF7-3’UTR or ORF7of type I wasfunctional in the type II PRRSV genomic background, which lays the foundation for furtherinvestigations of the structure-function relationships of the N protein.The N protein sequence alignment revealed that all type I PRRSV sequences contain an alanineresidue corresponding to position90of the N protein of type II viruses which is an cysteine.Cysteine-to-serine mutational analysis of type II N protein had previously shown that Cys90and Cys23,but not Cys75, are essential for virus infectivity. Based on our chimera study, we hypothesized thatCys90of the type II N protein could be successfully replaced by alanine. Results demonstrated that allof the cysteines in N protein are not essential for infectivity of either PRRSV genotype, and thesecysteine mutants were genetically stable and their growth properties similar to those of the parentalviruses. This is an remarkable finding that challenges previous reports.To investigate the biological significance of the N protein cysteines and resolve this apparentlyconflicting data, we repeated the biochemical work to mimic that conducted in the previous study fortypes II and I PRRSV. Results showed that Cys23in type II and Cys27in type I PRRSV participated in disulfide-linked N protein homodimerization that could be prevented by lysing cells in the presence ofNEM, suggesting that the disulfide bond formation observed had occurred when the N proteins wereexposed to the oxidizing extracellular environment after cell lysis. However, Both the monomeric anddimeric N proteins were observed in the mature virions in the presence of NEM, suggesting thatdisulfide bonds participate in virus particle assembly and may play a role in stabilization of the capsidstructure, but not essential.The fact that disulfide-linked N dimers are present in the mature virions seems to be contradictoryto our genetic evidence, which indicated that all of the cysteines can be knocked out without effect.Thus, the BiFC assay was employed to further study of specific N-N interactions, which can greatlycontribute to our understanding of the N protein dimerization and virus particle assembly. Itdemonstrated that not only intermolecular interactions between the full-length N proteins but alsointeractions among the N-or C-termini as well as possible inter-or intramolecular interactions betweenthe N-and C-terminal regions of N proteins, either with or without cysteine residues, implying animportant role of noncovalent interactions in particle assembly.We have studied the stucture-function relationships of the PRRSV N protein, and our findings willbe of great significance in further understanding the role of N dimerization and the mechanism ofarterivirus virion assembly.After the study on “small” nidovirus structural protein N, our work was extended to the “large”nidovirus, the coronavirus MHV. The accessory protein ns2of MHV is a non-structural proptein andnon-essential for replication in culture cell lines. But it has been shown that it’s required for effecientinfection on bone marrow derived macrophages and liver hepatitis. It is further demonstrated to be a2’,5’-phosphodiesterase (PDE), which has the activity to cleave and inhibit the accumulation of2-5A,preventing activation of RNase L, and thereby blocking the innate immune defense pathway. ns2is amember of Group II (eukaryotic LigT) of the2H-phosphoesterase superfamily, which contain twoHxT/S catalytic motifs. Based on the structural modeling, the VP3(C-terminal domain, VP3-CTD) ofdisparate group A rotavirus (RVA), has the same catalytic motifs. Because this protein has also beenshown to be virulent factor, we hypothesized that RVA VP3contains an antagonist of RNase Lactivation, similar to MHV ns2. Thus, by using the in vitro biochemstry methods and the MHV reversegenetics to construct the chimeric MHVs expressing the VP3-CTD, the structural and functionalsimilarity of ns2and VP3-CTD was investigated. The study in details is as follows:Sequence alignments between ns2and VP3-CTD revealed a pair of HxT/S motifs. To determinewhether these motifs were likely to serve the same function for ns2and the VP3-CTD, the structuremodeling was conducted and the highest confidence models were generated for each molecule. The twocatalytic His residues were located in homologous locations in each structure, suggesting a similar PDEfunction.Amino acid homology and predicted structural similarity suggested that VP3-CTD, like MHV ns2,was capable of degrading2-5A. To verify this, the ns2, VP3-CTD and its mutant were expressed andpurified from E. coli. After incubation with2-5A, both ns2and VP3-CTD can degrade this molecule, except for the VP3-CTD mutant.To determine if VP3-CTD was capable of cleaving2-5A in intact cells, vectors encoding VP3-CTDor its mutant, or ns2, were individually transfected into cells for expression followed by poly(rI):poly(rC)treatment to activate2’,5’-oligoadenylate (2-5A) synthetase (OAS). Data indicated that there wasminimal accumulation of2-5A in cells transfected with VP3-CTD or ns2, but apparent with mutantVP3-CTD.Next, we hypothesized that VP3-CTD could compensate for the loss of ns2function in mutantMHV-infected B6BMM. To test this hypothesis, we constructed a series of mutant or chimeric MHVsusing reverse genetics. Data showed that when wt VP3-CTD was expressed within the background ofns2mutant, it could recover the replication of ns2mutant to wt level in B6BMM, but not for theVP3-CTD. In RNase L-/-BMM, all of recombinant viruses replicated equally well. Moreover, theVP3-CTD expression could inhibit the rRNA degradation in B6BMM which is a consequence ofactivation of the OAS-RNase L pathway.We further test whether the replication of ns2mutant virus in the liver would be restored byexpression of VP3-CTD from the viral genome. As expected, results showed that active VP3-CTD couldenhance the replication of ns2mutant virus in liver of B6mice, but not for the inactive VP3-CTD.When infected the RNase L-/-mice, the inactive VP3-CTD could recover to the active VP3-CTD level.By using the methods of biochemstry and reverse genetics in vitro and in vivo, like the MHV ns2,the RVA VP3-CTD was demonstrated to be a similar2’,5’-phosphodiesterase as MHV ns2, which candirectly cleave the2-5A to prevent the activation of RNase L, and thereby blocking the cellular and viralRNA degradation to antagonize the OAS-RNase L pathway. This finding will advance ourunderstanding of the virus evolutionary mechanisms and lay a fundation for new antiviral drugdiscovery.Taken together, in our study, the nidovirus reverse genetics in combination with the biochemistrymethods were employed to dissect the structural and functional relaitonship of the N protein from the“small” PRRSV and ns2protein from “large” MHV, as well as the VP3from a disparate rotavirus. Thiswill of great significance to understand the mechanisms of virion assembly and the virus evasion of thehost immune system.