Quantitative phosphoproteomic analysis reveals common regulatory mechanisms between effector- and PAMP-triggered immunity in plants

Abstract

Plant immunity consists of two arms: pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI), induced by surface-localized receptors, and effector-triggered immunity (ETI), induced by intracellular receptors. Despite the little structural similarity, both receptor types activate similar responses with different dynamics. To better understand phosphorylation events during ETI, we employed a phosphoproteomic screen using an inducible expression system of the bacterial effector avrRpt2 in Arabidopsis thaliana, and identified 109 differentially phosphorylated residues of membrane-associated proteins on activation of the intracellular RPS2 receptor. Interestingly, several RPS2-regulated phosphosites overlap with sites that are regulated during PTI, suggesting that these phosphosites may be convergent points of both signaling arms. Moreover, some of these sites are residues of important defense components, including the NADPH oxidase RBOHD, ABC-transporter PEN3, calcium-ATPase ACA8, noncanonical Gα protein XLG2 and H⁺ -ATPases. In particular, we found that S343 and S347 of RBOHD are common phosphorylation targets during PTI and ETI. Our mutational analyses showed that these sites are required for the production of reactive oxygen species during both PTI and ETI, and immunity against avirulent bacteria and a virulent necrotrophic fungus. We provide, for the first time, large-scale phosphoproteomic data of ETI, thereby suggesting crucial roles of common phosphosites in plant immunity.

Keywords: Arabidopsis; bacteria; effectors; fungi; pathogen-associated molecular patterns (PAMPs); plant immunity; protein phosphorylation; reactive oxygen species (ROS).

Figure 1.

Schematic representation of the large-scale phosphoscreen. Individual steps are highlighted and annotated. For further details see the experimental procedures in the Materials and Methods section.

Figure 2.

Heat-maps showing RESISTANT TO P. SYRINGAE-2 (RPS2)-regulated phosphorylation sites. The phosphorylation sites shown are differentially phosphorylated upon RPS2 activation (one-way ANOVA p≤0.05 + t-test p≤0.05). (a) Phosphorylation sites significantly upregulated by RPS2 in Dex:avrRpt2 plants. (b) Phosphorylation sites significantly downregulated by RPS2 in Dex:avrRpt2 plants. Dendrograms were obtained by hierarchical clustering to represent Euclidian distances of normalized expression profiles. The colored sidebar indicates the regulation at the protein level in total protein membrane fractions (Supporting Information Table S1): black, sites for which the protein levels were not significantly altered amongst treatments; magenta, sites for which protein levels were significantly increased in Dex:avrRpt2 plants upon application of dexamethasone (Dex); green, sites for which protein levels were significantly decreased in rpm1rps2/Dex:avrRpt2 plants upon Dex application; orange, sites for which no peptides were detected in total protein samples; black stars, the phosphorylation sites upregulated in both lines after Dex treatment.

Figure 2.

Heat-maps showing RESISTANT TO P. SYRINGAE-2 (RPS2)-regulated phosphorylation sites. The phosphorylation sites shown are differentially phosphorylated upon RPS2 activation (one-way ANOVA p≤0.05 + t-test p≤0.05). (a) Phosphorylation sites significantly upregulated by RPS2 in Dex:avrRpt2 plants. (b) Phosphorylation sites significantly downregulated by RPS2 in Dex:avrRpt2 plants. Dendrograms were obtained by hierarchical clustering to represent Euclidian distances of normalized expression profiles. The colored sidebar indicates the regulation at the protein level in total protein membrane fractions (Supporting Information Table S1): black, sites for which the protein levels were not significantly altered amongst treatments; magenta, sites for which protein levels were significantly increased in Dex:avrRpt2 plants upon application of dexamethasone (Dex); green, sites for which protein levels were significantly decreased in rpm1rps2/Dex:avrRpt2 plants upon Dex application; orange, sites for which no peptides were detected in total protein samples; black stars, the phosphorylation sites upregulated in both lines after Dex treatment.

Figure 3.

Avirulent bacteria induce phosphorylation of RESPIRATORY BURST OXIDASE HOMOLOGUE D (RBOHD) at specific residues. Selected reaction monitoring (SRM) analysis of the phosphorylation sites 6 h after infiltration with 10 mM MgCl₂ solution, Pseudomonas syringae pv tomato (Pto) DC3000 EV (empty vector), Pto DC3000 (avrRpm1) or Pto DC3000 (avrRpt2) using a triple quadruple mass spectrometer. Values are means ±SE of three biological replicates. Different letters indicate significantly different values at p≤0.05 for S163 and S347, or at p≤0.01 for S343 (one-way ANOVA, Tukey post hoc test).

Figure 4.

RESPIRATORY BURST OXIDASE HOMOLOGUE D (RBOHD) phosphorylation sites S343 and S347 are required for reactive oxygen species (ROS) production during effector-triggered immunity (ETI). (a) Immunoblot showing similar protein levels of FLAG-tagged RBOHD protein in rbohD mutants expressing 3xFLAG-RBOHD WT and the S343A/S347A variant. Coomassie stain (CBB) shows rubisco protein to demonstrate equal loading. (b) 3,3’-diaminobenzidine (DAB)-mediated H₂O₂ staining in rbohD and rbohD mutants expressing 3xFLAG-RBOHD WT and the S343A/S347A variant 8 h after infiltration with the bacteria. The right half of the leaves were infiltrated with 10 mM MgCl₂ solution, Pseudomonas syringae pv tomato (Pto) DC3000 EV, Pto DC3000 (avrRpm1) or Pto DC3000 (avrRpt2). We repeated three times with similar results (4–5 leaves per genotype were stained each time).

Figure 5.

The bik1pbl1 double mutant does not exhibit a defect in reactive oxygen species (ROS) accumulation during effector-triggered immunity (ETI). (a) BOTRYTIS-INDUCED KINASE-1 (BIK1) protein accumulates after infection with Pseudomonas syringae pv tomato (Pto) DC3000 (avrRpm1). pBIK1:BIK1-HA plants were inoculated with 5 mM MgC_l2, Pto DC3000 EV or Pto DC3000 (avrRpm1) (2.5 × 10⁷ cfu (colony-forming units) ml⁻¹) and BIK1-HA protein amount was determined by immunoblot analyses using anti HA antibody. (b) H₂O₂ accumulation in Col-0 and bik1pbl1 mutant after inoculation with 10 mM MgC_l2, Pto DC3000 EV or Pto DC3000 (avrRpm1) (2.5 × 10⁷ cfu ml⁻¹). H₂O₂ accumulation was detected by 3,3’-diaminobenzidine (DAB). The experiments were performed three times with similar results.

Figure 6.

RBOHD-S343A/S347A variant can complement overactivation of immune-related gene expression in rbohD and the semi-dwarf autoimmune phenotype of rbohDrbohF. (a) Loss of RESPIRATORY BURST OXIDASE HOMOLOGUE D (RBOHD) leads to increased expression of FERULOYL COA ORTHO-HYDROXYLASE 1 (F6’H1) upon bacterial perception. Gene expression of F6’H1 in the leaves infiltrated with Pseudomonas syringae pv tomato DC3000 (avrRpt2) (2.5 × 10⁷ cfu (colony-forming units) ml⁻¹) or 10 mM MgCl₂ solution for 24 h was measured by qPCR analysis. The relative transcript levels were calculated by normalization to the U-box housekeeping gene transcript (At5g15400). Data are mean ±SE of three technical replicates. Different letters indicate significantly different values at p ≤ 0.05 (one-way ANOVA, Tukey post hoc test). The experiments were performed three times with similar results. (b) Immunoblot showing equal FLAG-RBOHD protein level in leaves of rbohDrbohF/pRBOHD:3xFLAG-RBOHD (WT) and rbohDrbohF/pRBOHD:3xFLAG-RBOHD (S343A/S347A) lines. Leaf tissue of the rbohDrbohF double mutant was used as negative control. Coomassie stain (CBB) shows rubisco protein to demonstrate equal loading. (c) Growth phenotypes of six-week-old plants. Bars, 1 cm.

Figure 7.

RESPIRATORY BURST OXIDASE HOMOLOGUE D (RBOHD) phosphorylation sites S343 and S347 are required for effector-triggered immunity (ETI). Bacterial growth of Pseudomonas syringae pv tomato (Pto) DC3000 (avrRpm1) (a) and Pto DC3000 (avrRpt2) (b) in rbohD and rbohD expressing 3xFLAG-RBOHD WT or the S343A/S347A variant. Bacteria were syringe infiltrated in three leaves per one plant at a concentration of 1 × 10⁵ cfu (colony-forming units) ml⁻¹. Three days post infiltration, leaves were harvested to determine bacterial growth. Data are means ±SD of 8 replicates. Solid horizontal lines within the boxes show median. Different letters indicate significantly different values at p ≤ 0.01 (one-way ANOVA, Tukey post hoc test). These experiments were repeated three times with similar results.

Figure 8.

The double mutant rbohDrbohF shows weakened resistance to virulent Plectosphaerella cucumerina (PcBMM isolate). (a) Symptoms of 18-d-old plants after spray inoculation with PcBMM (4 × 10⁶ spores ml⁻¹). (b) Quantification of PcBMM biomass. Fungal DNA was quantified by qPCR at 6 d post-inoculation (dpi) using specific primers for PcBMM β-TUBULIN and normalized to Arabidopsis thaliana UBIQUITIN 10 gene. Bars represent averages (±SE) of fungal DNA levels relative to Col-0 plants from two replicates. Statistical analysis was performed by ANOVA, corrected with Bonferroni post hoc test. Different letters indicate significant differences (p ≤ 0.05). The experiment was repeated twice with similar results.