Effector-mediated relocalization of a maize lipoxygenase protein triggers susceptibility to Ustilago maydis

Abstract

As the gall-inducing smut fungus Ustilago maydis colonizes maize (Zea mays) plants, it secretes a complex effector blend that suppresses host defense responses, including production of reactive oxygen species (ROS) and redirects host metabolism to facilitate colonization. We show that the U. maydis effector ROS burst interfering protein 1 (Rip1), which is involved in pathogen-associated molecular pattern (PAMP)-triggered suppression of host immunity, is functionally conserved in several other monocot-infecting smut fungi. We also have identified a conserved C-terminal motif essential for Rip1-mediated PAMP-triggered suppression of the ROS burst. The maize susceptibility factor lipoxygenase 3 (Zmlox3) bound by Rip1 was relocalized to the nucleus, leading to partial suppression of the ROS burst. Relocalization was independent of its enzymatic activity, revealing a distinct function for ZmLox3. Most importantly, whereas Zmlox3 maize mutant plants showed increased resistance to U. maydis wild-type strains, rip1 deletion strains infecting the Zmlox3 mutant overcame this effect. This could indicate that Rip1-triggered host resistance depends on ZmLox3 to be suppressed and that lox3 mutation-based resistance of maize to U. maydis requires functional Rip1. Together, our results reveal that Rip1 acts in several cellular compartments to suppress immunity and that targeting of ZmLox3 by Rip1 is responsible for the suppression of Rip1-dependent reduced susceptibility of maize to U. maydis.

Figure 1

Rip1 is a secreted fungal protein that suppresses the PAMP-triggered ROS burst in several subcellular compartments of a plant cell. A, Secretion of Rip1 in maize plants infected with U. maydis. Images show areas of cells from maize cv. B73 infected with U. maydis strains expressing mCherry fusions of Rip1 either with (Cmu1_pro:Rip1_1–166) or without the signal peptide (Cmu1_pro:Rip1_27–166), both under the control of the Cmu1 promoter. mCherry fluorescence was observed 5 dpi. Whereas secreted Rip1 (containing the signal peptide) strongly accumulates at the periphery of hyphal cells and hyphal tip (white arrows in upper), Rip1 lacking the signal peptide is evenly distributed within the hyphae, forming aggregates throughout the cells (lower). Merge = combined bright-field and fluorescence images. Bars = 5 μm. B, Secretion of Rip1 in axenic culture. Rip1_1–166-3xHA was expressed in U. maydis strain AB33 under the control of the constitutive otef promoter (otef_pro). Total cellular proteins were extracted from the cell pellets and secreted proteins were precipitated from the culture supernatants. These samples were analyzed by immunoblot using anti-HA or anti-actin primary antibodies. Rip1_1–166-3xHA was detected in both pellet and supernatant, whereas the unsecreted control actin was detected only in the pellet fraction. C, PAMP-triggered ROS burst assays in Z. mays. ROS accumulation in Z. mays transiently overexpressing p19-p2A-Rip1_27–166-p2A-mCherry was monitored over 40 min after challenging with flg22. Plants expressing p19-p2A-mCherry-p2A-GFP were used as controls. Box plots indicate total accumulation of plant derived apoplastic ROS after PAMP treatment based on detection of chemiluminescent emissions as described in “Materials and methods”. Data show values from a pool of three biological replicates (seven plants per pool) calculated for the total area under the curve of the luminol-based PAMP-triggered ROS-burst assay. Asterisks indicate statistically significant differences from the control (**P < 0.01, Students t test, Supplemental Table S4). D, Localization of Rip1 (without the signal peptide) in Z. mays. Transient expression of 35S_pro:Rip1_27–166-myc-mCherry via biolistic bombardment in maize plant leaves was observed via confocal microscopy. Rip1 showed nucleocytoplasmic localization. Arrows mark plant cell nuclei. Merge = combined bright-field and fluorescence images. Bar = 20 μm. E, Subcellular localization assays of Rip1 reveal its ROS burst suppression activity to be independent of the subcellular localization. In N. benthamiana, transiently expressed Rip1_27–166 fused to different subcellular localization signals (MYR, NLS, and NES) shows ROS burst suppression activity in all subcellular compartments tested (x = either NLS or NES, MYR fused to the N-terminus of Rip1_27–166). Data are from a pool of three biological replicates, n = 15 plants per pool. Asterisks indicate statistically significant differences from the control (***P < 0.001, one-way ANOVA followed by Tukey’s test, Supplemental Table S4).

Figure 2

Rip1 orthologs suppress ROS burst in N. benthamiana plants. A, Localization of different Rip1 orthologs in N. benthamiana. Orthologs are from S. scitamineum (SPSC_05323) (host plant sugarcane), U. hordei (UHOR_13428) (host plant barley), U. bromivora (UBRO_13428) (host plant Brachypodium spp.), and M. pennsylvanicum (BN887_05943) (host plant Polygonum pennsylvanicum). Orthologs were fused to myc-mCherry, coexpressed with 35S_pro:GFP-NLS and localization was assessed by confocal microscopy. Localization was observed in the nucleus and the cytoplasm for all orthologs. Left: mCherry, middle: GFP, right: merged mCherry and GFP. Arrows mark plant cell nuclei. Bars = 20 μm. B, ROS activity of N. benthamiana plants expressing different Rip1 orthologs driven by the cauliflower mosaic virus constitutive 35S promoter (35S_pro). Box plots represent the area under the ROS burst curve after challenging with flg22 monitored over 40 min based on the luminescence assay. All orthologs of Rip1 were able to significantly suppress the ROS burst, except for plants expressing MpRip1. Data are from a pool of three biological replicates (nine plants per pool) calculated as the area under the curve of the luminol-based PAMP-triggered ROS-burst assay. Asterisks indicate statistically significant differences to the control (NS: not significant, *P < 0.05, ***P < 0.05, one-way ANOVA followed by Tukey’s test, Supplemental Table S4). C, Part of the C-terminal protein sequence CLUSTAL alignment of Rip1 orthologs. Differently shaded residues indicate global percentage identities of three or more identical residues (more = darker color) among the proteins. Highly conserved C-terminal region present for all, but MpRip1 is highlighted with a black box. Numbers at the top indicate residue positions based on signal peptide deleted sequences. D, Schematic representations of different deletion and insertion mutation constructs of Rip1 and MpRip1. The RIFL motif (KHLPDLSRP, residues 144–152) is important for ROS burst suppression by Rip1 in planta. Expression of UmRip1 lacking its endogenous C-terminal RIFL motif (UmRip1ΔRIFL) shows similar ROS burst activity levels as do mCherry control plants. Plants expressing MpRip1 with a C-terminal RIFL motif replacement (MpRip1-KHLPDLSRP)) show ROS burst suppressive activity comparable to the control plants. Curves represent plant ROS burst levels after challenging with flg22, monitored over 40 min using the luminescence assay. Data are from a pool of three biological replicates of nine plants per pool. Error bars represent se.

Figure 3

Rip1 binds directly to maize LOX3, a negative regulator of the PAMP-triggered ROS burst. A, Direct interaction of Rip1 and maize LOX3 (Zmlox3) in an Y2H assay. DNA encoding Zmlox3 was cloned into pGBKT7 bait vectors and transformed into the yeast strain Ah109. DNA encoding Rip1 was cloned into the pGADT7 activation vector and transformed into yeast strain Y187. Following mating, diploid yeast cells containing both plasmids were spotted onto selective synthetic dropout (SD) media minus the amino acid noted, and yeast growth was monitored 4 days later. Columns in each panel represent serial 10-fold dilutions. The positive control in the bottom row represents two strongly interacting proteins from our laboratory (another U. maydis effector and a maize plant protein, unpublished). The experiment was repeated 3 times independently with comparable results. Ponceau S staining shows equal protein loading. B, Co-IP assay showing that Rip1 interacts with ZmLox3 in N. benthamiana. Using the constitutive 35S_pro promoter for all constructs, we coexpressed Rip1_27–166-mCherry-myc or mCherry-mCherry-myc (controls) along with Zmlox3-GFP in N. benthamiana leaves. We then performed Co-IP assays using an anti-myc antibody to pull down Rip1_27–166-interacting proteins. Immunoblot analysis using an anti-Zmlox3 antibody showed that Zmlox3 copurified with Rip1_27–-166-mCherry-myc but not with mCherry-mCherry-myc. Ponceau S staining shows the amount of protein loaded per lane. C, Knockout lines of maize lox3 (lox3-4 mutant, Gao et al., 2007) show elevated ROS levels after flg22 treatment. Mutant lines of lox3 were grown for 14 days before leaves were used in flg22-triggered ROS burst assays. ROS accumulation was monitored over 40 min using the luminescence-based assay described in “Materials and methods”. As controls, maize plants of the B73 accession were used. Data are shown as box plots for the area under the curve of the luminol-based PAMP-triggered ROS-burst assay of two pooled biological replicates, n = 6 plants/replicate. Significant statistical differences compared to control plants: Student t test, *P < 0.05, Supplemental Table S4. D, ROS burst inhibition effects are synergistically enhanced in the presence of both, expressed Zmlox3 and Rip1 in N. benthamiana. Plants expressing 35S_pro:Zmlox3-GFP, 35S_pro:Rip1_27–166-mCherry, or both constructs together were infiltrated into N. benthamiana leaves and flg22-triggered ROS bursts were monitored over 40 min using the luminescence-based assay. The ROS burst suppressing activities of single infiltrated Zmlox3 and Rip1 overexpression constructs were further significantly enhanced by simultaneous infiltration of both constructs. The OD₆₀₀ for Zmlox3 was adjusted to 0.2 and for Rip1 to 0.05, respectively. Data are shown as box plots for the area under the curve of the luminol-based PAMP-triggered ROS-burst assay of three pooled biological replicates, n = 9 plants/replicate. Statistically significant differences of 35S_pro:Zmlox3-GFP relative to control plants 35S_pro:mCherry (Student t test, *P < 0.05, Supplemental Table S4). Statistically significant differences of 35S_pro:Zmlox3-GFP and 35S_pro:Rip1_27–166-mCherry relative to 35S_pro:Zmlox3-GFP/35S_pro:Rip1_27–166-mCherry: *P< 0.05, ***P < 0.05, one-way ANOVA followed by Tukey’s test, Supplemental Table S4.

Figure 4

Zmlox3 is translocated into the nucleus by Rip1 coexpression to increase its ROS suppressive activity. A, Zmlox3 translocalizes into the nucleus in the presence of Rip1 in Z. mays plants. Full-length Zmlox3 was fused to GFP and Rip1 lacking its signal peptide was fused to mCherry under control of 35S_pro, the cauliflower mosaic virus 35S promoter. Constitutively expressed 35S_pro:mCherry and 35S_pro:GFP were used as controls for the colocalization assay, respectively. Zea mays leaves were simultaneously bombarded with a pair of constructs: 35S_pro:Zmlox3-GFP/35S_pro:mCherry, 35S_pro:Rip1_27–166-mCherry/35S_pro:GFP, or 35S_pro:Zmlox3-GFP/35S_pro:Rip1_27–166-mCherry. After 2 days, fluorescence signals were observed by confocal microscopy. Upper: GFP fluorescence, Upper middle panels: mCherry fluorescence, Lower middle: merged mCherry and GFP fluorescence, Lower: brightfield images. Images are merged z-stack projections. Insets represent magnified areas of plant cell nuclei. Arrows indicate positions of the plant cell nuclei. Bars = 20 μm. B, Fluorescence intensity profiles of mCherry and GFP channel images along transection lines (green marking line crossing the closeup of the nuclear region) at nuclear regions are visualized as relative gray values (plots). Purple plot areas represent mCherry intensity levels, green plot areas represent GFP intensity levels. Zmlox3 in the presence of Rip1 shows increased nuclear GFP signals, whereas GFP signal intensities decrease within the nucleus when Zmlox3 is coexpressed with the mCherry control. Images show a single plane of a z-stack from a. Bars = 5 μm. C, Subcellular mislocalization ROS burst assay of Zmlox3 in N. benthamiana using different localization tags (MYR, NLS, and NES). The highest ROS burst suppression activity was observed in nuclear-targeted ZmLox3 plants. Box plots represent the area under the ROS burst curve after challenging with flg22, monitored over 40 min using on the luminescence assay described in “Materials and methods”. Data are from a pool of three biological replicates, n = 7 plants per pool. Asterisks indicate statistically significant differences from the control (*P < 0.05, **P < 0.05, one-way ANOVA followed by Tukey’s test, Supplemental Table S4). D, Glucocorticoid system-based nuclear translocation of Zmlox3 enables ROS burst suppression. Nicotiana benthamiana plants were infiltrated with 35S_pro:Zmlox3 or 35S_pro:mCherry (control) coupled to a binding domain of a GR and were treated with DEX or H₂O after 24 h. Box plots represent the area under the ROS burst curve monitored over 40 min with the luminescence-based assay. Data are from a pool of two biological replicates; n = 6. Statistically significant differences of 35S_pro:Zmlox3-GR or 35S_pro:mCherry-GR treated with DEX to control plants 35S_pro:Zmlox3-GR or 35S_pro:mCherry-GR treated with H₂O (Student t test, NS: non-significant, *P < 0.05, Supplemental Table S4).

Figure 5

Presence of rip1 in Zmlox3 mutant maize plants is causative for the reduced host colonization. A, Virulence assays carried out in 7-day-old maize seedlings cv. B73 and cv. B73ΔLox3 (Zmlox3 deletion mutant, lox3-4) infected with the U. maydis progenitor strain SG200, SG200Δrip1 (rip1 deletion strain), or the complementation strain for SG200Δrip1 (SG200Δrip1-rip1_1–166). The complementation strain was generated by using a full-length Rip1 expression construct under the control of the native Rip1 promoter. Data represent means ± sd from three independent experiments, n = the total number of plants scored. Significant differences between strains were analyzed by the Fisher’s exact test and multiple testing correction was performed using Benjamin–Hochberg. Asterisks indicate statistical significance (***P < 0.0001, Supplemental Table S4). Control plants (B73 and B73Δlox3 infected with SG200) for this experiment were done in collaborations and described in Pathi et al. (2020). In any single biological replicate, all genotypes were present and infected with indicated U. maydis strains. B, Microscopic analysis of the Rip1 virulence phenotype in B73 and B73ΔLox3 plants. Maize plants cv. B73 were infected with either the progenitor strain SG200, the deletion strain SG200Δrip1 or the complementation strain SG200Δrip1‐rip1_1–166, and were harvested at 3 dpi. Images show the intracellular proliferation of U. maydis in epidermal maize cells. Infected leaf tissue was stained with WGA‐AlexaFluor488 (green filaments) to visualize fungal chitin and propidium iodide (red) to observe plant cell walls. Note that the progenitor strain SG200, the deletion strain SG200Δrip1 and the complementation strain SG200Δrip1‐rip1_1–166 were able to penetrate inside maize tissue from cv. B73 and cv. B73 B73ΔLox3. z-stack pictures were produced using confocal microscopy. Bar = 100 µm.

Figure 6

Model summarizing the role of Rip1 in the context of the U. maydis/maize interaction (upper left) and the specific interaction with Lox3 (upper right). Lower left side: Rip1 suppresses PTI in the cytosol by a currently unidentified mechanism. In the U. maydis/maize interaction, Rip1 is leading to Rip1-effector-dependent loss of susceptibility of maize for U. maydis. However, in the presence of Lox3, compatibility is fully restored. Lower right: Rip1 is secreted by U. maydis and is translocated into the plant cell where it suppresses PTI responses. One of its targets is the maize lipoxygenase Lox3 which is shuttled by Rip1 into the nucleus where its lipoxygenase activity is not needed to contribute to suppression of the ROS burst. Rip1 = U. maydis fungal effector, Lox3 = maize lipoxygenase 3.