Arabidopsis Raf-Like Kinase Raf10 Is a Regulatory Component of Core ABA Signaling

Abstract

Abscisic acid (ABA) is a phytohormone essential for seed development and seedling growth under unfavorable environmental conditions. The signaling pathway leading to ABA response has been established, but relatively little is known about the functional regulation of the constituent signaling components. Here, we present several lines of evidence that Arabidopsis Raf-like kinase Raf10 modulates the core ABA signaling downstream of signal perception step. In particular, Raf10 phosphorylates subclass III SnRK2s (SnRK2.2, SnRK2.3, and SnRK2.6), which are key positive regulators, and our study focused on SnRK2.3 indicates that Raf10 enhances its kinase activity and may facilitate its release from negative regulators. Raf10 also phosphorylates transcription factors (ABI5, ABF2, and ABI3) critical for ABAregulted gene expression. Furthermore, Raf10 was found to be essential for the in vivo functions of SnRK2s and ABI5. Collectively, our data demonstrate that Raf10 is a novel regulatory component of core ABA signaling.

Keywords: ABI5; Raf10; SnRK2s; abscisic acid; phytohormone; signaling.

Fig. 1. Dimerization of Raf10 and Raf11

(A) Schematic diagram of Raf10 and Raf11 domain structure. Amino acid sequences of N-terminal portions are shown in the upper part. Leucine repeats are boxed. PAS, PAC domain (pink box), and Kinase domain are indicated. (B) Yeast two-hybrid assay to determine dimerization. Interactions between full-length and N-terminal deletion constructs were examined. Transformed yeast was grown on medium lacking histidine (-HLUW) supplemented with 3 mM (Raf10) or 1 mM (Raf11) 3-amino-1,2,4-triazole (3-AT). FL, full-length. -LZ, constructs lacking leucine repeats. (C and D) Kinase activity of Raf10 and Raf11 lacking the N-terminal leucine repeat regions. Kinase assays were conducted using the full-length (lanes 1 and 3) or N-terminal deleted recombinant proteins (lanes 2 and 4). Left panels, Coomassie blue-stained gel. Right panels, autoradiography showing ³²P-γ-ATP incorporation. MyBP, myelin basic protein.

Fig. 2. Interaction of Raf10 with ABA signaling components

(A) Interactions between full-length Raf10 and PP2Cs (ABI1, ABI2, HAB1, AHG1, and AHG3) were determined by two-hybrid assay. Full-length Raf10 was used as bait, and PP2Cs were used as prey. Yeast transformants were grown on SC-HWU, and LacZ reporter activity was detected by X-gal overlay assay. Top panel shows growth control on SC-LWU medium. (B) Liquid β-galactosidase assay. LacZ reporter activity was determined by liquid assay using O-nitrophenyl-β-D-galactopyrano-side (ONPG) as a substrate. The numbers indicate Miller units. (C) Interactions between full-length Raf10 and SnRK2s (SnRK2.2, SnRK2.3, and SnRK2.6) were examined by two-hybrid assay. Full-length Raf10 was employed as bait, and SnRK2s were used as prey. Yeast transformatnts were grown on nonselective SC-LW medium (top panel), and LacZ reporter activity was determined by X-gal overlay assay (bottom panel). (D) Interactions between full-length Raf10 and ABI3 or ABI5 were examined by two-hybrid assay. Full-length Raf10 was used as bait, and ABI3 and ABI5 were used as prey. Yeast transformants were grown on SC-LW medium, and LacZ reporter activity was determined by X-gal overlay assay. (E) Interactions between full-length Raf10 and selective ABA signaling components (ABI1, SnRK2.3, ABI3, and ABI5) were determined by coimmunoprecipitation assay. Protein extracts, prepared from tobacco (N. benthamiana) leaves infiltrated with 3HA-pBA and 6myc-pBA constructs, were used in the pull-down with 1 μg of c-myc antibody bound to Dynabeads. Proteins in the input and pulldown samples were detected using Western blot with either c-myc or HA antibody. The numbers in the left of each autoradiogram indicate the position and size of size markers (kDa). (F) Interactions between Raf10 and ABA signaling components (ABI1, SnRK2.2, SnRK2.3, SnRK2.6, ABI3, and ABI5) were determined by BiFC. Protoplasts prepared from tobacco (N. benthamiana) leaves infiltrated with pSPYNE-35S and pSPYCE-35S construct pairs were observed under fluorescent microscope. See also Supplementary Figures S1 and S2.

Fig. 3. Phosphorylation of SnRK2s

(A) Phosphorylation of SnRK2s (SnRK2.2, SnRK2.3, and SnRK2.6) by Raf10. Approximately 1 μg of full-length Raf10 and SnRK2s were used in the in vitro kinase assay. FL, full-length; KD, kinase domain; C, C-terminal domain. Left panel, CBB-stained gel. Right panel, autoradiogram. (B) Phosphorylation of ABI5 by Raf10. ABI5 recombinant proteins. FL, full-length protein (amino acid 1-442); N, N-terminal fragment (amino acid 1–350); bZIP, bZIP domain-containing C-terminal fragment (amino acids 341–442). Left panel, CBB-stained gel. Right panel, autoradiogram. (C) Phosphorylation of ABF2 and ABI3 by Raf10. The same amount of recombinant ABF2 proteins were used in lanes 1 and 2, respectively. FL, full-length (amino acids 1–416); bZIP, C-terminal bZIP-containing fragment (amino acids 321–416). In lanes 3 and 4, recombinant ABI3 proteins were used as substrates. FL, full-length (amino acids 1–720); B3, B3 domain (amino acids 561–720). Left panel, CBB-stained gel. Right panel, autoradiogram. The position of size makers is presented on the left of autoradiograms. The arrowheads and the asterisks indicate relevant phosphorylation band position. See also Supplementary Figures S3 and S4.

Fig. 4. Site-directed mutagenesis to determine phosphorylation sites

(A) Phosphorylation of SnRK2 Domain II (ABA box) by Raf10. Assays were performed as in Figure 3, using full-length (FL) or Domain II constructs. Arrowheads indicate substrate band position. Left panel, CBB-stained gel. Right panel, autoradiogram. (B) Site-directed mutagenesis of SnRK2.3 Domain II (ABA box). Wild-type and mutant recombinant proteins were used as substrates in the assay. The bottom panel shows amino acid sequences of the C-terminal regions of snRK2s. The bar indicates Domain II (ABA box). Potential phosphorylation sites in SnRK2.2 and SnRK2.3 are indicated in red fonts, and S353 is boxed. Left panel, CBB-stained gel. Right panel, autoradiogram. (C) Site-directed mutagenesis of ABI5 C-terminal region. Phosphorylation of two mutant proteins is shown. The bottom panel shows the amino acid sequence of the C-terminal region. Potential phosphorylation sites examined in our study are indicated by red fonts. Left panel, CBB-stained gel. Right panel, autoradiogram. Size markers are shown on the left of autoradiograms in (A–C). See also Supplementary Figure S5.

Fig. 5. Effect of SnRK2.3 phosphorylation by Raf10

(A) Yeast two-hybrid assay to determine S353 phosphorylation effect on ABI1-SnRK2.3 interaction. Interaction between the SnRK2.3 phosphomimic (S353D) mutant and ABI1 are indicated by the red box. (B) Coimmunoprecipitation assay to determine S353 phosphorylation effect on ABI1-SnRK2.3 interaction. HA- or myc-tagged proteins were transiently expressed in tobacco (N. benthamiana) leaves, and coimmunoprecipitation was conducted using tobacco leaf extracts. Pull-down was conducted using 500 μg of extracts and 1 μg of myc antibody. Proteins in the input and pull-down samples were analyzed by Western blot analysis. Size markers are shown beside the blots. The bands corresponding to ABI1 are highlighted by red box. (C) Kinase assay to examine S353 phosphorylation effect on SnRK2.3 kinase activity. Kinase activity of wild-type and the phosphomimic (S353D) mutant of SnRK2.3 was determined using an ABI5 N-terminal fragment (amino acids 1–250) as substrates. The arrowhead shows substrate band position. Exposure time for autoradiograph was at least 10 times longer than that in Figures 3 and 4. See also Supplementary Figure S6.

Fig. 6. In-gel kinase assay to determine Raf10 effect on ABI5 phosphorylation

(A) In-gel kinase assay was conducted using seed extracts (20 μg each) prepared from immature, green siliques and MBP-ABI5 C-terminal fragment (2 mg) embedded in the gel as a substrate. The band corresponding to SnRK2s is indicated by an arrowhead, and unknown kinase bands are indicated by asterisks. (B) In-gel kinase assay using an MBP-ABI5 N-terminal fragment as a substrate. Seed extracts (20 μg) and 3 mg of recombinant proteins were used. The arrowhead indicates the band corresponding to SnRK2s. Left panels, CBB-stained gel. Right panels, autoradiograph.

Fig. 7. Raf10 suppresses abi1-1 phenotypes and is required for ABI5 function

(A) Dormancy of the double mutant progeny from the abi1-1 × Raf10 OX line cross. Freshly harvested seeds were plated without stratification and grown for 5 days before pictures were taken. Cotyledon greening efficiency was determined 7 days after the plating. (B) ABA sensitivity of the double mutant progeny from the abi1-1 × Raf10 OX line cross was determined during seed germination/cotyledon greening stage. Mature, dry seeds were plated and grown for 5 days before pictures were taken, and the cotyledon greening efficiency at 1 μM ABA was determined. (C) ABA sensitivity of ABI5 OX lines was determined during germination/cotyledon greening stage. Seeds were plated and grown for 7 days, and cotyledon greening efficiency was determined.