Satellite RNA-derived small interfering RNA satsiR-12 targeting the 3' untranslated region of Cucumber mosaic virus triggers viral RNAs for degradation

Abstract

RNA silencing provides protection against RNA viruses by targeting both the helper virus and its satellite RNA (satRNA). Virus-derived small interfering RNAs (vsiRNAs) bound with Argonaute (AGO) proteins are presumed participants in the silencing process. Here, we show that a vsiRNA targeted to virus RNAs triggers the host RNA-dependent RNA polymerase 6 (RDR6)-mediated degradation of viral RNAs. We confirmed that satRNA-derived small interfering RNAs (satsiRNAs) could be associated with different AGO proteins in planta. The most frequently cloned satsiRNA, satsiR-12, was predicted to imperfectly match to Cucumber mosaic virus (CMV) RNAs in the upstream area of the 3' untranslated region (3' UTR). Moreover, an artificial satsiR-12 (asatsiR-12) mediated cleavage of a green fluorescent protein (GFP) sensor construct harboring the satsiR-12 target site. asatsiR-12 also mediated reduction of viral RNAs in 2b-deficient CMV (CMVΔ2b)-infected Nicotiana benthamiana. The reduction was not observed in CMVΔ2b-infected RDR6i plants, in which RDR6 was silenced. Following infection with 2b-containing CMV, the reduction in viral RNAs was not observed in plants of either genotype, indicating that the asatsiR-12-mediated reduction of viral RNAs in the presence of RDR6 was inhibited by the 2b protein. Our results suggest that satsiR-12 targeting the 3' UTR of CMV RNAs triggered RDR6-dependent antiviral silencing. Competition experiments with wild-type CMV RNAs and anti-satsiR-12 mutant RNA1 in the presence of 2b and satRNA demonstrate the inhibitory effect of the 2b protein on the satsiR-12-related degradation of CMV RNAs, revealing a substantial suppressor function of the 2b protein in native CMV infection. Our data provide evidence for the important biological functions of satsiRNAs in homeostatic interactions among the host, virus, and satRNA in the final outcome of viral infection.

Fig. 1.

Analysis of satsiRNA association with AGO-containing complexes. (A) SD-CMV infection symptoms in wild-type Arabidopsis (Col-0) and 6-Myc-AGO1 ago1-27 (6myc-AGO1) transgenic Arabidopsis plants. Photographs were taken at 10 dpi. (B) AGO-containing complexes were immunopurified from total protein extracted from SD-CMV-infected Col-0 and 6myc-AGO1 plants by using specific antibodies as indicated. Total protein extracts incubated with protein A were used as negative controls (−). (C) Small RNA detection of copurified satsiRNAs and miRNAs. Small RNAs were extracted from immunoprecipitates of AGO-containing complexes and protein A controls as indicated at the top. Membranes were hybridized with ³²P-labeled specific oligodeoxynucleotide probes for each satsiRNA and miR-159 and miR-390a. U6 RNA hybridization is shown as an input sample control.

Fig. 2.

Imperfect complementarity between satsiR-12 and the 3′ UTR of SD-CMV RNA and alignment of representative CMV satRNA strains with satsiRNAs. (A) Sequence alignment among satsiR-12 and the SD-CMV RNA1, RNA2, and RNA3 3′ UTRs upstream of the TLS region. 5′CAP represents the 5′-terminal cap structure. The three mismatches between satsiR-12 and SD-CMV RNA1 and RNA3 and four mismatches between satsiR-12 and RNA2 are shown. The differences among the three target regions are indicated in the dashed box. (B) Alignment of representative CMV satRNA isolates/strains with satsiRNAs. Three cloned satsiRNAs (17) mapped to the conserved region (highlighted with a blue background) are shown. Red numbers in parentheses indicate the clone frequencies for each satsiRNA in this conserved region (17). Black numbers at the ends of the sequences indicate nucleotide positions.

Fig. 3.

Analysis of the effect of asatsiR-12 on regulating expression of GFP-3′UT_R1. (A and C) Observation of GFP fluorescence in samples from wild-type N. benthamiana plants (Nb) (A) and RDR6i plants (C) coexpressing GFP-3′UT_R1 with asatsiR-12 or a vector control (−). Photographs were taken under UV light at 2, 3, and 4 dpi. (B and D) Detection of the expression of asatsiR-12 and the accumulation of GFP-3′UT_R1 RNA, 5′ cleavage products, 3′UT_R1-derived siRNAs (siR-3′UT_R1), and GFP-derived siRNAs (siGFP), as well as the accumulation of GFP, in wild-type N. benthamiana plants (B) and RDR6i plants (D). ³²P-labeled GFP DNA probes and GFP-specific antibodies were used. For small RNA blots, ³²P-labeled in vitro transcripts from RNA1 construct 3′UT_R1 and from GFP were used as probes. Methylene blue-stained rRNA, U6 RNA hybridization, and Coomassie blue-stained total proteins are shown as RNA and protein loading controls. Pools of samples collected from four plants were analyzed for each time point. Quantification of GFP-3′UT_R1, 5′ cleavage products, GFP, and siGFP relative to the loading control are shown to the right of each panel using ImageQuant TL (GE Healthcare Life Sciences). +V, in the presence of the control vector. (E) A sketch map for the possible cleavage of GFP sensor RNA mediated by asatsiR-12 and DCLs (18). * indicates an asatsiR-12-mediated 5′ cleavage product, and ** indicates DCL-mediated cleavage products (18).

Fig. 4.

Analysis of the effect of asatsiR-12 on regulating expression of GFP-3′UT_R2. (A and C) and GFP-3′UT_R3 (B and D) in wild-type N. benthamiana plants (A and B) and RDR6i plants (C and D). Photographs were taken under UV light at 2, 3, and 4 dpi. The expression of asatsiR-12 and the accumulation of GFP-3′UT_R2 and GFP-3′UT₃ RNAs, 5′ cleavage products, and 3′UT_R2- and 3′UT_R3-derived siRNAs (siR-3′UT_R2 and siR-3′UT_R3), as well as the expression of GFP, were detected as described in the legend to Fig. 3.

Fig. 5.

Detection of asatsiR-12-mediated cleavage of nucleotide substitution mutant GFP sensors (GFP-3′UT_R2m and GFP-3′UT_R3m) and GUS sensor cleavage. (A) Alignment of 5′-terminal sequences (∼180 bp) of three SD-CMV genomic RNA 3′ UTRs (3′UT) containing the satsiR-12 target site. satsiR-12 and its target region are marked in the red box. Identical and conserved nucleotides are highlighted in blue and yellow, respectively. (B and C) GFP-3′UT_R2m and GFP-3′UT_R3m, in which the satsiR-12 target site in 3′UT_R2 or 3′UT_R3 was replaced by that of 3′UT_R1, were constructed by oligonucleotide-directed mutagenesis. The bases A₅₇ in 3′UT_R2 and U₆₀ in 3′UT_R3 are changed to U and C. Coexpression of GFP-3′UT_R2m (B) or GFP-3′UT_R3m (C) with asatsiR-12 or a vector control (−) in RDR6i plants was examined. GFP fluorescence was photographed under UV light at 2, 3, and 4 dpi. There were no obvious differences in accumulation of GFP-3′UTR_2m, GFP-3′UT_R3m, and their 5′ cleavage products between samples in which the construct was coexpressed with asatsiR-12 and those in which it was coexpressed with the vector control at each time point. (D) Analysis of the asatsiR-12-mediated cleavage of the GUS sensor. The GUS sensor construct contains the exact satsiR-12 target site sequence of 3′UT_R1 (highlighted in black). The arrows indicate the cleavage sites of the GUS sensor RNA detected by 5′ RACE. RDR6i plants were coinfiltrated with the GUS sensor and a vector or the GUS sensor and asatsiR-12, and samples were subjected to RNA blot analysis of GUS RNA at 4 dpi. A ³²P-labeled DNA probe specific for GUS mRNA was used. The 5′ cleavage product is indicated.

Fig. 6.

Analysis of the inhibitory effect of SD-CMV 2b on asatsiR-12-mediated cleavage of GFP-3′UT_R1. (A) Coexpression of GFP-3′UT_R1 with asatsiR-12 (siR-12) or a vector control (V) in the absence or presence of SD-CMV 2b protein in wild-type N. benthamiana plants (left panel) and RDR6i plants (right panel). GFP fluorescence was photographed under UV light at 4 dpi. (B) Expression of asatsiR-12 and accumulation of SD-CMV 2b and GFP-3′UT_R1 RNA and 5′ cleavage products, as well as expression of GFP, were detected as described in the legend to Fig. 3B.

Fig. 7.

Morphology and infection symptoms of transgenic Nbsat-12 and RDR6isat-12 plants carrying 35S–asatsiR-12. (A) Morphology of transgenic Nbsat-12 and RDR6isat-12 T0 plants and detection of asatsiR-12 expression levels in these transgenic plants. The corresponding vector transgenic plants (Nbv and RDR6iv plants) were used as controls. A ³²P-labeled oligodeoxynucleotide probe specific for asatsiR-12 was used. (B) Infection symptoms of Nbsat-12 and RDR6isat-12 plants and corresponding control (Nbv and RDR6iv) plants inoculated with an SD-CMV infectious clone: CMVwt, containing SD-CMV RNA1, RNA2, and RNA3, or CMVΔ2b, containing CMV RNA1, RNA2Δ2b, and RNA3, in which 2b protein expression is abolished (23). Photographs were taken at 4 and 14 dpi, respectively. Leaves of RDR6isat-12 plants with wild-type CMV-related yellowing turned brown (brown arrow) and dried rapidly compared to leaves of the wild-type CMV-infected Nbsat-12 plants (yellow arrow) at 14 dpi.

Fig. 8.

Analysis of the effect of asatsiR-12 and satsiR-12 mutant satRNAm on accumulation of wild-type viral RNAs and anti-satsiR-12 mutant RNA1 (RNA1m) in viral infection. (A and B) Detection of expression of asatsiR-12 and accumulation of CMV RNAs, siR-RNA1, and siR-RNA2 in leaves of Nbsat-12, Nbv, RDR6isat-12, and RDR6iv plants with local CMVwt or CMVΔ2b infection at 4 dpi. siR-RNA1 and siR-RNA2 represent CMV RNA1- and RNA2-derived vsiRNAs respectively. (C) Detection of accumulation of CMV RNAs in wild-type N. benthamiana leaves in systemic CMVΔ2b coinfection with wild-type satRNA and satsiR-12 mutant satRNAm at 4 and 16 dpi. An asterisk denotes a replication form of satRNA. Alignment of satsiR-12 and mutant satsiR-12m with CMV RNA1 3′-UTR RNA is shown under the blot. The satsiR-12/RNA1 3′-UTR pair contains three mismatches in base pairing, whereas there are eight mismatches in the satsiR-12m/RNA1 3′-UTR pair. (D) Detection of accumulation of CMV RNAs and RNA1m at 15 dpi in plants inoculated with a mixture of RNA1m, RNA1, RNA2, and RNA3 with or without satRNA or satRNAm. Membrane for detection of viral RNAs was first hybridized with an RNA1m-specific oligodeoxynucleotide probe (pRNA1m) (left panel) and then stripped and rehybridized with an RNA1-specific oligodeoxynucleotide probe (pRNA1), which can also detect other CMV RNAs (right panel), but not RNA1m. Alignment of satRNA (satR) and mutant satRNA (satRm) with CMV RNA1 and RNA1m 3′-UTR RNA at the satsiR-12 complementation region is shown under the blot. The satR/RNA1m and satRm/RNA1m pairs each contain eight mismatches in base pairing in the 3′ UTRs. Pools of samples collected from four plants were used. Hybridizations were performed with ³²P-labeled oligodeoxynucleotide probes specific for asatsiR-12, satsiR-12m, RNA1, or RNA1m (D) or a mix of probes for ³²P-labeled SD-CMV genomic RNA 3′ UTRs (A to C) or satRNA or ³²P-labeled in vitro transcripts from RNA1 and RNA2. Methylene blue-stained rRNA and the U6 RNA hybridization are shown as loading controls. Quantification of each CMV RNA and siRNA relative to the loading control is shown to the right of each panel.