Negative-Strand Tospoviruses and Tenuiviruses Carry a Gene for a Suppressor of Gene Silencing at Analogous Genomic Positions

Abstract

Posttranscriptional silencing of a green fluorescent protein (GFP) transgene in Nicotiana benthamiana plants was suppressed when these plants were infected with Tomato spotted wilt virus (TSWV), a plant-infecting member of the Bunyaviridae. Infection with TSWV resulted in complete reactivation of GFP expression, similar to the case for Potato virus Y, but distinct from that for Cucumber mosaic virus, two viruses known to carry genes encoding silencing suppressor proteins. Agrobacterium-based leaf injections with individual TSWV genes identified the NS_S gene to be responsible for the RNA silencing-suppressing activity displayed by this virus. The absence of short interfering RNAs in NS_S-expressing leaf sectors suggests that the tospoviral NS_S protein interferes with the intrinsic RNA silencing present in plants. Suppression of RNA silencing was also observed when the NS3 protein of the Rice hoja blanca tenuivirus, a nonenveloped negative-strand virus, was expressed. These results indicate that plant-infecting negative-strand RNA viruses carry a gene for a suppressor of RNA silencing.

RNA silencing involves a sequence-specific degradation which is induced by overabundant RNA and by double-stranded RNA (dsRNA) molecules and which can target transgenes as well as homologous endogenous genes. RNA silencing was first described for plants (35, 50) and over recent years has been described for other organisms, where it is also referred to as cosuppression, posttranscriptional gene silencing (17), or RNA-mediated virus resistance (3, 11, 30) in plants, quelling in fungi (9), or RNAi in animals (19). Building blocks of the gene-silencing pathway proved to have remarkable similarities in the different organisms and hence suggest an ancient role of gene silencing in pathogen resistance or development (10, 25, 53). One of the key intermediary elements in the RNA silencing pathway is dsRNA, which is recognized by a dsRNA-specific nuclease (5) to yield small (21 to 23 nucleotides) short interfering RNAs (siRNAs) (21). These siRNAs subsequently serve as guides for cleavage of homologous RNA molecules. In plants, versions of transgenes that produce dsRNA molecules have been shown to be very potent activators of RNA silencing (47). As all RNA viruses replicate through formation of dsRNA intermediates, these are potential targets of the RNA silencing mechanism. Indeed, antiviral RNA silencing has been shown to occur in nature and has been proposed as a natural defense mechanism protecting plants against viruses, resulting in resistance (1, 43).

To counteract the RNA silencing mechanism of their host, plant viruses have developed ways to evade or neutralize this response. Over recent years, RNA silencing-inhibiting proteins have been identified in several plant viruses. Among the best-studied examples are the helper component-proteinase (HC-Pro) of the potyvirus Potato virus Y (PVY) and the 2b protein of Cucumber mosaic virus (CMV) (7, 40). Other plus-strand RNA (and some DNA) viruses also have been found to suppress gene silencing, and for some of them the viral protein involved was identified (16, 37, 46, 54). The viral suppressor proteins of PVY and CMV act differently by targeting different steps in the RNA silencing pathway. HC-Pro was shown to prevent degradation of dsRNA into siRNAs, which are considered a hallmark of RNA silencing (21); however, it did not prevent the silencing signal from becoming systemic (7, 33). The 2b protein of CMV, which is required for long-distance movement, is targeted to the nucleus (32) and prevents silencing in newly emerging tissues, but unlike HC-Pro it is not able to reverse RNA silencing once it is established (4). Interestingly, the required nuclear accumulation of 2b for efficient suppression of RNA silencing indicated a possible blocking of silencing from the nucleus while RNA degradation takes place in the cytoplasm. A third step of the silencing pathway, namely, systemic signaling of silencing, is targeted by the Potato virus X p25 protein (54). Recently, the P19 protein of tombusviruses was implicating in inhibiting RNA silencing by physically interacting with siRNAs, thus providing another mechanism to interfere with RNA silencing (46).

An increasing number of positive-strand RNA viruses of plants have been shown to counteract the RNA silencing defense system. Although they are often transmitted by insects or fungi, plant-infecting plus-strand RNA viruses replicate exclusively in plant hosts. In contrast, plant-infecting negative-strand RNA viruses, i.e., tospoviruses, tenuiviruses, and rhabdoviruses, have life cycles in both plants and their arthropod vectors. (18, 23, 56). Recently, RNA silencing has also been demonstrated in insects (29, 36). Negative-strand viruses therefore have to cope with the RNA silencing defense systems of both the plant and insect hosts. For this reason, it is of interest to investigate whether negative-strand RNA viruses infecting plants also encode RNA silencing suppressors, which enable them to overcome the plant and insect intracellular defense responses. To test this, two representatives of the genera Tospovirus and Tenuivirus, i.e., Tomato spotted wilt virus (TSWV) and Rice hoja blanca tenuivirus (RHBV), respectively, have been investigated for this feature.

Like those of all bunyaviruses, the genome of TSWV is tripartite, of which the fully negative-stranded L RNA encodes the viral RNA-dependent RNA polymerase (RdRP) (Fig. 1). Both the M and S RNAs have two genes, in an ambisense arrangement (12, 13, 27). The M RNA codes for the precursor to the membrane glycoproteins G1 and G2 and the viral movement protein NS_M (48). The S RNA codes for the nucleoprotein (N) and a nonstructural protein (NS_S). No clear function could be assigned to the NS_S protein, although as accumulation of this protein coincides with increase of symptom severity, it has been implicated in viral virulence (26, 28). Like for TSWV, the largest RNA segment of RHBV (RNA 1) is of complete negative polarity and encodes the putative viral polymerase. The other three RNA segments have an ambisense coding strategy, thus encoding a further six proteins, of which only the function of the nucleoprotein encoded on RNA 3 has been indicated (14, 15, 41, 42). The genome arrangements of TSWV and RHBV are indicated in Fig. 1.

FIG. 1.

Schematic representation of the coding strategies of the TSWV and RHBV genomes. The dashed line surrounding the tenuivirus RHBV indicates that no membranous particles have been found for these viruses, in contrast to the case for tospoviruses and other members of the Bunyaviridae. Viruses of both genera have a fully negative-stranded large RNA (L RNA and RNA 1, respectively), encoding the viral RdRP. The nucleoprotein is encoded on the virus cRNA strand of the third-largest segment (S RNA or RNA 3), in which the NS_S and NS3 proteins are encoded on the viral RNA strand. The remaining RNA segments of both viruses are all ambisense.

Using a testing system previously applied to identify silencing suppressors of positive-strand RNA viruses, i.e., a combination of green fluorescent protein (GFP)-silenced reporter plants and Agrobacterium-based transient transformation (55), we investigated the occurrence of possible silencing suppression genes of TSWV and RHBV.

MATERIALS AND METHODS

Plants and viruses.

Transgenic Nicotiana benthamiana plants harboring a GFP transgene expressed from a 35S promoter-nopaline synthase terminator expression cassette were used (22). Transgenic lines were selected for strong GFP fluorescence prior to self-pollination. Subsequent S1 plants were scored for gene silencing by checking for GFP expression in meristematic tissues in otherwise nonexpressing (RNA-silenced) plants. S2 progenies of these plants were homozygous, and all showed a silenced phenotype resulting in silencing of GFP expression in leaf tissue. Complete GFP silencing in veins and stems was reached after several weeks. These S2 plants were used in the inoculation experiments.

TSWV isolate BR-01, Groundnut ringspot virus (GRSV) isolate SA-05, and Impatiens necrotic spot virus (INSV) isolate NL-07 were inoculated in series on both GFP-silenced and nontransgenic N. benthamiana plants acting as controls. For reference, the potyviruses PVY and Cowpea aphid-borne mosaic virus (CABMV) (34), as well as two different CMV isolates (CMV-Lily and CMV-Alstroemeria, belonging to subgroups I and II, respectively [8]), were used in these experiments.

Inoculation was performed in the greenhouse at the four- to six-leaf developmental stage. Systemically infected top leaves were homogenized in 10 mM sodium phosphate (pH 7.2) with 0.1% NaSO₃ added by using a mortar and pestle. Each plant was inoculated on two leaves, using carborundum powder as an abrasive agent. A sponge was used to apply the inoculum on the leaf.

Agrobacterium clones and agroinfiltration.

An expression vector harboring an expression cassette consisting of a 35S promoter, Tobacco mosaic virus 5′ untranslated region, multiple cloning site, and nopaline synthase terminator was used. Expression cassettes carried individual TSWV genes; nucleoprotein N, movement protein NS_M, glycoprotein G1G2 precursor, and NS_S have been previously described (39). GFP, CABMV HC-Pro protein (34), and CMV (subgroup I) 2b protein also were used in similar expression cassettes. The NS3 gene of RHBV was cloned by reverse transcription-PCR, using primers containing the respective start and stop codons of the gene plus appropriate restriction sites for cloning. Clones thus obtained were verified by sequence analysis (results not shown). Expression cassettes were cloned in the binary vector pBIN19 and subsequently introduced in Agrobacterium tumefaciens (strain LBA4404) by using triparental mating. Agrobacterium T-DNA transient-expression assays in N. benthamiana plants were performed by (co)-infiltrating at least two locations on the basal side of the leaf with Agrobacterium suspensions by using a 5-ml syringe without a needle. Cultures were grown overnight at 28°C from individual colonies in 2 ml of YEB medium (0.5% beef extract, 0.1% yeast extract, 0.5% peptone, 0.5% saccharose, 2 mM MgSO₄) containing 20 μg of rifampin per ml and 50 μg of kanamycin per ml. Four hundred microliters of cell culture was pelleted by centrifugation and resuspended in 2 ml of induction medium [10.5 g of K₂HPO₄ per liter, 4.5 g of KH₂PO₄ per liter, 1.0 g of (NH₄)₂SO₄ per liter, 1 mM MgSO₄, 0.2% (wt/vol) glucose, 0.5% (vol/vol) glycerol, 50 μM acetosyringone, and 10 mM MES (morpholineethanesulfonic acid) (pH 5.6)]. After overnight incubation at 28°C, cells were pelleted again and washed in Murashige-Skoog (MS) medium containing 10 mM MES (pH 5.6). Cells were resuspended to an optical density at 600 nm of 0.5 in Murashige-Skoog medium-MES with 150 μM acetosyringone. Young, fully expanded leaves were used for agroinfiltration and covered with plastic for 2 to 3 days in the greenhouse. Plants were subsequently monitored for GFP reactivation by using a hand-held 125-W UV lamp. Expression generally reached a maximum level after 3 to 4 days.

UV photography.

Pictures of whole plants (as shown in Fig. 2A) were made with a digital camera (Kodak DCS professional series) by using a hand-held 125-W UV lamp (Philips HPW 125W-T) and a 30-s exposure time. UV pictures at leaf level were made with 35-mm Kodak 200 ASA film by using a black box carrying two small UV lamps (366 nm). For the leaves shown in Fig. 2B, the exposure time was 2 min and a Kodak Wratten no. 58 filter was used. The leaves in Fig. 2A were exposed for 1 min without a filter. Close-up UV pictures as shown in Fig. 3 were made with a digital camera (CoolSnap; combined red and green channels) using a binocular stereomicroscope (M3Z; Leica). The GFP imaging photographs in Fig. 4 and 5A were taken with a yellow 022 B+W filter from Proline. Variable exposure times were used, depending on the intensity of the fluorescence.

FIG. 2.

Virus inoculation of transgenic N. benthamiana plants containing a silenced GFP gene. (A) Suppression of RNA silencing by TSWV. The plant on the left is a nontransgenic plant infected with TSWV to show that the infection does not cause autofluorescence. The plant on the right is a noninfected, GFP-silenced control plant. The photograph was taken without a filter. (B) Recovery of GFP expression in GFP-silenced plants by infection with TSWV or PVY at 10 days postinoculation. (C) Suppression of GFP expression in older leaves. Photographs were taken at 3 weeks postinoculation. In panels B and C, a Kodak Wratten no. 58 filter was used.

FIG. 3.

Agrobacterium infiltration experiments with different TSWV genes in GFP-silenced N. benthamiana plants. Agrobacterium strains harboring TSWV genes were coinfiltrated with GFP. Only NS_S suppresses the silencing of GFP (panel D).

FIG. 4.

RNA silencing suppression activity displayed by HC-Pro (PVY), NS_S (TSWV), and NS3 (RHBV). The picture was taken with a yellow filter at 6 days after infiltration.

FIG. 5.

(A) GPF imaging of Agrobacterium-infiltrated leaves from nontransgenic plants. GFP expression was visualized by UV light in leaves coinfiltrated with GFP and an empty vector, CABMV HC-Pro, TSWV NS_S, RHBV NS3, or CMV 2b. (B) Total protein was extracted from corresponding infiltrated leaf sectors. Western blotting was performed with anti-GFP antibodies. (C) Northern blot analysis of total mRNA extracted from the infiltrated leaf parts. Ethidium bromide staining of the same gel shows the 25S rRNA as a loading control.

Molecular analyses.

Northern blot analyses were performed by standard protocols with ³²P-radiolabeled full-length GFP PCR products. Western blot analysis of GFP and NS_S was performed with polyclonal rabbit antiserum.

Isolation and enrichment of small RNAs was performed as described by Hamilton and Baulcombe (21) Detection of siRNAs was performed by RNase A/T₁ protection assays as described by Sijen et al. (45) with in vitro-transcribed GFP RNA probes.

RESULTS

TSWV infection counteracts RNA silencing.

TSWV is capable of infecting many host plants, in many cases causing severe disease symptoms (20, 38, 51). We therefore sought to investigate whether the virulence of TSWV may be enhanced by its ability to suppress gene silencing. For this purpose, a series of virus inoculation experiments were performed on N. benthamiana plants in which a GFP transgene was silenced. Transgenic N. benthamiana plants were inoculated with TSWV, and GFP fluorescence was monitored. Two positive-strand viruses known to effectively suppress gene silencing (PVY and CMV) were used as controls in these experiments. In the transgenic plant line used in these experiments, the silencing of GFP fluorescence occurs only in leaf parenchyma, whereas stems and major veins still express GFP (Fig. 2A). After further aging of the plant (several weeks), stems and veins also become silenced resulting, in totally silenced plants showing no fluorescence.

Starting at around 6 days postinoculation, and concomitant with the occurrence of virus symptoms, all TSWV-infected plants (26 replicates) showed a complete reversal of GFP silencing and became highly GFP fluorescent in infected tissue (Fig. 2A). GFP expression in mock-inoculated control plants remained silenced, indicating that the observed fluorescence is GFP specific and not related to fluorescence of necrotized tissue. Inoculation of the (GFP-transgenic) test plants with PVY and CMV also resulted in a reversion of the GFP-silenced phenotype, although the effect of CMV was markedly less than that of PVY or TSWV. A clear relationship between viral symptoms and GFP expression was visible by distinct sectoring of GFP reactivation. For both PVY and TSWV, this phenomenon was observed at an early stage of infection (Fig. 2B). At a later stage (3 weeks postinoculation) not only newly developed systemically infected leaves but also older leaves were completely suppressed in GFP silencing (Fig. 2C). Similar inoculation experiments performed with two additional tospovirus species, Groundnut ringspot virus and Impatiens necrotic spot virus, also revealed a reversal of gene silencing, indicating that silencing suppression is a general feature of tospoviruses (data not shown).

Inoculation of GFP-silenced plants with two different CMV isolates did not result in immediate suppression of GFP silencing. Strikingly, after 3 weeks, only older leaves infected with CMV subgroup I started to show some suppression of GFP silencing coinciding with viral symptoms. However, this was restricted to the older leaves and was not observed in young leaves. Only after 1 month did newly emerging leaves of plants infected with CMV subgroup II start to show RNA silencing suppression patterns similar to those observed by Brigneti and coworkers (7).

The NS_S protein of TSWV is necessary and sufficient to confer suppression of RNA silencing.

To investigate a possible function for the individual TSWV proteins in the suppression of gene silencing, four of its genes, i.e., those for N, NS_S, G1G2, and NS_M, were cloned under the control of a 35S promoter and tested. Clones of HC-Pro of the potyvirus CABMV (34) and of 2b of a subgroup I CMV isolate (8) also were introduced in transgenic N. benthamiana leaves expressing the GFP transgene by the Agrobacterium-based delivery system using leaf infiltration (7). The TSWV gene constructs were injected in leaves together with an Agrobacterium strain carrying the GFP gene to initiate and enhance RNA silencing (55). Suppression of GFP silencing was monitored during the following days, and leaves were photographed 6 days after injection (Fig. 3). Only coinfiltration with the NS_S gene led to suppression of GFP silencing in all of the 16 treated plants (Fig. 3D and 4). Expression of the TSWV proteins was confirmed by Western blot analysis (data not shown). In none of these cases did the suppression of silencing spread beyond the inoculation focus, indicating that the suppressor protein is unable to move from cell to cell or induce a mobile silencing suppression signal.

The other TSWV genes, when cointroduced with GFP (approximately 15 plants each), as well as GFP alone, did not show GFP fluorescence in the injected areas (Fig. 3A to C). This indicates that both the original and the A. tumefaciens-delivered GFP (transient) transgenes were completely silenced and that the TSWV proteins N, NS_M, and G1G2 are not involved in the observed inhibition of gene silencing during virus infection.

The NS3 protein of tenuivirus RHBV also is a suppressor of RNA silencing.

Like the NS_S gene of TSWV, the NS3 gene of RHBV is located on the third-largest RNA segment (RNA 3), which furthermore encodes the nucleoprotein on the viral complementary strand (Fig. 1). So far, no function has been assigned to this protein, and it does not contain any protein sequence homology to NS_S or any other protein in the National Center for Biotechnology Information database. In order to investigate whether the NS3 protein may also be involved in silencing suppression, its gene was cloned into a plant expression cassette, like its positional tospoviral analogue NS_S. As shown in Fig. 4, the transient expression of NS3 reversed the effect of gene silencing, resulting in a phenotype very similar to those obtained with HC-Pro and NS_S.

Both NS_S and NS3 interfere with the initiation of RNA silencing.

The efficient reversal of GFP silencing by HC-Pro, NS_S, and NS3 in GFP-transgenic plants prompted the question whether NS_S and NS3 may be involved in suppressing early stages of the RNA silencing cascade (33). To investigate this question, untransformed N. benthamiana plants were infiltrated with combinations of Agrobacterium strains containing GFP and strains harboring silencing suppressor genes. The effect of the other TSWV genes was used as a control. When introduced together with the TSWV N, G1G2, or NS_M gene, the GFP gene remained silenced, similar to the case for plants where only GFP was infiltrated (not shown). However, when introduced together with the NS_S or NS3 gene construct, local expression of the GFP transgene was significantly boosted, yielding a much higher fluorescence (Fig. 5A) and enhanced protein content (Fig. 5B). Northern blot analysis showed significantly higher levels of GFP mRNA when coexpressed with NS_S, NS3, and HC-Pro (Fig. 5C). Two CMV 2b genes belonging to different subgroups were also coinfiltrated in nontransgenic plants together with GFP but showed no markedly enhanced GFP expression compared to that seen with the coinfiltration of GFP with the empty control plasmid (Fig. 5A).

Analysis of siRNAs.

To further substantiate that reversion of GFP expression in transgenic, silenced plants was based on genuine suppression of RNA silencing, the occurrence of siRNAs in GFP-expressing tissues was investigated. RNA was extracted from Agrobacterium-infiltrated GFP-expressing leaf sectors and enriched for siRNAs. Subsequently, siRNAs were analyzed by RNase protection assays. GFP-specific small RNAs were readily detected in GFP-infiltrated transgenic plants. As previously reported (31, 33) the siRNAs did not appear when GFP was coinfiltrated with HC-Pro. Exactly the same was observed when NS_S was coinfiltrated (Fig. 6).

FIG. 6.

Analysis of the GFP siRNAs by RNase A/T₁ protection assay. Small RNA-enriched samples were extracted from Agrobacterium-infiltrated sectors of GFP-silenced leaves. Noninfiltrated nontransgenic plants were used as a control. All other leaves were coinfiltrated with GFP and (putative) silencing suppressor constructs. Twenty micrograms of RNA was hybridized to sense GFP RNA transcripts and subsequently treated with the RNases A and T₁. Sizes of RNA oligonucleotides are indicated on the right in bases.

The infiltration of NS3 protein results in enhanced GFP fluorescence (Fig. 4 and 5), due to a great increase in mRNA levels (Fig. 5C), suggesting a strong RNA silencing suppression. However, unlike NS_S and HC-Pro, NS3 did not prevent the accumulation of small RNAs, indicating that it may have a mode of action different from that of HC-Pro or NS_S in N. benthamiana plants.

DISCUSSION

Upon infection with the tospovirus TSWV, transgenic plants silencing the GFP transgene show a strong recovery of GFP fluorescence, suggesting that RNA silencing is suppressed in the infected plant cells. The suppression of silencing is similar to that observed with potyviruses, resembling the suppression capabilities of these viruses both in time and in intensity. Like TSWV, Groundnut ringspot virus and Impatiens necrotic spot virus also showed suppressor capabilities, indicating that the suppression of silencing is a general feature of tospoviruses.

As suppression of GFP silencing closely resembles that of the potyvirus PVY, it is tempting to speculate the underlying mechanism by which tospoviruses suppress RNA silencing may also be similar. PVY is known to counteract silencing by targeting a maintenance step in the gene silencing machinery (33), in contrast to CMV, which targets the initiation of silencing (7). The results strongly suggest that tospoviruses mimic the ability of PVY to revert established silencing of a GFP transgene by targeting the maintenance step of the gene silencing machinery.

Transient assays were used to identify the TSWV gene responsible for suppression, making it possible to introduce genes independent of virus or viral vector. Possible complications induced by virus infection or additional suppressors encoded by the viral vector used were circumvented in this way, resulting in a gene function assay free of cytopathogenic interference by the virus. Agrobacterium-mediated introduction of TSWV genes in silenced plants has demonstrated that the nonstructural protein NS_S is necessary and sufficient for the silencing suppression. Formally, though, it cannot be excluded that the tospoviral RNA-dependent RNA polymerase (L protein) may play a supporting role, as this large (9-kb) gene was not included in our experiments.

Hitherto, no function of the NS_S protein during the virus infection cycle in plants or insects could be assigned. In previous reports it had been shown to be highly expressed in the cytoplasm of infected cells and was suggested to play a role in the virulence of the virus, as more-virulent isolates accumulate larger amounts of this protein (26, 28). In addition, it accumulates to high levels in salivary glands of thrips, the insect involved in the transmission of the virus, in which the virus also replicates (56). Recently, it was demonstrated that the insect virus Flock house virus also carries a gene which can induce RNA silencing suppression, albeit this was tested in a plant assay (29). Hence, the NS_S protein may perform its suppression function in both its plant host and insect vector. High accumulation of this protein in the salivary glands of viruliferous thrips (56) may support this notion.

The results presented in this paper also demonstrate that RHBV, belonging to another taxon of negative-strand RNA viruses (the floating genus Tenuivirus), specifies an RNA silencing suppressor. The Agrobacterium infiltration experiments resulted in exactly the same suppressing phenotype for HC-Pro, NS_S, and NS3. To dissect the mode of action of these novel RNA silencing suppressors, infiltrated leaf tissues were analyzed for the presence of GFP transcripts and for the presence of GFP-specific siRNAs by Northern blot analysis. These experiments show that both HC-Pro and NS_S eliminate GFP-specific siRNAs, consistent with their proposed mode of action. However, NS3 does not eliminate the GFP-specific siRNAs, whereas the increased amount of GFP mRNA indicates that the GFP mRNAs are protected from RNA silencing. It can be concluded that NS3 achieves the same strong silencing suppression function as HC-Pro and NS_S, but by another mechanism. This is possibly realized either by directly protecting (a part of) the mRNA population from degradation or indirectly by sequestering the siRNAs or interfering with the action of RNA silencing complexes, similar to the P19 silencing suppressor of the tombusvirus CymRSV (46). Another rice virus, Rice yellow mottle sobemovirus, has been indicated to encode a silencing suppressor protein (P1) in a similar assay in tobacco (55). For P1 also, the increase of GFP mRNA was indicative of silencing suppression. The boost of transgene expression by NS_S, NS3, and HC-Pro may be achieved by the physical protection of mRNAs either directly or by interfering with later phases of the gene silencing machinery such as maintenance, causing the suppression of the silencing phenomenon. The latter would imply that gene silencing could be very rapidly evoked, even when a transgene is expressed transiently from an Agrobacterium inoculation, as also suggested by Johansen and Carrington (24).

Interestingly, Agrobacterium-mediated introduction of NS_S, NS3, and HC-Pro suppressor proteins, but not CMV 2b, together with GFP in nontransgenic N. benthamiana plants resulted in increased GFP protein expression and fluorescence compared to introduction of GFP alone. A similar observation can be made for the effect of the potyviral HC-Pro on GUS expression in results presented by Llave and coworkers (31).

Recent findings with silencing-indispensable host genes indicate that RNA silencing interacts or even participates in gene regulation and plant development (49). The influence of RNA silencing suppressors on the generation of micro-RNAs, regulatory RNA elements resembling siRNAs (reviewed in reference 44), also may play a specific role in the altered regulation of genes in these plants. This is supported by the observation that plants overexpressing HC-Pro show abnormal development (2, 34) and the fact that despite numerous attempts it has not been possible to regenerate plants or even shoots transgenically expressing detectable amounts of NS_S protein (39).

In this paper, representatives of two different taxa of negative-strand plant viruses, TSWV and RHBV, have been shown to carry genes for suppressors of gene silencing in plants, indicating that this feature is present also in negative-strand viruses. It will be interesting to investigate whether these proteins also play a role during virus replication in their insect vectors, even more so as RNA silencing has been demonstrated in insects (36).

The observation that both genes responsible for silencing suppression are carried on the same position of the TSWV and RHBV genomes, i.e., on the third-largest segment in an ambisense arrangement opposite the N gene (Fig. 1), further underscores the genetic interrelationship between tospo- and tenuiviruses. Whether this analogy can be further extended to related animal-infecting bunyaviruses that are also transmitted by insects but infect vertebrates remains an open question. Interestingly, similar to the case for TSWV, the NS_S genes of both phleboviruses and orthobunyaviruses have been implicated in virulence of the virus (6, 52). The NS_S gene of phleboviruses, a genus that shares an ambisense S RNA with the tospoviruses, is a particularly intriguing candidate for further study.