A pure culture of M. incognita race 1 was maintained on eggplant (Solanum melongena L. cv. Pusa Purple Long) in a glasshouse. Egg masses were extracted from the roots of 2-months-old plant using sterilized forceps and were hatched in a Petri dish containing sterile water via modified Baermann assembly (Whitehead and Hemming, 1965). Freshly hatched J2s were used for all the experiments. Other life stages were carefully dissected out of the infected tissue at different intervals for qRT-PCR assay.

Total RNA (400 ng) was isolated from the pre-parasitic J2 of M. incognita, their quality and quantity were assessed and converted to cDNA as described by Shivakumara et al. (2016). Fragments of five CWMEs (Mi-xyl-1: AF224342, Mi-xyl-3: EU475876, Mi-pg-1: AY098646, Mi-eng-1: AF100549, and Mi-pel: AF527788) and two pioneer genes (msp-18: AY134437 and msp-20: AY134439) were PCR amplified, cloned into pGEM-T vector and identity of the insert was confirmed via sequencing as described in Shivakumara et al. (2016). Primer details are documented in Shivakumara et al. (2016).

Differential expression of each of the msp-18, msp-20, Mi-xyl-1, Mi-xyl-3, Mi-pg-1, Mi-eng-1, and Mi-pel transcripts were analyzed by qRT-PCR at six different developmental stages (eggs, pre-parasitic J2, post-parasitic J2, J3/J4, young female and adult female) of M. incognita. RNA was extracted from different stages of M. incognita and reverse transcribed to cDNA as described above. All qRT-PCR primers (Shivakumara et al., 2016) were designed using IDT OligoAnalyzer 3.1 software¹ and optimized for working concentration and annealing temperature prior to use with a Realplex² thermal cycler instrument (Eppendorf). The 10 μl reaction mixture contained 1.5 ng of cDNA, 750 nM of each primer and 5 μl of SYBR Green PCR master-mix (Eurogentec). The amplification reaction conditions and melt curve program were set as described by Shivakumara et al. (2016). Cycle threshold (Ct) values were exported from Realplex² software (Eppendorf). Gene expression was normalized using 18S rRNA (Genbank accession number HE667742) as reference and fold change in expression was calculated according to the augmented comparative Ct method (Livak and Schmittgen, 2001) and log₂-transformed. At least three biological and three technical replicates were used for each of the samples.

The pK7GWIWG2(I) binary vector (RNAi Gateway ready) was obtained from VIB Department of Plant Systems Biology, Ghent University, Belgium². Partial sequences of msp-18 (456 bp) and msp-20 (598 bp) flanked by attB1 and attB2 sites were PCR amplified from respective pGEMT clones and sub-cloned into the entry vector (pDONR 221) followed by transfer to pK7GWIWG2(I) in sense and antisense orientation (Supplementary Figure S1). The recombinant clones were transformed into Agrobacterium tumefaciens (LBA4404) competent cells using electroporation technique, positive clones were maintained in selection medium following standard techniques. Detailed procedure is described in Papolu et al. (2013) and Dutta et al. (2015b). Primer details are given in Supplementary Table S1.

Eggplant (cv. Pusa Purple Long) seeds were surface sterilized and germinated on MS agar medium (pH 5.8). Leaf explants of 1 cm² cut from the fortnight-old eggplant seedling were used for Agrobacterium-mediated transformation. Pre-cultivated explants were infected with A. tumefaciens (LBA4404) cells harboring the RNAi constructs of msp-18 and msp-20. Agroinfected leaves were plated on co-cultivation media followed by transfer to the selection medium containing kanamycin. Shoots that emerged from the explants were excised and sub-cultured in fresh selection medium. Elongated shoots were cultured in rooting media supplemented with 0.1 mg L^-1NAA (Supplementary Figure S2). Plants with well-established roots were hardened and transferred to the transgenic glasshouse facility at ICAR-IARI, New Delhi till the production of T₀ seeds. Plants transformed with A. tumefaciens harboring empty construct, and thus underwent no kanamycin selection, served as the control.

Genomic DNA was isolated from the fresh leaves of all the primary events (T₀) using a Nucleospin plant II DNA extraction kit (Macherey-Nagel) following the manufacturer’s instructions. All the putative transformants were initially confirmed using standard PCR reaction with different sets of primers which are enlisted in Supplementary Table S1. PCR products were resolved on 1.2% agarose gel. In order to analyse the integration and copy number of msp-18 and msp-20 transgenes in PCR-confirmed T₀ plants, Southern hybridization was carried out. Genomic DNA (15 μg) of each event was digested with SacI enzyme (cuts once in the T-DNA), electrophoresed on 0.8% agarose gel and transblotted to a nitrocellulose membrane (Bio-Rad). For probing, a 456 bp fragment of the msp-18 gene and a 598 bp of the msp-20 gene were used independently. Probe labeling, hybridization and blot development were performed as described in Papolu et al. (2013) and Dutta et al. (2015b).

PCR-positive T₀ plants were allowed to self-pollinate and T₁ seeds were obtained. Surface-sterilized T₁ seeds were germinated in MS medium containing kanamycin (100 mg L^-1) followed by transfer to pots filled with autoclaved soil. To analyse the transcript abundance of msp-18 and msp-20 genes in T₁ plants, qRT-PCR experiment was carried out. Total RNA was isolated from the fresh leaves of PCR-positive T₁ events using a Nucleospin plant II RNA extraction kit (Macherey-Nagel) and reverse transcribed to cDNA using a cDNA synthesis kit (Superscript VILO, Invitrogen) following the manufacturer’s instructions. qRT-PCR experiment was conducted with three independent biological and technical replicates following the protocol mentioned above. Expression of msp-18 and msp-20 in different transgenic lines were depicted as average ΔCt, i.e., difference between the Ct mean of transgene and the normalizer gene, 18S rRNA (Gantasala et al., 2013; Papolu et al., 2016).

For detection of msp-18 and msp-20 siRNAs in T₁ events, Northern analysis was performed. Small RNAs were extracted from the fresh leaves of PCR-positive T₁ plants using a miRNA isolation kit (mirVana^TM, Ambion) following the manufacturer’s instructions. Small RNAs (10 μg) of each event were electrophoresed on 15% denaturing PAGE gel followed by transblotting to a nitrocellulose membrane (Bio-Rad). Probes for msp-18 and msp-20 were labeled through PCR using digoxigenin (DIG) probe synthesis system (Roche Diagnostics). Blot hybridization and development were carried out according to the manufacturer’s instructions (Roche Diagnostics).

In order to evaluate the T-DNA integration sites in transformed plants, msp-18 and msp-20 integration and flanking sequences in selected T₂ events were identified using genome walking technique (Universal genome walker 2.0, Clontech laboratories). Genomic DNA was isolated from fresh leaves of T₂ plants as described above. Two microgram DNA was used as the template in PCR reaction with T-DNA- and adapter-specific primers (enlisted in Supplementary Table S1) following the manufacturer’s instructions. The final PCR products were cloned and sequenced as described above. Sequences were analysed via NCBI-BLAST and sol genomic network³. In addition, event-specific PCR was performed to authenticate the specific site of integration in the progeny plants. The gene-specific and flanking primers are given in Supplementary Table S1.

Homozygous T₁ plants harboring the RNAi constructs of msp-18 and msp-20 were subjected to bioefficacy studies against M. incognita. The roots of a fortnight-old T₁ plants (growing in 250 ml pots containing equal mixture of soil and soilrite) were inoculated with approximately 300 J2s of M. incognita in the vicinity of the root tip. Nematode infected plants were grown at 28°C, 70% relative humidity and 16:8 h light:dark photoperiod in confinement. Plants were harvested at 30 dpi (days post inoculation). Different agronomic characters (root and shoot weight) of plants and nematode parasitic success on these plants [in terms of number of galls, egg masses, eggs/egg mass and multiplication factor (MF)] were recorded as described by Papolu et al. (2013, 2016) and Dutta et al. (2015b). The experiment was conducted twice with at least six replicates.

At 2 and 7 dpi, some of the plants from each treatment were harvested and roots were stained with acid fuchsin (Byrd et al., 1983) in order to estimate the nematode penetration ability on transgenic plants.

Total RNA was isolated from feeding females and reverse transcribed to cDNA as described above. qRT-PCR was performed and fold change in expression of target gene was calculated as explained above.

Additionally, the effect of HIGS on early parasitic J2s of M. incognita was analyzed. Seven-day-old transgenic (T₁) and wild-type seedlings were placed independently at the center of a 50 mm Petri dish containing 5 ml of 23% pluronic gel (PF-127, Sigma; Dutta et al., 2011; Kumari et al., 2016) and approximately 100 J2s of M. incognita were inoculated behind the root tip of each plant. Plants were harvested at 8 h post inoculation because maximum numbers of J2s were latched on to the root tip during that time point. J2s clinging around the root tip were carefully extracted under the microscope, collected in a microcentrifuge tube and rinsed with sterile water for several times to remove the traces of PF-127. RNA was extracted from J2s and qRT-PCR assay was conducted as described above.

Seeds of wild-type eggplant (cv. Pusa Purple Long) were surface sterilized with 0.1% HgCl₂ prior to sowing on a sterilized plastic bowl containing quartz sand in a growth chamber. At 7 days post germination of eggplant, bowls (each containing 70–80 seedlings) were transferred into an air tight Plexiglas chamber and exposed to ¹⁴CO₂ which was released due to acidification of sodium ¹⁴C-bicarbonate (specific activity – 2.1 mCi/mmole) by 1 N HCl (Supplementary Figure S3). The plants were incubated in ¹⁴CO₂ environment for 30 min followed by the pouring of KOH solution into the source Petri dish via PVC tubing to remove any unused ¹⁴CO₂ in the chamber. After another 30 min, bowls containing ¹⁴CO₂-labeled plants were removed from the chamber and transferred to the laboratory condition at room temperature. Detailed procedure is described in Pandey et al. (2013) and Singh et al. (2014).

In order to visualize the transport of carbon within the labeled plant, autoradiogram analysis was done. The plants were removed gently from the bowl and washed with deionized water to remove any external contaminants. Wrapped in filter paper plants were instantly dried at 80°C in a gel drier for 20 min. The plants were further flattened under a weight for 5 days. Subsequently, the plants were wrapped in transparent cling film and exposed to Fujifilm (Kodak) for 1 week.

The incorporation of ¹⁴C into the different metabolic pools (i.e., roots and shoots) of eggplant was determined by measuring an aliquot of the respective fraction using a liquid scintillation counter (TRICARB, Perkin Elmer). The labeled plants were dried at 60°C in an oven for overnight followed by crushing of the roots and shoots using sterilized pestle and mortar. Twenty milligram of Cab-O-Sil (stabilizes sample suspension by thixotropy, Cabot Corp. Inc.) and 4 ml of Ultima Gold Scintillation Cocktail (Perkin Elmer) were added to the known volume of the dried sample, incubated for 30 min in dark and scintillation counts were measured.

In parallel, the fresh labeled plants were put in the PF-127 medium in a Petri dish (each containing five plants; Supplementary Figure S3) and challenge inoculated with 100 J2s of M. incognita as described above. At 8 h post inoculation, J2s were gently separated from the root tip and washed with deionized water for at least 20 times in order to remove external contaminants. Afterward, J2s were pooled together in an eppendorf tube and dried in vacuo. The known volume of the dry residue was dissolved in Cab-O-Sil and scintillation liquid cocktail, and scintillation counts were measured. The total ¹⁴C assimilation was calculated as disintegrations per minute (dpm) per mg dry weight. The relative ¹⁴C assimilation was expressed as percentage in shoot, root, and nematodes separately.

Bioassay and expression data were analyzed by one-way ANOVA and Tukey’s HSD test (significance level at P < 0.05) using SAS version 9.3 for Windows (SAS Software, Inc.).

Sequencing analysis of the cloned fragments of msp-18 (456 bp), msp-20 (598 bp), Mi-xyl-1 (610 bp), Mi-xyl-3 (610 bp), Mi-pg-1 (547 bp), Mi-eng-1 (585 bp), and Mi-pel (452 bp) revealed 100% identity to the reported sequences of M. incognita. Simultaneously, comparative expression of these genes in six developmental stages of M. incognita was carried out through qRT-PCR. Using the expression level in eggs as reference, mRNA levels of msp-18 were significantly (P < 0.05) downregulated in pre-parasitic J2 in contrast to consistent upregulation in post-parasitic J2, J3/J4 mixed stages, young and adult females (Figure 1). Conversely, msp-20 was significantly (P < 0.05) overexpressed in pre- and post-parasitic J2s compared to its marked repression in young and mature females (Figure 1). Transcripts of all the selected CWMEs were considerably upregulated in the early parasitic stages of M. incognita compared to their marked downregulation in young and egg laying females (Supplementary Figure S4). Taken together, qRT-PCR data demonstrates the association of msp-20 and msp-18 with the early and late stage of RKN parasitism in eggplant. Role of CWMEs in early parasitic stage of RKN was also re-established.

Using the leaf disk-Agrobacterium co-cultivation method A. tumefaciens (LBA4404) harboring the HIGS constructs of msp-18 and msp-20 were used to transform eggplant (Supplementary Figures S1, S2). After kanamycin selection and regeneration, 25 primary transgenic events (T₀) for each gene were generated. Each event was genotyped through PCR and presence of gene-specific, sense, antisense, and selectable marker fragment was detected in 15 and 14 events for msp-18 and msp-20 genes, respectively (Supplementary Figure S5). No significant difference in the morphological characters of untransformed control plants grown on non-selective medium and T₀ plants grown on selection medium was observed (data not shown), indicating that neither the antibiotic marker nor the HIGS construct expression affected the growth of transgenic plants.

A total of 8 and 11 T₀ lines were subjected to Southern blot assay to analyze the integration patterns of msp-18 and msp-20 transgenes, respectively (because in a preliminary assay hybridized fragments were detected only in these events). For msp-18, single copy insertion of T-DNA was observed in event numbers 18.4 and 18.8, while double copy insertion was documented for event numbers 18.3, 18.5, 18.6, and 18.7 (Figure 2A). For msp-20, single copy integration of the transgene was observed in event numbers 20.3, 20.4, 20.6, 20.7, 20.8, 20.9, and 20.10, while double copy integration was recorded in event numbers 20.1, 20.5, and 20.11 (Figure 2B). Henceforth, 6 and 10 events harboring the msp-18 and msp-20 transgenes, respectively, were used for further study.

T₁ progeny plants were generated in the growth chamber by germinating the seeds of selected T₀ events in the presence of antibiotic kanamycin. T₁ plants were genotyped through PCR and expected fragments were detected (Supplementary Figure S6). This indicates the stable integration and inheritance of msp-18 and msp-20 RNAi transgenes in the progeny plants.

Quantitative reverse-transcription PCR assay indicated the overexpression of msp-18 and msp-20 mRNAs in all the selected T₁ events. By contrast, no transcripts corresponding to either of msp-18 or msp-20 were detected in the RNA extracted from untransformed control plants. Hence, the expression data of msp-18 and msp-20 for the transformed lines are simply presented (as ΔCt) relative to the normalizer gene, 18S rRNA. Since Ct value is inversely proportional to expression quantification, significantly (P < 0.05) greater quantitative expression of msp-18 was documented in event numbers 18.4, 18.8 (single copy events), and 18.5 compared to event numbers 18.3, 18.6, and 18.7. Likewise, comparatively greater induction of msp-20 was recorded in single copy events (20.3, 20.4, 20.6, 20.7, 20.8, 20.9, and 20.10) than the double copy ones (20.1, 20.5, and 20.11) (Figure 3A).

As an integral component of HIGS, expression of msp-18 and msp-20 siRNAs in selected transgenic lines was demonstrated by Northern analysis. SiRNAs specific to msp-18 gene were detected in both single (18.4 and 18.8) and double copy (18.5, 18.6, and 18.7) events. Similarly, msp-20 siRNAs were detected in both single (20.3, 20.4, 20.6, 20.7, 20.8, 20.9, and 20.10) and double copy (20.5) events. However, RNAs isolated from event numbers 18.3 (msp-18-specific), 20.1 and 20.11 (msp-20-specific) did not show any hybridization signal with their respective probes (Figure 3B).

T₂ progeny plants were generated from the selected T₁ events through kanamycin selection. T₂ plants were genotyped through PCR and expected fragments were detected (data not shown). In order to characterize the T-DNA integration sites in selected T₂ events, event numbers 18.4 (msp-18-specific, single copy) and 20.7 (msp-20-specific, single copy) were evaluated using genome walking analysis because these events showed the greatest expression of their respective transgenes in qRT-PCR assays. Using event-specific primers fragments flanking the T-DNA region were detected in events 18.4 and 20.7 as well as in their progeny plants. The absence of these fragments in other msp-18- and msp-20-specific events confirmed the independent integration of both the selected events (Supplementary Figure S7). The T-DNA region flanking the sequences of both msp-18 and msp-20 transgenes showed homology with the eggplant genome sequences. The sequence of the flanking PCR product corresponding to event 18.4 showed a maximum of 97% identity with Solanum pennellii chromosome 1 (Sequence ID: HG975440). Whereas the sequence obtained from event 20.7 showed a maximum of 84% homology with the S. melongena draft genome sequence (Sequence ID: Sme2.5_00742.1).

To assess whether HIGS of two target genes confers resistance to M. incognita, 6 and 10 independent T₁ events harboring the RNAi constructs of msp-18 and msp-20, respectively, were evaluated in terms of nematode invasion, development and reproduction in the host roots. At 30 dpi, number of root galls was considerably reduced in different T₁ events compared to wild-type plants. Additionally, T₁ plants have showed more robust root system than the wild-type plants (Supplementary Figure S8). The average number of galls per plant was significantly (P < 0.05) reduced by 30.81–47.29% and 28.22–46.77% in msp-18 and msp-20 RNAi lines, respectively, compared to the wild-type. Accordingly, a marked reduction in the number of egg masses (27.27–49.25% for msp-18 and 22.17–45.86% for msp-20) was recorded in all the events compared to wild-type ones (Figure 4 and Supplementary Table S2). Because each female produces its progeny in a single egg mass, the number of egg masses indicates the identical number of successfully reproducing females. This suggests that the HIGS of msp-18 and msp-20 genes resulted in the reduced root galling due to impeded developmental progression of M. incognita juveniles to females in transgenic plants. In corroboration, fecundity of M. incognita in terms of number of eggs per egg mass was substantially decreased by 29.66–40.20% and 25.21–38.48% in msp-18 and msp-20 RNAi lines, respectively, compared to wild-type plants (Figure 4 and Supplementary Table S2). Considering the importance of MF as determinant of nematode reproductive fitness and parasitic success on a host plant, our results demonstrated that MF was significantly (P < 0.05) decreased by 48.93–69.68% and 41.74–66.69% in msp-18 and msp-20 RNAi lines, respectively, compared to wild-type plants. Among msp-18 RNAi lines, single copy transformants (18.4, 18.8) exhibited significantly (P < 0.05) greater reduction in MF than some of the double copy transformants (18.6, 18.7). Similarly, in case of msp-20 RNAi lines, some of the single copy transformants (20.3, 20.4, 20.7, 20.8, 20.10) showed significantly (P < 0.05) greater reduction in MF than the double copy ones (20.1, 20.5, 20.11). Interestingly, within the single or double copy transformants no significant difference (P < 0.05) could be observed (Figure 4 and Supplementary Table S2). Comparable results were obtained when T₂ plants of both msp-18 and msp-20 RNAi lines were tested for their bioefficacy against M. incognita (Supplementary Figure S9 and Table S2) suggesting that there was no resistance breakdown in the progeny plants. The complete experiment was carried out twice and showed similar results on each occasion.

In addition, perturbed penetration ability of M. incognita J2 at 2 dpi and consequently, significantly (P < 0.05) less number of spike-tail stages (J3/J4) at 7 dpi was documented in the T₁ plants compared to wild-type ones (Figure 5). This indicated the detrimental effect of HIGS of msp-18 and msp-20 genes on the early parasitic stages of M. incognita.

In order to investigate the effect of HIGS on the effector gene abundance in nematodes, qRT-PCR was performed to analyse the possible change in transcript level of CWME (Mi-xyl-1, Mi-xyl-3, Mi-pg-1, Mi-eng-1, and Mi-pel) and pioneer genes (msp-18 and msp-20) in M. incognita females extracted from the roots of T₁ events (18.4 and 20.7) at 20 dpi (event 18.4 and 20.7 were found to be the best transgenic lines in suppressing RKN parasitic ability). Results showed that the mRNA levels of msp-18 and msp-20 were significantly (P < 0.05) downregulated in events 18.4 and 20.7, respectively, compared to females isolated from non-transgenic control plants (Figure 6A). Expectedly, no significant transcriptional alteration was documented for msp-18 and msp-20 in events 20.7 and 18.4, respectively, suggesting the target-specific silencing effect of pioneer genes. Intriguingly, HIGS of pioneer genes had led to the significant (P < 0.05) downregulation of CWME genes (Mi-pg-1 and Mi-pel) in females developing in the events 18.4 and 20.7, compared to control (Figure 6A).

To further confirm the possibility of effector crosstalk among pioneer and CWME genes, qRT-PCR was performed with the early parasitic J2s of M. incognita extracted from the root tip of T₁ plants (18.4 and 20.7) at 8 h post inoculation (in PF-127 medium). Significant (P < 0.05) repression of msp-18 and msp-20 genes in J2s extracted from the events 18.4 and 20.7, respectively, and their non-significant expression vice versa (Figure 6B), have exemplified the target-specific silencing effect of pioneer genes. Concomitantly, significant (P < 0.05) overexpression of Mi-xyl-1 and Mi-xyl-3, and repression of Mi-pg-1 and Mi-eng-1 was documented in J2s infecting plants (18.4) that express msp-18 siRNAs. Similarly, Mi-xyl-1 was significantly (P < 0.05) overexpressed and Mi-pg-1, Mi-eng-1, and Mi-pel were downregulated in J2s isolated from event 20.7 that expresses msp-20 siRNAs (Figure 6B). Taken together, these data reveal that the dsRNA/siRNA molecules corresponding to msp-18/msp-20 genes were ingested by the nematodes as early as during the process of penetration.

In order to validate the hypothesis that nematodes may ingest host-derived molecules during the earliest phase of plant–nematode interaction (i.e., prior to initiation of feeding cell), wild-type eggplants were radiolabelled with carbon isotope (¹⁴CO₂) and inoculated with M. incognita J2 in PF-127 medium for 8 h followed by the scintillation counting of ¹⁴C in penetrating J2s. The autoradiographs revealed that ¹⁴C was incorporated throughout the plant tissue indicating the successful labeling of the plant material (Figures 7A,B). Quantitative measurements based on liquid scintillation counts (after deducting the dpm counts for unlabelled plant material and freshly hatched J2 as control) showed that 66.5, 26.3, and 7.2% of the total label was assimilated in the shoot, root, and J2s, respectively (Figures 7C,D and Supplementary Figure S10). The less intense radioactive signal in the root tissues can be explained by the possibility that a portion of the labeled carbon was lost during the course of the experiment due to respiration in heterotrophic tissues such as root. Nevertheless, detection of a significant proportion of the labeled carbon in J2 suggests that nematodes can ingest the cell sap containing plant-derived molecules while thrusting its stylet to enter into the plant tissue.