Exogenous Application of IR-Specific dsRNA Inhibits Infection of Cucumber Green Mottle Mosaic Virus in Watermelon

Abstract

Cucumber green mottle mosaic virus (CGMMV) represents a serious threat in the production of watermelon. Small RNAs facilitate a mechanism known as RNA interference (RNAi), which regulates gene expression. RNAi technology employs foreign double-stranded RNAs (dsRNAs) to target and reduce the expression levels of specific genes in plants by interfering with their mRNAs. In this study, watermelon plants were treated with dsRNAs of CGMMV MET, IR, and HEL fragments that had been generated in E. coli HT115. We investigated variations in several factors, including viral accumulation, virus-derived small interfering RNAs (vsiRNAs), and symptom severity. MET-dsRNA, IR-dsRNA and HEL-dsRNA dramatically decreased the symptoms of CGMMV in plants in the growth chamber test. Plants treated with viral-derived dsRNA showed a considerable decrease in both virus titers and vsiRNA levels. We also explored the mobility of spray-on dsRNA-derived long dsRNA and discovered that it could be identified in both inoculated leaves and the systemic leaves. IR-dsRNA outperformed MET-dsRNA and HEL-dsRNA in dsRNA therapy. Illumina sequencing of small RNAs from watermelon plants treated with IR-dsRNA and those that were not treated showed that the decreased accumulation of vsiRNAs was consistent with interference with CGMMV infection in systemic leaves. dsRNA-treated plants showed a higher level of 24-nt viral siRNA and lower level of 22-nt viral siRNA accumulation, while 22-nt viral siRNA predominated in untreated plants, indicating that dsRNA treatment improved DCL3 activity. In conclusion, our research provides deeper insights into the mechanism of antiviral RNA interference and confirms the effectiveness of applying dsRNA locally to enhance plant antiviral activity.

1. Introduction

Watermelon (Citrullus lanatus (Thunb.) Matsum & Nakai) belongs to the genus Citrullus of the Cucurbitaceae family and is an annual vine herb [1]. Its fruit is sweet and refreshing, rich in nutrients, and is one of the important horticultural crops globally [2]. Cucumber green mottle mosaic virus (CGMMV), a member of the genus Tobamovirus (family Virgaviridae), primarily infects plants in the family Cucurbitaceae, causing symptoms such as mottling, mosaic patterns, leaf blistering, stunted growth, and brown necrosis on stems, leaves, and fruits. CGMMV was initially documented in Cucumis sativus in the United Kingdom in 1935 [3] and has subsequently disseminated globally via the international seed trade [4]. The virus is transmitted primarily via mechanical means and infected seeds [5]. Effective control measures focus on limiting field transmission through management strategies, including the removal of infected plant material and solarization of soil, as well as the use of certified disease-free seeds and healthy planting material.

CGMMV possesses a 6.4 kb single-stranded positive-sense RNA genome containing four open reading frames (ORFs) that encode a 186 K viral replicase, a 129 K replication-related protein, a 29 K movement protein (MP), and a 17.3 K coat protein (CP) [6]. The 186 K replicase is generated through read-through of the 129 K protein’s termination codon, occurring with a 5% probability. As a result, the 186 K protein shares an N-terminal region with the 129 K protein, both of which possess methyltransferase (MET) and RNA helicase (HEL) domains, as well as an Internal Region (IR). The C-terminal extension of the 186 K protein corresponds to the RNA-dependent RNA polymerase (RdRp) domain. Together, these proteins facilitate viral replication and are collectively referred to as replication-related proteins [7,8]. The 129K/186K protein exhibits silencing suppressor activity and binding affinity for short RNA duplexes (SSA) 26K proteins [7]. RNA interference (RNAi) is an inherent regulatory mechanism mediated primarily by two categories of molecules: microRNAs (miRNAs) and small interfering RNAs (siRNAs) [9,10,11]. In this process, exogenous double-stranded RNA (dsRNA) triggers the DICER and RNA-induced silencing complex (RISC) pathways, where dsRNA is cleaved into siRNAs that guide the degradation of complementary target RNA [11,12]. Leveraging this mechanism, direct application of exogenous dsRNA to plants via mechanical or spray inoculation could effectively confer resistance to viral infections. This approach has emerged as a promising plant protection strategy, and potentially transforming future crop protection methods against pathogens [13,14,15].

This study utilizes exogenous homologous CGMMV dsRNA in watermelon to elicit a RNAi response and ultimately establish disease resistance. In comparison to the transgenic antiviral approach, this technology is more straightforward, safer, environmentally sustainable, and comparatively economical [16,17].

2. Materials and Methods

2.1. Materials

The CGMMV-hn isolate was obtained from watermelon leaves and propagated in the sensitive watermelon cultivar Hongheping (Citrullus lanatus) [18].

2.2. Plasmid Construction and Bacterial Induction

Total RNAs were extracted from CGMMV-infected watermelon leaves using the RNA Simple Total RNA Kit (TIANGEN, Beijing, China). cDNA was synthesized using a high-capacity cDNA reverse transcription kit (Vazyme, Nanjing, China). The MET, IR, and HEL fragments were amplified via RT-PCR and cloned into the Sac I and Hind III sites of L4440 plasmid with the NEBuilder HiFi DNA Assembly Cloning Kit (New England Biolabs, Ipswich, MA, USA), generating the constructs L4440-MET, L4440-IR, and L4440-HEL. Meanwhile, the 1 kb GUS gene fragment was cloned into the L4440 vector using the above method as a control. These constructs were individually transformed into the Escherichia coli strain HT115(DE3). Bacteria carring L4440-MET, L4440-IR, L4440-HEL and L4440-GUS were cultured in LB medium supplemented with 100 μg/mL ampicillin at 37 °C and 200 rpm for 16 h. To activate T7 RNA polymerase in plasmid vectors, β-D-1-thiogalactopyranoside (IPTG) was added to the bacterial culture at a final concentration of 4 μmol L⁻¹. Then, the culture was incubated at 37 °C and 220 rpm for 4 h. Total RNA was extracted from the bacteria using the RNA Simple Total RNA Kit (TIANGEN) and resuspended in 20 µL of TE buffer. After RNase A digestion (in 0.5 mol L⁻¹ NaCl), dsRNA was analyzed by agarose gel electrophoresis. The remaining culture was centrifuged at 13,000× g for 20 min, and the pellet were suspended in 1/50 volumes of 25 mM Tris and 5 m mol L⁻¹ EDTA (pH 7.5) and stored at −40 °C as “dsRNA.” [19].

2.3. Exogenous Application of dsRNAs on Watermelon Leaves

To evaluate the defense effect of dsRNAs on CGMMV infection, 12-day-old healthy watermelon seedlings were inoculated with CGMMV in a controlled greenhouse, with 20~28 °C. Leaves infected with CGMMV were collected, crushed and homogenized with 1× PBS at a 1:10 volume ratio, and 1 mL of the previously synthesized dsRNA (126 ng dsRNA/μL) was added to 4 mL of the homogenate to create the inoculum. Rubbing the inoculums onto two fully expanded cotyledons dusted with carborundum powder, using CGMMV-infected homogenate as a positive control and 1× PBS as a negative control. There were at least three biological replicates of each treatment, and for each replicate, at least 6 seedlings were treated. Inoculated plants were kept under the above conditions and monitored for symptom development until 21 days post-inoculation (dpi). The disease grading criteria were established according to the method described previously [20].

2.4. DAS-ELISA

DAS-ELISA was employed to detect the virus in the CGMMV-inoculated watermelon at 10 dpi. The procedure described by Clark and Adams (1977) [21] was modified to use CGMMV-specific antiserum (Agdia, Elkhart, IN, USA) and anti-rabbit IgG conjugated with alkaline phosphatase (Sigma Aldrich, St. Louis, MO, USA). A crude extract of watermelon leaves was prepared in 0.1 mol L⁻¹ sodium bicarbonate (pH 9.6), and 200 μL of each extract was directly loaded into a microtiter plate. The absorbance was then measured at 405 nm using an ELISA reader (Dynatech MR50000, Chantilly, VA, USA) after 30 min of incubation with a tablet solution of phosphatase substrate.

2.5. qRT-PCR Analysis

The primers used in this study are listed in the Supplementary Table S1. Complementary DNA (cDNA) was synthesized from 1 μg of total RNA with Prime script Reverse Transcriptase following the manufacturers’ instructions (Tiangen, Beijing, China). Quantitative real-time PCR was performed on a Roche LightCycler 480 system (Roche, Basel, Switzerland) using SYBR Green Master Mix (Vazyme, Nanjing, China). Following heating of the sample to 95 °C for 30 s, the cycling conditions were 95 °C for 10 s, 60 °C for 20 s, 72 °C for 20 s, and fluorescence acquisition at 72 °C. The viral CP gene was used as targeted gene to analysis the transcript level of CGMMV. The actin gene (Cla016178) was served as reference gene for normalization, as it is stable even when infected with a virus (Supplementary File S1). Gene relative expression levels were assessed using the 2^−ΔΔCT method [22]. The relative transcript level was determined in three biological replicates for each sample.

2.6. dsRNA Stability and Transport in Watermelon Leaves

To assess the stability and systemic transport capacity of dsRNA in plants, we mixed 1 mL of the produced IR-dsRNA with 4 mL of water, and then applied it to two leaves coated with quartz sand through mechanical abrasion. Water-treated plants served as the negative control. Each treatment group consisted of 18 seedlings. Local and systemic leaves were harvested at 1, 3, 5, 7, 8, and 12 dpi. Total RNA was isolated from the samples, and cDNA was synthesized utilizing random primers. Semi-quantitative RT-PCR was performed with actin as the internal reference gene and CGMMV IR-F/CGMMV IR-R as the specific primers to detect IR-dsRNA in local and systemic leaves, following the protocol described by Konakalla et al. [23] The cycling conditions consisted of aninitial denaturation at 94 °C for 2 min, followed by 35 cycles at 95 °C for 15 s, 58 °C for 20 s, 72 °C for 40 s and a final extension step at 72 °C for 4 min.

2.7. Illumina Sequencing of Viral siRNAs

RNA extracts from IR-dsRNA-treated (DC) and mock (non-treated) (CG) plants from the greenhouse assay were prepared as two distinct samples. RNA samples (DS) were further collected from dsRNA-treated leaves derived from a consortium of six non-infected plants. A minimum of 3 μg of RNA extracts was dispatched to Lc-Bio Technologies (Hangzhou, China) for short RNA (sRNA) sequencing. The sRNA sequencing library was generated utilizing TruSeq Small RNA Sample Prep Kits (Illumina, San Diego, CA, USA), and the resultant library was sequenced on the Illumina Hiseq 2500 with a single-end read length of 50 bp (SE50). Illumina sequencing adapters were removed from the raw sequences, and reads ranging from 18 to 25 nucleotides in length were utilized in further analysis. Reads of 18–25 nucleotides were aligned (MAQ) to the reference sequences of CGMMV (NCBI GenBank accession number LY64983) and quantified according to size (20, 21, 22, 23, 24, 25, total 20–25 nt) and polarity (forward, reverse, total) utilizing proprietary Fasteris scripts to generate single-nucleotide resolution maps of viral siRNAs.

2.8. Statistical Analyses

Data are shown as the means ± SD. The statistical significance was determined via the two-tailed Student’s t-test. Variations were considered statistically significant if the p value < 0.05.

3. Results

3.1. Generation of dsRNAs for the Different Regions of CGMMV

A bacterial expression system for CGMMV MET, IR, and HEL dsRNA was established to produce dsRNA in vivo. CGMMV MET, IR, and HEL segments, with 1130, 1300, and 765 bp, respectively, were amplified via RT-PCR and subsequently cloned into the L4440 vector. Through Gateway cloning of the viral amplicons, we acquired L4440-derived plasmids containing partial segments of the MET, IR, and HEL sections of CGMMV, bordered by two IPTG-inducible T7 promoters. Subsequently, they were transformed into the RNase III-deficient E. coli strain HT115 (DE3) to the expressed dsRNAs. Bacteria harboring L4440-MET, L4440-IR, or L4440-HEL were cultured and stimulated with IPTG for seven hours. Following the extraction of total RNA from the bacteria, steady dsRNA synthesis was confirmed by electrophoresis. The dsRNAs with anticipated sizes, 1130, 1300, and 765 bp, were produced in HT115 cells with plasmids L4440-MET, -IR, and -HEL, respectively (Figure 1).

3.2. Exogenous Application of MET, IR and HEL-dsRNA

To assess the control efficacy of exogenous dsRNAs against CGMMV, the dsRNAs expressed in E. coli were exogenously applied with CGMMV onto watermelon plants. Results showed that the control plants infected with CGMMV alone or co-inoculated with non-target GUS-dsRNA developed systemic mosaic and mottling symptoms by 14 dpi. In contrast, plants treated with MET-, IR-, or HEL-dsRNA showed a reduction in disease incidence and severity. The disease incidence rates for MET-, IR-, and HEL-dsRNA treated plants was 66.67%, 38.89%, and 72.22% at 14 dpi, and 72.22%, 50.00%, and 83.33% at 21 dpi, respectively (Figure 2A). Furthermore, these plants exhibited markedly less severe symptoms compared to the control groups (Figure 3). At 14 dpi, dsRNA-treated plants exhibited a significantly reduced disease index (DI) of 13, 8, and 14, compared to 22 (CGMMV-only) and 25 (CGMMV + GUS-dsRNA) in the control groups (Figure 2B). By 21 dpi, the disease index in control groups increased markedly to 73 (CGMMV-only) and 75 (CGMMV + GUS-dsRNA), while the treated groups maintained lower indices of 22, 14, and 26, respectively (Figure 2B). These results demonstrate that the exogenous application of MET, IR, and HEL-dsRNA can mitigate CGMMV infection, with IR-dsRNA having the most pronounced effect.

Consistent with the phenotypic data, molecular analysis of systemic leaves at 21 dpi revealed a significant reduction in viral load in dsRNA-treated plants. At 21 dpi, the systemic leaves of the inoculated plants were harvested, and the relative expression of the CGMMV CP gene and viral accumulation were assessed by qRT-PCR and DAS-ELISA. qRT-PCR analysis showed no significant difference in the relative expression of the CGMMV coat protein (CP) gene between the CGMMV-only and GUS-dsRNA control groups. However, compared to the CGMMV-alone control group, the expression levels of the viral CP gene in plants treated with MET-, IR-, and HEL-dsRNA were significantly reduced, reaching only 16%, 10%, and 47% of the control levels, respectively (Figure 2C). Similarly, DAS-ELISA results confirmed that there was no significant difference in viral accumulation between the two control groups. However, a significant reduction in viral accumulation was observed in plants treated with target-specific dsRNAs, with MET-dsRNA, IR-dsRNA, and HEL-dsRNA treatments showing viral levels of only 48%, 20%, and 80% of the control, respectively (Figure 2D). In conclusion, the exogenous application of MET-, IR-, and HEL-dsRNA effectively decreased both the incidence and severity of CGMMV infection by reducing the accumulation of viral RNA and protein, with IR-dsRNA exhibiting the greatest protective effect.

3.3. Systemic Movement of CGMMV IR-dsRNA

Since IR-dsRNA elicited the most significant response, it was used to trace the systemic movement of dsRNA in watermelon plants. Semi-quantitative RT-PCR was employed to analyze its presence over a time course. The results showed that IR-dsRNA levels were detectable in applied leaves up to 7 days after application (Figure 4A), and in systemic leaves up to 5 days after application (Figure 4B). The results indicate that the externally applied dsRNA can stably persist in the plant for at least 7 days and is capable of systemic movement.

3.4. High-Throughput Sequencing (HTS) of vsiRNA in CGMMV-Infected Plants

The length distribution of small RNA reads was examined utilizing the C. sativus reference in sRNAtoolbox (Figure 5) (Rueda et al., 2015) [24]. The distribution of read lengths led to the classification of the analysis by source (Figure 5). The distribution of hotspots throughout the sense and antisense strands of the viral genomes across each size category was comparable among plants (Figure 6A,B), suggesting that dsRNA therapy may modify the relative activity of DCL4 and DCL2. Reads from the sample DS matched exclusively to the IR gene of CGMMV (1545 vsiRNAs), with the antisense sequences distinctly dominating the sense sequences (Figure 6C). Concerning the favoured 5′ and 3′ terminal nucleotides of vsiRNA, both dsRNA-treated and untreated plants exhibited a predominance of 21-nt, 22-nt, and 23-nt vsiRNA with 5′-U and 3′-U, whereas 24-nt vsiRNA with 5′-A and 3′-U were predominant (Figure 7A,B).

Total raw reads obtained from three samples, including CGMMV-inoculated and dsRNA-treated plants (DC), only CGMMV-inoculated plants (CG), as well as only dsRNA-treated plants (DS), were 23,519,842, 21,949,306 and 21,784,146, respectively. After adapter trimming and quality filtering, 6,448,048 (DC), 4,557,156 (CG) and 6,075,683(DS) clean reads within the 18–25 nucleotide (nt) size range were retained for analysis. The length distribution of small RNAs was analyzed against the C. sativus reference genome using sRNA toolbox (Figure 5) (Rueda et al., 2015) [24], revealing distinct profiles for each sample group. For CGMMV-infected samples, the dsRNA-treated and untreated plants showed a comparable vsiRNA hotspot distribution pattern along the viral genome; however, a severe reduction in total vsiRNA abundance was observed in the dsRNA-treated plants (DC) (Figure 6A,B). Furthermore, vsiRNAs from the uninfected, dsRNA-treated plants (DS) mapped exclusively to the IR gene region of CGMMV, yielding 1545 reads that were predominantly antisense in orientation (Figure 6C). This similarity suggests that dsRNA treatment may alter the relative activities of DCL4 and DCL2 without drastically changing the overall profile. The significantly reduced vsiRNA abundance in DC sample indicated the successful suppression of viral replication. This is attributed to the exogenous dsRNA being efficiently processed into vsiRNAs, which triggered the RNAi machinery of plants.

The alignment of vsiRNAs to the CGMMV genome revealed distinct differences between the DC (dsRNA-treated) and CG (control) samples. Mapping of the 18–24 nt reads to the CGMMV reference genome (GenBank Acc. No. LY64983) yielded 1,099,222 vsiRNAs in the DC sample, compared to 7,520,652 in the CG sample. This represented a substantial decrease in viral-derived reads, with CGMMV vsiRNAs constituting only 5.11% of the total small RNAs in the DC sample, in contrast to 37.23% in the CG sample (Figure 6). The size distribution analysis showed that the majority (74.14%) of CGMMV-matching vsiRNAs fell within the 21–22 nucleotide range (Figure 8). In both samples, the 22-nt class was the most abundant, comprising 32.71% (DC) and 38.70% (CG) of the vsiRNAs, followed closely by the 21-nt class at 28.34% (DC) and 35.44% (CG). A slight preference for sense strand polarity was observed, with 57% of the 21–22-nt vsiRNAs exhibiting positive polarity (DC: 58.3%; CG: 56.0%).

Notably, dsRNA treatment altered the relative accumulation of different vsiRNA size classes. While the 22-nt class remained dominant in treated (DC) plants, its accumulation relative to the 24-nt class changed significantly. In CG plants, 22-nt vsiRNAs were 2.86 times more abundant than 24-nt vsiRNAs. This ratio decreased to 1.57 in DC samples, reflecting a relative decrease in the 21-nt and 22-nt classes and a concomitant increase in the 24-nt class, indicating a significant shift in the size class ratio due to the treatment (Figure 8).

In addition, analysis of the terminal nucleotides revealed that 21-nt, 22-nt, and 23-nt vsiRNAs from both CGMMV-inoculated groups were predominantly initiated with a 5′-U and terminated with a 3′-U. In the 24-nt class, vsiRNAs with a 5′-A and 3′-U were most abundant (Figure 7A,B). Uracil was the predominant terminal nucleotide, being the most frequent at both the 5′-end (DC: 35.0%; CG: 40.7%) and the 3′-end (DC: 33.3%; CG: 36.3%) of the vsiRNAs.

4. Discussion

Plant virus infections result in substantial declines in agricultural yield and can impact both local and global food security. RNAi technology has proven to be one of the most efficient methods for inducing virus resistance in plants [25]. This research aimed to develop a safe and effective anti-CGMMV method for watermelon using a non-GMO RNA vaccine that has been demonstrated to provide resistance against invasive plant viruses [17,26]. CGMMV induces significant disease issues in watermelons, leading to substantial declines in crop quality and yield. This study examined the efficacy of long dsRNA molecules, as prior research indicated that short dsRNA molecules (300 bp or shorter) diminish the RNAi efficiency against plant viruses [27]. The efficacy of CGMMV MET, IR, and HEL dsRNAs in protecting watermelon plants from CGMMV invasion was evaluated. In the biological assay, dsRNAs derived from MET and IR of CGMMV showed different levels of anti-CGMMV efficacy. IR-dsRNA (1296 bp) was more effective than MET-dsRNA (1131 bp) and HEL-dsRNA (765 bp). The function of the protein encoded by the targeted RNA may affect the protective effect [28]. Mutations in the IR of the 126 K protein encoding gene constitute a significant pathogenicity determinant of pepper mild mottle virus (PMMoV) [29,30]. This study demonstrated that the exogenous application of dsRNA targeting the CGMMV genome conferred resistance to the virus in treated plants, thereby decreasing both the incidence rate and disease index. Concurrently, the reduction in viral accumulation in the plants indicated that vsiRNA generated by exogenous dsRNA diminished or obstructed the viral infection in the inoculated plants. The observations indicate that the exogenous application of dsRNA produced from the CGMMV genome may function as a management method for suppressing CGMMV infection. In this experiment, the protective period for watermelon lasted for 5–7 days. Although no complete protective effect against CGMMV was observed, watermelon plants exhibited a degree of protection in greenhouse circumstances. Investigating the systematic transport and stability of dsRNA molecules may facilitate the exploration of additional uses for dsRNA [13]. Following the application of dsRNA to the watermelon leaves, it was swiftly translocated to the systemic leaves at 1 dpi. Gan discovered that the localized application of dsRNA can protect plants from viral infection for a duration of 5 days, potentially influenced by the flushing action of precipitation [13]. Rapid systemic dissemination of dsRNA has been documented, with TMV dsRNA identified one hour post-inoculation in both inoculated and un-inoculated systemic leaves [17]. Delgado-Martin reported that long specific RNAs can be detected systemically in the distal part of the plant after the application of the dsRNAs [31]. Namgail reported the identification of dsRNA in tomatoes for a minimum of 9 days following exogenous treatment [32]. RNA molecules are capable of entering the phloem sieve tube for systemic transport [26]. The dissemination of dsRNA may resemble that of viroids, which lack protein-coding capabilities and are non-enveloped, yet propagate within plants with the assistance of plant factors. The dsRNA used in this study may possess motifs that interact with host elements to facilitate their transport; nevertheless, the mechanism necessitates more investigation. Due to the volatility of dsRNA during external application, frequent spraying of dsRNA is essential for sustaining plant resistance to viruses during crop development. Mitter used layered double hydroxide (LDH) nanosheets as a carrier for the delivery of dsRNA, which enhanced the stability of dsRNA on the leaf surface, enabled the slow release of dsRNA, and thus extended the effective period of crop protection [16]. The external use of exposed dsRNA, used as an eco-friendly agent to protect crops against viral infections and encapsulated in nanomaterials, will serve as a more effective technique for managing plant viruses, which will be the emphasis of our forthcoming research.

We performed a comprehensive sequencing investigation of the siRNA virus in leaves infected with CGMMV throughout the organism. In comparison to plants subjected to dsRNA treatment, such treatment not only markedly diminished the quantity of virus-derived siRNAs but also altered their size classifications, resulting in 22-nt and 24-nt siRNAs exhibiting equivalent representativeness. On the contrary, in cucumber, Delgado-Martin observed the 21 nt followed by the 22 nt as the predominant sRNAs in leaves [31]. This indicates that dsRNA treatment may enhance the synthesis of 24-nt siRNAs. This aligns with the primary functions of DCL4 and DCL2 in RNA interference-mediated cytoplasmic RNA virus defence, particularly against CGMMV, as well as prior research on the production of CGMMV-derived siRNA in Arabidopsis (Wang et al., 2010; Wang et al., 2011) [33,34]. In Arabidopsis plants infected with CGMMV, DCL4, responsible for generating 21-nt sRNAs, is pivotal in antiviral defence, whereas DCL2, which produces 22-nt sRNAs, serves as a secondary mechanism that assumes antiviral defence in the absence or inactivation of DCL4 [31,34,35]. The DNA virus is also targeted by dsRNA processed by nuclear DCL3, which is translated into 24-nt siRNA through both sense and antisense transcription (Pooggin.2018) [35]. After these siRNAs combine with the proteins of the Argonaute (AGO) family, they guide the formation of the RNA-induced silencing complex (RISC), thereby cleaving, degrading, or restricting the translation process of viral RNA. Using comprehensive small RNA sequencing and bioinformatics, we demonstrated that dsRNA-mediated interference against systemic CGMMV infection correlates with a substantial decrease in the accumulation and ratio of 22- and 24-nucleotide viral siRNA, signifying a modification in the relative activity of DCL2 and DCL3. The favoured 5′ and 3′ terminal nucleotides of vsiRNA in plants, whether treated or untreated with dsRNA, predominantly consist of 5′-A and 3′-U, corroborating findings on vsiRNA in plant viruses, which often initiate with A or U [36]. AGO1 and AGO2 preferentially associate with sRNAs possessing 5′ terminal of U and A, respectively, while many AGO proteins participate in the processing of vsiRNA inside the watermelon RISC [37].

5. Conclusions

This study presented the implementation of an RNA vaccine targeting CGMMV in watermelon. When dsRNA molecules are locally applied to the leaf surface, IR-dsRNA exhibits a protective effect. In watermelon subjected to dsRNA treatment, virus titers and vsiRNA were markedly diminished. DsRNA may need to be applied once a week in crop productivity. Double-stranded RNA was identified in both injected and systemic leaves. Deep small RNA sequencing study revealed that a limited quantity of vsiRNA accumulated in the leaves of plants treated with IR-dsRNA, with size class proportions shifting to 22- and 24-nt, indicating that dsRNA treatment may augment the antiviral efficacy of DCL3.