Control of tomato brown rugose fruit virus (ToBRFV) in tomato plants using in vivo synthesized dsRNA

Abstract

The tomato brown rugose fruit virus (ToBRFV) is an increasingly prevalent pathogen that poses a threat to the global tomato industry. Topical application of dsRNA has shown promise as an effective tool to control many pathogens, including viruses; however, it this has not yet been demonstrated for ToBRFV. In this study, ToBRFV-specific long dsRNA molecules were synthesized in vivo by incorporating parts of its genome into that of bacteriophage phi6, thereby enabling the amplification of the chimeric dsRNA in Pseudomonas syringae. Co-inoculation of ToBRFV and purified, high-quality (hq)-dsRNA onto tomato (Solanum lycopersicum) plants resulted in reduction of both viral RNA levels and disease symptoms. Functional analysis of the hq-dsRNA response against the virus revealed its independence of RNA-DEPENDENT RNA POLYMERASE 6 (RDR6) and SUPPRESSOR OF GENE SILENCING 3 (SGS3). In addition, non-infected plants showed a mild activation of innate immune responses upon hq-dsRNA treatment, including accumulation of callose at plasmodesmata. Overall, our results provide evidence for hq-dsRNA as a tool for controlling ToBRFV in tomato plants, and demonstrate the potential of in vivo produced dsRNA in the battle against crop pathogens.

Application of in vivo-produced dsRNA specific to tomato brown rugose fruit virus reduces both viral RNA levels and disease symptoms in tomato plants, and triggers mild activation of innate immune responses.

Introduction

One of the major global challenges for the 21st century is providing food security for the rapidly growing human population. To meet these challenges, it is crucial to develop sustainable crop protection strategies that reduce the environmental and ecological footprint and the health risks associated with chemical applications. A promising approach for plant protection is RNAi, which is characterized by high specificity while potentially minimizing effects on non-target organisms (Bachman et al., 2020; Chen and De Schutter, 2024). Transgenic expression of dsRNAs or siRNAs has been used for engineering plant resistance to various pathogens (Rosa et al., 2018; Qi et al., 2024); however, the application of transgenic approaches in crop protection is subject to regulatory limitations on the use of genetically modified organisms (GMOs). As an alternative, exogenous application of dsRNAs, which are biocompatible and biodegradable molecules, has emerged as a viable method in disease management (Das and Sherif, 2020; Vatanparast et al., 2024). Topical application of dsRNA has been proved to be effective against various viruses (Mitter et al., 2017b; Niehl et al., 2018; Dubrovina and Kiselev, 2019; Voloudakis et al., 2022), fungi (Koch et al., 2016; Šečić and Kogel, 2021; Zheng et al., 2025), and insects (Huvenne and Smagghe, 2010), demonstrating its enormous potential in plant protection.

The treatment of plants with dsRNA triggers host resistance through two distinct mechanisms. The first and best-characterized is RNA silencing, also known as RNAi, which leads to sequence-specific antiviral defense. In this process, the dsRNA is cleaved by DICER-LIKE proteins, into 21- to 24-nucleotide siRNA duplexes (Bouché et al., 2006; Deleris et al., 2006). Following cleavage, one siRNA strand of the duplex is integrated into ARGONAUTE (AGO) proteins to form the RNA-induced silencing complex (RISC), which targets the viral RNA. This complex is stabilized by SUPPRESSOR OF GENE SILENCING3 (SGS3), which protects it from degradation (Yoshikawa et al., 2013). RNA-DEPENDENT RNA POLYMERASE 1 (RDR1) and RDR6 are then recruited to the viral RNA cleavage products, which are used as templates for new dsRNA synthesis, thereby enhancing antiviral siRNA synthesis and further reinforcing the silencing process (Mourrain et al., 2000; Wang et al., 2010; Yoshikawa et al., 2021; Lopez-Gomollon and Baulcombe, 2022). In tomato, rdr6 mutants are hypersusceptible to tomato brown rugose fruit virus (ToBRFV; Vaisman et al., 2022). In addition, RDR6 requires SGS3, which stabilizes the RDR6 template RNA to enable the amplification process (Mourrain et al., 2000; Yoshikawa et al., 2021). However, the specific functions of both proteins upon exogenous application of dsRNA have not yet been elucidated.

The second mode of dsRNA action is via the induction of pattern-triggered immunity (PTI), a general plant defense response against pathogens (Kørner et al., 2013; Niehl et al., 2016; Yu et al., 2017; Niehl and Heinlein, 2019; Amari and Niehl, 2020). PTI involves the recognition of pathogen-associated molecular patterns (PAMPs) and the activation of downstream signaling cascades, which lead to the expression of host defense genes and the production of antimicrobial compounds (DeFalco and Zipfel, 2021). Unlike RNAi, PTI is based on protein phosphorylation signals and not tailored to a specific target. Application of dsRNA or its synthetic analog polyinosinic–polycytidylic acid [(poly(I:C)] activates PTI responses, resulting in activation of various defense pathways including increased callose deposition at plasmodesmata (PD), which restricts viral cell-to-cell movement and increases the plant’s resistance to viruses (Niehl et al., 2016; Huang et al., 2023). These processes are dependent on PTI regulators such as SOMATIC EMBRYOGENESIS RECEPTOR-LIKE KINASE 1 (SERK1) and BOTRYTIS INDUCED KINASE1 (BIK1), as well as various PD regulators. However, the conservation of dsRNA-mediated PTI responses in crop plants and the respective contributions of RNAi and PTI to target-specific dsRNA-mediated defense are largely unknown (Niehl and Heinlein, 2019).

Viruses are among the most significant groups of plant pathogens, causing extensive agricultural and horticultural losses. In recent years, dispersal of plant viruses has dramatically increased, most likely due to the rise in international commerce of fresh produce and seeds (Jones and Naidu, 2019). In addition, the emergence of new viruses and resistance-breaking isolates poses a severe threat to food security (Jones, 2021). Plant viruses of the Tobamovirus genus (family Virgaviridae), which include the tobacco mosaic virus (TMV) and tomato mosaic virus (ToMV), cause severe damage to global agriculture (Ishibashi et al., 2023; Spiegelman and Dinesh-Kumar, 2023). Tobamoviruses are single-stranded sense-RNA viruses, which are characterized by rigid, stable rod-shaped particles that enclose the viral genome. This genome contains several ORFs encoding the following proteins: two subunits of the viral replicase that are distinguished by length due to a read-through stop codon (Ishibashi and Ishikawa, 2016); the viral movement protein (MP) that is essential for cell-to-cell transport of the virus (Lucas, 2006; Heinlein, 2015); and the coat protein (CP), which polymerizes to form the viral particle, and is also required for long-distance movement of the virus via the phloem (Venturuzzi et al., 2021).

Tomato brown rugose fruit virus (ToBRFV) has been recognized as a tobamovirus that poses an increasingly serious threat to the global tomato industry (Salem et al., 2023). One of the main reasons for the ongoing ToBRFV pandemic is that it has overcome all known tobamovirus resistance genes in tomato, including the durable Tm-2² gene, which has protected tomato cultivars for almost 60 years (Luria et al., 2017). This has been due to the unique ToBRFV movement protein, which enables the virus to evade Tm-2²-derived recognition and immunity (Hak and Spiegelman, 2021; Hak et al., 2023).

Several approaches have recently been developed for ToBRFV resistance, including the identification of natural resistance genes (Zinger et al., 2021, 2025) and CRISPR/Cas9-mediated mutagenesis of host susceptibility factors (Ishikawa et al., 2022; Kravchik et al., 2022). However, many commercial tomato cultivars are still devoid of ToBRFV resistance. In addition, the recent emergence of resistance-breaking ToBRFV isolates (Zisi et al., 2024) prompts the need to develop new approaches to control this devastating virus.

Application of tobamovirus-specific dsRNA has been shown to reduce symptoms and RNA levels of various tobamoviruses in plants (Konakalla et al., 2016; Mitter et al., 2017a; Niehl et al., 2018; Rego-Machado et al., 2020), providing a potentially useful strategy to reduce ToBRFV infections. However, the broad application of dsRNA treatments is hindered by the lack of efficient and economical methods for dsRNA design, and large-scale production and purification. Bio-production of antiviral dsRNA molecules has been achieved by utilizing components of the bacteriophage phi6 for the in vivo production of dsRNA in Pseudomonas syringae (Niehl et al., 2018). This system, based on the phi6 dsRNA replication mechanism, enables the production of long dsRNA molecules targeting TMV. By replacing segments of the phi6 genome with TMV sequences and introducing these constructs into P. syringae, stable production of long, fully duplexed, high-quality (hq)dsRNA against TMV is achieved. The resulting hq-dsRNA molecules are effective by against TMV by topical application in Nicotiana benthamiana plants.

In this study, we utilized the in vivo hq-dsRNA synthesis system to confer ToBRFV protection in tomato plants. Segments of the ToBRFV genome were used to produce hq-dsRNA in P. syringae bacteria. These hq-dsRNAs were effective against ToBRFV, leading to a decrease in both viral symptoms and RNA levels. In addition to specific ToBRFV protection, dsRNA application induced mild activation of multiple plant immune responses, including callose accumulation at PD. However, the non-specific immune response did not significantly protect plants against the closely related virus ToMV. Our results provide evidence for the potential of in vivo produced hq-dsRNA in protecting crops against plant pathogens.

Materials and methods

Plant materials and growth conditions

Seeds of tomato (Solanum lycopersicum) cv. Moneymaker (LA2706) were obtained from the Tomato Genome Research Center (https://tgrc.ucdavis.edu/). Seedsof cv. M82^−sp/−sp, and the rdr6-1 (wiry-1; Solyc04g014870; LA0274) and sgs3 (wiry4-3; Solyc04g02530; n5767) mutants (Yifhar et al., 2012) were a kind gift from Prof. Yuval Eshed, Weizmann Institute of Science, Israel. Seeds were germinated in a controlled growth room under a 16/8 h light/dark photoperiod at 25 °C for 1 week. Following germination, seedlings were transferred to 12 cm pots filled with commercial soil and grown in a semi-controlled greenhouse under natural daylight at 25 °C.

Virus purification

ToMV infectious RNA was generated by in vitro transcription using the pTLW3 plasmid containing a complementary (c)DNA clone of the ToMV genome (Hamamoto et al., 1993). The vector was linearized using the SmaI restriction enzyme (Thermo Scientific™) and then purified using phenol:chloroform:isoamyl alcohol (25:24:1) extraction. The linearized plasmid served as the template for in vitro transcription using the T7 RiboMAX™ Express Large Scale RNA Production System (Promega). Finally, the infectious in vitro-transcribed RNA was mechanically inoculated on Nicotiana benthamiana plants using celite. ToBRFV was isolated from tomato plants infected with the Israeli ToBRFV isolate (accession no. KX619418.1).

ToBRFV and ToMV virions were purified following the protocol described by Niehl et al. (2012) with several modifications. Nicotiana benthamiana plants were inoculated with sap from ToBRFV- or ToMV-infected tomato or N. benthamiana leaves, respectively. At 4 days post-inoculation (dpi), inoculated and symptomatic leaves were ground into a fine powder in liquid nitrogen. Sodium-phosphate buffer (0.5 M NaP, pH 7.0, containing 0.1% 2-mercaptoethanol) was added at a ratio of 1 ml buffer per gram of ground powder and thoroughly mixed. Virions were extracted using an equal volume of butanol/chloroform (1:1), and the phases were separated by centrifugation at 13 000 g for 12 min. The upper aqueous phase containing the virions was precipitated with 40% (w/v) polyethylene glycol (PEG) 8000, followed by centrifugation at maximum speed (17 000 g) for 10 min. The pellet was resuspended in 10 mM sodium-phosphate buffer, and the solution was centrifuged again at 5000g for 10 min. The resulting supernatant was further precipitated with 4% PEG 8000 and 1% (w/v) NaCl, centrifuged at 13 000g for 12 min, and resuspended in 10 mM sodium-phosphate buffer. Virion concentration was determined by measuring absorbance at 260 nm (OD₂₆₀=3 corresponds to 1 mg ml^–1). Purified virions were stored at −20 °C.

In vitro production of ToBRFV and ToMV dsRNA

Total RNA from ToBRFV-infected leaves was used to produce viral cDNA by reverse-transcription (RT). Three fragments of the ToBRFV genome were subsequently amplified using PCR to produce templates for the in vitro synthesis of three dsRNA molecules. For ToMV-specific dsRNA production, a linear DNA template of 416 bp (nt 2729–3145) with sequence identity to the ToMV replicase was PCR-amplified from the pTLW3 infectious clone using primers with the T7 promoter sequence attached to the 5´-end (Yamanaka et al., 2002; see Supplementary Table S1). The DNA templates for dsRNA synthesis were purified using Nucleospin PCR Gel and PCR cleanup columns (Macherey-Nagel). In vitro dsRNA synthesis was performed using a MEGAscript™ RNAi Kit (ThermoFisher, cat. no. AM1626), according to the manufacturer’s protocol. The dsRNA preparations were stored at −20 °C until use.

In vivo production of ToBRFV dsRNA

For the in vivo production of ToBRFV-specific dsRNA (dsRNA^ToBRFV), ToBRFV cDNA sequences of 2550 bp (ToBRFV nt 2906–5455) and 3535 bp (ToBRFV nt 1921–5455) were cloned into the pLD18 and pMS2 plasmids, which harbor cDNA copies of the phi6 S and M segment specific 5´- and 3´-untranslated regions, respectively, and a multiple cloning site in between (Niehl et al., 2018). Amplification of ToBRFV fragments for cloning was performed using specific primers (Supplementary Table S1) and High-Fidelity Phusion DNA Polymerase (Thermo Fisher, F530S), using the cDNA of ToBRFV-IL genome as the template. Vectors were linearized either by generating blunt ends with SmaI and EcoRV restriction enzymes (FastDigest, Thermo Fisher) or by generating vector-fragment overlapping regions via PCR amplification (Phusion, Thermo Fisher, F530S) using specific primers for pLD18 and pMS2 (Supplementary Table S1). Both the ToBRFV fragments and linearized vectors were purified from agarose gels using a NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel, #740609.50). Ligation was carried out using a NEBuilder^® HiFi DNA Assembly kit (NEB, #E5520) with a vector:insert molar ratio of 1:4, followed by transformation into NEB^® 5-α competent Escherichia coli cells. The resulting plasmids pLD18-ToBRFV and pMS2-ToBRFV were validated by PCR amplification of the resulting chimeric constructs using M13 forward and reverse primers, followed by DNA sequencing.

To establish the in vivo dsRNA^ToBRFV production, electroporation-competent cells of Pseudomonas syringae LM2691 were transformed with pLD18-ToBRFV and pMS2-ToBRFV plasmids together with plasmid pLM991 (Sun et al., 2004), which contains the cDNA copy of the phi6 L segment that encodes the phi6 polymerase complex together with a kanamycin-resistance gene insert. After electroporation, the cells were recovered by incubation in SOC solution [2% (w/v) Bacto tryptone, 0.5% (w/v) Bacto yeast extract, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl₂, 10 mM MgSO₄, 20 mM glucose] at 28 °C for 2 h, and the transformed cells were selected by plating them on Luria plates containing kanamycin. The plates were incubated at 28 °C for 2 d, and bacterial clones producing the correct pattern of dsRNA molecules were selected for the large-scale production of dsRNA^ToBRFV. A bacterial strain producing phi6-specific dsRNA (Niehl et al., 2018) was grown for the extraction of control dsRNA^phi6. dsRNA was extracted from concentrated suspensions of liquid cultures grown overnight using TRIzol-chloroform extraction, stepwise LiCl precipitation, followed by ammonium acetate precipitation (Niehl et al., 2018), and the resulting dsRNA pellet was dissolved in milliQ-purified water. The concentration of the purified dsRNA was measured using a NanoDrop 2000 UV-Vis spectrophotometer (ThermoFisher Scientific). Samples were stored at −20 °C prior to application.

Virus inoculation and dsRNA application

Tomato seedlings at 2–3 weeks old with 2–3 true leaves were used for virus inoculation. The cotyledons of these plants were gently rubbed with 40 ng of ToBRFV virions (20 ng for each cotyledon). The hq-dsRNAs were applied at two time-points: on the day of virus inoculation (day 0), and 10 dpi. For hq-dsRNA co-inoculations, dsRNA^phi6 or dsRNA^ToBRFV was added (3.5 µg per cotyledon) and the solution was adjusted to a volume of 5 µl per cotyledon using double-distilled water (ddH₂O) supplemented with 5 mg ml^–1 of carborundum powder (Fisher Scientific, #C309353). The first treatment, containing 7 µg of dsRNA, was applied to both cotyledons, while the second treatment, consisting of 3.5 µg of dsRNA (without added virus), was applied to a single young leaf. Healthy control plants were treated with 5 µl of carborundum solution as a mock control.

Inoculation of tomato plants with ToMV was performed as follows: a mixture containing 10 µl of in vitro-produced purified ToMV virions (1 ng µl^–1) and 10 µl of water or in vitro-produced dsRNA^ToMV (200 ng µl^–1) was distributed equally on both cotyledons. Control plants were treated with 20 µl of milliQ water. The inoculation was performed by gently rubbing the cotyledons in the presence of celite, and after 30 min they were rinsed with distilled water. The plants were returned to the greenhouse and samples from systemic tissues were collected at 13 dpi for RT-qPCR analysis (8–9 biological replicates).

For quantification of the expression of SlPR1a, cotyledons of 4-week-old plants were rubbed with a solution containing 7 µg of dsRNA (3.5 µg per cotyledon), and 24 h later systemic leaves (third leaf from the bottom) were harvested for RT-qPCR analysis.

Evaluation of viral symptoms

The severity of symptoms in ToBRFV-infected plants was assessed according to the protocol outlined by González-Concha et al. (2023). Briefly, systemic leaves at a similar developmental stage (third or fourth leaf down from the apex) were collected at 14 dpi and 21 dpi from virus-, mock-, and virus+dsRNA co-treated plants. Symptoms were scored according to González-Concha et al. (2023) on a scale of 0–5, as follows: 0, no visible symptoms; 1, light mosaic pattern; 2, severe mosaic pattern; 3, severe mosaic pattern and mild distortion; 4, severe mosaic pattern and severe distortion; and 5, severe mosaic pattern, severe distortion, and reduction in size of the leaf blade.

RT-qPCR analysis

For quantification of ToMV, ToBRFV, and SlPR1a in the ToBRFV dsRNA experiments, relative viral RNA levels were determined by reverse-transcription quantitative PCR (RT-qPCR). Total RNA from 50–100 mg samples of cotyledons (4 dpi and 7 dpi) or systemic leaves (7, 14, and 21 dpi) of inoculated and mock-treated plants was extracted using a Plant Total RNA Mini Kit (Geneaid, RPD300) and on-column treated with DNAse, according to the manufacturer’s protocol. Total RNA (500 ng) from each sample served as a template for cDNA synthesis using an AzuraQuant II cDNA Synthesis Kit (Azura Genomics). Reactions without the template or reverse-transcriptase were included as controls. qPCR was performed with a QuantStudio Real Time PCR System (Applied Biosystems) using 2× GreenFast qPCR blue mix LR (Azura Genomics, AZ2305) with the ToBRFV primers published previously (Hak and Spiegelman, 2021; Vaisman et al., 2023; described in Supplementary Table S1). Viral RNA levels were calculated relative to the level of transcripts of the reference gene, TIP41-interacting protein (SlTIP41; SGN-U584254) (Lacerda et al., 2015; Supplementary Table S1). Quantification of SlPR1a (Solyc01g106620) was done using specific primers and SlACTIN (Solyc11g005330) as the reference (Supplementary Table S1).

For ToMV quantification in the ToMV dsRNA experiments, absolute amounts of the viral RNA transcript were determined by RT-qPCR. Three leaf discs per plant (0.5 cm diamter) were excised from systemic leaves at 13 dpi and total RNA was extracted using TRIzol™ reagent (Invitrogen) according to the manufacturer’s instructions. All RNA samples were diluted to a concentration of 10 ng µl^–1 and 5 µl of RNA (50 ng) was used for the amplification reaction using a qScript^® XLT One-Step RT-PCR Kit (Quantabio) and a LightCycler480 (Roche) according to the following protocol: cDNA synthesis at 50 °C for 10 min, followed by 40 cycles consisting of 95 °C for 1 min, 95 °C for 10 s, and 60 °C for 30 s. The primers and probes used are listed in Supplementary Table S1. Finally, the absolute quantification of viral RNA transcripts (copies ng^–1 of RNA) was determined using a standard curve generated from serial dilutions of a 70 nt transcript derived from the pTLW3 plasmid containing the ToMV infectious cDNA.

Callose quantification at plasmodesmata

Leaflets of 2-month-old cv. Moneymaker plants were infiltrated with either 3.5 µg of dsRNA^phi6 or dsRNA^ToBRFV, or ddH₂O as a mock control. After 6 h, the leaflets were infiltrated with Aniline Blue (Fisher Chemical) diluted 1:1000 in MES buffer (10 mM MgCl₂, 10 mM MES, pH 5.6) (Hak et al., 2023). Approximately 5 min after stain infiltration, two sections from each leaflet were imaged using an Olympus IX 81 inverted laser-scanning confocal microscope (UV laser, excitation 405 nm, emission 475–525 nm). Ten images were captured per leaf, with one leaf from each of six individual plants being used.

For callose deposition tests with poly(I:C) treatment, leaf discs from cv. Moneymaker plants were washed, vacuum-infiltrated with 200 µl of 0.1% Aniline Blue solution (pH 9) containing 500 ng µl^–1 of poly(I:C) (Sigma-Aldrich) and incubated in the dark for 30 min. Confocal images were acquired using a Zeiss LSM 780 confocal laser-scanning microscope with a 405 nm diode laser for excitation and a 40×/1.3 N.A. Plan Neofluar objective with oil immersion.

Callose fluorescence intensity was quantified using Method A in the ‘calloseQuant’ plugin in ImageJ, as described by Huang et al. (2023). Briefly, the method identifies spots of maximal fluorescence as regions of interest (ROI) and measures the fluorescence intensity within each one. Before measurement, the expected location of the ROIs at the cell wall was verified. For all conditions, the same ROI size was applied around the peak fluorescence intensity to ensure consistent measurements. Three biological replicates were used per treatment.

Determination of reactive oxygen species

Reactive oxygen species (ROS) were measured as previously described (Leibman-Markus et al., 2017). The fourth leaf from the bottom of 4–5-week-old cv. Moneymaker plants was harvested and subjected to petiole-feeding with 1 µg ml^–1 of dsRNA overnight. Leaf discs of 0.5 cm diameter were then collected, washed in distilled water (dH₂O), placed in a white 96-well plate (SPL Life Sciences, Korea), and 50μl of dH₂O was immediately added into each well to prevent the tissue from drying out. The final ROS measurement reaction contained 150μM luminol and 15μg ml^–1 horseradish peroxidase either with or without 1μM flg₂₂ (PhytoTechLabs #P6622). Light emission over time was measured using a microplate luminometer (GloMax^® Discover, Promega). The measurements were performed on at least eight leaf discs from five different plants per treatment.

Ethylene production

Ethylene biosynthesis was measured as previously described (Leibman-Markus et al., 2017). The fourth leaf from the bottom from 4–5-week-old tomato cv. Moneymaker plants was harvested and subjected to petiole-feeding with 1 µg ml^–1 of dsRNA overnight. Leaf discs (0.9 cm in diameter) were then taken and washed in dH₂O. For each biological repeat in each treatment, six discs were sealed in a 10 ml bottle containing 1 ml of assay medium with 1μg ml^–1 of the relevant dsRNA either with or without 1μg ml^–1 of the fungal elicitor ethylene-induced xylanase (EIX), and incubated overnight, with vertical agitation (100 rpm). Ethylene production was measured using an Agilent Intuvo 9000 GC system. The measurements were performed in three independent experiments. Each experiment was based upon six biological replicates per treatment, each replicate being composed of six leaf discs that were randomly collected from five different plants per treatment (a total of 15 plants per experiment).

Ion leakage assays

Ion leakage (conductivity) measurement was performed according to Leibman-Markus et al. (2017). The fourth leaf from the bottom from 4–5-week-old cv. Moneymaker plants was harvested and subjected to petiole-feeding with 1 µg ml^–1 of dsRNA overnight. Leaf discs (0.9 cm diameter) were then collected and washed in dH₂O. For each sample, six discs were placed in a 10 ml flask with 1 ml of distilled water for 48 h, with vertical agitation (100 rpm). Then, 1.5 ml of dH₂O was added to each sample, and conductivity was measured using a conductivity meter (AZ^® Multiparameter pH/Mv/Cond./Temp Meter 86505). The measurements were performed in two independent experiments. Each experiment was based upon five biological replicates per treatment, each replicate being composed of six leaf discs that were randomly collected from five different plants per treatment (a total of 15 plants per experiment).

Botrytis cinerea pathogenicity assays

To determine non-specific dsRNA-mediated pathogen defense, Botrytis cinerea pathogenicity assays were conducted as previously described (Gupta et al., 2020). The B. cinerea isolate Bc5-10 was maintained on potato dextrose agar (PDA; Difco) plates at 22 °C, and subcultured once a week by inoculating a single 0.5 cm mycelial plug at the center of a new plate. The fourth leaf from the bottom from 4–5-week-old cv. Moneymaker plants was harvested and subjected to petiole-feeding with 1 µg ml^–1 dsRNA overnight. Agar discs measuring 0.4 cm in diameter were pierced from the fungal colony margins of 3-day-old plates, and used to inoculate detached leaves (two fungal discs per leaflet). Following inoculation, the leaves were placed in a humid chamber at 22 °C under 16/8 h conditions (200 μmol m^–2 s^–1). At 3 dpi, necrotic lesions were imaged and their size was measured using the ImageJ software, with the pixels mm^–1 scale being set using a standard ruler.

Statistical analysis

For dsRNA-mediated protection assays, Student’s t-test was performed to compare each treatment with non-dsRNA-treated, virus-infected plants of the same genotype. Differences among three or more treatments were examined using Welch’s one-way ANOVA for groups with unequal variances, followed by Dunnett’s post-hoc test. All statistical analyses were conducted using Prism9™. Schematic diagrams were created in BioRender.

Results

Validation of target sequences for dsRNA-mediated control of ToBRFV infection

To examine their potential in controlling ToBRFV infection, three dsRNAs targeting different fragments of the ToBRFV genome were synthesized in vitro: dsRNA 1 targeted the replicase ORF (nt 3295–3693); dsRNA 2 targeted a region between the replicase and MP (nt 4760–5133); and dsRNA 3 targeted the MP and CP (nt 5529–5928) (Fig. 1A). The three dsRNAs were separately co-inoculated with ToBRFV virions on 2-week-old tomato plants and symptoms were observed at 21 dpi. Of the three, dsRNA 1 showed reduced viral symptoms, and a reduction in ToBRFV RNA levels was observed in the dsRNA 1-treated plants (Fig. 1B, C; Supplementary Fig. S1). Similarly, targeting a parallel region in the related ToMV with ToMV-specific dsRNA also protected tomato plants against this virus (Supplementary Fig. S2). These results suggested that dsRNA that shares sequence identity with the replicase sequence could be used to control tobamovirus infections, and that this region provides an attractive target for in vivo-synthesized dsRNA.

Fig. 1.

Application of in vitro-produced ToBRFV-specific dsRNA confers protection against ToBRFV in tomato. (A) Schematic illustration of the ToBRFV genome, with the areas selected for in vitro and in vivo dsRNA production. In vitro-produced dsRNA 1 (nt 3295–3693), dsRNA 2 (nt 4760–5133), and dsRNA 3 (nt 5529–5928) targeted three different regions in the viral genome. In vivo dsRNA^ToBRFV was generated by inserts of 3535-bp and 2550-bp cloned for production of phi6M and S segments, respectively. (B) Representative images of leaves of a healthy, uninfected tomato cv. Moneymaker plants, a ToBRFV-inoculated plant, and a plant co-inoculated with ToBRFV and dsRNA 1 at 14 days post inoculation. Scale bar is 5 cm. (C) Levels of ToBRFV RNA in the leaves shown in (B) as determined by qRT-PCR. Levels are relative to transcripts of TIP41, which served as a reference gene. Significant differences between means were determined using Student’s t-test: *P<0.05; ***P<0.001; ns, not significant (n=7). Spiegelman, Z. (2025) https://BioRender.com/bhgmnfa.

In vivo production of ToBRFV-specific dsRNA

To enable scalable and more efficient production of ToBRFV-specific dsRNA (dsRNA^ToBRFV), we established a biosynthesis system in P. syringae LM2691 using the in vivo hq-dsRNA production platform based on dsRNA bacteriophage phi6. The L-segment (6374 bp) of the phi6 tripartite genome encodes the components of the phi6 polymerase complex that is essential for the amplification of heterologous dsRNA molecules in bacteria. Previously, TMV-specific dsRNA production in vivo has been successfully established by replacing the coding sequences of the phi6M (4063 bp) and S (2948 bp) segments by fragments of the TMV genome covering parts of replicase and MP sequences (Niehl et al., 2018). The dsRNA produced using this platform is of high quality (hq) because the phi6 RNA-dependent RNA polymerase produces fully duplexed dsRNA. To produce hq-dsRNA against ToBRFV, 3535 nt and 2250 nt fragments of the ToBRFV genome sequence (nt 1921–5455 and nt 2906–5455), both covering parts of the replicase and MP ORFs, were cloned between the 5´-packaging and 3´-replication signals of phi6M and S segment, respectively (Fig. 1A). The selected sequences encompassed similar parts of the ToBRFV genome as the in vitro-produced ToBRFV dsRNA 1, and were homologous to the TMV sequences of hq-dsRNA that have been shown to be effective against TMV (Niehl et al., 2018). Pseudomonas syringae cells were transformed with the phi6 constructs harboring the ToBRFV sequences, and dsRNA was extracted and purified from the bacterial cells (Fig. 2). Integration of the fragments resulted in the production of ToBRFV hq-dsRNA molecules that were 3207 bp (S_ToBRFV) and 4218 bp (M_ToBRFV) in length, together with the phi6 L_kan segment containing the kanamycin resistance gene insert (7599 bp) (Supplementary Fig. S3). The highly purified hq-dsRNA^ToBRFV, which consisted of bacterial dsRNA extract including all three segments, was applied on tomato plants and tested for its ability to confer protection against ToBRFV (Fig. 2).

Fig. 2.

Schematic diagram of the production of ToBRFV-specific dsRNA in vivo. The pLD18-ToBRFV and pMS2-ToBRFV plasmids harboring the ToBRFV sequences S and M, respectively (Fig. 1), were transformed into Pseudomonas syringae LM2691 together with the phi6 L segment cDNA that encodes the components of the phi6 dsRNA replication machinery. The liquid bacterial cultures were grown overnight, followed by high-quality dsRNA extraction and purification. The resulting dsRNA^ToBRFV was applied on tomato plants to test for protection against ToBRFV. Spiegelman, Z. (2025) https://BioRender.com/6ilxl3h.

Application of in vivo-produced dsRNA protects tomato plants against ToBRFV

To test the ability of in vivo-synthesized dsRNA^ToBRFV to confer protection in tomato plants, 7 µg of hq-dsRNA were co-inoculated with 40 ng of ToBRFV virions on the cotyledons of 2-week-old plants of cv. Moneymaker. As controls, plants were inoculated either with a mock solution, with a solution that contained ToBRFV virions alone (ToBRFV), or with ToBRFV together with non-specific phi6 hq-dsRNA (ToBRFV+dsRNA^phi6). Plants were treated again with 3.5 µg dsRNA at 10 dpi. At 14 dpi, the symptoms of ToBRFV infection were evident, but plants treated with either dsRNA^ToBRFV or dsRNA^phi6 were significantly less symptomatic (Fig. 3A, B). At 21 dpi, severe ToBRFV symptoms such as plant stunting and leaf malformations were observed in plants inoculated with ToBRFV alone or with ToBRFV together with dsRNA^phi6, but co-treatment of plants with ToBRFV and dsRNA^ToBRFV reduced viral symptoms significantly (Fig. 3C, D). Consistent with the viral symptom data, RT-qPCR analysis showed that at the early stages of 4 dpi and 7 dpi, ToBRFV RNA levels were significantly reduced in both the dsRNA^ToBRFV- and dsRNA^phi6- treated plants (Fig. 3E); however, at 14 dpi the ToBRFV levels remained low only in the plants treated with dsRNA^ToBRFV.

Fig. 3.

In vivo-produced dsRNA^ToBRFV confers protection against ToBRFV infection in tomato plants. (A) Representative images of shoots and the third leaf down from the apex at 14 days post inoculation (dpi) for mock-inoculated control plants, plants inoculated with ToBRFV alone, and plants inoculated with ToBRFV together with either in vivo-produced dsRNA^ToBRFV (ToBRFV+dsRNA^ToBRFV) or dsRNA^phi6 (ToBRFV+dsRNA^phi6). (B) ToBRFV disease score for the leaves on a scale of 0–5 at 14 dpi. Data are means (±SE), n≥30. (C) Representative images of shoots and the third and fourth leaves down from the apex at 21 dpi, and (D) the overall disease scores for the leaves. Data are means (±SE), n≥14. (E) Levels of ToBRFV RNA in inoculated leaves at 4 dpi and in systemic leaves at 7, 14, and 21 dpi as determined by RT-qPCR analysis. Levels are relative to transcripts of TIP41, which served as a reference gene, and are normalized to the plants inoculated with ToBRFV only, the values of which were set as 1. Data are means (±SE, shading), n≥6. Scale bars are 5 cm. Significant differences compared with the ToBRFV-only treatment were determined using Student’s t-test: *P<0.05, **P<0.01, ***P<0.001.

To corroborate these results, a parallel experiment was performed using a different tomato cultivar, M82. Here again, protection at 21 dpi was only conferred by dsRNA^ToBRFV treatment (Supplementary Fig. S4A–D). However, a transient decrease in viral levels at 7 dpi was observed for both dsRNA^ToBRFV and dsRNA^phi6 treatments (Supplementary Fig. S4E). Similar to the results for cv. Moneymaker, reduced viral RNA levels were observed at 14 dpi only in cv. M82 plants treated with dsRNA^ToBRFV (Supplementary Fig. S4F). These results suggested that in vivo produced dsRNA^ToBRFV is an effective tool for controlling ToBRFV levels and symptoms. In addition, during the first days of infection there was a transient, non-sequence-specific dsRNA-mediated effect that delayed the progression of viral infection.

Further experiments were performed to investigate the robustness and specificity of in vivo-produced hq-dsRNAs. First, a time-course experiment was conducted to examine the durability of hq-dsRNA-mediated defense against ToBRFV. For this, the plants were treated with dsRNA^ToBRFV up to 2 d before or after ToBRFV inoculation. A protective effect of dsRNA^ToBRFV was observed when applied either 1 d before (−1 dpi) or on the day of virus inoculation (0 dpi), suggesting limited hq-dsRNA stability and/or activity under greenhouse conditions (Supplementary Fig. S5). To investigate the specificity of in vivo-produced hq-dsRNA, the protective effects of dsRNA^ToBRFV and dsRNA^phi6 were examined in plants inoculated with ToMV (Supplementary Fig. S6). No significant protective effect was observed at 14 dpi or 21 dpi, suggesting that dsRNA^ToBRFV-mediated defense might be specific to ToBRFV.

RNA-DEPENDENT POLYMERASE 6 and SUPPRESSOR OF GENE SILENCING 3 are not essential for dsRNA-mediated defense against ToBRFV

The sequence-specific effect of dsRNA^ToBRFV that was seen at 14 dpi probably resulted from the activation of the antiviral RNA-silencing pathway. To test whether RDR6 and SGS3 are required for dsRNA-mediated protection against ToBRFV in tomato, the mutants rdr6 (LA0274) and sgs3 (n5767) were inoculated with ToBRFV, or co-inoculated with ToBRFV together with either dsRNA^ToBRFV or dsRNA^phi6. Wild type cv. M82 plants served as the control. As expected, both rdr6 and sgs3 plants were more susceptible to ToBRFV when inoculated with the virus only, showing stronger symptoms and higher viral RNA levels as compared to the controls (Fig. 4); however, the presence of dsRNA^ToBRFV conferred protection in both the rdr6 and sgs3 mutants. These results suggested that while both RDR6 and SGS3 are required for the natural plant antiviral defense, they are not essential for dsRNA-mediated ToBRFV resistance in tomato.

Fig. 4.

RDR6 and SGS3 are not essential for dsRNA-mediated ToBRFV resistance in tomato. (A) Representative images of plants of cv. M82 wild type and the rdr6 and sgs3 mutants that were either uninfected (healthy), inoculated with ToBRFV only, or inoculated with ToBRFV and co-treated either with in vivo-produced dsRNA^ToBRFV or with dsRNA^phi6. (B) Levels of ToBRFV RNA in systemic leaves of plants of all the different genotypes and treatments at 14 days post inoculation, as determined by RT-qPCR analysis. Levels are relative to transcripts of TIP41, which served as a reference gene. Data are means (±SE), n≥6. Significant differences between pairs of means were determined using Students t-test: *P<0.05, **P<0.01.

Activation of plant innate immunity by in vivo-produced dsRNA

It has been shown that dsRNA functions as a PAMP to activate plant immune responses, regardless of its sequence identity (Niehl et al., 2016; Huang et al., 2023), so therefore it is possible that the primary dsRNA-induced non-sequence-specific response against ToBRFV at 4 dpi and 7 dpi (Fig. 3E) involved activation of plant innate immunity. To investigate this, activation of the immune system was assayed in tomato leaves treated with either a mock solution (ddH₂O) or ddH₂O containing dsRNA^ToBRFV or dsRNA^phi6.

In plants, PTI activation results in a signaling cascade that triggers several responses, including ROS production, ethylene biosynthesis, ion leakage, and the induction of pathogen-related genes. These responses, which are typically triggered within minutes to hours after recognition of PAMPs, serve as hallmarks for PTI activation. No increase in basal levels of ROS were observed in leaves fed with hq-dsRNA alone (Fig. 5A), which was consistent with a previous study (Huang et al., 2023). However, when supplemented with the bacterial elicitor flg₂₂, a 22–44% increase in ROS production was observed in leaves treated with dsRNA^ToBRFV or dsRNA^phi6. Moreover, a small but significant increase of 30% in ethylene production was observed in dsRNA^phi6-treated leaves (Fig. 5B). Ethylene production was observed to increase 5-fold in leaves exposed to the fungal elicitor EIX, as reported previously (Sharfman et al., 2014; Zaid et al., 2022); interestingly, co-treatment of EIX with dsRNA led to a further 49–54% increase in production.

Fig. 5.

Application of in vivo-produced dsRNA triggers plant innate immunity in tomato. Leaves of cv. Moneymaker were treated via petiole feeding with ddH₂O (mock), or ddH₂O containing either 1 µg ml^–1 of dsRNA^ToBRFV or dsRNA^phi6, and immune responses were quantified at 24 h later. (A) Levels of reactive oxygen species burst, expressed as relative luminescence units (RLU), measured in leaf samples with or without the addition of the bacterial elicitor flg₂₂. RLU is relative to the level in the mock treatment in samples with flg₂₂, the value of which was set as 100%. (B) Ethylene emission measured in leaf samples with or without the addition of the fungal elicitor ethylene-induced xylanase (EIX). The ethylene concentration (ppm) is expressed relative to the mock treatment without EIX, the value of which was set as 100. (C) Ion leakage in the leaf samples. Conductivity was measured (μS) and is expressed relative to the mock treatment, the value of which was set as 100. (D) Representative images of leaves of plants from the mock, dsRNA^ToBRFV, and dsRNA^phi6treatments following inoculation with Botrytis cinerea. After the petiole-feeding, detached leaves were inoculated with the fungus and the images were taken 3 d later. (E) The area of necrotic lesions was quantified 3 d after B. cinerea inoculation. All data are means (±SE). In (A) n≥30; in (B) n≥10; in (C) n≥5; in (E) n≥19. Significant differences compared with the mock treatment were determined using Welch’s ANOVA followed by Dunnett’s post-hoc test: *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001; n.s., not significant.

In addition, a 32–42% increase in ion leakage was observed upon treatment of the leaves with either of the dsRNAs (Fig. 5C), indicating increased immune-mediated programmed cell death. In addition, a 2–3-fold increase in the transcript levels of SlPR1a was observed in systemic leaves after both the dsRNA treatments (Supplementary Fig. S7). These results are consistent with previous observations that dsRNA acts as a PAMP to activate PTI responses (Niehl et al., 2016).

To further assess the non-sequence-specific immune capacity of in vivo-produced dsRNA application, hq-dsRNA- and mock-treated leaves were challenged with the fungal pathogen Botrytis cinerea by mycelial infection, and the disease response was quantified by measuring the gray mold lesion size at 3 dpi. Treatment with dsRNA^ToBRFV and dsRNA^phi6 reduced the lesion size by 68% and 30%, respectively (Fig. 5D, E), suggesting that in addition to their sequence-specific defense, the in vivo-produced dsRNAs also primed or moderately activated plant innate immunity.

A recent study has demonstrated that the dsRNA-mediated PTI response restricts the progression of viral infection in Arabidopsis and N. benthamiana by inducing callose accumulation at PD, thereby limiting the cell-to-cell movement of the virus (Huang et al., 2023). To determine whether in vivo-produced dsRNA also triggers callose accumulation in tomato, Aniline Blue staining was performed on leaves treated with dsRNA^ToBRFV, dsRNA^phi6, or the dsRNA analogue poly(I:C). In agreement with the induced immune responses that were observed, plasmodesmal callose accumulation was increased by 79% and 47% in dsRNA^ToBRFV- and dsRNA^phi6- treated leaves, respectively (Fig. 6). Similarly, increased callose deposition was also observed in leaves treated with poly(I:C) (Supplementary Fig. S8). These results suggested that the non-specific PTI effect of dsRNA in tomato might include callose-mediated restriction of plasmodesmatal transport at the early stages after application.

Fig. 6.

Application of in vivo-produced dsRNA increases callose accumulation at plasmodesmata of tomato leaf epidermal cells. Leaves of cv. Moneymaker were infiltrated with ddH₂O (mock), or ddH₂O containing 3.5 µg ml^–1 of dsRNA^ToBRFV or dsRNA^phi6 and 6 h later they were infiltrated with Aniline Blue. (A) Images of stained callose at plasmodesmata (PD) in epidermal cells in the different treatments. The images are representative of six replicate plants. (B) Quantification of callose at PD, expressed as arbitrary fluorescence units (AFU). Individual data points are shown for n≥30 images per leaf from three replicate plants. Significant differences compared with the mock treatment were determined using Welch’s ANOVA followed by Dunnett’s post-hoc test: ***P<0.001, ****P<0.0001.

Discussion

The ToBRFV pandemic poses a substantial threat to tomato cultivation worldwide (Salem et al., 2023). Whilst recent advances in tomato breeding for ToBRFV resistance have provided effective and promising tools to cope with this disease, there is also a need to develop alternative solutions that can confer protection in current non-resistant elite varieties. Here, we provide evidence for the potential of dsRNA as a tool for the control of ToBRFV infections (Fig. 1). A scalable in vivo production system based on bacteriophage phi6 replication in P. syringae bacteria (Niehl et al., 2018; Levanova and Poranen, 2024) was utilized to produce long, high-quality (hq-)dsRNA molecules targeting ToBRFV (Fig. 2). Our results demonstrated the effectiveness and specificity of in vivo-produced dsRNA^ToBRFV against ToBRFV (Fig. 3; Supplementary Figs S4, S6), highlighting its potential as a tool for combating this virus.

Our preliminary screening of suitable ToBRFV and ToMV target dsRNA sequences (Supplementary Figs S1, S2), combined with previously published studies using in vitro-produced dsRNA and in vivo-produced hq-dsRNA against TMV (Konakalla et al., 2016; Niehl et al., 2018), identified the replicase-encoding region of the viral genomic RNA as an effective target for dsRNA-mediated defense. This effectivity might be explained by the ability of these dsRNAs to specifically target the viral genome, rather than the subgenomic RNAs (as do dsRNA2 and 3), which do not contain replicase sequences (Ishibashi and Ishikawa, 2016). In addition, as replication is a critical step in the viral life cycle, the viral genomic RNA might be more sensitive to silencing than the subgenomic RNAs. Moreover, the tobamovirus replicase functions as the viral suppressor of RNA silencing (VSR), which enables the virus to overcome the host RNA-silencing pathway (Csorba et al., 2007; Vogler et al., 2007). Therefore, by targeting the viral genomic RNA, the expression of the VSR is also directly affected and allows RNA-silencing to be more effective against the virus.

While P. syringae is a known plant pathogen, the dsRNA extraction process involves complete bacterial lysis and hence it is unlikely that its use would result in unexpected negative outcomes on plant health. However, there are still several restrictions that limit dsRNA application under commercial growth conditions. While the inherent instability of dsRNA offers an advantage as a biodegradable solution for control of agricultural pests and diseases, effective protection requires the dsRNA to be resistant to degradation caused by various environmental factors including heat, UV radiation, and microbes (Kalinina et al., 2025). In this regard, a formulation that increases hq-dsRNA stability and effectiveness against fungi has been developed (Moorlach et al., 2025; Zheng et al., 2025). In addition, the timing and the method of dsRNA application play critical roles, influencing its efficacy as a control treatment. Our time-course analysis demonstrated that successful protection against ToBRFV required dsRNA application close to the time of viral inoculation, with co-inoculation being most effective (Supplementary Fig. S5). Accordingly, we co-inoculated dsRNA with virion particles using a method involving mild mechanical disruption to facilitate the simultaneous penetration of both the virus and dsRNA. It is possible that non-simultaneous application of hq-dsRNA and ToBRFV would decrease the chance of different cells being inoculated by the virus and dsRNA, and that the short time-frame of our experiment was not sufficient for the evaluation of any cell-to-cell/systemic spread of the silencing signal that might affect the virus.

Apart from the need to increase its stability, dsRNA for antiviral protection must be able to overcome structural barriers such as the plant’s cuticle and cell wall in order to effectively penetrate into the host cells (Bennett et al., 2020). Conjugation of in vivo-produced dsRNA to various nanoparticles such as layered double-hydroxide nanosheets, carrier peptides, lipid-modified polyethylenimine, or alginate–chitosan might enhance the ability of ToBRFV dsRNA to survive and penetrate the host cell effectively (Numata et al., 2014; Mitter et al., 2017a; Avital et al., 2021; Komarova et al., 2023; Moorlach et al., 2025; Zheng et al., 2025). In addition, applying dsRNA using different techniques such as high-pressure spraying, trunk injection, and petiole feeding might improve its efficacy (Dalakouras et al., 2018; Bennett et al., 2020; Kalinina et al., 2024; Vatanparast et al., 2024) and enable its application to be further optimized in commercial greenhouse and field environments.

Application of dsRNA is known to activate the host RNA-silencing pathway for antiviral protection (Kalinina et al., 2024); however, the specific silencing factors involved are largely unknown. As expected, both the rdr6 and sgs3 tomato mutants were hypersensitive to ToBRFV (Fig. 4); however, interestingly, the application of dsRNA^ToBRFV was effective against the virus in both of them. This could be due to various reasons. One possibility is that the pleiotropic effect caused by the loss of RDR6 or SGS3 might synergistically impact the effects of dsRNA by up-regulating immune signaling pathways. For example, several plant immune receptors have been shown to be regulated by microRNAs (Li et al., 2012; Lopez-Gomollon and Baulcombe, 2022), and loss of this control might enhance plant antiviral immunity. Another possibility is that application of external dsRNA might further promote RNA-silencing by recruiting additional host proteins into the process. For example, while RDR6 seems to be essential for natural tobamovirus resistance in tomato, dsRNA-mediated defense might involve its homologue, RDR6b (Solyc08g075820) (Bai et al., 2012). In addition, this process might also involve additional RDR proteins (i.e. RDR1), thereby producing functional redundancy. Functional redundancy of RDR1, RDR2, and RDR6 in the production of antiviral RNAi has already been observed in Arabidopsis, suggesting that these enzymes act redundantly to restrict systemic infection (Garcia-Ruiz et al., 2010; Wang et al., 2010; Ding, 2023).

In addition to the induction of RNAi, dsRNA can also induce PTI responses (Niehl et al., 2016); however, the relative contributions of RNAi and PTI to defenses triggered by external dsRNA application are unclear. Here, we observed a non-sequence-specific effect of externally applied hq-dsRNA on viral RNA levels and symptoms during the first days of infection, including reduction in ToBRFV RNA levels at 4 dpi and 7 dpi, and a visible decrease in symptoms at 14 dpi (Fig. 3). In addition, hq-dsRNA application resulted in mild activation of innate immune responses (Fig. 5), which resulted in increased callose deposition at PD (Fig. 6). However, at 21 dpi only sequence-specific dsRNA^ToBRFV-mediated protection was observed (Fig. 3). These results suggest that dsRNA-mediated PTI inhibits systemic infection only transiently. This effect can be explained by the cell-autonomous nature of PTI responses, which might transiently inhibit the virus only at the site of inoculation. In contrast, the mobility of antiviral siRNAs and RDR-mediated siRNA amplification can result in a systemic response. These results are consistent with previous findings showing that non-specific dsRNA application results in only minor induction of antiviral immunity (Necira et al., 2024). Nevertheless, PTI might also play a cumulative or synergistic role with RNAi in dsRNA-mediated antiviral protection. It is important to note that while these responses might be specific to dsRNA, it is possible that they could be a general result of exogenous application of nucleic acids such as DNA or single-stranded RNA.

Overall, our study provides evidence for dsRNA-mediated antiviral responses in tomato and highlights the potential of dsRNA to protect against the devastating ToBRFV pandemic. These findings can pave the way for more sustainable approaches for dsRNA-mediated crop protection against plant pathogens.

Supplementary data

The following supplementary data are available at JXB  online.

Table S1. Oligonucleotides used in this work.

Fig. S1. Images of plants used for screening of functional dsRNA targets in the ToBRFV genome.

Fig. S2. Application of dsRNA confers ToMV protection in tomato plants.

Fig. S3. Agarose gel electrophoretic analysis of in vivo-produced ToBRFV dsRNA.

Fig. S4. In vivo-produced dsRNA reduces ToBRFV symptoms and accumulation in cv. M82.

Fig. S5. Impact of dsRNA application time on ToBRFV protection.

Fig. S6. In vivo-produced ToBRFV dsRNA does not protect against ToMV.

Fig. S7. Increased SlPR1a transcript levels in plants treated with dsRNA.

Fig. S8. Application of poly(I:C) increases callose accumulation at plasmodesmata of tomato leaf epidermal cells.

Author contributions

DW, ARS, MB, MMP, MH, and ZS conceptualized and designed the experiments, and wrote the manuscript; DW, ARS, ALL, ML-M, RG, RM, and HH performed the experiments and analyzed the data.

Funding

This work was supported by the ‘Maïmonide-Israel’ research program of the Israeli-French High Council for Scientific & Technological Cooperation (no. 47718YB/0002427) to MH, MB, and ZS, and a grant from the Research Council of Finland (no. 331627) to MMP. This project has also received funding from L’Agence National de la Recherche (ANR, grant no. ANR-21-SUSC-0003-01) to MH and the Ministry of Agriculture and Forestry Finland (grant no. VN/8755/2021) to MMP as part of the project BioProtect co-ordinated by MH and carried out under the second call of the ERA-NET Cofund Action ‘SusCrop’, which is part of the Joint Programming Initiative on Agriculture, Food Security and Climate Change (FACCE-JPI). SusCrop has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 771134.

Data availability

All data supporting the findings of this study are available within the paper and within its supplementary materials published online. Constructs, materials, and protocols used in this study will be made available by the corresponding author, Ziv Spiegelman, upon request.