Like other eukaryotes, plants use DICER-LIKE (DCL) proteins as the central enzymes of RNA silencing, which regulates gene expression and mediates defense against viruses. But why do plants like Arabidopsis express four DCLs, a diversity unmatched by other kingdoms? Here we show that two nuclear DNA viruses (geminivirus CaLCuV and pararetrovirus CaMV) and a cytoplasmic RNA tobamovirus ORMV are differentially targeted by subsets of DCLs. DNA virus-derived small interfering RNAs (siRNAs) of specific size classes (21, 22 and 24 nt) are produced by all four DCLs, including DCL1, known to process microRNA precursors. Specifically, DCL1 generates 21 nt siRNAs from the CaMV leader region. In contrast, RNA virus infection is mainly affected by DCL4. While the four DCLs are partially redundant for CaLCuV-induced mRNA degradation, DCL4 in conjunction with RDR6 and HEN1 specifically facilitates extensive virus-induced silencing in new growth. Additionally, we show that CaMV infection impairs processing of endogenous RDR6-derived double-stranded RNA, while ORMV prevents HEN1-mediated methylation of small RNA duplexes, suggesting two novel viral strategies of silencing suppression. Our work highlights the complexity of virus interaction with host silencing pathways and suggests that DCL multiplicity helps mediate plant responses to diverse viral infections.

DNA geminiviruses are thought to be targets of RNA silencing. Here, we characterize small interfering (si) RNAs—the hallmarks of silencing—associated with Cabbage leaf curl begomovirus in Arabidopsis and African cassava mosaic begomovirus in Nicotiana benthamiana and cassava. We detected 21, 22 and 24 nt siRNAs of both polarities, derived from both the coding and the intergenic regions of these geminiviruses. Genetic evidence showed that all the 24 nt and a substantial fraction of the 22 nt viral siRNAs are generated by the dicer-like proteins DCL3 and DCL2, respectively. The viral siRNAs were 5′ end phosphorylated, as shown by phosphatase treatments, and methylated at the 3′-nucleotide, as shown by HEN1 miRNA methylase-dependent resistance to β-elimination. Similar modifications were found in all types of endogenous and transgene-derived siRNAs tested, but not in a major fraction of siRNAs from a cytoplasmic RNA tobamovirus. We conclude that several distinct silencing pathways are involved in DNA virus-plant interactions.

Bipartite geminiviruses encode a small protein, AC2, that functions as a transactivator of viral transcription and a suppressor of RNA silencing. A relationship between these two functions had not been investigated before. We characterized both of these functions for AC2 from Mungbean yellow mosaic virus-Vigna (MYMV). When transiently expressed in plant protoplasts, MYMV AC2 strongly transactivated the viral promoter; AC2 was detected in the nucleus, and a split nuclear localization signal (NLS) was mapped. In a model Nicotiana benthamiana plant, in which silencing can be triggered biolistically, AC2 reduced local silencing and prevented its systemic spread. Mutations in the AC2 NLS or Zn finger or deletion of its activator domain abolished both these effects, suggesting that suppression of silencing by AC2 requires transactivation of host suppressor(s). In line with this, in Arabidopsis protoplasts, MYMV AC2 or its homologue from African cassava mosaic geminivirus coactivated >30 components of the plant transcriptome, as detected with Affymetrix ATH1 GeneChips. Several corresponding promoters cloned from Arabidopsis were strongly induced by both AC2 proteins. These results suggest that silencing suppression and transcription activation by AC2 are functionally connected and that some of the AC2-inducible host genes discovered here may code for components of an endogenous network that controls silencing.

The frontline of plant defense against non-viral pathogens such as bacteria, fungi and oomycetes is provided by transmembrane pattern recognition receptors that detect conserved pathogen-associated molecular patterns (PAMPs), leading to pattern-triggered immunity (PTI). To counteract this innate defense, pathogens deploy effector proteins with a primary function to suppress PTI. In specific cases, plants have evolved intracellular resistance (R) proteins detecting isolate-specific pathogen effectors, leading to effector-triggered immunity (ETI), an amplified version of PTI, often associated with hypersensitive response (HR) and programmed cell death (PCD). In the case of plant viruses, no conserved PAMP was identified so far and the primary plant defense is thought to be based mainly on RNA silencing, an evolutionary conserved, sequence-specific mechanism that regulates gene expression and chromatin states and represses invasive nucleic acids such as transposons. Endogenous silencing pathways generate 21-24 nt small (s)RNAs, miRNAs and short interfering (si)RNAs, that repress genes post-transcriptionally and/or transcriptionally. Four distinct Dicer-like (DCL) proteins, which normally produce endogenous miRNAs and siRNAs, all contribute to the biogenesis of viral siRNAs in infected plants. Growing evidence indicates that RNA silencing also contributes to plant defense against non-viral pathogens. Conversely, PTI-based innate responses may contribute to antiviral defense. Intracellular R proteins of the same NB-LRR family are able to recognize both non-viral effectors and avirulence (Avr) proteins of RNA viruses, and, as a result, trigger HR and PCD in virus-resistant hosts. In some cases, viral Avr proteins also function as silencing suppressors. We hypothesize that RNA silencing and innate immunity (PTI and ETI) function in concert to fight plant viruses. Viruses counteract this dual defense by effectors that suppress both PTI-/ETI-based innate responses and RNA silencing to establish successful infection.

In plants, RNA silencing-based antiviral defense is mediated by Dicer-like (DCL) proteins producing short interfering (si)RNAs. In Arabidopsis infected with the bipartite circular DNA geminivirus Cabbage leaf curl virus (CaLCuV), four distinct DCLs produce 21, 22 and 24 nt viral siRNAs. Using deep sequencing and blot hybridization, we found that viral siRNAs of each size-class densely cover the entire viral genome sequences in both polarities, but highly abundant siRNAs correspond primarily to the leftward and rightward transcription units. Double-stranded RNA precursors of viral siRNAs can potentially be generated by host RDR-dependent RNA polymerase (RDR). However, genetic evidence revealed that CaLCuV siRNA biogenesis does not require RDR1, RDR2, or RDR6. By contrast, CaLCuV derivatives engineered to target 30 nt sequences of a GFP transgene by primary viral siRNAs trigger RDR6-dependent production of secondary siRNAs. Viral siRNAs targeting upstream of the GFP stop codon induce secondary siRNAs almost exclusively from sequences downstream of the target site. Conversely, viral siRNAs targeting the GFP 3′-untranslated region (UTR) induce secondary siRNAs mostly upstream of the target site. RDR6-dependent siRNA production is not necessary for robust GFP silencing, except when viral siRNAs targeted GFP 5′-UTR. Furthermore, viral siRNAs targeting the transgene enhancer region cause GFP silencing without secondary siRNA production. We conclude that the majority of viral siRNAs accumulating during geminiviral infection are RDR1/2/6-independent primary siRNAs. Double-stranded RNA precursors of these siRNAs are likely generated by bidirectional readthrough transcription of circular viral DNA by RNA polymerase II. Unlike transgenic mRNA, geminiviral mRNAs appear to be poor templates for RDR-dependent production of secondary siRNAs. RNA silencing directed by small RNAs (sRNAs) regulates gene expression and mediates defense against invasive nucleic acids such as transposons, transgenes and viruses. In plants and some animals, RNA-dependent RNA polymerase (RDR) generates precursors of secondary sRNAs that reinforce silencing. Most plant mRNAs silenced by miRNAs or primary siRNAs do not spawn secondary siRNAs, suggesting that they may have evolved to be poor templates for RDR. By contrast, silenced transgenes often produce RDR-dependent secondary siRNAs. Here we demonstrate that massive production of 21, 22 and 24 nt viral siRNAs in DNA geminivirus-infected Arabidopsis does not require the functional RDRs RDR1, RDR2, or RDR6. Deep sequencing analysis indicates that dsRNA precursors of these primary viral siRNAs are likely generated by RNA polymerase II-mediated bidirectional readthrough transcription on the circular viral DNA. Primary viral siRNAs engineered to target a GFP transgene trigger robust, RDR6-dependent production of secondary siRNAs, indicating that geminivirus infection does not suppress RDR6 activity. We conclude that geminiviral mRNAs, which can potentially be cleaved by primary viral siRNAs, are resistant to RDR-dependent amplification of secondary siRNAs. We speculate that, like most plant mRNAs, geminiviral mRNAs may have evolved to evade RDR activity.