Inhibition of RNA recruitment and replication of an RNA virus by acridine derivatives with known anti-prion activities

Abstract

Background: Small molecule inhibitors of RNA virus replication are potent antiviral drugs and useful to dissect selected steps in the replication process. To identify antiviral compounds against Tomato bushy stunt virus (TBSV), a model positive stranded RNA virus, we tested acridine derivatives, such as chlorpromazine (CPZ) and quinacrine (QC), which are active against prion-based diseases.

Methodology/principal findings: Here, we report that CPZ and QC compounds inhibited TBSV RNA accumulation in plants and in protoplasts. In vitro assays revealed that the inhibitory effects of these compounds were manifested at different steps of TBSV replication. QC was shown to have an effect on multiple steps, including: (i) inhibition of the selective binding of the p33 replication protein to the viral RNA template, which is required for recruitment of viral RNA for replication; (ii) reduction of minus-strand synthesis by the tombusvirus replicase; and (iii) inhibition of translation of the uncapped TBSV genomic RNA. In contrast, CPZ was shown to inhibit the in vitro assembly of the TBSV replicase, likely due to binding of CPZ to intracellular membranes, which are important for RNA virus replication.

Conclusion/significance: Since we found that CPZ was also an effective inhibitor of other plant viruses, including Tobacco mosaic virus and Turnip crinkle virus, it seems likely that CPZ has a broad range of antiviral activity. Thus, these inhibitors constitute effective tools to study similarities in replication strategies of various RNA viruses.

Figure 1. Inhibition of TBSV gRNA accumulation in N. benthamiana protoplasts by treatment with CPZ or QC.

Northern blot analysis with a 3′ end specific probe was used to detect the accumulation levels of TBSV gRNA and subgenomic (sg)RNAs. The N. benthamiana protoplasts were treated with the shown concentrations of CPZ and QC either before or after electroporation. The samples were harvested 40 hours after electroporation. The ethidium-bromide stained gel at the bottom shows the ribosomal (r)RNA levels as loading controls. As a control, DMSO, the solvent for CPZ and QC, concentrations were the following: 0.04%, lanes 2,6; 0.005%, lanes 3,7, 12; 0.01%, lanes 4,8,13; and 0.02%, lanes 5 and 9. Note that the gRNA can reach rRNA levels in N. benthamiana protoplasts. The survival of the plant cells (after electroporation and treatment) was checked by measuring rRNA levels in total RNA extracts.

Figure 2. Inhibition of in vitro replication of TBSV repRNA in a cell-free extract by CPZ and QC.

(A) The first in vitro replication assay contained a yeast cell-free extract, purified recombinant p33 and p92^pol, a mixture of radiolabeled and unlabeled nucleotides and was programmed with TBSV DI-72 (+)repRNA. CPZ, QC and DMSO were added to the assay at 0, 20, 40 or 60 min after the start of the assay, which lasted for 3 hours. The newly made repRNA products were analyzed on denaturing PAGE. Note that the repRNA goes through a full replication cycle in this replication assay, producing mostly (+)-stranded progeny RNAs. (B) The second in vitro replication assay was similar to that described in panel A, except that the membrane fraction of the yeast cell-free extract was treated with CPZ, QC or DMSO first, followed by a washing step to remove the unbound compounds. This was followed by the addition of the soluble fraction of the yeast cell-free extract, purified recombinant p33 and p92^pol, a mixture of radiolabeled and unlabeled nucleotides and was programmed with TBSV DI-72 (+)repRNA. DMSO concentrations were the following: 0.05%, lanes 1, 2, 3, 7, 9, 10 and 0.02% in lanes 4, 5, 6, 8, 11, and 12. The experiments were done in duplicates and the right panel shows reproducibility. The data show the averages and standard deviations including both sets. (C) The third in vitro replication assay contained a partially purified tombusvirus replicase preparation, a mixture of radiolabeled and unlabeled nucleotides and was programmed with TBSV DI-72 (+)repRNA. Note that the repRNA can only produce the complementary (minus-stranded RNA) product in this assay. (D) The fourth in vitro replication assay was similar to that described in panel C, except (−)-stranded repRNA was used to generate (+)-stranded RNA products. Note that the solubilized/purified tombusvirus replicase also generates internally initiated (marked as “ii”) products on (−)RNA template.

Figure 3. Inhibition of in vitro binding of p33 replication co-factor to TBSV (+)RNA by QC and CPZ.

(A) The in vitro band-shift assay included the radiolabeled RII(+)-SL sequence, which is present within the p92 ORF in the TBSV gRNA, and 1.0, 0.4, 0.2 or 0.1 µg of purified recombinant p33C as well as CPZ (50 µM) or DMSO (0.01% final concentration). p33C is a soluble, N-terminally truncated version of p33 (contains the RNA-binding domain as well) fused to MBP. Purified recombinant MBP was used as a negative control. The binding was analyzed in 5% nondenaturing PAGE. The bound RNA in the DMSO control (lane 8) was chosen as 100%. Quantitation showed that the free RNA negatively correlated with the bound RNA. (B) Similar in vitro band-shift assay using QC (20 µM final concentration) as an inhibitor.

Figure 4. Inhibition of in vitro translation of TBSV RNA by QC.

(A) The wheat germ translation assay was programmed with artificial capped p33 mRNA carrying a poly(A) tail in the presence of CPZ, QC and DMSO as control. The radiolabeled p33 product was analyzed on SDS-PAGE. Each experiment was repeated three times. DMSO sample (0.03%) was chosen as 100%. (B) Similar in vitro translation assay was programmed with the uncapped wt TBSV gRNA, which does not carry a poly(A) tail.

Figure 5. Inhibition of TBSV gRNA accumulation in N. benthamiana plants treated with CPZ.

(A) Leaves were first infiltrated with CPZ (600 µM), followed by inoculation of the same leaves with TBSV virion preparation. Samples for viral RNA analysis were taken from the infiltrated leaves at 4 dpi. Northern blotting (top panel) shows the level of TBSV gRNA and sgRNAs accumulation in individual samples using a 3′ end specific probe. The bottom panel shows the ethidium bromide stained gel indicating the levels of rRNA and TBSV gRNA. Each experiment was repeated three times. DMSO sample was chosen as 100%. (B–C) The delay in symptom development due to TBSV infections in the CPZ treated plant (shown on the left) 4 and 14 dpi indicates the potent antiviral activity of CPZ. Comparable DMSO treatment of plant leaves prior to inoculation with TBSV did not protect the plants from infection.

Figure 6. Inhibition of TBSV gRNA accumulation in N. benthamiana plants treated with QC.

(A) Infiltration of leaves with QC (1600 µM), inoculation with TBSV and analysis of RNA samples were done as described under Fig. 5. Note that QC treatment resulted in delay of TBSV symptom formation when compared with DMSO-treated control plants. Note the 40% accumulation of TBSV RNA in one sample (marked with a star), which is likely due to areas in the leaf not (or partly) soaked by the compound during infiltration. Each experiment was repeated three times. DMSO sample was chosen as 100%. (B) The delay in symptom development due to TBSV infections in the QC treated plant (shown on the left) 10 dpi indicates an antiviral activity for QC when compared with the DMSO treatment of plant, which shows more severe symptoms (coloring and stunting).

Figure 7. Inhibition of the distantly related TMV RNA and the closely related TCV RNA accumulation in N. benthamiana protoplasts by treatment with CPZ or QC.

(A) Northern blot analysis with a 3′ end specific probe was used to detect the accumulation levels of TMV gRNA. N. benthamiana protoplasts were treated with the shown concentrations of CPZ and QC either before or after electroporation. See further details in the legend to Fig. 1. Note that the % of TMV RNA accumulation was normalized based on rRNA levels in the same samples to correct for sample-to-sample variation. DMSO concentrations were the following: 0.03%, lanes 3, 6, 11; 0.04%, lanes 4, 7, 12; 0.06%, lanes 5, 8, 9, 13, 14; and 0.08% in lanes 10 and 15. (B) Ethidium-bromide stained agarose gel electrophoretic analysis of the accumulation levels of TCV gRNA. N. benthamiana protoplasts were treated with the shown concentrations of CPZ and QC either before or after electroporation. See further details in the legend to Fig. 1. DMSO concentrations were the following: 0.005%, lanes 1, 4; 0.01%, lanes 2, 3, 5, 6; 0.02%, lanes 7, 10; and 0.06% in lanes 8, 9 and 11. (C) Comparison of the inhibitory effect of CPZ on TCV versus TBSV RNA accumulation. Ethidium-bromide stained agarose gel electrophoretic analysis of the accumulation levels of TCV and TBSV gRNAs from the same N. benthamiana protoplasts preparations. The protoplasts were treated with the shown concentrations of CPZ 30 min after electroporation. See further details in the legend to Fig. 1. DMSO concentrations were the following: 0.02%, lanes 2, 4, 6, 8; 0.04%, lanes 3, 5, 7 and 9.

Figure 8. The proposed inhibitory effects of CPZ and QC treatments on various steps of TBSV RNA replication.

Five early steps of TBSV replication , are shown schematically, including translation of the viral RNA, template selection, viral RNA/protein recruitment into replication, assembly of the viral replicase complex and viral RNA synthesis. While the inhibitory effect of QC has been indicated in several steps, CPZ has been shown to affect mostly one step. The viral cis-acting RNA structures, such as p33RE (p33 recognition element), RSE (replication silencer element) and gPR (genomic promoter), which are likely bound by QC, are magnified.