Arabidopsis SOMATIC EMBRYOGENESIS RECEPTOR KINASE proteins serve brassinosteroid-dependent and -independent signaling pathways

Abstract

The Arabidopsis (Arabidopsis thaliana) SOMATIC EMBRYOGENESIS RECEPTOR KINASE (SERK) genes belong to a small family of five plant receptor kinases that are involved in at least five different signaling pathways. One member of this family, BRASSINOSTEROID INSENSITIVE1 (BRI1)-ASSOCIATED KINASE1 (BAK1), also known as SERK3, is the coreceptor of the brassinolide (BR)-perceiving receptor BRI1, a function that is BR dependent and partially redundant with SERK1. BAK1 (SERK3) alone controls plant innate immunity, is also the coreceptor of the flagellin receptor FLS2, and, together with SERK4, can mediate cell death control, all three in a BR-independent fashion. SERK1 and SERK2 are essential for male microsporogenesis, again independent from BR. SERK5 does not appear to have any function under the conditions tested. Here, we show that the different SERK members are only redundant in pairs, whereas higher order mutant combinations only show additive phenotypes. Surprisingly, SERK members that are redundant within one are not redundant in another pathway. We also show that this evolution of functional pairs occurred by a change in protein function and not by differences in spatial expression. We propose that, in plants, closely related receptor kinases have a minimal homo- or heterodimeric configuration to achieve specificity.

Figure 1.

Phenotypic analyses of the multiple serk mutants. A and C, Root growth measurements of seedlings grown on medium containing different BL concentrations using various double-mutant combinations (A) or multiple mutant combinations (C). Each measurement is represented as a percentage of the root elongation of the control plants grown on medium containing the same volume of 80% (v/v) ethanol used to dilute BL. B and D, Quantitative analysis of the hypocotyl length of various double-mutant combinations (B) and multiple serk mutant combinations (D) grown in the dark for 5 d. Each measurement represents an average of hypocotyl lengths of 20 seedlings. Error bars indicate sd. E, RT-PCR analysis to evaluate the feedback regulation of the CPD gene in the different double-mutant backgrounds after treatment with different concentrations of BL. The constitutively expressed cyclophilin gene ROC5 (Chou and Gasser, 1997) was used as control. *, Significant differences from Col-0 wild type (P ≤ 0.05); **, significant differences from the serk3-1 mutant (P ≤ 0.05); s1-1, serk1-1 allele; s1-3, serk1-3; s2, serk2-2; s3, serk3-1; s4, serk4-1; s5, serk5-1; H, heterozygote. [See online article for color version of this figure.]

Figure 2.

35S:bes1-D rescues the serk1-1 serk3-1 phenotype, but fails to rescue the serk1-1 serk2-2 and serk3-1 serk4-1 phenotypes. A, The serk3-1 serk4-1 phenotype is not rescued by the 35S:bes1-D construct. B, Roots of transgenic plants transformed with the 35S:bes1-D-GFP fusion are visualized by confocal microscopy. C, Western analysis of the serk3-1 serk4-1 double mutant transformed with the 35S:bes1-D-GFP constructs (1) as compared to the serk3-1 serk4-1 double mutant (2). D, serk1-1 serk2-2 anther showing no pollen grain and bri1-201 anther showing pollen grain. E, Quantitative analysis of the hypocotyl length of serk3-1 serk4-1 double mutants transformed with the 35S:bes1-D-GFP, serk3-1, and Col-0 wild-type plants grown in the dark for 5 d. Each measurement represents an average of hypocotyl lengths of 20 seedlings. Error bars indicate sd. F, Root growth measurements of seedlings grown on medium containing different BL concentrations. Each measurement is calculated as a percentage of the root elongation of the control plants grown on medium containing the same volume of 80% (v/v) ethanol used to dilute BL. *, Significant differences from Col-0 wild type (P ≤ 0.05); s1, serk1-1 allele; s3, serk3-1; s4, serk4-1; kd, kilodalton. [See online article for color version of this figure.]

Figure 3.

SERK3 and SERK4 are partially redundant in cell death control, but not SERK1, SERK2, and SERK5. A, Infection phenotypes of representative Col-0 wild type and various single and double serk mutants at 7 DAI with A. brassicicola. B, Quantitative analysis of the growth of A. brassicicola in wild-type Col-0 and various single and double serk mutants. Results represent means ± sd (n ≥ 8). *, Significant differences from Col-0 wild type (P ≤ 0.05). [See online article for color version of this figure.]

Figure 4.

SERK overexpression only partially rescues the weak allele bri1-301. A, Overexpression of SERK genes partially suppresses the rosette phenotype of the weak bri1-301 mutation. B, Western analysis to confirm the elevated expression of SERK proteins in the transgenic lines. C, Quantitative analysis of the hypocotyl length of plants overexpressing SERK proteins, bri1-301 mutant, and Col-0 wild-type plants grown in the dark for 5 d. Each measurement represents an average of hypocotyl lengths of 20 seedlings. Error bars indicate sd. D, Root growth measurements of plants grown on medium containing different BL concentrations. Mq, Molecular size marker in daltons (da); P1, P2, and P4, SERK1, SERK2, and SERK4 promoters, respectively; 35S, 35S promoter; wt, wild type. [See online article for color version of this figure.]

Figure 5.

SERK genes are involved in several independent pathways. Model of pathways involving the five SERK genes (SERK1–SERK5) containing five LRRs (stripes) and one SPP (red square) domain as defined by Hecht et al. (2001). Shown are: BR pathway involving BRI1, SERK1, and SERK3 (Li et al., 2002; Nam and Li, 2002; Karlova et al., 2006); FLS2 pathway involving SERK3 essential to innate immunity (Chinchilla et al., 2007); pathway involving likely EMS1/EXS (Canales et al., 2002; Zhao et al., 2002), TPD1 (Yang et al., 2003), and SERK1 and SERK2 mediating male sporogenesis (Albrecht et al., 2005; Colcombet et al., 2005; C. Albrecht and S.C. de Vries, unpublished data); and pathway involving SERK3, SERK4, and probably an unidentified RLK leading to cell death control (He et al., 2007; Heese et al., 2007; Kemmerling et al., 2007). [See online article for color version of this figure.]