We used the M. truncatula Jemalong-A17 genotype. M. truncatula seeds were scarified according to Djébali et al. (2009). After stratification overnight at 4°C, they were germinated in phytochambers with 16 h light under 350 μmol m^-2 s^-1 photons at 23°C/8 h night at 21°C. One day after germination, the seedlings were transferred to 12 cm × 12 cm square Petri dishes containing modified M medium (Bécard and Fortin, 1988). This modified medium was sugar-free, enriched in phosphate (1.3 mM final concentration), and contained either 3.2 mM nitrate (complete medium) or no nitrate (NØ medium). The Petri dishes were sealed with parafilm and the roots were protected from light with aluminum foil, and then placed vertically in the culture chamber (16 h light under 350 μmol m^-2 s^-1 photons at 23°C/8 h night at 21°C) for 7 days. The strain Aphanomyces euteiches Drechs ATCC 201684 was used to inoculate the seeds one day after germination. Zoospores were produced as described in Rey et al. (2013), and each root was inoculated with 500 zoospores.

The pENTR4 vector carrying the MtNR1 or the MtNR2 fragment (Horchani et al., 2011) was recombined with the pK7GWIWG2d vector using LR clonase II enzyme mix (Invitrogen, France) to create RNA interference expression vectors. The MtGSNOR gene (M. truncatula Gene code Medtr7g099040) (1,143 bp) was amplified using M. truncatula cDNA as a template and the specific primers GSNOR-F 5′-AAAAAGCAGGCTTCACCATGGCATCGTCGACTGAAGGT-3′ and GSNOR-R 5′- AGAAAGCTGGGTGTCAATGCAATGCAAGCACAC containing the corresponding attB recombination sites. The PCR product was recombined into the pDONR entry vector (Invitrogen) and checked by sequencing. The pDONR vector carrying the MtGSNOR gene was recombined with pK7WG2d plasmids¹ to create the overexpression vector. The constructs pK7GWIGW2d-MtNR1-2/GFP (RNAi::MtNIA1/2) and pK7WG2d-MtGSNOR/GFP (35S::GSNOR) were introduced into A. rhizogenes strain Arqua1 (Quandt et al., 1993). M. truncatula plants were transformed with A. rhizogenes according to Boisson-Dernier et al. (2001). Control plants were transformed with A. rhizogenes containing the pK7GWIGW5D or the pK7WG2d empty vectors. Hairy roots were selected based on the fluorescent marker GFP 21 days after transformation.

Total RNA was extracted from transformed roots using TRIzol^® Reagent (Life Technologies) according to the manufacturer’s recommendations. To carry out the qPCR reaction, RNAs (0.5–1 μg) were reverse-transcribed in a final volume of 20 μL in the presence of RNasin (Promega, Charbonnières, France), and oligo(dT)₁₅, with M-MLV reverse transcriptase (Promega, Charbonnières, France), as recommended by the manufacturer.

Quantitative PCR was performed on reverse-transcribed RNAs from four independent biological replicates per condition and from two independent plant cultures. Quantitative PCR reactions were performed in an ABI PRISM 7900 sequence detection system (Applied Biosystems^®, Saint-Aubin, France), in a final volume of 15 μL containing Absolute SYBR green ROX Mix (Thermo Scientific, Surrey, UK), 0,3 μM of gene-specific primers, and 5 μL of cDNA template diluted 60-fold. The reference gene used for normalization was MtEF1α. Relative expression was expressed as 2^{-ΔCt test genes-reference gene}. The primers used for the qPCR all displayed a high amplification efficiency (90–100%). They were the following:

Roots were cultured on Shb10 medium (Boisson-Dernier et al., 2001) and transferred on modified Fahraeus medium enriched in ammonitrate (1 mM NH₄NO₃ final) one day before inoculation. Inoculation of the root cultures with A. euteiches strain ATCC 201684 was carried out by adding 10 mL of an A. euteiches zoospore suspension containing 80,000 zoospores.mL^-1 in sterilized Volvic (Colditz et al., 2007) water. Zoospore production was initiated as described in Rey et al. (2013). Control root cultures were inoculated with 10 mL of sterile Volvic water. After 4 h of incubation in the dark, the zoospore solution was drained off the roots, and the Petri dishes were placed back into the growth room and left there for 7 days in the dark.

Assessment of A. euteiches development in roots was performed by ELISA, using rabbit polyclonal serum raised against A. euteiches, and a mouse anti-rabbit IgG alkaline phosphatase conjugate as described by Slezack et al. (1999), on protein extracts from roots from pooled plants. Alkaline phosphatase activity was monitored by recording the increase in absorbance at 405 nm for 2–3 h, and was expressed as the slope of the resulting curve per mg of root fresh weight.

H₂O₂ concentration was measured using an Amplex Red^®/peroxidase-coupled fluorescence assay adapted from Ashtamker et al. (2007). Roots were ground on ice and in the dark, in 1 mL of KRPG buffer (145 mM NaCl; 5.7 mM K₂HPO₄; 4.86 mM KCl; 0.54 mM CaCl₂; 1.22 mM MgSO₄; 5.5 mM glucose; pH 7.35) with 10 μM Amplex Red^® and 0.2 U/mL of Horse Radish Peroxidase (HRP) per 100 mg of fresh weight. Catalase, an H₂O₂ scavenger, was used as a control. After 10 min of incubation at 4°C with catalase (1 unit/μL), 10 μM Amplex Red^® and 0.2 U/mL of HRP were added to the samples. After centrifugation (10,000×g, 15 min, 4°C), 100 μL of supernatant were used to quantify resorufin (λ_ex = 560 nm; λ_em = 584 nm) by spectrofluorimetry (Mithras, Berthold Technology). The relative fluorescence units were converted into μmol of H₂O₂ mg^-1 root fresh weight on the basis of a standard curve established from known concentrations of H₂O₂.

ONOO^- and NO concentrations were determined using A17 or transformed roots ground on ice and in the dark, with 1 mL of Tris-HCl (10 mM, pH 7.5), KCl (10 mM) buffer with 5 μM aminophenyl fluorescein (APF) or 10 μM 4,5-diaminofluorescein (DAF), respectively, per 100 mg of fresh weight. Epicatechin, an ONOO^- scavenger, was used as a control. After 10 min of incubation at 4°C with epicatechin (1 mM), APF was added to the samples at a final concentration of 5 μM. cPTIO, an NO scavenger, was used as a control. After 10 min of incubation at 4°C with cPTIO (500 μM), DAF was added to the samples at a final concentration of 10 μM.

After centrifugation (10,000×g, 15 min, 4°C), 100 μL of supernatant were used to quantify ONOO^- or NO (λ_ex = 485 nm; λ_em = 535 nm) by spectrofluorimetry (Mithras, Berthold Technology).

S-nitrosothiol quantification was performed using the Saville–Griess assay (Gow et al., 2007). A17 roots or transformed roots were ground, on ice and in the dark, in extraction buffer (1 mL/100 mg of fresh weight, 0.1 M Tris-HCl, pH 7.5; 1 mM PMSF). After centrifugation (10,000×g, 15 min, 4°C), 100 μL of supernatant were incubated with 100 μL of buffer A (0.5 M HCl; 1% sulfanilamide) or 100 μL of buffer B (0.5 M HCl; 1% sulfanilamide; 0.2% HgCl₂). After incubation (15 min at room temperature), 100 μL of Griess reagent[(0.5 M HCl; 0.02% N-(1-naphtyl)-ethylenediamine dihydrochloride] were added. After 15 min, SNOs were quantified by measuring absorbance at 540 nm. A standard curve was obtained using different concentrations of GSNO.

Nitrate determination was performed according to Miranda et al. (2001), based on the reduction of nitrate to nitrite by vanadium and colorimetric detection at 540 nm of nitrite in the presence of sulfanilamide and N-1-naphthylethylenediamine. Approximately 100 mg of 7-day-old plant roots were collected, flash-frozen in liquid N₂, and ground into powder. Three hundred micro liter of ultra-pure water were added to 20 mg of frozen sample, thoroughly vortexed, and incubated with occasional mixing for 15 min on ice. After centrifugation 15 min at 13,000×g and 4°C, the supernatant was recovered and used for nitrate determination.

Transformed root samples were frozen in liquid nitrogen and ground using pestle and mortar. Extraction was performed in MOPS buffer (1 mL per 100 mg of fresh weight, 50 mM MOPS-KOH buffer, pH 7.6; 5 mM NaF; 1 μM Na₂MoO₄; 10 μM FAD; 1 μM leupeptin, 0.2 g/g FW polyvinylpolypyrrolidone; 2 mM β-mercaptoethanol; 5 mM EDTA). After centrifugation (20,000×g, 5 min, 4°C), the supernatant was used to measure NR activity. The reaction mixture consisted of 50 mM MOPS-KOH buffer, pH 7.6, containing 1 mM NaF, 10 mM KNO₃, 0.17 mM NADH, and 5 mM EDTA. After incubation 15 min at 30°C, the reaction mixture was stopped by adding an equal volume of sulfanilamide (1% w/v in 3 N HCl) followed by N-naphtylethylenediamine dihydrochloride (0.02%, w/v), and the A₅₄₀ was measured. A standard curve was obtained based on different concentrations of nitrite.

To measure GSNOR activity, roots were ground in liquid nitrogen and proteins were extracted in 50 mM Tris-HCl buffer, pH 8, 0.5 mM EDTA, and 1 mg/mL of a protease inhibitor cocktail (1 mL of buffer per 100 mg of fresh weight). GSNOR activity was assayed from the rate of NADH oxidation by measuring the decrease in absorbance at 340 nm at 25°C using 25 μg of proteins in a total volume of 200 μL of extraction buffer containing 350 μM NADH with or without 350 μM GSNO. GSNO reductase activity was determined by subtracting NADH oxidation values in the absence of GSNO from values in the presence of GSNO. All samples were protected from light during the assay and tested for linearity. A standard curve was obtained using different concentrations of NADH.

Statistical analyses were performed using one- or two-way analysis of variance (ANOVA) followed by Fisher’s test. Data were considered as significantly different when p < 0.05.

To investigate the putative role of NO homeostasis in the M. truncatula/A. euteiches interaction, roots were transformed to inactivate the NR-encoding MtNIA1/2 genes or to overexpress GSNOR-encoding genes. Quantification of gene transcripts in transformed roots using RT-qPCR confirmed that the two NIA genes were repressed (Figure 1A) while GSNOR was overexpressed (Figure 1B). To perform functional validation of the different constructs, we quantified NO and SNO levels in transformed roots. The two genetic manipulations modulated NO or SNO levels (Figure 1). SNO levels remained unchanged in RNAi::MtNIA1/2 roots as compared to the controls, whereas NO levels clearly decreased (Figure 1A). This was in accordance with the downregulation of NR, a major enzymatic source of NO. Conversely, NO levels in the 35S::GSNOR roots did not significantly change, but SNO significantly increased as compared to control roots (Figure 1B). This was surprising because in most previous experiments a negative correlation was described between SNO levels and GSNOR activity (Feechan et al., 2005; Rusterucci et al., 2007; Yun et al., 2011). However, it is interesting to note that in pea (a legume closely related to M. truncatula), higher SNO levels induced by wounding were correlated with higher GSNOR activity (Corpas et al., 2008).

We studied the impact of these genetic transformations on the M. truncatula/A. euteiches interaction. ELISA tests using antibodies raised against A. euteiches (Slezack et al., 1999) were performed to quantify the presence of the pathogen in roots. In RNAi::MtNIA1/2 roots (Figure 2A), A. euteiches colonization was significantly greater than in control transformed roots (Control NR roots). These data reaffirm the role of the NR enzyme in the plant immune response. In A. thaliana, the NR-deficient double mutant (nia1 nia2) failed to exhibit a hypersensitive response and was hyper-susceptible to P. syringae (Modolo et al., 2006; Oliveira et al., 2009) and to the necrotrophic fungal pathogens Sclerotinia sclerotiorum or Botrytis cinerea (Perchepied et al., 2010; Rasul et al., 2012). Although these effects were attributed to the substantially reduced NO levels in this mutant, a side effect of N metabolism on plant defense cannot be excluded as NR stands at the crossroads between N metabolism and NO production.

Our results using GSNOR-transformed roots showed that pathogen levels were lower in GSNOR-overexpressing roots (Figure 2B) than in control transformed roots (Control LacZ roots). GSNOR could therefore be considered as a positive regulator of M. truncatula resistance to A. euteiches. Previous works already investigated the physiological roles of GSNOR in plant-pathogen interactions, using transgenic A. thaliana plants (Feechan et al., 2005; Rusterucci et al., 2007; Yun et al., 2011). Results are sometimes contradictory, as modulation of AtGSNOR expression enhanced or decreased plant disease resistance depending on the pathosystem. GSNOR could play a significant role in plant immunity because GSNO is considered as a mobile reservoir of NO, is more stable than NO, and is a transnitrosylation agent of proteins. The contrasted results obtained in our study with NR and GSNOR constructs could be attributed to the specific roles of the corresponding proteins in NO homeostasis. NR is involved in NO synthesis, whereas the primary role of GSNOR is to regulate GSNO contents. The recent results from Yun et al. (2016) confirm that GSNO and NO may play distinct roles in plant immunity by acting on different molecular targets. In addition, GSNOR indirectly affects NO, GSH, ROS, and total intracellular nitrosothiol (SNO) levels, indicating that GSNOR might be more globally involved in the regulation of the cell redox state (Espunya et al., 2006; Yun et al., 2011).

Nitric Oxide partly regulates N metabolism. Therefore we also investigated the effects of GSNOR overexpression on root NO₃^- contents and NR activity in transformed roots. GSNOR overexpression increased basal NO₃^- content and NR activity (Figures 3A,B). Modulation of N metabolism by GSNO and NO in A. thaliana has been described (Frungillo et al., 2014), and was explained by the effect of NO and GSNO on NR activity and on the expression of the AtNRT2.1 high-affinity NO₃^- transporter gene. Similarly to our data, that study shows that GSNOR overexpression is correlated with higher NR activity and NO₃^- content. Interestingly, we noted that pathogen colonization reduced NO₃^- concentrations in roots by approximately 65%, suggesting an effect of A. euteiches on nitrate transport and/or NO₃^- assimilation. Although we found a higher NO₃^- content in 35S::GSNOR-infected roots than in control infected roots, the amplitude of the pathogen-induced decrease in NO₃^- level was not impacted in 35S::GSNOR roots, suggesting that this process is independent of GSNO homeostasis. This reduced level of NO₃^- is unlikely to result from consumption of NO₃^- by the pathogen: data mining of the A. euteiches database revealed that no homologs of the NR, NIR, and NO₃^- transporter (NRT2) genes were detected in the genome of this pathogen², confirming earlier observations that NO₃^- is unfavorable for A. euteiches development (Huber and Watson, 1974). Alternatively, we cannot exclude that the decreased NO₃^- content in infected roots could be due to nitrate leakage from the roots related to developing necrosis.

To analyze the role of N availability on NO, H₂O₂, and ONOO^- accumulation, M. truncatula plants were cultivated in complete medium or NO₃^--deficient medium (NØ), and inoculated or not with A. euteiches. The NO scavenger cPTIO and the ONOO^- scavenger epicatechin were used as controls to check the specificity of the fluorescence probes. We observed that NO₃^- deficiency caused a significant increase in ONOO^- content on NØ medium (Figure 4C), whereas NO and H₂O₂ levels decreased (Figures 4A,B), highlighting a link between NO₃^- content and production of these reactive species. A clear effect of pathogen colonization was only evidenced for H₂O₂ contents (Figure 4A), and this increase was abolished on NØ. Surprisingly, although NO production is considered as a common response to pathogens, no increase in NO levels was detected in response to A. euteiches (Figure 4B). More generally, whereas NO, ROS, or ONOO^- production has been widely described in response to pathogens, the literature does not give a clear picture of the cross-talks between these molecules. For instance, we observed a negative correlation between NO and ONOO^- contents in response to NO₃^- deficiency, but in other models high NO levels are often correlated with high ONOO^- levels (Abramowski et al., 2015; Kulik et al., 2015). These conflicting observations raise some questions. Are these discrepancies due to plant models or due to the difficulty in measuring and precisely localizing these molecules? Differences in the stability of these molecules or their specific scavenging by plants during pathogen attack could explain why we did not detect changes in ONOO^- or NO levels in response to A. euteiches. Moreover, NO could also be used by the pathogen to activate its own metabolism, an important step in plant infection by fungi (Sedlářová et al., 2016).

We also measured root SNO levels and GSNOR activity in the biological conditions of interest. Root SNO contents, determined using the Saville–Griess method, significantly increased on NØ medium as compared to the complete medium (Figure 4D). In response to A. euteiches, no significant change in SNO levels was highlighted (Figure 4D). Therefore, on NØ medium, the SNO content evolved in an opposite way to the NO content, similarly to the ONOO^- content. This result is in accordance with results reported in Helianthus annuus (Chaki et al., 2011), and can be attributed to the fact that NO is the source for ONOO^- and SNO. By contrast, a high NO content can be correlated with a high SNO content when plants are grown on culture medium containing NO₃^- (Abramowski et al., 2015; Pietrowska et al., 2015). Our data also suggest that NO₃^- nutrition impacts the overall balance between NO, ONOO^-, and SNO. Regarding GSNOR, no changes in its activity was detected upon inoculation, in line with the absence of change in SNO levels in infected roots. In the roots of plants cultivated on NØ (Figure 5), higher GSNOR activity was correlated with higher SNO levels, confirming the positive correlation between GSNOR activity and SNO levels observed in 35S::GSNOR-transformed Medicago roots (Figure 1) and in pea, a closely related legume (Corpas et al., 2008). The positive or negative correlation between GSNOR activity and SNO levels or between NO and SNO levels depending on plant species and experimental conditions can be explained by several hypotheses. The SNO level is regulated through nitrosylation and denitrosylation; GSNOR, by controlling the level of GSNO, indirectly affects the level of S-nitrosylation. However, the TRX (thioredoxin)/NTR (NADPH-dependent TRX reductase) enzymatic system also controls S-nitrosylation (Kneeshaw et al., 2014). Interestingly, these activities were also identified in roots and activated by NO, leading to denitrosylation of specific proteins (Correa-Aragunde et al., 2015). Thus, these results, together with our study, illustrate the complex relationships between NO production/GSNOR activity and total SNO levels. Abiotic stresses also increase GSNOR activity (Kubienova et al., 2014), and this appears to be the case for M. truncatula plants under NO₃^- deficiency. Higher GSNOR activity in N-deprived roots (Figure 5) could lead to a physiological state inducing higher resistance to A. euteiches, as observed in the 35S::GSNOR-transformed roots (Figure 2B). This could partly explain the enhanced resistance to this oomycete on NØ medium despite the low levels of NO in the roots. Thus, altogether our data highlight the possible positive and non-redundant roles of NO (Figures 1A and 2A) and SNO (Figures 1B, 2B, and 4D) in mediating M. truncatula resistance to A. euteiches.