A Pseudomonas syringae ADP-ribosyltransferase inhibits Arabidopsis mitogen-activated protein kinase kinases

Abstract

The successful recognition of pathogen-associated molecular patterns (PAMPs) as a danger signal is crucial for plants to fend off numerous potential pathogenic microbes. The signal is relayed through mitogen-activated protein kinase (MPK) cascades to activate defenses. Here, we show that the Pseudomonas syringae type III effector HopF2 can interact with Arabidopsis thaliana MAP KINASE KINASE5 (MKK5) and likely other MKKs to inhibit MPKs and PAMP-triggered immunity. Inhibition of PAMP-induced MPK phosphorylation was observed when HopF2 was delivered naturally by the bacterial type III secretion system. In addition, HopF2 Arg-71 and Asp-175 residues that are required for the interaction with MKK5 are also necessary for blocking MAP kinase activation, PAMP-triggered defenses, and virulence function in plants. HopF2 can inactivate MKK5 and ADP-ribosylate the C terminus of MKK5 in vitro. Arg-313 of MKK5 is required for ADP-ribosylation by HopF2 and MKK5 function in the plant cell. Together, these results indicate that MKKs are important targets of HopF2.

Figure 1.

HopF2 Suppresses flg22-Induced Resistance Responses. (A) HopF2 inhibits flg22-induced FRK1 expression. Left: Protoplasts were transfected with ProFRK1:LUC reporter construct along with HopF2-FLAG, HopF2^R71A-FLAG, or HopF2^D175A-FLAG and induced with 100 nM flg22 for 2 h before ProFRK1-LUC activity was measured. Right: Pro35S:LUC reporter control. Each data point represents the mean of three replicates. (B) HopF2 inhibits the flg22-induced oxidative burst. Four-week-old wild-type (WT), HopF2 transgenic line 5, and line 22 T3 plants were sprayed with 50 μM β-estradiol 48 h before leaves were excised. Leaf strips were treated with water or 1 μM flg22, and the production of H₂O₂ was measured at the indicated time points and expressed as relative luminescence units. Each time point represents the data from eight replicates. Error bars indicate sd. Different letters indicate significant difference at a P value <0.01 (Student's t test). The experiments were repeated twice with similar results. (C) HopF2 inhibits flg22-induced callose deposition. Six-week-old wild-type and HopF2 transgenic line 5 T3 plants were sprayed with water (−) or 50 μM β-estradiol (+) 48 h before infiltration with 1 μM flg22. Leaves were stained for callose 6 h after flg22 treatment. The number below each microscopy photograph indicates the average number of callose deposits/0.1 mm² leaf from eight independent leaves.

Figure 2.

HopF2 Function Is Linked to the Inhibition of MAP Kinase Cascades. (A) HopF2 transgenic plants display clustered stomatal cells. The micrographs of the abaxial epidermis were taken 2 weeks after germination on plates containing estradiol. Numbers below the photographs are stomatal density/mm² in HopF2 and wild-type (WT) plants, which are significantly different at a P value <0.01 (Student's t test). The results shown are representative of four independent experiments. (B) HopF2 inhibits flg22-induced activation of MPK4 and MPK6. Protoplasts were transfected with the MPK4-FLAG or MPK6-FLAG construct in the presence (+) or absence (−) of the HopF2-FLAG construct, treated with water (−) or 1 μM flg22 (+) for 10 min, and the total protein extract was subjected to anti-FLAG immunoprecipitation. The purified MPK4-FLAG and MPK6-FLAG protein was then subjected to an in-gel kinase assay using myelin basic protein as a substrate and anti-FLAG immunoblot analysis. (C) Bacterial PAMP-induced MAPK activation is suppressed by TTSS-delivered HopF2. Leaves of 6-week-old Col-0 plants were infiltrated with the indicated bacteria at 1 × 10⁷ cells/mL. Leaf total protein was extracted at the indicated time points and subjected to immunoblot analysis with antiphospho-ERK antibodies. H, water control; V, P. fluorescens strain (pLN1965) bacteria lacking effector genes but containing an empty vector; F, P. fluorescens strain (pLN1965) carrying the ShcF-HopF2 construct. Equal loading was indicated by immunoblot with anti-HSP90 antibodies.

Figure 3.

HopF2 Blocks MKK5 Function. (A) HopF2 inactivates MKK5 in protoplasts. Protoplasts were transfected with MKK5^DD-FLAG alone or in combination with HopF2-HA, and MKK5^DD-FLAG protein was isolated by immunoprecipitation. MPK6-FLAG was expressed alone in protoplasts and isolated by immunoprecipitation. The isolated MPK6 and MKK5^DD were incubated in a kinase reaction system before an in-gel kinase assay was performed. The amount of MPK6 and MKK5^DD in the in vitro kinase reaction assay was indicated by immunoblot (IB). An aliquot of total protein extract was subjected to immunoblot analysis with anti-HA antibody to determine the expression of HopF2-HA. (B) HopF2 inhibits the callose deposition triggered by the constitutively active form of MKK5. Six-week-old Col-0, MKK5^DD, and MKK5^DD/HopF2 transgenic plants were induced with both 30 μM dexamethasone and 50 μM β-estradiol for 24 h before callose deposition was examined. The number below each microscopy photograph indicates the average number of callose deposits/0.1 mm² leaf from eight independent leaves. Error bars indicate sd. Different letters indicate significant difference at a P value <0.01 (Student's t test). The results shown are representative of two independent experiments.

Figure 4.

HopF2 Targets MKKs. (A) HopF2 interacts with MKK5 but not MPK6 in E. coli. GST or GST-HopF2 was coexpressed with His-MKK5 or MPK6-His in E. coli, purified with glutathione agarose, and subjected to immunoblot analysis using anti-His antibody to detect His-MKK5 or MPK6-His. Coomassie blue (CBB) staining indicates equal amounts of bait proteins. (B) HopF2 Arg-71 and Asp-175 are required for the interaction with MKK5 in E. coli. His-MKK5 was coexpressed with GST, GST-HopF2, GST-HopF2^R71A, or GST-HopF2^D175A in E. coli, and protein was subjected to a GST pull-down assay. (C) HopF2 interacts with MKK5 in plants. Transgenic plants carrying the HopF2-FLAG and MKK5-HA transgenes were analyzed by a co-IP assay. (D) HopF2 Arg-71 and Asp-175 are required for the interaction with MKK5 in protoplasts. Protoplasts were transfected with MKK5-HA alone or in combination with HopF2-FLAG, HopF2^R71A-FLAG, or HopF2^D175A-FLAG, and protein–protein interaction was analyzed by co-IP.

Figure 5.

HopF2 ADP-Ribosylates MKK5 in Vitro. (A) HopF2 inactivates MKK5 kinase activity in vitro. FLAG-MKK5^DD was coexpressed with GST or GST-tagged wild-type HopF2, HopF2^R71A, or HopF2^D175A in E. coli and purified with anti-FLAG M2-agarose. The activity of FLAG-MKK5^DD was measured by its ability to phosphorylate MPK6-His in vitro, as indicated by autoradiography. Amounts of protein were detected by immunoblot analysis with anti-FLAG and anti-MPK6 antibodies. (B) HopF2 ADP-ribosylates MKK5 in vitro. Purified HopF2, HopF2^R71A, or HopF2^D175A was incubated with purified FLAG-MKK5^DD or GST-MKK5^DD at a substrate-to-enzyme ratio of 10:1 in the presence of biotinylated NAD. ADP-ribosylation was detected by immunoblot using horseradish peroxidase–conjugated streptavidin. Amounts of MKK5^DD protein are indicated by CBB staining. (C) MKK5^DD is a better substrate for HopF2 than is WT MKK5. FLAG-tagged MKK5^DD or WT MKK5 was incubated with HopF2 or HopF2^D175A in the presence of biotinylated NAD, and ADP-ribosylation was detected as in (B). Amounts of MKK5 protein are indicated by Coomassie blue staining. (D) HopF2 ADP-ribosylates MKK5 in a NAD concentration–dependent manner. Purified HopF2 or HopF2^D175A was incubated with FLAG-MKK5^DD in the ADP-RT reaction system containing the indicated concentrations of biotinylated NAD.

Figure 6.

Arg-313 of MKK5 Is Required for ADP-Ribosylation and Function. (A) and (B) The C terminus of MKK5^DD is required for ADP-ribosylation by HopF2. FLAG-tagged MKK5^DD full-length (1-348Aa) or deletion constructs (1-270Aa and 1-310Aa) were incubated with wild-type HopF2 or HopF2^D175A in an ADP-RT reaction. Aa, amino acids. ADP-ribosylated MKK5 and total MKK5 protein were detected as in Figure 5. (C) The C-terminal 38 amino acids of MKK5 are sufficient for ADP-ribosylation. The GST-MKK5(311-348Aa) fusion protein was incubated with the wild-type HopF2 or HopF2^D175A in an ADP-RT reaction. (D) Arg-313 of MKK5^DD is required for ADP-ribosylation by HopF2. The MKK5^DD mutant protein containing the R313A substitution was incubated with wild-type HopF2 or HopF2^D175A in an ADP-RT reaction. (E) The C-terminal 38 amino acids of MKK5^DD are required for the activation of FRK1 expression. FRK1-LUC and the indicated constructs were cotransfected into Arabidopsis protoplasts, and LUC activity was measured 18 h later. An empty vector (Vec) was included as a negative control. Each data point represents the average of four replicates. Error bars indicate sd.

Figure 7.

HopF2 Arg-71 and Asp-175 Are Required for Virulence and ETI Functions. (A) HopF2 Arg-71 and Asp-175 are required for its ability to induce nonhost HR in tobacco W38 plants. Tobacco leaves were infiltrated with the P. syringae pv tabaci strain 1152 containing the indicated constructs at 2 × 10⁸ colony-forming units/mL. Photographs were taken at 14 h after inoculation. The numbers on the side indicate areas showing HR/total injected areas. Infiltrated leaf areas are indicated by circles. The data shown are representative of two independent experiments. (B) HopF2 Arg-71 and Asp-175 are required for virulence function in tomato plants. Tomato plants were infiltrated with the mutant strain lacking hopF2 (hopF2⁻), the hopF2⁻ strain complemented with the wild-type schF-hopF2 locus, or the hopF2⁻ strain complemented with the two mutant forms of schF-hopF2 at 5 × 10⁴ colony-forming units/mL. Leaf bacterial numbers were determined at the indicated time points (four replicates/data point). Error bars indicate sd. ** Significant difference compared with the hopF2⁻ strain carrying wild-type hopF2 at a P value <0.01 (Student's t test). The experiment was repeated three times with similar results.