RNA interference in insects: the link between antiviral defense and pest control

RNA interference (RNAi) is a form of gene silencing triggered by double‐stranded RNA (dsRNA) that operates in all eukaryotic cells. RNAi has been widely investigated in insects to determine the underlying molecular mechanism, to investigate its role in systemic antiviral defense, and to develop strategies for pest control. When insect cells are infected by viruses, viral dsRNA signatures trigger a local RNAi response to block viral replication and generate virus‐derived DNA that confers systemic immunity. RNAi‐based insect pest control involves the application of exogenous dsRNA targeting genes essential for insect development or survival, but the efficacy of this approach has limited potency in many pests through a combination of rapid dsRNA degradation, inefficient dsRNA uptake/processing, and ineffective RNAi machinery. This could be addressed by dsRNA screening and evaluation, focusing on dsRNA design and off‐target management, as well as dsRNA production and delivery. This review summarizes recent progress to determine the role of RNAi in antiviral defense and as a pest control strategy in insects, addressing gaps between our fundamental understanding of the RNAi mechanism and the exploitation of RNAi‐based pest control strategies.


Introduction
The term RNA interference (RNAi) was coined to describe the potent and specific post-transcriptional gene silencing observed when double-stranded RNA (dsRNA) was introduced into the nematode worm Caenorhabditis elegans (Fire et al., 1998).Subsequent studies in other organisms, including the model insect Drosophila melanogaster (Zamore et al., 2000;Tomari et al., 2004), gradually revealed the underlying molecular mechanism and the crosstalk between post-transcriptional and transcriptional silencing.Briefly, dsRNA is recog-Correspondence: Jinzhi Niu and Jin-Jun Wang, Key Laboratory of Entomology and Pest Control Engineering, College of Plant Protection, Southwest University, Chongqing 400715, China.Email: jinzhiniu@swu.edu.cn and wangjinjun@swu.edu.cnnized by the endonuclease Dicer, which processes the long molecule into fragments of 21-24 bp, with short overhangs, known as short interfering RNA (siRNA).One siRNA strand is then captured by the protein Argonaute (Ago) to form the RNA-induced silencing complex (RISC).RISC scans for complementary mRNA molecules and targets them for degradation, ultimately suppressing gene expression.RNAi was later recognized as a conserved biological response to viruses (Li et al., 2002).The ability of RNAi to cause sequence-dependent silencing has been exploited for the functional analysis of genes in Drosophila (Flockhart et al., 2006) and in many other insects (Bellés, 2010).The concept of RNAibased insect pest control was first applied in beetles and moths, aiming to use ingested dsRNA to evoke the RNAi response against genes that are essential for insect development or survival (Baum et al., 2007;Mao et al., 2007).
Fig. 1 The link between RNAi in antiviral defense and as a strategy for pest control.As a natural antiviral defense system, the RNAi machinery is activated by viral dsRNA signatures (e.g., viral dsRNA structure formed by viral genome/RNA replication intermediate).Such viral dsRNAs are converted to siRNAs by Dicer, and single siRNA strands then assemble with Ago into the RNA-induced gene silencing complex (RISC), resulting in the targeted degradation of viralgenomic RNAs and transcripts.Exogenously delivered insecticidal dsRNA also trigger RNAi via the same RNAi machinery, but in this case the target is the transcript of a gene essential for development or survival.The degradation of target mRNAs causes developmental arrest and mortality, thus reducing the size of pest populations.
RNAi research in insects has diverged into two major fields, one focusing on understanding the RNAi mechanism, and its role in antiviral defense, and the other aiming to exploit RNAi for pest control.Insects are reservoirs for many viruses (Shi et al., 2016;Wu et al., 2020;An et al., 2022), and have therefore evolved a sophisticated immune system in which RNAi plays a key role (Bonning & Saleh, 2021).However, the RNAi-based antiviral defense mechanism in insects is not as well understood as other conserved immunity-related pathways, such as Toll, Imd, and JAK-STAT (Swevers et al., 2018).Similarly, although RNAi is a promising pest control strategy, the potency varies among insect species through rapid dsRNA degradation, inefficient dsRNA uptake/processing, and ineffective RNAi machinery, which limits its efficacy against many insect pests (Zhu & Palli, 2020).
As insects are major viral reservoirs, the antiviral RNAi response can be triggered when insecticidal dsRNA is applied exogenously as a pest control strategy, and both viruses and exogenous dsRNA activate the same RNAi machinery (Fig. 1).For effective RNAi-based pest control, it is therefore essential to understand the response to both triggers.Accordingly, this review summarizes recent progress to determine the role of RNAi as an antiviral defense mechanism and as a pest control strategy in insects, addressing gaps between the fundamental understanding of the RNAi mechanism and the application of RNAi in pest control.These studies will improve our understanding of antiviral immunity and RNAi-based pest control.

Systemic RNAi-based immunity in Drosophila and mosquito
The local RNAi response to viruses is known as cell-autonomous RNAi, but host tolerance to viruses typically involves the spreading of RNAi signals from local to distal tissues, thus requiring the activation of an environmental RNAi response as well as, eventually, a systemic RNAi response.In recent years, evidence has emerged for a systemic antiviral response in Drosophila as well as in the mosquito Aedes aegypti (Fig. 2).
A systemic antiviral RNAi response was first observed in the insect model Drosophila (Goic et al., 2013).Autonomous immunity was triggered in infected cells by viral dsRNA, but systemic immunity was then observed in uninfected cells, demonstrating the replication and spread of an unknown RNAi signal.Later, Drosophila hemocytes were shown to take up dsRNA from infected cells and, using an endogenous transposon reverse transcriptase, produce virus-derived complementary DNAs, known as vDNAs (Tassetto et al., 2017).The vDNAs provide a template for the de novo synthesis of secondary viral short interfering RNAs (vsiRNAs) which are secreted into exosome-like vesicles to confer systemic immunity.During infections with RNA viruses, defective viral genomes serve as templates for the synthesis of vDNA and circular vDNA (cvDNA), in a process regulated by the DExD/H helicase domain of Dicer-2 (Poirier et al., 2018).These vDNAs provide the source of secondary vsRNAs, and thus pass the RNAi signals from infected cells to noninfected cells.Long-term antiviral immunity can be established when these vDNAs integrate into the host genome as endogenous viral elements (EVEs).In A. aegypti, piwi-interacting RNA (piRNA)based immunity relies on the acquisition of EVEs, which produce antisense piRNAs that are preferentially loaded onto Piwi4 to inhibit virus replication (Tassetto et al., 2019).
Local RNAi not only confers systemic antiviral immunity by inducing a secondary RNAi response in distant cells, but can also integrate with other antiviral immunity-related pathways (Fig. 2).For example, Dicer interacts with the protein Vago, a ligand that induces JAK-STAT signaling (Paradkar et al., 2012).A nucleic acid-based response, distinct from cellular and humoral immunity based on peptide recognition, can therefore mediate antiviral defense.In this way, insects organize a more sophisticated antiviral immunity via crosstalk between RNAi and non-RNAi pathways.

Key factors contribute to RNAi efficiency variation in agricultural insect pests
The application of exogenous dsRNA to insects induces a potent RNAi response that can be used to silence, and therefore determine the function of, any genes.However, many RNAi-based gene functional studies in insects have revealed broad variations in RNAi efficacy among different insect pests (Zhu & Palli, 2020).This variation can be understood as bottlenecks at any stage of the RNAi process, and is also influenced by the presence of viruses that may induce a parallel antiviral response once the RNAi pathway has been triggered (Bonning & Saleh, 2021).
The first hurdle that must be overcome to achieve efficient RNAi is the avoidance of dsRNA degradation.Exogenously applied dsRNA can be rapidly degraded by enzymes present in insects, especially dsRNA-specific nucleases (dsRNases).The knockout or silencing of these dsRNases has been shown to improve RNAi efficiency in species that are already highly sensitive to RNAi, such as the Colorado potato beetle Leptinotarsa decemlineata (Spit et al., 2017), but also in species with medium sensitivity, such as the whitefly Bemisia tabaci (Luo et al., 2017), and in species with low sensitivity, such as the tobacco cutworm Spodoptera litura (Peng et al., 2021) and Asian corn borer Ostrinia furnacalis (Fan et al., 2021).This implies that dsRNases act generally against all dsR-NAs in insects, and their physiological function is therefore not entirely clear.Importantly, the level and activity of dsRNases varies among insect species.
Another key hurdle is the delivery of dsRNA, or more specifically the uptake of dsRNA by insect cells (Saleh et al., 2009).The uptake of dsRNA is mediated primarily by endocytosis (Xiao et al., 2015;Cappelle et al., 2016;Ye et al., 2021), a mechanism also utilized by plant viruses to enter insect midgut cells (Zhao & Lei, 2020).In nematodes, dsRNA uptake is mediated by RNAi defective protein 1 (CeSid1), but homologs are only found in some insects, suggesting it is not the general insect dsRNA uptake channel (Luo et al., 2012;Cappelle et al., 2016;Ye et al., 2023).Tissue-specific RNAi efficiency was enhanced in transgenic Spodoptera frugiperda (fall armyworm) lines expressing CeSid1 (Chen et al., 2021b).It is unclear which other mechanisms control the uptake of dsRNA in insects, but it is necessary to achieve the up-take of sufficient quantities of dsRNA to ensure effective pest control.
Interestingly, the core components of the RNAi machinery also vary in different insect species, reflecting differences in the siRNA, microRNA (miRNA), and piRNA pathways across different insect orders (Arraes et al., 2021).In addition, viral infections and the application of exogenous dsRNA result in different responses from the core RNAi genes across insect species (Niu et al., 2016;Ye et al., 2019).Overexpression of the core RNAi machinery in Bombyx mori (silkworm) larvae increased their susceptibility to RNAi (You et al., 2020).Interactions among the 3 RNAi pathways may also influence the efficiency of the siRNA pathway, because multiple Ago family genes contribute to the siRNA-mediated RNAi pathway in Locusta migratoria (locust) (Gao et al., 2020).It is likely that the efficacy of gene silencing improves when the RNAi machinery is more active, which is consistent with the behavior of the inducible RNAi machinery following dsRNA/virus challenge (Niu et al., 2016;Ye et al., 2019;Fan et al., 2022a).
The production of siRNA is also a key bottleneck affecting the efficacy of RNAi.During viral infections, the characteristics of virus-derived small RNA (vsiRNA) is mainly dependent on the activity of Dicer-2, resulting in different vsiRNA and siRNA sizes in different insect species (Santos et al., 2019).The dsRNA cleavage pattern also affects RNAi efficacy (Fan et al., 2022b).However, RNAi as an antiviral defense has been observed in fruit flies (Zhang et al., 2021), aphids (An et al., 2022), and whiteflies (Huang et al., 2021), among others, suggesting the mechanism is pervasive even though the efficacy of RNAi varies among species.
As discussed above, systemic RNAi-based virus tolerance requires both the spread and amplification of local RNAi signals, conferring uninfected cells with the ability to suppress viral replication.This phenomenon appears more limited in response to exogenous dsRNAs.Exosomes have been shown to mediate the spread of dsRNA in the Colorado potato beetle (Yoon et al., 2020), and similar extracellular vesicles appear to spread the RNAi signal in the red flour beetle Tribolium castaneum (Mingels et al., 2020).Transgenerational RNAi was also observed in Tribolium following the maternal transmission of long dsRNA (Horn et al., 2022).These data suggest that RNAi signals are systemically transmitted in RNAi-sensitive species such as beetles, resulting in high RNAi efficacy, but the amplification of these signals has yet to be reported following the application of exogenous dsRNA.The absence of RNA-dependent RNA polymerase (RdRp) in insects suggests that systemic RNAi in insects may involve a distinct mechanism, such as vDNA-mediated antiviral defense (Tassetto et al., 2017).Therefore, it seems that in these species, the viral infection can easily induce systemic RNA but the process does not appear to be robust for insecticidal dsRNA.Our improved understanding of insect systemic RNAi, which can be learned from insect-virus interactions, could help to better design the RNAi-based insect pest control approach, through dsRNA delivery, dsRNA modification, and dsRNA structure, etc.

Pipeline for the development of RNAi-based pest control products
The first RNAi-based product for the control of insect pests was a transgenic Zea mays (maize) variety expressing dsRNA targeting the snf7 gene in the western corn rootworm Diabrotica virgifera virgifera, as well as the Bt toxins Cry3Bb1 and Cry34Ab1/Cry35Ab1 (Darlington et al., 2022).A dsRNA spray-based product has also been approved to control the Colorado potato beetle (Rodrigues et al., 2021).Both products target RNAisensitive beetles, but other products are in development targeting more recalcitrant insect pests.The pipeline for the development of RNAi-based pest control products can be divided into 3 major stages: dsRNA screening and evaluation; dsRNA design and off-target management; and dsRNA production and delivery.

Larger-scale dsRNA screening and evaluation
Insecticidal dsRNAs are often discovered during the large-scale RNAi-based analyses of gene functions, revealing phenocopies of mutations that affect survival or development.Such genes are then selected for further evaluation as RNAi targets.For example, a largescale screen in Tribolium was carried out by preparing and injecting dsRNAs corresponding to 5000 randomly screened genes (Ulrich et al., 2015).The most promising lethal candidates were then used to identify orthologs in the western corn rootworm, which were tested by dsRNA ingestion (Knorr et al., 2018).Large-scale screening has also been carried out in less RNAi-sensitive species, such as the termite Reticulitermes flavipes and the western flower thrips Frankliniella occidentalis, each of which was used to evaluate 57 dsRNAs (Raje et al., 2018;Han et al., 2019).

dsRNA design and off-target management
One key advantage of RNAi-based pest control is that the sequence-dependent mechanism of action allows biosafety to be integrated during the dsRNA design stage.Target genes that have diversified among insect species can be selected, thereby restricting the effect to target pests alone, without risk to beneficial insects (Taning et al., 2021).Therefore, dsRNA design combines the identification of a lethal insecticidal effect (determined during initial screening and evaluation, as described above) with biosafety prediction in silico, based on the exclusion of off-target effects.The off-target effects of RNAi correlate with the number of mismatches between the dsRNA and nontarget mRNAs.For example, offtarget effects are triggered when dsRNAs share >80% overall sequence identity with target genes, contain ≥16 bp of contiguous, perfectly matched sequence, or contain segments of >26 bp of almost perfectly matched sequence, with single mismatches inserted between ≥5 bp of matching segments or mismatched couplets inserted between ≥8 bp of matching segments (Chen et al., 2021a).When combined with the prediction of Dicer activity, these parameters can be used for off-target risk prediction as the basis of rational dsRNA design (Wang et al., 2023).

dsRNA production and delivery
As dsRNA molecules are much larger than typical chemical insecticides, the application method must be designed to facilitate their uptake by insect pests, usually by oral ingestion (Nwokeoji et al., 2019).For herbivorous insects, it is possible to express the dsRNA in transgenic plants so that molecules are ingested as the insects feed on plant tissue or take in sap.For some insect pests, dsRNA can also be expressed in microbes or is chemically synthesized and then applied topically at their typical feeding sites.
For the protection of crops from insect pests, insecticidal dsRNAs can be expressed in transgenic or transplastomic plants, with the transgenic approach involving gene transfer to the nuclear genome and with the transplastomic approach involving gene transfer to the plastid genome.Examples of the nuclear transgenic approach include the expression of dsRNA to control the whitefly B. tabaci (Gong et al., 2022).The nuclear transgenic strategy has thus been proven suitable to control both chewing and sap-sucking pests.Interestingly, recent studies also showed that insecticidal dsRNA expressed in transplastomic plants also controls chewing and sap-sucking pests (Dong et al., 2022;Wu et al., 2022).However, dsRNA expressed from the plastid genome was found to be intact when ingested, followed by rapid degradation in the digestive system of the cotton bollworm Helicoverpa armigera, thus limiting the RNAi response (Fu et al., 2022).In contrast, dsRNA expressed from the nuclear genome was already processed into siRNA by the plant RNAi machinery, and the siRNA was more stable in the insect digestive system, thus enhancing the RNAi response (Fu et al., 2022).Insecticidal dsRNA can also be expressed in microbes, which are subsequently applied to plants (or other surfaces) where insects feed, resulting in the effective triggering of RNAi responses (Bento et al., 2020).
Larger-scale dsRNA or siRNA production can be achieved based on chemical synthesis, which is becoming more accessible through the falling costs per base.However, the delivery of dsRNA molecules can be challenging because the molecules are large and very sensitive to nucleases.These issues have been addressed using synthetic nanoparticle formulations, which not only protect the dsRNA from degradation but also facilitate its penetration across the insect body wall and cell membranes.For example, nanoclay-based dsRNA formulations can be applied as a foliar spray to control sap-sucking pests (Jain et al., 2022), transdermal dsRNA delivery has been shown to facilitate penetration in aphids (Zhang et al., 2022c), guanylated polymer formulations were shown to protect dsRNA from degradation in the beet armyworm Spodoptera exigua, and lipid complexes were shown to facilitate the endosomal escape of dsRNA taken up in the fall armyworm (Gurusamy et al., 2020).Galanthus nivalis lectin (GNA) has been fused to a dsRNA-binding domain to accelerate the delivery of dsRNA in lepidopteran midgut cells (Martinez et al., 2021).Furthermore, trunk injection supported the persistence of dsRNA in Malus domestica (apple) trees (Yan et al., 2020).

RNAi-based biological control
The intensive application of conventional insecticides is detrimental to the natural enemies of pests, such as ladybeetles, lacewings, and parasitoids, thus disrupting the natural biological control achieved in the food web.This has the unwanted effect of shifting pest management tools away from biological control and towards chemical control.In contrast, the specificity of RNAi can enhance pest control by achieving synergies with biological control agents, a concept known as RNAi-based biological control (Fig. 3).This inhibits pest populations with-out off-target effects on biological control agents in the ecosystem, enhancing their overall efficiency.
RNAi-based insect pest control triggers the RNAi response to silence target genes, with a direct lethal effect; in this sense, acting in a similar fashion to the application of conventional insecticides (Niu et al., 2018).The active principles of RNAi-based insect control are insecticidal dsRNAs, and these large biomolecules can synergize with other biological control agents, for example, enhancing the effect of Bt toxins (Caccia et al., 2020), increasing the virulence of entomopathogenic microbes by silencing immunity-related genes (Zhang et al., 2022b), or by working in concert with gut microbes (Xu et al., 2021).Biological control can be enhanced by RNAi in several ways: (1) insecticidal dsRNAs can specifically target the insect pests, with no nontarget effects on biological control organisms, such as predator and parasitoid; (2) insecticidal dsRNAs can target insect genes responsive to infection with entomopathogenic microbes, to enhance their virulence; and (3) insecticidal dsRNAs can manage the insect pest by indirect effects, adding to the biological control effects of the ecosystem.

Concluding remarks and future prospects
RNAi-based insect pest control has already been successful against RNAi-sensitive species such as the western corn rootworm and the Colorado potato beetle (Rodrigues et al., 2021;Darlington et al., 2022).As we learn more about the natural role of RNAi in viral immunity, the underlying mechanisms, and the factors that influence RNAi efficacy against agricultural pests, these studies will lead to more effective dsRNA screening and evaluation, dsRNA design and biosafety management, and dsRNA production and delivery.This will support the implementation of more RNAi-based insect pest control strategies, together with the RNAi-based control of pathogens.However, we should approach RNAi-based insect pest control and conventional insecticides differently, focusing on the following aspects of RNAi.
First, RNAi-based pest control differs from conventional chemical insecticides in terms of the mode of action.Furthermore, RNAi-based pest control is highly reliant on the RNAi response of the target insect, which is in turn dependent on the efficiency of dsRNA uptake, intracellular transport and processing into siRNA, and target mRNA degradation.This RNAi response overlaps with the endogenous defense system for sensing viral dsRNA.As insects act as viral reservoirs, the response to exogenous dsRNA is also influenced by their infection status (Bonning & Saleh, 2021).left) involves the use of conventional insecticides that always affect natural enemies of pests, such as parasitoids and predators, as well as beneficial insects such as pollinators.This type of control strategy reduces biodiversity, especially via non-target effects against beneficial organisms.In contrast, RNAi-based biological control (right) combines RNAi and biocontrol agents, not only avoiding off-target effects against natural enemies and beneficial insects, but also achieving synergy to enhance the overall biological control capacity in the ecosystem, reducing the need to apply conventional insecticides, and maintaining the service of pollinators.
Second, the efficacy of RNAi varies across insect species and the effects are exerted at different levels.The principal effect is to deplete target mRNA levels mainly by post-transcriptional gene silencing (the cleavage of mRNAs that have already been transcribed), but siRNAs can also form complexes that modify histone proteins to suppress gene expression at the transcriptional level, also promoting the co-translational cleavage of nascent mRNA.These effects are rarely 100% efficient, and gene expression can recover, so the effects of RNAi tend not to achieve the same potency as conventional insecticides.Even so, the efficacy of RNAi can be optimized by rational dsRNA design and off-target management, and efficient dsRNA production and delivery.Interestingly, environmental RNAi has been observed in insects following the administration of exogenous dsRNA, but systemic RNAi has not been reported (in contrast to the natural systemic antiviral RNAi response observed in Drosophila and mosquitoes).It is unclear whether exogenous dsRNA can trigger a systemic RNAi response, which may require factors that are unique to replicating viruses.Notably, systemic RNAi in plants is dependent on DCL2, which recruits RdRP6 to Ago1derived cleavage products, resulting in the more efficient amplification of secondary and transitive dsRNAs and siRNAs (Taochy et al., 2017).These RNAi signals can be trafficked from cell to cell by binding to members of the small RNA-binding protein 1 family (Yan et al., 2020).Intriguingly, the influx of Ca 2+ in response to tissue damage caused by mechanical injuries is sufficient to prime the antiviral RNAi defense response during insect feeding (Wang et al., 2021).Accordingly, systemic RNAi in plants should be investigated to enhance the efficiency of insecticidal dsRNA (Zhang et al., 2022a).
Finally, regarding the perception of RNAi as a pest control strategy, it is unsurprising that some insects exposed to dsRNA have already evolved resistance, including the western corn rootworm (Khajuria et al., 2018) and the Colorado potato beetle (Mishra et al., 2021).This reflects the mode of action of RNAi and the response triggered by dsRNA, and it is therefore important to justify the use of RNAi-based insect control.One key justification is that RNAi provides an inherent biosafety advantage, which is achieved by rational dsRNA design.RNAi can therefore be developed as a general pest control strategy for crop protection.Another important justification is that RNAi is compatible with other pest control strategies, which may help to address both the limited efficacy of RNAi and the potential for target pests to evolve resistance.For example, the integration of RNAi with biological control (RNAi-based biological control) is a promising strategy to maintain the size of pest populations below a given economically important threshold.

Fig. 2
Fig. 2 The RNAi response to viral infection in Drosophila and mosquito.During viral infections, viral dsRNA signatures (e.g., viral dsRNA structure formed by viral genome/RNA or replication intermediate) trigger the local RNAi response (middle) which results in the targeted degradation of viral dsRNA.However, the viral dsRNA is also reverse transcribed into virus-derived complementary DNAs (vDNAs), which act as templates for the de novo synthesis of secondary viral short interfering RNAs (vsiRNAs) which are secreted in exosome-like vesicles to confer systemic immunity (left).Furthermore, there is cross-talk between the RNAi pathway and other immunity-related pathways, as seen when Vago transduces the RNAi-based antiviral response to the Jak-STAT pathway (right).

Fig. 3
Fig.3Concept of RNAi-based biological control.Insecticide-based chemical control (left) involves the use of conventional insecticides that always affect natural enemies of pests, such as parasitoids and predators, as well as beneficial insects such as pollinators.This type of control strategy reduces biodiversity, especially via non-target effects against beneficial organisms.In contrast, RNAi-based biological control (right) combines RNAi and biocontrol agents, not only avoiding off-target effects against natural enemies and beneficial insects, but also achieving synergy to enhance the overall biological control capacity in the ecosystem, reducing the need to apply conventional insecticides, and maintaining the service of pollinators.