SEARCH

SEARCH BY CITATION

Keywords:

  • ARGONAUTE;
  • auxin response factor;
  • post-transcriptional gene silencing;
  • RNAi;
  • root development

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The cucumber mosaic virus (CMV) 2b protein suppresses RNA silencing and determines viral symptoms. Among Arabidopsis thaliana lines expressing 2b proteins from mild (LS and Q CMV) or severe (Fny CMV) strains, only Fny 2b-transgenic plants displayed strong symptom-like phenotypes in leaves, stems and flowers, together with stunting of main root growth and increased emergence of lateral roots. However, LS and Fny 2b proteins both enhanced lateral root length. Micro (mi)RNA-mediated cellular mRNA turnover was inhibited in Fny 2b-transgenic plants, but there was no evidence for this in LS 2b-transgenic plants. Both 2b proteins efficiently suppressed small interfering (si)RNA-mediated RNA silencing, suggesting that 2b proteins can target the siRNA pathway without disrupting miRNA-regulated RNA turnover. Thus, symptom induction is not an inevitable consequence of RNA silencing suppression. For CMV, strain-specific differences between the 2b silencing proteins determine whether only one or both small RNA-guided RNA destruction pathways are disrupted.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

MicroRNAs (miRNAs) play vital roles in regulating plant development (Kidner and Martienssen, 2005). Formation of diverse structures, including the embryo, roots, vasculature and floral tissue, is in large part controlled by the activities of miRNAs (Guo et al., 2005; Kim et al., 2005; Mallory et al., 2004, 2005). Furthermore, a number of miRNAs operate as part of hormone response pathways, for example in the pathway controlling auxin-mediated lateral root development (Guo et al., 2005; Mallory et al., 2005; Sunkhar and Zhu, 2004). Thus, miRNAs are probably key factors in the integration of the hormonal and developmental control programs in plants.

miRNAs are short (21–24 nt), single-stranded RNA molecules derived from transcripts of endogenous plant loci (Llave et al., 2002a; Park et al., 2002; Reinhart et al., 2002). They are processed from polyadenylated precursors of approximately 1 kb length (pri-miRNAs), by the enzyme dicer-like 1 (DCL1; Park et al., 2002; Reinhart et al., 2002). Mature miRNAs, incorporated into an RNA-induced silencing complex (RISC), target endogenous plant transcripts for degradation or translational repression in a sequence-specific manner (Aukerman and Sakai, 2003; Llave et al., 2002b). The miRNA ‘pathway’ for post-transcriptional gene regulation overlaps to some extent with that controlling cytoplasmic RNA silencing. This latter sequence-specific RNA degradation pathway is directed by short interfering RNAs (siRNAs), which are generated de novo from double-stranded RNAs in the cytoplasm. It is thought that the primary function of RNA silencing is to protect eukaryotic organisms from viruses and retrotransposons (Ratcliff et al., 1997, 1999; Voinnet, 2005).

To overcome siRNA-directed destruction of viral RNA, most plant viruses express proteins that possess at least some ability to counteract RNA silencing (Palukaitis and MacFarlane, 2006; Voinnet, 2005). The first RNA silencing suppressors to be identified were the 2b protein of cucumber mosaic virus (CMV) and the potyviral P1/HC-Pro protein (Anandalakshmi et al., 1998; Brigneti et al., 1998; Kasschau and Carrington, 1998). Several RNA silencing suppressors, including P1/HC-Pro, the 2b protein, p19 (encoded by tombusviruses) and p38 (encoded by turnip crinkle virus), also function as symptom determinants during viral infection (Scholthof et al., 1995; Thomas et al., 2003; Voinnet et al., 1999). When expressed in transgenic plants, several of these RNA silencing suppressors disrupt normal plant development by interfering with miRNA-mediated regulation of turnover of specific host mRNAs (Chapman et al., 2004; Chellappan et al., 2005; Chen et al., 2004; Dunoyer et al., 2004). It is assumed that they target common elements of the miRNA and siRNA ‘pathways’. Surprisingly, although the CMV 2b protein is also a strong suppressor of RNA silencing, Chapman et al. (2004) found that expression of a 2b gene in transgenic plants caused no apparent disruption of the miRNA ‘pathway’ and little or no obvious alteration in phenotype.

The 2b protein plays crucial roles in the evasion of host defenses by CMV. It prevents induction of RNA silencing (Béclin et al., 1998; Brigneti et al., 1998), possibly by inhibiting the propagation of a silencing signal (Guo and Ding, 2002), and may interfere with siRNA-mediated translational repression (Qi et al., 2004). Additionally, the 2b protein allows CMV to overcome the inhibition of viral replication and cell-to-cell movement induced by salicylic acid (Ji and Ding, 2001; Naylor et al., 1998).

The CMV 2b protein exerts other effects on the viral infection process. A number of groups have shown, by deletion or mutation of the 2b gene, that the 2b protein determines the rate of viral movement over long distances through the host, at least in certain plant species (Ding et al., 1995a; Soards et al., 2002; Wang et al., 2004). Local cell-to-cell movement dynamics are also altered by deletion of the 2b gene (Shi et al., 2003; Soards et al., 2002). Symptom severity is also influenced by the 2b protein, and is dependent upon specific domains within the protein sequence. The disease symptoms induced by CMV range from mild to severe depending upon the host species and viral strain combination, and can include stunting and developmental abnormalities (Palukaitis and García-Arenal, 2003; Palukaitis et al., 1992). Replacing the 2b gene of a mild CMV strain with that of a more virulent strain created a hypervirulent virus, whilst deletion of the 2b gene rendered the virus symptomless (Ding et al., 1996; Shi et al., 2002; Soards et al., 2002). Furthermore, deletion of nuclear localization signals within the 2b protein sequence attenuated the symptoms induced by the mutant virus, as well as compromising its ability to inhibit RNA silencing (Lucy et al., 2000; Wang et al., 2004).

In contrast to an earlier report (Chapman et al., 2004), we have found that constitutive expression of a CMV 2b protein in transgenic Arabidopsis thaliana plants can induce symptom-like phenotypes in the aerial tissues as well as alterations in root development. Interestingly, the severity of the phenotype induced by the 2b transgene correlated with the severity of the viral strain from which the 2b protein sequence had been cloned and its ability to perturb miRNA-regulated transcript turnover, but not with its activity as a suppressor of siRNA-mediated silencing.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Development is disrupted in transgenic plants expressing 2b transgenes

It appeared paradoxical that while work with engineered CMV strains containing site-directed mutations or deletions in the 2b gene showed that the 2b protein is a symptom determinant (Ding et al., 1995a; Soards et al., 2002; Wang et al., 2004), expression of a 2b transgene in A. thaliana did not induce any apparent phenotype alterations (Chapman et al., 2004). This suggested either that the 2b protein induces symptoms only when expressed in conjunction with other viral protein(s) or RNAs, or that the 2b proteins of different CMV strains vary in their ability to elicit symptoms. To distinguish between these possibilities, we generated transgenic A. thaliana lines harboring 2b genes derived from three CMV strains: the severe Fny strain and two mild strains, LS and Q. These sequences were placed under the control of the constitutive 35S RNA promoter from cauliflower mosaic virus. In some cases, transgenes included the sequence of the CMV 3′ non-translated region (3′ NTR) in addition to the 2b open reading frame (ORF).

The aerial parts of plants harboring Fny 2b or Fny 2b + 3′ NTR transgenes exhibited striking phenotypes that clearly distinguished them from non-transformed A. thaliana: stunting, twisted petioles, serration and upturning of leaf borders (Figure 1a,b versus Figure 1f). Furthermore, the sepals and petals of flowers were narrowed and separated (Figure 1k–m). Some trichome development was noted on the underside of leaves, indicating potential disruption of adaxial/abaxial patterning or of salicylic acid/jasmonic acid-based signaling (Traw and Bergelson, 2003). In contrast, no clearly altered phenotypes were apparent in the corresponding organs of plants harboring the LS 2b, LS 2b + 3′ NTR or Q 2b transgenes (Figure 1c–e versus Figure 1f). The difference in phenotype severity is consistent with the difference in symptom induction between Fny and LS CMV infections, where the Fny strain causes greater stunting and deformation (Figure 1g versus Figure 1h). Severe stunting and deformation of leaves was most consistently seen in plants harboring the Fny 2b + 3′ NTR transgene (Figure 1n).

image

Figure 1.  Fny 2b protein induces strong phenotypes in aerial tissues of Arabidopsis thaliana. A. thaliana (ecotype Col-0) was transformed with Fny 2b, Fny 2b + 3′ NTR, LS 2b, LS 2b + 3′ NTR or Q 2b transgene sequences taken from the appropriate strain (Fny, LS or Q) of cucumber mosaic virus (CMV) and expressed under the control of the constitutive cauliflower mosaic virus 35S promoter. (a–j) Phenotypes or symptoms expressed in vegetative tissues by 2b transgenic or CMV-infected A. thaliana plants. (a) A typical Fny 2b transgenic plant exhibiting stunting, leaf distortion and twisting of petioles. (b) A typical Fny 2b + 3′ NTR transgenic plant exhibiting severe stunting, leaf distortion and twisting of petioles. (c) A typical LS 2b transgenic plant, exhibiting no phenotype alteration. (d) A typical LS 2b + 3′ NTR transgenic plant, exhibiting no phenotype alteration. (e) A typical Q 2b transgenic plant, exhibiting no phenotype alteration. (f) A wild-type, non-transgenic, non-infected A. thaliana plant. (g) A wild-type, non-transgenic A. thaliana plant infected with LS CMV. (h) A wild-type, non-transgenic A. thaliana plant infected with Fny CMV, showing stunting deformation compared to the plant shown in (f). (i) A typical Fny unt2b transgenic plant in which the three potential start codons within the 2b ORF had been mutated to stop codons. These plants exhibited no phenotype alteration. (j) A typical Fny unt2b + 3′ NTR transgenic plant in which the three potential start codons for the 2b ORF had been mutated to stop codons. These plants exhibited no phenotype alteration. (k–m) Floral phenotypes exhibited by plants transformed with Fny 2b transgenes. Flowers of plants harboring 2b-derived transgenes exhibited narrowing and separation of petals and sepals, as well as shortening of petals, compared with wild-type flowers. (k) A wild-type A. thaliana flower, photographed at 20 × magnification. (l) A flower from an Fny 2b transgenic plant, photographed at 30 × magnification. (m) A flower from an Fny 2b + 3′ NTR transgenic plant, photographed at 20 × magnification. (n) Independent, confirmed T0 generation transformants harboring the Fny 2b (top tray) and Fny 2b + 3′ NTR (bottom tray) transgenes exhibited a range of phenotype severities. However, the phenotypes exhibited by plants with the Fny 2b + 3′ NTR transgene were consistently more severe than those transformed with the Fny 2b transgene. Plants in panels (a)–(j) and (n) were photographed at approximately 2 months old. Scale bars in these panels represent 1 cm.

Download figure to PowerPoint

Roots of 2b-transgenic plants also exhibited specific phenotypes (Figure 2a,b). We examined nine independent transformed lines in detail, and data from one experiment (from a total of three similar experiments) are presented in Table 1. Of the 2b-transgenic lines examined, three of five lines harboring the Fny 2b-derived transgenes exhibited significantly shorter main roots, as well as significantly more lateral roots per centimeter of main root, when compared statistically to wild-type controls. None of four lines examined harboring LS 2b-derived transgenes exhibited statistically significant changes in main root length or in numbers of lateral roots per centimeter of main root. In contrast, the Fny 2b, Fny 2b + 3′ NTR and LS 2b + 3′ NTR transgenes all induced statistically significant increases in lateral root length (in five of eight lines examined). These data suggest that the severe strain Fny 2b gene induces stunting of main roots and an increase in the number of lateral roots initiated, whilst the mild strain LS 2b gene does not. However, both mild (LS) and severe (Fny) strain 2b genes can induce an increase in lateral root length. Data from all three replicates, including measurements from wild-type control plants on each plate and P-values derived from statistical tests, are shown in Tables S1–S3.

image

Figure 2.  Root development is altered in plants expressing 2b-derived transgenes. (a, b) Typical root growth phenotypes of A. thaliana plants harboring 2b-derived transgenes (2b ORF or 2b ORF plus 3′ non-translated region; 3′ NTR) after 12 days of growth on vertical plates. The presence of transgenes derived from either the Fny (a) or LS (b) strains of CMV caused an increase in lateral root length relative to wild-type plants in all the examples shown. Plants harboring Fny 2b-derived transgenes also had increased numbers of lateral roots per centimeter of main root, and their main roots were stunted relative to wild-type (WT). An untranslatable LS 2b-derived construct (LS unt2b + 3′ NTR) did not cause any significant changes in root morphology compared to wild-type controls. Line numbers are indicated in parentheses, and WT indicates wild-type, non-transformed A. thaliana controls. (c) 16-day-old WT seedlings, 10 days after inoculation with Fny CMV or mock inoculation. Viral infection caused increased lateral root length and numbers of lateral roots per centimeter of main root compared to mock-inoculated controls. Scale bars represent 1 cm.

Download figure to PowerPoint

Table 1. 2b-derived transgenes cause main root stunting and increased lateral root initiation in A. thaliana
TransgeneLineMean main root length after 12 days (cm)Mean number of lateral roots per cm of main rootMean lateral root length (mm)
TransgenicWTaTransgenicWTTransgenicWT
  1. aWT columns show averages for control plants grown on the same plates as transgenic plants from the lines indicated.

  2. *Significantly different to relevant wild-type controls with a probability of ≥95%, when assessed statistically using two-tailed Mann–Whitney tests. For further information, see Tables S1–S3.

Fny 2b2.12C2.99*3.472.99*0.664.33*0.92
2.25E1.07*2.072.59*0.671.871.59
2.30F4.223.511.791.143.34*1.06
Fny 2b + 3′ NTR3.7 H3.813.111.451.612.641.59
3.13F1.52*3.203.64*1.514.99*2.06
LS 2b4.31A2.062.300.771.502.052.31
LS 2b + 3′ NTR5.7D2.362.471.881.323.352.00
5.16A3.493.241.771.913.88*2.14
5.25A3.092.791.851.793.51*1.37

We carried out an experiment to examine root morphology in wild-type plants infected with Fny CMV (Figure 2c). Infected plants had longer lateral roots and a greater number of lateral roots per centimeter of main root than mock-inoculated controls, although no differences in main root length were observed within the time frame of these experiments (Figure 2c).

Only translatable 2b gene sequences induce altered phenotypes

The role of 2b protein in elicitation of the phenotypes seen in plants harboring 2b transgenes was examined using two approaches: immunoblot detection of the 2b protein using an anti-Fny 2b serum, and expression of untranslatable, mutant 2b transgenes. A cross-reacting band corresponding in size to the 2b protein was detected in extracts from plants harboring the Fny 2b + 3′ NTR transgene, although the amount of 2b protein that accumulated was clearly far less than that seen in CMV-infected plants (Figure 3a). No cross-reacting protein was detectable in plants harboring the Fny 2b transgene (Figure 3a), possibly as a result of the lower levels of 2b transcript accumulation seen in these plants compared with the levels typically seen in transgenic plants harboring the Fny 2b + 3′ NTR transgene (Figure 3b).

image

Figure 3.  The Fny CMV RNA 2 3′ NTR increases accumulation of Fny 2b protein in transformant plants. (a) The 2b protein detected by immunoblotting in transgenic plants harboring Fny 2b-derived transgenes. Protein was extracted from two independent primary transgenic plants harboring the Fny 2b transgene (plant lines 2.28 and 2.30) and the Fny 2b + 3′ NTR transgene (plants lines 3.33 and 3.34), as well as a wild-type plant infected with Fny CMV (Fny-CMV infected WT) and an uninfected wild-type plant (WT). Proteins were separated by SDS–PAGE, blotted to nitrocellulose membrane, and 2b protein detected using polyclonal rabbit anti-2b serum. Faint bands corresponding to the 2b protein (arrow) were visible in extracts from plants belonging to lines 3.33 and 3.34, but not in 2.28 or 2.30. (b) Steady-state accumulation of 2b and 2b + 3′ NTR transcripts was examined in second (T1) and third (T2) generation independent lines harboring either the Fny 2b transgene or the Fny 2b + 3′ NTR transgene by blotting of total RNA and probing with a 32P-labeled DNA probe complementary to the Fny 2b transcript. Transgenic lines harboring the Fny 2b transgene typically accumulated the Fny 2b transcript (lower arrow) to lower levels than lines harboring the Fny 2b + 3′ NTR transgene (Fny 2b + 3′ NTR position indicated by upper arrow). Note that the transgene in line 3.19 was silenced. To confirm equal loading, rRNA was visualized by ethidium bromide staining transcript.

Download figure to PowerPoint

These results suggest that plants harboring the Fny 2b transgene do accumulate sufficient 2b protein to have a biological effect on the plant, but that it is below the limit of detection for immunoblotting. Plants harboring the Fny 2b + 3′ NTR transgene accumulated 2b mRNA to higher levels (Figure 3b), allowing the synthesis of greater (and detectable) amounts of 2b protein (Figure 3a), and therefore a stronger phenotype (Figure 1n).

Site-directed mutagenesis was used to insert stop codons immediately after all three potential start codons in the Fny strain 2b ORF and after the one potential start codon of the LS 2b ORF to yield untranslatable 2b gene sequences (Fny unt2b, Fny unt2b + 3′ NTR, and LS unt2b + 3′ NTR). Plants transformed with either Fny unt2b or Fny unt2b + 3′ NTR expressed 2b-specific RNA transcripts (data not shown), but exhibited no apparent altered phenotypes in aerial tissues (Figure 1i,j). The roots of plants expressing LS unt2b + 3′ NTR did not show an altered phenotype (Figure 2b). Importantly, lateral root length, the only developmental phenotype altered in LS 2b-transgenic plants, did not differ significantly between LS unt2b + 3′ NTR and wild-type plant roots in three independent experiments (Table 2). The results show that phenotypes seen in 2b-transgenic plants are elicited by the 2b proteins and not by their corresponding RNA transcripts or 3′ NTRs.

Table 2.   An untranslatable 2b-derived transgene does not cause a root phenotype
Transgene Mean main root length after 12 days (cm)Mean number of lateral roots per cm of main rootMean lateral root length (mm)
  1. aDid not differ significantly from wild-type when assessed statistically (P = 0.05).

LS unt2b + 3′ NTRa2.001.550.22
Wild-type2.011.650.25

RNA silencing mediated by small-interfering RNAs is inhibited by 2b protein expressed in transgenic plants

The 2b protein performs an important counter-defense function for the virus by inhibiting siRNA-mediated RNA silencing. To determine whether the 2b proteins expressed in the 2b-transgenic plants were functional as silencing suppressors, we crossed plants harboring either LS 2b or Fny 2b transgenes with A. thaliana Amp243 plants. Amp243 plants harbor a self-silencing amplicon derived from potato virus X that contains the gene for green fluorescent protein (GFP; PVX:GFP; Dalmay et al., 2000). In the progeny plants, Amp243 × LS2b and Amp243 × Fny2b, the presence of either the LS 2b or Fny 2b transgenes alleviated the silencing of the amplicon, as indicated by restoration of GFP fluorescence (Figure 4a) and accumulation of PVX:GFP RNAs (Figure 4b). Crosses involving two independent LS and Fny 2b-transgenic lines were performed. In all four plants, the abundance of PVX:GFP RNAs was restored to an approximately equal extent by the LS and Fny 2b transgenes (Figure 4b). Thus, both 2b proteins are equally effective RNA silencing suppressor molecules.

image

Figure 4.  LS and Fny 2b transgenes are both equally capable of suppressing siRNA-mediated silencing of a virus. (a) GFP fluorescence in Amp243, Amp243 × LS 2b + 3′ NTR and Amp243 × Fny 2b + 3′ NTR plants. Amp243 plants contain a self-silencing PVX:GFP amplicon and do not accumulate GFP. The F1 progeny of crosses between Amp243 plants and LS and Fny 2b-transgenic plants exhibited strong GFP fluorescence in the leaves due to suppression of silencing by the 2b protein. The images were taken using a digital camera attached to an epifluorescence microscope. Scale bars = 200 μm. (b) PVX:GFP RNA accumulation in Amp243 plants and progeny of crosses between Amp243 and 2b-transgenics. Total RNA was isolated from 4-week-old rosettes of Amp243, Amp243 × LS 2b and Amp243 × Fny 2b plants. RNA gel blot analysis was performed using a random-primed 32P-labeled DNA probe complementary to the GFP sequence in PVX:GFP. RNA species detected were the genomic (G) and sub-genomic (sg) RNAs of PVX:GFP. Equal loading was confirmed by staining of rRNA with methylene blue.

Download figure to PowerPoint

Increased stability of miRNA target transcripts induced by expression of the 2b protein

Several viral RNA silencing suppressors have been shown to alter the stability of miRNA-targeted transcripts that play roles in plant development. However, a previous study found that the 2b protein did not affect the abundance of three of these key mRNAs: auxin response factor (ARF) 8, ARF10 and scarecrow-like (SCL) 6-IV(Chapman et al., 2004). Nevertheless, we hypothesized that the effects of the 2b protein on miRNA-mediated mRNA regulation may depend upon the strain of CMV from which the 2b protein gene was obtained. Therefore, we examined the abundance of these mRNAs, as well as ARF17, squamosa promoter-binding protein-like (SPL) 3 and NAC1 in various tissues of transgenic plants expressing either LS or Fny 2b (Figure 5).

image

Figure 5.  LS and Fny 2b transgenes have differential effects on the miRNA pathway. (a–d) RNA blot analysis of the accumulation of known miRNA targets and miRNAs in plants harboring CMV 2b gene-derived transgenes constitutively expressing the Fny CMV 2b ORF (Fny 2b), the Fny 2b ORF plus the viral 3′ non-translated region (Fny 2b + 3′), a mutated, untranslatable version of the Fny 2b ORF (UNT), the LS CMV 2b ORF (LS 2b), or the LS 2b ORF plus the viral 3′ NTR (LS 2b + 3′). RNA extracted from untransformed wild-type A. thaliana is indicated by WT. (a, b) Accumulation of ARF10 (At2g28350) and ARF17 (At1g77850) full-length mRNAs (>) and 3′ cleavage products (*) in whole seedlings with roots removed (seedling aerial parts) (a) and mature leaves (b). (c) Accumulation of full-length transcripts and 3′ cleavage products of the mRNAs ARF8 (At5g37020), ARF10, ARF17, SCL6-IV (At4g00150), SPL3 (At2g33810) and the miRNAs miR167 and miR171 in flowers. Loading controls: U6 RNA, 18S rRNA, and actin mRNA. (d) Abundance of NAC1 (At1g56010) full-length mRNA and miR164 in leaves and roots.

Download figure to PowerPoint

We found that, in the aerial parts of seedlings and in the leaves and flowers of older plants harboring Fny CMV 2b transgenes, full-length ARF10 and ARF17 mRNAs were stabilized, with a concomitant decrease in levels of their 3′ cleavage products (Figure 5a–c). By contrast, in equivalent samples from plants harboring LS CMV-derived 2b transgenes, the levels of the same full-length mRNAs and the corresponding 3′ cleavage products were similar to those seen in non-transformed plants and plants harboring the Fny unt2b construct (Figure 5a–c). For NAC1, there was a clear increase in the level of the full-length transcript in the roots and leaves of plants harboring Fny CMV 2b transgenes that was not seen in wild-type plants or in plants expressing the Fny unt2b construct (Figure 5d). However, the levels of ARF8, SCL6-IV and SPL3 were unaffected in flowers of plants harboring Fny 2b transgenes (Figure 5c).

To further investigate the effects of the 2b proteins on the miRNA-mediated RNA degradation pathway, we investigated the accumulation of miR164 (which targets NAC1; Guo et al., 2005), miR167 (which targets ARF8; Ru et al., 2006) and miR171 (which targets SCL6-IV; Llave et al., 2002b). miR164 was of particular interest as its target, NAC1, encodes a transcription factor involved in lateral root emergence (Guo et al., 2005). miR164 levels were increased in roots and leaves of plants harboring Fny CMV 2b transgenes (Figure 5d). In contrast, miR164 levels in roots and leaves of plants harboring LS CMV-derived 2b transgenes were similar to those in wild-type plants and Fny unt2b-transgenic plants. Levels of miR167 and miR171 were unaltered in flowers of plants harboring Fny or LS CMV-derived 2b transgenes (Figure 5c), consistent with the fact that the levels of their target mRNAs do not change in these tissues (Figure 5c). In combination, these data indicate that the 2b protein of the severe Fny strain is able interfere with the miRNA ‘pathway’, whereas the LS strain 2b cannot.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We have shown that the CMV 2b protein is the major factor determining whether or not viral disease symptoms become apparent in CMV-infected A. thaliana plants. The degree to which the 2b protein disrupts plant development, i.e. induces disease symptoms, is predominantly due to its ability to interfere with miRNA-regulated turnover of host mRNAs. It is not related to the ability of the 2b protein to interfere with siRNA-directed RNA silencing. Therefore, the 2b proteins of mild strains of CMV can be strong suppressors of RNA silencing but weak inducers of symptoms.

In a previous study, transgenic plants expressing a 2b protein were found to have little or no obvious phenotype (Chapman et al., 2004). However, the 2b gene sequence used by Chapman et al. (2004) was derived from the mild strain, Q CMV, that has little or no effect on miRNA regulated gene expression compared with the 2b sequence derived from a severe CMV strain, Fny. Our data suggests that 2b protein from a severe CMV strain is able to attack more than one target within the post-transcriptional regulatory system(s): one target is a component of the miRNA-regulated ‘pathway’, while the other is a component of the siRNA-mediated RNA breakdown ‘pathway’. Thus, inhibition of RNA silencing by a viral counter-defense protein need not lead inevitably to disruption of host miRNA-mediated mRNA turnover.

The viral 3′ NTR proved to be an important additional factor that accentuated the phenotype engendered by 2b transgenes. Interestingly, greater numbers of plants transformed with the Fny 2b + 3′ NTR transgene exhibited severe stunting and deformation than those harboring the Fny 2b transgene. Furthermore, expression of Fny 2b protein reached immunologically detectable levels only in plants transformed with the Fny 2b + 3′ NTR transgene. Phenotype induction (enhanced lateral root elongation) by LS 2b transgenes was also more pronounced in plants harboring the LS 2b + 3′ NTR construct. The viral 3′ NTR enhanced accumulation of the 2b transgene mRNA. We conclude that this led to increased 2b protein synthesis, causing stronger phenotypes in transgenic plants with the 2b + 3′ NTR transgenes.

Potential targets for the 2b protein in miRNA-regulated mRNA turnover

So far, efforts to identify cellular targets for the CMV 2b protein have been inconclusive. The protein localizes to the host cell nucleus and mutation of its nuclear localization sequences abolishes both its counter-silencing activity and ability to induce symptoms (Lucy et al., 2000; Mayers et al., 2000; Wang et al., 2004). However, it has been difficult to determine what molecules or structures the 2b protein might interact with inside the nucleus. Identifying 2b-interacting proteins in the yeast two-hybrid system is problematic: the 2b protein self-activates reporter gene expression unless its C-terminal sequence is deleted (Ham et al., 1999). Although the same study identified a potential 2b-interacting cellular protein, a plant homolog of a bacterial stringent response factor, it is unlikely to be a component of either the siRNA or miRNA pathways.

Several well-studied viral silencing suppressors can bind double-stranded RNA molecules of various size classes. It has been proposed that this RNA-binding activity underlies their mode of action, allowing them to sequester double-stranded siRNAs or longer precursor molecules (Mérai et al., 2006). Conceivably, the differential effects of the 2b proteins of the Fny and LS strains of CMV may be due to different affinities for dsRNAs or for different dsRNA size classes. However, there is currently some confusion as to whether or not the CMV 2b protein possesses RNA-binding activity. Although unpublished work by S.-W. Ding communicated toPalukaitis and García-Arenal (2003) indicated that it does, the results of a recent study using extracts of CMV-infected plants in gel retardation assays suggest that the 2b protein has no dsRNA-binding activity (Mérai et al., 2006). Furthermore, a range of size classes of siRNAs (21, 22 and 24 nt) appears to be involved in viral RNA silencing in a functionally redundant manner (Deleris et al., 2006). Consequently, it appears unlikely that differing dsRNA-binding affinities can fully explain the differences between the effects of the LS and Fny CMV 2b proteins on the miRNA pathway.

We propose that 2b proteins interact differentially with the ‘slicing’ components of silencing. ARGONAUTE 1, an RNA-binding protein that slices mRNA in a sequence-specific manner, associates with miRNAs, endogenous trans-acting siRNAs and transgene-derived siRNAs, but not with virus-specific RNAs (Baumberger and Baulcombe, 2005). Ten ARGONAUTE orthologs exist in A. thaliana, so it is feasible that different 2b proteins interact differentially with these or other proteins involved in loading of small RNAs into the slicing complex, thereby effecting selective inhibition of RNA degradation.

Development of roots and other tissues in plants harboring CMV 2b transgenes

CMV spreads extensively in roots, where it infects all cell types and inhibits growth (Dufour et al., 1989; Navas et al., 1998). The effects of 2b transgene expression on miRNA-regulated root development that we have observed indicate that the 2b protein is an important cause of CMV-induced decreases in root biomass observed previously in infected plants. Also, the inhibitory effect of the 2b protein on RNA silencing is likely to be an important factor permitting penetration of the virus to all parts of the root (Dufour et al., 1989; Navas et al., 1998). Furthermore, the activity of 2b as a suppressor of salicylic acid-mediated defense (Ji and Ding, 2001) may explain why CMV infection in crop plants is frequently associated with increased susceptibility to soil-borne fungal infections (Department of Crop Sciences, University of Illinois, 1999). In Fny CMV-infected A. thaliana, we observed changes in the root morphology that were similar to, but less extreme than, the effect of constitutive expression of the Fny 2b protein, and that were less severe than those reported previously for CMV-infected plants (Dufour et al., 1989; Navas et al., 1998). In particular, the inhibition of main root growth was less apparent.

The 2b transgenes derived from the severe (Fny) strain of CMV induced the strongest developmental anomalies in aerial organs and roots of transformed plants. Stunting of main roots and increased lateral root emergence on plants harboring Fny 2b transgenes was strikingly similar to phenotypes exhibited by mutant plants disrupted in auxin signaling (Guo et al., 2005). We suspected that the Fny 2b protein was disrupting auxin signaling by interfering with the miRNA pathway. This was borne out by our demonstration that the stability of full-length mRNAs encoding the auxin responsive transcription factors ARF10, ARF17 and NAC1 was altered in transgenic plants expressing the Fny 2b protein. Furthermore, the level of miR164, which targets NAC1, was altered in transgenic plants expressing the Fny 2b protein. Consistent with this, these mRNAs and miR164 were unaffected in LS 2b-transgenic plants, which did not display phenotypes indicative of disruption of auxin signaling.

Flowers of Fny 2b-transgenic plants had unaltered levels of miR167 and miR171 and the miRNA target mRNAs ARF8, SCL6-IV and SPL3. Tissue-specific silencing suppressor effects have been observed previously (Kasschau et al., 2003), and may relate to spatio-temporal variation in expression of individual miRNAs (Válóczi et al., 2006) or result from tissue-specific transcription of miRNA target transcripts (Kasschau et al., 2003).

Increased lateral root length was the only phenotype common to transgenic plants expressing Fny or LS 2b protein. We are not aware of any known miRNA-mediated control of lateral root extension. Instead, we speculate that, in the specific case of lateral roots, phenotype induction results from inhibition of siRNA-mediated RNA silencing and not inhibition of the miRNA pathway. The idea is consistent both with our finding that both 2b proteins are equally effective silencing suppressors, and the discovery by Valentine et al. (2002) that lateral root meristem activation triggers a wave of RNA silencing along the lateral root. This may allow re-programming of gene expression in new lateral roots, a process that would be disrupted by the presence of a strong suppressor of silencing induction such as 2b protein, leading to altered growth.

Symptom induction: train wreck or surgical strike?

Differential effects of 2b proteins on miRNA metabolism may explain the variation in symptom severity induced by various strains of CMV (Palukaitis et al., 1992; Zhang et al., 1994). In contrast, and as noted above, 2b proteins of mild and severe strains were equally efficient at relieving silencing of a PVX:GFP amplicon. This is consistent with studies showing that 2b proteins inhibit long-range signaling in systemic RNA silencing (Guo and Ding, 2002), and localized silencing, as seen in rescue of GFP mRNA in a protoplast model (Qi et al., 2004).

A growing body of evidence indicates that disruption of miRNA metabolism is an effect of many plant silencing suppressor molecules and that it may be one of the most common mechanisms underlying viral symptom induction (Voinnet, 2005). It is not known whether symptom induction caused by inhibition of miRNA activity evolved as a genuine role for RNA silencing suppressors or as a mere side-effect of their inhibition of siRNA-mediated RNA silencing; for example by targeting shared components of the siRNA and miRNA pathways (discussed by Bucher and Prins, 2006). Nevertheless, our data indicate that suppression of siRNA-mediated silencing does not inevitably lead to inhibition of miRNA-directed RNA cleavage, and suggest at least two potential models by which 2b proteins could affect either one or both pathways. In the first, severe strain CMV 2b proteins may possess separate functional domains that have different targets in the siRNA-mediated silencing pathway and in miRNA-directed mRNA regulation. In the second, all 2b proteins may possess an ‘active site’ that can target components that are common to the two pathways, for example ARGONAUTE proteins. However, mild strain 2b proteins may have an affinity for ortholog(s) involved in siRNA-directed RNA degradation, while severe strain 2b proteins may have evolved to interact with a wider range of orthologs. Both models are plausible as the LS and Fny 2b proteins share conserved regions but are only about 50% homologous overall (Lucy et al., 2000).

We conclude that the ability to suppress siRNA-mediated silencing does not vary greatly between CMV strains, but that the ability to interact with component(s) of the miRNA pathway varies considerably. The presence of a specific symptom-inducing function in a silencing suppressor protein strongly implies that, for certain viruses or strains, the induction of disease in the host need not be a side-effect of defense suppression. The survival of what appears to be a ‘dispensable’ activity in the 2b proteins of certain CMV strains, such as Fny, implies that symptom induction must have some selective advantage for these strains.

Note added in proof

Following acceptance of this paper, another study of the effects of CMV 2b proteins on small RNA pathways was published (Zhang et al., 2006). These workers also showed that Fny 2b-transgenic Arabidopsis thaliana plants display a symptom-like phenotype in the aerial tissues, resulting from perturbation of the miRNA pathway. Importantly, they also showed that the 2b and AGO1 proteins interact directly, consistent with a suggestion made in this paper. Zhang et al. also proposed that the different effects of Fny versus Q 2b protein on the miRNA pathway may be explained by differences between the two proteins with respect to their stability in vivo.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

DNA constructs

The 2b genes of CMV strains Fny, LS and Q (NCBI accession numbers D00355, AF416900 and X00985) were cloned from vectors pFny209 (Rizzo and Palukaitis, 1990), pLS-CMV2 (Zhang et al., 1994) and pQCD2 (Ding et al., 1995b) into the vector pBI121 (Jefferson et al., 1987), placing them under the control of the cauliflower mosaic virus 35S promoter. To aid subcloning, PCR was used to incorporate appropriate restriction sites. The plasmids generated were named pBI121 Fny 2b, pBI121 LS 2b and pBI121 Q 2b. Two further constructs were generated that harbored the Fny or LS 2b genes with the 3′ NTRs from their respective RNAs (plasmids pBI121 Fny 2b + 3′ NTR and pBI121 LS 2b + 3′ NTR). Two untranslatable 2b constructs, pBI121 Fny unt2b and pBI121 Fny unt2b + 3′ NTR, were generated using pBI121 Fny 2b and pBI121 Fny 2b + 3′ NTR. Fny 2b contains three start codons at positions 1, 8 and 18 in the amino acid sequence. Amino acids 2 (E), 9 (T) and 19 (V) were converted to stop codons by site-directed mutagenesis using the Quick-Change II XL kit (Stratagene; http://www.stratagene.com/). Primers CMVFnyd2b_A4F (5′-CTAGAGGATCCATGTAGTTGAACGTAGGTGCAATGTAGAACGTCGAACTCC) and CMVFnyd2b_A4R (5′-GGAGTTCGACGTTCTACATTGCACCTACGTTCAACTACATGGATCCTCTAG) were used to mutate nucleotides 4 (G[RIGHTWARDS ARROW]T), 6 (A[RIGHTWARDS ARROW]G), 25 (A[RIGHTWARDS ARROW]T), 26 (C[RIGHTWARDS ARROW]A) and 27 (A[RIGHTWARDS ARROW]G). Primers CMVFnyd2b_A3F (5′-CCAACTGGCTCGTATGTAGGAGGCGAAGAAGC) and CMVFnyd2b_A3R (5′-GCTTCTTCGCCTCCTACATACGAGCCAGTTGG) were used to mutate nucleotides 55 (G[RIGHTWARDS ARROW]T) and 56 (T[RIGHTWARDS ARROW]A). The untranslatable LS 2b construct pBI121 LS 2b + 3′NTR was produced by mutation of the codon encoding amino acid 2 (D) into a stop codon using primers CMVLSd2b_AIF (5′CTCTAGAGGATCCATGTAGGTGTTGACAGTAGTGG) and CMVLSd2b_AIR (5′-CCACTACTGTCAACACCTACATGGATCCTCTAGAG). Nucleotide numbers refer to positions in the 2b ORF.

Plant transformation

A. thaliana ecotype Columbia-0 seeds were planted on a 4:1 soil/sand mixture and maintained at 21°C with an 8 h photoperiod. Agrobacterium tumefaciens-mediated transformation was performed by floral dipping (Clough and Bent, 1998). Primary-transformed (T0 generation) plant seed was selected by using a germination medium [0.05% 2-(N-morpholino)ethanesulfonic acid, pH 5.7, 0.46% Murashige and Skoog salts] containing 50 μg ml−1 kanamycin. The presence of the intended transgene was confirmed by PCR with appropriate 2b gene-specific primers.

Viral strains and inoculation procedures

Fny CMV was reconstituted from infectious clones as described previously (Soards et al., 2002) and propagated in tobacco. LS CMV was maintained by passaging in tobacco. Infectious virus was purified from tobacco by the method described by Ng and Perry (2004). To investigate the effects of virus infection on aerial plant organs, A. thaliana plants at the four- to six-leaf stage were inoculated by using a cotton bud to apply 100 μg ml−1 CMV to a carborundum-dusted leaf. To examine the effects of virus infection on roots, 2-day-old wild-type seedlings grown on plates were inoculated with 40 μg ml−1 virus suspension. Virus-infected and mock-inoculated controls (eight plants per plate) were transferred to vertical plates at 4 days post-inoculation.

RNA blotting

Total RNA was extracted using TRIzol reagent, and poly(A) mRNA was purified from total RNA using Dynabeads (Invitrogen; http://www.invitrogen.com/). Total RNA or poly(A) mRNA was subjected to 1.2% agarose–formaldehyde gel electrophoresis, transferred to Hybond N+ membrane and UV cross-linked (12 000 μJ). Pre-hybridization, hybridization and washing of blots were performed using the ‘Never-fail’ protocol (http://www.bioon.com/experiment/rna3/62885.shtml). 32P-labeled DNA probes for mRNAs were synthesized by random priming of DNA sequences generated from genomic DNA by PCR using primers selected from those described by Kasschau et al. (2003). Primers were selected from those described by Kasschau et al. (2003). For RNA blot analysis of miRNAs, 20–40 μg of total RNA was separated on a 15% polyacrylamide gel and electroblotted to nylon membrane. Membranes were probed using 32P-labeled probes produced by end labeling of complementary oligonucleotides. Bound probe was visualized by phosphoimaging. Root tissue for RNA extraction was prepared according to the method described by Hétu et al. (2005).

Immunoblotting

Extraction and detection of Fny 2b protein was carried out using polyclonal rabbit anti-2b serum as described by Mayers et al. (2000). Primary antibody binding was detected by chemiluminescence (Mayers et al., 2000).

Root analyses

Roots of transgenic plants were analyzed by growth on vertical germination medium plates (growth chamber at 20°C, 16 h photoperiod, light level 47 μmol m−2 sec−1). On each plate, 5–10 transgenic and five wild-type plants were grown simultaneously. Roots were photographed digitally 12 days after the end of stratification, and root lengths were measured from the junction with the hypocotyl to the tip of the primary root using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Statistical analyses were conducted within Microsoft Excel using Mann–Whitney test methodology. Considerable variation can be observed even between individual wild-type (control) plants grown on different plates. Consequently, all statistical comparisons were conducted only between transgenic plants and control plants grown on the same plate. For the purpose of comparison, composite averages of measurements taken from all wild-type control plants are given in Table 1, but it should be noted that these may differ from the exact average on any individual plate. Three independent replicate experiments were carried out, detailed data for which are presented in Tables S1–S3. One independent replicate is presented in Table 1. Roots of CMV-infected plants on plates were imaged at 10 days post-inoculation and roots assessed as above, but statistical comparisons were made between treatment and control plates. Approximately 24 plants were examined per treatment (mock- versus virus-inoculated).

Silencing suppression assay

The ability of transgenes to suppress RNA silencing was assessed by crossing them into the Amp243 background (Dalmay et al., 2000). Restoration of GFP fluorescence in the progeny was observed by epifluorescence microcopy and imaged with a Nikon Coolpix (http://www.nikondigital.com/main.html) digital camera. Rescue of PVX:GFP-derived replicon RNA accumulation was analyzed by Northern blotting (as detailed above) using appropriate 32P-labeled DNA or oligonucleotide probes.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We are grateful to Shou-Wei Ding and Li Feng (University of California, Riverside, USA) for help and advice on small RNA techniques, and to David Baulcombe (Sainsbury Laboratory, Norwich, UK) for Amp243 seeds. We thank Caroline York for technical assistance and Alex Murphy and Piers Hemsley (Bristol University, UK) for useful discussions. M.L. was supported by a Biotechnological and Biological Sciences Research Council (BBSRC) studentship, and F.C.R. by a studentship from the Gatsby Charitable Trust. J.P.C.’s laboratory is funded by BBSRC grant BB/D008204/1, and work by P.P. and T.C. is funded by the Scottish Executive Environment and Rural Affairs Department.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Table S1. Detailed analysis of alterations in root phenotypes caused by CMV 2b-derived transgenes (Experiment 1) Table S2. Detailed analysis of alterations in root phenotypes caused by CMV 2b-derived transgenes (Experiment 2) Table S3. Detailed analysis of alterations in root phenotypes caused by CMV 2b-derived transgenes (Experiment 3)

FilenameFormatSizeDescription
TPJ_3042_sm_Tables.doc68KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.