The receptor-like kinase SERK3/BAK1 is a central regulator of innate immunity in plants

Abstract

In pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI), plant cell surface receptors sense potential microbial pathogens by recognizing elicitors called PAMPs. Although diverse PAMPs trigger PTI through distinct receptors, the resulting intracellular responses overlap extensively. Despite this, a common component(s) linking signal perception with transduction remains unknown. In this study, we identify SOMATIC EMBRYOGENESIS RECEPTOR KINASE (SERK)3/brassinosteroid-associated kinase (BAK)1, a receptor-like kinase previously implicated in hormone signaling, as a component of plant PTI. In Arabidopsis thaliana, AtSERK3/BAK1 rapidly enters an elicitor-dependent complex with FLAGELLIN SENSING 2 (FLS2), the receptor for the bacterial PAMP flagellin and its peptide derivative flg22. In the absence of AtSERK3/BAK1, early flg22-dependent responses are greatly reduced in both A. thaliana and Nicotiana benthamiana. Furthermore, N. benthamiana Serk3/Bak1 is required for full responses to unrelated PAMPs and, importantly, for restriction of bacterial and oomycete infections. Thus, SERK3/BAK1 appears to integrate diverse perception events into downstream PAMP responses, leading to immunity against a range of invading microbes.

Fig. 1.

Elicitor-dependent formation of an AtSERK3/BAK1–AtFLS2 complex and requirement for AtSERK3/BAK1 in flg22-dependent responses in A. thaliana. (A) Specificity of the αAtFLS2 antibody. Solubilized microsomes of Col-0 and Δfls2 seedlings were subjected to immunoprecipitation and immunoblot analysis by using αAtFLS2 (αFLS2) antibodies. (B) Coimmunoprecipitation of a 70-kDa protein (open arrowhead) with AtFLS2 (closed arrowhead) by using αAtFLS2 antibody after 100 nM flg22 elicitation from La-er cultured cells for the indicated time (in minutes). After separation on SDS/PAGE, immunoprecipitates were stained with Sypro ruby for total proteins. Proteins coimmunoprecipitated independent of flg22-elicitation are indicated by asterisks. (C) Dose-dependent AOS production at 14 min after elicitation with increasing flg22 concentrations in WT Col-0, serk1-1, serk2-1, bak1-4, and fls2Δ mutant lines, as indicated. Every experiment (n = 5) was repeated at least three times with similar results. For this experiment, n = 5 samples were analyzed, with an SEM (±) as indicated (error bars). Two sample t tests were performed for every experiment comparing each mutant to Col-0 at the same flg22 concentrations. At each flg22 concentration, the P values were highly significant (P < 0.01) for bak1-4 and Δfls2 mutant but not for serk1-1 and serk2-1 (P ≥ 0.097). Statistical significance is indicated by asterisks. (D) Dose-dependent MAPK activation after increasing flg22 concentrations at 15 min in Col-0 (WT), serk1-1, serk2-1, bak1-4, and Δfls2 seedlings as shown by immunoblot analysis by using an αp44/42-ERK antibody (αP-MAPK^act). Individual MAPKs are identified by mass. The same blots were stripped and probed with αMPK6 for equal loading. (E) flg22-dependent coimmunoprecipitation of AtFLS2 with AtSERK3/BAK1-GFP after 100 nM flg22 elicitation in bak1-4 seedlings expressing AtBAK1-GFP (b1-4/BAK1-GFP) but not in nonelicited bak1-4/AtBAK1-GFP or in elicited nontransformed bak1-4 or Col-0 seedlings. Solubilized membrane proteins were subjected to immunoprecipitation with αGFP antibodies followed by immunoblot analysis with antibodies indicated to the right to detect AtFLS2 (αFLS2), AtBAK1/SERK3-GFP (αGFP), and AtCalnexin (CNX), a membrane ER protein (αCNX).

Fig. 2.

Flagellin-triggered immunity in N. benthamiana. (A) Quantitative RT-PCR of NbFls2, NbSerk2, and NbSerk3 expression in N. benthamiana tissue silenced for RNA2:EV, RNA2:NbFls2, or RNA2:NbSerk3. The samples were normalized against the housekeeping gene NbEF1α. n = 3 samples were analyzed per experiment with the SEM (±), as indicated (error bars). For every experiment, two sample t tests were performed and showed a significant reduction of the respective gene with P ≤ 0.01. (B) AOS burst in N. benthamiana tissue silenced for EV, NbFls2, NbSerk3, and NbSerk2 induced by 50 nM flg22. For this experiment, n = 12 samples were analyzed, with an SEM (±) as indicated (error bars). Two sample t tests show that EV is significantly different from NbFls2 and NbSerk3 by P ≤ 0.05. (C) Quantitative RT-PCR analysis of defense gene expression induced by 100 nM flg22 for 30 min in tissue silenced for EV, NbFls2, or NbSerk3 relative to untreated tissue. All samples were normalized against the housekeeping gene NbEF1α. n = 3 samples were analyzed per experiment with the SEM (±), as indicated (error bars). For every experiment, two sample t tests were performed and showed that EV is significantly different from NbFls2 and NbSerk3 by P ≤ 0.01. (D) Bacterial growth curves with three P. syringae strains on N. benthamiana silenced for EV, NbFls2, or NbSerk3 at 0 and 3 days post infiltration. n = 3 samples were analyzed per experiment with the SEM (±), as indicated (error bars). For every experiment, two sample t tests were performed and showed that EV is significantly different from NbFls2 and NbSerk3 by P ≤ 0.05. Every experiment of Fig. 2 was repeated at least three times with similar results. Statistical significance is indicated by asterisks.

Fig. 3.

NbSerk3 is required for multiple PAMP-mediated responses in N. benthamiana. (A) AOS burst in N. benthamiana tissue silenced for EV (black), NbFls2 (gray), or NbSerk3 (white) treated with 100 nM CSP22 or 10 μg/ml INF1. For each treatment, n = 12 samples were analyzed, with an SEM (±) as indicated (error bars). Two sample t tests show that EV and NbFls2 are significantly different from NbSerk3 by P ≤ 0.05 in each treatment. Statistical significance is indicated by asterisks. (B) Trypan blue staining indicative of cell death of EV, NbFls2, or NbSerk3 silenced N. benthamiana treated with 100 μg/ml INF1 or inf1. (C) Trypan blue staining was used to visualize cell death and oomycete structures on EV, NbFls2, or NbSerk3 silenced N. benthamiana tissue infected with H. parasitica. A. thaliana Col-0 tissue infected with H. parasitica was used as an infection control. The pictures were taken 3 weeks after inoculation. The experiment was performed three times with similar results.