Control of the pattern-recognition receptor EFR by an ER protein complex in plant immunity

Abstract

In plant innate immunity, the surface-exposed leucine-rich repeat receptor kinases EFR and FLS2 mediate recognition of the bacterial pathogen-associated molecular patterns EF-Tu and flagellin, respectively. We identified the Arabidopsis stromal-derived factor-2 (SDF2) as being required for EFR function, and to a lesser extent FLS2 function. SDF2 resides in an endoplasmic reticulum (ER) protein complex with the Hsp40 ERdj3B and the Hsp70 BiP, which are components of the ER-quality control (ER-QC). Loss of SDF2 results in ER retention and degradation of EFR. The differential requirement for ER-QC components by EFR and FLS2 could be linked to N-glycosylation mediated by STT3a, a catalytic subunit of the oligosaccharyltransferase complex involved in co-translational N-glycosylation. Our results show that the plasma membrane EFR requires the ER complex SDF2-ERdj3B-BiP for its proper accumulation, and provide a demonstration of a physiological requirement for ER-QC in transmembrane receptor function in plants. They also provide an unexpected differential requirement for ER-QC and N-glycosylation components by two closely related receptors.

Figure 1

Identification of sdf2 mutants. (A) Growth phenotype of elfin (elf18-insensitive) seedlings. Five-day-old Arabidopsis seedlings were covered with liquid MS medium containing 1% sucrose (MS 1%) supplemented with 50 nM elf18 peptide. Seedling growth inhibition was scored qualitatively 1 week after treatment. The red arrowhead indicates an elfin mutant. (B) Schematic representation of the SDF2 gene (At2g25110) with positions of the T-DNA insertions and point mutations. Exons are depicted as black boxes. (C) SDF2 protein organisation. Top, schematic representation of SDF2 with the predicted signal peptide (SP) and MIR domains represented as black boxes. Bottom, SDF2 protein sequence. SP, dashed underlines; MIR domains, underlined. The positions of the sdf2 EMS alleles are indicated in bold. (D) Seedling growth inhibition triggered by elf18 in wild type (Col-0) and sdf2 alleles. Five-day-old Arabidopsis seedlings were transferred to liquid MS 1% without (white bars) or with 100 nM elf18 (black bars). Seedling fresh weight was quantified 1 week after treatment. Sdf2C corresponds to sdf2-2/SDF2p::SDF2-3xHA. Results are average±s.e. (n=6). Similar results were observed in at least three independent experiments.

Figure 2

Sdf2 mutant is compromised in PTI responses triggered by elf18 and, to a certain extend, flg22. (A) Seedling growth inhibition triggered by elf18 or flg22 in wild-type Col-0 (white bars) and sdf2-2 (black bars) seedlings. Five-day-old Arabidopsis seedlings were transferred to liquid MS 1% supplemented with the indicated concentrations of peptides. Seedling fresh weight was quantified 1 week after treatment. Results are average±s.e. (n=6). (B) Oxidative burst triggered by 100 nM elf18, 100 nM flg22 or 100 mg/ml chitin in wild-type Col-0 (blue) and sdf2-2 (red) leaf discs measured in relative light units (RLU). Results are average±s.e. (n=12). (C) Activation of the MAP kinases MPK4 and MPK6 in response to 100 nM elf18 or 100 nM flg22 in Col-0, sdf2-2 and erdj3b-1 seedlings (2 weeks old). Two-week-old Arabidopsis seedlings in liquid MS 1% were treated with 100 nM elf18 or flg22. Seedlings were flash frozen in liquid nitrogen at time points indicated. MPK4 and 6 were affinity purified using specific antibodies and used for in vitro kinase reactions performed with the myelin basic protein (MBP) as a substrate in the presence of [γ-³²P]ATP.

Figure 3

Sdf2 mutant is more susceptible to bacterial pathogens. (A–C) Pre-invasive bacterial susceptibility assay. Five-week-old Col-0, fls2c, efr-1, fls2c efr-1 and sdf2-2 plants were sprayed with Pseudomonas syringae pv. tomato (Pto) DC3000 (A), Pto DC3000 COR⁻ (B) or Pto DC3000 ΔavrPto/ΔavrPtoB (C) (OD₆₀₀ of 0.02, supplemented with 0.04% Silwett L-77) and covered for the remaining of the experiment. Bacterial counts were assessed at 3 dpi. Results are average±s.e. (n=8). (D, E) Post-invasive bacterial susceptibility assay. Leaves of 5-week-old plants were infiltrated with Pto DC3000 (OD₆₀₀=0.0002) (D) or Pseudomonas syringae pv. tabaci (Pta) 6005 (OD₆₀₀=0.002) (E). Bacterial populations were determined at 3 dpi. Results are average±s.e. (n=4). For all above experiments, similar results were observed in at least three independent experiments and asterisks indicate P<0.05 by t-test.

Figure 4

Sdf2 mutant is more susceptible to fungal pathogens. (A) Susceptibility to A. brassicicola. Left, macroscopic symptoms and disease rating at 5 dpi. Right, lesion diameter at 5 dpi. Similar results were observed in four independent experiments and asterisks indicate P<0.05 by one-way ANOVA with Bonferroni post hoc test. (B) Susceptibility to P. cucumerina. Left, macroscopic symptoms 14 dpi. Right, average disease rating (DR±s.d.) of the indicated genotypes at 14 dpi. DR varies between 0 (no symptoms) and 5 (dead plants). Letters indicate P⩽0.05 by ANOVA with Bonferroni post hoc test. Data are from one of two independent experiments that gave similar results.

Figure 5

SDF2 exists in an ER complex with ERdj3b and BiP. (A) Subcellular localisation of SDF2 in N. benthamiana. The constructs SDF2p::SDF2-eYFP and ER-CK (35S::HDEL-CFP) were transiently co-expressed by Agrobacterium-mediated transformation in N. benthamiana leaves. Confocal analysis of transformed cells was performed at 2 dpi. The middle row of panels are zoom-ins. (B) Subcellular localisation of SDF2 in transgenic sdf2-2/SDF2p::SDF2-eYFP Arabidopsis leaves. The second column of panels corresponds to zoom-ins. (C) Co-immunoprecipitation of SDF2 with ERdj3B and BiP in vivo. Protein extract from sdf2-2/SDF2p::SDF2-3xHA or wild-type Col-0 seedlings were subjected to immunoprecipitation with anti-HA affinity matrix. Western blot analysis was performed with anti-ERdj3B, anti-BiP and anti-HA antibodies. (D) ERdj3B serves as an adaptor between SDF2 and BiP. Yeast cultures were grown overnight in the -His-Ura-Trp liquid SD medium supplemented with glucose. Cultures were spun down and resuspended in water to OD₆₀₀=1. A series of 10-fold dilutions was plated on -His-Ura-Trp-Leu SD medium supplemented with galactose and raffinose to assess the expression of the LEU reporter gene. Images were taken after 3 days incubation at 30°C.

Figure 6

Erdj3b mutants show a phenotype similar to sdf2. (A) Schematic representation of the ERdj3B gene (At3g62600) with position of the T-DNA insertions and point mutations. Exons are depicted as black boxes. (B) Seedling growth inhibition triggered by elf18 or flg22 in wild-type Col-0 (white bars), sdf2-2 (black bars) and erdj3b-1 (grey bars). Results are average±s.e. (n=6). (C) Oxidative burst triggered by 100 nM elf18 or 100 nM flg22 in wild-type Col-0 (blue), sdf2-2 (purple) and erdj3b-1 (cyan) leaf discs measured in relative light units (RLU). Results are average±s.e. (n=12).

Figure 7

SDF2 controls EFR protein accumulation. (A) Effect of the sdf2-2 mutation on EFR protein levels. Total protein extracts were prepared from T2 transgenic lines (the numbers above the lanes indicate independent lines). Col-0 was used as a negative control. Equal amounts of plant tissue were used in all cases. (B) Relative abundance of EFR expressed under its native promoter in efr-1 and sdf2-2 genetic backgrounds. (C) Effect of the sdf2-2 mutation on EFR glycosylation state. Endo H assays were performed with the EFR protein from efr-1/- and efr-1 sdf2-2/EFRp::EFR-eGFP-HA. EFR was immunoprecipitated using anti-GFP agarose beads (Caltag Medsystems) and incubated with (+) or without (−) Endo H for 1 h at 37°C. (D) Immunoblot of FLS2. Total protein extracts were prepared from Col-0 and sdf2-2 plants as described in (A). FLS2 was detected with a polyclonal anti-FLS2 antibody. (E) Effect of sdf2-2 on the FLS2 glycosylation state. FLS2 was immunoprecipitated using the polyclonal anti-FLS2 antibody and protein G sepharose beads (Sigma). The Endo H assay was performed as in (C). (F) EFR expressed in the sdf2-2 background is stabilised by kifunensine (Kif). Despite the increase in the protein level on treatment with Kif, EFR expressed in sdf2-2 remains sensitive to Endo H. Two-week-old seedlings were incubated for 20 h in the 1/2 MS medium with (+) or without 50 μM Kif. The protein extracts were prepared as in (A) and the Endo H assay was performed as in (C). In (A–C, F) EFR was detected with the anti-GFP antibody (TP401, AMS Biotechnology).

Figure 8

Stt3a affects EF-Tu but not flagellin responsiveness. (A) Schematic representation of the STT3A gene with positions of the single nucleotide deletion and the T-DNA insertion. Exons are depicted as black boxes. stt3a-3 and stt3a-2 (Koiwa et al, 2003) are indicated. (B) Seedling growth inhibition triggered by elf18 or flg22 in wild-type Col-0 (white bars) and stt3a-2 (black bars) seedlings. Five-day-old Arabidopsis seedlings were transferred into liquid MS 1% supplemented with the indicated concentrations of peptides. Seedling fresh weight was quantified 1 week after treatment. Results are average±s.e. (n=6).