Oxylipins are implicated as communication signals in tomato–root-knot nematode (Meloidogyne javanica) interaction

Throughout infection, plant-parasitic nematodes activate a complex host defense response that will regulate their development and aggressiveness. Oxylipins—lipophilic signaling molecules—are part of this complex, performing a fundamental role in regulating plant development and immunity. At the same time, the sedentary root-knot nematode Meloidogyne spp. secretes numerous effectors that play key roles during invasion and migration, supporting construction and maintenance of nematodes' feeding sites. Herein, comprehensive oxylipin profiling of tomato roots, performed using LC–MS/MS, indicated strong and early responses of many oxylipins following root-knot nematode infection. To identify genes that might respond to the lipidomic defense pathway mediated through oxylipins, RNA-Seq was performed by exposing Meloidogyne javanica second-stage juveniles to tomato protoplasts and the oxylipin 9-HOT, one of the early-induced oxylipins in tomato roots upon nematode infection. A total of 7512 differentially expressed genes were identified. To target putative effectors, we sought differentially expressed genes carrying a predicted secretion signal peptide. Among these, several were homologous with known effectors in other nematode species; other unknown, potentially secreted proteins may have a role as root-knot nematode effectors that are induced by plant lipid signals. These include effectors associated with distortion of the plant immune response or manipulating signal transduction mediated by lipid signals. Other effectors are implicated in cell wall degradation or ROS detoxification at the plant–nematode interface. Being an integral part of the plant's defense response, oxylipins might be placed as important signaling molecules underlying nematode parasitism.


M. javanica infection induces expression of tomato oxylipin-biosynthesis genes
RNA-seq analysis of M. javanica J2 following exposure to 9-HOT and tomato protoplasts. To identify transcripts and effectors that are subject to plant host lipid-signaling regulation, the oxylipin 9-HOT, which-among other oxylipins-is induced in tomato roots by RKN infection (Fig. 1), was chosen as the inducer. Similarly, in an attempt to mimic exposure to endogenous metabolites secreted within plant tissue during the parasitic stage, J2 were exposed to tomato protoplasts. Freshly hatched J2 incubated in MES buffer or MES + ethanol served as controls for tomato protoplasts and the studied oxylipin, respectively. Altogether, eight cDNA libraries were constructed, generating a total of 170,786,458 reads, 121,161,460 reads after quality and adaptor trimming by Trimmomatic version 0.35 74 . The eight libraries represented: (1) biological duplicates of freshly hatched J2 with 15,208,788 reads; (2) biological duplicates of J2 exposed to protoplasts for 3 h with 15,572,666 reads; (3) biological duplicates of J2 exposed to 9-HOT for 3 h with 14,677,503 reads; (4) biological duplicates of J2 in MES + ethanol with 15,121,772 reads. All RNA-Seq raw data reads were uploaded to NCBI under BioProject Accession PRJNA480605. From the 112,732,357 high-quality paired-end reads, 72.01% of the freshly hatched J2, 67.11% of the J2 exposed to protoplasts, 52.78% of the J2 exposed to 9-HOT and 73.01% of the J2 exposed to MES + ethanol were mapped to the reference genome of M. javanica, available on WormBase ParaSite BioProject PRJEB8714 and sequenced by Blanc-Mathieu et al. 75 (Table 2).
Uncovering transcriptomic changes in M. javanica J2 upon exposure to oxylipin 9-HOT and other plant signals. To measure transcript regulation of M. javanica J2 by 9-HOT and plant signals, we measured changes in gene expression of infective J2 exposed to tomato protoplasts compared to their control (MES), and of infective J2 exposed to 9-HOT compared to their control (MES + ethanol). Statistical analysis of the differentially expressed genes (DEG) was performed using the DESeq2 package 76 . The threshold for DEG was FDR ≤ 0.001 and log2 fold change (FC) smaller than − 2 or greater than 2. Overall 7530 DEG were identified among the treatments-J2 exposed to protoplasts and J2 exposed to 9-HOT-compared to their respective controls. Principal component analysis (PCA) was conducted to determine and visualize the significant correlation between the different treatments using R (version 3.0.0) (http://www.R-proje ct.org) and the FactoMineR R package 77 . PCA of the transcriptomic data was performed for all expressed gene profiles (Fig. 3A). The two dimensions made up 98.51% of the total variance, indicating that most of the factors included in the data were responsible for the significant variation in DEG between treatments. Dimension 1 accounted for 97.46%, and dimension 2 for 1.05% of the variation. The first and second PC axis separated the freshly hatched J2 exposed to 9-HOT from all other groups. Taken together, these results suggest that the different treatments can be divided into two major expression profiles of DEG: (1) freshly hatched J2, J2 exposed to protoplasts and MES + ethanol and (2) J2 exposed to 9-HOT, the latter demonstrating the most variation (Fig. 3A). Next, all 7512 DEG were subjected to Venn diagram analysis illustrating DEG distribution among treatments (Fig. 3B). A total of 6085 DEG (81%) and 1057 DEG (14.1%) were found to be expressed exclusively in the 9-HOT and protoplast treatments, respectively, and 370 DEG (4.9%) were expressed in both treatments (common DEG) (Fig. 3B.1). A total of 4580 DEG were found to be upregulated (83.3%), with 3813 (83.3%) and 751 (16.4%) expressed in the 9-HOT and protoplast treatment, respectively, and 16 (0.3%) expressed in both treatments (Fig. 3B.2). GO enrichment analysis of the annotated DEG revealed several enriched biological processes ( Fig. 3C.1) and molecular functions (Fig. 3C.2). Detailed analysis of identified DEG following exposure to protoplast and 9-HOT revealed that the key enriched biological processes were negative regulation of endopeptidase (GO:0061135), peptidyl-proline modification (GO:0031543), glycosylceramide catabolic process (GO:0004348), cellulose metabolic process (GO:0030243), galactosylceramide metabolic process (GO:0006683).
To further classify the observed changes, all 7512 DEG were analyzed for Pathway enrichment using the webserver of KOBAS 3.0 78 KEGG, GO and PANTHER database enrichment analysis. When compared to Loa loa for glutathione metabolism (loa00480), which plays important roles in antioxidant defense 79 , 15 DEG were upregulated and 7 DEG downregulated, for a total of 22 out of 25 known genes involved in this pathway (Fig. S1A). In the Wnt signaling pathway (loa04310), known to regulate crucial aspects of cell-fate determination during embryonic development 80 , 24 DEG were upregulated and 5 DEG were downregulated-29 out of the 65 known genes represented in this pathway (Fig. S1B). In the fatty acid biosynthesis pathway (loa00061), which is a precursor for a variety of important building blocks 21,81,82 , 8 DEG were upregulated , 10 out of the 10 known genes. For retinol metabolism (loa00830), a total of 6 DEG were upregulated out of 6 known genes in this pathway. All of these DEG were enriched following exposure to 9-HOT. Following exposure of J2 to protoplast treatment, the calcium signaling pathway (loa04020) was represented by 27 upregulated DEG out of a total of 57 genes known to be involved in this pathway. In the inositol phosphate metabolism pathway (loa00562), 9 DEG were upregulated out of a total of 38 known genes in this pathway. In the phosphatidylinositol signaling system (loa04070), 9 DEG were upregulated out of 57 known genes in this pathway.

9-HOT application regulates the expression of genes encoding carbohydrate-active enzymes (CAZymes) related to cell wall modification and degradation.
To further evaluate the effect of 9-HOT and protoplasts on CAZymes, we investigated families of structurally related catalytic and carbohydrate-binding modules (CBM) of enzymes that degrade, modify or create glyosidic bonds. We focused on the DEG encoding CAZymes related to cell wall biosynthesis, modification and remodeling 82 (Fig. S2). Differential expression of four categories of CAZYmes were represented following exposure to 9-HOT and protoplasts, i.e., genes encoding carbohydrate esterase (CE), glycoside hydrolase (GH), glycosyl transferase (GT) and polysaccharide lyase (PL) ( Figure S2A). Among the CAZyme categories involved in cellulose degradation (Fig. S2B), www.nature.com/scientificreports/ two families of GH were differentially expressed following protoplast and 9-HOT treatments. These included GH5 and GH7 presented by endo-1,4-β-glucanase/cellulase and β-glucosylceramidase and chitosanase. Among the hemicellulose-degrading genes (Fig. S2C), two categories were represented by GH31 and CE1, such as α-galactosidase, α-mannosidase, glucosyltransferase acetyl transferase and carboxylesterase.

9-HOT induces major differences in nematode's effector-encoding gene expression.
Given that we were interested in genes involved in governing parasitism, i.e., effectors, our next step was an in-silico analysis to identify differentially expressed transcripts that might encode secreted effectors. DEG that contained a predicted signal peptide according to SignalP5.0 83 and which do not carry TMHMM (transmembrane alpha helix motifs) were subjected to Venn diagram analysis illustrating DEG distribution among treatments (Fig. 4A). A total of 913 DEG with a signal peptide were identified (12.2% of total DEG). Among these, 707 DEG (77.4%) and 116 DEG (9.9%) were expressed in the 9-HOT and protoplast treatments, respectively, and 90 DEG (12.7%) were expressed in both treatments (Fig. 4A.1). A total of 367 DEG with a signal peptide were found to be upregulated (33.7%) (Fig. 4A.2), and 597 (54.9%) DEGs were downregulated (Fig. 4A.3). All DEG that were upregulated by 9-HOT treatment were analyzed for GO terms (Fig. 4B) in each of the three main categories (biological process, molecular function and cellular component classification) of the GO classification. The GO cellular component classification of the DEG indicated 12% of extracellular region and 88% of the membrane part. Biological process was represented by 10 categories: 31% by negative regulation of peptidase activity (GO:0010466), 30% by regulation of endopeptidase activity (GO:0052548), about 4-7% each by chitin, ceramide, glucosamine, glycosaminoglycan and glycolsphingolipid catabolic process (GO:0006032, GO:0046514, GO:1901072, GO:0006027 and GO:0046479, respectively) (Fig. 4B). Molecular function was represented by 11 categories: 31% by serine-type endopeptidase inhibitor activity (GO:0004867), 18% by metalloendopeptidase activity (GO:0004222), 9% by carboxypeptidase activity (GO:0004180). Next, we validated the expression profile of predicted effectors in protoplast-and 9-HOT-treated M. javanica J2. For that purpose, seven selected DEG from the M. javanica J2 transcriptome were further confirmed and validated by quantitative reverse transcription (qRT)-PCR. Four downregulated genes and three upregulated genes were selected for quantitative analyses, on the basis of being potentially secreted and involved in the pathogenic process and carrying a signal www.nature.com/scientificreports/ peptide. One of the DEG was found in J2 exposed to protoplasts: protoplasts#1 encoding DB domain-containing protein (M.javanica_Scaff1102g012786); six were found in J2 exposed to 9-HOT: 9-HOT#1-unknown gene (M.javanica_Scaff10526g059067), 9-HOT#2-SCP domain-containing protein (M.javanica_Scaff139g002482), 9-HOT#3-unknown gene (M.javanica_Scaff8981g053951), 9-HOT#4-putative esophageal gland cell secretory protein 3 (M.javanica_Scaff2606g024064), 9-HOT#5-calycin-like domain (M.javanica_Scaff24242g089056), 9-HOT#6-triacylglycerol lipase (M.javanica_Scaff6853g045742), all of which are shown in Fig. S3. For all qRT-PCR analyses, two housekeeping genes were chosen as reference genes for M. javanica: endogenous reference genes 18S and EF-1α (Table 1S). For all tested transcripts, our analysis remained similar to the transcriptomic trend of the FC data.
Differentially regulated effector-encoding genes are implicated in cell wall modifications, stress response, plant immune suppression and nematode development, enabling parasitism. Among the 346 upregulated DEG in the 9-HOT treatment were genes implicated in nematode growth and development, such as MLT-10, the cuticlin-1, epicuticulin gene family and collagen, all of which participate in various cellular and developmental processes required for nematode molting and fecundity. Within the root tissues, RKN undergo three molting stages; in each molt, the makeup of the cuticle surface coat's compounds changes, one among many strategies acquired by plant-parasitic nematodes to avoid plant immunity 84 . Similarly, previous studies have shown that hormones and different compounds secreted by the roots trigger changes in the surface cuticle of sedentary plant-parasitic nematodes 10,85 .
In addition, a group of genes implicated in oxidation-reduction activity required for coping with oxidative stress response were induced by 9-HOT (i.e., glutaredoxin, thioredoxin-like domain). Similarly, effectors containing a C-type lectin, which was found to delay the oxidative burst in tobacco leaves following infection by M. graminicola 86 , were upregulated following exposure to 9-HOT (M.javanica_Scaff6180g042809) and protoplasts (M.javanica_Scaff16387g074472) (Fig. 4A). In addition, genes involved in lipid modification (several genes of triacylglycerol lipase, calycin domain) were also differentially expressed following 9-HOT treatment. Differential regulation of several proteases was observed: a carboxypeptidase, serine carboxypeptidase (SCP), was studied in depth and was shown to play a critical role in the development, invasion, and pathogenesis of certain parasitic nematodes and other animal pathogens 87 . Another protease, a serine proteinase, that was induced has also been shown to be involved in mediation of host invasion by the parasitic nematode Steinernema carpocapsae 87 . Similarly, alteration in the expression of chorismate mutase (CM) and venom allergen-like protein (Vap2), well-studied effectors that are involved in suppression of defense reactions of the host cell during the infection stages, were detected 14,87-91 . Papain inhibitor, which might be related to pathogen effectors that inhibit apoplastic papain-like cysteine protease (PLCP) which are strongly associated with the effector triggered immunity (ETI) response 92 , were strongly upregulated upon 9-HOT treatment. An extensive representation of genes involved in cell wall modification and remodeling, carrying a signal peptide, were altered upon 9-HOT and protoplast exposure, e.g., effectors carrying a Rare lipoprotein A domain found in several effectors, such as Mc-EXP1, GrEXPB1 and GrEXPB2, associated with cell wall extension in M. chitwoodi and G. rostochiensis, respectively 93,94 , were upregulated upon 9-HOT application. Additional observed differentially regulated genes were involved in cell wall degradation and modification (Fig. S2), including cellulose binding protein (CBP)-a nematode excretion protein that appears to be associated with the breakdown of cellulose present in the plant cell wall 95,96 -and pectate lyases, known to play a key role in pectin degradation by catalyzing the random cleavage of internal polymer linkages (endopectinases). Similarly, pectate lyases have been isolated from several sedentary plantparasitic nematodes, such as species of Heterodera, Globodera, and Meloidogyne [97][98][99] , and have been shown to be released into the plant tissue through the stylet of the nematode. GH family 38 (GH38) α-mannosidase was upregulated by 9-HOT; this protein is involved in α-mannose cleavage, carrying hemicellulose activity, and has been identified in several phytopathogenic nematode species 100 . In addition, several endoglucanases belonging to the GH5 family were differentially expressed upon 9-HOT treatment (Fig. S2); these have been shown to facilitate penetration and migration into root tissue and were localized to the esophageal glands of infective juveniles 95,101,102 . All of these genes are part of a cocktail of cell wall-degrading and modifying enzymes that are thought to soften and degrade the structure of plant cell walls during nematode migration and to facilitate infection 95,103 . Their fluctuation upon 9-HOT and protoplast treatment might indicate tight regulation governed by oxylipin signals, among others.
Triacylglycerol lipase and MLT10-like, predicted effectors, are exclusively localized to the M. javanica esophageal glands. Additional effectors that may facilitate plant-nematode interactions through manipulation of the plant defense system, or are required for nematode developmental processes, and which were localized to the esophageal glands upon 9-HOT application, are triacylglycerol (TAG) lipase, which functions in TAG release from lipid droplets by lipolysis in the peroxisome 82 . Interestingly, modulation of plant peroxisomes in giant cells by sedentary RKN has been described previously 104 ; and molting cycle MLT-10-like, required for nematode development. Using fluorescence in situ hybridization (FISH), we designed a Cy5-probe to specifically target the spatiotemporal expression of several putative effector-encoding genes derived from the above DEG carrying a signal peptide. FISH results, shown in Fig. 5, localized TAG lipase exclusively to the dorsal and two subventral glands upon J2 exposure to 9-HOT (Fig. 5A1-4), compared to its control (i.e., J2 exposed to MES + ethanol) that showed no fluorescent signal (Fig. 5A5-8). Similarly, MLT-10 was localized to the subventral glands upon J2 exposure to 9-HOT (Fig. 5B1-4), compared to its control (J2 exposed to MES + ethanol) with no fluorescent signal. www.nature.com/scientificreports/ These results strengthened our assumption that the DEG containing a signal peptide are potential secreted effectors. Additional analysis of TAG lipase and MLT-10-like by qRT-PCR at different stages of M. javanica development further correlated their expression with parasitism (Fig. S4).

Conclusions
Despite enormous progress in the discovery and identification of nematode effectors in the last decade 7,15,[105][106][107][108][109][110] , less is known about their function and the specific signals required for their induction. Our transcriptomic studies of M. javanica provide evidence of transcripts with homology to previously reported plant-parasitic nematode effectors, as well as unknown secreted proteins, all induced by 9-HOT. Together with the oxylipin profile analysis, it seems that oxylipins, while part of the plant's defense response, might play an important signaling role in regulation of the nematode transcriptome. Among the differentially regulated predicted effectors, several were further confirmed by FISH analysis as effectors located within the esophageal glands. Taken together, these results placed oxylipins as early modulators of plant defense signals, as well as important signals regulating nematode parasitism. The implication of G-protein coupled receptor (GPCR) as an oxylipin receptor, shown previously by Affeldt et al. 111 and Lahvic et al. 112 , should place this group of nematode receptors as important mediators of parasitic behavior. This interaction remains to be studied. www.nature.com/scientificreports/

Methods
Nematode culture and inoculum preparation. M. javanica were multiplied on tomato plants (Solanum lypopersicum cv. Avigail 870) in a greenhouse. Nematode egg masses were extracted by cutting roots into pieces and macerating in 0.05% (v/v) sodium hypochlorite (NaOCl) in a blender 113 . The resulting suspension was passed through a set of three sieves (120, 60 and 30 µm). The debris was discarded, while the eggs deposited on the 30-µm sieve were transferred to a 50-mL test tube. Centrifugal flotation with 40% (w/v) sucrose at 6000 rpm for 10 min was performed; the supernatant, containing the eggs, was poured onto a 30-µm sieve and washed with tap water, and eggs were collected in MES buffer 113 and sterilized as described by van Vuuren and Woodward 114 . Subsequently, eggs were transferred onto a 30-µm sieve and suspended in 5 mL MES buffer in a petri dish. The petri dish was placed in a growth chamber at 26 °C under dark conditions till hatching (5-6 days) 115  Differential expression analysis. The sequences were trimmed for adaptor and low-quality sequence removal using Trimmomatic software 74 . Cleaned sequences were mapped using Bowtie2 118 and quantified using the RSEM method 119 to the reference genome of M. javanica (accession no. GCA_900003945.1). The annotated proteins of the nematode were searched for signal peptides using the software SignalP 5.0 83 . DEG were identified using the DESeq2 R package 76 . To create the Venn diagrams, we used Venny website http://bioin fogp.cnb.csic.es/tools /venny / 120 . All RNA-Seq datasets were uploaded to the SRA NCBI database under BioProject Accession No. PRJNA480605. CAZyme annotation. The search for and functional annotation of CAZymes (automated carbohydrateactive enzyme annotation) was performed using the CAZY database (http://www.cazy.org/) according to Lombard et al. 121 . We assigned the M. javanica (accession no. GCA_90003945.1) proteins to the CAZY database using the dbCAN2 meta server (http://bcb.unl.edu/dbCAN 2inde x.php).
Plant material and growth conditions. Tomato (Solanum lycopersicon cv. Avigail 870) seeds were sterilized with 1.4% (v/v) NaOCl for 10 min, washed three times with sterile water for 5 min each, and then planted on standard strength Gamborg's B5 salts medium (DUCHEFA), supplemented with 2% sucrose and solidified with 0.8% Gelrite agar (DUCHEFA). Seeds were kept in a growth chamber at 26 °C under a 16/8-h photoperiod at 120 µmol/m 2 s for 2 weeks until cotyledons appeared.
Plasmid construction and generation of transgenic hairy roots. All  www.nature.com/scientificreports/ GCC CGG GAC TTG ACA ACT AAA A-3′, with Kpn1 and SmaI restriction sites; αDOX1 FOR:5′-TTG GGA GAG AGG AGC TCG ACA ATT TTT-3′, REV-5′-AGG TAC CTA GCC CGG GTG TTT ATA CGA-3′, with SacI and SmaI restriction sites; all for ≈ 1500 pb amplicons. LOX1.2, AOS1, OPR2 and αDOX1 promoters were then cloned into the pUC19_Y vector 122 . The whole cassettes containing the specific gene promoters and the GUS reporter gene were then isolated by restriction digestion with SacI, EcoRI, KpnI and SacI (respectively) and SalI, to be cloned into the pCAMBIA2300 binary vector. The identity, orientation, and junctions of the resulting constructs pLOX1.2:GUS, pAOS1:GUS, pOPR2:GUS and pαDOX:GUS were confirmed by their digestion patterns. The pCAMBIA2300 empty-vector control and the four constructs were subsequently used for Rhizobium rhizogenes-mediated transformation 123 .
Rhizobium rhizogenes-mediated root transformation. R. rhizogenes ATCC 15,834 strain was used for the transformation by heat-shock method 124 . Individual cotyledons were excised from 15-to 20-day-old tomato seedlings grown as described above and immersed in a 2-day-old R. rhizogenes suspension for incubation at 28 °C for 2 h, with agitation at 100 rpm. The excised cotyledons then were placed on standard-strength Gamborg's B5 salt media for 3 days for co-cultivation, and then transferred to B5 agar media supplemented with the antibiotics kanamycin (50 mg/mL) (DUCHEFA, Haarlem, the Netherlands) and timentin (15:1) at 300 mg/mL (DUCHEFA, Haarlem, the Netherlands). After 7-10 days of incubation in the dark at 25 °C, roots emerged from the wounded surface of the cotyledons. Hairy roots were transferred to Gamborg's B5 medium containing 0.8% Gelrite and kanamycin (50 mg/mL). For nematode-infection experiments, transformed roots were subcultured in antibiotic-free media for 2 weeks, and 300 freshly hatched sterile M. javanica juveniles were used to inoculate the transgenic root lines, and root samples were taken at the designated time points for GUS assessment.
GUS bioassay. Two-week-old hairy root lines carrying the promoter GUS constructs were inoculated as described by Chinnapandi et al. 125 , and assayed histochemically for GUS activity at the designated times after infection with 300 sterile freshly hatched pre-parasitic M. javanica J2.
Plant oxylipin and hormone extraction. Tomato (Solanum lycopersicon cv. Avigail 870) seeds were sterilized and planted as described. Emerging roots were cut into 2-cm long segments, subcultured on Gamborg's B5 medium and kept in the dark for 2 weeks. Root systems were inoculated with freshly hatched M. javanica J2 and galls were collected 5, 15 and 28 DAI. The noninoculated roots were collected as controls. Five independent biological replicates with a total of 100-130 mg of gall samples from inoculated and noninoculated tomato roots at the selected time points were collected into a 1.5-mL tube and kept at − 80 °C until further analysis. Samples were weighed and all of the data were normalized to the relative weight of those frozen tissues. Oxylipins/hormones were extracted from each sample in liquid N 2 using the phytohormone-extraction protocol reported by Yang et al. 126 127 . The gall tissue was homogenized at 6000 rpm for 30 s, twice. Samples were agitated for 30 min at 4 °C in the dark and then 500 µL dichloromethane was added and samples were agitated again for 30 min at 4 °C in the dark. Samples were centrifuged at 13,000 g for 5 min and the lower layer was collected into a glass vial for complete evaporation under a N 2 gas stream. Samples were resuspended in 150 µL methanol, shaken for 1 min and then centrifuged in a 1.5-mL microcentrifuge tube at 14,000 g for 2 min to pellet any debris. Supernatant (100 µL) was collected into an autosampler vial for injection into a SCIEX API 3200 LC-MS/MS with a C18 column for chromatography and electrospray ionization. Peaks were integrated using Analyst 1.6.2 software and metabolites were quantified against internal standards 41,127 . Identification of fatty acid peaks was verified by comparison of the mass spectra to authentic standards. Noninoculated tomato roots served as controls.

Real-time qPCR analysis.
For the qRT-PCR experiments, we removed contaminating genomic DNA from RNA with the Turbo DNA-Free Kit from AMBION (APPLIED BIOSYSTEMS). DNA-free RNA (1 μg) was converted to first-strand cDNA using the Verso cDNA Synthesis Kit (ABGENE, Epsom, UK), and reactions were performed using ABsolute SYBR Green ROX Mix (ABGENE). Primers for qRT-PCR experiments were designed with Primer Express software (APPLIED BIOSYSTEMS, Table 1S). A total volume of 10 μL contained 3.4 µL cDNA, consisting of 1 × SYBR-Green ROX Mix (ABGENE) and 150 nM forward primer and 150 nM reverse primer subjected to real-time PCR (Rotor-Gene RG-3000, CORBETT RESEARCH) using 0.1 mL 4-tube strips & caps (AXYGEN, Union City, CA, USA). All PCR cycles began with 2 min at 50 °C, then 10 min at 95 °C, followed by 40 cycles of 10 s at 95 °C and 1 min at 60 °C. After the PCR, a melting curve was generated by gradually increasing the temperature to 95 °C to test for amplicon specificity. For qRT-PCR, a mixture of all cDNAs was used for all treatments as a template for calibration curves designed for each pair of primers. Each reaction was performed in triplicate and the results represent the mean of two independent biological experiments. Two reference genes, EF-1α (GenBank accession no. U94493.1) and 18S (GenBank accession no. AF442193.1), were used as endogenous controls for gene-expression analysis. Transcript levels were normalized for each sample with the geometric mean of the corresponding selected reference genes. All of the reference genes were confirmed to display minimal variation across the treatment and were the most stable reference genes from a set of tested genes in a given cDNA sample. Values were expressed as the increase or decrease in level relative to a calibration sample. A negative control PCR without cDNA template was also run to confirm the absence of www.nature.com/scientificreports/ nonspecific PCR products (NTC) No Template Control 127 . The same process was performed for qRT-PCR of the expression of the nematode effectors TAG lipase and MLT-10.

FISH.
Freshly hatched pre-parasitic M. javanica J2 were exposed to 9-HOT diluted in MES buffer to a final concentration of 10 mM, or to 0.01 M MES buffer, for 3 h; all samples were washed with 0.01 M MES buffer. The FISH procedure followed the method of Sakurai et al. 128 with slight modifications 129 . The J2 were dissected manually with a razor blade and transferred to Carnoy's fixative (chloroform:ethanol:glacial acetic acid, 6:3:1, v/v) and fixed overnight. The samples were then decolorized in 6% (v/v) hydrogen peroxide in ethanol for 2 h and hybridized overnight in hybridization buffer (20 mM Tris-HCl pH 8.0, 0.9 M NaCl, 0.01% w/v SDS, 30% v/v formamide) containing 10 pmol fluorescent probe/mL. Based on the transcriptome sequences of interest, DNA probes were designed using Primer Express 3.0.1 software and checked for specificity using BLASTn (NCBI); TAG lipase (nematode effector) Cy5 (5′-Cy5-AAT TGA TGT TCG TGC AGA CCAT-3′) and MLT-10-like Cy5 (5′-Cy5′-AGA CAA AAG GGT GCA GAA CGA-3′) were used as probes to target M. javanica J2. The stained samples were submerged in hybridization buffer supplemented with DAPI (0.1 mg/mL in 1X PBS) and transferred to a slide with liquid blocker, covered, sealed with nail polish and viewed under a confocal microscope.Detection specificity was confirmed using M. javanica exposed to 0.01 M MES buffer only as a control. www.nature.com/scientificreports/