Detection of Cereal yellow dwarf virus using small interfering RNAs and enhanced infection rate with Cocksfoot streak virus in wild cocksfoot grass (Dactylis glomerata).

Small RNA sequences were obtained from leaf extracts of wild Dactylis glomerata (cocksfoot grass) using deep sequencing (454 Life Sciences, Roche Diagnostics), and were screened against virus sequences in GenBank using a local BLASTn search program (BioEdit). Putative small interfering (si)RNAs complementary in sequence to Cereal yellow dwarf virus (CYDV, genus Luteovirus) genomes were identified. Primer sequences were made against the "high scoring" siRNA sequences and RT-PCR was used to amplify a 438 bp CYDV fragment in total RNA extracts from D. glomerata leaves. Sequencing of the RT-PCR product confirmed the occurrence of a previously undescribed CYDV population with phylogenetic affinity to CYDV-RPS. In D. glomerata the CYDV infection rates were 42.3% (n=78) in 2008 and 50.0% (n=48) in 2009. Specific RT-PCR tests also showed that this D. glomerata population harboured Cocksfoot streak virus (CSV, genus Potyvirus). Dual infections by these viruses were observed in 20.5-22.9% of all plants tested in 2008-2009. Interestingly, infections of either CYDV or CSV enhanced the occurrence of the other virus in individual grasses. Opportunities are discussed for using siRNA sequencing approaches in virus survey and other ecology studies under field conditions.

Cocksfoot mottle sobemovirus coat protein contains two nuclear localization signals.

Cocksfoot mottle virus (CfMV) coat protein (CP) localization was studied in plant and mammalian cells. Fusion of the full-length CP with enhanced green fluorescent protein (EGFP) localized to the cell nucleus whereas similar constructs lacking the first 33 N-terminal amino acids of CP localized to the cytoplasm. CP and EGFP fusions containing mutations in the arginine-rich motif of CP localized to the cytoplasm and to the nucleus in plant cells indicating the involvement of the motif in nuclear localization. In mammalian cells, mutations in the arginine-rich region were sufficient to completely abolish nuclear transport. The analysis of deletions of amino acid residues 1-11, 1-22, and 22-33 of CP demonstrated that there were two separate nuclear localization signals (NLS) within the N-terminus--a strong NLS1 in the arginine-rich region (residues 22-33) and a weaker NLS2 within residues 1-22. Analysis of point mutants revealed that the basic amino acid residues in the region of the two NLSs were individually not sufficient to direct CP to the nucleus. Additional microinjection studies with fluorescently labeled RNA and CP purified from CfMV particles demonstrated that the wild-type CP was capable of transporting the RNA to the nucleus. This feature was not sequence-specific in transient assays since both CfMV and GFP mRNA were transported to the cell nucleus by CfMV CP. Together the results suggest that the nucleus may be involved in CfMV infection.

Nucleotide bias of DCL and AGO in plant anti-virus gene silencing.

Plant Dicer-like (DCL) and Argonaute (AGO) are the key enzymes involved in anti-virus post-transcriptional gene silencing (AV-PTGS). Here we show that AV-PTGS exhibited nucleotide preference by calculating a relative AV-PTGS efficiency on processing viral RNA substrates. In comparison with genome sequences of dicot-infecting Turnip mosaic virus (TuMV) and monocot-infecting Cocksfoot streak virus (CSV), viral-derived small interfering RNAs (vsiRNAs) displayed positive correlations between AV-PTGS efficiency and G+C content (GC%). Further investigations on nucleotide contents revealed that the vsiRNA populations had G-biases. This finding was further supported by our analyses of previously reported vsiRNA populations in diverse plant-virus associations, and AGO associated Arabidopsis endogenous siRNA populations, indicating that plant AGOs operated with G-preference. We further propose a hypothesis that AV-PTGS imposes selection pressure(s) on the evolution of plant viruses. This hypothesis was supported when potyvirus genomes were analysed for evidence of GC elimination, suggesting that plant virus evolution to have low GC% genomes would have a unique function, which is to reduce the host AV-PTGS attack during infections.

Proof of concept pilot study: prevalence of grass virus infection and the potential for effects on the allergenic potency of pollen.

BACKGROUND: Wild plants harbour a variety of viruses and these have the potential to alter the composition of pollen. The potential consequences of virus infection of grasses on pollen-induced allergic disease are not known. METHODS: We have collected pollen from Dactylis glomerata (cocksfoot; a grass species implicated as a trigger of allergic rhino-conjunctivitis) from Wytham Wood, Oxfordshire UK. Extracts were prepared from pollen from uninfected grass, and from grass naturally infected by the Cocksfoot streak potyvirus (CSV). Preparations of pollen from virus-infected and non-infected grasses were employed in skin testing 15 grass pollen-allergic subjects with hayfever. Allergen profiles of extracts were investigated by Western blotting for IgE with sera from allergic subjects. RESULTS: The prevalence of CSV infection in cocksfoot grasses sampled from the study site varied significantly over an eight-year period, but infection rates of up to 70% were detected. Virus infection was associated with small alterations in the quantities of pollen proteins detected by polyacrylamide gel electrophoresis, and in the patterns of allergens identified by Western blotting with IgE from grass pollen allergic subjects. For individual subjects there were differences in potencies of standardised extracts of pollen from virus-free and virus-infected plants as assessed by skin testing, though a consistent pattern was not established for the group of 15 subjects. CONCLUSION: Infection rates for CSV in cocksfoot grass can be high, though variable. Virus-induced alterations in components of grass pollen have the potential to alter the allergenic potency.

The three-dimensional structure of cocksfoot mottle virus at 2.7 A resolution.

Cocksfoot mottle virus is a plant virus that belongs to the genus Sobemovirus. The structure of the virus has been determined at 2.7 A resolution. The icosahedral capsid has T = 3 quasisymmetry and 180 copies of the coat protein. Except for a couple of stacked bases, the viral RNA is not visible in the electron density map. The coat protein has a jelly-roll beta-sandwich fold and its conformation is very similar to that of other sobemoviruses and tobacco necrosis virus. The N-terminal arm of one of the three quasiequivalent subunits is partly ordered and follows the same path in the capsid as the arm in rice yellow mottle virus, another sobemovirus. In other sobemoviruses, the ordered arm follows a different path, but in both cases the arms from three subunits meet and form a similar structure at a threefold axis. A comparison of the structures and sequences of viruses in this family shows that the only conserved parts of the protein-protein interfaces are those that form binding sites for calcium ions. Still, the relative orientations and position of the subunits are maintained.