In-depth analysis of isochorismate synthase-derived metabolism in plant immunity: Identification of meta-substituted benzoates and salicyloyl-malate

Isochorismate-derived metabolism enables biosynthesis of the plant defense hormone salicylic acid (SA) and its derivatives. In Arabidopsis thaliana, the stress-induced accumulation of SA depends on ISOCHORISMATE SYNTHASE1 (ICS1) and also requires the presumed isochorismate transporter ENHANCED DISEASE SUSCEPTIBILITY5 (EDS5) and the GH3 enzyme avrPphB SUSCEPTIBLE3 (PBS3). By comparative metabolite and structural analyses, we identified several hitherto unreported ICS1- and EDS5-dependent, biotic stress-inducible Arabidopsis metabolites. These involve meta-substituted SA derivatives (5-formyl-SA, 5-carboxy-SA, 5-carboxymethyl-SA), their benzoic acid (BA) analogs (3-formyl-BA, 3-carboxy-BA, 3-carboxymethyl-BA), and besides the previously detected salicyloyl-aspartate (SA-Asp), the ester conjugate salicyloyl-malate (SA-Mal). SA functions as a biosynthetic precursor for SA-Mal and SA-Asp, but not for the meta-substituted SA- and BA-derivatives, which accumulate to moderate levels at later stages of bacterial infection. Interestingly, Arabidopsis leaves possess oxidizing activity to effectively convert meta-formyl- into meta-carboxy-SA/BAs. In contrast to SA, exogenously applied meta-substituted SA/BA-derivatives and SA-Mal exert a moderate impact on plant immunity and defence-related gene expression. While the isochorismate-derived metabolites are negatively regulated by the SA receptor NON-EXPRESSOR OF PR GENES1, SA conjugates (SA-Mal, SA-Asp, SA-glucose conjugates) and meta-substituted SA/BA-derivatives are oppositely affected by PBS3. Notably, our data indicate a PBS3-independent path to isochorismate-derived SA at later stages of bacterial infection, which does not considerably impact immune-related characteristics. Moreover, our results argue against a previously proposed role of EDS5 in the biosynthesis of the immune signal N-hydroxypipecolic acid and associated transport processes. We propose a significantly extended biochemical scheme of plant isochorismate metabolism that involves an alternative generation mode for benzoate- and salicylate-derivatives.

Supporting Figure 1.The meta-substituted salicylic acid-and benzoic acid-derivatives detected in plant extracts and authentic compounds possess identical GC retention times and mass spectra.A-E, Overlays of ion chromatograms of a GC-MS sample from extracts of Psm-inoculated Arabidopsis npr1 leaves (red) and the same leaf extract sample supplemented with 100-200 ng of the indicated authentic substances.The co-injection of extract sample and authentic compounds yield single peaks with identical retention times.A, 5-formyl-salicylic acid (5-FSA, 1).B, 5-carboxy-salicylic acid (5-CSA, 2).C, 3-formyl-benzoic acid (3-FBA, 4).D, 3-carboxy-benzoic acid (3-CBA, 5).E, 3-carboxymethylbenzoic acid (3-CMBA, 6).F-J, Mass spectra of authentic, trimethylsilyl-diazomethane-derivatised 5-FSA (F), 5-CSA (G), 3-FBA (H), 3-CBA (I), and 3-CMBA (J) are identical to the mass spectra of the plant-derived meta-substituted SA/BA-derivates (Fig. 1).The bars represent the mean ± SD of three biological replicates from different plants, each replicate consisting of six leaves from two plants.Individual data points of biological replicates are super-imposed on the bar graphs (small circles).Please note the similar accumulation patterns of the analytes in the data sets derived from GC-MS-(Fig. 4) and LC-qTOF-MS-based analysis.Due to the identical masses and similar retention times of 5-FSA and 3-CBA and a consequent overlap of peaks in the EICs, an exact quantification of 3-CBA was not possible via LC-qTOF-MS and therefore not presented in B.  Racemic malic acid (1) (3.35 g, 25.0 mmol, 1.00 equiv) was placed under nitrogen atmosphere in a screw cap Schlenk tube with magnetic stir bar and dissolved in toluene (40 mL).To the solution was added benzyl alcohol (2) (5.41 g, 50.0 mmol, 2.00 equiv), p-toluene sulfonic acid (0.047 g, 0.025 mmol, 1.00 mol%) and the mixture was heated heated to reflux for 24 h.After cooling to room temp, the solvent was removed under reduced pressure and the crude product was purified by column chromatography on silica gel (n-hexane/ethyl acetate 5:1) to give dibenzyl rac-2-hydroxysuccinate (rac-3) (6.01 g, 19.1 mmol, 71%) as a pale-yellow oil, Rf (nhexane/ethyl acetate 5:1): 0.13.

(C14H12O3 [228.24])
Methyl 2-hydroxybenzoate (28.0 g, 180 mmol, 1.00 equiv) was placed in a 1000 mL round bottom flask with magnetic stir bar and a reflux condenser and dissolved in acetonitrile (750 mL).Potassium carbonate (49.5 g, 360 mmol, 2.00 equiv) was added under vigorous stirring and the reaction mixture was stirred at room temp for 10 min.Benzyl bromide (32.4,193 mmol, 1.07 equiv) was then added to the mixture with constant stirring and the reaction mixture was heated to reflux for 14 h.After cooling to room temp, the solvent was removed under reduced pressure.The crude product was dissolved in water (150 mL) and extracted with ethyl acetate (3 × 25 mL).The collected organic layers were dried with anhydrous sodium sulfate, filtered and then the solvents were removed under reduced pressure.The crude product was dissolved in 1,4-dioxane (50 mL) and aqueous sodium hydroxide solution (1.5 M, 60.0 mL) was added.The mixture was heated to reflux for 2 h and then the solvent was removed under Fig.S3 reduced pressure.The aqueous phase was then extracted with ethyl acetate (3x 50 mL).The organic phase was dried with anhydrous sodium sulfate, then filtered and the solvent removed under reduced pressure to give O-benzyl salicylic acid (4) (32.5 g, 143 mmol, 80%) as a yellowish solid, Rf (n-hexane/ethyl acetate 3:1) = 0.90.In a Schlenk tube with a rubber septum and magnetic stir bar, rac-2-{[2-(benzyloxy)benzoyl]oxy}succinate (rac-5) (2.00 g, 3.81 mmol, 1.00 equiv) were placed 10 wt% Pd/C (40 mg, 40 mmol) and ethyl acetate (20 mL).The reaction mixture was purged with a stream of nitrogen for 1 min.Via a cannula, hydrogen gas was passed through the septum into the reaction mixture (approx.180 mL of gas until the solution was saturated).The reaction mixture was stirred at room temp for 14 h.The reaction mixture was filtered, the solvent was removed under reduced pressure, and the crude solid product was crystallized from dichloromethane/acetone (50:1) to give rac-O-salicyloyl malic acid (rac-6) (0.92 g, 3.61 mmol, 95%) as a colorless solid, Rf (n-hexane/ethyl acetate 3:1) = 0.41.and 3).Note that the samples were derivatised with trimethylsilyl-diazomethane to convert carboxylic acid groups into methyl ester groups prior to GC-MS analysis.The methyl group introduced by derivatisation is indicated in grey.

6 .Supporting Figure 8 .
meta-Substituted SA/BA-derivatives and other SA-related metabolites accumulate in both compatible and incompatible plant-bacterial interactions.Levels of different ICS1-dependent metabolites at 24 and 48 hours post inoculation (hpi) with a suspension of OD 600 = 0.005 of the compatible (virulent) Psm strain (Psm), a suspension of OD 600 = 0.005 of the incompatible (avirulent) Psm avrRpm1 strain (Psm avrRpm1), or mock-inoculation (mock) by infiltrating an aqueous, 10 mM MgCl 2 solution.Three rosette leaves per plant were infiltrated, and metabolite contents determined by GC-MS analysis as described in the text.Metabolite levels are given in µg per gram FW.Bars represent means ± SD of three biological replicates.One replicate sample consisted of six leaves from two plants.Individual data points of biological replicates are super-imposed on the bar graphs (small circles).The presence of asterisks above the lines indicates whether significant differences between mock-and bacterial inoculations exist (***P < 0.001, **P < 0.01, *P < 0.05; two-tailed t-test).Absence of asterisks indicates no significant difference (P > 0.05).A, Meta-substituted SA/BA-derivatives: 5-FSA, 5-CSA, 5-CMSA, 3-FSA, 3-BSA, and 3-CMBA.B, SA-Mal, SA-Asp, Pyr-Glu, and IC-Glu.Related to Fig.5.Time course of accumulation of pipecolic acid and NHP glucose ester in P. syringae-inoculated Col-0, sid2, eds5, and pbs3 plants.A-B, Levels of (A) pipecolic acid (Pip) and (B) N-hydroxypipecolic acid glucose ester (NHPGE) at 6, 10, 24, and 48 hours post treatment (hpt) in mock-infiltrated (mock) and Psm-inoculated (Psm) leaves of Arabidopsis Col-0, sid2, eds5, and pbs3 plants.Three rosette leaves per plant were treated.Metabolite contents were determined by GC-MS analysis of trimethylsilylated analytes and are given in µg per gram FW.Bars represent means ± SD of four biological replicates.One replicate sample consisted of six leaves from two plants.Individual data points of biological replicates are super-imposed on the bar graphs (small circles).Different letters denote significant differences (p < 0.05, Kruskal-Wallis H test). nd: not detected.Related to Fig.7.Expression levels of VAS1 in Arabidopsis leaves inoculated with Psm according to publicly available microarray analyses.Means ± SD of Affymetrix expression values of probe 260328_at (At1g80360, VAS1) originating from three biological replicates are given.Individual expression values of biological replicates are super-imposed on the bar graphs (small circles).The presence of asterisks above the lines indicates whether significant differences between mock-and Psm-treatment exist for the particular timepoint (***P < 0.001, **P < 0.01, *P < 0.05; two-tailed t-test).The data originate from NASCARRAYS-414: Gene expression in Arabidopsis leaves inoculated with Pseudomonas syringae pv.maculicola ES4326.Individual leaves of Col-0 plants were infiltrated with Psm (OD 600 = 0.002) or mock-infiltrated and harvested 9, 24 or 32 hpi (Nafisi et al., 2007).

derivatives, SA, and the SA conjugates SA-Mal, SA-Asp, SAG, and SGE by an LC- qTOF-MS-based method. A
, Table listing retention times (RT), sum formulars, [M-H] -ions, and other characteristic fragment ions of the compounds analysed in Arabidopsis leaf extracts by LC-qTOF-MS in negative ionization mode (see experimental procedures for details).
A Supporting Figure 2. Analysis of meta-substituted SA-and BA--ions or characteristic fragment ions were determined and related to the fresh weight (FW) of the leaf material.Each sample is derived from the extraction of Psm-(dark blue bars) and mock-inoculated (light blue bars) leaves at 48 h post treatment.