Fig 1.
Expression of AtRH30 DEAD-box helicase inhibits tombusvirus genomic (g)RNA replication in N. benthamiana plant and in yeast surrogate host.
N. benthamiana plants expressing AtRH30 were inoculated with (A) TBSV, (B) CNV, (C) CIRV, respectively. Top panel: Northern blot analyses of tombusvirus gRNA using a 3’ end specific probe shows reduced accumulation of gRNA and subgenomic RNAs in plants expressing RH30 than in control plants. Bottom panel: Ethidium-bromide stained gel shows 18S ribosomal RNA as a loading control. (D-E) Expression of the helicase core mutant of RH30 (RH30m, F416L) inhibited TBSV or CIRV replication, respectively, to a lesser extent, demonstrating the requirement of the helicase/ATPase function of RH30 for its full virus restriction function. See further details in panel A. Each experiment was repeated at least three times. (F) Expression of RH30 inhibits TBSV replication in yeast. Top panel: Northern blot analysis of TBSV repRNA using a 3’ end specific probe shows reduced accumulation of repRNA in WT yeast strain expressing RH30. Viral proteins His6-p33 and His6-p92pol were expressed from plasmids from the CUP1 promoter, while DI-72(+) repRNA was expressed from the GAL1 promoter. His6-RH30 was expressed from a plasmid. Middle panel: Northern blot with 18S ribosomal RNA specific probe was used as a loading control. Bottom images: Western blot analysis of the level of His6-p33, His6-p92pol and His6-RH30 with anti-His antibody.
Fig 2.
Knockdown of NbRH30 gene expression leads to enhanced tombusvirus replication in N. benthamiana plants.
(A) Top panel: Accumulation of the TBSV genomic (g)RNA and sgRNAs in RH30-silenced N. benthamiana plants 1.5 days post-inoculation (dpi) was measured by Northern blot analysis. Inoculation of TBSV gRNA was done 12 days after silencing of RH30 expression. VIGS was performed via agroinfiltration of tobacco rattle virus (TRV) vector carrying 5’ or 3’-terminal NbRH30 sequences, whereas as a control, 3’-terminal GFP sequences. Second panel: Ribosomal RNA is shown as a loading control in an ethidium-bromide stained agarose gel. Third panel: Northern blot analysis shows the knock-down level of NbRH30 mRNA in the silenced and control plants. Fourth panel: Northern blot analysis shows 18S ribosomal RNA as a loading control. Fifth and seventh panels: RT-PCR analysis of NbRH30 mRNA level in the silenced and control plants. Sixth and eighth panels: RT-PCR analysis of TUBULIN mRNA level in the silenced and control plants. Each experiment was repeated. Bottom panel: Accelerated and more severe TBSV-induced symptom development is observed in RH30-silenced N. benthamiana plants as compared with the control plants. Note the mild growth defect phenotype in RH30-silenced N. benthamiana plants. The picture was taken 5 dpi. (B-C) Top panel: Accumulation of the CNV or CIRV gRNA in RH30-silenced N. benthamiana plants 2 days post-inoculation (dpi) was measured by Northern blot analysis. See further details in panel A.
Fig 3.
Confocal microscopy shows the retargeting of the mostly nuclear RH30 into the large replication compartment in plant protoplasts and whole plants infected with CNV.
(A) Most of RH30 is re-targeted into the replication compartment marked by the BFP-tagged p33 replication protein (pointed by arrows) in N. benthamiana protoplasts. Second panel: in the absence of viral components, GFP-tagged RH30 is mostly present in the nucleus, as marked by the histone protein (RFP-H2B). Third panel: The re-targeted GFP-RH30 is present in the viral replication compartment, marked by p33-BFP replication protein and RFP-SKL peroxisomal matrix marker. Arrows point at the viral replication compartment. Fourth panel: RH30 is not co-localized to the peroxisomes in the absence of tombusvirus replication. Fifth panel: The re-targeted GFP-RH30 is partially co-localized with the ER marker within the viral replication compartment, marked by p33-BFP replication protein. The leaves of N. benthamiana plants (transgenic plants expressing nucleus marker RFP-H2B or ER marker RFP-ER) were agro-infiltrated to express p33-BFP, GFP-RH30, and CNV20KSTOP gRNA. Leaves without the expression of p33-BFP and CNV20KSTOP gRNA were used as controls. The agro-infiltrated leaves were collected to isolate protoplasts for confocal imaging 2.5 days post agro-infiltration. Scale bars represent 10 μm. (B) Confocal microscopy images show co-localization of TBSV p33-BFP or CIRV p36-BFP replication proteins and the GFP-RH30 in planta. The large replication compartment was visualized via expression of TBSV p33-BFP or CIRV p36-BFP. Expression of the above proteins from the 35S promoter was done after co-agroinfiltration into N. benthamiana leaves. The leaves of N. benthamiana plants were agro-infiltrated to express TBSV p33-BFP or the CIRV p36-BFP, GFP-RH30, and CNV20KSTOP or CIRV gRNAs. Leaves without the expression of p33-BFP or p36-BFP and the viral RNAs were used as controls. The agro-infiltrated leaves were collected for confocal imaging 2.5 days post agro-infiltration. Scale bars represent 10 μm. Each experiment was repeated.
Fig 4.
Enrichment of AtRH30 in the nucleus nullifies its antiviral effect against TBSV.
(A) Northern blot analysis of TBSV gRNA using a 3’ end specific probe shows lack of inhibition of gRNA accumulation in plants expressing RH30 fused to an NRS. Bottom panel: Ethidium-bromide stained gel to show 18S ribosomal RNA as a loading control. (B) NRS-RH30-GFP is not re-targeted into the replication compartment marked by the TBSV BFP-tagged p33 replication protein (pointed by an arrow) in N. benthamiana protoplasts. Second panel: in the absence of viral components, NRS-RH30-GFP is present in the nucleus, as marked by the histone protein (H2B-RFP). The leaves of N. benthamiana plants (transgenic plants expressing nucleus marker RFP-H2B) were agro-infiltrated to express p33-BFP, GFP-RH30, and CNV20KSTOP gRNA. Leaves without the expression of p33-BFP and CNV20KSTOP gRNA were used as controls. The agro-infiltrated leaves were collected to isolate protoplasts for confocal imaging 2.5 days post agro-infiltration. (C) Confocal microscopy images show different localization of TBSV p33-BFP replication protein and NRS-RH30-GFP in N. benthamiana cells infected with CNV. The large replication compartment was visualized via expression of TBSV p33-BFP. Expression of the above proteins from the 35S promoter was done after co-agroinfiltration into N. benthamiana leaves. See further details in Fig 3B. Scale bars represent 10 μm. Each experiment was repeated.
Fig 5.
Co-purification of RH30 helicase with the viral replicase from membranous fraction of yeast.
(A) Co-purification of His6-tagged RH30 with Flag-p33 and Flag-p92pol replication proteins from subcellular membranes. Top panels: Western blot analysis of co-purified His6-RH30 (lanes 1, 2, and 3) with Flag-affinity purified replicase, Flag-p33 and Flag-p92pol replication proteins, respectively as shown. His6-p33, His6-p92pol and His6-RH30 were detected with anti-His antibody, while Flag-p33 and Flag-p92pol replication proteins were detected with anti-FLAG antibody. The negative control was from yeast expressing His6-RH30, His6-p33 and His6-p92pol purified in a FLAG-affinity column (lane 4). Bottom panel: blot of total His6-p33 and His6-p92pol and His6-RH30 in the total yeast extracts detected with anti-His antibody. (B) Pull-down assay including TBSV GST-p33 replication protein and the MBP-tagged RH30. Note that we used the soluble C-terminal region of TBSV p33 replication protein, which lacked the N-terminal sequence, including the trans-membrane TM domain. Top panel: Western blot analysis of the captured GST-p33C with the MBP-affinity purified MBP-RH30 or the helicase core mutant of RH30 (RH30mut, F416L) was performed with anti-His antibody. The negative control was MBP (lane 1). Middle panel: Coomassie-blue stained SDS-PAGE of the captured MBP-RH30 and MBP. Bottom panel: Western blot analysis of GST-p33C in total E. coli lysates. Each experiment was repeated three times. (C) Interactions between TBSV p33 replication protein and the RH30 helicase was detected by BiFC. The TBSV p33-cYFP replication protein and the nYFP-RH30 and the RFP-SKL peroxisomal marker protein were expressed via agro-infiltration. The merged image shows the efficient co-localization of the peroxisomal RFP-SKL with the BiFC signals, indicating that the interaction between the tombusvirus replication protein and the recruited RH30 helicase occurs in the large viral replication compartments, which consist of aggregated peroxisomes. Scale bars represent 5 μm.
Fig 6.
Inhibition of TBSV repRNA accumulation by RH30 in in vitro replication assay based on CFE obtained from WT yeast.
(A) The purified recombinant tombusvirus p33 and p92 replication proteins from E. coli were added in combination with the template (+)repRNA to program the in vitro tombusvirus replication assay. Increasing amounts (1.9 and 5.7 μM) of purified recombinant MBP-RH30 or MBP, as a control, were added to the reactions. Non-denaturing PAGE shows the accumulation of 32P-labeled (+)repRNAs and the dsRNA replication intermediate products made by the reconstituted replicases. Heat treatment, as shown, was applied to demonstrate the dsRNA nature of the replication intermediate. (B) Scheme of the two-step CFE-based in vitro replication assay. Step #1 promotes the assembly of the functional tombusvirus replicase, whereas step #2 supports viral RNA synthesis in the presence of all four ribonucleotides. Note that MBP-RH30 or MBP (1.9 and 5.7 μM), as a control, were added to the reactions either at step #1 or step #2, as shown. The 32P-labeled TBSV repRNA products of the reconstituted replicases were detected by denaturing PAGE. (C) The in vitro RdRp activation assay is based on (+)repRNA and p92-Δ167N RdRp protein in the presence of the soluble fraction of yeast CFE. Purified MBP-RH30 and MBP were added in increasing amounts. Denaturing PAGE analysis of the 32P-labeled RNA products obtained in an in vitro assay with recombinant p92-Δ167N RdRp. (D) In vitro translation assay with wheat germ extract programmed with CIRV gRNA. Purified MBP-RH30 and MBP were added in increasing amounts (1.9 μM and 3.8 μM). The 35S-methionine-labeled p36 replication protein translation product is detected by SDS-PAGE. Tdh2 mRNA was used as a control. Each experiment was repeated three times.
Fig 7.
RH30 binds to the RII(+)-SL cis-acting element involved in RNA template selection.
(A-B) RNA gel mobility shift analysis shows that MBP-RH30 binds to 32P-labeled (+)repRNA and (-)repRNA, respectively, in vitro. Purified MBP-RH30 or MBP were added in increasing amounts (0.4, 1.9 μM and 5.7 μM) to the assays. The MBP-RH30—32P-labeled ssRNA complex was visualized on nondenaturing 5% polyacrylamide gels. Each experiment was repeated at least three times. (C) In vitro RNA binding assay with purified RH30. The assay contained 5.7 μM of purified MBP-RH30 or MBP in combination with the 32P-labeled (+)repRNA template (~0.1 pmol) or (-)repRNA template (~0.1 pmol) and unlabeled competitor RNAs (2 and 4 pmol) representing one of the four regions of TBSV DI-72 RNA from both RNA strands (see panel D) were used in the competition assay. The MBP-RH30—32P-labeled ssRNA complex was visualized on nondenaturing 5% acrylamide gels. Each experiment was repeated at least three times. (D) Schematic representation of the four regions carrying cis-acting sequences in the DI-72 (+)repRNA. In vitro RNA binding assay with purified MBP-RH30 (1.9 and 5.7 μM) and the 32P-labeled RII(+)-SL was performed in the presence or absence of 1 mM ATP. MBP-p33C (1.9 and 5.7 μM) representing the C-terminal soluble portion of TBSV p33 replication protein was used as a positive control, whereas MBP was the negative control. See further details in panel A. (E-F) Top: Schematic representation of the partial RNA/RNA duplexes used in the strand separation assay. The unlabeled template consists of DI-72 (+)repRNA and a short 32P-labeled complementary (-)RNA (representing either RI or RII in DI-72), which anneals to the 621 nt DI-72 (+)repRNA. Increasing amounts of purified recombinant MBP-RH30, a helicase core mutant of MBP-RH30m or MBP, as a control, were added to the reactions in the presence or absence of ATP. Bottom: Representative native gel of 32P-labeled RNA products after the in vitro strand separation assay. Quantification of the partial dsRNA probe was done with a Phosphorimager. This experiment was repeated two times. (G) Increasing amounts (1.9 and 3.8 μM) of purified MBP-fusion protein or MBP (as a control) were added to the in vitro CFE assay #1. The 32P-labeled RNA products were detected by nondenaturing PAGE. The bottom image shows the contrasted image of the dsRNA bands of the top image.
Fig 8.
RH30 DEAD-box helicase inhibits the template recruitment by p33 and promotes the release of the viral (+)RNA from p33 replication protein in vitro.
(A) Top: Scheme of the in vitro assay with biotinylated RII(+) RNA from TBSV bound to streptavidin-coated magnetic beads. The scheme shows the order of addition of biotin-labeled RII(+) RNA, MBP-RH30 and MBP-p33C to the in vitro assay. The RNA probe and MBP-RH30 was allowed to form an RNP complex for 15 min, followed by addition of MBP-p33C protein, and incubation for 15 min. Then, the biotin-labeled RII(+) RNA—protein complex was captured on streptavidin-coated magnetic beads and washed the beads with a buffer. We eluted the proteins from the beads and measured the amounts of MBP-p33C in the eluates by Western blotting using anti-p33 antibody. Reduced amounts of MBP-p33C in the eluates mean that RH30 prevented the binding of p33C to the viral RNA, likely due to remodelling the RNA structure that could not be recognized by p33 any longer. Nonbiotinylated RNA (lane 1) was used as a control. (B) The scheme shows that the biotin-labeled RII(+) RNA, MBP-RH30 and MBP-p33C were added simultaneously to the in vitro assay. See additional details in panel A. (C) Top: The scheme shows that the biotin-labeled RII(+) RNA probe and MBP-p33C was allowed to form an RNP complex for 30 min, followed by capturing the biotin-labeled RII(+) RNA—protein complex on streptavidin-coated agarose beads. Then, we added MBP-RH30 protein with or without ATP, followed by incubation for 15 min and washing the beads with a small amount of buffer. Then, we measured the amount of MBP-p33C in the eluates by Western blotting using anti-p33 antibody. Increased amounts of MBP-p33C in the eluates mean that RH30 displaced p33C from the viral RNA, likely due to remodelling the RNA structure that could not be recognized by p33 any longer. Nonbiotinylated RNA (lane 6) was used as a control. Each experiment was repeated four times.
Fig 9.
Confocal microscopy shows co-localization of RH30 with the viral repRNAs in whole plants infected with CNV.
(A-B) Most of RH30 is re-targeted into the replication compartment where RNA synthesis takes place. The viral (-)repRNA and (+)repRNA carried six copies of the MS2 phage RNA hairpin (MS2hp) recognized by MS2 CP fused with RFP. The replication compartment was also marked by the BFP-tagged p33 replication protein in N. benthamiana. Note that RFP-MS2CP contains a weak nuclear localization, therefore this protein ends up in the nucleus in the absence of target RNAs in the cytosol. Expression of the above proteins from the 35S promoter was done after co-agroinfiltration into N. benthamiana leaves. The leaves of N. benthamiana plants were agro-infiltrated to express TBSV p33-BFP, GFP-RH30, RFP-MS2CP, repRNA(-)MS2hp or repRNA(+)MS2hp and the helper virus CNV20KSTOP gRNA. The repRNA(+)MS2hp consists of the repRNA(+) carrying six copies of cis-MS2 hairpin, which can be bound by RFP-MS2CP to show the subcellular localization of repRNA(+). The repRNA(-)MS2hp consists of repRNA(+) carrying six copies of trans-MS2 hairpin, which can only be recognized by RFP-MS2CP when viral RNA replication produces the complimentary strand repRNA(-) by the helper virus CNV20KSTOP. The absence of transient expression of GFP-RH30, repRNA(+)/(-)MS2hp or CNV20KSTOP were used as controls. The agro-infiltrated leaves were collected for confocal microscopy imaging 3.5 days post infiltration. Scale bars represent 10 μm. Each experiment was repeated.
Fig 10.
Co-localization of the viral double-stranded gRNA with RH30 in whole plants infected with CNV.
The CNV genomic dsRNA replication intermediate was detected via a dsRNA detector assay based on dsRNA binding-dependent fluorescence complementation assay [69]. The assay was performed with two dsRNA binding proteins (i.e., vp35 and B2), which are fused to N- and C-terminal halves of the yellow fluorescence protein (YFP), respectively. Simultaneous binding of the two fusion proteins to the same CNV dsRNA replication intermediate leads to the restoration of YFP fluorescence, allowing the visualization of the viral dsRNA replication intermediate location via confocal microscopy. The dsRNA sensor B2YN and VP35YC plasmids were agro-infiltrated into N. benthamiana leaves at OD600 of 0.15, respectively, together with RFP-RH30 and p33-BFP at OD600 of 0.5. CNV infection was initiated via agro-infiltration (OD600 of 0.15). Leaves were harvested and then immediately subjected to confocal microscopic analysis 2 days after agro-infiltration. The fluorescence complementation was detected via the GFP channel (excitation/emission: 488nm/500-530nm). Top panel: viral dsRNA replication intermediate is co-localized with RFP-RH30 within the replication compartment, which is marked by TBSV p33-BFP. Middle panel: no expression of RFP-RH30 was used as control. Bottom panel: N. benthamiana leaves with no viral components expressed were used as control. Expression of the above proteins from 35S promoter was done after co-agroinfiltration into N. benthamiana leaves. Scale bars represent 10 μm. Each experiment was repeated.
Fig 11.
Expression of AtRH30 DEAD-box helicase inhibits TCV and RCNMV genomic (g)RNA replication in N. benthamiana plants.
N. benthamiana plants expressing AtRH30 were inoculated with (A) TCV, and (B) RCNMV, respectively. Expression of the above proteins from the 35S promoter was done via co-agroinfiltration into N. benthamiana leaves. Top panel: Northern blot analyses of TCV gRNA and RCNMV RNA1 using 3’ end specific probes show reduced accumulation of TCV gRNA and RCNMV RNA1, respectively, in plants expressing RH30 than in control plants. Bottom panel: Ethidium-bromide stained gel to show 18S ribosomal RNA as a loading control. (C) Increased accumulation level of TCV in Arabidopsis RH30 knockout mutants based on Northern blot analysis. Samples in lanes 1 and 2 are from mock- inoculated Arabidopsis RH30 knockout mutants. See further details in panel A. (D) Semi-quantitative RT-PCR shows the induction of RH30 mRNA expression in Arabidopsis plants infected with TCV when compared the mock-inoculated plants. Each experiment was repeated. (E) Expression of RH30 and its mutant protein together with the cDNA of full-length TMV from the 35S promoter was done via co-agroinfiltration into N. benthamiana leaves. Top panel: Northern blot analysis of TMV gRNA and subgenomic RNA using a 3’ end specific probe shows reduced accumulation of TMV RNAs in leaves expressing RH30, but not RH30m in comparison with the control plants. Bottom panel: Northern blot analysis shows the 18S ribosomal RNA as a loading control. Each experiment was repeated.
Fig 12.
Models showing the antiviral functions of the plant RH30 DEAD-box helicase during TBSV replication.
Based on our current and previous data, we propose that the DDX17-like RH30 helicase interferes with several major steps during TBSV replication. First, RH30 interferes with the recruitment of the viral (+)RNA through unwinding RII(+)-SL cis-acting RNA element, which specifically binds to p33 replication protein only when the stem-loop structure is formed. Also, RH30 can potentially remodel the p33-(+)RNA complex, thus displacing p33 from the complex. Second: Inhibition of p33-(+)RNA complex formation by RH30 also leads to blocking the activation of the p92 RdRp, which requires the (+)RNA with the stem-loop structure in RII(+)-SL formed. Third, displacing p33 from the p33-(+)RNA complex by RH30 inhibits VRC assembly as well. This is because the stem-loop structure in RII(+)-SL is essential part of the VRC assembly platform. The cytosolic pool of RH30 is essential for the antiviral activity.