Arabidopsis SENESCENCE-ASSOCIATED GENE101 stabilizes and signals within an ENHANCED DISEASE SUSCEPTIBILITY1 complex in plant innate immunity

Abstract

Plant innate immunity against invasive biotrophic pathogens depends on the intracellular defense regulator ENHANCED DISEASE SUSCEPTIBILITY1 (EDS1). We show here that Arabidopsis thaliana EDS1 interacts in vivo with another protein, SENESCENCE-ASSOCIATED GENE101 (SAG101), discovered through a proteomic approach to identify new EDS1 pathway components. Together with PHYTOALEXIN-DEFICIENT4 (PAD4), a known EDS1 interactor, SAG101 contributes intrinsic and indispensable signaling activity to EDS1-dependent resistance. The combined activities of SAG101 and PAD4 are necessary for programmed cell death triggered by the Toll-Interleukin-1 Receptor type of nucleotide binding/leucine-rich repeat immune receptor in response to avirulent pathogen isolates and in restricting the growth of normally virulent pathogens. We further demonstrate by a combination of cell fractionation, coimmunoprecipitation, and fluorescence resonance energy transfer experiments the existence of an EDS1-SAG101 complex inside the nucleus that is molecularly and spatially distinct from EDS1-PAD4 associations in the nucleus and cytoplasm. By contrast, EDS1 homomeric interactions were detected in the cytoplasm but not inside the nucleus. These data, combined with evidence for coregulation between individual EDS1 complexes, suggest that dynamic interactions of EDS1 and its signaling partners in multiple cell compartments are important for plant defense signal relay.

Figure 1.

Identification of SAG101 as an EDS1-Interacting Protein in Arabidopsis Leaf Soluble Extracts. (A) Protein gel blot analysis showing levels of HA-tagged EDS1 in a transgenic eds1-1 line used for affinity purification of EDS1 complexes. Ponceau S staining of the membrane shows equal loading. (B) EDS1-interacting proteins were purified from 5-week-old leaves of the HA-tagged EDS1 line or from Ws-0 as a control. Interacting proteins were eluted, separated by SDS-PAGE, and stained with colloidal Coomassie blue. Differential bands (arrowheads) were isolated and identified by mass spectrometry. Molecular mass markers (kilodaltons) are shown at left. (C) SAG101 protein sequence from accession Col-0 showing peptides identified by Q-TOF tandem mass spectrometry analysis of the protein band isolated in (B). (D) Sequence alignment of the N-terminal lipase-like domains of EDS1, PAD4, and SAG101. A predicted signal peptide cleavage position in SAG101 is indicated with an arrow. A potential SAG101 nuclear localization sequence is underlined. Open circles show the positions of predicted Ser hydrolase catalytic residues in EDS1 and PAD4 and their apparent absence in SAG101.

Figure 2.

Characterization of Arabidopsis sag101 Mutants. (A) Scheme of the SAG101 protein showing the positions of two independent dSpm transposon insertions isolated in accession Col-0. (B) Protein gel blot analysis of SAG101 in sag101 knockout lines. Total leaf protein was isolated from unchallenged 4-week-old plants and analyzed with anti-SAG101 antibodies. Ponceau S staining of the membrane shows equal loading.

Figure 3.

Loss of RPP2 and Basal Resistance in pad4 sag101 Mutants. Two-week-old seedlings were spray-inoculated with P. parasitica conidiospores (4 × 10⁴/mL), and pathogen development was recorded. (A) Infection phenotypes of leaves inoculated with P. parasitica isolate Cala2. Leaves were stained with lactophenol trypan blue 7 d after inoculation to visualize pathogen mycelium and necrotic plant cells. HR, hypersensitive response; M, mycelium; TN, trailing necrosis. (B) Sporulation levels of P. parasitica isolates Cala2 and Noco2 on Arabidopsis wild-type and mutant lines. pad4 sag101 double mutants permit pathogen sporulation to levels equivalent to those on eds1-1 and eds1-2. Spores were harvested from leaves and counted 6 d after inoculation. Top, Cala2 is recognized by RPP2 in Col-0 and by RPP1A in Ws-0 but is virulent on Ler. Bottom, Noco2 is virulent on Col-0 but recognized by RPP5 in Ler and by RPP1 in Ws-0. Backgrounds are Ler for eds1-2 and pad4-2 and Ws-0 for eds1-1 and pad4-5. Experiments were repeated twice with similar results. Bars represent means + sd.

Figure 4.

Growth of P. syringae pv tomato Strains in Leaves of Wild-Type and Mutant Arabidopsis. Five-week-old plants of the indicated plant lines were vacuum-infiltrated with a bacterial suspension (5 × 10⁵ colony-forming units [cfu]/mL) of avirulent P. syringae pv tomato (Pst) strain DC3000 expressing avrRps4 (A), Pst DC3000 expressing avrRpm1 (B), or virulent Pst DC3000 without an avr gene (C). Bacterial titers were measured at d 0 (d0) and d 3 (d3). Bacterial growth is expressed as mean values of viable bacteria per cm² of leaf tissue ± sd resulting from two replicate samplings for d0 and three replicate samplings for d3.

Figure 5.

EDS1, PAD4, and SAG101 Proteins Are Stabilized by Their Interacting Partners. Protein gel blot analysis of total protein extracts derived from 4-week-old unchallenged leaves of different Arabidopsis lines. Equal loading is shown by Ponceau S staining of the membranes. (A) Accumulation of Myc-PAD4 requires EDS1. (B) SAG101 protein requires EDS1 but not PAD4 for accumulation. Equal amounts of total soluble protein of the indicated lines were separated by gel filtration, and SAG101-containing fractions were pooled and analyzed by protein gel blotting. (C) Maximal Myc-PAD4 accumulation depends on SAG101. (D) EDS1 protein is depleted incrementally in pad4, sag101, and pad4 sag101 backgrounds. Numbers below the blot indicate band intensities relative to the EDS1 signal obtained for wild-type Col-0, as measured by ImageQuant 5.2 software.

Figure 6.

Infection Phenotypes of Arabidopsis Mutants Depleted in EDS1. (A) EDS1 abundance in total protein extracts from 4-week-old unchallenged leaves of the indicated Arabidopsis lines. All mutants are in Col-0, except eds1-2 (Ler). Equal loading is shown by Ponceau S staining of the membrane. (B) Sporulation levels of P. parasitica isolates Cala2, recognized by RPP2 (left), and virulent Noco2 (right) on Arabidopsis lines tested in (A). Two-week-old seedlings were spray-inoculated with P. parasitica conidiospores, and spores were counted as described for Figure 3. Experiments were repeated twice with similar results. Bars represent means + sd.

Figure 7.

Distinct EDS1 Complexes Are Present in Soluble Leaf Extracts. Size exclusion chromatography was used to separate total soluble protein extracted from 5-week-old unchallenged leaves of the indicated lines. Individual fractions from a Superdex 200 16/60 column were analyzed for the presence of EDS1-, Myc-PAD4–, and/or SAG101-containing complexes. Schemes of possible monomeric, dimeric, and trimeric protein associations are shown at top. Equal amounts of total protein per line were separated for each gel filtration experiment. (A) Profiles of EDS1, PAD4, and SAG101 protein complexes in wild-type and mutant lines. (B) Effect of sag101-2 and eds1-1 mutations on apparent Myc-PAD4 complex size. Removal of SAG101 or EDS1 protein does not significantly alter apparent Myc-PAD4 complex size. The top gel was exposed for 1 min, the middle gel was exposed for 5 min, and the bottom gel was exposed for 10 min to compensate for overall reduced Myc-PAD4 protein levels in sag101 and eds1 mutants.

Figure 8.

Subcellular Localizations and FRET Interaction Studies of EDS1, SAG101, and PAD4. (A) Arabidopsis epidermal cells were cotransfected with fluorescently tagged EDS1 and SAG101 (top row) or EDS1 and PAD4 (bottom row) and analyzed by confocal laser scanning microscopy. Images shown are three-dimensional reconstructions from individual image stacks. (B) Protein gel blot analysis of EDS1, SAG101, and Myc-PAD4 in subcellular fractions of unchallenged leaf tissues. Histone H3 was used as a nuclear marker, and cytosolic Hsc70s served as a cytosolic marker. N, nuclear protein extracts; S, total protein extracts depleted of nuclei. (C) FRET-APB analysis of the interaction in nuclei between EDS1-CFP and SAG101-YFP. Mean FRET efficiencies ± sd from individual sample sites (>30 for EDS1–SAG101 and 10 to 20 for controls) are shown. Representative images of pseudocolored nuclei show donor fluorescence before and after bleaching for each cotransfection. An increase of donor fluorescence (red) is seen only if protein–protein interaction occurs. (D) FRET-APB analysis of the interaction between EDS1-CFP and EDS1-YFP. Mean FRETefficiencies ± sd from individual sample sites (20 for EDS1–EDS1 and 15 for controls) are shown.