Glutathione contributes to plant defence against parasitic cyst nematodes

Abstract Cyst nematodes (CNs) are an important group of root‐infecting sedentary endoparasites that severely damage many crop plants worldwide. An infective CN juvenile enters the host's roots and migrates towards the vascular cylinder, where it induces the formation of syncytial feeding cells, which nourish the CN throughout its parasitic stages. Here, we examined the role of glutathione (l‐γ‐glutamyl‐l‐cysteinyl‐glycine) in Arabidopsis thaliana on infection with the CN Heterodera schachtii. Arabidopsis lines with mutations pad2, cad2, or zir1 in the glutamate–cysteine ligase (GSH1) gene, which encodes the first enzyme in the glutathione biosynthetic pathway, displayed enhanced CN susceptibility, but susceptibility was reduced for rax1, another GSH1 allele. Biochemical analysis revealed differentially altered thiol levels in these mutants that was independent of nematode infection. All glutathione‐deficient mutants exhibited impaired activation of defence marker genes as well as genes for biosynthesis of the antimicrobial compound camalexin early in infection. Further analysis revealed a link between glutathione‐mediated plant resistance to CN infection and the production of camalexin on nematode infection. These results suggest that glutathione levels affect plant resistance to CN by fine‐tuning the balance between the cellular redox environment and the production of compounds related to defence against infection.

Glutathione biosynthesis generally occurs via two ATPdependent steps (Meister, 1995). In the first reaction, a peptide bond forms between the α-amino group of l-cysteine (l-Cys) and the γ-carboxyl of l-glutamate (l-Glu) via a process catalysed by glutamate-cysteine ligase (GSH1). In the second step, glutathione synthetase (GSH2; Hell & Bergmann, 1990) adds glycine (Gly) to the dipeptide γ-glutamyl-cysteine (γ-EC) to form GSH. In Arabidopsis thaliana, these two enzymes are encoded by single-copy genes; GSH1 is exclusively localized to chloroplasts, whereas GSH2 is predominantly located in the cytosol and to a lesser extent in plastids (Pasternak et al., 2008;Wachter et al., 2005). Under stress conditions, reduced glutathione is rapidly converted into the oxidized form glutathione disulphide (GSSG), leading to an imbalance in the glutathione redox potential (E GSH ), which is considered an indicator of oxidative stress. The accumulation of GSSG is counteracted by glutathione reductases, which reduce GSSG back to GSH at the expense of electrons provided by NADPH (Marty et al., 2009(Marty et al., , 2019. Glutathione also functions as a co-substrate in conjugation and detoxification reactions catalysed by glutathione transferases (GSTs), and in glutathionylation of protein thiols as a posttranslational protein modification (Dixon et al., 2005). In addition, glutathione acts in plant perception systems to activate basal defence responses that help counter microbial attack (Glazebrook & Ausubel, 1994;Parisy et al., 2007). In coordination with cysteine, glutathione plays a role in the establishment and signalling of pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI) (Alvarez et al., 2011;Jones & Dangl, 2006). GSH1 is thought to be the rate-limiting enzyme for GSH synthesis (Arisi et al., 1997). Null mutants of GSH1 in Arabidopsis are embryo-lethal (Cairns et al., 2006). Less severe mutations within this gene appear in five distinct mutant alleles (rml1, rax1, pad2, cad2, and zir1) that were identified through forward genetic screens and found to show a partial decrease in glutathione production (Ball et al., 2004;Bangash et al., 2019;Cobbett et al., 1998;Glazebrook & Ausubel, 1994;Shanmugam et al., 2012;Vernoux et al., 2000). Studies of these GSH1-deficient mutants have shed light on the diverse roles of glutathione in many cellular processes, including plant development and responses to abiotic and biotic stimuli (Foyer & Noctor, 2011;Noctor, 2006;Potters et al., 2002).
The Arabidopsis root-meristemless1 (rml1) mutant possesses <5% of wild-type levels of foliar glutathione and exhibits severe defects in plant development (Vernoux et al., 2000). Another Arabidopsis GSH1 mutant, zinc tolerance induced by iron 1 (zir1), contains <15% of wild-type levels of glutathione in leaves, leading to deficits in Fe-dependent Zn tolerance and impaired nitric oxide (NO)mediated Fe-deficiency signalling (Shanmugam et al., 2012(Shanmugam et al., , 2015. The cadmium-sensitive2 (cad2) mutant contains c.30% of wild-type glutathione levels and shows sensitivity to cadmium and enhanced susceptibility to Phytophthora brassicae (Cobbett et al., 1998;Howden et al., 1995;Parisy et al., 2007). The regulator of ascorbate peroxidase2 1 (rax1) mutant contains c.40% of wild-type levels of glutathione in leaves and shows increased sensitivity to light stress (Ball et al., 2004). However, the altered glutathione content in rax1 does not appear to affect plant resistance to Pseudomonas syringae or P. brassicae (Parisy et al., 2007). The phytoalexin defi-cient2 (pad2) mutant, the most extensively studied mutant in an allelic series of GSH1 mutants, possesses only c.20% of wild-type levels of glutathione in leaves and shows enhanced susceptibility to many pests and pathogens (Dubreuil-Maurizi & Poinssot, 2012).
Despite substantial evidence for the role of glutathione in different plant pathosystems, little is known about how glutathione mediates the molecular dialogue during plant-nematode interactions.
Plant-parasitic cyst nematodes (CNs) are among the most damaging plant pests and pathogens, causing substantial yield losses globally (Savary et al., 2019). Infective-stage juveniles (J2) of CNs hatch from eggs on stimulation by mostly unknown host triggers and migrate toward the roots. The CNs make numerous perforations in the epidermal cell wall via back-and-forth movements with their stylets and invade the roots of the host plant near the root tip.
Subsequently, they migrate intracellularly through the root cortical cells to the vascular cylinder. On reaching the vascular cylinder, the nematodes probe single cambial or procambial cells to induce the formation of an initial syncytial cell (ISC) as a feeding site (Sobczak et al., 1999). The nematodes inject a cocktail of secretions into the ISC through their hollow stylets to modify plant morphogenetic pathways towards the development of the feeding site (Hewezi & Baum, 2013;Wyss & Zunke, 1986). Hundreds of adjacent root cells successively fuse with the ISC via local cell wall dissolution to form a hypertrophied, multinucleate, hypermetabolic syncytial nurse cell (Wyss & Grundler, 1992).
The juvenile nematode rapidly becomes sedentary due to cellspecific muscle atrophy and starts feeding on the syncytium, which acts as a nutrient sink throughout its parasitic stages (Han et al., 2018). Syncytium development is accompanied by extensive metabolic, transcriptomic, and proteomic changes in the infected root tissues (Hofmann et al., 2010;Hütten et al., 2015;Siddique et al., 2009;Szakasits et al., 2009). The juveniles feed, enlarge, and moult three times to differentiate into males or females. Female nematodes grow rapidly and burst through the root surface, while males regain a vermiform body shape and mobile form by remodelling their neuromuscular structures, leave the roots, and search for females (Han et al., 2018). The female dies after fertilization, and the body wall tans to form a typical brown cyst that envelops and shields the next generation of eggs. The eggs are able to survive for prolonged periods (up to 20 years) in the soil until a suitable host is found growing nearby (Grainger, 1964 (Sijmons et al., 1991). Here, using the Arabidopsis and H. schachtii model system, we investigated the role of glutathione in orchestrating host defence responses to CN infection. We conclude that glutathione contributes to plant defence against CN infection through modulation of cellular redox homeostasis and camalexin production in host roots.

| Cyst nematode infection activates GSH biosynthesis genes in Arabidopsis
To investigate the roles of glutathione biosynthetic genes during different phases of Arabidopsis parasitism by CN, we examined the expression patterns of GSH1 and GSH2 in publicly available transcriptomic data Siddique et al., 2021;Szakasits et al., 2009). These surveys revealed a significant increase in GSH1 expression during the migratory (10 h postinoculation, hpi) and sedentary (5 and 15 days postinfection, dpi) stages of cyst nematode infection (Table S1 and Figure S1). In comparison, although GSH2 transcript abundance increased significantly during the migratory stage, it remained unchanged during the sedentary stage of infection.
To validate these microarray data, we analysed the expression of GSH biosynthesis genes in Arabidopsis root segments infected with CN. We collected several hundred root segments (c.0.2 cm) containing infection sites at 10 hpi (migratory stage) and syncytia at 10 dpi (sedentary stage), and analysed GSH1 and GSH2 expression in these tissues compared with uninfected wild-type Columbia (Col-0) roots by reverse transcription-quantitative PCR (RT-qPCR). The results confirmed that GSH1 transcript levels increased during both the migratory and sedentary stages of CN infection (Figure 1a), whereas GSH2 was significantly induced at 10 hpi but not at 10 dpi ( Figure 1b). Overall, these results suggest that GSH biosynthesis genes are activated on CN infection, and that the regulation of GSH1 is more pronounced than that of GSH2 during CN infection; therefore, we focused our further analysis on GSH1.

| Altered glutathione levels in plants influence cyst nematode infection and development
To assess whether glutathione levels in Arabidopsis affect CN parasitism, we screened an allelic series of GSH1 and mutants in nematode infection assays by measuring multiple nematode susceptibility parameters. We grew plants under sterile conditions in agar medium, and when the roots spread through the agar we inoculated them with 60-70 J2 nematodes per plant. At 14 dpi, we counted the average number of females, a widely accepted parameter for nematode susceptibility under in vitro conditions. The average number of females per centimetre of root length was significantly higher in zir1 and cad2 than in the Col-0 control ( Figure 2a). However, rax1 and pad2 plants showed no difference in the number of females as compared to Col-0 ( Figure 2a). Next, we measured the size of female nematodes and female-associated syncytia at 14 dpi. There was no significant difference in either parameter in any lines carrying allelic mutations in GSH1, except for rax1, which, surprisingly, displayed a slight decrease in the size of female-associated syncytia relative to the wild type (Figure 2b,c). Finally, we compared cyst egg contents and cyst size at 42 dpi. We detected substantially more eggs per cyst in pad2, cad2, and zir1 than Col-0, but not in rax1 plants ( Figure 2d). However, the cyst size was unaffected in all gsh1 mutants examined ( Figure 2e).

| Changes in the susceptibility of glutathionedeficient mutants are unrelated to CN attraction
To elucidate whether the changes in the susceptibility of the GSH1 mutants to H. schachtii were associated with the attractiveness of F I G U R E 1 Glutathione (GSH) biosynthesis genes are induced by infection with Heterodera schachtii. Validation of changes in expression of genes coding for glutathione biosynthesis enzymes GSH1 (a) and GSH2 (b) during the migratory and sedentary stages of nematode infection via reverse transcription-quantitative PCR (RT-qPCR). Values represent relative expression levels on CN infection, with the value of the first replicate in uninfected control roots (Col-0) set to 1. The mRNA levels were measured in three technical replicates per sample. The transcript level of each gene was normalized to that of the Arabidopsis housekeeping gene 18S rRNA and UBQ10 with three biological replicates each (blue circle). Data are presented as the mean ± SE. Each biological replicate contained a pool of hundreds of small root segments with infection sites at 10 h postinoculation (hpi) ('Migratory') or syncytia at 10 days postinoculation (dpi) ('Sedentary'). Data were analysed with a two-tailed Student's t test (p < .05) and asterisks represent significant differences compared to the uninfected control root the plant roots, we performed nematode attraction assays using agar discs containing root exudates from 12-day-old mutants and wild-type control plants. Agar discs containing root exudates of Col-0 plants attracted more J2s than control agar discs (without roots), indicating that nematodes were able to sense the signal from the root exudates ( Figure S2). However, we observed no significant difference in attractiveness towards the root exudates for any GSH1 mutant compared with control plants ( Figure S2). Thus, the changes in the glutathione-deficient mutants' susceptibility to CN infection are probably not associated with any factor(s) mediating the plants' attractiveness to nematodes.

| Thiol levels are altered in GSH1 mutants, regardless of CN infection
Arabidopsis plants harbouring mutant alleles of GSH1 contain constitutively reduced foliar glutathione levels (Parisy et al., 2007;Shanmugam et al., 2012). We explored the extent to which glutathione accumulation is affected in roots due to mutations in GSH1. For this purpose, we collected uninfected root tissues from 12-day-old plants and measured glutathione levels. We observed a similar trend in root glutathione levels in GSH1 mutants as in the foliar levels reported previously (Parisy et al., 2007; Shanmugam F I G U R E 2 Effect of glutathione (GSH) biosynthesis mutations on the susceptibility of Arabidopsis to Heterodera schachtii. (a) Nematode infection assay showing the number of females/cm root length of GSH1 mutant lines versus wild-type Col-0. Approximately 60-70 H. schachtii infective juveniles were inoculated onto 12-day-old Arabidopsis plants. The number of females per root system was counted at 14 days postinoculation (dpi), and the infection rate per centimetre of root length was determined after scanning the roots using the WinRhizo root image analysis system. (b) Average size of female nematodes at 14 dpi. Approximately 30-40 female nematodes were randomly selected and their outlines measured for each biological replicate. (c) Average size of syncytia at 14 dpi. Approximately 30-40 female-associated syncytia in each replicate were measured using a stereomicroscope. (d) Average number of eggs per cyst at 42 dpi. For each biological replicate, 40-45 cysts were randomly selected and crushed between slides, their contents transferred to a counting dish, and the number of J2s/eggs counted under a stereomicroscope. (e) Average cyst size (50-55 per repetition) at 42 dpi. Bars represent mean ± SE. All data were analysed using one-way analysis of variance followed by Tukey's HSD post hoc test. Different letters indicate significantly different means (95% confidence). Data represent one of three independent experiments with similar results et al., 2012), but these levels differed: rax1 had the highest glutathione level (54% of Col-0), followed by pad2 (46%), cad2 (31%), and zir1 (30%; Figure 3a). Next, we measured glutathione levels in root pieces with infection sites at 10 hpi (migratory stage) and 10 dpi (sedentary stage). Glutathione levels were significantly higher in infected Col-0 at 10 hpi than in uninfected control roots at 10 hpi. However, glutathione returned to uninfected control levels at 10 dpi. In contrast to Col-0, no increase was detected in glutathione levels in the roots of any of the GSH1 mutant on nematode infection (Figure 3a).
Cysteine functions as a metabolic precursor for numerous essential biomolecules, such as glutathione, glucosinolates, and phytoalexins (Meyer & Hell, 2005;. Because

| Glutathione-depleted mutants have impaired basal defence to CN
We hypothesized that the hypersusceptibility of GSH-deficient mutants might be due to impaired expression of genes in defencerelated pathways. Therefore, we assessed the transcript abundance of following six plant basal defence marker genes that are strongly up-regulated during the migratory stage of infection Shah et al., 2017): NPR1, a salicylic acid signalling gene (Yan & Dong, 2014); ACS2, which is involved in ethylene signalling (Yamagami et al., 2003); JAZ10, a jasmonic acid signalling gene (Chini et al., 2007); and three genes associated with indole-glucosinolate and camalexin biosynthesis pathways, namely, the genes encoding GSTF6 (mediates the conjugation between indole-3-acetonitrile stages were collected for measurement of glutathione (a) and cysteine (b) contents by high-performance liquid chromatography (HPLC). Data points represent four independent experiments (means ± SE). Different lowercase letters denote significant differences, as determined by analysis of variance (p < .05) followed by Tukey's HSD post hoc tests. C, control; M, migratory; S, sedentary those of the GSH1 mutants ( Figure S3). Next, we sampled small root segments (c.0.2 cm) containing infection sites at the migratory stage of CN infection (10 hpi) and analysed gene expression by RT-qPCR.
As previously observed Shah et al., 2017), expression of all tested genes was significantly increased in Col-0 on infection compared with uninfected control roots (Figure 4a-f). In comparison to Col-0, normal increase in transcript levels of all tested genes was impaired in pad2, cad2, and zir1 (Figure 4a-f). Interestingly, rax1showed a normal or even more pronounced increase in defence marker gene expression on infection (Figure 4a-f). Together, these findings suggest that glutathione is involved in activation of plant defence responses, including indole-glucosinolate and camalexin biosynthesis pathways, on CN infection in Arabidopsis roots.

| CN parasitism causes shifts in the redox state of glutathione
The glutathione-dependent fluorescent probe Grx1-roGFP2 is commonly used to monitor E GSH at the subcellular level in vivo (Meyer et al., 2007;Müller-Schüssle et al., 2021;Schwarzländer et al., 2008). To assess whether CN parasitism causes changes in E GSH , we performed ratiometric analysis of Grx1-roGFP2 (Gutscher et al., 2008;Wagner et al., 2019) expressed in the cytosol via confocal microscopy following excitation at 405 and 488 nm. We grew plants in 35-mm Petri dishes with a 14-mm glass microwell and inoculated them with 50 J2s of CNs. For calibration, we exposed uninfected roots to reducing (10 mM dithiothreitol [DTT]) and oxidizing (5 mM 2,2′-dipyridyldisulfide [DPS]) reagents. We found that the 405/488 nm fluorescence ratio in Col-0 roots was slightly increased at 10 hpi, indicating a less negative E GSH (Figure 5a).
Together with the increase in total GSH during the migratory state ( Figure 3a) this result suggests a slight oxidation of the glutathione pool (Figure 5a). A similar trend albeit with larger variance towards an even more pronounced oxidation was found for rax1 ( Figure 5a). In both pad2 and cad2 under noninfected control conditions, an increased ratio indicated a less reducing E GSH , which was expected from the pronounced decrease in total glutathione (Meyer et al., 2007). While cad2, which also displayed enhanced susceptibility to H. schachtii, showed no change in Grx1-roGFP2 F I G U R E 4 Reduced glutathione levels lead to impaired basal defence responses. (a-f) Expression of defence marker genes in response to Heterodera schachtii infection during the migratory stage in roots of glutathione (GSH)deficient mutants. Several hundred root segments (c.0.2 cm) containing infection sites were collected at 10 h postinoculation (hpi) and the expression levels of NPR1, JAZ10, ACS2, CYP81F2, GSTF6, and PAD3 were analysed by reverse transcription-quantitative PCR (RT-qPCR). Expression levels of the genes of interest were normalized to the Arabidopsis housekeeping gene 18S rRNA. Data represent relative expression levels of the indicated genes, with the value for uninfected roots of each genotype set to 1. Asterisks represent significant difference in expression levels from the corresponding uninfected control roots. Data represent means ± SE (n = 3 biological replicates, *p < .05, **p < .01, ***p < .001, using two-tailed Student's t test). C, control; M, migratory stage of infection; ns, nonsignificant oxidation on infection, pad2 surprisingly showed a more reducing E GSH after infection (Figure 5a).

| GSH-mediated camalexin levels in roots determine plant susceptibility to CN infection
To explore whether there is a link between glutathione and nematodeinduced camalexin biosynthesis, we used high-performance liquid chromatography (HPLC) to measure camalexin levels in root segments at 10 hpi and in uninfected roots. In uninfected roots, camalexin levels were low and no significant differences in camalexin content were detected between Col-0 and glutathione-deficient mutants. However, we observed an almost 200-fold increase in camalexin content on CN infection in Col-0 ( Figure 6). Interestingly, an even more pronounced increase was found in the camalexin level in rax1 roots on CN infection ( Figure 6). However, in both cad2 and zir1 roots, the increase in camalexin levels was significantly lower than for Col-0 following CN infection, and in pad2 roots induction was the same (Figure 6).

| DISCUSS ION
In the present work, we use Arabidopsis GSH biosynthesis mutants to examine how glutathione contributes to plant susceptibility to CN infection. We determined that the expression of GSH1 and GSH2 is significantly increased in response to CN infection during the migratory stage of CN infection. CNs cause extensive damage to host root tissues during their migration inside the roots (Shah et al., 2017). In turn, reactive oxygen species (ROS) production and the detoxification machinery are simultaneously switched on to fine-tune the activation of a network of plant defence responses (Ithal et al., 2007;Mazarei et al., 2011;Molinari & Miacola, 1997;Siddique et al., 2014).
Increased expression of GSH1 and GSH2 therefore may be a response to the high demand for glutathione to re-establish ROS homeostasis after an infection or for synthesis of the sulphur-containing phytoalexin camalexin during the migratory stage of infection (Parisy et al., 2007;Siddique et al., 2014).
While null mutants of GSH1 in Arabidopsis are embryo-lethal (Cairns et al., 2006), five distinct mutant alleles (rml1, rax1, pad2, cad2, and zir1) show a partial decrease in glutathione production (Ball et al., 2004;Bangash et al., 2019;Cobbett et al., 1998;Glazebrook & Ausubel, 1994;Shanmugam et al., 2012;Vernoux et al., 2000). Of these, rml1 contained <5% of the wild-type levels of glutathione and showed aborted root growth (Vernoux et al., 2000). Because this severe phenotype compromised the analysis of rml1, we analysed plant-CN interactions in the four other GSH1 mutants (rax1, pad2, cad2, and zir1). We found that glutathione content increased signifi- in vivo, we found a more oxidized probe during the migratory stage of nematode infection in Col-0. Interestingly, rax1 mutants showed an oxidative shift of E GSH , as Col-0, and partially retained induction of defence gene expression. In contrast, this oxidative E GSH shift at 10 hpi was absent in the more susceptible cad2 and pad2 plants. This finding suggests that glutathione deficiency interferes with pathogentriggered signalling events that are crucial for successful nematode parasitism, leading to increased plant susceptibility to CN.
The importance of GSH for camalexin accumulation and for disease resistance has been previously demonstrated (Glazebrook & Ausubel, 1994;Parisy et al., 2007). Furthermore, loss-of-function camalexin biosynthesis mutants have been shown to display enhanced susceptibility to CN (Ali et al., 2014;Shah et al., 2017). Here, we found that the up-regulation of two key camalexin biosynthesis genes in response to CN infection was impaired in GSH-deficient mutants as compared with Col-0 control plants. Furthermore, camalexin accumulation in response to CN infection was significantly lower in the roots of cad2 and zir1 plants. Based on these data, we propose that an adequate level of GSH is required for camalexin accumulation, which in turn plays a role in the plant's defence against CN infection.
Surprisingly, the rax1 mutant showed a different biochemical phenotype than the other GSH1 mutants, with significantly higher levels of camalexin during early nematode parasitism. Moreover, the cellular redox state was more oxidized in rax1 than in the other mutants, pointing to oxidative stress conditions, which might ultimately influence camalexin accumulation in these plants. In our experiments, rax1 was noted to contain 54% of the wild-type level of glutathione, the highest amount produced by any GSH1 mutant, indicating that it is able to utilize cysteine to maintain glutathione biosynthesis at a level sufficient to resist nematode invasion.
Indeed, rax1 showed a significant reduction in the size of femaleassociated syncytia. Moreover, rax1 shows increased ASCORBATE PEROXIDASE2 (APX2) expression in response to wounding (Ball et al., 2004). Loss-of-function apx2 mutants are compromised in ROS production during oxidative stress (Suzuki et al., 2013). APX2 expression is probably triggered by H. schachtii during its destructive invasion and migration, leading to excessive ROS accumulation in the roots of rax1 plants. Although the rax1 mutant displays lower ROS accumulation under nonstress conditions, and CNs can utilize ROS for successful parasitism, the mutant's overproduction of ROS on nematode infection (caused by the misregulation of the ascorbateglutathione cycle, resulting in oxidative stress and enhanced camalexin production) might hamper nematode development (Ball et al., 2004;Noctor & Foyer, 1998;Siddique et al., 2014;Tierens et al., 2002;Zhao et al., 1998).
In summary, our findings suggest that the precise regulation of glutathione homeostasis is crucial for mounting an effective plant defence response to CN infection. We propose that glutathione depletion is positively correlated with camalexin accumulation in roots on CN infection. Accumulation of camalexin in turn perturbs the cellular redox state and activates signalling events that negatively influence plant susceptibility to CNs.

| Plant material and growth conditions
A. thaliana seeds were disinfected by washing in 2% sodium hypochlorite (wt/vol) for 3 min, followed by washing with 70% (vol/vol) ethanol for 5 min and rinsing three times consecutively with sterile water. After being dried on sterile Whatman filter paper for 2-4 h, seeds were stored at 4°C before plating. Sterilized seeds were sown in Petri dishes with agar medium enriched with modified Knop's nutrient solutions as previously described (Sijmons et al., 1991). Plants were grown under long-day conditions with 16 h of light and 8 h of darkness in a growth chamber at 23°C for CN infection .

| Validation of mutant lines
The mutant lines pad2, cad2, rax1, and zir1 used in our study were validated by PCR. The area surrounding the predicted point mutation was amplified by PCR using the primers listed in Table S2  Cyst sizes were measured at 42 dpi.

| Nematode attraction assay
Nematode attraction assays were conducted on all glutathionedeficient mutants, according to Dalzell et al. (2011), with some modifications. Uniform circular counting wells 6 mm in diameter, attached through cylindrical tunnels (2.5 mm depth × 20 mm length), were constructed in a 2% (wt/vol) water agar plate. The cylindrical channels were created by placing a 20-mm long plastic tube constructed from the handle of a regular inoculation loop onto the agar surface immediately after pouring. With the aid of fine forceps, the tubular plastic was removed once the medium had solidified. The wells were cut with a small transfer glass pipette at either side of the central channels to create a 6-mm diameter well. Agar plugs were excised from the plate close to the roots carrying root exudates from Col-0 and mutant plants raised on Knop medium and subsequently transferred into the counting wells. Around 60-80 J2s were put in the centre of the cylindrical linking channel and kept at room temperature in a dark place. Over a 2-h period, the number of nematodes that moved to either one or the other well were counted and considered as attracted by the exudates of the respective agar plug. Experiments were repeated three times independently with six plates each (n = 18). The attraction rate (%) was calculated from the total number of nematodes applied.  The qPCR analysis was conducted using a StepOne Plus Real-Time PCR (Applied Biosystems) system based on a two-step amplification protocol, with the following cycling conditions: 95°C for 3 min followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. For each sample run, a melt-curve analysis was done following 95°C for 15 s, 65-95°C with 0.5°C incremental progress yielding a single peak. As a negative control, a water-containing nontemplate reaction was included. The transcript abundance of targeted genes was computed from three biological replicates per treatment, with three technical replicates for each biological sample. Relative expression was calculated by normalizing target gene expression to the abundance of the Arabidopsis housekeeping genes 18S rRNA and UBQ10 (Pfaffl, 2001).

| RT-qPCR
The gene-specific primer pairs used for RT-qPCR are provided in Table S2.

| Confocal laser scanning microscopy imaging and ratiometric analysis
Confocal laser scanning microscopy (CLSM) imaging and ratiometric analyses in Col-0, rax1, pad2, and cad2 root tissues expressing Grx1-roGFP2 were conducted according to Schwarzländer et al. (2008). Plants were grown in Knop medium (0.2% wt/vol) in 35-mm Petri dishes with a 14-mm glass microwell (Mattek) and 10-day-old seedlings were inoculated with 50 J2s of CNs. While we were able to grow and inoculate all gsh1 alleles, zir1 showed extremely poor growth under these conditions. We therefore excluded zir1 from roGFP experiments.
The ratio of fluorescence intensity after excitation at 405 and 488 nm of the glutathione redox potential sensor Grx1-roGFP2 was examined in uninfected root tissues and nematode infection sites at 10 hpi. The dynamic range of the sensor (405/488 nm ratio) was determined by in situ calibration (0.5-2.0) and was set as the minimum/maximum value of the ratio false colour scale. For calibration, 5 mM DPS solution as oxidizing and 10 mM DTT solution as reducing agents were used. Images were captured using a confocal microscope (LSM 780; Zeiss) with a 25× lens using the multitrack mode, with line switching and averaging two frames. The fluorescence of Grx1-roGFP2 was collected at 505-530 nm. RRA v. 1.2 software (Fricker, 2016)

| Statistical analysis
All data were analysed using the statistical software GraphPad Prism v. 8.4.3 for Windows. Infection assays were carried out with 20-30 plants for each genotype and replicated at least three times independently. The normality of data was assessed using the Kolmogorov-Smirnov test/Shapiro-Wilk test (α < 0.05). The specific statistical method used for the data set of each experiment is described in the

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