Inositol hexakisphosphate biosynthesis underpins PAMP‐triggered immunity to Pseudomonas syringae pv. tomato in Arabidopsis thaliana but is dispensable for establishment of systemic acquired resistance

Abstract

Phytic acid (inositol hexakisphosphate, InsP ₆) is an important phosphate store and signal molecule necessary for maintenance of basal resistance to plant pathogens. Arabidopsis thaliana (‘arabidopsis’) has three genes encoding myo‐inositol phosphate synthases (IPS1–3), the enzymes that catalyse conversion of glucose‐6‐phosphate to InsP, the first step in InsP ₆ biosynthesis. There is one gene for inositol‐(1,3,4,5,6)‐pentakisphosphate 2‐kinase (IPK1), which catalyses the final step. Previously, we showed that mutation of IPS2 and IPK1 but not IPS1 increased susceptibility to pathogens. Our aim was to better understand the InsP ₆ biosynthesis pathway in plant defence. Here we found that the susceptibility of arabidopsis (Col‐0) to virulent and avirulent Pseudomonas syringae pv. tomato was also increased in ips3 and ips2/3 double mutants. Also, ipk1 plants had compromised expression of local acquired resistance induced by treatment with the pathogen‐derived molecular pattern (PAMP) molecule flg22, but were unaffected in other responses to flg22, including Ca²⁺ influx and the oxidative burst, seedling root growth inhibition, and transcriptional up‐regulation of the PAMP‐triggered genes MITOGEN‐ACTIVATED PROTEIN KINASE (MPK) 3, MPK11, CINNAMYL ALCOHOL DEHYDROGENASE 5, and FLG22‐INDUCED RECEPTOR‐LIKE KINASE 1. IPK1 mutation did not prevent the induction of systemic acquired resistance by avirulent P. syringae. Also, ips2 and ips2/3 double mutant plants, like ipk1, were hypersusceptible to P. syringae but were not compromised in flg22‐induced local acquired resistance. The results support the role of InsP ₆ biosynthesis enzymes in effective basal resistance and indicate that there is more than one basal resistance mechanism dependent upon InsP ₆ biosynthesis.

Mutating Inositol‐Pentakisphosphate 2‐Kinase (IPK1) compromises PAMP‐triggered immunity. Activity of inositol phosphate synthases IPS2 and IPS3 supports other basal resistance phenomena.

1. INTRODUCTION

Phytic acid (myo‐inositol‐1,2,3,4,5,6‐hexakisphosphate: InsP ₆) is a phosphorylated derivative of myo‐inositol that is ubiquitous in eukaryotes. In plants, InsP ₆ is found abundantly in seeds and is thought to be an important reserve for phosphorus, myo‐inositol, and minerals (Lott et al., 2000; Otegui et al., 2002). InsP ₆ is also involved in the mediation of drought stress responses triggered by the hormone abscisic acid by controlling release of Ca²⁺ from endomembrane stores in guard cells (Lemtiri‐Chlieh et al., 2003). Subsequently, it was shown that the InsP ₆ biosynthesis pathway is required for defence against pathogens (Murphy et al., 2008). InsP ₆ is a cofactor for the auxin receptor TIR1 (transport inhibitor response 1) in conjunction with the SCF^TIR1 adaptor ASK1 (Tan et al., 2007), whilst other inositol polyphosphate species, Ins(1,2,4,5,6)P ₅ and InsP ₈, have been proposed to be cofactors for the jasmonic acid (JA) co‐receptor COI1 (Sheard et al., 2010; Laha et al., 2015). These findings indicate that InsP ₆ and related inositol polyphosphates have important biological functions in plants.

InsP ₆ biosynthesis can occur in either a lipid‐dependent or a lipid‐independent manner (Figure S1). In the lipid‐dependent pathway, phosphatidylinositol species are synthesized, and hydrolysis of phosphatidylinositol‐(4,5)‐bisphosphate by phospholipase C releases Ins(1,4,5)P ₃ (Brearley et al., 1997), which is subsequently phosphorylated by inositol polyphosphate kinases (Stevenson‐Paulik et al., 2002). Within the lipid‐independent pathway described in plants, the enzyme myo‐inositol phosphate synthase (IPS) catalyses the conversion of glucose‐6‐phosphate to InsP (Loewus et al., 1980, 1982). InsP is sequentially phosphorylated by inositol polyphosphate kinases, leading to the final phosphorylation of the 2‐OH of the myo‐inositide, catalysed by inositol‐(1,3,4,5,6)‐pentakisphosphate 2‐kinase (IPK1) to generate InsP ₆ (González et al., 2010; Baños Sanz et al., 2012).

InsP ₆ has a strong affinity for divalent metal ions due to its six phosphate groups. This property gives dietary InsP ₆ antinutrient properties in monogastric animals, including humans, because it chelates minerals and prevents their uptake by the gut. This is especially deleterious for people with mineral‐deficient diets (Cheryan and Rackis, 1980). InsP ₆ can also cause environmental damage. This arises when InsP ₆ in feed passes undigested through monogastric livestock (pigs, for example) and enters nearby watercourses, in which it can promote excessive algal growth, leading to eutrophication (Turner et al., 2002). The negative health and environmental impacts associated with InsP ₆ have provided an incentive to breed low phytic acid crops (Raboy, 2002). Hence, a variety of low phytic acid mutant lines have been investigated in maize, barley, and rice (Raboy et al., 2001).

However, given the important biological functions of InsP ₆ in plants, deployment of crops with total depletion of InsP ₆ should be done cautiously. Plants disrupted in the InsP ₆ biosynthesis pathway exhibit defects in plant development, abiotic stress tolerance, and biotic stress responses. Arabidopsis thaliana (hereafter ‘arabidopsis’) mutant plants with a T‐DNA insertion within the IPK1 gene (locus At5g42810) appear stunted under long day conditions, exhibit an early flowering phenotype (Lee et al., 2015), and have low levels of InsP ₆ in seeds (Stevenson‐Paulik et al., 2002; Kim and Tai, 2011) and vegetative tissue (Kuo et al., 2018). In arabidopsis, IPS enzymes are encoded by a small family of three genes: IPS1 (At4g39800), IPS2 (At2g22240), and IPS3 (At5g10170). Arabidopsis ips1 mutants are depleted in myo‐inositol (Donahue et al., 2010) but not in InsP ₆ (Kuo et al., 2018). The ips1 mutants, but not ips2 or ips3 mutants, have defects in plant development, characterized by curling and spontaneous lesions in leaves and distorted root caps (Meng et al., 2009; Chen and Xiong, 2010; Donahue et al., 2010). Arabidopsis ips1 mutants also exhibit impaired auxin transport in roots in a PIN2‐dependent manner (Chen and Xiong, 2010). Furthermore, arabidopsis ips2, but not ips1, mutant plants are hypersusceptible to a range of pathogens, including several RNA viruses and one DNA virus, the hemibiotrophic bacterial pathogen Pseudomonas syringae pv. tomato (Pst), and the necrotrophic fungus Botrytis cinerea (Murphy et al., 2008). Likewise, ipk1 mutant arabidopsis plants are also compromised in basal resistance to viruses, bacteria, and the cyst nematode Heterodera schachtii (Murphy et al., 2008; Jain et al., 2015; Ma et al., 2017). The weakened resistance seen in ips2 and ipk1 mutant plants supported a role for the InsP ₆ biosynthetic pathway in pathogen defence, yet it was unknown whether IPS3 is also involved in biotic stress responses.

Plants have evolved layers of defence to combat microbial pathogens. The initial layer is characterized by the perception and recognition of pathogen‐associated molecular patterns (PAMPs) by surface‐localized pattern recognition receptors (PRRs), leading to PAMP‐triggered immunity (PTI) (Boller and Felix, 2009). A classical PAMP recognized by a PRR is a sequence of 22 conserved amino acid residues in bacterial flagellin (flg22), which is perceived by the LRR‐RLK receptor FLS2 (flagellin sensing 2) (Boller and Felix, 2009). FLS2‐mediated perception of flg22 and its downstream signalling effects have become a paradigm for PTI in plants (reviewed in Boller and Felix, 2009). For instance, PAMP treatment triggers the rapid release of reactive oxygen species (ROS) (Felix et al., 1999; Kunze et al., 2004), a characteristic influx of cytosolic Ca²⁺ (Ma and Berkowitz, 2011), a cascade of sequential MAPK phosphorylation (Jagodzik et al., 2018), and a series of transcriptional changes (Clay et al., 2009). In turn, virulent pathogens have evolved specific avirulence (Avr) effectors to suppress a variety of PTI‐related processes. In the case of Pst these are secreted into the host cell through the Type III secretion injectisome (Arnold and Jackson, 2011; Block and Alfano, 2011).

Resistant host plants can recognize Avr effectors if they possess appropriate cytoplasmic NB‐LRR resistance (R) proteins. Recognition results in an accelerated, amplified resistance response termed effector‐triggered immunity (ETI) or the hypersensitive response (Carr et al., 2010). Elicitation of ETI can activate throughout the plant an enhanced state of readiness to inhibit infection upon any subsequent pathogen attack, a state called ‘systemic acquired resistance’ (SAR) (Carr et al., 2010). Effector recognition can occur indirectly, as is the case when NB‐LRR proteins detect effector‐induced changes in the host cell (Carr et al., 2010). For instance, in arabidopsis the Pst AvrRPT2 effector cleaves RIN4 (RPM1‐interacting protein), which can be detected by the R proteins, RPS2 and RPM1 (Kim et al., 2009). Another Pst effector, AvrB, phosphorylates the host protein RIN4 (Mackey et al., 2002). This leads to enhanced host susceptibility by up‐regulating JA‐dependent responses (He et al., 2004) in a manner dependent upon the JA co‐receptor, COI1 (Shang et al., 2006), and by activating MPK4 via an HSP90 interaction (Cui et al., 2010). Results from these and similar studies are exemplary of an evolutionary ‘zigzag’ arms race between the host and pathogen (Jones and Dangl, 2006).

Our previous study pointed to a role for InsP ₆ biosynthesis in maintaining basal defence responses (Murphy et al., 2008). It was thus the aim of our study to understand better the underlying molecular mechanisms of this role in plant defence. In this study, we present new findings on the effects of mutating the IPS3 and IPK1 genes on defence responses to Pst in arabidopsis. Interestingly, low‐InsP ₆ ipk1 mutant plants can establish SAR but are compromised in flg22‐induced resistance to localized infection, reinforcing the idea that the ability to synthesize InsP ₆ is important in the maintenance of basal resistance mechanisms such as PTI. However, ipk1 mutant plants appear to be unaffected in several well‐studied PAMP‐triggered responses, including the oxidative burst and Ca²⁺ influx, expression of early and late PTI marker genes, and inhibition of seedling growth.

2. RESULTS

2.1. Knockdown of IPS3 resulted in hypersusceptibility to virulent and avirulent Pst

In arabidopsis there is a family of three IPS genes. We previously described T‐DNA insertion mutants for AtIPS1 and AtIPS2 (Murphy et al., 2008), designated here as ips1 and ips2. In this work we identified an AtIPS3 mutant (corresponding to locus At5g10170) from the Salk Institute Genome Analysis Laboratory population of mapped insertions (Alonso et al., 2003: SALK_ 097,807) from a segregating population. Transcript levels for IPS1, IPS2, and IPS3 were compared in the three ips mutants and to the transformation control by reverse transcription coupled to the quantitative polymerase chain reaction (RT‐qPCR). IPS1 expression was enhanced in ips2 but unaffected in ips3 and ipk1 (Figure S2). IPS2 expression was enhanced in ips1 and ips3 and was unaffected in ipk1 while IPS3 expression was enhanced in ips2 but unaffected in ips1 and ipk1 (Figure S2). This indicated that the ips3 T‐DNA insertion mutant could be used as a loss‐of‐function mutant in pathology experiments.

We examined the responses of ips3 plants to virulent Pst and found that they were significantly more susceptible to the bacterial pathogen than transformation control plants (Figure 1). The susceptibility of ips3 to Pst was similar to that of ips2, previously shown to be hypersusceptible to Pst (Murphy et al., 2008), and the sid2 mutant (Wildermuth et al., 2001; Macaulay et al., 2017), which is impaired in salicylic acid (SA) biosynthesis and highly susceptible to pathogen attack (Figure 1). As would be anticipated, plants carrying both the ips2 and ips3 mutations showed increased susceptibility to virulent Pst (Figure S3). Assessing bacterial growth in the mutant and control plants in relation to infected tissue fresh weight (Figure 1a) or according to infected leaf area (Figure 1b) produced comparable results. Mutant ips3 plants also showed increased susceptibility to avirulent (harbouring the AvrB effector gene) and virulent Pst cells, as did ips2 and sid2 plants (Figures 2 and S3).

Figure 1.

The arabidopsis ips3 mutant is hypersusceptible to virulent Pseudomonas syringae pv. tomato (Pst). Leaf tissues from arabidopsis plants were infiltrated with a suspension of virulent Pst (10⁵ cfu/ml) and samples harvested at 2 days post‐inoculation, followed by serial dilution to determine accumulation of viable bacteria. (a) CFU in leaf samples (two leaves per plant, n = 4 plants) normalized to plant tissue fresh weight. (b) CFU in leaf samples (two leaves per plant, n = 4–5 plants) normalized to leaf disc area. Values not sharing the same lowercase letter are significantly different (p < .05: ANOVA and Tukey's test). Error bars represent standard error around the mean (SEM)

Figure 2.

The arabidopsis ips3 mutant is hypersusceptible to avirulent Pseudomonas syringae pv. tomato (Pst) expressing AvrB. Arabidopsis plants were infiltrated with a suspension of avirulent Pst (5 × 10⁵ cfu/ml). At 2 days post‐inoculation, leaf samples were harvested according to fresh weight (mg) for bacterial serial dilution assays. Results were pooled from three experiments for statistical analysis (three leaves per plant, n = 14–17 plants). Values not sharing the same lowercase letter are significantly different (p < .05: ANOVA and Tukey's test). Error bars represent SEM

2.2. Mutations affecting InsP ₆ biosynthesis did not inhibit the establishment of SAR

SAR can be induced in plants following inoculation with an avirulent pathogen and triggering of ETI/the hypersensitive response. As disruption of InsP ₈ interaction with the COI‐JAZ receptor diminishes JA‐regulated defences (Laha et al., 2015) and JA signalling is important in long‐distance signalling for SAR (Truman et al., 2007), it was possible that lesions in InsP ₆ biosynthesis may affect SAR. To determine if InsP ₆ biosynthesis is vital for establishment of SAR, ips1, ips2 ips3, and ipk1 mutant plants were inoculated on lower leaves either with Pst cells harbouring a plasmid expressing the AvrB gene or with mock inoculum (10 mM MgCl₂). Four days after inoculating the lower leaves, upper leaves were challenge inoculated with a suspension of virulent Pst and bacterial growth in these challenged leaves was determined 2 days later (Figure 3). Arabidopsis plants of the Col‐0 accession possess the R gene RPM1 that mediates indirect recognition of the bacterial effector AvrB and induces ETI (Boyes et al., 1998; Mackey et al., 2002). Non‐mutant (transformation control plants) exhibited induction of SAR, as indicated by decreased bacterial growth in the upper, Pst‐challenged leaves of plants that had previously been inoculated with cells of the avirulent bacterial strain (Figure 3). As expected, sid2 mutant plants, which are compromised in their ability to synthesize the SAR‐inducing signal SA, did not exhibit SAR (Figure 3). All four of the InsP ₆ biosynthetic mutants exhibited SAR, indicating that while inhibiting expression of IPS2, IPS3, and IPK1 diminishes basal resistance at the site of inoculation (Figures 1 and 2), it has no effect on defensive systemic signalling following ETI and the consequent establishment of SAR.

Figure 3.

Systemic acquired resistance is not dependent on InsP ₆ biosynthesis. Lower, fully expanded leaves of arabidopsis plants were initially inoculated with a mock solution (10 mM MgCl₂) or a 5 × 10⁵ cfu/ml suspension in 10 mM MgCl₂ of avirulent Pseudomonas syringae pv. tomato (Pst) harbouring a plasmid encoding the AvrB effector to stimulate systemic acquired resistance. Upper, noninoculated leaves were challenge‐inoculated with virulent Pst cells 4 days later. Two days following challenge, leaf samples were harvested for serial dilution assays. Results were pooled from five independent experiments for statistical analysis (two leaves per plant, n = 14–33 plants). Asterisks denote statistically significant differences between treatments of the respective genotypes (unequal variances with Welch's ANOVA, and Games–Howell post hoc test, *p < .05 and **p < .01)

2.3. A mutation affecting the final step of InsP ₆ biosynthesis inhibited flg22‐induced immunity

The best‐studied basal immunity mechanism is PTI, therefore we investigated the responses of mutants in InsP ₆ biosynthesis to one of the best studied experimental PAMPS, the synthetic oligopeptide flg22. Arabidopsis plants were infiltrated with water (control treatment) or a solution of flg22 1 day prior to a challenge with virulent Pst. As expected, transformation control plants exhibited flg22‐induced resistance to virulent Pst, whereas plants carrying a mutant allele for the flg22 receptor, FLS2, did not (Figure 4). The ips1, ips2, and ips3 mutants all exhibited flg22‐induced resistance to Pst, indicating that PTI was not impaired in these mutants (Figure 4). However, ipk1 mutant plants, like fls2 mutants, did not respond to flg22 (Figure 4). Thus, PTI is inhibited by mutation of IPK1 but not by mutation of IPS genes, suggestive that InsP ₆ is an important factor underpinning PTI.

Figure 4.

Local acquired resistance induced by flg22 was dependent on the IPK1 gene. Leaves of arabidopsis plants were locally infiltrated with a control (water) treatment or a 1 μM flg22 solution and 1 day later the same leaves were challenge‐inoculated with virulent Pseudomonas syringae pv. tomato (Pst) (10⁵ cfu/ml). Leaf discs were sampled at 3 days post‐inoculationand leaf extracts serially diluted to determine bacterial titres. (a) Results for transformation control (TC) plants and plants of the mutant lines ips1, ips2, ips3, ipk1 and the flg22‐insensitive mutant, fls2. (b) Data from additional experiments that included plants of the double ips2/3 mutant line. Each panel shows results pooled from three independent experiments for statistical analysis (two leaves per plant, n = 9–18 plants). Asterisks denote significant differences between the indicated samples (unequal variances with Welch's ANOVA, Games–Howell post hoc test, p < .05)

2.4. Double ips2/3 mutant plants were compromised in basal resistance but still exhibited flg22‐induced resistance

The ips2, ips3, and ips2/3 mutant plants were, like ipk1, hypersusceptible to Pst infection (Figure S3), but responded differently from ipk1 plants to flg22. IPS2 transcript accumulation was enhanced in ips3 and IPS3 expression was enhanced in ips2, suggesting that some regulatory cross‐talk occurs between the systems controlling IPS2 and IPS3 gene regulation (Figure S2), but both IPS2 and IPS3 expression was knocked down in the double ips2/3 mutant (Figure S4). We therefore tested the ability of the ips2/3 double mutant to exhibit flg22‐induced resistance to Pst (Figure 4b). The ips2/3 double mutant was hypersusceptible to Pst (Figure S3), but it was not impaired in its ability to exhibit flg22‐induced resistance (Figure 4b). Thus, although ips2, ips3, and ipk1 were compromised in their ability to resist Pst, only the ipk1 mutant was specifically compromised in flg22‐induced resistance.

2.5. The expression of flg22‐induced PTI marker genes was not affected in ipk1 mutant plants

To determine if InsP ₆ influences expression of PTI‐related genes, we investigated the steady‐state transcript levels of early (MPK3 and MPK11: maximal expression 30 min after treatment) and late (CAD5 and FRK1: maximal expression 3 hr after treatment) flg22‐dependent genes in the ipk1 mutant. Expression of MPK3 in transformation control plants increased by 2‐fold following flg22 treatment, and there was a similar (2.5‐fold) increase in ipk1 mutant plants (Figure 5a). Expression of MPK11 increased 7‐fold and 9‐fold in transformation control plants and ipk1 mutant plants, respectively (Figure 5b). Overall, the up‐regulation of MPK3 and MPK11 can be specifically attributed to flg22 treatment because the negative control, fls2 mutant, did not exhibit a transcriptional response to flg22 treatment (Figure 5).

Figure 5.

Mutation of the IPK1 gene did not inhibit flg22‐induced gene expression. Arabidopsis leaves (from transformation control TC, ipk1, and fls2 plants) were infiltrated with 1 μM flg22 or control solution (water) and sampled 30 min after treatment (for assay of MPK3 and MPK11) (a, b) or 3 hr after treatment for CAD5 and FRK1 (d, e). In each experiment, leaves from at least five plants per genotype were pooled for RNA extraction and RT‐qPCR with appropriate primers. Changes in transcript accumulation were normalized to the housekeeping gene GAPDH but accumulation of EF1a was used as an additional internal control (c, f). Error bars represent SEM. Asterisks denote significant differences from control treatments (ANOVA, Tukey's post hoc test, *p < .05). Three independent experiments were carried out with similar results

Expression of late‐expressed flg22‐induced genes, CAD5 and FRK1, was unaffected in ipk1 mutant plants (Figure 5d,e). As expected, the fls2 mutant did not exhibit any increase in the expression of flg22‐specific late marker genes (Figure 5d,e). Accumulation of the transcript for EF1a was used at both time points as an additional control and check for stability of expression of GAPDH (Figure 5c,f). Control gene expression was stable across treatments and genotypes, demonstrating that up‐regulation of flg22‐induced marker genes detected in ipk1 mutant plants can be specifically attributed to elicitation by the flg22 oligopeptide.

2.6. The flg22‐induced calcium influx, oxidative burst, and root growth inhibition were not affected in the ipk1 mutant

Recognition of PAMPs such as flg22 during PTI is associated with an influx of Ca²⁺ ions into the cell, which triggers production of superoxide anions catalysed by NADPH oxidase (Macho and Zipfel, 2014). As ipk1, ips2, and ips3 mutants were compromised in basal resistance and ipk1 was also compromised in flg22‐induced resistance, we examined the ability of mutants to exhibit the flg22‐triggered Ca²⁺ ion influx and oxidative burst using, respectively, an aequorin transgene and a luminol‐based assay. To examine whether plants depleted in InsP ₆ are affected with respect to the flg22‐triggered Ca²⁺ influx into the cytosol calcium influx signals, we generated transgenic plants expressing the Ca²⁺ reporter aequorin in the ipk1 mutant background (Dodd et al., 2010; Cheval et al., 2013). When treated with flg22, aequorin‐expressing transgenic ipk1 seedlings exhibited a similar influx of Ca²⁺ to aequorin‐transgenic arabidopsis possessing wild‐type IPK1 alleles, indicating that InsP ₆ is not a key factor in this PTI‐associated phenomenon (Figure S5). Using a luminol assay to detect reactive oxygen species generation, leaf discs from ips1, ips2, ips3, and ipk1 mutant plants responded to flg22 with an oxidative burst of the same magnitude and in the same time frame as control plants (Figure S6).

Seedling growth inhibition is another well‐characterized response of arabidopsis plants to flg22 (Gómez‐Gómez et al., 1999). When treated with varying concentrations of flg22, control plants displayed a decrease in root growth proportional to flg22 concentration (Figure S7) while the fls2 mutant (included as a control) did not exhibit flg22‐triggered growth inhibition. The ipk1 mutant plants also displayed flg22‐triggered inhibition of seedling root growth (Figure S7), indicating that the IPK1 gene is not involved in the molecular processes underlying growth inhibition induced by the PAMP flg22.

2.7. Hypersusceptibility to virulent Pst in ips2, ips3, and ipk1 mutant plants was lost after wounding

It was noted during experiments in which mock inoculation (infiltration of a control solution) was carried out on a lower leaf that subsequent bacterial growth on an upper leaf of the ips2, ips3, and ipk1 mutant plants was similar to the transformation control plants (Figures 3 and 4). This suggested that the mock inoculation rescued the impaired defence responses in the hypersusceptible InsP ₆ biosynthetic mutants. Experiments were then carried out to determine whether the cause of resistance induced by the mock treatment was due to wounding (as a result of the infiltration method, pressing a syringe against the tissue) or the infiltration of a solution into the apoplast (flooding of the intercellular space, potentially causing anoxia). The enhanced susceptibility to Pst exhibited by InsP ₆ biosynthesis mutants was lost after water infiltration (Figure S8) or wounding (Figure S9). These data indicate that the resistance induced by mock treatment is due to wounding and that wound‐induced resistance is not impaired in the InsP ₆ biosynthesis pathway mutants.

3. DISCUSSION

Previous results showed that InsP ₆ biosynthesis is required for the maintenance of basal resistance against viral, bacterial, and fungal pathogens in arabidopsis and potato (Murphy et al., 2008), and resistance to cyst nematode infestation (Jain et al., 2015). Interestingly, a recent study highlighted the importance of a higher‐order inositol polyphosphate, InsP ₈, in plant resistance to chewing herbivores (Laha et al., 2015). InsP ₈, like InsP ₆, is dependent on the enzymes IPS and IPK for its biosynthesis (Figure S1). To investigate the mechanism(s) underlying the role of InsP ₆ in pathogen resistance, we analysed the responses of ips3 mutant plants to microbial infection and dissected the various PAMP‐triggered responses in ipk1 mutant arabidopsis plants. We found that ips3 mutant plants were as hypersusceptible to virulent or avirulent Pst as ips2 mutant plants, which implicates IPS2 and IPS3 in synthesis of distinct pools of InsP ₆ that are required for effective basal defence. Furthermore, the vascular and hydathode‐specific expression of the IPS2 and IPS3 genes in arabidopsis (Donahue et al., 2010) strongly suggests a critical involvement of InsP ₆ biosynthesis in specific cells and tissues that can act as barriers to pathogen ingress (hydathodes) or dissemination through the host (vasculature). Hydathodes are open pores on the leaf margin that provide an entry point for bacterial pathogens such as Pst into the apoplast (Rufián et al., 2018). Elevated levels of defence‐related gene expression in cells surrounding hydathodes has been previously reported indicating a role of localized constitutive defence around these vulnerable openings (Macaulay et al., 2017; van den Burg et al., 2010).

Our finding that ipk1 mutant plants could exhibit SAR but not flg22‐induced resistance suggests a differentiation in the requirement for InsP ₆ biosynthesis between these two plant defence responses. We showed that basal levels of SA were not diminished in ips2 or ipk1 and that SA biosynthesis increased in response to inoculation with avirulent Pst in a similar manner to transformation control plants (Murphy et al., 2008). Taken together, our previous data and new findings demonstrate that the hypersusceptible ipk1 mutants are not impaired in SA‐mediated defensive signalling, nor in the ability to express SAR, but that normal synthesis of InsP ₆ is essential for PTI. However, an increase in resistance to Pst induced by injection of air or water into the apoplast (which we presume to be wound or stress induced: Figures S8 and S9) appears to be a form of defence that is distinct from PTI and not dependent on InsP ₆ biosynthesis.

The degree of impairment of flg22‐induced resistance in plants carrying a mutation in IPK1 was similar to that in plants with a mutation in FLS2. However, ipk1 mutant plants were not affected in a number of early flg22‐dependent responses, including the rapid oxidative burst, influx of Ca²⁺ ions, and the up‐regulation of flg22‐responsive transcripts. Our work contradicts certain previous findings on the effects of the ipk1 mutation on flg22‐elicited gene expression. We found that in plants grown in soil under a normal day/night regime flg22‐induced transcripts were up‐regulated to a similar extent in soil‐grown ipk1 mutant and control (IPK1) plants but that, nevertheless, flg22‐induced resistance to Pst was severely diminished in ipk1 mutant plants. However, Ma et al. (2017) found that in ipk1 mutant plants grown hydroponically and under continuous illumination, flg22‐induced transcriptional responses were diminished, although not abolished. The differences between these studies probably hinge on differences in growth conditions with, perhaps, the most important effect being photoperiod. Continuous light conditions disrupt the circadian system of the plant. This vital internal clock system regulates expression of defence‐related transcripts (Robertson et al., 2009) and responses of plants to infection (Genoud et al., 2002; Handford and Carr, 2007; Palukaitis et al., 2013).

Our results with ipk1 mutants suggest that the increased local resistance to Pst induced by flg22 is not dependent on the transcriptional up‐regulation of MAPKs, but still depends on InsP ₆ biosynthesis. Possible explanations may include redundancy in downstream defensive signalling, or that the system is not absolutely dependent on transcriptional changes in expression of these factors or that post‐transcriptional effects such as protein phosphorylation are more important than changes in the steady‐state accumulation of these proteins. Flg22‐induced resistance is also independent of signalling mediated by SA, JA, and ethylene (Zipfel et al., 2004). It was noted during this study that injecting air or water into the apoplast 24 hr before challenge with Pst induced an increase in resistance to this pathogen, presumably a form of wounding‐induced resistance. This form of resistance was not affected by the ipk1 mutation, which suggests that basal resistance involves multiple signalling pathways, not all of which require InsP ₆ biosynthesis for operation. In conclusion, it appears that InsP ₆ biosynthesis is required for flg22‐triggered resistance but it remains unknown whether resistance elicitation by other PAMPs also depends on this. However, based on the broad range of pathogens that are able to overcome basal resistance in InsP ₆‐depleted plants (Murphy et al., 2008), this would seem likely.

The hypersusceptibility to Pst of ips2 and ips3 single mutant plants indicates that both of the isoenzymes IPS2 and IPS3 are necessary for basal resistance to pathogen attack, that is, they are not redundant factors in maintenance of this form of defence. The double ips2/3 mutant was also hypersusceptible to Pst infection and, like the single ips2 and ips3 mutants, was not impaired in flg22‐induced resistance. In contrast, plants carrying a mutation in the IPK1 gene were compromised in basal resistance and in flg22‐stimulated resistance to Pst. Several possibilities might account for the difference between the ips and ipk1 mutants. The set of proteins contributing to the pools of synthesized InsP ₆ might be partitioned and differ for basal resistance or flg22‐induced resistance. In our experiments, we investigated the defence responses of mutants affected in steps in the InsP ₆ lipid‐independent pathway and found, for example, that ipk1 mutants were not compromised in flg22‐induced Ca²⁺ influx into the cytosol. However, it has been shown that plants depleted in lipid‐derived InsP ₃ have a diminished flg22‐induced Ca²⁺ influx response (Ma et al., 2017). It is possible, therefore, that both the lipid‐independent and the lipid‐dependent InsP ₆ biosynthetic pathway (Figure S1) might each be able to contribute to distinct InsP ₆ pools needed for different aspects of basal resistance and flg22‐induced responses.

In conclusion, our work confirms that normal InsP ₆ biosynthesis is important in maintaining basal defences against pathogens and also shows that it contributes to more than one defensive mechanism. In creating low‐phytate crops, it may be wise not to inhibit synthesis of InsP ₆ in all tissues but to do so only in those parts that are to be consumed in the human diet or processed as animal feeds, an approach shown to be feasible in, for example, soybean where levels of this metabolite were selectively decreased in the seeds (Kumar et al., 2019).

4. EXPERIMENTAL PROCEDURES

4.1. Plant material and growth conditions

A. thaliana accession Col‐0 plants were grown under short‐ or long‐day conditions as described below. Seeds for Atipk1, Atips1, and Atips2 mutants were from pools previously confirmed to be homozygous for the T‐DNA insertion (Murphy et al., 2008). AtIPS3 corresponds to locus At5g10170. Seeds for Atips3 (SALK_ 097,807) were obtained from the Nottingham Arabidopsis Stock Centre (NASC, http://www.arabidopsis.info), and a homozygous line for the T‐DNA insertion was obtained after verification by PCR genotyping and sequencing. Transformation control (indicated by TC in figures) plants do not have the T‐DNA insertion and were selected from the original segregating T‐DNA mutant populations to serve as controls in the experiments, as previously described by Murphy et al. (2008).

For infection, oxidative burst, and flg22‐induced transcription experiments, seeds were sown in a 4:1 compost:sand mixture and stratified for 2 days at 4 °C. Seeds were germinated and grown under short‐day conditions (8 hr light/16 hr dark cycles, 22 °C, 60% relative humidity, and 200 μmol⋅ m^{– 2}⋅s^{– 1} photosynthetically active radiation) in a Conviron growth room. For inhibition of seedling growth experiments, seeds were surface sterilized and stratified for 2 days at 4 °C on 0.5 × Murashige and Skoog (MS) basal salts, 1% (wt/vol) agar. Seeds were germinated and grown under long‐day conditions (16 hr light/8 hr dark cycles, 21 °C, and 200 μmol⋅m^{– 2}⋅s^–1 photosynthetically active radiation). For Ca²⁺ influx experiments, seedlings were grown in liquid 0.5 × MS basal salts medium.

4.2. Bacterial procedures

Virulent P. syringae pv. tomato (Pst) DC3000 was maintained on Luria–Bertani (LB) agar (Sambrook et al., 1989) containing 50 µg/ml rifampicin at 25 °C and avirulent Pst carrying the AvrB gene (Pst AvrB) was maintained in the same manner with additional antibiotic selection with 50 µg/ml kanamycin (Murphy et al., 2008). Bacterial inoculum was prepared by streaking out bacterial colonies 1 day prior to the experiment (Tornero and Dangl, 2001). On the day of the experiment, the colonies were initially suspended in 5 ml 10 mM MgCl₂ and diluted to achieve 10⁵ cells/ml for all infiltration experiments (unless otherwise stated). Leaves of 4‐week‐old arabidopsis plants (that had been grown under short‐day conditions) were inoculated using a needleless 1 ml syringe (Klement, 1963). Two days post‐inoculation (dpi), unless otherwise stated, leaf samples were taken to determine bacterial growth titres. Plant tissue samples were taken by recording fresh weight (mg) per leaf or taking 2‐mm diameter leaf discs with a cork borer. Samples were homogenized in 10 mM MgCl₂, serially diluted and plated on LB agar to determine the titre of viable bacterial cells, using counts of colony‐forming units (cfu).

4.3. Testing for establishment of SAR

A method modified from Cameron et al. (1999) was used to test for SAR induction in noninoculated leaves. Bacterial inoculum was prepared as described above. Lower leaves of arabidopsis plants were infiltrated with either 10 mM MgCl₂ (mock inoculum) or a suspension of avirulent Pst AvrB at 5 × 10⁵ cells/ml. Four days later, upper leaves of the same plants were challenge inoculated with a suspension of virulent Pst at 5 × 10⁵ cells/ml. Three days after the challenge inoculations, unless otherwise stated, leaf samples were taken and bacterial titres determined as described above.

4.4. Testing for flg22‐induced local acquired resistance

To test for flg22‐induced local acquired resistance, a modified method based on Zipfel et al. (2004) was used. Arabidopsis leaves were treated with either a mock treatment (water) or 1 μM flg22 24 hr before inoculation with a suspension of virulent Pst (10⁵ cells/ml). At 3 dpi, discs of tissue were harvested from the inoculated leaves and bacterial titres determined as described above.

4.5. Oxidative burst assay

A luminol–horseradish peroxidase (HRP) assay adapted from methods previously described (Whitehead et al., 1983; Keppler et al., 1989) was used to measure oxidative bursts in arabidopsis. Leaf discs (4.5 mm diameter) from 4‐week‐old arabidopsis plants grown under short‐day conditions were placed in opaque white 96‐well microtitre plates with 150 μl water per well overnight. The release of reactive oxygen species was measured using a luminol‐dependent assay by replacing the water with 100 μl of 100 μM luminol (Sigma) containing 10 μg/ml HRP (250 units/mg solid: Sigma) and the PAMP flg22 (100 nM) (Whitehead et al., 1983; Keppler et al., 1989; Kunze et al., 2004). Luminescence was measured immediately using a FLUOstar OPTIMA multimode microplate reader (BMG Labtech) for 20 min.

4.6. Seedling growth inhibition

Seedlings were grown vertically on 0.5 × MS agar plates containing varying concentrations of flg22 (10 nM, 1 μM, 2.5 μM). Root growth inhibition was assessed at 3 and 7 days after germination using ImageJ v. 1.43u (National Institutes of Health, Bethesda MD) to measure the lengths of primary roots from scanned images.

4.7. Reverse transcription‐coupled quantitative PCR and analysis of gene expression

Seeds were grown under short‐day conditions, and leaves from 4‐week‐old plants were infiltrated with 1 μM flg22 or water. At 30 min and 3 hr following treatment, infiltrated leaves were harvested and flash‐frozen in liquid nitrogen. Total RNA for reverse transcription‐coupled quantitative PCR (RT‐qPCR) analysis was extracted using TRIzol reagent (Invitrogen) according to the manufacturer's protocol, followed by lithium chloride precipitation, phenol–chloroform extraction, and ethanol precipitation (Berry et al., 1985). Total RNA samples were treated with TURBO‐DNase (Ambion) according to the manufacturer's protocol. Complementary DNA (cDNA) was synthesized from 0.5 μg total RNA using GoScript Reverse Transcription System (Promega) with oligo‐dT₁₅ primers following the manufacturer's instructions in the presence of RNaseOUT Recombinant Ribonuclease Inhibitor (Life Technologies). The cDNA was diluted 10‐fold and quantitative PCR (qPCR) was performed using SensiMix No ROX (Bioline) in 20 μl reactions, containing 250 nM each of the forward and reverse primers complementary to sequences of interest (Table S1). The transcripts for glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) and elongation factor 1A (EF1α) were used as internal standards (Westwood et al., 2013). PCRs were conducted in triplicate using the CFX96 Touch Real‐Time PCR Detection System (Bio‐Rad). Data were analysed using LinRegPCR v. 2014.8 (Hartfaal Centrum, the Netherlands) to calculate amplification efficiencies and C _t values. Expression levels of genes of interest relative to GAPDH and EF1α were calculated using the ΔΔC _t method (Livak and Schmittgen, 2001; Ramakers et al., 2003).

Supporting information

FIGURE S1 InsP ₆: its biosynthesis and turnover in plants

FIGURE S2 Expression of (a) IPS1, (b) IPS2, and (c) IPS3 was measured in ips and ipk1 mutant plants and compared to transformation control (TC) plants

FIGURE S3 Arabidopsis ips2/3 double mutant plants were hypersusceptible to virulent Pseudomonas syringae pv. tomato

FIGURE S4 Expression of IPS1, IPS2, and IPS3 in ips2/3 double mutant compared to transformation control (TC) plants

FIGURE S5 The flg22‐induced influx of Ca²⁺ ions was not affected by mutation of the IPK1 gene

FIGURE S6 The flg22‐induced oxidative burst was not affected in InsP ₆ biosynthetic mutants

FIGURE S7 Flg22‐induced root growth inhibition was unaffected in the ipk1 mutant

FIGURE S8 Pretreatment with water infiltration induced resistance to Pseudomonas syringae in the normally hypersusceptible ips2, ips3 and ipk1 mutants

FIGURE S9 Infiltration of leaves with air or water 1 day before challenge with Pseudomonas syringae induced resistance in nonmutant plants and ipk1 mutants but not in fls2 mutant plants

TABLE S1 List of primers used in this study

Poon JSY, Le Fevre RE, Carr JP, Hanke DE, Murphy AM. Inositol hexakisphosphate biosynthesis underpins PAMP‐triggered immunity to Pseudomonas syringae pv. tomato in Arabidopsis thaliana but is dispensable for establishment of systemic acquired resistance. Mol Plant Pathol. 2020;21:376–387. 10.1111/mpp.12902

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.