Regulation of Arabidopsis brassinosteroid receptor BRI1 endocytosis and degradation by plant U-box PUB12/PUB13-mediated ubiquitination

Abstract

Plants largely rely on plasma membrane (PM)-resident receptor-like kinases (RLKs) to sense extracellular and intracellular stimuli and coordinate cell differentiation, growth, and immunity. Several RLKs have been shown to undergo internalization through the endocytic pathway with a poorly understood mechanism. Here, we show that endocytosis and protein abundance of the Arabidopsis brassinosteroid (BR) receptor, BR INSENSITIVE1 (BRI1), are regulated by plant U-box (PUB) E3 ubiquitin ligase PUB12- and PUB13-mediated ubiquitination. BR perception promotes BRI1 ubiquitination and association with PUB12 and PUB13 through phosphorylation at serine 344 residue. Loss of PUB12 and PUB13 results in reduced BRI1 ubiquitination and internalization accompanied with a prolonged BRI1 PM-residence time, indicating that ubiquitination of BRI1 by PUB12 and PUB13 is a key step in BRI1 endocytosis. Our studies provide a molecular link between BRI1 ubiquitination and internalization and reveal a unique mechanism of E3 ligase-substrate association regulated by phosphorylation.

Keywords: Arabidopsis; BRI1; E3 ligase; endocytosis; ubiquitination.

Fig. 1.

PUB12 and PUB13 ubiquitinate BRI1. (A) Ubiquitination of different RLKs by PUB13 in vitro. The CDs of RLKs, including BRI1, FLS2, EFR, CERK1, and ERECTA, were purified as the MBP-fusion proteins with an HA tag at the C terminus. The ubiquitination of RLKs by GST-fused PUB13 was detected by a WB with an α-HA antibody after an in vitro ubiquitination reaction (Top). The total ubiquitinated proteins, including both RLK and PUB13, were detected by a WB with an α-ubiquitin (α-Ub) antibody (Middle). RLK inputs were indicated by Coomassie Brilliant Blue (CBB) staining. (B) PUB12 and PUB13 ubiquitinate BRI1 in vitro. The ubiquitination of MBP-BRI1_CD-HA was carried out by using GST-fused PUB12 or PUB13 as the E3 ligase. The E1 enzyme AtUBA1 was excluded (−) or included (+) in the ubiquitination reaction. (C) The PUB13 residues C262 and W289 are essential for its autoubiquitination and ubiquitination on BRI1. The in vitro ubiquitination of MBP-BRI1_CD-HA was performed using GST-PUB13 wild type or mutants as the E3 ligase. (D) PUB13 mediates BRI1 ubiquitination in vivo. Arabidopsis protoplasts were cotransfected with FLAG-tagged ubiquitin (FLAG-Ub), HA-tagged BRI1 (BRI1-HA) and together with a control vector or MYC-tagged PUB13 (PUB13-MYC) and incubated for 10 h followed by treatment with 1 μM BL for 3 h in the presence of 2 μM MG132. The ubiquitinated BRI1 was detected with an α-HA WB after IP with α-FLAG antibody (Top). The total ubiquitinated proteins were detected by an α-FLAG WB (Middle) and PUB13 proteins were detected by an α-MYC WB (Bottom). (E) Reduced BRI1 ubiquitination in pub12pub13. IP was performed using α-GFP antibodies on solubilized microsomal fraction protein extracts from homozygous plants expressing BRI1-mCitrine (mCit) in either wild-type (Col-0) or pub12pub13 mutant background, and from plants expressing BRI1-mCit and BRI1_25KR-mCit in bri1, and subjected to immunoblotting with α-Ub (P4D1) (Top) or α-GFP (Middle). The asterisk indicates nonspecific signals from the same blot as a loading control. (F) Quantification of BRI1 ubiquitination profiles. Error bars represent SD (n = 3). The asterisks indicate statistical significance by using t test (*P < 0.05). (G) BRI1 protein accumulation in pub12, pub13, and pub12pub13. Total proteins were isolated from 5-d-old seedlings and detected by WB using α-BRI1 antibodies. The protein inputs were equilibrated using α-Tubulin antibodies. (H) Quantification of BRI1 abundance (BRI1/Tubulin) (n = 3 biological replicates).

Fig. 2.

BRI1 interacts with PUB13. (A) BL application stimulates BRI1–PUB13 association in Arabidopsis protoplasts. Arabidopsis protoplasts from wild-type (Col-0), det2-1, or bak1-4 mutant plants were cotransfected with BRI1-HA and PUB13-FLAG or a control vector and incubated for 10 h followed by 1 μM BL treatment for 3 h. The association of BRI1–PUB13 was detected by an α-HA WB after α-FLAG IP (Top). The protein levels of BRI1 and PUB13 before IP were detected by α-HA (Middle) or α-FLAG (Bottom) WB, respectively. (B) BL-induced BRI1–PUB13 association in the presence of the proteasome inhibitor MG132 in protoplasts. Arabidopsis Col-0 protoplasts were cotransfected with BRI1-HA and PUB13-FLAG and incubated for 10 h. Protoplasts were pretreated with 2 μM MG132 for 2 h before 1 μM BL treatment for 1 or 3 h. The association of BRI1–PUB13 was detected by an α-HA WB after α-FLAG IP. (C) BL treatment induces BRI1–PUB13 association in Arabidopsis plants. Arabidopsis transgenic seedlings carrying 35S::PUB13-HA alone or with pBRI1::BRI1-GFP were used for Co-IP assay. Fourteen-day-old seedlings were pretreated with 50 µM MG132 for 5 h before treatment with 1 µM BL for 3 h. (D) BRI1 interacts with PUB13 in an in vitro pull-down assay. HA-tagged MBP-fusion proteins (MBP-PUB13-HA or MBP-BRI1_CD-HA) were incubated with glutathione beads coupled with GST, GST-PUB13, or GST-BRI1_CD and the beads were collected and washed for α-HA WB (Top). The protein inputs were determined by α-HA (Middle) or α-GST (Bottom) WB. The position of the corresponding proteins is labeled with an asterisk in α-GST WB.

Fig. 3.

BRI1 interacts with and phosphorylates the ARM domain of PUB13. (A) BRI1 interacts with the PUB13 ARM domain in an in vitro pull-down assay. MBP-BRI1_CD-HA fusion proteins were incubated with glutathione beads coupled with GST or GST-fused various PUB13 truncated proteins and the beads were collected and washed for an α-HA WB (Top). The protein inputs are indicated by CBB staining (Bottom). (B) GST-BRI1_CD phosphorylates GST-PUB13 in vitro. GST-PUB13 proteins were used as substrates and GST-BRI1_CD proteins were used as the kinase in an in vitro kinase assay. Phosphorylation was detected by autoradiography (Top), and the protein loading is shown by CBB staining (Bottom). (C) MBP-BRI1_CD-HA phosphorylates the ARM domain of PUB13. GST-fused various PUB13 truncated proteins were used as substrates and MBP-BRI1_CD-HA as the kinase in an in vitro kinase assay. (D) PUB13^S344 is phosphorylated by BRI1. MS analysis identified the phosphorylated S344 residue of PUB13 after an in vitro kinase assay using GST-BRI1_CD as the kinase. The graph indicates MS/MS spectrum of a doubly charged peptide of m/z 652.29. The 2+ indicates doubly charged and −98 indicates neutral loss of the phosphate group. The MS/MS spectrum was exported from the raw MS/MS file using Xcalibur and annotated manually. (E) PUB13^S344 is required for BRI1-mediated phosphorylation. The serine residues around PUB13^S344 (S³⁴³S³⁴⁴FS³⁴⁶S³⁴⁷) were individually mutated to alanine (A). GST-tagged wild-type or various mutated PUB13 proteins were subjected to an in vitro kinase assay using GST-BRI1_CD as the kinase. (F) PUB13^S344 is not required for BAK1-mediated phosphorylation. GST-fused ARM domains of PUB13 or PUB13^S344A proteins were subjected to an in vitro kinase assay with MBP-BRI1_CD or MBP-BAK1_CD as the kinases. (G) BL-induced phosphorylation of PUB13 in vivo. Arabidopsis protoplasts were cotransfected with BRI1-HA and PUB13-FLAG and incubated for 10 h followed by 1 μM BL treatment for 3 h. BRI1^KM is a kinase-dead mutant of BRI1. The phosphorylation of PUB13 was detected by WB using α-phospho-threonine/serine (α-pT/pS) after α-FLAG IP (Top). PUB13 and BRI1 were detected by WB using α-FLAG (Middle) or α-HA (Bottom) antibodies, respectively.

Fig. 4.

BRI1 kinase activity is required for PUB13-mediated ubiquitination and interaction. (A) The BRI1 kinase-dead mutant (BRI1^KM) blocks ubiquitination mediated by PUB13. The ubiquitination of MBP-BRI1_CD-HA or MBP-BRI1_CD^KM-HA by GST-PUB13 was detected by an α-HA WB after an in vitro ubiquitination reaction. The autoubiquitination of GST-PUB13 was detected by an α-GST WB. The protein loading is indicated by CBB staining. (B) BRI1^KM blocks ubiquitination in vivo. Arabidopsis protoplasts were cotransfected with FLAG-Ub and BRI1-HA or BRI1^KM-HA and incubated for 10 h followed by treatment with 1 μM BL for 3 h with 2 μM MG132. The ubiquitinated BRI1 was detected with an α-HA WB after an α-FLAG IP (Top). The total ubiquitinated proteins were detected by an α-FLAG WB (Bottom). The BRI1 proteins are shown by an α-HA WB (Middle). (C) BRI1^KM blocks interaction with PUB13. GST-PUB13_ARM proteins were incubated with amylose beads coupled with MBP, MBP-BRI1_CD, or MBP-BRI1_CD^KM and the beads were collected and washed for an α-GST WB. The protein inputs are shown by CBB staining. (D) PUB13^S344A blocks interaction with BRI1. MBP-BRI1_CD-HA was incubated with glutathione beads coupled with GST-PUB13_ARM or GST-PUB13_ARM^S344A (S344A) mutant proteins and the beads were collected and washed for an α-HA WB. (E) Reduced interaction between BRI1 and PUB13^S344A in Arabidopsis protoplasts. The Co-IP was performed in Col-0 protoplasts expressing BRI1-HA and PUB13-FLAG or PUB13^S344A-FLAG. (F) PUB13^S344A shows reduced ubiquitination on BRI1. The ubiquitination of MBP-BRI1_CD-HA by GST-PUB13 or GST-PUB13^S344A was detected by an α-HA WB after an in vitro ubiquitination assay. The protein inputs are shown by CBB staining. (G) PUB13^S344E enhances interaction with BRI1. MBP-BRI1_CD-HA was incubated with glutathione beads coupled with GST-PUB13_ARM or GST-PUB13_ARM^S344E (S344E) and the beads were collected and washed for an α-HA WB.

Fig. 5.

Loss of PUB12 and PUB13 causes BR hypersensitivity. (A and B) Seedlings expressing BRI1-mCitrine (mCit) in wild type (Col-0) and in pub12pub13 (two independent lines) were grown for 5 d and used for quantification of the root length. The representative picture is shown in A and quantification is shown in B (n > 20 seedlings). (Scale bar, 5 mm.) (C and F) Averaged hypocotyl length of 5-d-old seedlings grown in the presence of increasing concentrations of BL. Experiments were done in triplicate. Error bars indicate SE (n = 20). The asterisks indicate statistical significance by using t test (*P < 0.05, **P < 0.01). DMSO is the control for BL. Seedlings used in C and F were subjected to WB analysis of BES1 dephosphorylation using an α-BES1 antibody (D and G). The percentage of dephosphorylated BES1 relative to the total BES1 from three biological repeats is shown in E and H.

Fig. 6.

PUB12 and PUB13 regulate BRI1 endocytosis. (A) Analysis of BRI1 internalization in the pub12pub13 mutant. BRI1-mCitrine (mCit) and BRI1_25KR-mCit expressed in either Col-0 or bri1 background were used as controls. Five-day-old seedlings were pretreated with CHX (50 μM) for 1.5 h and root epidermal cells were imaged. (Scale bar, 5 μm.) (B) Relative PM to intracellular fluorescence intensity values were derived from the images shown in A using ImageJ. At least 15 cells from five roots were measured for each line. (C) Analysis of BRI1 internalization by using heat shock-induced BRI1-YFP in Col-0 or in pub12pub13 background. YFP intensity in 5-d-old HS::BRI1-YFP/Col-0 and HS::BRI1-YFP/pub12pub13 root epidermal cells chased for 30, 45, 60, and 90 min after 1 h 37 °C induction was analyzed. (Scale bar, 5 μm.) (D) Relative PM to intracellular fluorescence intensity values were determined from images shown in C with ImageJ. At least 15 cells from five roots were measured for each line. (E) Analysis of BRI1 dwell time in the PM in kymographs obtained from spinning-disk movies of BRI1-mCitrine/Col-0, BRI1-mCit/pub12pub13 (#1 and #2), and BRI1_25KR-mCit/bri1. Arrowheads mark the dwell-time tracks of the proteins at the PM. Timescale, 20 s. (F) Time of residency in the PM of BRI1-mCit in Col-0 and in pub12pub13. The plot graph was based on kymograph analyses of at least 150 tracks from 10 cells of at least five seedlings. The asterisks in D indicate statistical significance by using t test (**P < 0.01). The different letters in B and F indicate statistically significant difference analyzed with one-way ANOVA followed by Tukey’s test (P < 0.05).