Nuclear-cytoplasmic partitioning of cucumber mosaic virus protein 2b determines the balance between its roles as a virulence determinant and an RNA-silencing suppressor

Abstract

The Cucumber Mosaic Virus (CMV) 2b protein is an RNA-silencing suppressor that plays roles in CMV accumulation and virulence. The 2b proteins of subgroup IA CMV strains partition between the nucleus and cytoplasm, but the biological significance of this is uncertain. We fused an additional nuclear localization signal (NLS) to the 2b protein of subgroup IA strain Fny-CMV to create 2b-NLS and tested its effects on subcellular distribution, silencing, and virulence. The additional NLS enhanced 2b protein nuclear and nucleolar accumulation, but nuclear and nucleolar enrichment correlated with markedly diminished silencing suppressor activity in patch assays and abolished 2b protein-mediated disruption of microRNA activity in transgenic Arabidopsis. Nucleus/nucleolus-localized 2b protein possesses at least some ability to inhibit antiviral silencing, but this was not sufficient to prevent recovery from disease in younger, developing leaves in Arabidopsis. However, enhanced nuclear and nucleolar accumulation of 2b increased virulence and accelerated symptom appearance in older leaves. Experiments with Arabidopsis lines carrying mutant Dicer-like alleles demonstrated that compromised suppressor activity explained the diminished ability of 2b-NLS to enhance virus accumulation. Remarkably, the increased virulence that 2b-NLS engendered was unrelated to effects on microRNA- or short interfering RNA-regulated host functions. Thus, although nucleus- and nucleolus-localized 2b protein is less efficient at silencing suppression than cytoplasm-localized 2b, it enhances CMV virulence. We propose that partitioning of the 2b protein between the cytoplasmic and nuclear/nucleolar compartments allows CMV to regulate the balance between virus accumulation and damage to the host, presumably to maximize the benefit for the virus.

Importance: In this work, the main finding is that nucleus/nucleolus-localized 2b protein is strongly associated with CMV virulence, which is independent of its effect on small RNA pathways. Moreover, this work supports the contention that the silencing suppressor activity of CMV 2b protein is predominantly exerted by that portion of the 2b protein residing in the cytoplasm. Thus, we propose that partitioning of the 2b protein between the cytoplasmic and nuclear/nucleolar compartments allows CMV to regulate the balance between virus accumulation and damage to the host, presumably to maximize the benefit for the virus.

FIG 1

Subcellular distributions of the 2b proteins from subgroup IA strain Fny-CMV and subgroup II strain LS-CMV. Leaf epidermal cells of Nicotiana benthamiana in tissue agroinfiltrated to transiently express free EGFP, EGFP-Fny2b, or EGFP-LS2b were imaged by confocal scanning laser microscopy at 5 days postinfiltration. Bars, 200 μm (upper panels) and 10 μm (lower panels at higher magnification).

FIG 2

Subcellular distributions of Fny2b and its variants. (A) Schematic diagram of LS2b, Fny2b, and Fny2b-derivative proteins. The two NLS sequences (NLS1 and NLS2) are indicated in blue for Fny2b and in red for LS2b. Mutations to the NLS sequences in Fny2b are shown in red. Attachment of an NLS sequence (KRRRRR) or six histidine residues to the carboxyl end of Fny2b created the variant Fny2bNLS or Fny2b6His, respectively. (B to D) Transient expression of EGFP fusion proteins of Fny2b and its variants in N. benthamiana by agroinfiltration. Epidermal cells in infiltrated leaves were imaged at 5 days postinfiltration by confocal scanning laser microscopy. (B) Cytoplasmic and nuclear localization. (C) Magnified views of nuclei with nucleoli indicated by arrows. (D) Quantification of GFP fluorescence in nuclei and nucleoli as the mean value for 30 cells for each protein. Fluorescence intensities in nuclei and nucleoli were measured using the software LAS AF Lite (Leica). Bars represent standard errors of the means.

FIG 3

Effects of Fny2b and its variants on suppression of local RNA silencing. The binary vector pBI121-derived plasmid p35S:GFP expressing the reporter gene modified green fluorescence protein (mgfp5) was agroinfiltrated together with binary vectors expressing EGFP-Fny2b, its variants, or control proteins GUS and EGFP as indicated. (A) Infiltration patches photographed under UV light at 5 days postinfiltration. (B and C) Immunoblot analyses of accumulation of GFP and EGFP-fused 2b fusion proteins in agroinfiltrated patches at 5 days postinfiltration. A polyclonal anti-GFP serum was used to detect GFP and EGFP-fused 2b proteins. Ponceau S staining was used to monitor the equivalence of protein loading and transfer. (D) RNA gel blot analyses of steady-state accumulation of GFP and EGFP-2b transcripts in the agroinfiltrated patches. Total RNA was isolated from the agroinfiltrated patches at 5 days postinfiltration. GFP and EGFP-2b transcripts were detected using a digoxigenin-labeled GFP probe, which was made using the DIG high primer DNA labeling system with full-length GFP cDNA as the template. Equal loading was confirmed by staining of rRNA with ethidium bromide.

FIG 4

Translocation of Fny2b into the nucleus diminished its ability to disrupt microRNA functions in transgenic Arabidopsis. EGFP- or EGFP-2b-transgenic Arabidopsis plants were generated by transformation of binary vector pBI121-derived plasmids expressing free EGFP, EGFP-Fny2b, EGFP-Fny2bNLS (attachment of LS2b NLS to the C terminus of Fny2b), and EGFP-Fny2b6His (attachment of six histidine residues to the C terminus of Fny2b). (A) Typical phenotypes of EGFP- and EGFP-2b-transgenic lines at the cotyledon stage. EGFP-transgenic plants showed normal cotyledon morphology. EGFP-Fny2b-transgenic plants exhibited elongated and/or severely stunted cotyledons. EGFP-Fny2b6His-transgenic plants also exhibited elongated and severely unexpanded cotyledons. EGFP-Fny2bNLS-transgenic plants showed normal cotyledon development. (B) Typical phenotypes of EGFP- and EGFP-2b-transgenic lines in older plants (approximately 45 days after germination). The terms in parentheses indicate the line designations. (C) Detection of constitutive expression of EGFP and EGFP-fused 2b proteins in the transgenic lines using anti-GFP and anti-2b sera. (D) Confirmation of constitutive expression of EGFP or EGFP-2b proteins in the transgenic plants by observation of GFP fluorescence using a confocal microscope. Bar, 50 μm. (E) Northern blot analyses of accumulation of mature microRNAs and their star strands in transgenic plants as indicated. Total RNA was extracted from pooled aerial parts from five individual plants. U6 RNA was used as a loading control. (F) RT-qPCR analyses of accumulation in the transgenic lines of transcripts targeted by the microRNAs examined in panel E. PHABULOSA (PHB), AUXIN RESPONSE FACTOR 8 (ARF8), and ARGONAUTE 1 (AGO1) are targets of miR165/166, miR167, and miR168, respectively. ARABIDOPSIS ELONGATION FACTOR 1α (EF1α) was used as a reference transcript. The analysis of variance (ANOVA) Duncan's new multiple range test was used to analyze data for statistically significant differences. Different letters are assigned to results with statistically significantly different values (P < 0.05), while results assigned the same letter are not significantly different. Bars represent standard errors of the means.

FIG 5

Modification of the Fny2b protein by addition of either an additional NLS or a hexahistidine sequence did not compromise binding to double-stranded small RNAs. (A) SDS-PAGE analysis of GST and GST-2b fusion proteins synthesized in E. coli and purified by affinity chromatography confirmed successful expression of GST-Fny2b, GST-Fny2bNLS, and GST-Fny2b6His. Protein bands were visualized by staining with Coomassie blue. (Ladder) indicates a lane loaded with protein molecular mass markers. (B) Electrophoretic mobility shift assays for assessing the abilities of 2b protein variants to bind miRNA and siRNA duplexes. Biotin-labeled miRNA and siRNA duplexes (ds-miRNA and ds-siRNA) were produced by annealing Arabidopsis miR168 sequence-based, biotin-labeled sRNA-1 with either miR168 star sequence-based sRNA-2 or with sRNA-3, respectively. ds-miRNA or ds-siRNA (5 pmol) was incubated with 1 μg purified GST or GST-tagged 2b proteins, resolved by 8% acrylamide native PAGE, and blotted onto a nylon membrane. Biotin-labeled RNAs were detected using horseradish peroxidase conjugated to streptavidin. The positions of bands corresponding to free and protein-bound (Bound) RNAs are indicated.

FIG 6

Increased translocation of Fny2b into the nucleus enhanced virus virulence but did not increase virus accumulation. (A) Viral symptoms in wild-type Arabidopsis infected with Fny-CMV, Fny-CMV2bNLS, and Fny-CMV2b6His. Water-inoculated plants are indicated (Mock). Arrowheads indicate inoculated leaves. Photographs were taken at 7 dpi and 14 dpi as indicated on the right. (B and C) Northern blot analyses of accumulation of viral progeny RNAs in the plants as shown in panel A and the plants infected with Fny-CMVΔ2bpro, which expresses no 2b protein at all. Total RNA was extracted from pooled aerial parts from five individual plants at 7 dpi and from pooled top leaves from five individual plants at 14 dpi. A biotin-labeled DNA oligonucleotide complementary to the highly conserved sequence of the 3′ untranscribed region (UTR) of the CMV genomic and subgenomic RNAs was used to detect viral RNA. Equal loading was confirmed by ethidium bromide staining of rRNA. (D) Northern blot analyses of accumulation of viral siRNAs. Total RNA was extracted from plant samples as described in panels B and C. A mixture of 8 DIG-labeled DNA oligonucleotides corresponding to CMV RNA3 was used to probe viral siRNAs. U6 RNA was used as a loading control.

FIG 7

Fny-CMV2bNLS was rescued in Arabidopsis dcl2-1/dcl4-2 double mutant plants and killed the mutant hosts. (A) Disease symptoms on A. thaliana wild-type Col-0 (WT) and dcl2-1/dcl4-2 double mutant plants inoculated with Fny-CMV, Fny-CMV2b6His, Fny-CMV2bNLS, or water (Mock). Photographs were taken at 14 dpi and 30 dpi as indicated on the right. (B) Immunoblot analysis of CMV CP accumulation in the upper, still-living leaves at 14 dpi. CP was detected using a polyclonal serum against CMV CP. Ponceau S staining was used to monitor the equivalence of protein loading and transfer.

FIG 8

Systemic necrosis caused by infection with the mutant Fny-CMV2bNLS is unrelated to effects on small RNA pathways. Arabidopsis wild-type (WT) and mutants deficient in microRNA biogenesis (dcl1-9) or siRNA biogenesis (dcl2-1/dcl4-2 and dcl3-1) were inoculated with Fny-CMV, Fny-CMV2bNLS, or Fny-CMV2b6His. Mock, plants were mock inoculated with sterile water. Plants were photographed at 14 dpi with the exception of dcl1-9 mutant plants, which were imaged at 10 dpi.