Background

Tung tree, belonging to the Euphorbiaceae family, are important woody oil-producing trees in China. Tung oil produced from tung tree seeds is rich in the trivalent unsaturated fatty acid eleostearic acid [1, 2], which has excellent properties such as corrosion resistance and acid and alkali resistance. Vernicia fordii and Vernicia montana represent the two principal cultivars in China. Compared to V. montana, the seeds of V. fordii exhibit a more rapid maturation process and yield superior oil quality. However, the recent surge in Fusarium wilt occurrences has adversely affected the cultivation and industrial development of tung tree. Previous studies have revealed that V. fordii is susceptible to Fusarium wilt, while V. montana exhibits effective resistance to this disease [3, 4]. At present, there is no established cure for Fusarium wilt in V. fordii. The most effective strategy for managing this disease involves using disease-resistant V. montana as the rootstock and grafting it with V. fordii as the scion. Consequently, it is imperative to explore the disease resistance mechanisms in V. montana and understand the factors that contribute to the susceptibility of V. fordii to Fusarium wilt.

Plants encode multiple disease-resistance (R) genes that confer resistance to insects and pathogens [3,4,5]. In the past few decades, researchers have cloned over 300 R genes from plants [6]. The proteins encoded by these R genes exhibit diverse domain combinations [6, 7]. Among these domains, the nucleotide-binding site (NBS) and leucine-rich repeat (LRR) domains are the most widespread in known R genes, collectively referred to as NBS-LRR genes [8]. The NBS-LRRs originate from green plants [9], and their encoded proteins facilitate plant resistance primarily by recognizing receptors within pathogens themselves [10]. Considering the variations in their N-termini structures, NBS-LRRs can be further categorized into two types: Toll/interleukin-1 receptor (TIR)-NBS–LRR (TNL) types, characterized by a TIR domain, and non-TIR-NBS-LRR (non-TNL) types, which feature either leucine zipper (LZ) or coiled-coil (CC) domains in place of the TIR domain [8, 11, 12]. The C-terminal LRR domains of NBS-LRR proteins are crucial for protein-protein interactions and pathogens recognition specificity [13, 14]. Conversely, the N-terminal NBS domains bind GTP and ATP facilitating hydrolytic reactions that provide energy for downstream signaling processes [13, 14]. NBS-LRRs primarily confer disease resistance in plants by recognizing pathogens. For example, the NBS-LRR protein RPS5 is capable of identifying bacteria expressing the type III effector AvrPphB, thereby conferring resistance to downy mildew [15]. An NBS-LRR protein located at the Rpp1 locus counteracts the effectiveness of Rpp1-mediated resistance to Phakopsora pachyrhizi in soybean [16]. Through rapid evolution, R genes enable plants to detect avirulence genes in various pathogens, triggering downstream signaling cascades that culminate in programmed cell death, hypersensitive reactions, and defense responses [17,18,19,20].

Recent advancements in sequencing technologies have facilitated the identification of NBS-LRR genes across a broad spectrum of plant species, such as Arabidopsis, rice, cabbage, grape, and sunflower [12, 21,22,23,24]. These investigations have revealed variations in the size of the NBS-LRR family across different plant genomes. Here, we determined members of the NBS-LRR family and conducted a comparative analysis across the genomes of V. fordii and V. montana. Furthermore, we carried out functional analyses to elucidate the roles of these genes in Fusarium wilt resistance between V. montana and V. fordii. As a result, we established a potential contribution of VmNBS-LRR to the resistance of V. montana’s resistance to Fusarium wilt. These findings provide crucial insights into Fusarium wilt-responsive NBS-LRRs, which can serve as pivotal targets for molecular breeding aimed at enhancing disease resistance in tung tree.

Results and discussion

NBS–LRRs in V. montana and V. fordii

Utilizing HMMER software for analysis, we identified a combined total of 239 NBS-containing sequences in the two Vernicia species: 90 in V. fordii and 149 in V. montana. From the V. fordii NBS-LRRs, 90 VfNBS-LRRs were categorized into four subgroups: CC-NBS-LRR (12), NBS-LRR (12), CC-NBS (37), and NBS (29) (Table 1 and Additional file 1). Notably, 49 VfNBS-LRRs from V. fordii contained the CC domain, accounting for 54.4% of the VfNBS-LRRs. It is important to highlight that no TIR domains were found in VfNBS-LRRs, indicating that none of the resistance genes in V. fordii belonged to the TIR class. This is consistent with previous studies that have shown the absence of TIR domain-containing NBS-LRRs in monocots, whereas they are more commonly observed in eudicots [10]. In fact, the loss of NBS-LRRs containing TIR domains in eudicots has only been reported in Sesamum indicum [25] and is now also observed in V. fordii as reported in this study. However, among the 149 VmNBS-LRRs identified in V. montana, they were divided into seven subgroups: CC-NBS-LRR (9), TIR-NBS-LRR (3), CC-TIR-NBS (2), TIR-NBS (7), NBS-LRR (12), CC-NBS (87), and NBS (29) (Table 1 and Additional file 2). Among them, 98 VmNBS-LRRs contained the CC domains, accounting for 65.8% of the VmNBS-LRRs, 12 VmNBS-LRRs possessed the TIR domains (8.1%), and 2 VmNBS-LRRs contained both the CC and TIR domains. These findings provide valuable insights into the evolutionary aspects of plant disease-resistance genes.

Table 1 Classification of NBS-LRR genes in V. fordii and V. montana
Full size table

The LRR domain is vital for plant immune responses, facilitating both protein-ligand and protein-protein interactions [26,27,28]. In this study, we identified two distinct LRR domains, LRR3 and LRR8, among the 90 VfNBS-LRRs of V. fordii, with 4 and 23 numbers, respectively (Additional file 3). Conversely, V. montana displayed four types of LRR domains: LRR1 (2), LRR3 (4), LRR4 (2), and LRR8 (20) (Additional file 4). The LRR1 domain was exclusive to the NL and CNL proteins of V. montana, while the LRR4 domain was only present in the NL proteins of V. montana. Interestingly, these two LRR domains were not found in V. fordii, indicating the occurrence of LRR domain loss events in V. fordii during evolution.

Chromosomal distributions and evolutionary analysis of NBS-LRRs

The genomes of Vernicia species, comprising 11 chromosomes [2, 29], are designated as Vfchr.1-Vfchr.11 in V. fordii and Vmchr.1-Vmchr.11 in V. montana. To investigate the chromosomal distribution of NBS-LRR genes in these species, we first analyzed the homologous chromosome relationships between their genomes (Fig. 1a). One-to-one syntenic relationships were observed, such as Vfchr1 in V. fordii corresponded to Vmchr1 in V. montana, and Vfchr2 and Vfchr3 respectively aligned with Vmchr7 and Vmchr2 (Fig. 1a). Subsequently, we found significant differences in the distributions of NBS-LRR genes across these chromosomes between the two species (Fig. 1b–d). In V. fordii, a higher number of VfNBS-LRRs were located on Vfchr2, Vfchr3, and Vfchr9, while a lower number was found on Vfchr1, Vfchr4, and Vfchr10 (Fig. 1b, c). Similarly, in V. montana, a higher number of VmNBS-LRRs were present on Vmchr2, Vmchr7, and Vmchr11, while a lower number was observed on Vmchr9 and Vmchr10 (Fig. 1c, d). These results suggest that NBS-LRR genes are distributed non-randomly across all chromosomes, showing a clustered distribution. The enrichment of NBS-LRRs in corresponding genomic regions suggested that the evolution of resistance genes may involve tandem duplications of linked gene families, as reported by previous studies [8, 30].

Fig. 1
figure 1

Chromosomal distribution analysis of NBS-LRR genes from Vernicia. a Analysis of homologous chromosomes between V. fordii and V. montana. b VfNBS-LRR genes were mapped on the chromosomes of V. fordii. c Comparison of the location of NBS–LRRs on each chromosome between V. fordii and V. montana. d VmNBS-LRR genes were mapped on the chromosomes of V. montana

Full size image

In general, NBS-LRRs can be divided into various evolutionary groups based on the differences in their structural domains [12, 31]. In this study, we observed that V. fordii exhibited a smaller number of LRR domains compared to V. montana. Additionally, V. montana displayed a wider variety of LRR domains within its VmNBS-LRRs relative to VfNBS-LRRs. To investigate the relationship between NBS-LRRs in V. montana and V. fordii, we detected one-to-one orthologous gene pairs between these two sister species of Vernicia (Fig. 2a and Additional file 5). Totally, 43 one-to-one orthologous gene pairs were identified, and most of these pairs were located on corresponding sister chromosomes in the Vernicia genomes. Furthermore, we compared the domains and groups of the orthologous NBS-LRRs genes and found that some CNL sequences were orthologous genes with TNL genes. In Arachis, three TNL genes were identified within the CNL group, while one TNL gene was found clustering within the CNL gene [31]. Similar observations have been reported in other plant species such as Eucalyptus grandis, Vitis vinifera, and Medicago truncatula, where TNL genes are integrated within the CNL group [32,33,34]. These findings suggest that the recombination events have occurred within the NBS domain. For example, Innes et al. [35] confirmed that recombination occurred between some NBS domains from TNL and CNL sequences.

Fig. 2
figure 2

Gene duplication and environmental selection pressure analysis. a One-to-one orthologous NBS-LRR genes between V. fordii and V. montana. The green lines indicate orthologous NBS-LRR genes in a similar homologous chromosomal location between V. fordii and V. montana. b Gene duplication analysis of V. montana. The orange lines indicate NBS-LRR duplication genes in V. montana. c Scatter plots of the Ka/Ks ratios of orthologous NBS-LRR gene pairs between V. fordii and V. montana. The X-axes and Y-axes indicate the Ka and Ka/Ks ratio for each pair, respectively. d Frequency distribution histogram of Ka/Ks of orthologous NBS-LRR gene pairs between V. fordii and V. montana

Full size image

Gene duplication and environmental selection pressure analysis

Plants have undergone one or multiple polyploidization events during their long evolutionary history [36,37,38]. Tung tree has also experienced at least one polyploidization event(s) [2, 29]. To delve deeper into the impact of polyploidization on NBS-LRRs in V. montana and V. fordii, we assessed the intraspecific collinearity of their genomes using the MCScanX software [39]. This approach facilitated the identification of paralogous genes within the NBS-LRRs of V. montana and V. fordii, as these genes have the potential to generate novel resistant functions compared to ancestral genes [40]. Similar to NBS-LRRs found in other plants [21, 23, 24], the NBS-LRRs in the Vernicia genomes have undergone many tandem duplication events. Specifically, we identified a total of 10 tandemly duplicated NBS-LRR sequences in the V. fordii genome (Fig. 1b), and 20 tandemly duplicated NBS-LRR sequences in the V. montana genome (Fig. 1d). Chromosomes 2 in V. fordii and chromosomes 2 and 11 in V. montana exhibited the highest number of tandemly duplicated genes.

In domesticated plant species, duplication events have been observed to occur more frequently [41]. Segmental duplication events have significantly contributed to neo-functionalization or sub-functionalization processes, resulting in the acquisition of novel and distinct functions in comparison to the ancestral genes [42]. The expansion of NBS-LRR genes may be regarded as a plant-specific adaptation to extracellular signal perception [42, 43], such as the ability to recognize various PAMPs in Arabidopsis [44], given the abundance of pathogens in Vernicia [3, 45]. Further research revealed that five NBS-LRR gene pairs in V. montana (Fig. 2b and Additional file 6) likely originated from a duplication event, whereas no duplication events were detected in the VfNBS-LRR genes. These results indicate that species differentiation occurred after the divergence of V. montana and V. fordii from a common ancestor, and then the gene loss events occurred in VfNBS-LRRs of V. fordii during the long-term evolution process, which finally led to the V. fordii containing fewer members of the NBS-LRR family than the V. montana. These findings may offer an explanation for the resistance of V. montana to Fusarium wilt, while V. fordii remains susceptible to the disease.

In this study, we calculated the Ka/Ks values of all NBS-LRR gene pairs to investigate the evolutionary constraints acting on these genes. The results, as presented in Additional file 7, suggest that the Ka/Ks values of NBS-LRR gene pairs in V. montana were less than one, indicating a purifying selection on these gene pairs in V. montana. The number of duplicated pairs varied between V. montana and V. fordii, with the former exhibiting strong purifying selection and showing slow protein-level evolution. Subsequently, we compared orthologous gene pairs between V. montana (resistant to Fusarium wilt) and V. fordii (susceptible to Fusarium wilt) to identify disease resistance-related loci. The ratio of Ka to Ks was calculated for each NBS-LRR gene pair to determine the selective pressure on them (Additional file 8 and Fig. 2c). The peaks at Ka/Ks ratio of 0.3–0.5 and 0.6–0.7 (Fig. 2d) were observed indicating purifying selection between V. montana and V. fordii (Ka/Ks < 1). Previous studies on cultivated and wild Vigna angularis have also shown similar patterns, with peaks at Ka/Ks ratios of 0.4–0.6 and 0.6–0.9, suggesting the presence of novel disease resistance alleles in wild V. angularis that differ from those in the cultivated V. angularis, such as Vang02g14420, Vang0229s00140, Vang0291s00070, and Vang03g15160 [46].

Expression patterns of differentially expressed NBS-LRRs after infection of Fusarium wilt

The NBS-LRR family represents a widespread and ancient group of disease-resistance genes that play key roles in safeguarding plants against various pathogens [47]. The number of NBS-LRRs differs significantly between V. montana and V. fordii, with 149 and 90, respectively. Furthermore, only 33.8% of the NBS-LRRs were identified as orthologs between V. montana and V. fordii. Despite the similarity in genomic features between the two species [2, 29], this discrepancy in NBS-LRR counts suggests that they might have undergone gene gains or losses to adapt to distinct environments [30, 48]. This divergence might also explain the contrasting susceptibility of V. fordii to Fusarium wilt and the resistance of V. montana to this disease [3, 4, 49].

To determine the mechanism of NBS-LRRs in tung tree infected with Fusarium wilt, we further detected the expression levels of these genes during various infection stages in V. fordii and V. montana [45]. Among 43 orthologous NBS-LRR gene pairs in these two Vernicia species, five gene pairs were not detectable at each stage response to wilt disease. Specifically, we further focused on the expression patterns of the remaining 38 orthologous NBS-LRR gene pairs in response to wilt disease. Figure 3 displays the results, revealing divergent expression patterns among most NBS-LRR genes. For example, genes such as Vf09G1722, Vf02G0644, and Vf02G0653 were highly expressed in the roots of V. fordii uninfected and infected with Fusarium wilt, while their corresponding orthologous genes exhibited either low expression or were not expressed at all in V. montana. Notably, we found that one orthologous gene pair Vf11G0978-Vm019719 exhibited consistent upregulated expression in V. montana, in contrast to persistent downregulation in V. fordii. These results suggest that this gene pair may be responsible for the resistance to Fusarium wilt in V. montana and the susceptibility to Fusarium wilt in V. fordii.

Fig. 3
figure 3

Expression of one-to-one orthologous NBS-LRR genes in root tissue of V. fordii and V. montana under Fusarium wilt infection. The RNA-seq data obtained from Chen et al. [45]. The heat map depicts expression profiles of one-to-one orthologous NBS-LRR genes in V. fordii (left) and V. montana (right) in response to Fusarium wilt during various infection stages: 0, uninfected stage; 1, 2 days after Fusarium wilt infection (dpi); 2, 8 dpi. F0–F2 indicated the expression of VfNBS-LRRs in V. fordii during various infection stages (0, 1, 2) by the pathogen Fusarium wilt; M0-M2 indicated the expression of VmNBS-LRRs in V. montana during the infection stage (0, 1, 2) by the pathogen Fusarium wilt

Full size image

Vm019719 positively regulates the resistance to Fusarium wilt in V. montana

Pathogen attack by Fusarium wilt induced the expression of Vm019719 in V. montana but cannot induce the expression of its allele Vf11G0978 in V. fordii (Fig. 3). Amino acid sequence alignment revealed that the sequences were basically identical to the allele Vm019719, except for a non-functional region (i.e., excluding NBS, TIR, CC and LRR domains) at the N-terminal insertion end of Vf11G0978. The primers used in the experiment are listed in Additional file 9. To further investigate the role of Vm019719 in disease resistance, we employed the VIGS method [50] to downregulate its expression in V. montana. The qRT-PCR analysis confirmed the successful downregulation of Vm019719 expression in the VIGS plants, while the expression remained unchanged in the control plants (Fig. 4a, b). To verify whether the paralogs of Vm019719 were cross-silenced, we selected three genes (VmNBS-LRR1: Vm016833, VmNBS-LRR2: Vm012761, and VmNBS-LRR3: Vm026869) that had the closest expression pattern to Vm019719 for evaluation (Fig. 4b). The expression of these selected paralogs was unaffected in Vm019719 VIGS plants, as revealed by qRT-PCR, indicating the specific downregulation of Vm019719 by VIGS. To assess the potential attenuation of disease resistance V. montana RNAi Vm019719, three transgenic lines, RNAi1, RNAi2, and RNAi3, and control plants were selected and placed under Fusarium wilt challenge (Fig. 4c). The RNAi plants displayed typical symptoms of Fusarium wilt, including necrosis, chlorosis, and wilting, whereas the control V. montana plants exhibited significant resistance to the disease with only minor symptoms.

Fig. 4
figure 4

Silencing of Vm019719 reduces Fusarium wilt resistance in V. montana. a Diagram of the VIGS technique used for infecting the V. montana leaves. b The silencing efficiency of the Vm019719 using qRT-PCR. Standard deviations (SDs) were calculated from three biological replicates. c Phenotypic identification of Fusarium wilt resistance in V. montana after silencing of Vm019719

Full size image

WRKY conferred disease resistance to Fusarium wilt in tung tree

Since the CDSs of Vm019719 and Vf11G0978 were basically identical, we cloned their promoter sequences from the genomes of V. montana and V. fordii, respectively. Subsequently, we constructed AD bait vectors respectively and identified their upstream regulatory transcription factors from V. fordii and V. montana yeast libraries. Interestingly, we identified a potential regulator, WRKY64 (Vm005641), from the V. montana library, but no corresponding transcription factor was screened in another yeast library. Promoter sequence analysis showed that the promoter of Vm019719 (ProVm019719) had two conserved WRKY binding W-box elements (C/T)TGAC(T/C) [51,52,53], while ProVf11G0978 had no corresponding elements (Fig. 5a). WRKY64, an ortholog of the Arabidopsis transcription factor WRKY70 [54], has been reported to positively regulate defense against Fusarium infection [53, 54]. To confirm the direct binding of WRKY64 to the Vm019719 promoter and its impact on the transactivation of Vm019719, we performed a yeast single-hybrid experiment. The results confirmed that WRKY64 can indeed bind to the promoter of Vm019719 to activate its expression (Fig. 5b). Furthermore, we overexpressed the Vm019719 using its own promoter, which led to enhanced resistance against Fusarium wilt in V. fordii. As depicted in Fig. 5c, the control plants exhibited typical symptoms of Fusarium wilt infection, including necrosis, chlorosis, and wilting, while the overexpression V. fordii plants were significantly resistant to this Fusarium wilt disease by showing only minor symptoms. Taken together, these results demonstrate that WRKY64 can indeed combine with the promoter of Vm019719 to regulate its expression, thereby providing tung trees with the capacity to resist Fusarium wilt.

Fig. 5
figure 5

Overexpression of Vm019719 driven by Vm019719 promoter enhances Fusarium wilt resistance in V. fordii. a Promoter cis-acting element analysis of Vf11G0978 and Vm019719. b WRKY64 directly bound to the Vm019719 promoter. c Overexpression of Vm019719 driven by Vm019719 promoter enhances Fusarium wilt resistance in V. fordii

Full size image

A proposed mechanism for V. montana’s resistance to Fusarium wilt

Through our examination of the genomic characteristics of the Fusarium wilt-susceptible V. fordii and Fusarium wilt-resistant V. montana, we successfully identified the presence of NBS-LRR genes. Among these genes, the allele pair Vm019719-Vf11G0978 stood out as potential key players in conferring resistance to Fusarium wilt in tung tree. Notably, Vm019719-Vf11G0978 displayed downregulated expression in V. fordii but showed upregulated expression in V. montana. Extending our analysis, we conducted various experiments including cis-acting element analysis, VIGS, overexpression studies, and yeast one-hybrid experiments. This comprehensive investigation led us to unveil a novel disease defense mechanism, wherein the transcription factor WRKY64 binds to the W-box elements to effectively defend against Fusarium wilt infection (as depicted in Fig. 6).

Fig. 6
figure 6

The mechanism of V. montana significantly resistant to Fusarium wilt. Upon Fusarium infection, the expression level of Vm019719 rapidly increases in V. montana, which is driven by WRKY64

Full size image

Conclusions

In this study, we identified NBS-LRRs in two tung tree genomes: Fusarium wilt-susceptible V. fordii and Fusarium wilt-resistant V. montana. Interestingly, we observed a preferential loss of LRR domains in V. fordii compared to V. montana. The expression patterns of NBS-LRRs indicated their potential contributions to disease resistance in tung tree. Notably, the orthologous gene pair Vf11G0978-Vm019719 displayed consistent downregulated expression in V. fordii while being consistently upregulated in V. montana, which correlated with the resistance of V. montana to Fusarium wilt. We confirmed that VmWRKY64 can bind to the W-box promoter element of Vm019719, enabling V. montana to resist Fusarium wilt. Remarkably, the deletion of the W-box elements in the promoter of the Vf11G0978 allele prevented its binding to WRKY64, resulting in the loss of resistance to Fusarium wilt in V. fordii. These findings not only provide insights into the roles of NBS-LRRs in tung tree and their response to Fusarium wilt infection but also shed light on the genetic mechanisms underlying resistance to Fusarium wilt and the potential application of marker-assisted breeding in tung tree.

Methods

Identification of NBS-LRR genes

For identification of NBS-LRRs, the proteins and coding sequences of V. montana and V. fordii were obtained from the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov/), as published by Zhang et al. (2019) and Cui et al. (2018) [2, 29]. The hidden Markov models (HMM) for the TIR (PF01582) and NBS (PF00931) domains were obtained from the Pfam (https://pfam.xfam.org/). The NBS-containing sequences were identified in V. montana and V. fordii using NBS domain by employing HMMER (version 3.0) [55] with an E-value of 1e–3, and then these genes were extracted according to their sequencing ID. Among these sequences, the same method was used to determine the TIR-containing sequences. The different types of HMMs of the LRR domain, including LRR1-9, LRV, LRR_adjacent, LRR19-TM, LRRC37, LRRC37AB_C, LRRCT, LRRFIP, LRRNT, and LRRNT_2, were obtained from Pfam and then scanned these LRR domains in NBS-containing sequences in V. montana and V. fordii. The Paircoil2 (http://cb.csail.mit.edu/cb/paircoil2/paircoil2.html) was used to find the coiled-coil (CC) domains in NBS-containing sequences with a P-score cutoff 0.03 [56]. TBtools (version 2.029) [57] was used to determine the chromosomal locations of NBS-LRRs in the genomes of both V. montana and V. fordii. TBtools (version 2.029) [57] was further utilized to compare the chromosomal locations of NBS-LRRs between V. montana and V. fordii.

Homology in Vernicia genomes

The paralogs of NBS-LRRs in V. montana and V. fordii, as well as the orthologs of NBS-LRRs between V. montana and V. fordii, were identified using local BLAST analyses with specific evaluation criteria. These criteria included an E-value threshold of ≤ 10–10, identity > 80%, and alignment coverage > 80% [8]. The software TBtools (version 2.029) [57] was utilized to extract the paralogs and orthologs of NBS-LRRs pairs, and we then further used this software to calculate the non-synonymous to synonymous per site substitution rates (Ka/Ks), Ks, and Ka values. A Ka/Ks value less than 1 denotes purifying selection, equal to 1 indicates neutral selection, and greater than 1 suggests positive selection [58].

Gene expression analysis

The RNA-seq data with accession numbers PRJNA445068, PRJNA483508, and PRJNA318350 were obtained from NCBI for the purpose of acquiring the expression levels of NBS-LRRs [45]. The RNA-seq data underwent a thorough quality assessment using FastQC (version 0.11.7). Subsequently, low-quality reads and bases were meticulously removed with Trimmomatic (version 0.39) [59]. Finally, the processed reads were aligned using HISAT2 (version 2.2.1) [60] with default parameters, and fragments per kilobase million (FPKM) values were calculated using StringTie (version 2.1.7) [61] with default parameters, following the methodology described by Li et al. [49].

Yeast one-hybrid (Y1H) assays

To confirm the interactions between the W-box element in the promoter of Vm019719 and WRKY64, we performed Y1H assays using the Matchmaker® Gold Yeast One-Hybrid System (Clontech; TaKaRa). The full-length coding sequence of WRKY64 was cloned and inserted into the pGADT7. The pAbAi vector underwent modifications to incorporate the sequences of the Vm019719 promoter. The resulting constructs were linearized using BstBI (New England Biolabs, Beverly, MA, USA) digestion and subsequently transformed into the Y1HGold yeast strain. To determine the minimum inhibitory concentration of Aureobasidin A (AbA) for the pAbAi bait constructions, the transformed yeast strains were cultured on SD medium without uracil. Finally, the pGADT7-WRKY64 construct was introduced into the transformed Y1H yeast strains containing the promoter regions. The interactions between the pAbAi bait constructs and pGADT7-WRKY64 were assessed based on the growth of positive yeast cells on SD/-Ura/AbA medium.

Agrobacterium-mediated VIGS

In accordance with the methodology described by Cao et al. [50], a similar approach was adopted for virus-induced silencing (VIGS) experiments. Gene-specific fragment primers for WRKY64, Vf11G0978, and Vm019719 were designed using Primer-BLAST and utilized for cloning and introduction into a pTRV2 vector. Three recombinant plasmids (pTRV2-WRKY64, pTRV2-Vf11G0978, and pTRV2-Vm019719) and control (an empty pTRV2 plasmid) were individually transformed into Agrobacterium tumefaciens GV3101. The Agrobacterium cells containing pTRV1 or the recombinant plasmids (or pTRV2) were prepared and cultured as described by Cao et al. [50]. The suspensions of the recombinant plasmids (or pTRV2) or pTRV1-containing cells were mixed thoroughly and then inoculated onto the leaves of tung tree plants. Each experiment was carried out at least three times before using the inoculated tung tree plants for functional evaluations after 2 weeks.

Real-time quantitative PCR (qRT-PCR) assays

Total RNA was extracted from tung tree leaves using the TRIzol reagent (TaKaRa, Tokyo, Japan). The cDNA synthesis was carried out utilizing the PrimerScript™ RT reagent Kit with gDNA Eraser (Takara, Dalian, China). For expression analysis, a qRT-PCR assay was performed in a total reaction volume of 20 μl, which included 2 μl of cDNA template, 10 μl of HieffTM qPCR SYBR® Green Master Mix (Yeasen, Shanghai, China), and 0.2 μl of each primer. The relative gene expression levels were determined using the cycle threshold (Ct) 2−ΔΔCt method. This study included three biological replicates for each sample.