Translocation of Tomato Bushy Stunt Virus P19 Protein into the Nucleus by ALY Proteins Compromises Its Silencing Suppressor Activity

Cloning of additional N. benthamiana ALY genes.

Previously two ALY genes (ALY615 and ALY617) were identified and cloned from N. benthamiana (23) and a partial ALY gene, referred to as Hin19, was cloned from Nicotiana tabacum (13). BLAST searches of The Institute for Genomic Research (Rockville, MD) gene indices (http://www.tigr.org/tigr-scripts/tgi/T_reports.cgi?species=N.benthamiana) identified two additional N. benthamiana ALY homologues (TC1693 and TC916). PCR amplification of cDNA synthesized from total RNA isolated from leaves of N. benthamiana was carried out using primers 871 (5′-AAACATGTCAAATCTCGATGTATCCCTCG-3′) and 828 (5′-TTTCAGTTAGTCTCCATTGCTTCTGAATG-3′) to clone NbALY1693 and primers 833 (5′-AACCATGGCTTCCTTGGATATGACCCTCG-3′) and 834 (5′-TTTTAGGTCTGCATAGCATCTGCAGT-3′) to clone NbALY916.

Construction and expression of ALY and P19 proteins in plants.

The Arabidopsis and N. benthamiana ALY genes and TBSV P19 were cloned by restriction fragment ligation into the binary vector pROK2 to allow expression in plants. These genes were also cloned in pROK2, in sequential or three-fragment ligations with the green fluorescent protein (GFP) gene and, for P19, also with the monomeric red fluorescent protein (mRFP) gene (3), to allow expression in plants of the ALY and P19 proteins each with a C-terminal GFP or mRFP extension. The ALY and P19 genes were also cloned into the gateway entry vectors pDONR201 and pDONR207 (Invitrogen) to allow their subsequent mobilization by recombination into vectors pLX222, pBAT-TL-K-GFP, pBAT-TL-K-sRFP-N, and pBAT-TL-K-sRFP-C for expression in plants as either unfused proteins or proteins fused at the C terminus with GFP or fused at the C terminus with N- or C-truncated mRFP for bimolecular fluorescence complementation, respectively (2). Plasmids were transformed into Agrobacterium tumefaciens strains LBA4404 and C58C1 for delivery into N. benthamiana plants by agroinfiltration (27). 4′,6′-Diamidino-2-phenylindole (DAPI) staining of nuclei was achieved by infiltration of a 1-μg ml⁻¹ aqueous solution as described previously (6).

Imaging of fluorescent proteins.

All imaging was conducted using either a Leica TCS-SP1 confocal laser-scanning microscope (CLSM) or a Leica TCS-SP2 acoustical optical beam splitter (AOBS) (Leica Microsystems Heidelberg GmbH, Germany). All images were obtained by sequential scanning, either frame by frame (SP1) or line by line (SP2) using the following excitation (EX) and emission (EM) settings: for GFP, EX of 488 nm and EM of 500 to 530 nm; for DsRed and mRFP, EX of 568 nm and EM of 580 to 600 nm (SP1) or EX of 561 nm and EM of 590 to 630 nm (SP2); for chlorophyll-associated autofluorescence, EX of 488 nm and EM of 660 to 700 nm; and for DAPI, EX of 405 nm and EM of 440 to 475 nm.

Silencing assays.

N. benthamiana plants were infiltrated with a culture of Agrobacterium containing the binary plasmid pGPTV^(+35,+NosT) expressing β-glucuronidase (GUS) (4) either alone or mixed with other cultures expressing the ALY proteins, either with or without a GFP tag, or P19 either untagged or with GFP or mRFP tags. Where possible, expression of these proteins was confirmed by confocal fluorescence microscopy. Infiltrated tissue was collected after 6 days, and RNA samples were extracted, processed for mRNA and siRNA extraction, and analyzed by Northern blotting using digoxigenin-labeled GUS-specific probes as described before (4).

Arabidopsis ALY chimeras.

For the synthesis of chimeras between AtALY2 and AtALY3, the ALY protein was divided into three domains: an N-terminal domain (amino acids [aa] 1 to 100); a central, RNA recognition motif (RRM) domain (aa 101 to 195 of AtALY2; aa 101 to 191 of AtALY3); and a C-terminal domain (aa 196 to 292 of AtALY2; aa 192 to 293 of AtALY3). Each gene was modified by PCR to introduce a SpeI restriction site between the N-terminal and central domains and a NgoMIV restriction site between the central and C-terminal domains. Thus, for AtALY2, aa 101 and 102, Gly (GGC) and Ser (TCG) became Thr (ACT) and Ser (AGT), and aa 196 and 197, Ala (GCT) and Thr (ACA) became Ala (GCC) and Gly (GGC). For AtALY3, aa 100 and 101, Val (GTG) and Ser (TCA) became Thr (ACT) and Ser (AGT), and aa 192 and 193, Ala (GCA) and Ala (GCT) became Ala (GCC) and Gly (GGC). Using these new restriction sites and existing, unique sites (HpaI and XbaI) located upstream or downstream, respectively, of the ALY genes, AtALY2 and AtALY3 chimeras were assembled carrying all eight possible combinations of the three domains of the ALY protein. The chimeras were named according to their construction; thus, chimera 2-2-2 carries all three domains derived from AtALY2, whereas, for example, chimera 2-3-2 carries the N- and C-terminal domains from AtALY2 together with the central domain from AtALY3. The chimeras were cloned into either Tobacco rattle virus or pROK2 as described before (23) for expression in plants and analysis by confocal microscopy.

To create chimeras of the central domain, an EagI site (CGGCCG) was introduced at nucleotides 432 to 437 (spanning aa 144 to 146) in the AtALY2 gene and nucleotides 429 to 434 in the AtALY3 gene (spanning aa 143 to 145). These nucleotide changes did not alter the amino acid sequences of the ALY proteins but separated the N-terminal part of the RRM domain, including the highly conserved RNP2 motif, from the C-terminal part containing the RNP1 motif. The SpeI-EagI and EagI-NgoMIV fragments of the central domain were assembled with the N- and C-terminal domains of AtALY3 to construct two chimeras. Chimera 3-23-3 has the N-terminal part of the central domain from AtALY2 and the C-terminal part of the central domain from AtALY3, whereas chimera 3-32-3 has the N-terminal part of the central domain from AtALY3 and the C-terminal part of the central domain from AtALY2. The chimeras were cloned into pROK2 for analysis in plants.

Nucleotide sequence accession numbers.

The sequences of two N. benthamiana ALY genes were deposited with the EMBL/GenBank data libraries (NbALY1693, accession number AM167905; NbALY916, accession number AM167906).

Effect of ALY expression on P19 suppressor function.

In our earlier work we did not observe any link between the interaction of P19 protein with ALY protein and the known function of P19 as a silencing suppressor protein (23). As we had cloned only two ALY genes (ALY617 and ALY615) from N. benthamiana, the plant species that was used for the virus-induced gene silencing and agroinfiltration experiments, we suggested that examination of additional genes would be useful. Consequently, for this work we have cloned two more N. benthamiana ALY genes (ALY1693 and ALY916). ALY916 is the orthologue of N. tabacum Hin19 (13) and is 275 aa long, whereas ALY1693 is 246 aa long. Sequence comparison of the four N. benthamiana ALY proteins shows that ALY615 and ALY1693 are 93% identical, ALY615 and ALY617 are 57% identical, ALY615 and ALY916 are 38% identical, ALY617 and ALY916 are 41% identical, ALY617 and ALY1693 are 39% identical, and ALY916 and ALY1693 are 25% identical.

To test whether any of the ALY proteins could affect silencing suppression by P19, we used agroinfiltration of leaves to transiently express the P19 protein together with or without each of the four Arabidopsis or four N. benthamiana ALY proteins, and examined whether P19 was able to suppress silencing of a coexpressed marker gene (β-glucuronidase). As expected, accumulation of the GUS mRNA increased when coexpressed with P19 due to the silencing suppressor activity of P19 (Fig. 1A and B, lane 6 versus lane 1). Without P19, transient expression of each of the ALY genes had no effect on the level of GUS mRNA accumulation (Fig. 1A and B, lanes 2 to 5 versus lane 1). However, when both ALY and P19 were expressed simultaneously, four of the ALYs (AtALY1, AtALY3, NbALY615, and NbALY1693) partially inhibited the suppressor activity of P19, resulting in reduced levels of GUS mRNA accumulation compared to that found with P19 alone (Fig. 1A, lanes 7 and 9 versus lane 6; Fig. 1B, lanes 7 and 10 versus lane 6), whereas the other four ALYs (AtALY2, AtALY4, NbALY617, and NbALY916) did not reduce the level of accumulated GUS mRNA (Fig. 1A, lane 8 and 10 versus lane 6; Fig. 1B, lanes 8 and 9 versus lane 6). This result was obtained consistently when coexpressing native ALY and P19 (six independent experiments), ALY-GFP and P19-mRFP (two experiments), and native ALY with P19-mRFP (one experiment).

Fusion of GFP to P19 affects its function.

Our previous studies, focusing on the subcellular distribution of ALY, have all used native P19; however, to examine in greater detail the interaction between P19 and the different ALY proteins, we decided to tag P19 at the C terminus by fusion either to GFP or mRFP. To assess whether tagging compromised the activity of P19, we examined whether P19-GFP and P19-mRFP retained the ability to suppress RNA silencing. Plants were agroinfiltrated with a binary plasmid encoding GUS, together with P19, either untagged, GFP tagged or mRFP tagged, or with an empty binary plasmid as a control. Initially, GUS mRNA accumulated, but after 3 or 4 days, it began to decline in amount as RNA silencing was initiated, and GUS-specific siRNAs accumulated instead. In the presence of native P19, full-length GUS mRNA was not degraded and siRNAs accumulated only to a low level (Fig. 2, lane 2). However, Western blot experiments showed that P19-GFP accumulated poorly in infiltrated tissue (data not shown), and there was a significant impairment in the suppression of silencing of GUS, leading to the degradation of GUS mRNA and accumulation of GUS-derived siRNAs (Fig. 2, lane 3). P19-mRFP accumulated to the same level as native P19 in infiltrated tissue (data not shown) and was an effective suppressor of silencing, allowing full-length GUS mRNA to accumulate (Fig. 2, lane 4). Consequently, all other experiments included mRFP-tagged P19, rather than GFP-tagged P19.

Subcellular relocalization of P19 by ALY.

P19 is considered primarily to be a cytoplasmic protein (17), although some P19-GFP was also found to be present in the nucleus (but not the nucleolus) of agroinfiltrated plants (23) in a pattern similar to the distribution of free (unfused) GFP. In this work, fluorescence derived from P19-mRFP had the same distribution (Fig. 3A). We showed previously that when coexpressed with native P19, three of the ALY-GFP proteins (AtALY2, AtALY4, and NbALY617) were localized to the cytoplasm, whereas the remaining three ALY-GFP proteins (AtALY1, AtALY3, and NbALY615) were localized to the nucleus (23). The two newly identified N. benthamiana ALY proteins, NbALY916 and NbALY1693, were both found to be nuclear (Fig. 3B, left panels).

However, NbALY916-GFP was relocalized to the cytoplasm by P19, whereas NbALY1693-GFP was retained in the nucleus (Fig. 3B, right panels). The distribution patterns of the ALY-GFP proteins remained consistent when P19-mRFP was substituted for P19 (data not shown).

An unexpected observation in these experiments was that an alteration in the subcellular distribution pattern of P19-mRFP also occurred. When coexpressed with those ALY-GFP proteins that were delocalized to the cytoplasm, P19-mRFP was also localized mostly to the cytoplasm (Fig. 3C and data not shown). In contrast, a significant proportion of P19-mRFP accumulated inside the nucleus, and specifically in the nucleolus, when coexpressed with those ALY-GFP proteins that themselves remained nuclear/nucleolar (Fig. 3C and data not shown). The same pattern of P19 distribution occurred when untagged ALY rather than ALY-GFP was included in the infiltrations (data not shown). A summary of the localization patterns of each of the ALY proteins and P19 is shown in Table 1.

Mapping of ALY domains involved in nucleolar targeting and delocalization induced by P19.

In order to map the ALY domains involved in nucleolar targeting and the nucleus-to-cytoplasm delocalization induced by P19, chimeras were made between two GFP-tagged A. thaliana ALY proteins, AtALY2, which does not accumulate in the nucleolus and is relocated to the cytoplasm in the presence of P19 (or P19-mRFP), and AtALY3, which accumulates in the nucleolus and remains there in the presence of P19 (or P19-mRFP). For synthesis of the chimeras, the ALY protein was divided into three domains: the two N- and C-terminal domains containing variable, nonconserved sequences and a more conserved central domain containing the RNA-binding motif (RRM) (Fig. 4). Eight different GFP-tagged chimeras, resulting from all the possible permutations between the three domains of these two ALYs were analyzed by infiltration in N. benthamiana plants either with or without P19, followed by confocal microscopy.

While all the chimeras were nuclear when expressed alone, some displayed the nucleus-to-cytoplasm delocalization phenotype when coexpressed with P19. This phenotype was determined solely by the presence of the central RRM domain of AtALY2 (Fig. 4, field panels). Further mapping of the delocalization motif was made using additional, intra-RRM chimeras, revealing that the AtALY3 chimera carrying the C-terminal RRM fragment from AtALY2 (ALY3-32-3-GFP) relocated to the cytoplasm in the presence of P19 (or P19-mRFP), whereas the AtALY3 chimera carrying the N-terminal RRM fragment from AtALY2 (ALY3-23-3-GFP) remained nuclear (Fig. 5A). Thus, the C-terminal half of the AtALY2 RRM domain was solely responsible for the delocalization phenotype induced by P19.

Nucleolar targeting of AtALY3 did not map to a single domain. While the presence of the C-terminal domain from AtALY3 in a chimera (e.g., AtALY2-2-3-GFP) was sufficient for it to be targeted to the nucleolus, replacing the C-terminal domain of AtALY3 with that of AtALY2 (ALY3-3-2-GFP) did not prevent accumulation of this protein in the nucleolus (Fig. 4, nucleus panels). Replacing both the RRM and C-terminal domains (AtALY3-2-2-GFP) did prevent nucleolar accumulation. In addition, chimera AtALY2-3-2-GFP displayed a mixed phenotype, with nonnucleolar targeting in the majority of epidermal nuclei (Fig. 4, nucleus panel), but some nucleolar targeting in the nuclei of occasional epidermal cells was also observed (data not shown).

Effects of ALY RRM domain swaps on P19 suppressor function.

The experiments described above showed that the four Arabidopsis and N. benthamiana ALY proteins that reduced P19 suppressor function in an agroinfiltration assay were those that remained nuclear during coexpression experiments with P19 (this work and earlier work [23]) and that relocated the P19 protein to the nucleus (AtALY1, AtALY3, NbALY615, and NbALY1693). Conversely, those ALY proteins that relocalized to the cytoplasm together with P19 (AtALY2, AtALY4, NbALY617, and NbALY916) did not interfere with P19 suppressor activity. These data show that the accumulation of P19 in the nucleus mediated by ALY had an inhibitory effect on silencing suppression by P19. To dissect the roles of the different ALY domains on the inhibition of P19 suppressor function, the AtALY2 and AtALY3 RRM domain chimeras described above were analyzed in the GUS mRNA silencing/suppression assay.

As was shown before, AtALY3-GFP, which remains nuclear in the presence of P19 or P19-mRFP, reduced the suppressor activity of P19 (Fig. 1A, lane 9 versus lane 6). Similarly, the AtALY3 RRM chimera carrying the N-terminal RRM fragment from AtALY2 (ALY3-23-3-GFP), which also remains nuclear in the presence of P19 or P19-mRFP, reduced the activity of the suppressor (Fig. 5B, lanes 4 and 6 versus lanes 5 and 7). However, the chimera carrying the C-terminal RRM fragment from AtALY2 (ALY3-32-3-GFP), which relocalizes to the cytoplasm, did not have an inhibitory effect on the suppression function (Fig. 5B, lanes 5 and 7 versus lanes 4 and 6). These results demonstrate that the effect of ALY on the suppressor activity of P19 is linked to the nuclear versus cytoplasmic localization phenotype that maps to the C-terminal domain of the central RRM motif of the ALY protein.

Localization of the site of P19-ALY interaction.

The results so far revealed that in some combinations P19 and ALY both localized to the nucleus. To confirm that the colocalization reflected actual physical interaction between these proteins, we used the technique of bimolecular fluorescence complementation (BiFC) whereby one of the proteins under study is fused to a nonfunctional truncated fluorescent reporter, while the second assay protein is fused to a complementing fragment of the same reporter (9). Our studies examined plant cells coinfiltrated to express P19 fused at the C terminus to an N-terminal fragment of mRFP and each of the four Arabidopsis ALY proteins fused at the C terminus to a C-terminal fragment of mRFP. For P19 with AtALY1, AtALY2, or AtALY4, we observed diffuse patterns of mRFP fluorescence in the nuclei of cells, the location of which was confirmed by counterstaining with DAPI (Fig. 6a to d). Cells coexpressing the P19 and AtALY3 BiFC constructs appeared to show mRFP fluorescence in nuclei, but these nuclei were distorted in shape and difficult to counterstain with DAPI. Nevertheless, these results suggest that P19 and all four Arabidopsis ALY proteins interact in the nucleus but not, to any detectable extent, in the cytoplasm. As a control, plants were infiltrated with BiFC constructs of P19 and an Arabidopsis 19S proteosome subunit protein (At1g29150), which is located in the nucleus and cytoplasm and is functional in other, unrelated BiFC experiments (data not shown). No mRFP fluorescence was detected with this combination of proteins (Fig. 6e).