The specific binding to 21-nt double-stranded RNAs is crucial for the anti-silencing activity of Cucumber vein yellowing virus P1b and perturbs endogenous small RNA populations

Abstract

RNA silencing mediated by siRNAs plays an important role as an anti-viral defense mechanism in plants and other eukaryotic organisms, which is usually counteracted by viral RNA silencing suppressors (RSSs). The ipomovirus Cucumber vein yellowing virus (CVYV) lacks the typical RSS of members of the family Potyviridae, HCPro, which is replaced by an unrelated RSS, P1b. CVYV P1b resembles potyviral HCPro in forming complexes with synthetic siRNAs in vitro. Electrophoretic mobility shift assays showed that P1b, like potyviral HCPro, interacts with double-stranded siRNAs, but is not able to bind single-stranded small RNAs or small DNAs. These assays also showed a preference of CVYV P1b for binding to 21-nt siRNAs, a feature also reported for HCPro. However, these two potyvirid RSSs differ in their requirements of 2-nucleotide (nt) 3' overhangs and 5' terminal phosphoryl groups for siRNA binding. Copurification assays confirmed in vivo P1b-siRNA interactions. We have demonstrated by deep sequencing of small RNA populations interacting in vivo with CVYV P1b that the size preference of P1b for small RNAs of 21 nt also takes place in the plant, and that expression of this RSS causes drastic changes in the endogenous small RNA populations. In addition, a site-directed mutagenesis analysis strongly supported the assumption that P1b-siRNA binding is decisive for the anti-silencing activity of P1b and localized a basic domain involved in the siRNA-binding activity of this protein.

FIGURE 1.

P1b binds specifically double-stranded small RNAs. (A) Increasing amounts of NTAP–P1b purified by affinity chromatography (30, 60, 120, 240, and 480 nM), as well as bovine serum albumin (BSA; 250 nM) or just buffer (-), were incubated with the indicated ³²P-labeled small RNAs. (B) Increasing amounts of NTAP–P1b purified by affinity chromatography (100 and 300 nM), as well as BSA (500 nM) or just buffer (-), were incubated with the indicated ³²P-labeled nucleic acids. Double-stranded molecules had 2-nt 3′ protruding ends. Complexes were resolved in polyacrylamide gels and revealed by autoradiography. Upper and lower arrows indicate bound and free ³²P-labeled probes, respectively. The asterisk indicates the presence of nonspecific shift.

FIGURE 2.

P1b binds ds-siRNAs with size selectivity. (A) Crude protein extracts from N. benthamiana leaves infiltrated with agrobacteria carrying p35S–NTAP–P1b or the empty pBIN19 vector (lane V), and harvested at 6 dpi, were incubated with ³²P-labeled 21-nt ds-siRNAs in the presence of the indicated unlabeled ds-siRNA competitors added in increasing molar excess (10-, 40-, 80-, 160-, 320-, 640-, 1280-, and 2560-fold) or in the absence of them (lanes V and -). Complexes were resolved in polyacrylamide gels and revealed by autoradiography. Upper and lower arrows indicate bound and free ³²P-labeled 21-nt ds-siRNAs, respectively. (B) Densitometric analysis of the autoradiographic signals. The ratio bound RNA/total RNA of each lane was plotted as a function of logarithm of molar excess of competitors. All ds-siRNAs had 2-nt 3′ protruding ends.

FIGURE 3.

P1b and HCPro display different structural requirements for efficient siRNA recognition. Crude protein extracts from N. benthamiana leaves infiltrated with agrobacteria carrying p35S–NTAP–P1b (A,C), p35S-P1HC (B,D), or the empty pBIN19 vector (lanes V), and harvested at 6 dpi, were incubated with ³²P-labeled 21-nt ds-siRNAs in the absence (lanes V and -) or in the presence of increasing molar excesses (A, 10-, 40-, 80-, 160-, 320-, 640-, 1280- and 2560-fold; B, 100-, 500-, 2000-, 8000-, 16,000-, 36,000-, and 72,000-fold; C,D, 100-, 500-, 2000-, 8000-, 16,000- and 32,000-fold) of unlabeled competitors: nonphosphorylated 21-nt ds-siRNAs with 2-nt 3′ protruding ends (black triangles, A–D); nonphosphorylated blunt-ended 19-nt ds-siRNAs (dark gray triangles, A,B); phosphorylated 21-nt ds-siRNAs with 2-nt 3′ protruding ends (light gray triangles, C,D). Complexes were resolved in polyacrylamide gels and revealed by autoradiography (left panels). Upper and lower arrows indicate bound and free ³²P-labeled ds-siRNAs, respectively. For all the competition experiments, densitometric analyses of the autoradiographic signals are shown in the right panels. The ratio bound RNA/total RNA of each lane was plotted as a function of logarithm of molar excess of competitors.

FIGURE 4.

P1b binds siRNAs in vivo. N. benthamiana plants were co-infiltrated with agrobacteria carrying p35S:GFP and p35S:GF-IR plus p35S–NTAP–P1b or p35S–NTAP–P1b RK68,69AA. The infiltrated leaves were harvested at 6 dpi. (A) Northern blot analysis of GFP mRNA extracted from infiltrated leaves. Agarose gel stained with ethidium bromide is shown as loading control (rRNA). (B) Small RNAs separated by polyacrylamide gel electrophoresis and stained with ethidium bromide (Small nucleic acids) and GFP-specific siRNAs detected by Northern blot analysis (GFP siRNAs) from either infiltrated leaves (Inputs) or NTAP–P1b, wild-type or RK68,69AA mutant, purified by affinity chromatography (Pull down). Numbers above each line (30X, 5X, 2.5X) indicate the times of enrichment relative to Inputs. Bands corresponding to 21 nt/22 nt, and 24 nt, are indicated. 5S rRNA and tRNA stained with ethidium bromide are shown as a loading control. The samples of purified NTAP–P1b proteins used for the siRNA extraction were subjected to SDS-polyacrylamide gel electrophoresis to assess their protein amounts (NTAP–P1b).

FIGURE 5.

Enrichment in 21-nt small RNAs associated to NTAP–P1b expression. (A) Flowchart followed for the analysis of small RNA sequences obtained by deep sequencing of small RNAs from N. benthamiana (Nb), N. benthamiana infiltrated with agrobacteria carrying p35S–NTAP–P1b (Nb + P1b), or a fraction of NTAP–P1b purified by affinity chromatography from these plants (CoP–P1b). The number of reads passing each step and the percentage with respect to the previous step are shown for each sample. (B) Size-distribution histograms of endogenous (planta) small RNAs of each sample, expressed as percentage with respect to the planta small RNA population. (C) Size-distribution histograms of plasmid-derived (plasmid) small RNAs of each sample, expressed as percentage with respect to the plasmid small RNA population.

FIGURE 6.

Basic domains in P1b-like proteins. (A) Schematic representation of CVYV P1b showing the location of conserved domains and a plot of charge density along the protein. (B) Details of two basic domains placed at the N-terminal half of CVYV P1b. An amino acid alignment of the second domain, which is partially conserved in P1b-like proteins, is shown. Proteins included in the alignment are as follows: P1b from the ipomoviruses CVYV and Squash vein yellowing virus (SqVYV), and P1 from the ipomoviruses Cassava brown streak virus (CBSV) and Sweet potato mild mottle virus (SPMMV), from the tritimoviruses Wheat streak mosaic virus (WSMV), Oat necrotic mottle virus (ONMV), Wheat eqlid mosaic virus (WEqMV) and Brome streak mosaic virus (BStMV), and from the still unclassified potyvirid Triticum mosaic virus (TriMV), Sugarcane streak mosaic virus (SCSMV), and Blackberry virus Y (BVY). Boxed amino acids are identical or chemically similar between at least two ipomoviruses (green boxes), between at least three nonipomoviruses (yellow boxes), or between at least two ipomoviruses plus two nonipomovirus (black boxes). Dashes represent gaps. The position of the first amino acid of each aligned segment is indicated on the left side of the sequence. The asterisks indicate the residues of CVYV P1b that were mutated in this work.

FIGURE 7.

Silencing suppression activity and siRNA binding capacity of P1b mutant proteins. N. benthamiana plants were co-infiltrated with agrobacteria carrying p35S:GFP and p35S:GF-IR plus empty pBIN19 plasmid (Vector), wild-type p35S–NTAP–P1b or derivatives of this plasmid with the indicated mutations. The infiltrated leaves were harvested at 4 dpi. (A) GFP fluorescence pictures taken under a fluorescence stereomicroscope. (B) Northern blot analysis of GFP mRNA extracted from infiltrated leaves. Agarose gel stained with ethidium bromide is shown as loading control (rRNA). (C) siRNA binding analysis by EMSA. Crude protein extracts of infiltrated leaves (three doses: 1, 3, and 9 μL) were incubated with ³²P-labeled 21-nt ds-siRNAs. Complexes were resolved by polyacrylamide gel electrophoresis and revealed by autoradiography. The amount of NTAP–P1b proteins present in each extract was estimated by Western blot analysis with Peroxidase anti-Peroxidase complex (on top of the panel). Upper and lower arrows indicate bound and free ³²P-labeled 21-nt ds-siRNAs, respectively. (D) Northern blot analysis of both GF- and P-derived siRNAs from either co-infiltrated leaves (Inputs) or NTAP–P1b proteins purified by affinity chromatography (Pull down). Pull down samples are enriched 20X (WT), 30X (K61A), 30X (R68L), 50X (R68A), 35X (K69A), and 50X (RK68,69AA) relative to the Inputs. Bands corresponding to 21 nt/22 nt, and 24 nt, are indicated. 5S rRNA and tRNA stained with ethidium bromide are shown as loading control. The samples of purified NTAP–P1b proteins used for the siRNA extraction were subjected to SDS-polyacrylamide gel electrophoresis to assess their protein amounts (NTAP–P1b).