Perturbations in nitric oxide homeostasis promote Arabidopsis disease susceptibility towards Phytophthora parasitica

Abstract Phytophthora species can infect hundreds of different plants, including many important crops, causing a number of agriculturally relevant diseases. A key feature of attempted pathogen infection is the rapid production of the redox active molecule nitric oxide (NO). However, the potential role(s) of NO in plant resistance against Phytophthora is relatively unexplored. Here we show that the level of NO accumulation is crucial for basal resistance in Arabidopsis against Phytophthora parasitica. Counterintuitively, both relatively low or relatively high NO accumulation leads to reduced resistance against P. parasitica. S‐nitrosylation, the addition of a NO group to a protein cysteine thiol to form an S‐nitrosothiol, is an important route for NO bioactivity and this process is regulated predominantly by S‐nitrosoglutathione reductase 1 (GSNOR1). Loss‐of‐function mutations in GSNOR1 disable both salicylic acid accumulation and associated signalling, and also the production of reactive oxygen species, leading to susceptibility towards P. parasitica. Significantly, we also demonstrate that secreted proteins from P. parasitica can inhibit Arabidopsis GSNOR1 activity.

examined effectors from P. infestans suppress immune responses in tobacco, indicating that Phytophthora effectors target key elements of the defence system during infection to aid colonization (Zheng et al., 2014). In this context, PcAvh103 and RxLR48 from P. capsici interacted with enhanced disease susceptibility 1 (EDS1) and nonexpressor of PR1 (NPR1), respectively, both key regulatory nodes in the plant defence signalling network, promoting pathogen virulence by disrupting signalling via the immune activator, salicylic acid (SA) (Li, Chen, et al., 2019;Li, Wang et al., 2020).
A key feature of attempted pathogen infection is the rapid production of small, redox active molecules, including nitric oxide (NO) and ROS. These redox molecules orchestrate a plethora of immune responses in plants, including the accumulation of SA (Feechan et al., 2005;Grant & Loake, 2000;Lindermayr et al., 2010;Tada et al., 2008). A major route for the transfer of NO bioactivity is S-nitrosylation, the covalent attachment of NO to a cysteine thiol (SH) to form an S-nitrosothiol (SNO) (Jahnová et al., 2019;Yu et al., 2014). This redox-based posttranslational modification (PTM) is established with exquisite specificity (Astier et al., 2011;Umbreen et al., 2018), largely due to the unique properties of the sulphur atom component of the cysteine thiol (Umbreen et al., 2018). The enzyme S-nitrosoglutathione reductase 1 (GSNOR1) is a key determinant in indirectly controlling the total levels of cellular S-nitrosylation by depleting S-nitrosoglutathione (GSNO), the major cellular NO donor (Chen et al., 2009;Feechan et al., 2005;Lee et al., 2008). Arabidopsis GSNOR1 is required for multiple modes of plant disease resistance (Feechan et al., 2005) and also some aspects of plant development (Kwon et al., 2012;Lee et al., 2008) and this function is conserved in tomato (Gong et al., 2019;Hussain et al., 2019;Matamoros et al., 2020). However, a potential role for GSNOR1 in resistance to Phytophthora infection has not been uncovered.
It has recently been demonstrated, however, that exogenous application of the NO donor sodium nitroprusside (SNP) reduced resistance against Phytophthora (El-Beltagi et al., 2017). Moreover, ROS and SA accumulation are also thought to be required for resistance against this pathogen in both tobacco and Arabidopsis (Li, Wang, et al., 2020;Pan et al., 2016;Wi et al., 2012). In this context, cata-lase2 (CAT2) is directly targeted by PsCRN63 from P. sojae, leading to increased H 2 O 2 levels, triggering plant cell death (Zhang et al., 2015).
SA signalling is also thought to be targeted by the Phytophthora sojae effector PsICS1 and the P. capsici effector RxLR48 (Li, Chen, et al., 2019;Liu et al., 2014). However, evidence that effectors might manipulate immune-related redox signalling remains to be established.
In this study, we employed the Arabidopsis-Phytophthora pathosystem to explore the role of redox signalling in this interaction.
Our findings suggest that NO accumulation within a given concentration range supports Arabidopsis basal resistance against P. parasitica. Furthermore, GSNOR1 is required for full basal resistance against this pathogen. Transcriptomic and associated genetic analysis suggested that gsnor1 plants have impaired SA signalling and ROS production, leading to enhanced susceptibility to P. parasitica.
Significantly, we have also demonstrated that secreted proteins from P. parasitica may inhibit GSNOR1 activity.

| NO is required for basal disease resistance of Arabidopsis against P. parasitica
A key feature following pathogen recognition in eukaryotes is the engagement of a nitrosative burst, which leads to the accumulation of the gaseous signalling molecule NO and activation of cognate defence systems (Delledonne et al., 1998;Yu et al., 2014). However, the potential role of NO in basal disease resistance is not well documented. Thus, we explored if the NO burst was engaged during basal disease resistance against P. parasitica. Arabidopsis seedlings were challenged by P. parasitica and endogenous changes in NO levels were analysed by real-time imaging of NO accumulation using 4-amino-5-methylamino-2',7'-difluorofluorescein diacetate (DAF-FM DA). The endogenous NO level in wildtype Col-0 plants displayed a dynamic change postinoculation with a small increase at 3 hr postinoculation (hpi), together with a more significant increase at 6 hpi, followed by a decrease to unchallenged levels at 9 hpi (Figure 1a,b). Thus, NO accumulates during the early stages of basal disease resistance against P. parasitica. Subsequently, we checked additional Arabidopsis genotypes, Ler and Ws, which are also susceptible to P. parasitica. These genotypes showed a similar NO accumulation pattern in response to P. parasitica (Figure 1a,b).
Collectively, these results suggest that P. parasitica can induce an NO burst in Arabidopsis.
A major route for the transfer of NO bioactivity is S-nitrosylation, the addition of an NO moiety to a protein cysteine thiol, forming an SNO (Yu et al., 2014). Therefore, we examined total SNO levels in Arabidopsis after P. parasitica inoculation employing the biotinswitch assay (Jaffrey & Snyder, 2001). The total SNO level was significantly increased in Col-0 plants following P. parasitica inoculation compared to mock-treated plants (Figure 1c,d). These results suggest that P. parasitica triggered NO accumulation and subsequent SNO generation at early stages during the deployment of basal disease resistance.
To further explore a possible role for the nitrosative burst in Arabidopsis basal disease resistance against P. parasitica, we employed the NO scavenger 2-4-carboxyphenyl-4,4,5,5-tetra-methyli midazoline-1-oxyl-3-oxide (cPTIO) and the mammalian NOS inhibitor L-NG-nitro-arginine methyl ester (L-NAME), which has been shown to inhibit a NOS-like activity in plants (Delledonne et al., 1998). The application of cPTIO and L-NAME both enhanced the susceptibility of Col-0 plants to P. parasitica (Figures 2a and S1a). We next checked the nitrate reductase (NR)-deficient double mutant nia1 nia2, which has been reported to exhibit reduced NO accumulation (Modolo et al., 2006) (Figure S1b,c) to P. parasitica inoculation. The nia1 nia2 line exhibited enhanced disease susceptibility to P. parasitica compared with Col-0 plants (Figure 2b,d). This enhanced susceptibility phenotype was confirmed by determining the biomass of P. parasitica ( Figure 2c). Collectively, our data suggest that NO accumulation contributes to Arabidopsis basal resistance against P. parasitica. Arabidopsis. (a, b) NO levels determined in given genotypes by DAF-FM staining. Roots of living plants after P. parasitica infection or mock treatment were stained with DAF-FM DA and photographed with a Leica SP5 fluorescence microscope (a) and the relative intensity of DAF-FM stain was determined by ImageJ software (b) at indicated time points. Water was used as control. Eight seedlings of each line were used. Error bars represent the SD of eight replicates from each line. Asterisks indicate statistically significant differences compared to 0 hr postinoculation (hpi) (one-way analysis of variance [ANOVA], ***p < .001). (c, d) Total SNO detection by biotin switch assay followed by quantification. An equal amount of total protein extracts was subjected to the biotin-switch assay after P. parasitica infection or mock treatment. Ascorbate (ASC) was employed as negative control for SNO formation. The signal was quantified by ImageJ software. Experiments were repeated three times with similar results. *p < .05, ***p < .001, one-way ANOVA F I G U R E 2 The role of nitric oxide in Phytophthora parasitica-Arabidopsis interactions. (a) Disease severity index (DSI) of Col-0 seedlings in response to P. parasitica in the presence of 150 μM L-NAME (L-NG-nitroarginine methyl ester) or 200 μM cPTIO (2-4carboxyphenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide) for 4 days. (b) Phenotype of detached leaves inoculated with Pp016 for 3 days. Pp, infected by P. parasitica; Mock, treated by water. (c) Pathogen biomass analysis by quantitative PCR (qPCR). The DNA ratio of P. parasitica compared to Arabidopsis thaliana (PpUBC/AtUBC9) was determined by qPCR, total DNA extracted from inoculated leaves was used as a template. Error bars represent SD of six biological replicates. (d) The means of disease severity index of Col-0 and nia1 nia2 seedlings inoculated with Pp016. (e) Phenotype of indicated genotypes of detached leaves inoculated with Pp016. (h) Pathogen biomass analysis determined by qPCR. (f) Incidence of P. parasitica on wildtype Col-0 and nox1 seedlings 4 days postinoculation (dpi). For the mean of disease severity index assay, 16 seedlings were used for each experiment, error bars represent SD from three replicates. Asterisks indicate statistically significant differences compared to wildtype Col-0. One-way analysis of variance, ***p < .001, **p < .005, *p < .05

F I G U R E 1 Phytophthora parasitica infection elevates (S)NO levels in
To further explore the contribution of NO to Arabidopsis basal disease resistance against P. parasitica, we employed the NO hyperaccumulating line, no overexpression 1 (nox1). Unexpectedly, the leaves of nox1 plants also exhibited enhanced disease susceptibility to P. parasitica relative to Col-0 ( Figure 2e). A biomass assay also confirmed significantly higher growth and proliferation of P. parasitica in nox1 plants as compared to Col-0 plants ( Figure 2f).
Counterintuitively, nox1 seedlings exhibited an increased disease severity index relative to wildtype plants (Figure 2g), whereas application of cPTIO partially rescued this phenotype ( Figure S1d). In aggregate, our data suggest that either reduced or enhanced NO levels negatively impact basal resistance against P. parasitica.

| GSNOR1 is required for basal resistance against P. parasitica in Arabidopsis
GSNOR1 plays a critical role in governing protein-SNO levels during plant immune responses and GSNOR1 activity controls the level of both GSNO and global protein-SNOs (Chen et al., 2009;Feechan et al., 2005;Lee et al., 2008). First, the transcript level of GSNOR1 on P. parasitica inoculation was determined. GSNOR1 transcripts were significantly increased after 6 hr of infection by P. parasitica, but had decreased by 9 hpi (Figure 3a). This was accompanied by a concomitant increase in GSNOR1 activity following pathogen challenge (Figure 3b), suggesting the involvement of GSNOR1 in plant resistance to P. parasitica. To further elucidate the role of GSNOR1, loss-of-function mutants gsnor1-3 and par2-1 were challenged by P. parasitica. Water-soaked lesions were produced as a result of P. parasitica inoculation in both gsnor1-3 and par2-1 detached leaves at 2 days postinoculation (dpi) ( Figure S2a). The pathogen:host biomass ratio assay also showed significantly higher and faster proliferation of P. parasitica in gsnor1-3 and par2-1 plants as compared to wildtype Collectively, these results indicate that GSNOR1 plays a key role in Arabidopsis basal resistance against P. parasitica. P. parasitica is a typical soilborne pathogen and mainly infects the roots and crown area of the stem. We challenged in vitro grown seedlings with P. parasitica by inoculating them with culture plugs placed on the crown area. As expected, gsnor1-3 and par2-1 plants exhibited F I G U R E 3 GSNOR1 is required for Arabidopsis resistance to Phytophthora parasitica infection. (a) GSNOR1 transcript accumulation on infection was assayed by quantitative reverse transcription PCR. (b) GSNOR activity was measured at the indicated time points after inoculation with P. parasitica. (c) Pathogen biomass analysis by quantitative PCR for indicated genotypes after P. parasitica inoculation. Error bars in a-c indicate SD of three biological replicates. (d) Representative image of pathogen colonization in infected leaves by trypan blue staining. Schematic showing the corresponding observation site on the leaf: upper row (inoculation site) and lower row. Scale bars: upper row, 200 μm; lower row, 100 μm. Red arrow: wound site to enable colonization. Black arrow: hypha. Blue arrow: plant vein. (e) Relative intensity quantification of trypan blue staining in lower row of (d). Error bars represent SD of eight replicates. (f) Representative images of green fluorescent protein (GFP) expression P. parasitica H1121 hyphae colonization in detached leaves 3 days postinoculation. Upper row shows inoculation site, while lower row represents the middle of infected leaves. Scale bars: 100 μm. (g) Relative intensity of GFP signal of (e) was determined by ImageJ software. Error bars represent SD of eight replicates. Left: upper row in (f). Right: lower row in (f) an enhanced death rate relative to Col-0 at 4 dpi ( Figure S2d). These results show that GSNOR1 plays a key function in basal resistance of Arabidopsis against P. parasitica challenge.

| GSNOR1 is required for SA signalling during basal resistance against P. parasitica
A potential role for SA signalling in Arabidopsis-P. parasitica interactions has not been explored in detail to date. Thus, we examined SArelated defence gene expression in Arabidopsis following attempted P. parasitica infection. As shown in Figure 4a-c, expression of the SA defence marker genes PR1, PR5, and WRKY62 was significantly lower in the gsnor1-3 line relative to Col-0 plants on P. parasitica infection ( Figure 4a-c). These results indicate that SA signalling is engaged during basal disease resistance against P. parasitica and that this response is regulated by GSNOR1 function.
These findings prompted us to investigate if SA biosynthesis is associated with basal disease resistance in response to attempted P. parasitica infection. First, we measured the expression of CBP60g, SARD1, ICS1, and PAD4, which encode key SA biosynthesis and regulatory proteins, respectively (Zhang & Li, 2019). Interestingly, the induction of all these SA-related genes following P. parasitica inoculation was significantly lower in gsnor1-3 plants (Figure 4d-g), implying that GSNOR1 might also regulate SA biosynthesis. As anticipated, the total SA content in gsnor1-3 plants was significantly lower than Col-0 plants following P. parasitica inoculation ( Figure 4h). Collectively, these data demonstrate that GSNOR1 is a positive regulator of both SA accumulation and signalling during basal disease resistance against P. parasitica.
NPR1 is a key regulator of SA signalling and associated SAdependent gene expression (Cao et al., 1997). Leaves from 4-week-old Col-0 and npr1 plants were therefore subjected to P. parasitica inoculation. Phenotypic observations showed the appearance of severe symptoms on the leaves of npr1 plants but not on those of Col-0 plants, indicating enhanced disease susceptibility of npr1 to P. parasitica ( Figure 4i). The npr1 line also supported significantly higher and faster proliferation of P. parasitica as determined by biomass analysis ( Figure 4j). Furthermore, npr1 plants also exhibited a higher level of disease severity index relative to Col-0 plants ( Figure 4k). Collectively, these data show that SA-dependent responses are required for immunity against P. parasitica and GSNOR1 is required for both SA biosynthesis and signalling during attempted P. parasitica infection.

| Loss of GSNOR1 function reduces P. parasiticatriggered ROS production
The oxidative burst is an early immune response and it has been reported that the timing of ROS production is an important F I G U R E 4 GSNOR1 is required for salicylic acid (SA) synthesis and signalling during Phytophthora parasitica infection. (a-f) Expression of SA marker genes PR-1 (a), PR5 (b), and WRKY62 (c), and SA synthesisrelated genes SARD1 (d), CBP60g (e), ICS1 (f), and PAD4 (g) were analysed. Ten-day-old wildtype (Col-0) and gsnor1-3 seedlings were inoculated with P. parasitica or mock-treated for 12 hr and then collected for quantitative reverse transcription PCR. Relative gene expression was normalized against the constitutively expressed gene, UBQ10. (h) Total SA levels in the leaves of wildtype (Col-0) or gsnor1-3 plants 24 hr after P. parasitica inoculation. (i) Wildtype Col-0 and npr1 plants were infected with P. parasitica and disease symptoms recorded at 3 days postinoculation. (j) Pathogen biomass analysis by quantitative PCR for (i). (k) Incidence of P parasitica on wildtype Col-0 and npr1 seedlings on P. parasitica inoculation was analysed by disease severity index (DSI). The mean of the DSI was obtained from 16 seedlings, error bars represent SD from three replicates. Error bars indicate ± SD of three biological replicates. *p < .05, ***p < .001 determinant during the plant defence response to P. parasitica (Grant & Loake, 2000;Wi et al., 2012). Therefore, transcript accumulation of Respiratory Burst Oxidase Homolog D (RBOHD), which encodes a key enzyme for pathogen-triggered ROS production (Torres et al., 2002), was analysed in response to P. parasitica. RBOHD was induced in wildtype plants following P. parasitica inoculation. Surprisingly, RBOHD expression in gsnor1-3 plants was higher compared to Col-0 after infection with P. parasitica (Figure 5a). However, RBOHD activity was lower in the gsnor1-3 line compared to Col-0 on infection ( Figure 5b). Subsequently, we monitored ROS production in plants challenged with P. parasitica. Reduced diaminobenzidine (DAB) staining, a ROS-sensitive marker (Thordal-Christensen et al., 1997), was observed at gsnor1-3 plant inoculation sites (Figure 5c) relative to the wild type and quantification of DAB staining also showed significantly lower ROS accumulation in the gsnor1-3 mutant in response to P. parasitica infection (Figure 5d). To investigate the potential involvement of RBOHD, a major source of apoplastic ROS (Torres et al., 2002), in response to P. parasitica infection, we inoculated the Arabidopsis rbohd loss-of-function mutant (Torres et al., 2002) with P. parasitica. The rbohd line was highly susceptible to infection and exhibited enhanced disease symptoms as compared to wildtype plants ( Figure 5e). Furthermore, relative biomass quantification of P. parasitica was undertaken via quantitative PCR (qPCR): enhanced levels of P. parasitica DNA was detected in rbohd plants relative to the wild type (Figure 5f), demonstrating the enhanced disease susceptibility of rbohd plants to P. parasitica.

| GSNOR1 has a global impact on ROS and SAmediated gene expression
To determine the impact of loss of GSNOR1 function on global gene expression following P. parasitica inoculation, we performed RNA-Seq-mediated transcriptomic analysis of wildtype and gsnor1-3 seedlings. Significant changes in the expression of 1,778 differentially expressed genes (DEGs) were found in wildtype plants after 24 hr of infection, of which 1,045 were upregulated, whereas 733 were downregulated (Figure 6a). On the other hand, 1,591 DEGs were identified in gsnor1-3, of which 1,232 were upregulated and 359 were downregulated (Figure 6a). This suggests that GSNOR1 function is required for significant gene expression reprogramming during P. parasitica infection.

Gene ontology (GO) analysis identified enriched categories for
DEGs related to biotic and abiotic stress responses (Figure 6b), indicating GSNOR1 has a global impact on the Arabidopsis hormone, immune function, and oxidation-associated gene expression profile.
Specifically, we found that the oxidative stress response and response to hydrogen peroxide were more significantly affected in the gsnor1-3 line in response to P. parasitica as compared to wildtype plants ( Figure 6c, Table 1). However, genes related to the oxidationreduction process were more strongly impacted in wildtype plants relative to the gsnor1-3 line. These ROS-related genes include ALCOHOL DEHYDROGENASE 1, ADH1 (De la Rosa et al., 2019) and zinc finger protein 7, ZAT7 (Ciftci-Yilmaz et al., 2007). This suggests that GSNOR1 is a key regulator of ROS regulation in response to P. parasitica. This again highlights the importance of ROS regulation by GSNOR1 in response to P. parasitica infection. Furthermore, we found that genes integral to SA-related signalling were strongly differentially expressed in wildtype plants but not in the gsnor1-3 line in response to P. parasitica inoculation (Table 1).
This data suggests that SA signalling plays an important role in resistance against P. parasitica and GSNOR1 is a key regulator of these responses. Interestingly, jasmonic acid (JA) signalling seems not to be regulated by GSNOR1 on P. parasitica inoculation. Collectively, these results suggest that GSNOR1 has significant impact on SA signalling in response to attempted P. parasitica infection.

| Secreted proteins from P. parasitica inhibit GSNOR1 activity
As GSNOR1 function is important for basal resistance against P. parasitica, we analysed whether the secreted proteome of P. parasitica might target GSNOR1 activity. Freshly secreted total protein from P. parasitica was collected as previously described (Kamoun et al., 1993). Recombinant GSNOR1 was incubated with the P. parasitica secreted proteome. Interestingly, the secreted proteome of

| D ISCUSS I ON
Our findings show that the level of NO is crucial for establishing basal resistance against P. parasitica in Arabidopsis. Thus, counterintuitively either a relatively low or relatively high level of NO accumulation leads to reduced basal resistance against P. parasitica.
Thus, perturbations in NO homeostasis promote Arabidopsis disease The 20 most enriched GO terms in biological process. There are 1,233 differentially expressed genes (DEGs) and the percentage of each GO term is indicated. (c) Heat map of the normalized RNA-Seq data for genes involved in biotic stress response after inoculation with P. parasitica susceptibility towards P. parasitica. Furthermore, loss-of-function mutations in GSNOR1 disable both SA accumulation and signalling, and also ROS accumulation, leading to enhanced susceptibility towards P. parasitica. Significantly, we have demonstrated that an extract of the secreted proteome from P. parasitica inhibits GSNOR1 activity. Because GSNOR1 activity is required for SA signalling and ROS production, P. parasitica may target GSNOR1 to aid pathogenesis, facilitating colonization.
Based on our data, we propose a model for the role of GSNOR1 in basal disease resistance against P. parasitica (Figure 8). We have shown that attempted P. parasitica infection induces NO and subsequently GSNO accumulation. Subsequently, these molecules activate defence responses, such as SA signalling, leading to a restriction of P. parasitica infection. In addition, RBOHD is activated, enhancing ROS production, resulting in inhibition of P. parasitica infection. However, as NO and GSNO concentrations increase, their homeostasis is perturbed, and excessive S-nitrosylation might result in defence suppression and associated pathogen susceptibility.
For example, increasing NPR1 S-nitrosylation sequesters this transcriptional co-activator in the cytosol, blunting SA-dependent gene expression (Tada et al., 2008). Also, S-nitrosylation of SA-binding protein 3 (SABP3) reduces the SA binding of this protein and also the carbonic anhydrase activity of this enzyme, increasing disease susceptibility . In parallel, the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, RBOHD, is S-nitrosylated at Cys890, decreasing its activity (Yun et al., 2011) and suppressing ROS production. Collectively, these molecular events in combination may curb the immune response. Thus, perturbations in NO/ GSNO homeostasis may promote increased susceptibility towards P. parasitica.

| NO homeostasis is a critical component of plant resistance against P. parasitica
Previous studies have revealed that NO is induced during the establishment of resistance against P. capsici challenge (Requena et al., 2005) or by defence-activating elicitors from Phytophthora cryptogea (Foissner et al., 2000). Our data extend these findings by suggesting that P. parasitica infection (in addition to resistance) also triggered a NO burst in Arabidopsis. Application of the NO donor SNP has been reported to enhance potato susceptibility to P. infestans infection (El-Beltagi et al., 2017). In contrast, we found that application of either NO scavengers or nitric oxide synthase (Ichinose et al., 2003) inhibitors blunted Arabidopsis resistance to P. parasitica. Also, the NR-deficient mutant nia1 nia2 displayed reduced P. parasitica-induced NO production and enhanced susceptibility to P. parasitica. This is consistent with previous data that suggests that NR-dependent NO production is required for resistance to P. infestans (Floryszak-Wieczorek et al., 2016).
The NO overproduction mutant, nox1, also showed increased susceptibility to P. parasitica relative to the wild type ( Figure 1).
Furthermore, loss of GSNOR1 function in the gsnor1-3 mutant, which elevates both NO and GSNO, also showed enhanced susceptibility to P. parasitica (Figure 2), suggesting excessive endogenous NO and/ or GSNO negative feedback regulates immunity against P. parasitica.
In this context, it has been shown that endogenous NO and GSNO, two pivotal redox signalling molecules, may have both distinct and overlapping functions during the development of immunity (Yun et al., 2016). Collectively, our data imply there is a breadth of NO  (Yun et al., 2011). In parallel, NPR1 is S-nitrosylated, promoting its cytosolic localization and blunting SA-dependent gene expression (Tada et al., 2008). Collectively, these molecular events in combination may curb the immune response. During infection P. parasitica might deliver secreted proteins into host cells to target GSNOR1 leading to excessive (S)NO accumulation to promote increased susceptibility

| GSNOR1 is required for basal resistance against P. parasitica
The mutant lines gsnor1-3 and par2-1, another allele of GSNOR1 (Chen et al., 2009) that exhibits a high level of SNO, both showed enhanced susceptibility to P. parasitica. Consistent with our study, previous research also found that the absence of GSNOR1 activity might be involved in potato susceptibility to P. infestans (Abramowski et al., 2015). Interestingly, we also found that the NO overproduction mutant nox1 also showed enhanced susceptibility to P. parasitica. It has been shown that GSNOR1 is S-nitrosylated, inhibiting its activity (Frungillo et al., 2014). Thus, high NO levels in this mutant could inhibit GSNOR1 activity through S-nitrosylation, leading to excessive (S)NO levels. Consistently, GSNOR activity is required for plant basal immunity (Feechan et al., 2005;Yun et al., 2016). Furthermore, GSNOR1 activity has been shown to be differently modulated in lettuce against downy mildew (Tichá et al., 2018;Yun et al., 2016).

Collectively, our data establish that GSNOR1 function is required for
Arabidopsis resistance against Phytophthora.

| NO-mediated regulation of SA signalling integral for P. parasitica resistance
The immune regulator SA plays a central role in the plant immune response, with SA biosynthesis and signalling under redox regulation (Feechan et al., 2005;Lindermayr et al., 2010;Tada et al., 2008).
Although SA function has been shown to be important for immunity against a number of diverse plant pathogens Pan et al., 2016;Yang et al., 2017), a potential role for SA in resistance against P. parasitica has not been thoroughly investigated. We have previously reported that the Arabidopsis mutant, resistance to P. parasitica1 (rtp1), engaged SA signalling and exhibited resistance to P. parasitica (Pan et al., 2016). Recently, the Arabidopsis rtp5 line has also been shown to be associated with increased SA synthesis and enhanced protection against P. parasitica infection . Here we show SA-mediated transcriptional reprogramming is impaired in gsnor1-3 plants in response to P. parasitica. The SA-signalling mutant npr1 also showed susceptibility to P. parasitica. Furthermore, in gsnor1-3 plants, increased SNO levels have been shown to compromise SA signalling by both inhibiting SABP3 function and controlling the nuclear translocation of NPR1 (Tada et al., 2008;Wang et al., 2009). Thus, GSNOR1 may regulate SA-dependent basal resistance against P. parasitica in a similar fashion.

| P. parasitica-induced NO regulates the ROS burst
The timing of ROS production is an important determinant in plant compatible interactions towards P. parasitica (Pan et al., 2016;Wi et al., 2012). NADPH oxidases have been uncovered as an essential enzyme for ROS production in plant defence and previous reports have shown that attempted Phytophthora infection could induced RBOHD expression and enhance the ROS burst. Moreover, it has been shown that resistance in the root of Arabidopsis against P. parasitica requires an NADPH oxidase-mediated oxidative burst: NADPH oxidase knockdown or knockout plants showed increased susceptibility to oomycetes (Shibata et al., 2010). Recently, the effector RxLR207 from P. capsici was shown to target BPA1 (Binding partner of ACD11), which may control ROS production . Also, we reported a T-DNA insertion mutant, rtp1-1, that disables a nodulin-related MtN21 family protein, resulting in increased ROS production and enhanced resistance to P. parasitica (Pan et al., 2016). Our findings here suggest that rbohd mutants, disabled in the most important NADPH oxidase isoform associated with the leaf oxidative burst (Torres et al., 2002), showed increased susceptibility to P. parasitica. Furthermore, the oxi-

| P. parasitica-secreted proteins target GSNOR1 activity
Oomycete pathogens secrete a series of proteins that interfere with plant immune function by targeting key regulators of plant immunity (Jiang & Tyler, 2012;Wang et al., 2019). Interestingly, it has recently been reported that host plants can disarm the bacterial effector HopAI1 by S-nitrosylation, demonstrating a function for NO production in the neutralization of pathogen effectors (Ling et al., 2017). Furthermore, it has been reported that Phytophthora effectors can interfere with the host plant immune system. For example, Phytophthora effector PsCRN63 can target CAT2 to regulate ROS homeostasis (Li et al., 2016). RxLR48, an effector from P. capsici, has been reported to target NPR1, triggering its proteasome-mediated degradation . Our data suggests that one or more secreted proteins from P. parasitica might inhibit GSNOR1 activity. Thus, our study reveals a potential novel molecular mechanism where secreted P. parasitica proteins may target GSNOR1 for inactivation to promote pathogenesis.
In conclusion, we provided genetic evidence that NO and GSNOR1 are required for Arabidopsis basal resistance against P. parasitica. Furthermore, our study suggests that NO-mediated Snitrosylation regulates both SA signalling and the ROS burst against P. parasitica. We also provide preliminary evidence that one or more secreted proteins from P. parasitica can inhibit GSNOR1 activity.
Future work will endeavour to uncover the identity of protein(s) that target GSNOR1 and the underpinning mechanism that promotes the inhibition of this key immunity-related enzyme. This work may provide new avenues for molecular breeding strategies to convey improved resistance in crop plants against oomycete infection.
Zoospores were collected and the concentration was adjusted to 10 5 zoospores/ml for all experiments unless otherwise specified.

| Biotin switch for total SNO measurement
Arabidopsis seedlings were inoculated with Pp016 zoospores and whole seedlings were collected after 0, 3, 6, and 9 hr. The sample was homogenized in HEN buffer (25 mM HEPES, pH 7.7, 1 mM EDTA, 0.1 mM neocuproine) containing 1× complete protease inhibitor cocktail (Roche) and centrifuged at 4 °C for 20 min at 10,000 × g. Total protein concentration in the supernatant was measured using the Bradford assay.
About 200 µg of total protein from each sample was subjected to the biotin-switch assay (Jaffrey & Snyder, 2001)

| GSNOR activity assay
GSNOR activity of total protein from Arabidopsis lines was measured at 25 °C by decomposition of NADH at 340 nm (Feechan et al., 2005). Samples were collected at indicated times and associated activity was determined by incubating 100 µg of plant extracted total protein in 1 ml of reaction mix containing 20 mM Tris-HCl (pH 8.0), before resuspending in 500 μl of extraction buffer. The homogenate was transferred to a prefiltration column and spun down for 2 min at 16,000 × g before adding 500 μl of 2-propanol to the flow-through.
The mixture was centrifuged through a mini-isolation column and washed twice with 600 μl of wash buffer at 16,000 × g for 1 min.
Finally, RNA was eluted with RNase-free water and quantified with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific).
cDNA was synthesized from about 2 μg of total RNA using oligo(dT) primers and reverse transcriptase (First-Strand cDNA Synthesis Kit; Invitrogen).

| Quantitative PCR
To check the pathogen colonization levels a quantitative real-time PCR analysis was conducted to compare P. parasitica UBC (PpUBC) DNA levels (to measure the pathogen biomass) to Arabidopsis UBC (AtUBC9) DNA levels (Pan et al., 2016;van Esse et al., 2008).
qPCR analysis for biomass assay and for checking the expression of all other genes was performed as described below, using the primers given in Table S1. qPCR was performed using the LightCycler 480 Real-Time PCR System (Roche). Gene expression levels were quantified by LightCycler DNA Master SYBR Green I mix and LightCycler system and gene expression values were determined using UBQ10 as reference. All experiments were repeated at least three times.

| DAB and trypan blue staining
To quantify the various ROS species in the Arabidopsis lines before and after pathogen inoculation, leaf samples were submerged in 0.5 mg/ml nitroblue tetrazolium (DAB; Sigma) staining buffer for 3 hr at room temperature in the dark. Leaves were then destained in 70% ethanol until the green colour was completely removed. Leaves were observed with a microscope (Olympus) and photographed.
To observe the proliferation of P. parasitica hyphae during infection, uninoculated control and inoculated Arabidopsis leaves were stained in trypan blue solution (10 g phenol, 10 ml glycerol, 10 ml lactic acid, 10 ml water, 10 mg trypan blue). For this purpose, leaves were dipped in the staining solution inside small glass bottles submerged in boiling water for 2 min. After cooling to room temperature, the samples were destained with 2.5 g/ml chloral hydrate solution until the samples were clean of any residual stain. The samples were rinsed with water and viewed under a microscope (Olympus).
4.10 | Purification of total secreted protein from P. parasitica Pp016 P. parasitica strain Pp016 was grown on V8 agar for 5 days. Then small discs were transferred to 500 ml of synthesis liquid medium and total secreted proteins were collected as described (Kamoun et al., 1993). The total precipitated protein was dissolved in 2 ml of TE buffer and the protein concentration was quantified with the Bradford assay.

| RNA sequencing and GO enrichment analysis
For transcriptome analysis, 10-day-old seedlings of Col-0 and gsnor1-3 mutant were inoculated by P. parasitica or water and collected in liquid N 2 at 12 hpi. Total RNA was extracted as described followed. Briefly, about 100 mg of sample homogenized in liquid N 2 was added to 1 ml TRIzol reagent (Invitrogen) and the homogenate separated by centrifugation, followed by mixing the upper aqueous layer with chloroform. After centrifugation the aqueous layer was mixed with ethanol and added to RNase-Free Columns (CR3) followed by two washes with buffer RW1 and one wash with RW before RNA was resolved in RNase-free water as described

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 on reasonable request.