Targeting Magnaporthe oryzae effector MoErs1 and host papain-like protease OsRD21 interaction to combat rice blast

Abstract

Effector proteins secreted by plant pathogenic fungi are important artilleries against host immunity, but there is no precedent of such effectors being explored as antifungal targets. Here we demonstrate that MoErs1, a species-specific effector protein secreted by the rice blast fungus Magnaporthe oryzae, inhibits the function of rice papain-like cysteine protease OsRD21 involved in rice immunity. Disrupting MoErs1-OsRD21 interaction effectively controls rice blast. In addition, we show that FY21001, a structure-function-based designer compound, specifically binds to and inhibits MoErs1 function. FY21001 significantly and effectively controls rice blast in field tests. Our study revealed a novel concept of targeting pathogen-specific effector proteins to prevent and manage crop diseases.

Fig. 1 |. MoErs1 is a cytoplastic effector required for the virulence of M. oryzae.

a, MoErs1 is required for the virulence of M. oryzae. Conidial suspensions (5 × 10⁴ conidia per ml in 0.2% gelatin) of Guy11, ΔMoers1 mutants (#1, #7 and #18) and the complement strains were sprayed onto 2-week-old rice seedlings (CO39). Diseased rice leaves were photographed after 7 days post inoculation (dpi). b, Fungal growth measured by quantifying M. oryzae genomic 28S rDNA relative to rice genomic Rubq1 DNA. Mean ± s.d. of 3 determinations. Significant differences were determined using two-sided Duncan’s new multiple-range tests and marked with different letters. c, The disease lesion area (DLA) was assessed using Image J software. The biologically independent DLA is displayed as boxes with individual datapoints (n = 9). The error bars represent maximum and minimum values. Centre line, median; box limits, 25th and 75th percentiles. Significant differences were determined using two-sided Duncan’s new multiple-range tests and marked with different letters. d, The ΔMoers1 mutant cannot produce typical lesions on rice leaves. Conidiation lesions on surface-sterilized rice leaves in a were counted and photographed. The lesions producing conidia (left) are typical lesions and the lesions that fail to produce conidia (right) are necrotic lesions. e, Statistical analysis of typical and necrotic lesions on rice leaves in a. f,g, MoErs1 is a cytoplasmic effector regulated by MoSyn8 (f). The fungal transformants ΔMoers1 and ΔMosyn8 expressing MoErs1:RFP or MoErs1:RFP:NLS/Slp1:GFP at 30 hpi in the sheath cells of rice cultivar CO39 treated with or without Brefeldin A (BFA) are shown as a projection under a confocal microscope, the left column is RFP, the middle is DIC, and the right is Merge (g). The experiments were repeated independently at least 3 times with similar results. Arrows indicate the biotrophic interface complex and the white asterisks indicate rice nuclei. Scale bar, 10 μm.

Fig. 2 |. MoErs1 interacts with OsRD21 in the plasma membrane in vivo.

a, OsRD21 is mainly concentrated in intracellular components but not in the apoplast. GFP-FLAG, Slp1-FLAG and proOsRD21-FLAG were transiently expressed in N. benthamiana. The proOsRD21 and MoERS1 gene-overexpressing rice lines (OsRD21-OX and MoERS1^ΔSP-OX) with a C-terminal FLAG tag were used to determine the distribution of OsRD21 in rice cells; the OsAO4-OX overexpressing rice line was used as a positive control. The apoplastic and intracellular leaf extracts were separated and stained with CBB. Immunoblots with appropriate anti-FLAG antibodies showed apoplast Slp1 and OsAO4 levels, but not GFP or OsRD21 levels. b,c, BiFC assay in N. benthamiana (b). Co-expression of MoErs1^ΔSP-nYFP and proOsRD211-cYFP with a PM marker Remorin-RFP, treated with (plasmolysis) or without (natural) 1 M NaCl (c) showed that MoErs1 and OsRD21 co-localized in the PM. The relevant negative controls in b showed no fluorescence. Scale bar, 5 μm. d,e, BiFC assays in rice protoplast cells (d). Co-expression of MoErs1^ΔSP-nYFP and proOsRD211-cYFP with Remorin-RFP (e) showed that MoErs1 and OsRD21 co-localized in the PM in rice protoplasts. The relevant negative controls in d showed no fluorescence. Scale bar, 10 μm. All experiments were repeated independently at least 3 times with similar results.

Fig. 3 |. MoErs1 functions as a PLCP inhibitor to inhibit the activity of OsRD21.

a, A structural model for the MoErs1–OsRD21 interaction predicted by ClusPro. Top: the surface of the MoErs1–OsRD21 complex. Middle: the interaction interface with amino acid residues shown as ribbon diagrams. Bottom: table showing three loops (L2, L4 and L8) and one β-strand (β11) with interaction sites of MoErs1. The sticks in different colours indicate corresponding interacting amino acids between MoErs1 and OsRD21. The yellow dotted line indicates hydrogen bonding and the orange dotted line indicates non-covalent binding. b, MoErs1 inhibits the PLCP activity of OsRD21. Top: model of the structure and processing of OsRD21. OsRD21 maturation comprising signal peptide release resulting in ProRD21, prodomain cleavage and final granulin domain removal to produce mature RD21 (mRD21). Bottom: GFP-tagged MoErs1 or the interaction site mutations in a were transiently expressed in N. benthamiana together with a FLAG fusion of OsRD21. Total proteins were extracted and purified, and labelled with 2 μM DCG-04 for 4 h. Proteins were separated by SDS–PAGE, detected with streptavidin-HRP, and chemiluminescence and immunoblotting with anti-FLAG and anti-GFP antibodies. The experiments were repeated independently at least 3 times with similar results. c–e, The interaction sites of MoErs1 contribute to virulence. Conidial suspensions (5 × 10⁴ conidia per ml in 0.2% gelatin) of Guy11, the ΔMoers1 mutant, the point mutation mutants and the complement strains were sprayed onto 2-week-old CO39. Diseased rice leaves were photographed after 7 dpi (c, top). The total protein of all the strains was extracted and detected with anti-GFP and anti-actin antibodies (c, bottom). DLA was assessed using Image J (d) and fungal growth was evaluated by quantifying M. oryzae genomic 28S rDNA relative to rice genomic Rubq1 DNA (e). Mean ± s.d. of 3 determinations. Significant differences were determined using two-sided Duncan’s new multiple-range tests and marked with different letters.

Fig. 4 |. The diphenyl ether ester compound inhibits the inhibitory activity of MoErs1.

a, The interaction region of MoErs1 and OsRD21. b, The molecular docking model of diaryl ether–MoErs1. c, The interaction model of MoErs1 (green)–OsRD21(blue), with the red area representing the compound FY21001. The key residues surrounding the active site are shown as green sticks. The yellow dotted line with the marked distance indicates hydrogen binding between FY21001 and amino acid residues Phe72 in MoErs1. The orange dotted line indicates non-covalent binding. d, FY21001 exhibits a stronger MoErs1 binding affinity than OsRD21, as assessed using MST. GST-MoErs1 (10 μM) was labeled with RED-NHS. The raw data were integrated and fitted to a binding model using the MST analysis software. The recombinant proteins were contained in NT standard capillaries. The solid curve is the fit of the datapoints to the standard K_d-fit function. Each binding assay was repeated independently three times (n = 3) and error bars represent s.d. e, BiFC assay showed that FY21001 inhibits the interaction between MoErs1 and OsRD21 in vivo. Co-expression of MoErs1-nYFP and OsRD211-cYFP, with PM marker Remorin-RFP, treated with DMSO or 500 μM FY21001 in rice protoplast cells, showed that fluorescence was detected when treated with DMSO, but not with FY21001. Scale bar, 10 μm. f, Co-IP assay showed that FY21001 inhibits the interaction between MoErs1 and OsRD21 in a dose-dependent manner in vivo. Co-expression of MoErs1-GFP and OsRD21-FLAG in rice protoplast cells treated with FY21001 in different concentrations. Immunoprecipitates obtained with the anti-FLAG antiserum and total protein extracts were immunoblotted with anti-FLAG and anti-GFP antibodies. g, GST-tagged MoErs1 was expressed and purified from E. coli strain BL21. FLAG-tagged OsRD21 was transiently expressed in N. benthamiana. Total proteins were extracted, purified and labelled with 2 μM DCG-04 for 4 h in the presence of compound FY21001 in a dose-dependent manner. Proteins were separated using SDS–PAGE and detected with streptavidin-HRP, and chemiluminescence and immunoblotting with anti-FLAG and anti-GST antibodies. All experiments were repeated independently at least 3 times with similar results.

Fig. 5 |. Diphenyl ether ester compounds are effective against rice blast.

a,b, FY21001 is effective against rice leaf blast. Rice leaves were sprayed with 500 μM of compounds with either 24 h co-inoculation, pre-inoculation or post-inoculation with Guy11 spores. Diseased rice leaves were photographed after 7 dpi (a). The biologically independent DLA (b, top) is displayed as boxes with individual datapoints (n = 21). The error bars represent maximum and minimum values. Centre line, median; box limits, 25th and 75th percentiles. Fungal growth (b, bottom) was measured by quantifying M. oryzae genomic 28S rDNA relative to rice genomic Rubq1 DNA. Mean ± s.d. of 3 determinations. Significant differences were determined by two-sided Duncan’s new multiple-range tests and marked with different letters. c,d, Conidiation lesions on surface-sterilized rice leaves in a were counted and photographed (c). The lesions producing conidia (typical lesions) and those that fail to produce conidia (necrotic lesions) are quantified (d). Error bars represent s.d. e,f, FY21001 is effective against rice neck blast in the field. e, Neck blast severity was evaluated using the standard 0–9 scale, rated on six levels defined as follows: level 0: no visible lesion or observed lesions on only a few pedicels; level 1: lesions on several pedicels or secondary branches; level 3: lesions on a few primary branches or the middle part of the panicle axis; level 5: lesion partially around the base (node) or the uppermost internode or the lower part of panicle axis near the base; level 7: lesion completely around panicle base or uppermost internode or panicle axis near the base with more than 30% of filled grains; level 9: lesion completely around panicle base or uppermost internode or the panicle axis near the base with less than 30% of filled grains (International Rice Research Institute Standard Evaluation System for Rice). f, Field resistance to neck blast was assessed in a natural rice blast nursery (Jiangsu Province, China). Three plots were established in the field. One thousand susceptible rice plants were planted in each plot. One group treated with 500 μM DMSO was used as the negative control, one group treated with 500 μM tricyclazole (TCZ) was used as the positive control and the trial group was treated with 500 μM FY21001. g, Statistics of disease index. Disease index = 100 × Σ (number of diseased leaves at all levels × representative value at all levels)/(total leaves investigated × highest representative value) (n = 3 biologically independent samples). h, Grain yield of Nip treated with FY21001, tricyclazole or DMSO under natural blast nursery conditions. Significant differences were determined using Duncan’s new multiple-range tests and marked with different letters (n = 3 biologically independent samples). i, Wild-type TP309 and OsRD21-KO transgenic rice leaves sprayed with or without FY21001 (500 μM) for 24 h were inoculated with Guy11 and the ΔMoers1 mutant (1 × 10⁵ spores per ml). j, The biologically independent DLA is displayed as boxes with individual datapoints (n = 10). The error bars represent maximum and minimum values. Centre line, median; box limits, 25th and 75th percentiles. k, Fungal growth measured by quantifying M. oryzae genomic 28S rDNA relative to rice genomic Rubq1 DNA. Mean ± s.d. of 3 determinations. Significant differences were determined using two-sided Duncan’s new multiple-range tests and marked with different letters.

Fig. 6 |. A proposed model of MoErs1 function to suppress host immunity.

There are 3 states of OsRD21 during the M. oryzae–rice interaction when treated with compound FY21001. (1) Rice cells perceive the infection via the functions of transmembrane receptors and, at the same time, PLCP OsRD21-mediated substrate degradation allows the activation of the host immune response. (2) M. oryzae secretes the effector protein MoErs1 during its interaction with the rice host. MoErs1 targets OsRD21 and suppresses its PLCP activities. (3) FY21001, a diphenyl ether ester, specifically binds MoErs1 to inhibit its function, which relieves the inhibition of OsRD21 protease activities to promote host immunity against M. oryzae infection.