Zoospore exudates from Phytophthora nicotianae affect immune responses in Arabidopsis

Zoospore exudates play important roles in promoting zoospore communication, homing and germination during plant infection by Phytophthora. However, it is not clear whether exudates affect plant immunity. Zoospore-free fluid (ZFF) and zoospores of P. nicotianae were investigated comparatively for effects on resistance of Arabidopsis thaliana Col-0 and mutants that affect signaling mediated by salicylic acid (SA) and jasmonic acid (JA): eds16 (enhanced disease susceptibility16), pad4 (phytoalexin deficient4), and npr1 (nonexpressor of pathogenesis-related genes1). Col-0 attracted more zoospores and had severe tissue damage when flooded with a zoospore suspension in ZFF. Mutants treated with ZFF alone developed disease symptoms similar to those inoculated with zoospores and requirements of EDS16 and PAD4 for plant responses to zoospores and the exudates was apparent. Zoospore and ZFFs also induced expression of the PR1 and PDF1.2 marker genes for defense regulated by SA and JA, respectively. However, ZFF affected more JA defense signaling, down regulating PR1 when SA signaling or synthesis is deficient, which may be responsible for Arabidopsis mutant plants more susceptible to infection by high concentration of P. nicotianae zoospores. These results suggest that zoospore exudates can function as virulence factors and inducers of plant immune responses during plant infection by Phytophthora.


Introduction
Phytophthora species use a variety of strategies to infect plants, causing severe losses for food crops and ornamental plant production, as well as damage in forest ecosystems [1]. For example, P. nicotianae (syn. P. parasitica) can colonize foliar structures and roots of over 200 plant species including Arabidopsis [1][2][3][4]. This pathogen can use exudates from zoospores and bacteria to promote successful infection [3,5,6]. The exudates from zoospores of Phytophthora and Pythium, termed zoospore-free fluids (ZFF), function as quorum-sensing signals promoting zoospore homing and triggering infection by single zoospores [3,5,7]. However, it is not clear whether ZFF also affects plant immunity. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Plant immune responses are activated by two major classes of surveillance proteins: transmembrane pattern recognition receptors (PRRs) and plant nucleotide-binding, leucine-rich repeat (NLR) proteins [8,9]. PRRs recognize conserved, pathogen-associated molecular patterns (PAMPS) to activate PAMP-triggered immunity (PTI). NLR proteins recognize pathogen effectors to activate effector-triggered immunity (ETI). Activation of PTI and ETI in plants involves a complex signaling network regulated by small plant hormone molecules such as salicylic acid (SA), jasmonic acid (JA) and ethylene (ET), as well as proteins responsible for production and signal transduction of these hormones [10,11].
Arabidopsis thaliana is a popular model plant for studying plant immunity and there are many loss-of-function Arabidopsis mutants available for dissection of immune responses to different pathogens. For example, eds 16-1 contains a loss of function mutation in a gene encoding isochorismate synthase, a major enzyme in the biosynthesis [12] of salicylic acid (SA), which plays a central role in immune responses to biotrophic pathogens. Similarly, pad4-1 is a mutant of the Phytoalexin Deficiency 4 gene that encodes a lipase-like protein operating upstream of SA accumulation to regulate various aspects of immune responses, including production of the anti-microbial phytoalexin camalexin, and SAdependent immune signaling, and other sectors of the immune response network [13][14][15][16]. Npr-1 is a mutant of Nonexpresser of Pathogenesis-Related Protein 1 (NPR1), encoding a receptor of SA and a central regulatory protein of SA-mediated responses in plant defense [11,17,18]. Another useful tool in Arabidopsis is marker genes that are assayed to report activation of specific defense signaling pathways. For example, PR1 [19] and PDF1.2 [20] are widely used to determine activation of pathways mediated by SA and JA/ET signaling, respectively.
Current models of defense signaling indicate that SA and JA/ET pathways are activated by pathogens with different lifestyles [21]. For example, activation of JA/ET signaling is often associated with the defense against necrotrophic such as Pythium species [22][23][24][25]. In contrast, activation of SA signaling is a key component of defense against biotrophs such as Hyaloperonospora arabidopsidis [26][27][28]. The SA and JA/ET pathways are mutually antagonistic in Arabidopsis, such that activation of one pathway triggers repression of the other. It is believed that this mutual antagonism provides for adaptive tailoring so that the most effective defense is deployed against the type of pathogen that is invading the plant [29]. However, it is unclear how these pathways are regulated during infection by hemi-biotrophic pathogens that colonize the plants as biotrophs and later activate a necrotrophic phase.
Many Arabidopsis-based pathosystems have been studied for plant defense mechanisms against oomycetes of importance, including Hyaloperonospora arabidopsidis, Phytophthora capsici, P. cinammomi, P. infestans, P. nicotianae, P. porri, P. sojae, and Pythium irregulare [22,28,[30][31][32][33][34]. However, only H. arabidopsidis is a natural pathogen of Arabidopsis, while P. porri is a natural pathogen of Brassica plants related to Arabidopsis. Most of other oomycetes either cause a hypersensitive response in Arabidopsis under controlled conditions or have a restricted host range [2,24,31,[35][36][37][38]. In contrast, P. nicotianae has a wide host range causing diseases of over 200 plant species and can infect Arabidopsis at a high inoculum concentration in the laboratory [2,3]. Moreover, P. nicotianae is relatively amenable to biological, molecular and genetic approaches. Thus, the Arabidopsis-P. nicotianae phytopathosystem is considered as a useful model for studying interaction between plants and oomycete pathogens.
In this study, we compared how P. nicotianae zoospores and ZFF interacted with A. thaliana Col-0 and mutants in SA signaling and biosynthesis. While previous studies focused on roles of ZFF in Phytophthora quorum sensing and behaviors, this study focused on the impact of zoospore exudates on plant immune responses during infection by P. nicotianae.

Pathogen growth conditions and ZFF preparation
Phytophthora nicotianae isolate 1B11 from annual vinca (Catharanthus roseus cv. 'Little Bright Eye') was used in this study. Growth, maintenance and preparation of zoospore-free fluid (ZFF) of the organism followed previous protocols unless otherwise stated [3,5,39]. Specifically, high concentrations of zoospore suspensions were made by growing mycelium plugs in 10% clear V8 juice broth at 23˚C for one week in the dark followed by inducting sporangia under the lights for two days, rinsing and flooding of the grown mycelium/sporangium mats with cold sterilized distilled water (SDW) to remove the nutrient residues from media and to release zoospores, respectively. Zoospore suspensions were measured for concentrations with a hemocytometer under a microscope. To make ZFF, suspensions were filtered through a 0.2 μm filter after vortexed for 5 min to allow zoospore to encyst and release chemicals into the solution. The filtrate was referred as ZFF, and its concentration was determined by the concentration of zoospores suspension used. In this study, all ZFF were from zoospore suspensions at 10 5 /ml or higher.

Plant growth conditions
Arabidopsis thaliana (Col-0) and mutants, eds16-1, npr1-1 and pad4-, were seeded in Metro Mix 360 top-layered with Metro Mix 200 (Scotts, Marysville, OH) and grown for 5 weeks in a growth chamber at 25˚C with a 14/10 h light/dark cycle. All plants were fertilized in the third and fourth week with 20-20-20 liquid fertilizer (Scotts, Marysville, OH). Four-to 6-week-old plants were used in assays including microscopy, disease assessment and gene expression.

Examination of effects of ZFF on plant response to zoospores
Detached cauline leaves from 5-week old Col-0 plants were placed individually in the wells of a 24-well plate and flooded with 250 μl of ZFF or sterile distilled water (SDW), each containing zoospores at a concentration of 1,600/ml. The plate was kept in the dark overnight at 23˚C, and then transferred to growth chamber. Leaves in the wells were examined for zoospore aggregation and plant tissue damage under a Nikon Labophot-2 or Olympus 1X71 inverted microscope after 5, 24, 48 and 120 h incubation. Each treatment (ZFF or SDW) included three replicated wells.

Assessment of effects of ZFF on disease development
Plants of four to five weeks were removed from MetroMix, cleaned in tap water and rinsed with SDW before treatment. Six plants per genotype were flooded with 20 ml ZFF or controls SDW and zoospore suspension at 200,000/ml, respectively in a 40-ml beaker and placed in moisture containers. Plants were kept at 23˚C in the dark for 16 h and then placed in a growth chamber under the plant growth conditions. Disease was rated by the percentage of symptomatic leaves in total leaves at 88 h after treatment. Yellowing and rotting symptoms were rated separately. Each experiment included four biological replicates and was repeated once.
Analysis of effects of ZFF on plant gene expression ZFF was also assessed for activation of SA and JA signaling mediated defense marker genes PR1 and PDF1.2 in a time course experiment. The experiment was set up as that for disease assessment and included three replicates for each treatment: ZFF, SDW or zoospores at 200,000/ml. At [40] and PR1 [41] as well as for the endogenous control, reference gene Actin-1 (ACT1) [40]. All reactions were run at the 9600 mode Emulation with 50˚C for 2 min at first stage, 95˚C for 10 min at second stage followed by 40 cycles of 15 sec at 95˚C, 1 min at 60˚C at the third stage. Each sample had three PCR runs for each of the genes. The mean value of Ct (threshold cycle) of each gene in a sample of the PCRs were generated with Relative Quantification Study in the software SDS v.1.3.1 (Applied Biosystems, Foster City, CA). Target gene expression was measured with Relative transcription levels (RQ) which was calculated using the formula 2 -dCt after Ct of target gene was normalized with that of the reference gene ACT1 in the same sample (dCt target = Ct target −Ct ACT1 ).

Statistical analysis
Standard errors were calculated from standard deviation and number of replicates with Microsoft Excel. T-test at equal variances in Excel was used for evaluating statistical significance between the treatments. P values <0.05 were considered significant.

Effect of zoospore exudates on Arabidopsis infection by Phytophthora nicotianae
Detached cauline leaves of Col-0 were flooded in ZFF containing zoospores at 1, 600/ml for 5 h (Fig 1A). At this time point, zoospores aggregated at the base of trichome. The affected trichomes became macerated 24 h after the exposure. Meanwhile, there was zoospore germination and mycelium growth on and in plant tissues (Fig 1C). Leaf tissue damage and substance discharge from trichomes was evident at 48 hours after the treatment (Fig 1E). Leaf yellowing was also evident at this time point. On the 5th day of the treatment, rotting became apparent and further tissue damage and production of numerous sporangia was observed under the microscope (Fig 1G). Same effect was found for ZFF from zoospore suspensions at ! 10 5 /ml as used in this study. In contrast, there was sparse zoospore attachment to leaves treated with SDW ( Fig 1B). SDW-treated leaves/sustained little damage in trichomes or leaf tissue and supported limited hyphal growth with no sporangia production (Fig 1B, 1D, 1F and 1H). These results indicate that ZFF promoted susceptibility of Arabidopsis plants to P. nicotianae.

Effect of zoospore exudates on Arabidopsis defense components
To better understand why plants become more susceptible to zoospores with addition of ZFF, Arabidopsis Col-0 and mutant plants eds16-1; npr1-1 and pad4-1(all in the Col-0 background) were treated with a zoospore suspension at 200,000/ml; ZFF alone, or a SDW control (Fig 2,  Table 1). Visual symptoms were compared in response to all three treatments. Zoospore treatment induced visible symptoms in >40% of leaves from Col-0. Contrastingly, ZFF treatment induced only a minor increase in symptoms compared to water-treated control. Interestingly, eds16-1 plants developed severe leaf yellowing and rotting, compared to Col-0, after treatment with ZFF (Fig 2B). Similar symptoms were displayed after inoculation with zoospores. The pad 4-1 mutant also displayed slightly enhanced symptoms, compared to wild-type Col-0, after treatment with ZFF or zoospore inoculum. Interestingly, responses to ZFF and zoospores were slightly attenuated in npr1-1 compared to Col-0. These results indicate that a deficiency in SA  biosynthesis/signaling rendered Arabidopsis more susceptible to P. nicotianae, and that ZFF treatment triggered disease-like symptoms in an SA-deficient mutant.

Effect of zoospore exudates on regulation of SA and JA signaling in the plants
To determine how ZFF may affect defense signaling pathways, expression of SA and JA marker gene PR1 and PDF1.2 was examined in wild type and mutant plants treated with zoospores, ZFF, or SDW. In Col-0, both genes were induced by ZFF and zoospores although the effect of ZFF was weaker than zoospores. ZFF treatment appeared to have a stronger effect on PDF1.2 expression than on PR1 expression, compared to the water treatment (Fig 3A). eds16-1 plants displayed substantially different patterns of transcript accumulation, compared to Col-0: In zoospore-and ZFF-treated plants, expression of PDF1.2 was much lower at 4 and 8 hours after treatment, compared to Col-0, but then increased to a level higher than Col-0 by 72 hours. Zoospore-responsive expression of PR-1 was lower than Col-0 in eds16-1 at early time points, but increased at later time points. As in Col-0, ZFF effect on both marker genes in eds16-1 was roughly equivalent to that of water. The npr1-1 and pad4-1 mutants were similar to each other and to eds16-1: Expression of PDF1.2 was equivalent or somewhat higher to Col-0, with zoospore treatment inducing stronger expression than ZFF or water. PR-1 expression in mutants was much lower than in Col-0 in response to all treatments although zoospore treatment induced a much stronger response in the late time.

Discussion
Zoospore exudates of Phytophthora have been shown as chemical signals for Phytophthora quorum sensing leading to zoospore aggregation, homing and infection initiation as well as communications between zoosporic oomycetes and with bacteria [3,5,7,39,42]. This study reveals new roles of zoospore exudates as virulence factors of the pathogen and inducers of plant immunity, contributing to the success of plant infection by Phytophthora zoospores.
Most importantly, zoospore pre-treatment with ZFF rendered Arabidopsis plants more susceptible to infection by P. nicotianae zoospores (Fig 1). Additionally, ZFF treatment induced disease-like symptoms of yellowing and rotting that were similar to those induced by treatment with zoospores. Finally, ZFF treatment influenced expression of marker genes for both SA-and JA-dependent immune responses. These data indicate that zoospore exudates can affect the function of the host immune system, in addition to mediating intra-and interspecies chemical communication, promoting zoospore homing and plant infection described in previous studies. The ability of ZFF to promote disease-like symptoms, independent of pathogen colonization, is particularly intriguing. We propose that the chlorosis and rotting are due to one or more toxins present in ZFF. It would be of great interest to identify the precise compounds in ZFF that induces these symptoms. It will also be interesting to determine if the toxins promote immune activation as part of the necrotrophic phase of plant colonization. Another interesting aspect is that plant responses to zoospores and ZFF treatment appear linked to SA signaling, based on differential responses observed during treatment of mutants that disable SA biosynthesis (eds16-1) or signaling (pad4-1, npr1-1). When SA biosynthesis is disabled (eds16-1), the disease symptoms triggered by zoospores or ZFF are substantially enhanced. A similar but weaker enhancement was observed in the pad4-1 mutant. These data indicate that SA signaling plays a role in resistance to pathogen colonization (zoospores) and/ or toxicity mediated by ZFF. Because pad4-1 was less responsive to ZFF, PAD4 may partially involve the resistance to the toxins in zoospore exudates despite the fact that it can provide amplified defense through phytoalexin camalexin production or interaction with EDS [14,15]. Interestingly, npr1-1 mutants displayed reduced symptoms to both treatments, suggesting that the resistance to P. nicotianae and ZFF-induced disease symptoms occurs through a mechanism that requires SA and PAD4, but operates independently of NPR1. Although npr1-1 has been associated with SA signaling in many interactions [43,44] including the oomycete Hyaloperonospora arabidopsidis [11], genetic studies have provided evidence for immune responses that are SA-dependent but to not require NPR1 [45][46][47][48].
The responsiveness of PR-1 and PDF1.2 to P. nicotianae zoospores indicates that SA and JA pathways are responsive, in Arabidopsis, to Phytophthora pathogens [2,31]. Interestingly, ZFF treatment had a more pronounced effect on PDF1.2 expression than on PR-1 expression, suggesting that a component of ZFF was sufficient to trigger JA-dependent immune responses (Fig 3). It is conceivable that ZFF may be more involved in necrotrophic activities, or that activation of JA defenses may function to suppress SA defenses during biotrophic growth. This is apparent when SA function is attenuated; ZFF induced yellowing and rotting on the treated plants while induced no or very low PR-1 response in the mutants, and high concentration of zoospores invoked much more potent responses on defense signaling and symptoms (Figs 2 and 3B-3D). This regulatory feature of ZFF on plant defense may explain why addition of ZFF in a low concentration of zoospore inoculum which is technically equivalent to high concentration of zoospores in the term of the impact on quorum sensing mediated infection [3], enhanced susceptibility of plants (Fig 1). Similar complex responses have been well documented for necrotrophs with a broad host range which can produce diverse PAMPs that activate plant immune responses as well as virulence factors that suppress immune responses [49,50]. Understanding the exact molecular mechanisms that drive these responses will be a significant step towards understanding the basis of resistance and susceptibility to necrotrophic and hemi-biotrophic pathogens that cause substantial crop loss.
Using an external control is uncommon in a time course study on plant-pathogen interaction with qRT-PCR. In this study, we used SDW as the control to compare zoospore inocula and exudates at the same condition. However, SDW treatment caused responses of SA and JA defense signaling marker gene which sometimes was equivalent to those triggered by zoospores or ZFF, indicating that flooding may induce similar SA or JA responses. Net effects of zoospores and ZFF can be obtained by calibrating transcripts of these target genes from the same sampling time in the individual treatments with transcripts of the control. Alternatively, calibration may be done with samples at zero exposure time. However, reduction of such background may cover up real effects of zoospore inoculum or exudates on plant defense which gave rise to visual symptoms distinct from that of SDW (Fig 2). Nevertheless, to include external control in qRT-PCR for analyses of plant defense gene expression, it is warranted to use SA and JA marker genes that are not responsive to flood or to develop alternative feasible inoculation or treatment methods.
Supporting information S1