Preventing inappropriate signals pre- and post-ligand perception by a toggle switch mechanism of ERECTA

Abstract

Dynamic control of signaling events requires swift regulation of receptors at an active state. By focusing on the Arabidopsis ERECTA (ER) receptor kinase, which perceives peptide ligands to control multiple developmental processes, we report a mechanism preventing inappropriate receptor activity. The ER C-terminal tail (ER_CT) functions as an autoinhibitory domain: Its removal confers higher kinase activity and hyperactivity during inflorescence and stomatal development. ER_CT is required for the binding of a receptor kinase inhibitor, BKI1, and two U-box E3 ligases, PUB30 and PUB31, that trigger activated ER to degradation through ubiquitination. We further identify ER_CT as a phosphodomain transphosphorylated by the coreceptor BAK1. The phosphorylation impacts the tail structure, likely releasing ER from autoinhibition. The phosphonull version enhances BKI1 association, whereas the phosphomimetic version promotes PUB30/31 association. Thus, ER_CT acts as an off-on-off toggle switch, facilitating the release of BKI1 inhibition, enabling signal activation, and swiftly turning over the receptors afterward. Our results elucidate a mechanism that fine-tunes receptor signaling via a phosphoswitch module, maintaining the receptor at a low basal state while ensuring robust yet transient activation upon ligand perception.

Keywords: inflorescence growth; peptide hormone; phosphoswitch; receptor kinase; stomatal development.

Fig. 1.

The conserved C-terminal tail in ER-family LRR-RKs inhibits the kinase activity. (A) Alignment of the C-terminal tail regions of ER and ERLs. Conserved residues (Cons) are filled in black. The kinase domain is highlighted in dark gray, and ER_CT is in red. The location of the α-Helix is superimposed. Deletions used in this study are indicated in blue: ER_ΔCTα-Helix, ER_ΔC43, ER_ΔC56, and ER_ΔC62. (B) Structural modeling of the cytoplasmic domain of ER (ER_CD), with the juxtamembrane domain in sand, the kinase domain in gray, and the C-terminal tail in red. (C) In vitro phosphorylation assay of ER_CD. Left panels, control ER_CD and kinase-dead version (ER_CD_K676E). Right panels, ER_CD and ER_CD_ΔC43. Top panels, autoradiography. Middle panels, mean and S.D. of densitometry plotted for the experiments performed three times. * t test P < 0.05. Bottom panels, Coomassie Brilliant Blue staining (CBB) as a loading control.

Fig. 2.

Removal of the ER C-terminal tail leads to hyperactivity in promoting pedicel elongation and inhibiting stomatal development. (A) Representative pedicels and mature siliques of WT, er-105 (er), ER-FLAG er, ER_ΔCTα-Helix-FLAG er, and ER_ΔC43-FLAG er plants. For each transgenic construct, two representative lines were subjected to analysis. Images were taken under the same magnification. (Scale bar, 1 cm.) (B) Morphometric analysis of pedicel length from each genotype. 6-wk-old mature pedicels (n = 10) were measured. One-way ANOVA followed by Tukey’s HSD test was performed and classified their phenotypes into categories (e.g., a, b, c). (C) Representative confocal microscopy of cotyledon abaxial epidermal at 7 d postgermination (DPG7) from wild type (WT), er-105 (er), ER-FLAG er, ER_ΔCTα-Helix-FLAG er, and ER_ΔC43-FLAG er. The same representative transgenic lines were observed. Images were taken under the same magnification. (Scale bar, 50 µm.) (D) Quantitative analysis. Stomata + meristemoid index (number of stomata and meristemoid per 100 epidermal cells) of the cotyledon abaxial epidermis from 7-d-old seedlings of respective genotypes (n = 10). One-way ANOVA followed by Tukey’s HSD test was performed and classified their phenotypes into categories (e.g., a, b, c).

Fig. 3.

The C-terminal tail is essential for the binding of ER with BKI1 and PUB30. (A) Quantitative analysis of interactions between BKI1 and ER_CD variants (ER_CD, ER_CD_ΔCTα-Helix, and ER_CD_Δ43) using BLI. In vitro binding response curves for recombinantly purified GST-BKI1 and MBP-ER_CD variants at seven concentrations (156.25, 312.5, 625, 1,250, 2,500, 5,000, and 10,000 nM) are shown. Kd values are indicated. Data are representative of three independent experiments. (B) Deletion of the C-terminal tail decreases the association of ER with BKI1 in vivo. Proteins from double transgenic lines carrying ERpro::BKI1-YFP ERpro::ER-FLAG, ERpro::BKI1-YFP ERpro::ER_ΔCTα-Helix-FLAG, and ERpro::BKI1-YFP ERpro::ER_ΔC43-FLAG were immunoprecipitated with anti-FLAG beads (IP). The immunoblots (IB) were probed with anti-FLAG and anti-GFP antibodies, respectively. (C) Quantitative analysis of interactions between PUB30 and ER_CD variants (ER_CD, ER_CD_{ΔCTα-Helix,} and ER_CD_ΔC43) using BLI. In vitro binding response curves for recombinantly purified GST-PUB30 and MBP-ER_CD variants at seven concentrations (156.25, 312.5, 625, 1,250, 2,500, 5,000, and 10,000 nM) are shown. Kd values are indicated. Data are representative of three independent experiments. (D) Deletion of the C-terminal tail decreases the association of ER with PUB30 in vivo. Proteins from PUB30pro::PUB30-YFP; ERpro::ER-FLAG, PUB30pro::PUB30-YFP; ERpro:: ER_ΔCTα-Helix-FLAG, and PUB30pro::PUB30-YFP; ERpro:: ER_ΔC43-FLAG plants were immunoprecipitated with anti-GFP beads (IP), and the immunoblots (IB) were probed with anti-GFP and anti-FLAG antibodies, respectively. (E) Deletion of the C-terminal tail decreases the ubiquitination of ER by PUB30 in vivo. Arabidopsis protoplasts were cotransfected with PUB30-MYC, FLAG-UBQ, together with ER-HA, ER_ΔCTα-Helix-HA, and ER_ΔC43-HA. Five micromolar EPFL6 was used for treatment for 1 h. After immunoprecipitation using anti-FLAG beads, the ubiquitinated ER variants were probed with anti-HA antibody. The total ubiquitinated proteins were probed by anti-FLAG antibody and PUB30 proteins were probed by anti-MYC antibody. The inputs of ER were probed with anti-HA antibody. (F) Representative EPFL6 treatment destabilizes ER variants in Arabidopsis protoplasts coexpressing PUB30-MYC. Protoplasts expressing the indicated proteins were treated with 50 μM CHX and 5 μM EPFL6 for 3 h. before the total protein was examined with immunoblot. The experiment was repeated independently three times with similar results. The quantification of average ER protein abundance (ERs-FLAG/Rubisco) among these three repeats was labeled.

Fig. 4.

The C-terminal tail of ER is phosphorylated. (A) Phosphorylation sites in the C-terminal tail of ER. In vitro auto- and transphosphorylation sites identified by LC–MS/MS analysis are marked in blue and lime green, respectively. In vivo localized phosphorylation sites of ER found in this study are marked with green asterisks. Potential in vivo phosphorylation sites of ER are marked with orange asterisks. Juxtamembrane and C-terminal tail are highlighted in sand and red, respectively. Underline, the Activation Loop. (B and C) MS/MS spectra for selected in vivo phosphorylation sites of ER: Thr947 (B) and Ser955 (C).

Fig. 5.

Phosphorylation of the ER C-terminal tail evicts BKI1 and recruits PUB30/31. (A) Quantitative analysis of interactions between BKI1 and ER_CD variants (ER_CD, ER_CD_T/S8A, and ER_CD_T/S8E) using BLI. In vitro binding response curves for recombinantly purified GST-BKI1 and MBP-ER_CD variants at seven concentrations (156.25, 312.5, 625, 1,250, 2,500, 5,000, and 10,000 nM) are shown. Kd values are indicated. Data are representative of three independent experiments. (B) Phosphorylation of the C-terminal tail decreases the association of ER with BKI1 in vivo. Proteins from ERpro::BKI1-YFP; Col-0, ERpro::BKI1-YFP; ERpro::ER-FLAG, ERpro::BKI1-YFP; ERpro::ER_T/S8A-FLAG, and ERpro::BKI1-YFP; ERpro::ER_T/S8E-FLAG plants were immunoprecipitated with anti-FLAG beads (IP), and the immunoblots (IB) were probed with anti-FLAG and anti-GFP antibodies, respectively. (C) Quantitative analysis of interactions between PUB30 and ER_CD variants (ER_CD, ER_CD_T/S8A, and ER_CD_T/S8E) using BLI. In vitro binding response curves for recombinantly purified GST-PUB30 and MBP-ER_CD variants at seven concentrations (156.25, 312.5, 625, 1,250, 2,500, 5,000, and 10,000 nM) are shown. Kd values are indicated. Data are representative of three independent experiments. (D) Phosphorylation of the C-terminal tail promotes the association of ER with PUB30 and PUB31 in vivo. Arabidopsis protoplasts were cotransfected with PUB30-MYC or PUB31-MYC, together with ER-FLAG, ER_T/S8A-FLAG, and ER_T/S8E-FLAG. Five micromolar EPFL6 was used for treatment for 1 h. After immunoprecipitation using anti-FLAG beads, the immunoblots (IB) were probed with anti-FLAG and anti-MYC antibodies, respectively. (E) Representative pedicels and mature siliques of WT, er, ER-FLAG er, ER_T/S8A-FLAG er, and ER_T/S8E-FLAG er plants. Images were taken under the same magnification. (Scale bar, 1 cm.) (F) Morphometric analysis of pedicel length from each genotype. 6-wk-old mature pedicels (n = 10) were measured. One-way ANOVA followed by Tukey’s HSD test was performed and classified their phenotypes into categories (a, b, c, and d). (G) Confocal microscopy of 7-d-old abaxial cotyledon epidermis of WT, er, ER-FLAG er, ER_T/S8A-FLAG er, and ER_T/S8E-FLAG er plants. Images were taken under the same magnification. (Scale bar, 50 μm.) (H) Quantitative analysis. Stomata + meristemoid index of the cotyledon abaxial epidermis from 7-d-old seedlings of respective genotypes (n = 10). One-way ANOVA followed by Tukey’s HSD test was performed and classified their phenotypes into categories (a, b, c, and d).

Fig. 6.

Mechanism preventing inappropriate signals pre- and postactivation of ER. (A) Structural modeling of ER_CD (Left). Structural modeling of ER_CTα-Helix domain (Right, Top), with the Ser972 residue shown as sticks in red, and ER_CTα-Helix_S972p (Right, Bottom), with the phosphorylated Ser972 residue shown as sticks in cyan. (B) CD spectra of ER_CTα-Helix (black) and ER_CTα-Helix_S972p (cyan). (C) Alignment of ER_KD and BRI1_KD. Based on the importance for the phosphorylation efficiencies toward BKI1, the crucial and minor residues in BRI1 are labeled with “^” and “#”, respectively. Conserved amino acids between ER_KD and BRI1_KD are highlighted in green for identical residues and yellow for residues with similar properties. Nonconserved amino acids are marked with magenta brackets. (D) Structural superimposition of the BRI1–BIM complex (PDB, 4OH4) and the ER AlphaFold2 model, highlighting BKI1_BIM in blue, BRI1_KD in yellow orange, ER_KD in gray, ER_CT in red, and residues in ER that are not conserved in magenta. (E and F) Regulation of stomatal development (E) and inflorescence/pedicel elongation (F) by ER_CT. (Left, the 1st state, OFF) BKI1 (blue) associates with ER (green) in the absence of ligand and ER is in the basal state. (Middle, the 2nd state, ON) Upon perception of EPF2 (E, violet) or EPFL6 (F, pink), ER signaling becomes activated. Both the ER kinase domain and C-terminal tail are phosphorylated by its coreceptor BAK1/SERKs (orange) during the transphosphorylation events. The phosphorylation of the C-terminal tail evicts BKI1. The activated ER-BAK1/SERKs receptor complex transduces signals most likely via BSKs (sand) before the activation of a MAPK cascade and subsequent inhibition of stomatal development (E) or promotion of inflorescence/pedicel elongation (F). TMM (gray) biases the signal activation for stomatal development (E). (Right, the 3rd state, OFF) The phosphorylated ER_CT recruits the E3 ligases PUB30/31 (cyan), which, after being phosphorylated by activated BAK1/SERKs, ubiquitinate ER (dimmed green) for eventual degradation to ensure the robust yet appropriate signaling activation upon ligand perception.