Fig 1.
CMV CP negatively regulates viral virulence in newly emerging Nicotiana benthamiana leaves.
(A) Schematic representations of pCass-RNA1, pCass-RNA2, and pCass-RNA3 for agroinfiltration. The full-length cDNAs of the CMV Fny strains genomic RNA1, RNA2, and RNA3 were inserted independently into pCass4-RZ between the double CaMV 35S promoter and a ribozyme sequence (Rz) derived from tobacco ringspot virus satellite RNA. The CPWT and CPRA harboring viruses were designated as CMVWT and CMVRA, respectively. (B) Mild and severe symptoms elicited by CMVWT and CMVRA, respectively, in newly emerging N. benthamiana leaves at 7 dpi. (C) Severe leaf distortion, stunting and delayed emergence of CMVRA infected leaves compared with mild mosaic symptoms of CMVWT at 14- and 21- dpi. Note: Compare size differences with Mock inoculated plants (D) Comparison of loss of apical dominance of CMVRA versus CMVWT infected plants at 42 dpi.
Fig 2.
Distribution of CMVRA or CMVWT in the apical meristems of infected plants.
In situ hybridization of longitudinal sections of shoot apices in N. benthamiana plants inoculated with CMVRA and CMVWT at 7-, 14-, and 21- dpi. Dark areas indicate the presence of viral RNA revealed by digoxigenin-labeled riboprobe corresponding to the CMV CP. CMV signals were not observed in the shoot meristem from uninfected samples (left panel). Only a low signal density was present in CMVWT infected meristems at 7 dpi and these disappeared from the meristem by 14 dpi (middle panels). In contrast, CMVRA infection contained abundant signal densities beneath the meristems at 7 dpi, and indicated partially invaded meristems by 14 dpi, and completely invaded meristems at 21 dpi (right panels). Bars = 100 μm.
Fig 3.
CMVWT induces more potent antiviral silencing than CMVRA in emerging N. benthamiana leaves.
(A) Accumulation of CMV gRNAs and sgRNAs and (B) RNA3-vsiRNAs in emerging N. benthamiana leaves at 7 dpi with CMVRA and CMVWT. Loading controls for the high and low molecular weight RNAs were rRNA and U6 RNAs, respectively. (C) Ratios of accumulation levels of CMV gRNA3 and RNA3-vsiRNAs calculated from signal intensities in three independent hybridization experiments. The vsiRNAs/RNA3 values refer to the relative ratios of RNA3-vsiRNAs versus viral genomic RNA3. The values of RNA3 and RNA3-vsiRNAs in CMVRA-infected leaves were set as 1. ** P-value < 0.01; *** P-value < 0.001.
Fig 4.
CMVWT induced more potent RDR6/SGS3-dependent antiviral silencing than CMVRA in newly emerging tissues of infected Arabidopsis thaliana.
(A) CMVWT and CMVRA symptoms in newly emerging tissues of Col-0, sgs3, and rdr6 mutant plants at 21 dpi. CMVWT caused severe symptoms in emerging tissues of sgs3 and rdr6 mutant plants, but not in those of Col-0 plants. In contrast, CMVRA caused similar severe symptoms in Col-0, sgs3, and rdr6 plants. (B) Accumulation of CMV gRNAs/sgRNAs and RNA3-vsiRNAs in emerging leaves of A. thaliana plants infected with CMVRA or CMVWT at 21 dpi. Loading controls for the high and low molecular weight RNAs were rRNA and U6 RNAs, respectively. (C) Comparison of CMVWT and CMVRA symptoms Col-0, sgs3 and rdr6 plants at 56 dpi. CMVWT elicited mild mosaic and lack of stunting in Col-0 plants, whereas CMVRA plants were severely stunted, had reduced apical dominance and bolted much later than CMVWT plants. Both sgs3 and rdr6 plants developed severe symptoms after infection with CMVRA and CMVWT. (D) Accumulation of CMV CP and 2b protein in the A. thaliana emerging leaves shown in panel A. Anti-CP, and -2b polyclonal antibody were used to detect the accumulation of CP and 2b protein accumulation, respectively. Coomassie brilliant blue (CBB) staining was used as the protein loading control.
Fig 5.
CMV CP attenuates 2b-mediated suppression of local GFP silencing.
(A) GFP fluorescence in regions of N. benthamiana leaves after agroinfiltration of sGFP reporter gene (OD600 = 0.4), in combination with different amounts of the pGD empty vector (V, OD600 = 0.4), the pGD-2b (OD600 = 0.2) and the pGD-CP vectors (OD600 = 0.4), as indicated. Photographs were taken under UV light at 5 dpi. (B) Protein gel blot analysis of samples extracted from infiltrated region shown in panel A. (C) Accumulation of GFP mRNA and siRNAs in the region shown in panel A. Methylene blue-stained rRNA and U6 RNA were used as loading controls for high and low molecular weight RNAs, respectively. The values under GFP mRNA (RA1) and siRNAs (RA2) represent the relative accumulation (RA) of GFP mRNA and GFP-derived siRNAs, respectively. The RA values of sGFP with 2b and V were set as 1. RA2/RA1 ratios under U6 detection represent the relative production of GFP-derived siRNAs versus GFP mRNA. (D) GFP fluorescence (left panel) in local patches agroinfiltrated with the sGFP vector (OD600 = 0.4), empty pGD vector, and the 2b vectors (OD600 = 0.2), either alone or in combination with CPWT or CPRA vectors (OD600 = 0–0.8), as indicated in middle panel. A protein gel blot analysis of samples extracted from the infiltrated regions is shown in the right panel. Anti-GFP, -CP, and -2b polyclonal antibodies were used to assess the GFP, CP, and 2b, accumulations, respectively. Mock-infected plants were used as the negative control. Coomassie brilliant blue (CBB) staining was used as the protein loading control.
Fig 6.
RNA binding and translation inhibition by the CPWT and CPRA proteins.
(A) RNA-binding abilities of the GST-CPWT and GST-CPRA proteins as assessed by digoxigenin-labeled CMV RNA4 or Luciferase (Luc) mRNA (See Materials and methods for details). GST served as a negative control. Coomassie brilliant blue (CBB) staining was used as the protein loading control. (B) Translation inhibition of Luciferase mRNA by CPWT and CPRA in vitro. The Luc mRNA was transcribed and incubated with different concentrations of the GST, GST-CPRA, or GST-CPWT in wheat germ extract at 25°C for two hours. Then, the activity of translated Luciferase in vitro was measured with a luminometer. Luciferase activities from mRNA without GST, GST-CPRA, or GST-CPWT was set as 100%. Error bars represent the standard error of the mean. Data points are the mean value of three independent experiments. *P-value < 0.05; ** P-value < 0.01; *** P-value < 0.001.
Fig 7.
CMV CP is associated with RDR6/SGS3 complex.
(A) Interactions detection of SGS3 with CPWT or CPRA in N. benthamiana leaves epidermal cells using bimolecular fluorescence complementation (BiFC). Leaves were infiltrated with pairs of Agrobacterium strains harboring tested proteins tagged with the different halves of YFP. Rubisco protein (Rub) was negative control. Images were taken at 2 dpi by confocal scanning laser microscopy. Bars represent 20 μm. (B) Co-immunoprecipitation (Co-IP) assays showing associations of SGS3 with CPs in vivo. N. benthamiana leaves coexpressing Flag-CPWT and Flag-CPRA proteins in combination with the GFP-SGS3 protein were homogenized and incubated with anti-Flag M2 antibodies for co-IP assays. GFP was included as a negative control. Anti-GFP and anti-Flag antibodies were used to detect accumulations of GFP-SGS3/GFP and CP, respectively. The positions of GFP-SGS3 and GFP are indicated by black and white triangles, respectively.
Fig 8.
A model explaining the CMV CP-modulated conflict between RNA silencing and 2b-induced suppression in the shoot apices infected with CMVWT.
During the early infection stages of CMVWT, levels of CPWT accumulation are not sufficient to affect accumulation of 2b protein and antiviral silencing. However, at intermediate stages of replication, higher CPWT levels accumulate and bind to viral RNAs to result in virion assembly, reduced 2b protein accumulation and protection of aberrant RNAs from degradation. The CPWT and bounded RNA complex are postulated to recruit RDR/SGS3 for amplification of vsiRNAs that increase antiviral RNA silencing and antagonize 2b protein suppression of host RNA silencing. In summary, CPWT RNA binding inhibits the translation and accumulation of the 2b protein to favor RNA silencing that contributes to viral self-attenuation and long-term symptom recovery.