BxCDP1 from the pine wood nematode Bursaphelenchus xylophilus is recognized as a novel molecular pattern

Abstract The migratory plant‐parasitic nematode Bursaphelenchus xylophilus is the causal agent of pine wilt disease, which causes serious damage to pine forests in China. Plant immunity plays an important role in plant resistance to multiple pathogens. Activation of the plant immune system is generally determined by immune receptors, including plant pattern recognition receptors, which mediate pattern recognition. However, little is known about molecular pattern recognition in the interaction between pines and B. xylophilus. Based on the B. xylophilus transcriptome at the early stages of infection and Agrobacterium tumefaciens‐mediated transient expression and infiltration of recombinant proteins produced by Pichia pastoris in many plant species, a novel molecular pattern (BxCDP1) was characterized in B. xylophilus. We found that BxCDP1 was highly up‐regulated at the early infection stages of B. xylophilus, and was similar to a protein in Pararhizobium haloflavum. BxCDP1 triggered cell death in Nicotiana benthamiana when secreted into the apoplast, and this effect was dependent on brassinosteroid‐insensitive 1‐associated kinase 1, but independent of suppressor of BIR1‐1. BxCDP1 also exhibited cell death‐inducing activity in pine, Arabidopsis, tomato, pepper, and lettuce. BxCDP1 triggered reactive oxygen species production and the expression of PAMP‐triggered immunity marker genes (NbAcre31, NbPTI5, and NbCyp71D20) in N. benthamiana. It also induced the expression of pathogenesis‐related genes (PtPR‐3, PtPR‐4, and PtPR‐5) in Pinus thunbergii. These results suggest that as a new B. xylophilus molecular pattern, BxCDP1 can not only be recognized by many plant species, but also triggers innate immunity in N. benthamiana and defence responses of P. thunbergii.


| INTRODUC TI ON
The pine wood nematode (PWN) Bursaphelenchus xylophilus is an extremely damaging migratory plant-parasitic nematode that infects most Pinus species, including P. densiflora, P. massoniana, P. thunbergii, and P. pinaster, resulting in massive economic losses in eastern Asia and Europe, especially in China and Japan (Hunt, 2009;Vicente et al., 2011;Jones et al., 2013;Zhou et al., 2017;Li et al., 2018b;Huang et al., 2019). Pine wilt disease (PWD) caused by PWN is now considered the most serious threat to pine forests worldwide and is very difficult to control.
B. xylophilus exhibits phytophagous and mycophagous stages during feeding. At the phytophagous stage, the nematode migrates to the xylem resin and ray canals and feeds on parenchyma cells, leading to the death of these cells (Mamiya, 2012). As part of a strong defence response during the early stages of infection, the tree releases reactive oxygen species (ROS), polyphenolic compounds, terpenoids, and lipid peroxides (Fukuda, 1997). As the tree dies, the nematode switches to the mycophagous stage and feeds on the fungi that colonize the tree (Jones et al., 2008).
Similar to most pathogens and parasites, B. xylophilus must overcome plant immunity to achieve successful host colonization.
In addition to PAMPs and effectors that can trigger cell death and plant immunity, various cytolytic toxins produced by phytopathogenic microorganisms have also been shown to trigger plant immunity-associated responses (van't Slot et al., 2002). These toxins include Fusarium spp.-derived fumonisin B1, Phomopsis amygdali-derived fusicoccin or Cochliobolus victoriae-derived victorin (van't Slot et al., 2002;Qutob et al., 2006). Toxin-mediated host immune activation is likely to be the result of toxin-mediated interference with host integrity (Ottmann et al., 2009). For example, various necrosis and ethylene-inducing peptide 1 (Nep1)-like proteins (NLPs) produced by bacterial, oomycete, and fungal microbes are cytolytic toxins that trigger plant immunity-associated defences by toxin-induced host cell damage (plasma membrane destruction and cytolysis) (Ottmann et al., 2009). A subsequent study further found that NLPs triggered plant immunity-associated defences by two different mechanistic modes (i.e., toxin action and a classical PAMP motif [nlp20]) (Böhm et al., 2014).
In the present study, we selected 15 candidate effectors that were up-regulated in the B. xylophilus transcriptome at the early stages of infection (Hu et al., 2019). We then carried out transient expression in Nicotiana benthamiana using a potato virus X (PVX) expression vector and identified a B. xylophilus transcript (BxCDP1) that was highly abundant at the early stages of pine infection and that encoded a protein that triggered cell death in N. benthamiana.
BxCDP1 was also able to induce cell death in pine and many other plants. Moreover, BxCDP1-triggered cell death was dependent on NbBAK1 but did not require NbSOBIR1 and induced immune responses in N. benthamiana. BxCDP1 could also induce a defensive reaction in P. thunbergii. These data suggest that BxCDP1 may act as a novel molecular pattern in B. xylophilus. Furthermore, PhCDP1 from Pararhizobium haloflavum, a protein similar to BxCDP1, also induced cell death in N. benthamiana.

| BxCDP1 induces cell death in N. benthamiana
Among the 247 up-regulated genes from the B. xylophilus transcriptome (accession number: PRJNA397001) at the early stages of infection, we screened 69 candidate effectors (Hu et al., 2019). Fifteen of the 69 candidate effectors with signal peptides were cloned separately into the binary PVX vector pGR107, which adds a C-terminal 3 × hemagglutinin (HA) tag to the protein.  reading frame (ORF) is 708 bp and encodes a 236-amino acid polypeptide that contains three cysteine residues. We carried out further studies on this protein.
To determine whether apoplastic localization is required for BxCDP1-triggered cell death, the signal peptide from PR protein 1 (PR1) of Nicotiana tabacum (Table S2) was added to the N-terminus of BxCDP1nsp to produce PR1 SP-BxCDP1nsp (Figure 1b). Transient assays showed that BxCDP1 and PR1 SP-BxCDP1nsp, but not BxCDP1nsp, triggered cell death in N. benthamiana (Figure 1c), and the expression of these proteins was validated by western blot analysis ( Figure 1d). Ion leakage measurements were then employed to quantify the degree of cell death triggered by the five tested proteins. The electrolyte leakage induced by INF1, BxCDP1, and PR1 SP-BxCDP1nsp was significantly higher than that induced by F I G U R E 1 BxCDP1 triggers cell death in Nicotiana benthamiana. (a) Representative N. benthamiana leaves at 7 days after inoculation with Agrobacterium tumefaciens GV3101 carrying the BxCDP1 gene in the vector pGR107. Representative N. benthamiana leaves were also placed under UV illumination to observe cell death. The infiltration assay was performed three times, and three different plants with three inoculated leaves were used in each assay. (b) Regions of BxCDP1 examined for cell death activity. (c) Representative N. benthamiana leaves at 7 days after agroinfiltration carrying BxCDP1nsp (no signal peptide) and PR1 SP-BxCDP1nsp. The infiltration assay was performed three times, and three different plants with three inoculated leaves were used in each assay. (d) Immunoblot analysis of proteins from N. benthamiana leaves transiently expressing target proteins fused with 3 × HA tags. The experiment was repeated three times with similar results. (e) Quantification of cell death by measuring electrolyte leakage in N. benthamiana leaves at 7 days post-infiltration with constructs encoding the indicated proteins. GFP, green fluorescent protein. The experiment was repeated three times with similar results. The data are the means, and the error bars represent ± SD from three independent experiments. Different letters on top of the bars indicate statistically significant differences (p < .05, t test), as measured by Duncan's multiple range test BxCDP1nsp or GFP ( Figure 1e). These data indicate that BxCDP1 triggers cell death in N. benthamiana when secreted into the apoplast.

| Pararhizobium haloflavum contains a similar protein to BxCDP1 that also induces cell death in N. benthamiana
Querying the BxCDP1 amino acid sequence against the National Center for Biotechnology Information (NCBI) protein database resulted in the identification of only one similar protein, which was present in P. haloflavum (denoted as PhCDP1), recently identified as a novel species of Pararhizobium (Shen et al., 2018). The nucleotide and amino acid sequences of BxCDP1 share 87% and 77% identity, respectively, with that of PhCDP1 ( Figure S1a,b). In order to check that the B. xylophilus gene was of nematode origin, it was cloned from genomic DNA of B. xylophilus ( Figure S4a,b). In addition, the sequence was identified in the genome assembly for B. xylophilus and was also present in a sequence present in a transcriptome analysis of this species (Kikuchi et al., 2011;Tsai et al., 2016).
To determine whether PhCDP1 also induces cell death in N. benthamiana, PhCDP1 was cloned separately into the pBINRFP vector, which adds a C-terminal red fluorescent protein (RFP) tag to the protein. Then, the protein was transiently expressed in N. benthamiana.
The results show that PhCDP1 also triggered cell death in N. benthamiana ( Figure S1c). The expression of PhCDP1 was validated by western blot analysis ( Figure S1d).

| BxCDP1 induces cell death in N. benthamiana, Arabidopsis, tomato, pepper, and lettuce
To further confirm that the BxCDP1 protein is able to induce cell death in N. benthamiana, BxCDP1 was produced in the yeast Pichia pastoris using the pPICZαA vector (pPICZαA:BxCDP1), which contains a 6 × His tag and targets the protein for secretion into the culture medium. As described in previous studies (Ma et al., 2015;Wang et al., 2018), P. pastoris culture supernatant from the pPICZαA empty vector (EV) control strain was used as the control in this assay and was purified in the same way as BxCDP1. After recovery from the culture supernatant using Ni-NTA resin, the recombinant protein BxCDP1rec and EV were detected by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and western blot analysis, respectively. The results showed that BxCDP1 was successfully induced and purified ( Figure S2).
BxCDP1rec was tested for cell death activity by infiltrating 100 pM to 3 µM protein solution into the mesophyll of N. benthamiana leaves. EV and bovine serum albumin (BSA) were used as controls. In this system, BxCDP1rec induced cell death at 3 days after infiltration ( Figure 2). Moreover, cell death was enhanced with increasing concentrations of BxCDP1 from 100 pM to 3 µM.
To examine the host specificity of BxCDP1, we infiltrated the purified protein into expanded leaves of various plant species. EV was used as the control. BxCDP1 at 300 nM induced cell death in

| NbBAK1, but not NbSOBIR1, is required for BxCDP1-triggered cell death in N. benthamiana
As coreceptors for different PRRs, BAK1 and SOBIR1 partici- Furthermore, the degree of cell death triggered by these three proteins was measured, with BxCDP1-triggered cell death being significantly reduced in NbBAK1-silenced plants (Figure 4b).
The silencing constructs we used (TRV:NbBAK1 and TRV:NbSOBIR1/NbSOBIR1-like) were the same as previously used in similar studies (Liebrand et al, 2013;Wang et al., 2018), which showed that NbBAK1 and its homologs NbBAK1a and NbBAK1b, and NbSOBIR1 and its homolog NbSOBIR1-like were all silenced (Liebrand et al, 2013;Wang et al., 2018). The silenced parts of the receptors (NbBAK1 and NbSOBIR1) are shown in Table S3. The nucleic acid sequences of NbBAK1a and NbBAK1b are too similar to design specific primers to detect their individual expression. Thus, we detected the relative expression of NbBAK1, NbSOBIR1, and NbSOBIR1-like by quantitative reverse transcription PCR (RT-qPCR).
The results confirmed that NbBAK1, NbSOBIR1, and NbSOBIR1like were all markedly silenced by the VIGS constructs ( Figure 4c).
According to the principle of VIGS, the NbBAK1 dsRNA can be degraded by Dicer enzyme into siRNA of 21-23 nucleotides. Thus, the TRV:NbBAK1 could also silence NbBAK1a and NbBAK1b. The nucleic acid sequences of NbBAK1a, NbBAK1b, and NbSOBIR1like are provided in Table S4. Moreover, western blotting verified that all assayed proteins were expressed in the silenced N. benthamiana plants (Figure 4d). Taken together, these results suggest that NbBAK1 participated in BxCDP1-triggered cell death but that NbSOBIR1 did not.

| BxCDP1 induces immune responses in N. benthamiana
To further determine whether BxCDP1 induces cell death, the purified BxCDP1 protein (300 nM) and EV were infiltrated into GFP-, NbBAK1-, and NbSOBIR1-silenced plants. At 3 days after infiltration, the purified BxCDP1 protein triggered cell death in the GFP-

| BxCDP1 is highly induced at early infection stages
According to our previous transcriptome data for B. xylophilus, BxCDP1 is up-regulated at the early infection stages (Hu et al., 2019). Thus, we detected the relative expression of PR genes (PtPR-3, PtPR-4, and PtPR-5) at 6 hr post-inoculation by RT-qPCR and found that PtPR-3, PtPR-4, and PtPR-5 were significantly up-regulated in P. thunbergii compared with the expression levels in EV-treated samples (Figure 7a).
Approximately 2,000 nematodes (a mixture of juveniles and adults) were inoculated into P. thunbergii seedlings, which were inoculated with purified BxCDP1 protein and EV for 4 hr. The results show that the early symptoms occurred later and the degree of morbidity of P. thunbergii seedlings inoculated with the former were significantly lower than those of seedlings inoculated with the latter (Figure 7b-e) at 15 days post-inoculation. A number of studies have demonstrated that most known PRRs require BAK1 for their function (Heese et al., 2007). In this study, BxCDP1-triggered cell death was dependent on NbBAK1. Thus, the result suggests that BxCDP1 is not a general toxin and its detec- Arabidopsis, tomato, pepper, and lettuce. The results indicate that the cell death that occurred in these plants was induced through the same signalling pathway. Taken together, these data indicate that BxCDP1 may be a PAMP.
Typically, PAMPs are highly conserved within a class of microbes.
For example, homologs of the oomycete PAMP PsXEG1 are widely present in prokaryotic and eukaryotic microorganisms, especially in plant-associated microbes (oomycetes, fungi, and bacteria) (Ma et al., 2015). In addition, homologs of the R. commune PAMP RcCDI1 are distributed widely across ascomycetes (Franco-Orozco et al., 2017). However, there are also some PAMPs that are less conserved (Brunner et al., 2002;Lee et al., 2009). For example, Pep-13, a surface-exposed fragment of a calcium-dependent cell wall transglutaminase (TGase) that is conserved among Phytophthora species, can activate various plant defences (Brunner et al., 2002). In this study,

F I G U R E 6
The expression pattern of BxCDP1 at the early stages of Bursaphelenchus xylophilus infection by RT-qPCR analysis. Relative expression of BxCDP1 at the mycetophagous stage (0 hr) and at the early stages of infection (6, 12, and 24 hr). Data are the means, and the error bars represent ± SD from three biological replicates. Different letters on top of the bars indicate statistically significant differences (p < .05, t test) as measured by Duncan's multiple range test F I G U R E 7 BxCDP1 induces pathogenesis-related marker genes expression of Pinus thunbergii. (a) Relative transcript levels of pathogenesis-related genes in P. thunbergii infected with the purified BxCDP1 protein and empty vector (EV). Stems c.2 cm in length were used for RNA extraction at 6 hr post-inoculation. The inoculation assay was repeated three times, and in each assay three different seedlings for each treatment were used. Data are the means, and the error bars represent ± SD from three independent experiments. Different letters on top of the bars indicate statistically significant differences (p < .05, t test) as measured by Duncan's multiple range test. (b) and (c) Representative photographs of P. thunbergii at 1 day post-inoculation. (d) and (e) Representative photographs of P. thunbergii at 15 days postinoculation. The infection assay was repeated three times we attempted to match the BxCDP1 amino acid sequence with sequences in the NCBI protein database using BLASTP to identify homologs of BxCDP1, but only one similar protein, denoted as PhCDP1, was identified, which is from a bacterium (P. haloflavum). It showed that the nucleotide and amino acid sequences of BxCDP1 shared 87% and 77% identity, respectively, with the sequences of PhCDP1.
Moreover, the nucleotide sequence of BxCDP1 was queried against genomes of all other nematodes in the WormBase ParaSite database and the transcriptome of B. mucronatus in NCBI, yet no homolog was retrieved. We undertook extensive analysis to confirm that this gene is derived from the nematode rather than from a bacterial contaminant. These analyses indicate that BxCDP1 is only present in B. xylophilus but not in other nematodes and is most similar to a bacterial sequence.
In this study, BxCDP1 induced the expression of some PR genes  in P. thunbergii. Moreover, compared with the control, the onset of purified BxCDP1 protein in pines was delayed after inoculation with B. xylophilus. This finding indicates that BxCDP1 might trigger the defence response of the pines. We speculate that the enhancement of the defence response of the pines might further prevent B. xylophilus infection to some degree such that the onset of pine wilt was delayed.
In conclusion, BxCDP1 is a new molecular pattern that is recognized by the host pine. This study represents an important step toward understanding the interaction mechanisms between B. xylophilus and pine.

| Nematodes and plant materials
A highly virulent strain of B. xylophilus, AMA3, was transferred into a mycelial mat of Botrytis cinerea growing on potato dextrose agar (PDA) plates and cultured at 25 °C for 7 days. The nematodes were extracted using the Baermann funnel technique (Qiu et al., 2013). A. thaliana, S. lycopersicum, C. annuum, and L. sativa var. romana were grown in a glasshouse at 25 °C with a relative humidity of 60% under 16:8-hr light:dark conditions. P. thunbergii

| RNA isolation and cDNA synthesis
A suspension of approximately 10,000 mixed-life-stage nematodes was collected from PDA plates and inoculated onto P. thunbergii (3-year-old) seedlings for 6, 12, and 24 hr. The whole seedlings were cut into approximately 2-cm segments of stems, which were split into halves. The nematodes were extracted from these segments of stems using a Baermann funnel. Total RNA from the nematodes was extracted using TRIzol reagent (Invitrogen). Leaves of N. benthamiana and stems of P. thunbergii were sampled and frozen in liquid nitrogen. Total RNA was extracted using the Plant Total RNA Kit (Zoman). First-strand cDNA for RT-qPCR was synthesized from 1 μg of total RNA using HiScript II Q RT SuperMix for qPCR (+gDNA wiper) (Vazyme) according to the manufacturer's protocol.

| Plasmid constructs
The transcriptome of B. xylophilus at the early stages of infection (6, 12, and 24 hr post-inoculation) was obtained in our previous study, with 69 candidate effectors (Hu et al., 2019). Fifteen randomly selected candidate effectors with highly abundant expression were cloned using B. xylophilus cDNA and matched against the SWISS-PROT database using BLAST (Table S1). The gene encoding PhCDP1 was synthetized (GenScript). These fragments of the 15 candidate effectors were ligated into PVX (pGR107-3 × HA), and BxCDP1 and PhCDP1 were ligated into pBINRFP (pCAM1300-RFP) using the F I G U R E 8 BxCDP1 induces cell death in pine. (a)-(c) Representative photographs of the cell morphology of Pinus thunbergii treated with empty vector (EV), as assessed by transmission electron microscopy (TEM). (d)-(f) Representative photographs of the cell morphology of P. thunbergii treated with the purified BxCDP1 protein, as assessed by TEM Clone Express II One Step Cloning Kit (Vazyme). Individual colonies for each construct were examined by PCR for insertions, and the selected clones were verified by sequencing. The primers used are listed in Table S1.

| Expression and purification of the recombinant BxCDP1 protein
The ORF for BxCDP1 was amplified from B. xylophilus cDNA using primers with vector-specific extensions (Table S1). The purified PCR product was inserted into the linearized vector pPICZαA (Invitrogen), which contains a 6 × His tag. The construct was transformed into Escherichia coli TOP10 competent cells. Individual colonies from the construct were tested by PCR for insertions, and the selected clones were verified by sequencing. The constructed plasmid was linearized with SacI (New England Biolabs). The pPICZαA vector containing BxCDP1-6 × His and EV were transformed into the P. pastoris KM71H (Invitrogen) via electroporation. Positive clones were grown in yeast extract-peptone-dextrose (YPD) medium containing 100 µg/ml zeocin at 30 °C for 3 days. BMGY (buffered glycerolcomplex medium) and BMMY (buffered methanol-complex medium) were used for protein expression. After 4 days, the cultures were centrifuged at 6,500 rpm for 10 min to procure the supernatant containing BxCDP1-6 × His and EV, respectively, as confirmed by SDS-PAGE. Purification of the recombinant protein BxCDP1-6 × His and EV from the culture supernatant was performed by affinity chromatography using Ni-NTA Superflow resin (Qiagen).

| A. tumefaciens and protein infiltration assays
Infiltration assays were performed according to a previous report (Hu et al., 2019). In brief, constructs were transformed into A. tumefaciens GV3101 by electroporation, and the cells were grown on Luria Bertani agar plates with kanamycin and rifampicin. For agroinfiltration assays, recombinant A. tumefaciens strains were grown at 30 °C in a shaking incubator at a rotation speed of 200 rpm for 12 hr.  (Heese et al., 2007). GFP was used as the negative control.
The purified BxCDP1 protein was diluted in phosphate-buffered saline (PBS, pH 7.2). To assess the induction of cell death by the recombinant protein produced in P. pastoris, solutions of 100 pM to 3 µM purified BxCDP1 protein were infiltrated into N. benthamiana leaves; 100 pM to 3 µM purified BSA protein and EV solution were used as negative and blank controls, respectively. The purified BxCDP1 protein (300 nM) was also infiltrated into the leaves of A. thaliana ecotype Columbia, tomato (S. lycopersicum), and pepper (C. annuum var. CM334). The purified BxCDP1 protein (1 µM) was also infiltrated into the leaves of lettuce (L. sativa var. romana) and wheat (T. aestivum). We photographed N. benthamiana leaves at 3 days post-infiltration, A. thaliana, and S. lycopersicum leaves at 5 days post-infiltration, and C. annuum, L. sativa, and T. aestivum leaves at 7 days post-infiltration. The infiltration experiment was performed three times, and three different plants with three inoculated leaves were used for each assay.

| Electrolyte leakage assay
Ion leakage from leaf discs of N. benthamiana was measured as described previously (Yu et al., 2012). Five N. benthamiana leaf discs (9 mm diameter) were collected 7 days after agroinfiltration and floated on 5 ml deionized water for 3 hr with continuous shaking (100 rpm) at room temperature. The initial and final electrolyte leakage values after 30 min of boiling were measured using a conductivity meter (S470 Seven Excellence; Mettler Toledo). Relative electrolyte leakage was calculated by comparing the initial and final values. The experiment was performed three times.

| VIGS for NbBAK1 and NbSOBIR1 in N. benthamiana
VIGS experiments were conducted using A. tumefaciens GV3101 harbouring the pTRV1 vector and pTRV2:NbBAK1, pTRV2:NbSOBIR1/ NbSOBIR1-like, and pTRV2:GFP, as described in previous studies (Wang et al., 2018). At 3-4 weeks post-infiltration with VIGS constructs, N. benthamiana was infiltrated with PVX:BxCDP1, PVX:INF1, and PVX:NPP1. The oomycete PAMP INF1 and effector NPP1 were used as negative and positive controls, respectively. Silenced plants were also infiltrated with 300 nM purified BxCDP1 protein, with EV used as the control. The N. benthamiana phenotypes were scored at 5 days post-infiltration. VIGS effectiveness was assessed according to the phenotype of PDS, as described previously (Liu et al., 2002).
The infiltration experiment was performed three times, and three different plants with three inoculated leaves were used for each assay.

| Sequence analysis
Sequences similar to BxCDP1 were retrieved by querying the BxCDP1 protein against the NCBI protein database using BLASTP (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The signal peptide and transmembrane helices were predicted using the SignalP v. 5.0 server (http://www.cbs.dtu.dk/servi ces/Signa lP/) and TMHMM v. 2.0 server (www.cbs.dtu.dk/servi ces/TMHMM /), respectively. BxCDP1 domains were analysed using SMART (http://smart.embl-heide lberg.de/) (Hu et al., 2019). In addition, the nucleotide sequence of BxCDP1 was queried against the genomes of all other nematodes in the WormBase ParaSite database (https://paras ite.wormb ase.org/). The P. haloflavum protein that most closely matched BxCDP1 was also queried in the genomes of species of closely related genera (Pararhizobium, Rhizobium, and Ensifer) and other microorganisms using local BLAST analysis with a BLAST E-value cut-off of 1e−10.
Alignment of the amino acid and nucleotide sequences of BxCDP1 and PhCDP1 was performed using BioEdit with ClustalW multiple alignment (Li et al., 2018a).

| RT-qPCR assays
RT-qPCR assays were carried out using ChamQ SYBR qPCR Master  (Hirao et al., 2012;Ma et al., 2015). All assays were performed three times. Primer sequences are provided in Table S1.

| Protein extraction and western blotting
Total protein extraction and immunoblotting were performed according to a previous report (Yin et al., 2013). Briefly, agroinfiltrated N. benthamiana leaves were harvested at 2 days post-inoculation.
Total protein extracts were prepared by grinding 400 mg of leaf tissue in 1 ml of radioimmunoprecipitation assay (RIPA) lysis and extraction buffer (Beyotime) in the presence of 0.1 mM phenylmethanesulfonyl fluoride (PMSF) and protease inhibitor cocktail (Beyotime).
Total proteins were separated by 12% SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane (Bio-Rad). The membranes were blocked with 5% (wt/vol) nonfat dry milk for 1 hr at room temperature followed by three washes with PBS containing 0.1% Tween-20. Transient protein expression in N. benthamiana was assessed by incubating the membrane with a 1:5,000 dilution of a primary mouse anti-HA antibody (Abmart) or anti-RFP antibody (Abcam), followed by incubation with a goat anti-mouse secondary antibody at a 1:10,000 dilution (IRDye 800, 926-32210; LI-COR Biosciences). The purified BxCDP1 protein and EV were analysed by incubating the membrane with a 1:5,000 dilution of a primary rabbit anti-His polyclonal antibody, followed by incubation with a peroxidase-conjugated goat anti-rabbit IgG secondary antibody at a 1:10,000 dilution (Zhongshan Bio-technique). The proteins were visualized using an Odyssey LI-COR imaging system. Equal protein loading was confirmed by Ponceau S staining.

| Measurement of ROS
Measurement of ROS was performed according to a previous report (Yin et al., 2013). ROS production was monitored by a luminol/peroxidase-based assay using leaf discs (

| Inoculation assay and P. thunbergii cell morphology observation
Three-year-old P. thunbergii seedlings were inoculated with 1 ml purified BxCDP1 protein (final concentration 50 μg/ml) and EV as described previously (Zhong et al., 2013). At 4 hr post-inoculation, approximately 2,000 nematodes (a mixture of juveniles and adults) were inoculated into these seedlings. The levels of B. xylophilus infection were tested, and the morbidity degrees of P. thunbergii seedlings were recorded. The infection assay of B. xylophilus was performed three times.
Meanwhile, at 10 days post-inoculation with purified BxCDP1 protein and EV, c.3 mm of seedling stem was collected from 1 cm below the inoculation site, and transversal 1-mm thick sections were cut using a surgical blade. The specimens were processed as follows: a. Double fixation: The specimens were fixed with 2.5% glutaraldehyde in phosphate buffer (PB, pH 7.0) for at least 4 hr, washed three times in PB, postfixed with 1% OsO 4 for 1 hr and washed three times in the PB.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.