Viral suppressors of RNA silencing hinder exogenous and endogenous small RNA pathways in Drosophila

Abstract

Background: In plants and insects, RNA interference (RNAi) is the main responder against viruses and shapes the basis of antiviral immunity. Viruses counter this defense by expressing viral suppressors of RNAi (VSRs). While VSRs in Drosophila melanogaster were shown to inhibit RNAi through different modes of action, whether they act on other silencing pathways remained unexplored.

Methodology/principal findings: Here we show that expression of various plant and insect VSRs in transgenic flies does not perturb the Drosophila microRNA (miRNA) pathway; but in contrast, inhibits antiviral RNAi and the RNA silencing response triggered by inverted repeat transcripts, and injection of dsRNA or siRNA. Strikingly, these VSRs also suppressed transposon silencing by endogenous siRNAs (endo-siRNAs).

Conclusions/significance: Our findings identify VSRs as tools to unravel small RNA pathways in insects and suggest a cosuppression of antiviral RNAi and endo-siRNA silencing by viruses during fly infections.

Figure 1. Flies expressing insect and plant VSRs accumulate high levels of FHV RNA1ΔB2 transcripts.

(A) RNA1 and RNA1ΔB2 constructs. The GAL4 inducible UAS-RNA1 transgene (left hand side) expresses a RNA1 of (+) polarity. RNA1 encodes an RdRp protein, which directs autonomous replication of RNA1 of (−) polarity, then of both RNA1 and RNA3 of (+) polarity. Two point mutations (triangles) interrupt the B2 ORF in the RNA1ΔB2 (right hand side) without affecting the RdRp ORF . (B) Fold changes in RNA1+RNA3 levels. The RNA1 transgenic line was crossed with the da>GAL4 line (control). The RNA1ΔB2 transgenic line was crossed with the da>GAL4 line (control) or with da>GAL4; UAS-VSR lines as indicated. RNA1+RNA3 expression was measured by quantitative RT-PCR from the progenies of these crosses. Error bars indicate the standard deviation from triplicate qPCR experiments.

Figure 2. Hypersensitivity to DCV infection.

(A) An Act5C>GAL4 driver line was crossed to UAS>VSR transgenic lines (▪) or to w1118 control strain (□). 20 females from the progeny of these crosses were challenged by an intrathoracic injection of a control Tris solution (---) or a DCV suspension corresponding to 10² LD50 (–) and survival was monitored daily. (B) An hsp>GAL4 driver line was crossed to UAS>VSR transgenic lines. 20 females from the progeny of these crosses were challenged by an injection of the control Tris solution (---) or a the DCV suspension (–) under heat shock (▪) or non heat shock (□) conditions. Plotted values represent the mean±SEM (standard error to the mean) of three independent experiments.

Figure 3. Suppression of silencing induced by injection of long dsRNAs or siRNAs in embryos.

RNAi of the fushi tarazu gene (ftz) in wild type syncytial embryos results in loss of denticle belts in the cuticle of prehatching larvae (A). Early embryos homozygous for the ago2⁴¹⁴ mutant allele or expressing the indicated VSR under the control of the da>GAL4 driver were injected with long dsRNAs (B) or siRNAs (C) targeted to the ftz gene and scored 48 h after injection for the ftz mutant phenotype. Confidence intervals (alpha risk 0.05) for the observed ratio are indicated.

Figure 4. VSRs inhibit silencing induced a white inverse-repeat transgene.

Silencing of the white gene induced by an IR[w] inverted-repeat transgene in flies expressing insect and plant VSRs in eyes. (A) In all heterozygous combinations GMR>GAL4; UAS>GFP or GMR>GAL4; UAS>VSR, the mini-white markers of the two transgenes produce equivalent strong red eye pigmentation. (B) Silencing of the mini-white markers by one copy of the GMR>IR[w] transgene (note that the GMR>IR[w] construct has no mini-white marker [28]). The effect of the GMR>IR[w] transgene on the two mini-white markers in GMR>GAL4; UAS>VSR genetic combinations depends on the efficiency of the suppressor to inhibit white silencing. Eye pigment levels were measured in 3 separate experiments from 2 days-old adult eyes and are expressed as a percentage of pigment levels measured in the GMR>GAL4; UAS>GFP combination. Error bars correspond to standard deviations.

Figure 5. VSRs inhibit retrotransposon silencing by endo-siRNAs.

297 and ZAM retrotransposon silencing is impaired by B2, 1A and P19 VSRs. RNA levels of 297 and ZAM were measured in heads and ovaries of 1 day-old female flies heterozygous for the da>GAL4 driver and the indicated UAS>VSR transgene. Fold changes in RNA levels were calculated relative to the 297 and ZAM RNA levels measured in heterozygous da>GAL4 control flies. Fold changes in homozygous ago2⁴¹⁴ mutants were calculated relative to the 297 and ZAM RNA levels measured in heterozygous ago2⁴¹⁴ flies. Therefore, the weaker effect of the ago2 mutation on 297 expression observed in adult heads may result from variations in copy number of 297 between the ago2 mutant stock and the VSR transgenic stocks. Error bars indicate the standard deviation from triplicate qPCR experiments.