GSK3-like kinases positively modulate abscisic acid signaling through phosphorylating subgroup III SnRK2s in Arabidopsis

Abstract

Arabidopsis glycogen synthase kinase 3 (GSK3)-like kinases have versatile functions in plant development and in responding to abiotic stresses. Although physiological evidence suggested a potential role of GSK3-like kinases in abscisic acid (ABA) signaling, the underlying molecular mechanism was largely unknown. Here we identified members of Snf1-related kinase 2s (SnRK2s), SnRK2.2 and SnRK2.3, that can interact with and be phosphorylated by a GSK3-like kinase, brassinosteroid insensitive 2 (BIN2). bin2-3 bil1 bil2, a loss-of-function mutant of BIN2 and its two closest homologs, BIN2 like 1 (BIL1) and BIN2 like 2 (BIL2), was hyposensitive to ABA in primary root inhibition, ABA-responsive gene expression, and phosphorylating ABA Response Element Binding Factor (ABF) 2 fragment by in-gel kinase assays, whereas bin2-1, a gain-of-function mutation of BIN2, was hypersensitive to ABA, suggesting that these GSK3-like kinases function as positive regulators in ABA signaling. Furthermore, BIN2 phosphorylated SnRK2.3 on T180, and SnRK2.3(T180A) had decreased kinase activity in both autophosphorylation and phosphorylating ABFs. Bikinin, a GSK3 kinase inhibitor, inhibited the SnRK2.3 kinase activity and its T180 phosphorylation in vivo. Our genetic analysis further demonstrated that BIN2 regulates ABA signaling downstream of the PYRABACTIN RESISTANCE1/PYR1-LIKE/REGULATORY COMPONENTS OF ABA RECEPTORS receptors and clade A protein phosphatase 2C but relies on SnRK2.2 and SnRK2.3. These findings provide significant insight into the modulation of ABA signaling by Arabidopsis GSK3-like kinases.

Keywords: kinase activation; phosphorylation cascades; signal transduction.

Fig. 1.

BIN2 interacts with subgroup III SnRK2s and phosphorylates SnRK2.2 and SnRK2.3. (A) BIN2 interacts with SnRK2.2, SnRK2.3, and SnRK2.6 in BiFC assay. nYFP-BIN2 or nYFP was cotransformed into pavement cells of N. benthamiana with cYFP (a and e), SnRK2.2-cYFP (b and f), SnRK2.3-cYFP (c and g), or SnRK2.6-cYFP (d and h). (Scale bar, 50 μm.) (B) SnRK2.2, SnRK2.3, and SnRK2.6 interact with BIN2 in vitro. (C) SnRK2.2, SnRK2.3, and SnRK2.6 can pull down BIN2-FLAG in whole-protein extract from BIN2-FLAG transgenic plants. (D) BIN2 phosphorylates SnRK2.2 and SnRK2.2^K52N (K52N) in vitro. (E) BIN2 phosphorylates SnRK2.3 and SnRK2.3^K51N (K51N) in vitro. (F) BIN2 cannot phosphorylate SnRK2.6 or SnRK2.6^K50N (K50N) in vitro. K69R is the abbreviation for BIN2^K69R.

Fig. 2.

BIN2 plays positive roles in modulating ABA signaling outputs. (A) Primary root phenotype of wild-type (Ws-2) and bin2-3 bil1 bil2 grown on medium with or without (Mock) 10 μM ABA. (B) Statistic analysis of relative primary root length of Ws-2 (n = 25) and bin2-3 bil1 bil2 (n = 25). (C) Induction of RAB18, RD26, and RD29B expression by ABA in Ws-2 and bin2-3 bil1 bil2. Seedlings were treated with 50 μM ABA or solvent only (Mock) for 1 h. (D) Primary root phenotype of wild-type (Col-0) and bin2-1 grown on medium without (Mock) or with 10 μM ABA. (E) Statistic analysis of relative primary root length for Col-0 (n = 20), bin2-1 heterozygote (HT) (n = 35), and bin2-1 homozygote (HM) (n = 20). (F) Induction of RAB18, RD26, and RD29B expression by ABA in Col-0 and bin2-1. Seedlings were treated with 50 μM ABA or solvent only (Mock) for 1 h. (G) Expression levels of RAB18, RD26, and RD29B in bin2-3 bil1 bil2 and Ws-2. Expression level of each gene in Ws-2 was normalized to “1.” (H) Expression level of RAB18, RD26, and RD29B in bin2-1and Col-0. Expression level of each gene in Col-0 was normalized to “1.” (I) Bikinin inhibits expression of RAB18, RD26, and RD29B in Col-0. The expression level of each gene in Col-0 grown on medium with or without (Mock) 10 μM bikinin. Expression level of each gene in Col-0 grown on medium without bikinin was normalized to “1.” (J) Signal intensity of in-gel kinase assay on phosphorylating the ABF2 fragment after treatment with 50 μM ABA or solvent only (Mock) for 1 h. (K) Signal intensity of in-gel kinase assay on phosphorylating the ABF2 fragment after treatment with 50 μM ABA or solvent only (Mock) for 1 h. In J and K, the arrowheads indicate the ABA-induced bands representing the activated SnRK2s. “*” indicates unknown, none ABA-induced bands, and right graphs in J and K show relative radioactivity intensity of ABA-inducible bands (means ± SE, n = 4). The relative intensity of ABA-inducible bands in both Ws-2 and Col-0 after ABA treatment was normalized to “1.” Coomassie brilliant blue R250 staining (CBB) of the large subunit of Rubisco (RbcL) was used as the loading control. Also see replicates 2–4 in Fig. S3 E and F. (Scale bars, 1 cm in A and D.) Primary root length of each material in mock treatment was normalized to 100% in B and E. Expression level of each gene for each material under mock treatment was normalized to “1” in C and F. Values are means ± SE for B and C and E–I. Student's t test was used to determine the significance of the indicated comparisons. Significant levels: **P < 0.01; ***P < 0.001.

Fig. 3.

BIN2 phosphorylates SnRK2.3 and enhances its kinase activity. (A) In vitro kinase assays of various mutant forms of SnRK2.3 with BIN2 kinase. Potentially phosphorylated Ser or Thr residues of SnRK2.3 were mutated to Ala, and each mutation was combined with a kinase dead mutation K51N. SnRK2.3^{K51N S168A} is abbreviated for K51N S168A, and the same rule also applied for other mutant forms. (B) In vitro phosphorylation assays of SnRK2.3 and SnRK2.3^T180A on the ABF2 fragment. (C) Relative expression levels of RAB18, RD26, and RD29B in Col-0, SnRK2.3-OX, and SnRK2.3^T180A-OX after treatment with 50 μM ABA or solvent only (Mock) for 1 h. Expression level of each gene for each material under mock treatment was normalized to “1.” Values are means ± SE. Student's t test was used to determine the significance of the indicated comparisons. Significant levels: *P < 0.05; **P < 0.01; ***P < 0.001; and n.s. (not significant, P > 0.05). (D) In vitro phosphorylation assays of SnRK2.3 on ABF2 fragment in the presence of BIN2 or BIN2^K69R (K69R). (E) Immuno-kinase assays of SnRK2.3-MYC on ABF2 fragment. The SnRK2.3-MYC-OX plants were grown on 1/2 MS without (Mock) or with 5 μM bikinin for 10 d and treated with 50 μM ABA for 1 h. The IPed SnRK2.3-MYC was used to phosphorylate ABF2 fragment. (F) Quantitative mass spectrum analysis on phosphorylation of the IPed SnRK2.3-MYC. SnRK2.3-MYC was IPed from seedling as described in Fig. 3E and digested by trypsin. Phospho-peptides were enriched with TiO₂ beads. The sequence of phospho-peptide for quantification is shown. Phosphorylation sites are marked with “*.” Signal intensity of the target phospho-peptide from the mock treatment was normalized to “1.” Two independent experiments were conducted.

Fig. 4.

BIN2 promotes ABA signaling outputs through subgroup III SnRK2s. (A) Root phenotype of Col-0:er, bin2-1:er, pyr1 pyl1 pyl2 pyl4, and bin2-1 pyr1 pyl1 pyl2 pyl4 grown on medium with or without (Mock) 10 μM ABA. (B) Relative primary root length in Col-0:er (n = 25), bin2-1:er (n = 15), pyr1 pyl1 pyl2 pyl4 (n = 25), and bin2-1 pyr1 pyl1 pyl2 pyl4 (n = 20). (C) Induction of RAB18, RD26, and RD29B expression by ABA in pyr1 pyl1 pyl2 pyl4 and bin2-1 pyr1 pyl1 pyl2 pyl4. Seedlings were treated with 50 μM ABA or solvent only (Mock) for 1 h. (D) Root phenotype of Col-0, abi2-3, bin2-1, and abi2-3 bin2-1. (E) Relative primary root length in Col-0 (n = 25), bin2-1 (n = 15), abi2-3 (n = 25), and abi2-3 bin2-1 (n = 20). (F) Induction of RAB18, RD26, and RD29B expression by ABA in abi2-3 and abi2-3 bin2-1. Seedlings were treated with 50 μM ABA or solvent only (Mock) for 1 h. (G) Primary root phenotype of Col-0, bin2-1, SnRK2-RNAi/Col-0, and SnRK2-RNAi/bin2-1. (H) Relative primary root length in Col-0 (n = 20), SnRK2-RNAi/Col-0 (n = 20), bin2-1 (HM) (n = 20), and SnRK2-RNAi/bin2-1 (HM) (n = 20). (I) Induction of RAB18, RD26, and RD29B expression by ABA in bin2-1 and SnRK2-RNAi/bin2-1. Seedlings were treated with 50 μM ABA or solvent only (Mock) for 1 h. (Scale bars, 1 cm in A, D, and G). Primary root length of each material in mock treatment was normalized to 100%. Expression level of each gene for each material under mock treatment was normalized to “1” in C, F, and I. Values are means ± SE in B, C, E, F, H, and I. Student's t test was used to evaluate the significance of indicated comparisons. Significant levels: *P < 0.05; **P < 0.01; and ***P < 0.001.