Write a comment

Reviewer summary:

This preprint investigates the biogenesis, mobility, and functional loading of Cuscuta campestris trans-species microRNAs (miRNAs) that move into host plants and silence host gene expression. The authors combine spatial and temporal expression analyses, artificial miRNA–mediated knockdowns, RNA biochemistry, and Argonaute immunoprecipitation approaches to demonstrate that these miRNAs are produced by parasite DCL1, accumulate as miRNA/miRNA* duplexes within parasite tissue, and selectively avoid loading onto parasite Argonautes. Upon transfer into host tissue, the trans-species miRNAs become loaded predominantly onto host AGO1. The authors previously showed that these trans-species miRNAs then proceed to direct canonical RNA silencing activity of host genes.

The study provides a coherent mechanistic framework for how Cuscuta produces a class of “export-only” miRNAs that remain inactive within the parasite yet become functional upon host entry. By clarifying the timing of transcription, processing, movement, and AGO loading, this work significantly advances our understanding of cross-species RNA communication in parasitic plant–host interactions and establishes a foundation for future exploration of RNA-based virulence mechanisms.

General comments:

• The manuscript is clearly written, logically structured, and easy to follow.

• Figures are clear and well designed, with illustrations that strongly support interpretation of the results.

• Background information is helpful, accurate, and well-integrated into the narrative.

• There is relatively little to criticize; most comments are suggestions for clarification, interpretation, or future directions rather than major concerns.

Introduction:

• Lines 70–77: There is evidence suggesting that mobile Arabidopsis small RNAs are not exclusively secreted in extracellular vesicles. Mentioning this nuance could strengthen the discussion of RNA mobility mechanisms.

• Line 76: Please define “small RNAs (sRNAs)” at first usage or consider avoiding the term if it is not used consistently thereafter (instead using siRNAs and miRNAs explicitly).

• Line 82: Please confirm whether “MIRNA” is correctly formatted (uppercase italics).

Results:

Figure 1:

• The schematic illustration is clear and effective.

• Please re-check the logarithmic scaling of the Y-axis. The “log10” label may be redundant, as the values already appear log-transformed (e.g., log₁₀1e+00 = log₁₀1 = 0).

• Including an uninfected N. benthamiana negative control would help demonstrate that basal petiole tissue contains negligible amounts of Cc-miRNAs.

• Consider assaying infected N. benthamiana leaves, which would complement the basal petiole analysis and determine whether there is movement of mRNA towards the leaf instead of the stem.

• Are host targets of the Cc-miRNAs (e.g. NbSEOR1 or NbTIR1; lines 110–112) also reduced in areas 1–3 relative to area 4? Evidence of spatial regulation of host gene expression near the haustorium would be interesting.

• What host factors are involved in miRNA movement? Are plasmodesmata or vascular tissues implicated? Is it possible to use different hosts/mutants to assess the contributions of host tissue architecture to trans-species RNA movement?

• Have plant-derived RNAs been detected in Cuscuta tissues (areas 5–9), suggesting bidirectional RNA movement?

Figure 2:

• Lines 110–112: Were the two Cc-miRNAs selected due to particular biological relevance, abundance, or mobility?

• Please clarify why an additional Cc-miRNA was included in Figure 2 but not Figure 1.

• Lines 170–174: Additional experiments could help distinguish whether miR12463a is inherently less mobile or simply below detection limits. More generally, does miRNA abundance influence mobility?

• Figure 2D: Consider rearranging the X-axis order to place “Cc tip” and “host stem” adjacent to one another, emphasizing precursor accumulation in the parasite prior to maturation and transport.

• Please add a brief explanation of blue-light–induced haustoria, either in the main text, legend, or Methods.

Figure 3:

• It would be informative to test whether host target gene mutants (e.g. bik1 in Arabidopsis) or amiR-mediated knockdowns in N. benthamiana alter Cuscuta attachment efficiency.

• Extended Data 2: We appreciate that this graph includes all the information from these five trials, but perhaps its clarity could be improved. Successful and failed are difficult to distinguish. The percentages seem most important, and perhaps could be summarized in an additional graph or table?

• The apparent partial knockdown of CcDCL1 is intriguing. Could amiR-CcDCL1 act primarily through translational repression rather than transcript cleavage?

• Why is partial CcDCL1 knockdown sufficient to dramatically reduce Cc-miRNA accumulation?

• Could CcDCL1 expression be assessed spatially (e.g. enriched in haustoria versus other parasite tissues, similar to Figure 1)?

• Please clarify the meaning of “FALSE” and “TRUE” in panel F.

Figure 4:

• In panel B, consider using distinct colours or clearer legend formatting to distinguish the three ccm-miRNAs.

• Please clarify which miRNA families possess 5′U versus 5′C.

• In the TRAPR assay, could these RNAs be bound by non-AGO RNA-binding proteins?

• Trans-species miRNAs appear highly enriched in the TRAPR fraction relative even to miR159. Is there a mechanistic explanation for the remarkable association of trans-species RNA with RNA-binding proteins?

• In panel E, please label miR159 as ccm-miR159 and specify CcAGO for clarity, both in the figure and in lines 312–313, with an appropriate citation for that statement.

• Please correct the typo “Caluculate” in panel A.

• Additional Y-axis clarification in panel E would help improve clarity (that input and TRAPR fractions are being compared here).

Figure 5:

• As noted in the Discussion (lines 404-406), testing additional AGOs like AGO5. AGO10 might also be worth testing, as it binds to 5’U miRNAs like AGO1, though AGO10 is not highly expressed.

• Yellow text is difficult to read when printed; please consider improving contrast.

• Immunoblot control would help show that the IP works and that the plants express the protein properly, though this is already somewhat shown by the miRNA qRT-PCR controls.

Discussion:

• Are trans-species miRNAs potentially held as a duplex to avoid “self-targeting” of Cc genes? This is mentioned but could be discussed further.

• Please comment on the limitations of ΔCt versus ΔΔCt normalization for qRT-PCR experiments. How does this reflect on your results?

• Line 309: Consider whether “naïve” should include a diacritic (ï versus i), depending on journal style.

• Please expand discussion of possible mechanisms governing miRNA duplex export from Cuscuta. Is the duplex perhaps a key feature for export selection in Cc? What molecular mechanisms could be involved in the export of Cc miRNAs?

• How do trans-species RNA duplexes survive in the extracellular environment in the presence of RNAses? Are trans-species RNAs in the vasculature, or do they immediately translocate to plant cells proximal to infection, and travel through plasmodesmata? Do these miRNAs move primarily through vasculature, via plasmodesmata, or another mechanism?

• Line 368: Please correct the typo “produces”

• Do trans-species RNA travel by similar mechanisms to those described for plant mRNA and miRNA that travel long distances? In other words, has Cc co-opted existing plant mechanisms for RNA transport through plant tissue. How are miRNAs imported into host cells?

• How are trans-species miRNAs imported by plant cells?

• How is selective miRNA accumulation in haustoria regulated?

• What is the functional benefit of trans-species miRNAs for Cuscuta - do they directly facilitate infection, as suggested by the amiR-DCL experiments?

• Line 424: Consider removing the abbreviation introduced here, as it is not reused much.

• How does the miRNA duplex avoid binding to the host AGO? We like the discussion of RNA modifications.

Methods:

Please specify the amount of RNA used for cDNA synthesis and clarify whether it was held constant across samples. This is particularly important when using single ΔCt normalization.

Acknowledgements:

YN is credited with having performed “most experiments.” Please clarify whether this should instead read “all experiments,” if accurate. If not, consider crediting the individual of group of individuals responsible.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.