Viral RNA silencing suppressors inhibit the microRNA pathway at an intermediate step

Abstract

RNA silencing suppressors from different plant viruses are structurally diverse. In addition to inhibiting the antiviral silencing response to condition susceptibility, many suppressors are pathogenicity factors that cause disease or developmental abnormalities. Here, unrelated suppressors from multiple viruses were shown to inhibit microRNA (miRNA) activities and trigger an overlapping series of severe developmental defects in transgenic Arabidopsis thaliana. This suggests that interference with miRNA-directed processes may be a general feature contributing to pathogenicity of many viruses. A normally labile intermediate in the miRNA biogenesis/RNA-induced silencing complex (RISC) assembly pathway, miRNA*, accumulated specifically in the presence of suppressors (P1/HC-Pro, p21, or p19) that inhibited miRNA-guided cleavage of target mRNAs. Both p21 and p19, but not P1/HC-Pro, interacted with miRNA/miRNA* complexes and hairpin RNA-derived short interfering RNAs (siRNAs) in vivo. In addition, p21 bound to synthetic miRNA/miRNA* and siRNA duplexes in vitro. We propose that several different suppressors act by distinct mechanisms to inhibit the incorporation of small RNAs into active RISCs.

Figure 1.

Developmental defects induced by silencing suppressors from five viruses. (A) Diagram of constructs used for expression of HA epitope-tagged viral suppressor proteins. Each construct contained the gene for one of the suppressors shown. (P_35S) 35S promoter from Cauliflower mosaic virus; (TL) translational leader element from Tobacco etch virus; (T_35S) 35S terminator from Cauliflower mosaic virus. (B) Expression of epitope-tagged silencing suppressors in transgenic Arabidopsis plants. Immunoblot analysis was done by using total protein samples (amounts shown) from plants transformed with empty vector (V) or constructs encoding P1/HC-Pro (HC), p21, 2b, CP, or p19. Mobility positions of 16–62-kDa protein standards are shown. Note that HC-Pro migrates as ∼50-kDa protein, but that breakdown fragments migrate at ∼25-kDa and ∼35-kDa positions. (C) Effects of silencing suppressors on leaf and rosette morphogenesis. Bars, 10 mm. (D) Effects of silencing suppressors on flowers (stage 11–12). Bars, 1 mm.

Figure 2.

Blot analysis of miRNA targets in transgenic Arabidopsis plants. RNA samples from vector-transformed (lanes 1,2) and suppressor-expressing transgenic Arabidopsis plants (lanes 3–12) were analyzed in duplicate by hybridization with DNA probes. Ethidium bromide-stained 28S rRNA is shown for each blot. Mobility positions of relevant RNA size standards (kb) are shown. (A) Accumulation of ARF8 (At 5 g 37020) mRNA. Mean relative accumulation (RA) of ARF8 mRNA in suppressor expressing plants was calculated relative to that in vector-transformed plants, normalized against the accumulation of the control TyrAT (At 2 g 20610) mRNA. (B) Accumulation of ARF10 (At 2 g 28350) mRNA. Mean RA of ARF10 mRNA was calculated as in A.(C) Expression of full-length SCL6-IV (At 4 g 00150) mRNA. The mean ratio of full-length SCL6-IV mRNA (a) to the cleavage product (b) is shown. (D) Expression of TyrAT (At 2 g 20610) mRNA. This blot was first used to analyze ARF10 expression, then stripped and reprobed.

Figure 3.

miRNA and miRNA* accumulation in transgenic Arabidopsis plants. Small RNA samples from vector-transformed (lanes 1,2) and suppressor expressing transgenic Arabidopsis plants (lanes 3–12) were analyzed in duplicate by hybridization with oligonucleotide probes. Ethidium bromide-stained tRNA and 5S rRNA are shown below each blot, and mobility positions of 24- and 21-nucleotide RNA size standards are shown. Mean relative accumulation (RA) of miRNA or miRNA* signal relative to that in vector-transformed plants (lanes 1,2) is shown. (A) Accumulation of miR167 and miR167* from the miR167b locus. (B) Accumulation of miR160 and miR160* from the miR160c locus. (C) Accumulation of miR171 and miR171*.

Figure 4.

Interactions between suppressor proteins and small RNAs. (A) Co-IP of suppressor proteins, miRNAs, and miRNAs* from transgenic Arabidopsis inflorescence tissue. Immunoblot analysis was done using IP input (in) samples and immunoprecipitated fractions using HA or NIa/NIb monoclonal antibodies. Mobility positions of 16–62-kDa protein size standards are shown. Blot hybridization assays were done by using RNA recovered from immunoprecipitates. Mobility positions of 21- and 24-nucleotide RNA size standards are shown. (B) Co-IP of suppressor proteins, miRNAs, and miRNAs* from transgenic Arabidopsis total aerial tissue. (C) Co-IP of suppressor proteins and GFP-specific siRNAs from Agrobacterium-infiltrated N. benthamiana leaf tissue. (D) Electrophoretic mobility shift assays using purified p21 and siRNA duplex, miR171, miR171*, and miR171/miR171* duplex. Complexes were analyzed by native PAGE. The positions of p21:RNA complexes and free probes are shown.

Figure 5.

Mechanisms of suppression of miRNA and siRNA pathways by silencing suppressors. (A) microRNA pathway. (B) Antiviral siRNA pathway. p19 and p21 sequester miRNA/miRNA* and siRNA duplexes, stabilizing both strands. The point in the pathway at which p19 and p21 sequester small RNA duplexes is not known. P1/HC-Pro inhibits duplex unwinding and therefore stabilizes both strands, but by a mechanism that does not involve direct binding to small RNA duplexes. Protein X is a hypothetical protein with a role comparable to animal R2D2 (Liu et al. 2003). In both pathways, interference by HC-Pro, p19, and p21 prevents RISC assembly and subsequent target RNA degradation. The TCV CP is proposed to interfere with DCL2 or other DICER-LIKE activities required for TCV-derived siRNA formation (Xie et al. 2004), but not DCL1.