Antiviral RNA Silencing Initiated in the Absence of RDE-4, a Double-Stranded RNA Binding Protein, in Caenorhabditis elegans

Xunyang Guo, Rui Zhang, Jeffrey Wang, Rui Lu

Department of Biological Sciences, Louisiana State University, Baton Rouge, Louisiana, USA

Small interfering RNAs (siRNAs) processed from double-stranded RNA (dsRNA) of virus origins mediate potent antiviral de-fense through a process referred to as RNA interference (RNAi) or RNA silencing in diverse organisms. In the simple inverte-brate Caenorhabditis elegans, the RNAi process is initiated by a single Dicer, which partners with the dsRNA binding proteinRDE-4 to process dsRNA into viral siRNAs (viRNAs). Notably, in C. elegans this RNA-directed viral immunity (RDVI) also re-quires a number of worm-specific genes for its full antiviral potential. One such gene is rsd-2 (RNAi spreading defective 2), whichwas implicated in RDVI in our previous studies. In the current study, we first established an antiviral role by showing that rsd-2null mutants permitted higher levels of viral RNA accumulation, and that this enhanced viral susceptibility was reversed by ecto-pic expression of RSD-2. We then examined the relationship of rsd-2 with other known components of RNAi pathways and es-tablished that rsd-2 functions in a novel pathway that is independent of rde-4 but likely requires the RNA-dependent RNA poly-merase RRF-1, suggesting a critical role for RSD-2 in secondary viRNA biogenesis, likely through coordinated action with RRF-1.Together, these results suggest that RDVI in the single-Dicer organism C. elegans depends on the collective actions of both RDE-4-dependent and RDE-4-independent mechanisms to produce RNAi-inducing viRNAs. Our study reveals, for the first time, anovel siRNA-producing mechanism in C. elegans that bypasses the need for a dsRNA-binding protein.

In fungi, plants, insects, and nematodes, double-stranded RNAs(dsRNAs), formed by complementary viral transcripts or

through intramolecular base pairing, trigger potent antiviral de-fense through a process often referred to as RNA interference(RNAi) (1). Increasing evidence suggested that this RNA-directedviral immunity (RDVI) is initiated upon the processing of viraldsRNA into small interfering RNAs (siRNAs) by an RNase III-likeenzyme called Dicer (2). Subsequently, these virus-derivedsiRNAs are incorporated into RNA-induced silencing complexes(RISCs), formed by Argonaute protein and cofactors, and serve assequence guides for target viral transcript selection and destruc-tion (1). In diverse organisms, dsRNA binding proteins are essen-tial components of RDVI that contribute to the biogenesis orfunction of virus-derived siRNAs (viRNAs) (3–8). In plants,RDVI is further potentiated by secondary viRNAs processed fromdsRNAs converted from cleaved viral transcripts by RNA-depen-dent RNA polymerases (RdRPs) (9–12). Plant RDVI also involvesan intercellular signal that is believed to restrict systemic spread-ing of the invading virus by priming an antiviral status prior tovirus arrival (13, 14). Recently, siRNAs were found to serve as thephysical carrier of this mobile silencing signal in plants (15, 16). Inthe fruit fly RDVI also spreads systemically, although the mecha-nism involved is different (17).

Since RNAi triggered by artificial dsRNA mechanistically reca-pitulates RDVI, previous studies on artificial dsRNA-triggeredRNAi in C. elegans have significantly improved our understandingof RDVI in nematode worms. Current models envision that theworm RDVI is initiated upon the processing of viral dsRNA intoprimary viRNAs by the worm Dicer, called DCR-1 (4, 5, 18).DCR-1 also processes precursor microRNAs (miRNAs) into ma-ture miRNAs, a class of endogenous small RNAs with importantroles in development (19). Whereas efficient processing of viraldsRNA into primary viRNA requires a dsRNA binding protein,termed RDE-4 (RNAi defective 4), the processing of precursormiRNAs into mature miRNAs appears to be RDE-4 independent

(3, 5, 19–21). Interestingly, RDE-4, a key factor of RDVI, is alsoconserved in the fruit fly, whose genome is known to encode twoDicer proteins with dedicated function in the biogenesis of siRNAand miRNA, respectively (7, 20–23). So far, at least two closelyrelated AGO proteins, RDE-1 (RNAi defective 1) and C04F12.1,have been found to play an important role in RDVI in C. elegans(3–5, 24). RDE-1 has the slicer activity that is required for thematuration of RISC but not the cleavage of the target transcript(25). Currently, how C04F12.1 contributes to RDVI remainslargely unknown. Efficient RDVI in C. elegans also requires anRdRP, termed RRF-1, which produces 22-nucleotide (nt) single-stranded secondary siRNAs in a Dicer-independent manner (4, 5,26, 27) (28). The fact that RDE-1 is required for production ofsecondary siRNAs suggests that RRF-1 functions downstream ofRDE-1 in the same genetic pathway (28, 29).

In addition to Dicer, AGO, and RdRP proteins, the wormRDVI also requires some unique components, such as DRH-1(Dicer-related RNA helicase 1) and RSD-2 (RNAi spreading de-fective 2) (5). DRH-1 encodes a putative DEAD box RNA helicasethat shares significant sequence homology with RIG-I (retinoicacid-inducible gene 1), a cytosolic virus sensor that initiates inter-feron-mediated viral immunity upon virus detection in mammals(30). Intriguingly, although essential to RDVI, DRH-1 appears tobe dispensable in RNAi triggered by artificial dsRNAs, suggestinga dedicated role of DRH-1 in antiviral defense (5). Previously, wereported that viRNAs detected in drh-1 mutants became undetect-able in drh-1;rde-4 double mutants (5), suggesting that DRH-1,

Received 15 May 2013 Accepted 18 July 2013

Published ahead of print 24 July 2013

Address correspondence to Rui Lu, ruilu@lsu.edu.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/JVI.01305-13

October 2013 Volume 87 Number 19 Journal of Virology p. 10721–10729 jvi.asm.org 10721

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

together with RDE-1 and RRF-1, functions in the same geneticpathway as RDE-4.

RSD-2, together with SID-1 (systemic RNA interference defi-cient 1), was originally identified as one of the key factors thatenable systemic spreading of RNAi (31–33). A recent study sug-gested that RSD-2 contributes to the accumulation of secondarysiRNAs in both exogenous and endogenous RNAi pathways (34).Since rsd-2 mutants displayed defects in transposon silencing un-der unfavorable conditions, it is believed that RSD-2 plays an im-portant role in maintaining chromosome integrity (35). SID-1contains multiple putative transmembrane domains and has beenshown to be required for dsRNA intake (33, 36, 37). Interestingly,a recent study suggested that at least two classes of silencingdsRNAs that move between C. elegans tissues can act as or generatethe systemic silencing signal (38). Currently, whether SID-1 con-tributes to systemic RNAi by transporting these two classes ofsilencing RNAs remains largely unknown.

Previously, high concentrations of dsRNAs were found to trig-ger efficient RNAi in the absence of RDE-4 (39). Currently, thebiological significance of this observation remains largely un-known. By analyzing viral replication and viRNA accumulation inworm single and double mutants defective in rsd-2, we show herethat RSD-2, as an important component of worm RDVI, func-tions downstream of primary viRNA biogenesis. Most impor-tantly, with an increase in viral replication, primary viRNAs accu-mulated to a higher level in rsd-2;rde-4 double mutants than inrsd-2 single mutants, suggesting an rde-4-independent mecha-nism for the biogenesis of primary viRNAs in C. elegans. We fur-ther showed that RRF-1, together with RDE-1, also contributed tothe rde-4-indepedent RDVI. Notably, although playing an essen-tial role in facilitating the spreading of RNAi triggered by artificialdsRNAs, SID-1 appears dispensable in RDVI targeting a naturalviral pathogen of C. elegans.

MATERIALS AND METHODSWorm genetics. The Bristol isolate of C. elegans, N2, was used as thereference strain in this study. All other strains used in this study are de-rived from N2 and include drh-1 (tm1329), drh-2 (ok951), rde-1 (ne300),rde-4 (ne337), rsd-2 (tm1429 and pk3307), sid-1 (qt2), and rrf-1 (pk1417)mutants. The genotypes of drh-1, drh-2, rsd-2, and rrf-1 mutants wereconfirmed using PCR. The primers used to confirm tm1329 and drh-2alleles were described previously (5). The primer pairs used to confirm thealleles are tm1429intf (CTACTAAACCGGTTACGTGT) and tm1429intb(CCACAGGGATTTTGTAGGGA) for tm1429 and 1417EcoRV (AGGAGAGCATAGAAGGATATCA) and 1417PfoI (GTCACGGGGAGCTATTGTGAA) for pk1417. The genotypes of rde-1, sid-1, and rde-4 mutants wereconfirmed using skn-1 feeding RNAi combined with genomic DNA se-quencing. NGM plates seeded with Escherichia coli strain OP50 were usedto maintain all worm strains. Single and double mutants containing var-ious transgenes were created through standard genetic crosses.

Plasmid constructs and transgenic worms. The construction ofFR1gfp replicon was described previously (5). Psur5::RSD-2 was devel-oped by inserting RSD-2 coding sequence, amplified through reversetranscription-PCR (RT-PCR), into the LR50 vector described previously(40). The RSD-2 coding sequence in the resulting construct was con-firmed by DNA sequencing. The plasmid construct used to drive gfpdsRNA expression in E. coli was described previously (40). The chromo-somal integrant for Psur5::RSD-2 was generated using a protocol de-scribed previously (40).

RSD-2 function rescue assay. The RSD-2 function rescue assay uti-lized ectopic expression of RSD-2 coding sequence in rsd-2 null mutants(tm1429) carrying a nuclear transgene corresponding to the FR1gfp rep-

licon. The sur-5 promoter, known to be active in most of the worm cells(41), was used to drive the expression of RSD-2. A construct that directsmCherry expression in the pharynx tissue was used to produce a visualmark for the Psur5::RSD-2 transgene. The RSD-2 function rescue assaybegan with microinjection of Psur5::RSD-2 construct into the gonad ofrsd-2;FR1gfp worms together with the mCherry reporter construct. Be-cause of the nature of worm transformation through gonad injection,most of the transgenic extrachromosomal arrays are randomly passed onto the next generations. As a result, some progenies are free of the extra-chromosomal arrays and can serve as internal negative control. Upon heatinduction, which will initiate the replication of the FR1gfp replicon, trans-genic worms marked with red fluorescence in the head will show no ordecreased green fluorescence if the Psur5::RSD-2 transgene is able to re-store the function of RSD-2 in the rsd-2 mutants.

RNA preparation and Northern analysis. Total RNA was prepared 48h after heat induction using TRI reagent by following the manufacturer’sinstructions (Sigma-Aldrich, Inc.). Small RNA species in the total RNAsamples were further enriched using the mirVana kit (Ambion). For thedetection of high-molecular-weight viral RNAs, 4 �g total RNA of eachsample was fractionated in 1.2% agarose gel. For the detection of smallRNA, including viRNAs and miRNAs, 10 to 20 �g of small RNA persample was resolved in 15% acrylamide denaturing gel. After electropho-resis, all RNA samples were transferred onto a Hybond N� membrane(GE Healthcare Inc.) and UV cross-linked using 1.8 � 105 �J/cm2 as theoutput power (SpectroLinker).

The blots for the detection of high-molecular-weight viral RNAs werehybridized with alkaline phosphatase-labeled cDNA probes prepared us-ing an AlkPhos direct labeling module (GE Healthcare). The probes forFR1gfp transcript detection were prepared using cDNA fragments ampli-fied from the green fluorescent protein (GFP) coding sequence withinFR1gfp. The probes used to detect Orsay virus RNA1 were prepared usingcDNA fragments amplified, through RT-PCR, from the 3= end of Orsayvirus RNA1 genome. The hybridization was performed at 65°C overnight.After washing at 65°C three times, the labeled cDNA probes were detectedusing CDP-Star detection reagent by following the manufacturer’s in-structions (GE Healthcare).

The detection of siRNAs and miRNAs adopted a protocol describedpreviously (40). For the detection of FR1gfp-derived siRNAs, DNA oligo-nucleotides covering the entire subgenomic RNA sequence were labeledusing the digoxigenin (DIG) oligonucleotide tailing kit (Roche AppliedScience). DNA oligonucleotides with the sequence ATTGCCGTACTGAACGATCTCA were labeled and used for the detection of miR-58. FourDNA oligonucleotides with sizes of 19, 21, 23, and 25 nt, respectively, weredetected using DIG-labeled DNA oligonucleotides and, together withmiR-58, served as size references.

Infectious filtrate preparation and Orsay virus inoculation. Orsayvirus was maintained using the JU1580 isolate of C. elegans at room tem-perature by following a protocol described previously (6). To prepareOrsay virus inoculum, JU1580 worms infected with Orsay virus werewashed off, using M9 buffer, from slightly starved 10-cm agar plates. Afterspinning at low speed, the virus-containing supernatant was filteredthrough a 0.22-�m filter unit (Millipore), and the filtrate was used toresuspend pelleted OP50 E. coli for NGM plate seeding.

RNAi experiments. Feeding RNAi targeting the endogenous geneskin-1 or the transgene gfp was performed using a bacterial feeding pro-tocol described previously (42). NGM agar plates containing 5 mM iso-propyl-�-D-thiogalactopyranoside (IPTG) and 50 mg/ml carbenicillinwere used to prepare the feeding RNAi food. All feeding RNAi experi-ments were carried out at room temperature.

Imaging microscopy. The green and red fluorescence images wererecorded using the same exposure for each set of images. A Nikon digitalcamera p7000 mounted on a Nikon SMZ1500 microscope was used torecord all images.

Guo et al.

10722 jvi.asm.org Journal of Virology

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

RESULTSrsd-2 is not essential for the biogenesis of primary viRNAs. rsd-2of C. elegans was first implicated in RDVI through an RNAi screenthat selected for genes whose downregulation by dsRNA feedingled to loss of RDVI (5). To confirm that the loss of RDVI in wormsfed with rsd-2-targeting dsRNA was indeed caused by partial lossof rsd-2 function, we examined the replication of FR1gfp repliconin the mutant worms containing the tm1429 or pk3307 allele ofrsd-2. The tm1429 allele contains a 251-nt deletion, spanning fromnt 1376 to 1626 of rsd-2 cDNA, and is predicted to produce atruncated RSD-2 that contains only the first 458 amino acid (aa)residues of the 1,266-aa wild-type RSD-2. Conversely, the pk3307allele contains a premature stop codon in the RSD-2 coding se-quence that caused the truncation of the last 542 aa of RSD-2.Thus, both rsd-2 alleles are considered null alleles. The FR1gfpreplicon was developed from flock house virus (FHV) RNA1 byreplacing the coding sequence of B2, the FHV-encoded RNAi sup-pressor, with that of green fluorescence protein (GFP). As a result,the replication of FR1gfp is subdued by RDVI in wild-type N2worms but rescued in RDVI-defective worms, such as the rde-1mutants. As shown in Fig. 1A, similar to the rde-1 mutants, bothrsd-2 mutants supported elevated accumulation of FHV RNAs(Fig. 1A, compare lanes 3 and 4 to lanes 1 and 2), confirming acritical role of wild-type RSD-2 in FHV-targeting RDVI. En-hanced accumulation was also observed in the rsd-2 mutants forOrsay virus, a natural viral pathogen of C. elegans that is closelyrelated to FHV (Fig. 1B, compare lanes 2 and 3 to lane 1) (6). Theseresults together strongly suggested that RSD-2 plays an importantrole in RDVI.

To reconfirm that the loss of RDVI in rsd-2 mutants indeedresulted from the rsd-2 null alleles but not any other closely linkedgenetic alleles, we checked if ectopic expression of wild-typeRSD-2 restores RDVI in rsd-2 mutants (tm1429) containing theFR1gfp replicon. The ectopic expression of wild-type RSD-2 cod-ing sequence was achieved through gonad microinjection of plas-mid construct containing RSD-2 coding sequence driven by thesur-5 promoter (Fig. 1C) (41). A plasmid construct that directsmCherry expression in pharynx tissue was coinjected to produce avisual reporter for the transgene. As shown in Fig. 1C, FR1gfpreplication, manifested as green fluorescence, was suppressed inrsd-2 mutants containing the Psur-5::RSD-2 transgene (comparethe worms that showed red fluorescence in the heads to wormsthat did not). Consistent with this observation, FR1gfp transcriptswere detected at a reduced level in worms containing the Psur-5::RSD-2 transgene (Fig. 1D, compare lane 2 to lane 3). These resultsconfirmed that RSD-2 indeed plays an important role in wormRDVI.

rsd-2 is known to contribute to the accumulation of secondary,but not primary, endogenous siRNAs (34). To find out whetherRSD-2 contributes to RDVI through a similar mechanism, wechecked the accumulation of viRNAs in rsd-2 mutants (tm1429)using Northern blotting. Both rde-1 and rde-4 play essential rolesin worm RDVI. However, virus-derived siRNAs accumulate toreadily detectable levels in rde-1 but not rde-4 mutants (3, 5).Thus, as a control, the accumulation of viRNAs in rde-1 and rde-4mutants was also detected in this test. As shown in Fig. 1E, al-though not detectable in wild-type N2 worms or rde-4 mutants,FR1gfp-derived viRNAs were detected at high levels in rde-1 mu-tants, with a major band detected at the position corresponding to

23 nt. FR1gfp-derived siRNAs were also detected in rsd-2 mutantswith a pattern similar to that in rde-1 mutants but at a visibly lowerlevel. Previously, it has been shown that primary siRNAs pro-duced by worm Dicer are predominantly 23 nt in size and accu-mulate at high levels in rde-1 mutants (18, 29, 43, 44). This resultsuggested that rsd-2 is not essential for the biogenesis of primaryviRNAs.

rsd-2 contributes to rde-4-independent antiviral silencing.We previously reported that FR1gfp-derived siRNAs detected indrh-1 mutants become undetectable in drh-1;rde-4 double mu-tants, suggesting that drh-1 and rde-4 function in a linear geneticpathway (5). To find out whether rsd-2 also functions in the samegenetic pathway, we compared the accumulation of primaryviRNAs in double mutants corresponding to drh-1;rde-4, rsd-2;rde-4, and rsd-2;drh-1. As shown in Fig. 2A, although not detect-able in drh-1;rde-4 double mutants, FR1gfp-derived primarysiRNAs were detected in both rsd-2;drh-1 and rsd-2;rde-4 doublemutants (compare lane 3 to lanes 1 and 4). Most importantly, anenhanced replication for both Orsay virus and FR1gfp was ob-served in rsd-2;rde-4 and rsd-2;drh-1 double mutants compared torde-4 and drh-1 single mutants, respectively (Fig. 2B and C, left).These observations together suggested an rde-4-independentmechanism that directs antiviral silencing in an rsd-2-dependentmanner in C. elegans. RDE-1 specifically recruits primary siRNAs,but not secondary siRNAs, for target RNA selection and is be-lieved to function downstream of RDE-4 (29). Thus, we expectedto see enhanced viral replication in rsd-2;rde-1 double mutantscompared to rde-1 single mutants if rsd-2 indeed contributes toRDVI in an rde-4-independent manner. As shown in Fig. 2B andC, right, enhanced viral replication was observed, although to alesser extent, in rsd-2;rde-1 double mutants compared to rde-1single mutants. As expected, FR1gfp-derived siRNAs accumulatedin rsd-2;rde-1 double mutants (Fig. 2A, lane 2). Taken together,these results strongly suggest that there is an rde-4-independentbut rsd-2-dependent pathway contributing to RDVI in C. elegans.

rsd-2 and rrf-1 function in the same RDVI pathways. rrf-1encodes an RdRP that initiates de novo synthesis of secondarysiRNAs using the targeted transcript as the template in RNAi (26,27, 45, 46). To find out whether rsd-2 functions in the same geneticpathway as rrf-1, which is known to function downstream of rde-4and rde-1 (28, 29), we checked viral replication in rsd-2;rrf-1 dou-ble mutants compared to the single mutants. We expected to seeenhanced viral replication in the double mutants if rsd-2 and rrf-1function in separate genetic pathways. However, as shown in Fig.3A, an enhancement in the replication of FR1gfp or Orsay viruswas not observed in the double mutants compared to the singlemutants (compare lane 4 to lanes 2 and 3 in the left panel and lane5 to lanes 3 and 4 in the right panel). Consistently, FR1gfp-derived23-nt primary siRNAs accumulated to comparable levels in thersd-2;rrf-1 double mutants and in the single mutants (Fig. 3B,compare lanes 1 and 2 to lane 7). This result, together with theresults shown in Fig. 2, suggest that both rrf-1 and rsd-2 functionin two independent genetic pathways that mediate rde-4-depen-dent and rde-4-independent antiviral silencing, respectively.

rrf-1 contributes to rde-4-independent antiviral silencing.To confirm that rrf-1 indeed contributes to rde-4-independentRDVI, we first compared the replication of both FR1gfp and Orsayvirus in rrf-1;rde-4 and rrf-1;drh-1 double mutants to that in thesingle mutants corresponding to rrf-1, drh-1, and rde-4. We ex-pected to see enhanced viral replication in the double mutants if

RDE-4-Independent Antiviral Silencing

October 2013 Volume 87 Number 19 jvi.asm.org 10723

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

FIG 1 rsd-2 is not essential for the biogenesis of primary viRNA. (A) Northern blot showing the accumulation of FR1gfp genomic and subgenomic transcripts indifferent genetic backgrounds. Total RNA was extracted from heat-induced worms 48 h after heat induction. The viral transcripts were detected using probes preparedfrom full-length gfp cDNA. RNA1, genomic RNA of FR1gfp; RNA3, subgenomic RNA of FR1gfp; rRNA, methylene blue-stained rRNA serving as an equal loadingcontrol. (B) Accumulation of Orsay virus RNA1 (ovRNA1) in wild-type N2 worms and rsd-2 mutants carrying the tm1429 or NL3307 allele as indicated. Total RNA wasprepared 72 h after viral inoculation. Orsay virus RNA1 was detected using cDNA probes prepared using a cDNA fragment amplified from the 3= end of Orsay virusRNA1. (C) The upper panel shows the structure of Psur-5::rsd-2. Psur-5, the promoter for the endogenous gene sur-5; RSD-2, the coding sequence of wild-type rsd-2;UTR, the 3=-end untranslated region of unc-54. The lower panel shows visualization of green fluorescence in rsd-2 mutants (tm1429 allele) carrying the FR1gfp replicontransgene and extrachromosmal array corresponding to Psur-5::RSD-2 48 h after heat induction. Shown here are merged images recorded, using the same exposure,under red fluorescence and green fluorescence. Worms showing red color are transgenic for Psur-5::RSD-2. (D) Accumulation of FR1gfp transcripts in wild-type N2worms, drh-1 mutants, and rsd-2 mutants (tm1429) with or without the extrachromosomal assay derived from the Psur-5::RSD-2 construct. RSD-2wt denotes thePsur-5::RSD-2 transgenic array. (E) Accumulation of FR1gfp-derived siRNAs in wild-type N2 worms and RDVI-defective mutants as indicated. Lane M, four DNAoligonucleotides with sizes of 19, 21, 23, and 25 nt detected and used, together with miR-58, as size references.

Guo et al.

10724 jvi.asm.org Journal of Virology

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

rrf-1 indeed contributes to rde-4-independent RDVI. As shown inFig. 4A and B, both FR1gfp and Orsay virus replicated to higherlevels in the double mutants than in the single mutants. Mostimportantly, we detected FR1gfp-derived primary siRNAs in therrf-1;rde-4 and rrf-1;drh-1 double mutants, in contrast to what wasseen in drh-1;rde-4 double mutants (Fig. 4C, compare lanes 2 and4 to lane 1). As shown in Fig. 4A and B, an increase in FR1gfpreplication was observed in the rrf-1;rde-1 double mutants com-pared to the single mutants, although such an increase was notclear for Orsay virus. As expected, FR1gfp-derived primarysiRNAs also accumulated in rde-1;rrf-1 double mutants (Fig. 4C,lane 3). These observations confirmed that rrf-1 plays a role inrde-4-independnet RDVI.

rde-1 plays a role in rde-4-independent RDVI. Previously, en-hanced viral replication was observed in drh-1;rde-1, but not indrh-1;rde-4, double mutants compared to the single mutants (5),suggesting a role for rde-1 in rde-4-independent RDVI. To findout whether rde-1 indeed contributes to rde-4-independentRDVI, we compared both viral replication and primary viRNAaccumulation in rde-1;rde-4 double mutants and the single mu-tants. We expected to see enhanced viral replication and accumu-lation of primary viRNAs if rde-1 indeed plays a role in rde-4-independent RDVI. As shown in Fig. 5A, both Orsay virus and

FR1gfp replicated to higher levels in the double mutants than thesingle mutants. Most importantly, FR1gfp-derived primarysiRNAs, although not detectable in the drh-1;rde-4 double mu-tants, were readily detected in rde-1;rde-4 double mutants (Fig.5B, compare lane 3 to lane 2). These results together suggested thatrde-1 also plays a role in rde-4-independent RDVI.

sid-1 is not required for RDVI targeting Orsay virus. In C.elegans, RNAi triggered by artificial long dsRNA spreads systemi-cally and causes sequence-specific silencing in distant tissues thatare not exposed to the dsRNA trigger (32). This observation raisedthe question of whether, like that in plants (13, 14), worm RDVIinvolves an intercellular signal that prevents viruses from spread-ing systemically. Previously, we have shown that RDVI triggeredby an FHV replicon was not compromised in sid-1 mutants (40)defective in RNAi spreading (33, 36, 47). However, since potentRDVI will have been triggered in every individual cell that con-tains the replicon transgene, a systemic silencing signal may not beneeded to silence such a viral replicon that is not known to movesystemically. To find out whether worm RDVI indeed spreadssystemically to restrict virus spread under natural conditions, wecompared Orsay virus infection between wild-type N2 worms andworms defective in sid-1. As shown in Fig. 6, an increase in Orsayvirus replication was not detected in sid-1 mutants compared tothat in wild-type N2 worms (compare lane 4 to lane 2). Based on

FIG 2 rsd-2 contributes to rde-4-independent antiviral silencing. (A) Accu-mulation of FR1gfp-derived siRNAs in double mutants defective in RDVI.Experimental details are the same as those described in the legend to Fig. 1E.(B) Accumulation of Orsay virus RNA1 (ovRNA1) in wild-type N2 worms andRDVI-defective mutants carrying single or double mutations as indicated. Seethe legend to Fig. 1B for experimental details. (C) Accumulation of FR1gfptranscripts in various genetic backgrounds as indicated. See the legend to Fig.1A for experimental details.

FIG 3 rsd-2 and rrf-1 function in the same antiviral silencing pathways. (A)Accumulation of FR1gfp and Orsay virus transcripts in various genetic back-grounds as indicated. See the legend to Fig. 1A for experimental details. (B)The accumulation of FR1gfp-derived siRNAs in single or double mutants con-taining rsd-2 and/or rrf-1 null alleles. See the legend to Fig. 1E for experimentaldetails.

RDE-4-Independent Antiviral Silencing

October 2013 Volume 87 Number 19 jvi.asm.org 10725

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

this observation, we concluded that sid-1-mediated systemicRNAi is not required for restricting Orsay virus infection in C.elegans.

DISCUSSION

rsd-2 was first identified as a gene that enables RNAi spreadingfrom soma to germ line (31). A role of rsd-2 in RDVI was thenrevealed in a genetic screening that utilized RNAi-mediated geneknockdown to phenocopy genetic mutants (5). Recently, a geneticstudy further suggested that rsd-2 helps maintain normal chromo-somal function, such as transposon control, under unfavorableconditions (35). Currently, how rsd-2 contributes to those biolog-ical processes remains poorly understood. rsd-2 is unique to nem-atode worms, such as C. elegans, which are known to encode thesingle Dicer required for the biogenesis of both siRNAs andmiRNAs. Thus, the functional and mechanistic study of rsd-2 isexpected not only to shed light on the mechanism by which rsd-2helps maintain chromosomal integrity in response to environ-mental stresses but also to reveal the uniqueness of RDVI in single-Dicer invertebrates. Here, we show that rsd-2 functions in twoparallel genetic pathways that mediate antiviral silencing in rde-4-dependent and rde-4-independent manners. In Arabidopsis,dsRNA binding protein regulates RDVI by facilitating the process-

ing of viral dsRNA into siRNAs by one of the four Dicers (8, 48). InDrosophila, the RDE-4 homologue R2D2 contributes to the func-tion of viRNAs produced by one of the two drosophila Dicers (7,22, 49). Therefore, by demonstrating the rde-4-indpendent mech-anism for primary viRNA biogenesis, our study, for the first time,suggested that RDVI in single-Dicer invertebrates can be initiatedin the absence of a dsRNA binding protein.

Previously, it has been shown that dsRNA of high concentra-tion triggers RNAi in C. elegans in the absence of RDE-4 (39). Thisobservation suggests that viruses, as powerful replicators that rap-idly produce large amounts of dsRNA replication intermediates,trigger RDVI in the absence of RDE-4. Here, we show that with anincrease in viral replication level, primary viRNAs were detected athigh levels in rde-4 mutants containing either rsd-2 or the rrf-1null allele, suggesting an rde-4-independent mechanism for the

FIG 4 rrf-1 contributes to rde-4-independent antiviral silencing. (A) Accu-mulation of Orsay virus RNA1 (ovRNA1) in wild-type N2 worms and RDVI-defective mutants carrying single or double mutations as indicated. (B) Accu-mulation of FR1gfp transcripts in wild-type N2 worms and RDVI-defectivemutants carrying single or double mutations as indicated. (C) Accumulationof FR1gfp-derived siRNAs in double mutants, as indicated, defective in RDVI.

FIG 5 rde-1 contributes to rde-4-independent antiviral silencing. (A) Viralreplication is further enhanced in rde-1;rde-4 double mutants compared to thesingle mutants. Shown here is the accumulation of Orsay virus (left panel) andFR1gfp transcripts in single or double mutants containing rde-1 and/or rde-4null alleles. (B) Accumulation of FR1gfp-derived primary siRNAs in single ordouble mutants containing rde-1 and/or rde-4 null alleles.

FIG 6 sid-1 is not required for RDVI in C. elegans. Shown is the accumulationof Orsay virus RNA1 in wild-type N2 worms and sid-1 mutants. See the legendto Fig. 1B for experimental details.

Guo et al.

10726 jvi.asm.org Journal of Virology

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

generation of primary viRNAs (Fig. 2 and 4). Apparently such anincrease in viral replication and viRNA accumulation cannot beattributed to residual rde-4 function in the rde-4 allele used in thisstudy, since FR1gfp-derived siRNAs were not detected in rde-4single or drh-1;rde-4 double mutants under the same conditions(5) (Fig. 1E, 2A, 4C, and 5B). Thus, by demonstrating the exis-tence of an rde-4-independent pathway for antiviral silencing, weexplained why dsRNAs of high levels triggers RNAi in the absenceof RDE-4 in C. elegans (39). Notably, although readily detectablein worms, such as rde-1, rrf-1, and rsd-2 mutants, that are defectivein secondary viRNA biogenesis, virus-derived primary siRNAswere hardly detectable in wild-type N2 worms or worms defectivein rde-4 (Fig. 1E, 2A, 3C, and 4C). This may simply reflect the factthat antiviral silencing in both rde-4-dependent and rde-4-inde-pendent mechanisms is amplified by the secondary viRNAs and,as a result, much less viral dsRNA will be processed into primaryviRNAs by Dicer in the presence of secondary viRNAs.

Notably, our observations suggested that rde-1 also plays a rolein rde-4-independent RDVI similar to that of rsd-2 (Fig. 5). Inter-estingly, an increase in FHV replication level was observed in rde-1;rsd-2 double mutants compared to rde-1 single mutants. rde-1probably plays a major role in rde-4-dependent RDVI but a minorrole in rde-4-independent RDVI, and as a result, antiviral silencingin rde-1 mutants is further compromised in the presence of anrsd-2 null allele. Nevertheless, by analyzing the accumulation ofviral transcripts and primary viRNAs in various worm single anddouble mutants defective in RDVI, our study demonstrated, forthe first time, that in addition to the conventional rde-4-depen-dent pathway, there is an rde-4-independent pathway for antiviralsilencing in C. elegans (Fig. 7). Previously, we showed that anotherRDE-1-related worm AGO, termed C04F12.1, plays a role inRDVI. It would be interesting to test whether C04F12.1 plays arole in rde-4-independent RDVI. C. elegans is known to encode alarge number of AGO proteins. Thus, testing the rde-4 depen-dence of C04F12.1 may allow us to address the question ofwhether different worm AGO proteins specifically function inmechanistically distinct RDVI pathways in C. elegans.

rsd-2 is known to be required for the accumulation of second-ary, but not primary, endogenous siRNAs (34). Currently, it re-mains unclear whether rsd-2 contributes to RDVI through a sim-ilar mechanism. Here, we show that FR1gfp-deirved primarysiRNAs accumulated in rsd-2 with a pattern similar to that in rrf-1or rde-1 single mutants, which are known to accumulate primaryviRNAs (Fig. 1E, 2A, and 3B). This finding suggests that, similar tothat in endogenous RNAi, rsd-2 enhances cell-autonomous anti-viral silencing by contributing to the accumulation of secondary,but not primary, viRNAs. This finding explained why rsd-2mutants are resistant to low doses but sensitive to high doses ofdsRNAs (34). Presumably, when the trigger dsRNAs are present athigh concentrations, sufficient primary siRNAs will be producedto trigger efficient RNAi in the absence of secondary siRNAs.However, when the trigger dsRNAs are introduced at low levels,efficient RNAi will require those dose-sensitive RNAi factors, suchas RSD-2, RDE-10, and RDE-11, which amplify RNAi by promot-ing the production/accumulation of secondary siRNAs (34). No-tably, the replication of both FHV and Orsay virus was furtherenhanced in rsd-2 mutants compared to that in rrf-1 mutants,suggesting that rsd-2 also contributes to the function of primaryviRNAs. RSD-2 contains 3 tandem domains of unknown functionat the N terminus. Functional characterization of these domains

may yield further insight into the mechanistic basis of RSD-2 inRDVI.

Like that in plants, RNAi in C. elegans involves an intercellularsignal that is responsible for systemic spread of RNAi (32, 50, 51).A recent study further suggested that long dsRNAs or their pro-cessed intermediates are the physical carrier of this intercellularsignal (38). Consistent with this finding, systemic spreading of theintercellular signal requires the transmembrane protein SID-1that transports dsRNAs without a size preference and may facili-tate the intercellular transportation of the silencing signal (33, 36,47). Previously, it has been shown that SID-1 is dispensable forRDVI triggered by artificial viruses, such as vesicular stomatitisvirus and FHV (4, 40). Currently, whether SID-1 contributes toRDVI under natural conditions remains largely unknown. To ad-dress this question, we compared Orsay virus infection betweenwild-type N2 worms and sid-1 mutants and found that Orsay virusinfection was not further enhanced in the sid-1 mutants (Fig. 6).Worm RDVI probably involves an intercellular silencing signalthat is transported through an SID-1-independent mechanism.Consistent with this hypothesis, SID-1 was shown to be dispens-able for the export of the mobile silencing signal in C. elegans (52).Alternatively, SID-1, as a key component of systemic RNAi, maybe active in nonintestinal tissues, and as a result its role in systemic

FIG 7 Working model for rde-4-dependent and rde-4-independent antiviralsilencing in C. elegans. WAGO, worm-specific AGO (29). Here, we proposethat the processing of viral dsRNAs into primary viRNAs occurs in an rde-4-dependent and rde-4-independent manner. drh-1 contributes to the rde-4-dependent, but not rde-4-independent, antiviral silencing. We believe thatrsd-2 and rrf-1 contribute to both rde-4-dependent and rde-4-independentantiviral silencing, while rde-1 plays a major function in the rde-4-dependentantiviral silencing and a minor function in rde-4-independent antiviral silenc-ing. We also believe that whereas the major function of rrf-1 is to direct thesynthesis of secondary viRNAs, rsd-2 contributes to the function of primaryviRNAs, thereby initiating the synthesis of secondary viRNAs. As proposedpreviously, some WAGOs may direct the cleavage of cognate viral transcriptsusing the secondary viRNAs as a sequence guide (29).

RDE-4-Independent Antiviral Silencing

October 2013 Volume 87 Number 19 jvi.asm.org 10727

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

RDVI cannot be identified using Orsay virus, which mainly targetsthe intestine cells for replication (6). Moreover, probably due tothe lack of an RDVI suppression activity, Orsay virus is known tobe highly sensitive to RDVI (28). Thus, it is possible that the rep-lication of Orsay virus in sid-1 mutants is suppressed by intracel-lular antiviral silencing, making the intercellular antiviral silenc-ing redundant in keeping this particular virus under control.

Unlike RDE-4, which plays an important role in the biogenesisof endogenous siRNAs, DRH-1 appears to be a dedicated factor ofworm RDVI (5, 20). DRH-2 shares significant sequence homologywith DRH-1 but appears to negatively regulate worm RDVI (5).These observations together raised the interesting question ofwhether DRH-2 specifically targets and regulates the function ofDRH-1, thereby negatively regulating the rde-4-dependent RDVI.Apparently, by addressing this question we will be able to find outwhether worms have evolved such a mechanism that would allowfor specific regulation of RDVI without affecting the normal cel-lular function under unfavorable environments.

ACKNOWLEDGMENTS

We thank the Caenorhabditis Genetics Center (CGC) for some of theworm strains used in this study. This work would not have been possiblewithout the JU1580 worm strain and the Orsay virus from M.-A. Félix,E. A. Miska, and D. Wang.

This work was supported by the Board of Regents, Louisiana[LEQSF(2011-14)-RD-A-09 and LEQSF-EPS(2012)-PFUND-277], andthe NIH (1R03AI092159-01A1).

REFERENCES1. Ding SW, Voinnet O. 2007. Antiviral immunity directed by small RNAs.

Cell 130:413– 426.2. Roghiyh Aliyari S-WD. 2009. RNA-based viral immunity initiated by the

Dicer family of host immune receptors. Immunol. Rev. 227:176 –188.3. Wilkins C, Dishongh R, Moore SC, Whitt MA, Chow M, Machaca K.

2005. RNA interference is an antiviral defence mechanism in Caenorhab-ditis elegans. Nature 436:1044 –1047.

4. Schott DH, Cureton DK, Whelan SP, Hunter CP. 2005. An antiviral rolefor the RNA interference machinery in Caenorhabditis elegans. Proc. Natl.Acad. Sci. U. S. A. 102:18420 –18424.

5. Lu R, Yigit E, Li WX, Ding SW. 2009. An RIG-I-like RNA helicasemediates antiviral RNAi downstream of viral siRNA biogenesis in Caeno-rhabditis elegans. PLoS Pathog. 5:e1000286. doi:10.1371/journal.ppat.1000286.

6. Félix M-A, Ashe A, Piffaretti J, Wu G, Nuez I, Bélicard T, Jiang Y, ZhaoG, Franz CJ, Goldstein LD, Sanroman M, Miska EA, Wang D. 2011.Natural and experimental infection of Caenorhabditis nematodes bynovel viruses related to nodaviruses. PLoS Biol. 9:e1000586. doi:10.1371/journal.pbio.1000586.

7. Wang XH, Aliyari R, Li WX, Li HW, Kim K, Carthew R, Atkinson P,Ding SW. 2006. RNA interference directs innate immunity against virusesin adult Drosophila. Science 312:452– 454.

8. Jakubiec A, Yang SW, Chua N-H. 2012. Arabidopsis DRB4 protein inantiviral defense against Turnip yellow mosaic virus infection. Plant J.69:14 –25.

9. Qu F, Ye X, Hou G, Sato S, Clemente TE, Morris TJ. 2005. RDR6 has abroad-spectrum but temperature-dependent antiviral defense role inNicotiana benthamiana. J. Virol. 79:15209 –15217.

10. Dalmay T, Hamilton A, Rudd S, Angell S, Baulcombe DC. 2000. AnRNA-dependent RNA polymerase gene in Arabidopsis is required forposttranscriptional gene silencing mediated by a transgene but not by avirus. Cell 101:543–553.

11. Wang X-B, Wu Q, Ito T, Cillo F, Li W-X, Chen X, Yu J-L, Ding S-W.2010. RNAi-mediated viral immunity requires amplification of virus-derived siRNAs in Arabidopsis thaliana. Proc. Natl. Acad. Sci. U. S. A.107:484 – 489.

12. Garcia-Ruiz H, Takeda A, Chapman EJ, Sullivan CM, Fahlgren N,Brempelis KJ, Carrington JC. 2010. Arabidopsis RNA-dependent RNA

polymerases and Dicer-like proteins in antiviral defense and small inter-fering RNA biogenesis during turnip mosaic virus infection. Plant Cell22:481– 496.

13. Baulcombe D. 2002. Viral suppression of systemic silencing. Trends Mi-crobiol. 10:306 –308.

14. Guo HS, Ding SW. 2002. A viral protein inhibits the long range signalingactivity of the gene silencing signal. EMBO J. 21:398 – 407.

15. Molnar A, Melnyk CW, Bassett A, Hardcastle TJ, Dunn R, BaulcombeDC. 2010. Small silencing RNAs in plants are mobile and direct epigeneticmodification in recipient cells. Science 328:872– 875.

16. Dunoyer P, Schott G, Himber C, Meyer D, Takeda A, Carrington JC,Voinnet O. 2010. Small RNA duplexes function as mobile silencing sig-nals between plant cells. Science 328:912–916.

17. Saleh M-C, Tassetto M, van Rij RP, Goic B, Gausson V, Berry B,Jacquier C, Antoniewski C, Andino R. 2009. Antiviral immunity inDrosophila requires systemic RNA interference spread. Nature 458:346 –350.

18. Ketting RF, Fischer SE, Bernstein E, Sijen T, Hannon GJ, Plasterk RH.2001. Dicer functions in RNA interference and in synthesis of small RNAinvolved in developmental timing in C. elegans. Genes Dev. 15:2654 –2659.

19. Grishok A, Pasquinelli AE, Conte D, Li N, Parrish S, Ha I, Baillie DL,Fire A, Ruvkun G, Mello CC. 2001. Genes and mechanisms related toRNA interference regulate expression of the small temporal RNAs thatcontrol C. elegans developmental timing. Cell 106:23–34.

20. Lee RC, Hammell CM, Ambros V. 2006. Interacting endogenous andexogenous RNAi pathways in Caenorhabditis elegans. RNA 12:589 –597.

21. Tabara H, Yigit E, Siomi H, Mello CC. 2002. The dsRNA binding proteinRDE-4 interacts with RDE-1, DCR-1, and a DExH-box helicase to directRNAi in C. elegans. Cell 109:861– 871.

22. Zambon RA, Vikram NV, Louisa PW. 2006. RNAi is an antiviral im-mune response against a dsRNA virus in Drosophila melanogaster. Cell.Microbiol. 8:880 – 889.

23. Ghildiyal M, Zamore PD. 2009. Small silencing RNAs: an expandinguniverse. Nat. Rev. Genet. 10:94 –108.

24. Lu R, Maduro M, Li F, Li HW, Broitman-Maduro G, Li WX, Ding SW.2005. Animal virus replication and RNAi-mediated antiviral silencing inCaenorhabditis elegans. Nature 436:1040 –1043.

25. Steiner FA, Okihara KL, Hoogstrate SW, Sijen T, Ketting RF. 2009.RDE-1 slicer activity is required only for passenger-strand cleavage duringRNAi in Caenorhabditis elegans. Nat. Struct. Mol. Biol. 16:207–211.

26. Pak J, Fire A. 2007. Distinct populations of primary and secondary effec-tors during RNAi in C. elegans. Science 315:241–244.

27. Sijen T, Steiner FA, Thijssen KL, Plasterk RH. 2007. Secondary siRNAsresult from unprimed RNA synthesis and form a distinct class. Science315:244 –247.

28. Guo X, Lu R. 2013. Characterization of virus-encoded RNAi suppressorsin Caenorhabditis elegans. J. Virol. 87:5414 –5423.

29. Yigit E, Batista PJ, Bei Y, Pang KM, Chen CC, Tolia NH, Joshua-Tor L,Mitani S, Simard MJ, Mello CC. 2006. Analysis of the C. elegans Argo-naute family reveals that distinct Argonautes act sequentially duringRNAi. Cell 127:747–757.

30. Seth RB, Sun L, Chen ZJ. 2006. Antiviral innate immunity pathways. CellRes. 16:141–147.

31. Tijsterman M, May RC, Simmer F, Okihara KL, Plasterk RH. 2004.Genes required for systemic RNA interference in Caenorhabditis elegans.Curr. Biol. 14:111–116.

32. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC. 1998.Potent and specific genetic interference by double-stranded RNA in Cae-norhabditis elegans. Nature 391:806 – 811.

33. Winston WM, Molodowitch C, Hunter CP. 2002. Systemic RNAi in C.elegans requires the putative transmembrane protein SID-1. Science 295:2456 –2459.

34. Zhang C, Montgomery TA, Fischer SEJ, Garcia SMDA, Riedel CG,Fahlgren N, Sullivan CM, Carrington JC, Ruvkun G. 2012. The Caeno-rhabditis elegans RDE-10/RDE-11 complex regulates RNAi by promotingsecondary siRNA amplification. Curr. Biol. 22:881– 890.

35. Han W, Sundaram P, Kenjale H, Grantham J, Timmons L. 2008. TheCaenorhabditis elegans rsd-2 and rsd-6 genes are required for chromo-some functions during exposure to unfavorable environments. Genetics178:1875–1893.

36. Feinberg EH, Hunter CP. 2003. Transport of dsRNA into cells by thetransmembrane protein SID-1. Science 301:1545–1547.

Guo et al.

10728 jvi.asm.org Journal of Virology

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.

37. Shih JD, Hunter CP. 2011. SID-1 is a dsRNA-selective dsRNA-gatedchannel. RNA 17:1057–1065.

38. Jose AM, Garcia GA, Hunter CP. 2011. Two classes of silencing RNAsmove between Caenorhabditis elegans tissues. Nat. Struct. Mol. Biol. 18:1184 –1188.

39. Habig JW, Aruscavage PJ, Bass BL. 2008. In C. elegans, high levels ofdsRNA allow RNAi in the absence of RDE-4. PLoS One 3:e4052. doi:10.1371/journal.pone.0004052.

40. Guo X, Li W-X, Lu R. 2012. Silencing of host genes directed by virus-derived short interfering RNAs in Caenorhabditis elegans. J. Virol. 86:11645–11653.

41. Gu T, Orita S, Han M. 1998. Caenorhabditis elegans SUR-5, a novel butconserved protein, negatively regulates LET-60 Ras activity during vulvalinduction. Mol. Cell. Biol. 18:4556 – 4564.

42. Kamath RS, Ahringer J. 2003. Genome-wide RNAi screening in Caeno-rhabditis elegans. Methods 30:313–321.

43. Wu Q, Luo Y, Lu R, Lau N, Lai EC, Li W-X, Ding S-W. 2010. Virusdiscovery by deep sequencing and assembly of virus-derived small silenc-ing RNAs. Proc. Natl. Acad. Sci. USA 107:1606 –1611.

44. Parameswaran P, Sklan E, Wilkins C, Burgon T, Samuel MA, Lu R,Ansel KM, Heissmeyer V, Einav S, Jackson W, Doukas T, Paranjape S,Polacek C, dos Santos FB, Jalili R, Babrzadeh F, Gharizadeh B, GrimmD, Kay M, Koike S, Sarnow P, Ronaghi M, Ding S-W, Harris E, ChowM, Diamond MS, Kirkegaard K, Glenn JS, Fire AZ. 2010. Six RNAviruses and forty-one hosts: viral small RNAs and modulation of smallRNA repertoires in vertebrate and invertebrate systems. PLoS Pathog.6:e1000764. doi:10.1371/journal.ppat.1000764.

45. Sijen T, Fleenor J, Simmer F, Thijssen KL, Parrish S, Timmons L,Plasterk RH, Fire A. 2001. On the role of RNA amplification in dsRNA-triggered gene silencing. Cell 107:465– 476.

46. Aoki K, Moriguchi H, Yoshioka T, Okawa K, Tabara H. 2007. In vitroanalyses of the production and activity of secondary small interferingRNAs in C. elegans. EMBO J. 26:5007–5019.

47. Shih JD, Fitzgerald MC, Sutherlin M, Hunter CP. 2009. The SID-1double-stranded RNA transporter is not selective for dsRNA length. RNA15:384 –390.

48. Fukudome A, Kanaya A, Egami M, Nakazawa Y, Hiraguri A, MoriyamaH, Fukuhara T. 2011. Specific requirement of DRB4, a dsRNA-bindingprotein, for the in vitro dsRNA-cleaving activity of Arabidopsis Dicer-like4. RNA 17:750 –760.

49. Han Y-H, Luo Y-J, Wu Q, Jovel J, Wang X-H, Aliyari R, Han C, Li W-X,Ding S-W. 2011. RNA-based immunity terminates viral infection in adultDrosophila in the absence of viral suppression of RNA interference: char-acterization of viral small interfering RNA populations in wild-type andmutant flies. J. Virol. 85:13153–13163.

50. Palauqui J-C, Elmayan T, Pollien J-M, Vaucheret H. 1997. Systemicacquired silencing: transgene-specific post-transcriptional silencing istransmitted by grafting from silenced stocks to non-silenced scions.EMBO J. 16:4738 – 4745.

51. Voinnet O, Baulcombe DC. 1997. Systemic signalling in gene silencing.Nature 389:553.

52. Jose AM, Smith JJ, Hunter CP. 2009. Export of RNA silencing from C.elegans tissues does not require the RNA channel SID-1. Proc. Natl. Acad.Sci. U. S. A. 106:2283–2288.

RDE-4-Independent Antiviral Silencing

October 2013 Volume 87 Number 19 jvi.asm.org 10729

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Feb

ruar

y 20

22 b

y 95

.132

.249

.125

.