Pattern-Triggered Immunity Suppresses Programmed Cell Death Triggered by Fumonisin B1

Programmed cell death (PCD) is a crucial process for plant innate immunity and development. In plant innate immunity, PCD is believed to prevent the spread of pathogens from the infection site. Although proper control of PCD is important for plant fitness, we have limited understanding of the molecular mechanisms regulating plant PCD. Plant innate immunity triggered by recognition of effectors (effector-triggered immunity, ETI) is often associated with PCD. However pattern-triggered immunity (PTI), which is triggered by recognition of elicitors called microbe-associated molecular patterns (MAMPs), is not. Therefore we hypothesized that PTI might suppress PCD. Here we report that PCD triggered by the mycotoxin fumonisin B1 (FB1) can be suppressed by PTI in Arabidopsis. FB1-triggered cell death was suppressed by treatment with the MAMPs flg22 (a part of bacterial flagellin) or elf18 (a part of the bacterial elongation factor EF-Tu) but not chitin (a component of fungal cell walls). Although plant hormone signaling is associated with PCD and PTI, both FB1-triggered cell death and suppression of cell death by flg22 treatment were still observed in mutants deficient in jasmonic acid (JA), ethylene (ET) and salicylic acid (SA) signaling. The MAP kinases MPK3 and MPK6 are transiently activated and inactivated within one hour during PTI. We found that FB1 activated MPK3 and MPK6 about 36–48 hours after treatment. Interestingly, this late activation was attenuated by flg22 treatment. These results suggest that PTI suppression of FB1-triggered cell death may involve suppression of MPK3/MPK6 signaling but does not require JA/ET/SA signaling.


Introduction
Programmed cell death (PCD) is a genetic program for cellular suicide and plays an important role during immunity, biotic stress responses, development and senescence in multicellular organisms [1].The hypersensitive response (HR) to pathogen infection is a type of plant PCD that has been extensively studied [1].Although the biological relevance of HR-PCD is not fully understood, its timely activation and suppression are thought to be important for plant fitness.
Analysis of an Arabidopsis dde2/ein2/pad4/sid2 quadruple mutant showed that jasmonic acid (JA), ethylene (ET), and salicylic acid (SA) signaling all contribute positively to flg22-triggered immunity measured as growth inhibition of a virulent bacterial strain, Pseudomonas syringae pv.tomato DC3000 (Pto DC3000), in leaves [9].DDE2, EIN2 and SID2 are essential components of JA, ET, and SA signaling, respectively [10,11,12].The pad4 mutation affects SA signaling and other immune signaling [13,14].Although flg22-triggered immunity is only weakly affected in plants with mutations in any one of these genes, it was mostly abolished in the quadruple mutant [9].
PTI apparently suppresses HR triggered by inoculation of ETItriggering bacterial strains [15,16,17,18].Overlapping infiltration assays showed that HR triggered by avirulent bacterial pathogens was typically inhibited in the overlapping region that was pretreated with flg22 or the non-host pathogen, PTI-inducing Pseudomonas fluorescens [17].This suppression of HR by PTI is due to restriction of the type III protein secretion system's ability to inject type III effectors [18].It is host recognition of these effectors that triggers HR.Thus, it remains unknown whether PTI interferes with the plant signaling mechanisms that lead to induction of the HR.
Fumonisin B1 (FB1) is a mycotoxin produced by several species of Fusarium molds, including Fusarium verticillioides, which elicits PCD in both plant and animal cells [19,20,21].FB1 treated leaves develop lesions with characteristics reminiscent of HR, including the generation of reactive oxygen species, deposition of phenolic compounds and callose, accumulation of camalexin, and expression changes of defense-related genes [21].An Arabidopsis mutant deficient in vacuolar processing enzymes (VPEs), which are vacuole-localized cysteine proteases, is insensitive to FB1, indicating that VPEs are essential for FB1-triggered cell death [22].
Perception of flg22 triggers the activation of a mitogen activated protein kinase (MAPK) cascade consisting of MKK4 and MKK5 [mitogen-activated protein kinase kinase (MAPKK)] and MPK3 and MPK6 (MAPK) [23].MPK3 and MPK6 are also activated by other MAMPs such as elf18 and chitin [5,8].Activation of the endogenous MAPKs by induced expression of a constitutively active form of MKK4 or MKK5 leads to HR-like cell death [24].An mpk6 T-DNA insertion mutant shows attenuated seedling growth inhibition and cell features of programmed cell death upon FB1 exposure, indicating that MPK6 mediates signaling for FB1triggered cell death [25].
In this study, we hypothesized that PTI suppresses PCD and investigated the effect of PTI on FB1-triggered cell death.We used FB1 as an inducer of PCD since FB1 simulates HR-like PCD and highly purified FB1 is commercially available.FB1-triggered cell death was suppressed by PTI triggered by flg22 or elf18 (flg22-, and elf18-PTI).FB1-triggered cell death and its suppression by flg22-PTI were observed in wild-type Col-0 and the defense related signaling mutants, dde2, ein2, pad4, sid2 and dde2/ein2/pad4/ sid2 quadruple mutant.FB1triggered induction of MKK4 and MKK5 mRNA expression and phosphorylation of MPK3 and MPK6 at a late time point were inhibited by flg22 treatment.The suppression effect of flg22-PTI on FB1-triggered cell death was attenuated in mpk6 T-DNA insertion mutants.These results suggest that suppression of the MAPK cascade by PTI signaling results in suppression of FB1-triggered cell death and that prolonged activation of the MAPKs leads to eventually HR like cell death.

PTI suppresses FB1-triggered cell death
Flg22-PTI can strongly suppress the growth of a virulent Pseudomonas syringae strain if leaves are pretreated with flg22 one day prior to inoculation with bacteria, but no PCD is evident [9,26].Thus, strong immunity is not necessarily correlated with PCD, unlike an HR during ETI.We speculated that plants have a mechanism to avoid excessive PCD after sufficient activation of PTI.To investigate this hypothesis, we used FB1 as PCD inducer for the following reasons.(1) PCD triggered by bacterial strains expressing effectors that are recognized by R proteins was unsuitable to test the direct effect of PTI on PCD because delivery of effectors by the bacterial type III secretion system is suppressed by PTI [18].(2) In planta expression of such ETI-triggering effectors using an established chemical inducible expression system, such as dexamethasone [27] and estradiol [28] inducible systems, was inhibited by PTI (Figure S1A, B).This inhibitory effect of PTI was not transgene-specific since induced expression of an arbitrary reporter, the b-glucuronidase (GUS) gene, was also inhibited.(Figure S1C).Therefore, induced expression of ETItriggering effector genes was not suitable for the analysis of the PTI and PCD interaction.(3) FB1 develops lesions reminiscent of the HR [21,29].
FB1-triggered cell death was macroscopically observed as yellowing of leaves three to four days after inoculation of Arabidopsis mature leaves with 50 mM FB1 (Figure 1A).Leaves pretreated with 1 mM flg22 24 hours prior to FB1 treatment showed a reduction of FB1-triggered cell death visible as a lower level of leaf yellowing.This attenuation of PCD was confirmed when we analyzed the extent of cell death using trypan blue staining (Figure 1B) and electrolyte leakage measurements (Figure 1C).Flg22 pretreatment did not suppress FB1-triggered cell death in fls2, a flg22-receptor mutant, indicating that the flg22 effect depends on its specific receptor (Figure 1C).
We also tested PCD suppression by two other MAMPs, elf18 and chitin.We used chitosan, a partially de-acetylated, readily water-soluble derivative of chitin, in place of chitin.Chitosan induces responses very similar to those induced by chitin, and the response is dependent on CERK1, the chitin receptor, in Arabidopsis (Figure S2) [30].To ensure that chitosan treatment induced PTI, we examined the expression of PTI inducible genes.As shown in Figure S2, 1 mM flg22, 1 mM elf18, or 100 mg/ml chitosan treatment similarly induced the mRNA levels of Chitinase (At2g43620) and PR6 (At2g38870).Despite the similar induction of the marker genes, chitosan pretreatment did not suppress FB1triggered cell death, whereas elf18 pretreatment did (Figure 2).FB1 is a mycotoxin produced by several species of Fusarium molds which have cell walls containing chitin and other polysaccharides [31].The observation that FB1-triggered cell death is not affected by chitin-PTI suggests that the fungal toxin evolved to induce a PCD that is not inhibited by chitin-PTI.Otherwise FB1 would not be able to induce a PCD as fungi are coated with chitin containing cell walls.
Since flg22-PTI is well characterized, we focused on the flg22 effect for further analysis.Suppression of FB1-triggered cell death by flg22 was observed when flg22 was administered 48 or 0 hours prior to FB1 treatment (Figure S3A).To investigate whether the flg22 effect is dose-dependent, we co-infiltrated leaves with various concentrations of flg22 and FB1.As shown in Figure S3B, treatment with 50 nM flg22 was sufficient to suppress FB1triggered cell death.The fact that PCD suppression can be induced with very low amounts of flg22 suggests that the suppression is a biologically relevant phenomenon.

FB1-triggered cell death in hormone signaling mutants
The plant signaling molecules jasmonic acid (JA), ethylene (ET) and salicylic acid (SA) play important roles in plant immune responses [32,33].Interactions between the SA and the JA/ET signaling sectors can be both antagonistic and synergistic [9,34,35,36].These three signaling sectors are involved in PCD triggered by FB1 or oxidative stress [29,34,37].To investigate the involvement of these sectors in PTI suppression of FB1-triggered cell death, we measured electrolyte leakage after FB1 treatment in dde2, ein2, pad4, and sid2 single mutants as well as in a dde2/ein2/ pad4/sid2 quadruple mutant.DDE2, EIN2 and SID2 are essential components of the JA, ET and SA signaling sectors, respectively.The pad4 mutation affects the SA sector and expression of many SA-independent pathogen-induced genes [9,14,38].As shown in Figure 3A, FB1-triggered cell death estimated by electrolyte leakage was observed after FB1 treatment of each single and the quadruple mutant in mature leaves.A previous report indicated that FB1-triggered cell death requires JA, ET and SA-dependent signals in protoplasts [29].This suggests that FB1-triggered induction of cell death is not the same between intact leaves and protoplasts.Additionally, suppression of FB1-triggered cell death by flg22 was observed in each single and the quadruple mutant.The flg22 effect, which was estimated by subtracting the value of electrolyte leakage of flg22-pretreated leaves from the value of mock-pretreated leaves, was slightly different between Col-0 and pad4 (Figure 3B, P = 0.03), suggesting that PAD4 might play a minor role in mediating the flg22 effect.It has been reported that ethylene signaling controls FLS2 gene expression and responses to flg22 are impaired in ein2 [39].However, our previous work showed that early signaling events in response to flg22, such as MAPK activation and induction of MAMP-responsive gene expression, were intact [9] in the quadruple mutant.Taken together, the work cited above and our results that treatment with 50 nM flg22 suppressed FB1-triggered cell death (Figure S3b), indicate that early flg22-responses, which are JA, ET and SA independent, may be important for suppression of FB1-triggered cell death in intact adult leaves.

FB1 induced activation of MAPKs was affected by PTI
FB1 activates MAPK cascades in plant and animal cells [25,40,41].Growth suppression of seedlings on an FB1-containing medium was reduced in a T-DNA insertion mutant of MPK6 [25].Perception of flg22 or AvrRpt2, an ETI-triggering effector, by the cognate receptor triggers phosphorylation of MPK3 and MPK6 in a transient or sustained manner, respectively [9,42].Since FB1triggered cell death is macroscopically observed 3-4 days after FB1 treatment, we hypothesized that MPK3 and MPK6 are activated by FB1 at late time points in a sustained manner.As shown in Figure 4A, phosphorylation of MPK3, MPK6 and other MPKs were detected 36-48 hours after FB1 treatment.Interestingly, flg22 treatment inhibited FB1-triggered MPK3 and MPK6 phosphorylation but chitosan treatment did not (Figure 4B).To determine whether MPK3 and MPK6 signaling are essential for FB1-triggered cell death and its suppression by flg22, we measured electrolyte leakage after co-infiltration of FB1 and flg22 in mpk3-1 and mpk6-2 T-DNA insertion mutants.
In mpk6-2, but not in mpk3-1, FB1-triggered cell death was attenuated (P = 0.0055) (Figure 4C).This result is consistent with previous reports showing MPK6 plays a role in cell death triggered by FB1 and heat shock [25,43].The flg22 effect in mpk6-2, which was estimated by subtracting the ion conductivity value for FB1 with flg22 treatment from that for FB1 only treatment, was significantly smaller (23%) than that in Col-0 (Figure 4D).This was also observed with another mpk6 mutant allele (mpk6-3, Figure S4).Since both FB1-triggered cell death and the flg22 effect were not completely abolished in mpk6 mutants, we concluded that MPK6-dependent as well as MPK6-independent signaling pathways are involved in these responses.MPK3 may be involved, as MPK3 and MPK6 function redundantly in some biological processes [23,44].In the Arabidopsis genome, there are 10 predicted MAPKKs [45].Previous studies showed that MKK4 and MKK5 are the upstream kinases activating MPK3 and MPK6 during abiotic and biotic stress-response signaling [23,45,46].Our previous mRNA profiling data showed that mRNA levels of MKK4 increased during infection with an ETI-triggering bacterial strain, Pto DC3000 AvrRpt2 [47].The MKK4 and MKK5 mRNA levels increased 24 hours after flg22 treatment and 48 hours after FB1 treatment (Figure 4E).The induced MKK4 and MKK5 mRNA levels were higher with FB1 treatment than with flg22 treatment.Flg22 treatment reduced the FB1-induced levels of MKK4 and MKK5 mRNA, whereas chitosan treatment did not (Figure 4E).Induced expression of constitutively active forms of MKK4 and MKK5 leads to HR-like cell death [24].Therefore, late and prolonged MKK4 and MKK5 mRNA induction by FB1 treatment may be important for FB1-triggered cell death, and reduction of FB1-induced MKK4 and MKK5 mRNA levels by flg22 treatment may be a cause of attenuation of FB1-triggered cell death by flg22 treatment.
Vacuolar processing enzymes (VPEs) play an important role for both FB1-triggered cell death and heat shock triggered PCD (HS-PCD) [22,43].VPE mRNAs increase after FB1 and HS treatment [22,43].Heat shock activates the MAPK cascade, and MPK6 is important for heat shock induction of the VPE genes [43].We found that flg22 treatment did not reduce the FB1-induced level of cVPE gene expression (Figure S5A).In contrast with HS treatment, FB1 induction of cVPE gene expression was not suppressed by a knockout mutation of MPK6 (Figure S5B).These results indicate that FB1-and HS-PCD use different signaling mechanisms for cell death induction, and that induced expression of VPEs is not sufficient to induce cell death.
In conclusion, our observations suggest that suppression of FB1triggered cell death by flg22 treatment is partially mediated by flg22-triggered interference with late and prolonged induction of MKK4 and MKK5 gene expression (Figure 5), which affects signaling mediated by the downstream MAPK MPK6.
We hypothesize that suppression of PCD by PTI is not exclusive to FB1-triggered cell death since this system seems to benefit plants by avoiding excessive and wasteful PCD responses.It will be interesting to investigate the effect of PTI on ETI-associated PCD (i.e., the HR) and, if an effect is detectable, the factors that are involved in mediating the PTI effect.Comparison of such results with what we learned about the PTI effect on FB1-triggered cell death will deepen our understanding of signaling mechanisms involved in plant PCD.
Plants were grown in an environment-controlled chamber at 22uC with 75% relative humidity and a 12h/12h light/dark cycle.
Polymeric chitosan was used as inducer of chitin-PTI.Chitosan (C3646; SIGMA, St. Louis, MO, USA) was dissolved in 0.05 % (v/v) acetate to make 2 mg/ml solution and diluted in water for inoculation (final concentration: 100 mg/ml).Fumonisin B1 (F1147; SIGMA) was dissolved in methanol to make 5 mM FB1 and diluted in water for inoculation (final concentration: 50 mM).One percent methanol (v/v) was used as a negative control for FB1 treatment.

Trypan blue staining
Trypan blue staining stock solution was 0.1% (w/v) trypan blue in solvent consisting of equal volumes of water saturated-phenol, lactate, glycerol and water.One volume of this stock solution was diluted with 2 volumes of 95% ethanol before use.Leaves were boiled in the diluted trypan blue staining solution for 1 minute and destained in chloral hydrate solution (1.25 g/ml chloral hydrate in water) for 2 days.

Electrolyte leakage analysis
Three-or four-week old plants were infiltrated with 50 mM FB1. Forty-eight hours after infiltration, leaf disks (6 or 7 mm diameter) were made and floated in purified water for 48 hours under 12h/ 12h light/dark conditions.Electrolyte leakage values were determined by measuring conductivity with an electrical conductivity meter (B-173, Horiba, Japan).
Log 10 -transformed conductivity values were analyzed.The following mixed-effect model was fit to the log 10 -transformed conductivity data using the lme function of the nlme package in the R environment (http://www.r-project.org):C gptr = GPT gpt +R r +e gptr , where C, log 10 -transformed conductivity; GPT, genotype:pretreatment:treatment interaction and random factors; R, independent experiment, e, residual.The mean estimate of the genotype:pretreatment:treatment interaction was used as the modeled log 10 -transformed conductivity.The flg22-triggered suppression of electrolyte leakage was calculated by subtracting the electrolyte leakage values of the flg22-treated samples from the values of the mock-treated samples.Significance of differences was determined using two-tailed t-tests.For the t-tests, the standard error appropriate for each comparison was calculated using the variance and covariance values obtained from the model fitting.(A, B) Activation of MAPKs by FB1 treatment.The fifth and sixth true leaf of three-week old plants were treated with water (mock), 1 mM flg22 or 100 mg/ml chitosan and 1% methanol (mock) or 50 mM FB1, and samples were collected at 0 to 48 (A) or 48 (B) hours after treatment as indicated.Each sample was a pool of four to six leaves.Activated MAP kinases were detected by immunoblotting using the anti-phospho-p44/42 MAPK antibody.Equal loading of samples was confirmed by Ponceau S staining.Experiments were conducted twice with similar results.(C) Quantification of FB1-triggered cell death by electrolyte leakage.The sixth true leaf of three-week old plants was treated with water (mock) or 1 mM flg22 and 1% methanol (mock) or 50 mM FB1. Leaf disks (7 mm diameter) were prepared 48 hours after FB1 treatment.One leaf disk, from one leaf was floated in wells of 48-well plates each containing 0.5 ml of purified water.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: eight biological replicates were performed for each of two independent experiments.The vertical axis shows log 10 -transformed conductivity values.Yellow arrows indicate the levels of the flg22 effect.Asterisks indicate significant differences from mock (P,0.001,two-tailed t-tests).(D) Comparison of the flg22 effect in MAPK mutants for the FB1-triggered cell death.The flg22 effect was estimated by subtracting the electrolyte leakage values of flg22-treated samples from the values of mock-treated samples.The flg22 effect on each mutant was compared with that on Col-0 using a two-tailed t-test using the standard errors calculated based on the mixed linear model to obtain the P-values.(E) Induction of MKK4 and MKK5 mRNA expression by FB1.The fifth and sixth true leaves of three-week old plants were treated with water (mock), 1 mM flg22 or 100 mg/ml chitosan (chitin) and 1% methanol or 50 mM FB1, and samples were collected at 24 and 48 hours after treatment.Each sample was a pool of four leaves from two plants.The mRNA levels of MKK4 and MKK5 were measured by qRT-PCR.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: two or three biological replicates were performed foreach of two independent experiments.The vertical axis is the log 2 expression level relative to that of Actin2 (At3g18780).Asterisks indicate significant differences from mock treatment (P,0.001,two-tailed t-tests).doi:10.1371/journal.pone.0060769.g004 The raw data from the electrolyte leakage analysis are included in Table S2.

Quantitative RT-PCR analysis
Total RNA was isolated using TRIzol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol.Quantitative RT-PCR was performed using a Superscript III Platinum SYBR Green One-Step qRT-PCR kit (Invitrogen) and a LightCycler480 machine (Roche, Basel, Switzerland).Primers for qRT-PCR are listed in Table S1.The following mixed-effect model was fit to the Ct value data using the lme function of the nlme package in the R environment (http://www.r-project.org):Ct gtr = GT gt +R r +e gtr , where GT, gene:treatment interaction and random factors; R, independent experiment; e, residual.The mean estimates of the gene:treatment interaction were used as the modeled Ct values.The relative log 2 expression values were obtained by subtracting the Ct value of the genes from the Ct value of the Actin2 (At3g18780) gene and compared for each gene using two-tailed ttests.For the t-tests, the standard error appropriate for each comparison was calculated using the variance and covariance values obtained from the model fitting.

Measurement of MAPK activation
The activated MAPKs were detected as described previously [50].
The fifth and sixth leaves of three-week-old plants were infiltrated with 50 mM FB1. Samples, each of which was a pool of six leaves from three plants, were collected at each time point and flash frozen in liquid nitrogen.The frozen samples were ground in liquid nitrogen and homogenized in 100 ml of extraction buffer (100 mM HEPES, pH 7.5, 5 mM EDTA, 5 mM EGTA, 2 mM dithiothreitol, 10 mM Na 3 VO 4 , 10 mM NaF, 50 mM ßglycerolphosphate, 1 mM phenylmethylsulfonyl fluoride, 1 tablet/ 10 ml extraction buffer of proteinase inhibitor cocktail (11 836 153 001; Roche Applied Science, Indianapolis, IN, USA) and phosphatase inhibitor cocktail (04 906 845 001; Roche Applied Science), 10% glycerol, 1% (w/v) polyvinylpolypyrrolidone). The protein concentration was determined using a Bradford assay (Bio-Rad, Hercules, CA, USA) with BSA as a standard.Sixty micrograms of protein per lane was separated in a 10% polyacrylamide gel.Immunoblot analysis was performed using anti-phospho-p44/42 MAPK (1:2000, Cell Signaling Technology, Danvers, MA, USA) as the primary antibody, peroxidaseconjugated goat anti-rabbit IgG (1:15,000, A6154; Sigma) as the secondary antibody, and Pierce Protein-Free T20 (TBS) Blocking Buffer (37351; Thermo Scientific, Waltham, MA, USA).

Supporting Information
Figure S1 Attenuation of chemical inducible gene expression system in plant.Transgenic plants carrying the transgenes for estradiol (Ed)-inducible AvrRpt2 (parent vector: pER8), dexamethazone (Dex)-inducible AvrRpm1 (parent vector: pTA7002, gift from Jeff Dangl's Lab) or estradiol (Ed)-inducible bglucuronidase (GUS, parent vector: pER8) were used for this experiment.Twenty-four hours after pretreatment with water (mock) or 1 mM flg22, 10 mM estradiol (A, C) or dexamethazone (B) was infiltrated into the pretreated leaves of four-week old plants.The mRNA levels of AvrRpt2 (A), AvrRpm1 (B) or GUS (C) were measured by qRT-PCR.Bars represent means and standard errors for four (A) or three (B, C) biological replicates, calculated using a mixed linear model.The vertical axis is the log 2 -transformed expression level relative to that of Actin2 (At3g18780).The level of expression in flg22-pretreated leaves was compared with that in mock-pretreated leaves using a two-tailed t-test using the standard errors calculated based on the mixed linear model to obtain the Pvalues.Expression levels below the dection limit of the qPCR system are shown as n.d.(not determined).(PDF) Figure S2 Induction of Chitinase and PR6 mRNA by flg22, elf18 and chitosan.Four-week old plants were treated with 1 mM flg22, 1 mM elf18 or 100 mg/ml chitosan, and samples were collected 3 hours after MAMP treatment.Each sample was a pool of six mature leaves.The mRNA levels of Chitinase (At2g43620) and PR6 (At2g38870) were measured by qRT-PCR.The receptor deficient mutants (fls2: SAIL_691C4, efr-2: SALK_068675, and cerk1-2: GABI_096F09) were included as a negative control for each MAMP.Bars and error bars represent means and standard deviation for three independent experiments.The vertical axis is the log 2 expression level relative to that of Actin2 (At3g18780).(PDF) Figure S3 During and concentrations of flg22 treatment required for suppression of FB1-triggered cell death.(A, B) Quantification of FB1-triggered cell death by electrolyte leakage.This experiment was performed as described in Figure 1C except for the following differences.Forty-eight hours (A) or 0 hour (A, B) after pretreatment with water (mock) or 50 nM (B) to 1 mM flg22 (A, B), 50 mM FB1was infiltrated into the pretreated leaves of four-week old plants.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: four to six biological replicates were performed for each of two independent experiments.(PDF) Figure S4 Quantification of FB1-triggered cell death by electrolyte leakage in mpk6-3.This experiment was performed as described in Figures 4C and D. Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: five to six biological replicates were performed for each of two independent experiments.(PDF) Figure S5 Induction of cVPE mRNA by FB1.The fifth and sixth true leaves of three-week old plants were treated with water (mock), 1 mM flg22 or 100 mg/ml chitosan (chitin) and 1% methanol (mock) or 50 mM FB1, and samples were collected at 24 and 48 hours after treatment.Each sample was a pool of four leaves, from two plants each.The mRNA levels of cVPE were measured by qRT-PCR.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: two biological replicates were performed for each of two independent experiments.The vertical axis depicts the log 2 transformed expression values relative to that of Actin2 (At3g18780).Asterisks indicate significant differences between mock and MAMPs (A) or Col-0 and mpk mutants (B) (P,0.001,two-tailed t-tests).

(PDF)
Table S1 The sequences of the primers used in quantitative RT-PCR.

(XLSX)
Table S2 The raw ion conductivity data used in the figures. (XLSX)

Figure 1 .
Figure 1.Flg22 pretreatment suppressed FB1-triggered cell death.Twenty-four hours after pretreatment with water (mock) or 1 mM flg22, 1% methanol (mock) or 50 mM FB1 were infiltrated into the pretreated leaves of four-week old plants.Three days after FB1 treatment, the leaves were subjected to macroscopic examination (A) or trypan blue staining (B) to visualizeFB1-triggered cell death.Scale bar = 1 cm.(C) Quantification of FB1-triggered cell death by electrolyte leakage.Leaf disks (6 mm diameter) were prepared 48 hours after FB1 treatment.Four leaf disks, from two leaves, were floated on in wells of 12-well plate each containing 2 ml of purified water.Bars represent means and standard errorsfor biological replicates, calculated using a mixed linear model: four to six biological replicates were performed for each of two independent experiments.The vertical axis shows the log 10 -transformed conductivity values.Asterisks indicate significant differences from the mock-pretreatment (P,0.001,two-tailed t-tests).doi:10.1371/journal.pone.0060769.g001

Figure 2 .
Figure2.Chitosan treatment did not suppress FB1-triggered cell death.FB1-triggered cell death was quantified by electrolyte leakage measurements.This experiment was performed as described in Figure1Cexcept for the following differences.Twenty-four hours after pretreatment with water (mock), 1 mM flg22, 1 mM elf18 or 100 mg/ml chitosan, 1% methanol (mock) or 50 mM FB1 or were infiltrated into the pretreated leaves of four-week old plants.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: three to six biological replicates wereperformed for each of two independent experiments.doi:10.1371/journal.pone.0060769.g002

Figure 3 .
Figure 3. Flg22 pretreatment suppressed FB1-triggered cell death in dde2, ein2, pad4, sid2 and quadruple mutants.(A) Quantification of FB1-triggered cell death by electrolyte leakage.This experiment was performed as described in Figure 1C except for the following differences.Twenty-four hours after pretreatment with water (mock) or 1 mM flg22, 1% methanol (mock) or 50 mM FB1 was infiltrated into the pretreated leaves of four-week old plants.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: four biological replicates were performed foreach of two independent experiments.Yellow arrows indicate the strength of the flg22 effect.Asterisks indicate significant differences from mock (P,0.001,two-tailed t-tests).(B) Comparison of the flg22 effect in the mutants for FB1-triggered cell death.The flg22 effect was estimated by subtracting the electrolyte leakage value of the flg22-treated sample from the value of the corresponding mock-treated sample.The flg22 effect in each mutant was compared with that in Col-0 using a twotailed t-test using the standard errors calculated based on the mixed linear model to obtain the P-values.Quadruple mutant of dde2/ein2/ pad4/sid2 was shown as quad.doi:10.1371/journal.pone.0060769.g003

Figure 4 .
Figure 4. Flg22 treatment inhibited the MKK4/MKK5-MPK6 pathway, which positively regulated FB1-triggered cell death.(A,B) Activation of MAPKs by FB1 treatment.The fifth and sixth true leaf of three-week old plants were treated with water (mock), 1 mM flg22 or 100 mg/ml chitosan and 1% methanol (mock) or 50 mM FB1, and samples were collected at 0 to 48 (A) or 48 (B) hours after treatment as indicated.Each sample was a pool of four to six leaves.Activated MAP kinases were detected by immunoblotting using the anti-phospho-p44/42 MAPK antibody.Equal loading of samples was confirmed by Ponceau S staining.Experiments were conducted twice with similar results.(C) Quantification of FB1-triggered cell death by electrolyte leakage.The sixth true leaf of three-week old plants was treated with water (mock) or 1 mM flg22 and 1% methanol (mock) or 50 mM FB1. Leaf disks (7 mm diameter) were prepared 48 hours after FB1 treatment.One leaf disk, from one leaf was floated in wells of 48-well plates each containing 0.5 ml of purified water.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: eight biological replicates were performed for each of two independent experiments.The vertical axis shows log 10 -transformed conductivity values.Yellow arrows indicate the levels of the flg22 effect.Asterisks indicate significant differences from mock (P,0.001,two-tailed t-tests).(D) Comparison of the flg22 effect in MAPK mutants for the FB1-triggered cell death.The flg22 effect was estimated by subtracting the electrolyte leakage values of flg22-treated samples from the values of mock-treated samples.The flg22 effect on each mutant was compared with that on Col-0 using a two-tailed t-test using the standard errors calculated based on the mixed linear model to obtain the P-values.(E) Induction of MKK4 and MKK5 mRNA expression by FB1.The fifth and sixth true leaves of three-week old plants were treated with water (mock), 1 mM flg22 or 100 mg/ml chitosan (chitin) and 1% methanol or 50 mM FB1, and samples were collected at 24 and 48 hours after treatment.Each sample was a pool of four leaves from two plants.The mRNA levels of MKK4 and MKK5 were measured by qRT-PCR.Bars represent means and standard errors for biological replicates, calculated using a mixed linear model: two or three biological replicates were performed foreach of two independent experiments.The vertical axis is the log 2 expression level relative to that of Actin2 (At3g18780).Asterisks indicate significant differences from mock treatment (P,0.001,two-tailed t-tests).doi:10.1371/journal.pone.0060769.g004

Figure 5 .
Figure 5.A proposed model for the FB1-triggered cell death and suppression by flg22-PTI.The MKK4/MKK5-MPK6 pathway partially contributes to FB1-triggered cell death.PTI may suppress FB1triggered cell death by suppression of MKK4 gene expression and by other so far unknown pathways.doi:10.1371/journal.pone.0060769.g005