Cross-kingdom affairs: Can microRNAs of host plants regulate genes of their insect herbivores?

Abstract

Plants often fend off insect herbivores using induced chemical defenses. Several small RNAs, including microRNAs (miRNAs), are also induced in plants after herbivory. Since the herbivore ingests the small RNAs induced in the leaf, we hypothesized that they regulate the insect herbivores' gene expression. Using miRNA as the small RNA model, we tested this hypothesis in the Arabidopsis thaliana-Plutella xylostella system. First, we experimentally tested the ingested miRNA-mediated gene regulation as a proof of concept in Plutella xylostella using its own let-7 miRNA. Pre-let-7 and mature-let-7 feeding led to target downregulations. Larvae also showed growth and developmental defects, validating the functionality of the trophically acquired miRNA. Then, we tested the hypothesis that the host plant's miRNAs regulate the herbivore's genes. In the in silico analysis performed with three different target prediction algorithms, 128 plant miRNA-target insect mRNA pairs (CP) were common. The mature CP miRNAs had structural features similar to positive control P. xylostella and AGO1-binding Drosophila melanogaster miRNAs (positive controls). 3D modeling showed high structural similarity between the plant (Ath-Dicer-like-1) and insect (Px-Dicer-1) dicer-1s. The average pre- miRNA, stem, and loop lengths and the dicing site distributions in CP and P. xylostella miRNAs were similar, hinting that the CP miRNAs were amenable to cleavage by Px-Dicer- 1. Thus, the in silico analysis indicated that the insect miRNA processing pathway can process plants' pre- and mature miRNAs. For the empirical validation, we first analyzed the response of CP miRNAs to herbivory. Thirteen CP miRNAs were induced after 1-hour of herbivory. Three CP miRNAs, miR-164, miR824, and miR447, were found to target glucosinolate sulfatase (GSS), P. xylostella's counter-defense protein against plant glucosinolates. We selected this target gene for further study. Larvae fed on miR164, miR824, and miR447 showed reductions in GSS transcripts. As miR164 showed continuous induction upon herbivory, we selected it as a candidate. Larvae fed on miR164 via a glucosinolate-replete diet showed ≥2.2-fold higher mortality, ≥1.5-fold longer first instar duration, and ≥0.44-fold lower mass than controls, respectively. Such an effect was not observed when these miRNAs were fed via a glucosinolate-deplete diet. Glucoraphanin was reduced by >2.1 folds in the miR164 fed larvae than controls; similarly, desulfoglucoraphanin, the larvae's glucoraphanin-detoxification product, was reduced by >5.9-folds. Sulforaphane, the insecticidal product, increased ≥4.5-fold in the miR164-fed larvae. These results suggested that larvae's differential fitness was associated with their differential glucosinolate detoxification efficiencies. Last, we studied the miR164-mediated GSS downregulation's ecological role by analyzing P. xylostella's natural enemies' preferences towards and performances of the miR164-fed larvae. Diadegma semiclausum (parasitoid) and Chrysoperla carnea (predator) did not show a differential preference for the miR164-fed larvae. However, we found that only D. semiclausum larvae showed >7.7-fold mortality in the pre-and mature-miR164-fed P. xylostella larvae via a glucosinolate-replete diet than controls, suggesting its association with the isothiocyanate content in larvae. Together, this study provides in silico and empirical evidence of plant 'miRNAs' cross-kingdom role in plant defense against an insect herbivore.