Write a comment

Summary:

This article explores the development of a novel RNA-based method for crop protection against plant viruses such as Cucumber Mosaic Virus (CMV), a widespread and economically damaging pathogen. The study centres around "effective small interfering RNAs" (esiRNAs), which are siRNAs specifically identified as being capable of efficiently silencing target viral RNA. By topically introducing select esiRNAs, the authors were able to enhanced Nicotiana benthamiana resistance to CMV infection. This builds upon previously described RNAi approaches, involving exposure of plants to large double-stranded segments of viral genomes or mRNA, often resulting in inefficient viral defence.

The authors utilised a previously-published in vitro screen (termed “slicer assay”) to compare the RNAi activity of siRNAs selected from a pool of candidates generated from N. tabacum BY2 cytoplasmic extracts. Then, the capability of putative esiRNAs to offer robust protection against CMV was tested by topical application alongside the virus in N. benthamiana leaves. With the aim of enhanced silencing efficiency further, the authors designed "effective dsRNAs" (edsRNAs), engineered to be preferentially processed into several potent esiRNAs. Such constructs could potentially offer cross-protection against different viral variants – an advantage for plants faced with rapidly mutating pathogens like CMV.

Overall, the study demonstrates the potential of esiRNAs and edsRNAs as powerful tools in RNA-mediated plant protection and represents a significant advancement over current dsRNA-based strategies, by potentially increasing both the efficacy and robustness of antiviral responses.

General comments:

·      The manuscript is a well-written body of work which describes the issues facing deployment of RNA actives in agriculture, and proposes a potential promising solution. esiRNAs and edsRNAs seem to be a genuine improvement over prior dsRNA approaches and this is convincingly demonstrated. Below we have suggested some improvements to the figure design and highlighted parts of the text which might need clarification.

·      A major limitation of this manuscript is that it only examines the effects of RNA actives when co-delivered with the virus in plants. We feel it is important to at least discuss this, if not investigate it experimentally, as this might pose a major hurdle to the implementation of such topically applied treatments in a field and/or glasshouse setting.

Introduction:

The authors did a very good job of explaining relevant RNAi concepts to non-expert audiences, and emphasising why topical application of RNAi products would be beneficial in agriculture. A few minor comments:

·      The abbreviation ‘a-sites’ is used only five times in the entire text. Perhaps it would be worth spelling out ‘accessible sites’ for clarity?

·      On page 4 – ‘However, the production of transgenes is time-consuming and costly, and their release is prohibited in many countries due to safety concerns.’ We would argue that release of transgenics is prohibited due to negative public perception, not safety concerns backed up by evidence. It is important to sustain an informed and evidence-based discussion of the matter.

·      On page 4 – ‘Furthermore, the plants are potentially susceptible to infection by variants of the pathogen against which they were bred for resistance.’ This sentence could be clarified to avoid misinterpretation e.g. emphasising that the plants are not potentially susceptible to the specific viral variants they were bred to resist, but rather that they might be susceptible to variants they were NOT bred to resist.

·      On page 10 – Could a reference be included to back up the statement that RNA 2 and 3 are essential for the CMV lifecycle? Or is there evidence that RNA 1 is not essential?

Figure 1 and S1:

·      No major comments – we thought the experimental approach and supporting data were presented clearly.

Figure S2:

·      (nt) label is missing from x-axes in S2B.

Figure 2:

·      It would be nice if the authors could clarify why particular esiRNA candidates were carried forward for further analysis above others. For example, siR1489 and siR1613 appear to have very similar efficiency in 2A, but only siR1489 was carried forward.

·      Some of us found the number of asterisks on slicer assay figures a little distracting, and thought labelling of the intact RNA (already highlighted with an arrow) would be sufficient for interpretation. This comment applies to 2A, 2C, 4C, S4A and S6.

·      A scale bar for schematics in 2B and 2D would be nice. This also applies to S1B.

Figure 3:

·      While flipping between Figures 2 and 3, we found it difficult to keep track of which numbered esiRNAs were efficient in the slicer assays, and which were effective in planta. Perhaps a copy of the relevant lanes from 2A and 2C could be included next to the corresponding plant photos in 3A and 3C, respectively? Additionally, colour-coding the photos in 2A and 2C (perhaps with a coloured border or marker in the corner) to match the lines in 2B and 2D, respectively, would make interpretation easier.

·      We liked that siR1844 and siR2634 were included as negative controls in 2A-B, and siR359 as a positive control in 2C-D, but felt these could be emphasised more e.g. with colour-coding or an additional label.

·       Some of the less protective esiRNAs in 3B seemed to cleave RNA 2 perfectly well in 2B – could authors comment on this?

·       On page 12 – ‘both the total number and the cleavage efficiencies of the esiRNA candidates identified for CMV RNA 3 were significantly lower than for CMV RNA 2’. Is this definitely the case? Across 2A and 2C, the number and cleavage efficiencies of esiRNA candidates look quite similar. Although the differences in efficiency are more obvious when comparing Tables 1-2, we are not sure the word ‘significantly’ should be used in the absence of statistical tests. Instead of the reported lower efficiency in slicer assays, alternative explanations for the superior protection offered by esiRNAs targeting RNA 2 could be:

o   RNA 2 encodes the viral polymerase (as discussed by the authors on page 18).

o   RNA 2 also encodes the viral suppressor of RNA silencing (VSR), so suppression of this genomic region may be especially beneficial for a plant protected via RNAi.

·       Scale bars on the photos in 3A and 3C would be nice, as would some close-up images of leaves on each plant to show symptoms (or lack thereof) in more detail e.g. ‘mosaic symptoms, leaf deformation, systemic necrosis, chlorosis’, as described on page 4.

·      Could the authors explain the difference in timescales between 3B and 3D? Were the plants represented in 3D too heavily infected to continue?

Figure S3:

·      How the numbering of plants corresponds to those presented in Figure 3 is a little confusing.

·      Plant 15 is used as a negative control without reverse transcriptase (w/o RT), yet this plant was not tested for viral cDNAs in subsequent lanes.

Figure 4:

·      In 4C, dsCMV6-21o appears to be less efficient in the slicer assay than the siR mix, especially on the AGO1 panel where some intact RNA 2 remains. Can the authors comment on this? Nevertheless, we understand that dsRNAs currently have more potential in the field, when applied to plants challenged with multiple pathogens/variants.

Figure S5:

·      “Courier New”  might be a better choice of font for nucleotide sequences, given the even spacing of letters.

Figure 5:

·      We wondered if rearranging the siRNAs in the long dsRNAs could impact efficiency in the slicer assay, and indeed in planta (Figure 6). Especially if there was clearly variation in the amount of 21 nt fragments produced at different positions along the dsRNA in 5B – is this a sequence or position effect?

Figure 6:

·      It is impressive that dsCMV6-21o clearly outperforms dsCMV in the plant protection experiment! Seeing as some of the most effective esiRNAs e.g. siR1172 (which prevented symptoms in 100% of plants in 3B) were included in dsCMV6-21o, this suggests that stringing together multiple esiRNAs does no harm in this system. However, what would happen if the authors designed a dsRNA with only less effective esiRNA candidates e.g. those targeting RNA 3? Together in one dsRNA, could they confer enhanced protection, as opposed to when individually applied?

Discussion:

The discussion did a very good job of summarising the results and placing them in context. Below are some additional comments and questions:

·       We think the main limitation of this study stems from the inoculation method used i.e. delivering both viral genomes and siRNAs to same host cells at the same time. There seems to be substantial uncertainty in the field surrounding the efficacy of RNA actives if treatment application is separated from infection in space and/or time. As the application of edsRNAs as a plant protection strategy is central to the narrative of the paper, especially as this is featured in the title, we feel it is important to thoroughly engage with the outstanding challenges and limitations of this technology. For example, would the siRNAs still provide protection if delivered days or centimetres apart from the virus? If experimentation was done to probe these questions, we feel it would greatly complement the existing body of work.

·       As rub-inoculation would likely be unfeasible for RNA delivery in an agricultural context, effective RNA encapsulation and delivery tools will be necessary to ensure successful future application of this research. Perhaps the potential/development of these tools could have been discussed in greater detail.

·       In the introduction, a transgenic approach was mentioned as a potential method to deploy siRNAs for the purposes of plant defence in the field. While acceptance of transgenic plants in Europe is lagging, it is worth noting the considerable progress achieved elsewhere in the world. We feel it is worth discussing transgenic plants as an alternative way to implement edsRNAs.

·       While targeting of RNA 2 seemed to be especially effective for reducing CMV symptoms in N. benthamiana, perhaps targeting of RNA 1 or 3 could be effective in other host species? Assuming they can they be infected and treated with exogenous RNA in a similar manner, it would be interesting to see if the esiRNAs and dsRNAs designed in this study could reduce symptoms on other CMV hosts. In future, it might also be interesting to test the efficiency of esiRNAs derived from other host species.

Competing interests

The authors declare that they have no competing interests.