We previously showed that biosynthesis of the ET precursor ACC is inhibited by CGMMV infection in a susceptible watermelon variety, enhancing its susceptibility to viral infection (Liu et al., 2023a). To explore the role of the ET pathway in plant defense against CGMMV infection, we measured the concentrations of ACC and ET in mock- and CGMMV-infected plants of a susceptible watermelon variety (Zhengkang No. 2) at 10 days post-inoculation (dpi) ( Figures 1A, B ). At this stage, we observed typical symptoms, including mottle and mosaic patterns on leaves, in CGMMV-infected watermelons. In addition, CGMMV RNAs accumulated to high levels in these plants (Liu et al., 2023a). ACC levels in CGMMV-infected plants were about half of those in mock-inoculated plants ( Figure 1A ), whereas ET levels were approximately twice as high in CGMMV-infected as in mock-inoculated plants ( Figure 1B ).

We wondered whether ACC and ET play opposite roles in antiviral defense in watermelons. To address this question, we sprayed watermelon leaves with ACC (2 mM), ET (500 µM), or H₂O (as a control) 72 h before inoculation of CGMMV. At 10 dpi, ACC-treated plants were more susceptible to CGMMV infection, displaying more severe disease symptoms, higher disease indices, and higher levels of viral accumulation than H₂O-treated plants ( Figures 1D-F ). However, ET-treated plants showed that resistance to CGMMV was enhanced, with the exhibition of milder disease symptoms, lower disease indices, and lower levels of virus accumulation, compared to H₂O-treated plants ( Figures 1D-F ). These results undoubtedly suggest that ACC increases the susceptibility of watermelon to CGMMV infection, whereas ET enhances its resistance to this virus.

To identify the genes of watermelon responsible for ET accumulation upon CGMMV infection, we analyzed the gene expression profiles of mock- and CGMMV-infected watermelon seedlings by RNA sequencing (RNA-Seq) (Liu et al., 2023a). The expression levels of >1,200 genes significantly altered upon CGMMV infection (Liu et al., 2023a). Gene Ontology (GO) analysis showed that these DEGs were enriched for biological processes in the ET response category ( Supplementary Figure S1A ). In addition, the expression levels of approximately two-thirds of the 30 ET-related genes were altered in CGMMV-infected watermelons, including genes encoding ET synthases (ACO genes), ET receptors (ETR genes), and ET-responsive TFs (ERF genes) ( Supplementary Figure S1B ). Specifically, the expression of most ERF, ACO, and ETR genes was inhibited ( Figure 1C and Supplementary Figure S1B ), whereas the expression of only one ACO (Cla97C07G144530) and two ERF genes (Cla97C06G122830 and Cla97C08G147180) was induced by CGMMV infection. Quantitative reverse-transcription PCR (RT-qPCR) further confirmed the results of RNA-Seq data ( Figure 1C ). All those results suggest that these ET-related genes might contribute to antiviral defense responses in watermelons.

Three DEGs annotated as ACO genes displayed marked changes in expression ( Figure 1C ), with only one (Cla97C07G144530) being upregulated upon CGMMV infection. We, therefore, explored whether this gene might be related to the decrease of ACC content and increase of ET content upon CGMMV infection. Sequence analysis showed that the cDNA amplified from the contig included a 1,038-bp open reading frame encoding a protein of 345 amino acids with a predicted molecular mass of 39.7 kDa and an isoelectric point of 6.07. BLAST analysis and sequence alignment of Cla97C07G144530 with its homologs revealed that its predicted amino acid sequence shares 87.54%, and 65.31% amino acid identity with CsACO5 (1-aminocyclopropane-1-carboxylate oxidase 5, XP_004137205) in cucumber (Cucumis sativus) and AtDMR6 (AT5G24530.1) in Arabidopsis thaliana ( Supplementary Figure S2 ), respectively. We named the Cla97C07G144530 gene as ClACO5.

Bioinformatics prediction revealed that ClACO5 might be localized to the cytoplasm ( Supplementary Figure S3 ). To confirm this prediction, 35S-ClACO5-GFP or 35S-GFP was co-transformed with a nuclear marker protein H2B (35S-H2B-mCherry) or a membrane marker protein OsMCA1 (35S-OsMCA1-mCherry) into Nicotiana benthamiana leaves. The GFP signals in leaf cells expressing 35S-ClACO5-GFP or 35S-GFP completely overlapped with the signals of mCherry ( Figure 2 ), suggesting that ClACO5 was localized in the cytoplasm and nucleus of N. benthamiana.

We generated ClACO5-knockdown watermelon plants (pV190-ClACO5) via agroinfiltration by our previously constructed pV190-based virus-induced gene silencing (VIGS) system ( Figure 3A ). ClACO5 expression levels of silenced plants were only 23% of that in control plants at 20 dpi ( Figure 3B ), as well as demonstrated less ET accumulation than controls ( Figure 3C ). To examine the impact of ClACO5 downregulation on CGMMV accumulation, we used systemically infected leaves from control (phytoene desaturase [PDS]-silenced) and ClACO5-silenced watermelon plants to measure CGMMV RNA levels. The accumulation of CGMMV gRNA in ClACO5-knockdown plants increased by 230% and 570% compared to that in PDS-knockdown plants, as revealed by RT-qPCR and RNA gel blot analysis, respectively ( Figures 3D, E ). These findings manifest that ClACO5 increases ET levels and enhances defense against CGMMV in watermelon plants.

To further confirm that ClACO5 functions in plant defense against CGMMV infection, we exogenously overexpressed ClACO5 in N. benthamiana (a CGMMV-susceptible species) and selected three transgenic lines to assess CGMMV resistance. Under normal growth conditions, the wild-type (WT) and transgenic tobacco plants did not exhibit phenotypic differences ( Figure 3F ), and ClACO5 transcript abundance was significantly higher in the transgenic plants ( Figure 3H ). However, upon CGMMV infection, these plants showed milder CGMMV disease symptoms, including milder mottle and mosaic symptoms in systemically infected leaves, than the WT plants ( Figure 3G ). In addition, transgenic N. benthamiana plants accumulated significantly less CGMMV than the WT plants ( Figure 3I ). Collectively, these results show that knocking down ClACO5 significantly enhanced CGMMV susceptibility, whereas overexpressing this gene enhanced resistance to this virus, indicating that ClACO5 plays a positive role in ET biosynthesis.

To identify TFs that regulate ClACO5 expression, we conducted a manual search of the NewPLACE database for the cis-elements in the ClACO5 promoter sequence (~2 kb) involved in responses to plant pathogens. Several W-boxes and MYB motifs in the ClACO5 promoter were identified, which are recognized by WRKY and MYB transcription factors, respectively. An abscisic-acid-response element (ABRE), an SA-responsive element, and a methyljasmonate (MeJA)-responsive element were also found in the ClACO5 promoter ( Supplementary Table S1 and Supplementary Figure S4 ). Analysis of the RNA-Seq data found that CGMMV also induced ClWRKY70 (Cla97C10G206240) expression and that this gene shared similar expression patterns with ClACO5 ( Figure 4A ). Next, we examined the subcellular localization of ClWRKY70 by transiently co-expressed 35S-ClWRKY6-GFP or 35S-GFP with the nuclear marker gene H2B fused to mCherry in N. benthamiana leaves. The GFP signal from the 35S-GFP vector was ubiquitously distributed throughout the cell. By contrast, the GFP signal from the 35S-ClWRKY70-GFP vector was exclusively localized to the nucleus and co-localized with mCherry ( Figure 4C ). To examine the transcription-activating activity of ClWRKY70, a transactivation reporter assay was performed. Yeast strain AH109 containing pGBKT7-ClWRKY70 grew well on SD/–Trp or SD/–Trp/–Ade/–His medium, and displayed GAL4 activity on SD/–Trp/–Ade/–His medium supplemented with X-α-Gal, suggesting that ClWRKY70 is a transcriptional activator ( Figure 4B ).

CGMMV infection increased the abundance of ClWRKY70 and ClACO5 transcripts; these genes play vital roles in plant responses to pathogen infection. We also identified several WRKY-binding motifs in the ClACO5 promoter ( Supplementary Table S1 ). These observations prompted us to examine whether ClWRKY70 interacts with ClACO5 in response to CGMMV infection. A yeast one-hybrid assay (Y1H) confirmed that ClWRKY70 could bind to the W-box element in the ClACO5 promoter ( Figure 5B ). We also conducted a GUS transactivation experiment in N. benthamiana leaves to further validate this interaction. Leaves co-transformed with pCNF-ClWRKY70, and ClACO5pro-GUS exhibited significantly higher GUS activities than leaves infiltrated with the empty pCNF effector and ClACO5pro-GUS, as shown by the deeper blue color staining image and higher expression level of the GUS gene ( Figure 5D ).

To further determine the interaction between ClWRKY70 and the ClACO5 promoter, we carried out a dual-luciferase reporter assay. The binding of ClWRKY70 to the ClACO5 promoter led to a nearly three-fold increase in the relative LUC/REN ratio vs. the control ( Figure 5C ). To verify the role of ClWRKY70 in regulating the expression of ClACO5 in watermelon, we transiently transformed the leaves of 7-day-old seedlings with the pCNF-ClWRKY70 construct or empty vector (pCNF-EGFP). ClWRKY70 transcript levels were significantly higher in leaves transformed with the pCNF-ClWRKY70 construct vs. the empty vector control, and the transcript levels of ClACO5 increased nearly two-fold in leaves transformed with pCNF-ClWRKY70 vs. the pCNF-EGFP control ( Figure 5A ). Collectively, our results implied that ClWRKY70 functions as a transcriptional activator of ClACO5 and promotes its transcription by binding to the ClACO5 promoter.

Since ClWRKY70 participated in plant responses to Acidovorax citrulli infection (Zhang et al., 2020), and ClWRKY70 is an upstream regulator of ClACO5, we reasoned that ClWRKY70 might promote systemic acquired resistance to CGMMV infection. Next, we generated ClWRKY70-overexpressing transgenic N. benthamiana plants and selected three independent lines that had significantly higher ClWRKY70 expression levels than the WT plants ( Figure 6B ). Under normal growth conditions, the WT and transgenic plants did not display phenotypic differences ( Figure 6A ). However, after CGMMV infection, transgenic plants exhibited milder mosaic and mottle symptoms in systemically infected leaves compared to WT plants ( Figure 6C ). In addition, WT plants had higher CGMMV CP transcript levels than transgenic plants ( Figure 6D ). The aforementioned results suggest that the heterologous expression of ClWRKY70 enhanced CGMMV resistance.

To further probe the function of ClWRKY70 in regulating the host’s susceptibility to CGMMV, we used the pV190 VIGS system to silence ClWRKY70 in watermelon plants. Knockdown of ClWRKY70 resulted in the simultaneous downregulation of ClACO5 ( Figures 6E, F ). Finally, we analyzed CGMMV accumulation in ClWRKY70-knockdown and control (PDS-silenced) watermelon plants by RT-qPCR, finding that it was 62% higher in the ClWRKY70-knockdown plants ( Figure 6G ). This indicated that the knockdown of ClWRKY70 increased the susceptibility of watermelons to CGMMV infection.

Plants have developed complex systems to defend themselves against viral attacks, in which phytohormones play essential roles in facilitating defense signaling in plants (Mishra et al., 2020). ET plays a dual role in plant-virus interactions; it promotes resistance in some interactions and contributes to disease progression in others (Chen et al., 2013; Zhu et al., 2014b; Zhao et al., 2017). For example, the rice dwarf virus (RDV) increases ET production by interacting with OsSAMS1, an essential component of the ET biosynthesis, resulting in an increase in the enzyme’s activity, thereby increasing susceptibility to RDV (Zhao et al., 2017). Southern rice black-streaked dwarf virus (SRBSDV) encoded P6 protein subcellular localization is changed during different stages of SRBSDV infection differentially altering ethylene signaling to support SRBSDV infection and transmission (Zhao et al., 2022). In the current study, we observed a marked increase in ET production in watermelon plants after CGMMV infection. When exogenous ET was applied prior to the mechanical inoculation of CGMMV, watermelon leaves had milder disease symptoms, lower disease indices, and reduced virus accumulation compared with mock plants ( Figure 1 ). These results imply that ET accumulation in watermelon plants might promote resistance to CGMMV infection. The opposite roles of ET in plant responses to viral infection might be related to different virus-host combinations as well as crosstalk between ET and other phytohormones.

ACS and ACO are major enzymes that catalyze the two key steps of ET biosynthesis (Kim et al., 2003). Eight ACS and eight ACO genes of watermelon plants were previously identified (Zhou et al., 2016). In the current research, we found the expressions of three ClACO genes (Cla97C07G140100, Cla97C07G144530, and Cla97C10G205440) were altered by CGMMV infection by analyzing the transcript levels of these genes under CGMMV infection. Among them, only Cla97C07G144530 (ClACO5) was upregulated. Whereas the transcript levels of all ClACS genes were unchanged ( Supplementary Figure S1 ). Thus, ClACO genes play more important roles than ClACS genes in the response of watermelon plants to CGMMV infection.

The expression pattern of ClACO5 is largely consistent with the changes in ET levels in watermelon leaves during CGMMV infection ( Figure 3C ), and ClACO5 shares high sequence similarity with CuACO5 and AtDMR6 ( Supplementary Figure S2 ). Based on these results, we speculate that ClACO5 may be vital for ET production during CGMMV infection in watermelon leaves. AtDMR6 encodes a putative 2OG-Fe(II) oxygenase associated with defense responses but required for susceptibility to downy mildew (Zeilmaker et al., 2015); it also participates in plant responses to bacteria and oomycetes, flavonoid biosynthesis, and SA catabolism (Falcone Ferreyra et al., 2015; Zhang et al., 2017), but it is unclear whether this gene participates in plant responses to viruses. In the present study, the knockdown of ClACO5 in watermelons conferred enhanced susceptibility to CGMMV infection, whereas heterologous overexpression of this gene in N. benthamiana enhanced resistance to CGMMV infection ( Figure 3 ). Taken together, we conclude that ClACO5 functions in plant responses to CGMMV infection and enhances CGMMV resistance.

WRKY TFs regulate gene expression by binding directly to the W-box in their target genes’ promoters (Rushton et al., 2010). WRKY70s in some species, such as Arabidopsis, wheat (Triticum aestivum), and tomato (Solanum lycopersicum), play vital roles in plant responses to pathogen infection (Atamian et al., 2012; Wang et al., 2017; Zhou et al., 2018; Liu et al., 2021). However, far less is known about their roles in defense against viral pathogens as compared to fungal or bacterial pathogens. In the present study, we report that ClWRKY70 is induced by CGMMV infection. We observed a marked difference between ClWRKY70-overexpressing transgenic N. benthamiana and WT plants in leaves systemically infected with CGMMV, suggesting that ClWRKY70 might inhibit CGMMV accumulation.

WRKY70 regulates the expression of numerous defense-related genes (Machens et al., 2014), including SARD1, PR1, PR2, and PR5 ( Li et al., 2004). WRKY70 may also inhibit ET- and JA-responsive gene expression (Li et al., 2004; Li et al., 2006; Li et al., 2019), but whether it regulates the expression of genes related to phytohormone biosynthesis remains unresolved. In the current study, we found that both ClWRKY70 and ClACO5 were consistently upregulated after CGMMV infection ( Figures 1C , 4A ). ClWRKY70 encodes a TF that positively regulates watermelon resistance to CGMMV ( Supplementary Figures 4 , 6 ). We also showed that the ClACO5 promoter contains a WRKY-binding site. ClWRKY70 directly bound to the ClACO5 promoter and enhanced its activity in N. benthamiana leaves and upregulated this gene in watermelon leaves ( Supplementary Figure S4 and Figure 5 ), implying that ClACO5 is a target gene of ClWRKY70. Based on these results, we propose that ClWRKY70 functions as a major positive regulator of ET biosynthesis by activating ClACO5 expression in watermelon leaves and contributes to the resistance of watermelon plants to CGMMV infection.

SA can promote ClWRKY70 expression (Zhang et al., 2020), and we previously determined that CGMMV infection can induce SA accumulation and that treatment with exogenous SA significantly enhances the CGMMV resistance of watermelons (Liu et al., 2023a). Based on these and the current findings, we propose that SA affects ET biosynthesis by inducing the expression of ClWRKY70 and therefore activation of ClACO5 enhanced resistance of watermelon plants to CGMMV infection. In addition, we previously showed that SA-induced flavonoid biosynthesis plays a vital role in watermelon development and CGMMV resistance (Liu et al., 2023a), whereas AtDMR6, which is highly homologous to ClACO5, has flavone synthase activity (Falcone Ferreyra et al., 2015). Therefore, it would be worth investigating whether SA-induced flavonoid biosynthesis also involves the regulation of ClACO5 expression by ClWRKY70 in watermelon leaves defense against CGMMV infection and whether SA, ET, and flavonoids form a complex transcriptional regulatory network during the defense response of watermelons against CGMMV. Furthermore, we previously found that ET could induce ClAGO5 expression, thus speculating that ethylene may enhance watermelon defense against CGMMV via activating RNA silencing pathway (Liu et al., 2023b).

Taken together, our results demonstrate that ClACO5, localized to the nucleus and cytoplasm, and involved in ET biosynthesis, enhances CGMMV resistance. We propose that CGMMV infection increases SA biosynthesis in watermelon plants, resulting in the activation of downstream signaling pathways. ClWRKY70, involved in SA signaling, regulates ClACO5 expression by interacting with the W-box element in its promoter, thereby increasing ET levels ( Figure 7 ). This model unveils the molecular mechanisms and transcriptional regulatory network of ET accumulation in response to CGMMV infection.

The susceptible watermelon variety Zhengkang No. 2 and tobacco (N. benthamiana) were used in this research work. To analyze the effects of exogenous ET and ACC on plant responses to CGMMV, 7-day-old watermelon seedlings were sprayed with water, 500 μM ET, or 2 mM ACC for 3 days before CGMMV inoculation. Three biological replicates of leaves were collected 10 days after CGMMV treatment and used for RNA extraction.

ET production was measured according to a previous method (Kawano and Shimokawa, 1994). Approximately 5 g of leaf tissue was placed into a glass bottle. The bottle was sealed and incubated at room temperature for at least 5 hours. A 1ml sample of gas was removed from the bottle and analyzed by gas chromatography (Shimadzu GC-2010).

Phytohormones ACC contents were detected by MetWare (http://www.metware.cn/) based on the AB Sciex QTRAP 6500 LC-MS/MS platform.

Multiple alignment of the putative protein sequences encoded by ClACO5 and its homologs was performed using ClustalW software in conjunction with MEGA 7.0; then, the molecular weight and theoretical isoelectric point of the deduced ClACO5 protein was predicted using ExPASy (http://www.expasy.org), an online tool. The NewPLACE (https://www.dna.affrc.go.jp/PLACE/) database was used to identify the cis-elements in the ClACO5 promoter.

Total RNAs from watermelon leaf tissues were extracted using the method listed in the kit manual (Tiangen, Beijing, China). After the first-strand cDNA was synthesized, five-fold diluted cDNA products were subjected to qPCR using SYBR Green Master Mix (Vazyme, Nanjing, China) on a Roche Real-Time PCR system. We analyzed gene expression levels by the 2^–△△CT method (Livak and Schmittgen, 2001). The primers used for RT-qPCR are included in Supplementary Table S2 .

The ClACO5 and ClWRKY70 coding sequences (CDSs) without the stop codons were amplified via 2×Fast Pfu Master Mix (novoprotein, Suzhou, China) and cloned into the pCNF-GFP vector. The recombinant 35S-ClACO5-GFP, 35S-ClWRKY70-GFP, and 35S-GFP (control) plasmids were transferred into the A.tumefaciens strain GV3101. N. benthamiana leaves were infiltrated with the bacterial suspensions, together with a plasmid expressing nuclear or membrane markers, 35S-H2B-mCherry or 35S-OsMCA1-mCherry (OD600~0.6). We observed the subcellular localizations of 35S-H2B-mCherry, 35S-OsMCA1-mCherry, 35S-ClACO5-GFP, 35S-ClWRKY70-GFP, and 35S-GFP under a confocal laser-scanning microscope.

The ClWRKY70 CDS was cloned into the pGBKT7 vector. Yeast strain AH109 cells were transformed with the resulting construct and plated on SD/–Trp, SD/–His/–Trp/–Ade, or SD/–Trp/–Ade/–His supplemented with X-α-Gal. The pGBKT7 + pGADT7-T vector was used as a negative control.

The ClWRKY70 CDS was ligated into the pGADT7 vector. The 789-bp promoter region containing a W-box cis-element was amplified with primers AD-ClWRKY70-F/R. Then, the amplified fragment was ligated into the pAbAi vector to generate the bait construct. An interaction assay of the ClWRKY70 and ClACO5 promotor fragments was conducted using a Matchmaker™ Gold Yeast One-Hybrid Library Screening System Kit (Clontech, San Francisco, USA).

The PCR-amplified 789-bp fragment of the ClACO5 promoter was cloned into pCAMBIA3301 and pGreenII 0800-LUC to generate reporter constructs. The ClWRKY70 CDS was cloned into pCNF to produce the effector construct. The reporter and effector constructs were transferred into A. tumefaciens GV3101 with the pSoup helper vector and A. tumefaciens strain GV3101, respectively, which were then used to coinfect 4-week-old tobacco leaves. At 48-72 h after infiltration, LUC and REN activities and GUS activity were measured in the leaves of these plants.

The expression construct pCNF-ClWRKY70 was transferred into A. tumefaciens GV3101; the empty vector pCNF was used as a negative control. Watermelon leaves were infiltrated with these cultures, collected 3 days after infiltration, and used to analyze the expression level of ClWRKY70 and ClACO5.

A 300-bp fragment of ClACO5 or ClWRKY70 was amplified with gene-specific primers pV190-ClWRKY70-F/R or pV190-ClACO5-F/R ( Supplementary Table S2 ), and cloned into pV190. Then, according to the previous description, all recombinant plasmid were separately transformed and injected into watermelon leaves (Liu et al., 2020). At 20 dpi, systemically infected leaves were collected for qPCR and RNA gel blot analyses.

The ClACO5 or ClWRKY70 CDS was amplified and ligated into the expression vector pBWA(V)HS-ccdB-GFP (Biorun, Wuhan, China). The A. tumefaciens strain GV3101 harboring the ClACO5 or ClWRKY70 ORF recombinant plasmids was transformed into N. benthamiana. We selected three positive T₂ transgenic lines for further experiments.

N. benthamiana plants

N. benthamiana plants from the WT line and three overexpressed ClACO5 transgenic lines (OE#4, OE#9, and OE#21) or three overexpressed ClWRKY70 transgenic lines (OE#1, OE#2, and OE#3) were inoculated with CGMMV. At 10 dpi, the plants were observed, and leaves were collected for measurement of target gene expression and CGMMV accumulation.

The leaves of watermelon plants from the VIGS experiment were collected, and total RNA was extracted and used to analyze CGMMV accumulation with a DIG Northern Starter kit (Roche, Mannheim, Germany). The detailed method refers to a previously described method (Liu et al., 2017).

The inoculations of CGMMV were repeated at least twice with at least three replicates for each line and each time point. GraphPad Prism 9.0 was used to process all data and analyze statistical differences using Student’s t-tests at significance levels of p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***), and p ≤ 0.0001 (****).