Activation of SnRK2 by Raf-like kinase ARK represents a primary mechanism of ABA and abiotic stress responses

Abstract

The Raf-like protein kinase abscisic acid (ABA) and abiotic stress-responsive Raf-like kinase (ARK) previously identified in the moss Physcomitrium (Physcomitrella) patens acts as an upstream regulator of subgroup III SNF1-related protein kinase2 (SnRK2), the key regulator of ABA and abiotic stress responses. However, the mechanisms underlying activation of ARK by ABA and abiotic stress for the regulation of SnRK2, including the role of ABA receptor-associated group A PP2C (PP2C-A), are not understood. We identified Ser1029 as the phosphorylation site in the activation loop of ARK, which provided a possible mechanism for regulation of its activity. Analysis of transgenic P. patens ark lines expressing ARK-GFP with Ser1029-to-Ala mutation indicated that this replacement causes reductions in ABA-induced gene expression, stress tolerance, and SnRK2 activity. Immunoblot analysis using an anti-phosphopeptide antibody indicated that ABA treatments rapidly stimulate Ser1029 phosphorylation in the wild type (WT). The phosphorylation profile of Ser1029 in ABA-hypersensitive ppabi1 lacking protein phosphatase 2C-A (PP2C-A) was similar to that in the WT, whereas little Ser1029 phosphorylation was observed in ABA-insensitive ark missense mutant lines. Furthermore, newly isolated ppabi1 ark lines showed ABA-insensitive phenotypes similar to those of ark lines. Therefore, ARK is a primary activator of SnRK2, preceding negative regulation by PP2C-A in bryophytes, which provides a prototype mechanism for ABA and abiotic stress responses in plants.

Figure 1

Analysis of P. patens lines expressing ARK-GFP constructs. (A) Schematic representation of ARK-GFP showing positions of Ser532 (S532) in the non-kinase domain and putative phosphorylation sites Ser1029 (S1029) and Ser1030 (S1030) in the activation loop of the kinase domain. (B) Growth responses to ABA of WT, AR7, and AR7 lines expressing ARK-GFP without mutation (ARK) and with S532F, S1029A, S1030A, and S1029A/S1030A mutations. Protonemata of these lines were cultured with or without 10-µM ABA for 1 week. (C) ABA-induced gene expression in transgenic lines. Protonemata were bombarded with plasmid constructs of the Em promoter fused with the beta-glucuronidase (proEm-GUS) gene and the rice ubiquitin promoter fused with the luciferase gene (proUbi-LUC) and were cultured with or without 10-µM ABA for 1 d before GUS and LUC assays. Error bars indicate standard error of the mean. *P < 0.05 in the t test (n = 3) compared with the ABA-treated ARK line. (D) Freezing tolerance of ARK-GFP lines. Protonemata were cultured with or without 10-µM ABA for 1 d and subjected to freezing at 10°C. After thawing, electrolyte leakage was determined to estimate the damage caused by freezing. Error bars indicate standard error of the mean. *P < 0.05 in the t test (n = 3) compared with the ABA-treated ARK line. (E) In-gel kinase assays for detection of SnRK2 activity. Proteins extracted from protonemata were electrophoresed using SDS-polyacrylamide gel polymerized with histone IIIS as a substrate. After denaturation and renaturation processes, the proteins were reacted with ³²P-ATP and radioactive signals were detected by exposure to X-ray film. Staining of the large subunit of ribulose bisphosphate carboxylase (rbcL) is shown as a control.

Figure 2

ABA-stimulated phosphorylation of Ser1029 detected by anti-phosphopeptide antibody. (A) Proteins from protonemata of transgenic P. patens AR7 lines expressing ARK-GFP (ARK) and ARK-GFP with S532F, S1029A, and S1030A mutations with or without treatment with 10-µM ABA for 15 min were subjected to immunoblot analysis using an anti-phospho-Ser1029 (P-S1029) antibody and an anti-GFP antibody. Since levels of ARK-GFP accumulation vary among these lines, the amount of proteins loaded per lane was adjusted using the anti-GFP antibody. Note that ARK^S1030A-GFP tends to give enhanced signals for an unknown reason. (B) Detection of native ARK in WT P. patens. Protonemata were treated with 10-µM ABA for the indicated time periods, and extracted proteins were reacted with the anti-P-S1029 antibody and an antibody that recognizes the C-terminal 15-amino-acid peptide of ARK (anti-ARKc).

Figure 3

ABA response of AR144 of P. patens with a mutation in the catalytic core of the kinase domain of ARK. (A) Comparison of amino acid sequences in the catalytic core of protein kinase A (PKA), cyclin-dependent protein kinase2 (CDK2), p38, casein kinase2 (CK2), and epidermal growth factor receptor (EGFR) (adapted from Nolen et al., 2004) with ARK. The conserved Asp (D) has been changed to Asn (N) in AR144. (B) Growth responses to ABA. Protonemata of WT, AR7, and AR144 were grown on a medium with 10-µM ABA for 1 week. (C) Comparison of freezing tolerance. Protonemata incubated with or without 10-µM ABA for 1 d were frozen to −4°C. After thawing, electrolyte leakage was measured to estimate the extent of freezing injury. Error bars indicate standard error of the mean. (D) Transient gene expression assays of WT, AR7, and AR144. Protonemata were bombarded with plasmid constructs of proEm-GUS and proUbi-LUC and were cultured with or without 10-µM ABA for 1 d for GUS and LUC assays. Error bars indicate standard error of the mean. **P < 0.01 in the t test (n = 3) compared with control WT without ABA treatment for (C) and (D). (E) Effect of ABA on Ser1029 phosphorylation in WT, AR7, and AR144. Protonemata were treated with 10-µM ABA for different times and used for immunoblot analysis with anti-phospho-Ser1029 (P-S1029) and anti-ARKc antibodies. Staining of the large subunit of ribulose bisphosphate carboxylase (rbcL) is shown as a control. Results of an in-gel kinase assay for detection of SnRK2 are also shown.

Figure 4

Cold responses mediated by ARK in P. patens. (A) Changes in freezing tolerance during cold acclimation in WT and AR7 (ark^S532F). Protonemata that had been cold-acclimated for the indicated days were frozen to −3°C and injury rate was estimated by measurement of electrolyte leakage after thawing. Error bars indicate standard error of the mean. *P < 0.05 in the t test (n = 3) compared with non-acclimated protonemata of each line. (B) Immunoblot analysis with anti-phospho-Ser1029 (P-S1029) and anti-ARKc antibodies. Coomassie Brilliant Blue staining of the large subunit of ribulose bisphosphate carboxylase (rbcL), results of immunoblot analysis for cold-induced 17B9 protein, and SnRK2 activity analyzed by an in-gel kinase assay are also shown.

Figure 5

ABA response of the ppabi1 line of P. patens. (A) Growth responses to ABA in protonemata of WT and ppabi1. The protonemata were grown on medium with or without 10-µM ABA for 1 week. (B) Effect of ABA treatment on Ser-1029 phosphorylation in WT and ppabi1. Protonemata were treated with 10-µM ABA for different times and used for immunoblot analysis with anti-ARKc and anti-phospho-Ser1029 (P-S1029) antibodies. Coomassie Brilliant Blue staining of the large subunit of ribulose bisphosphate carboxylase (rbcL) and SnRK2 activity analyzed by an in-gel kinase are also shown.

Figure 6

ABA response of ppabi1 ark lines of P. patens. (A) Positions of mutations found in the ABA-insensitive ppuv4 line isolated by ultraviolet mutagenesis of ppabi1. (B) Effects of ABA on growth of WT, ppabi1, AR7, and ppuv4. (C) ABA-induced accumulation of boiling-soluble LEA-like proteins in WT, ppabi1, AR7, and ppuv4. Total soluble proteins of these lines were boiled for 1 min and centrifuged, and the boiling-soluble fraction (BSF) in the supernatant was analyzed by SDS-PAGE. The proteins were either stained with Coomassie Brilliant Blue or used for immunoblot analysis using an antibody against the LEA-like 17B9 protein. Staining of the large subunit of ribulose bisphosphate carboxylase (rbcL) of the total soluble proteins is shown as a control. (D) SnRK2 activity in WT and the mutant lines analyzed by in-gel kinase assays using histone IIIS as a substrate. (E) Freezing tolerance tests of WT and mutant lines. Protonemata were treated with or without 10-µM ABA for 1 d and frozen to −4°C. Electrolyte leakage (%) was measured after thawing to determine freezing injury. Error bars indicate standard error of the mean. *P < 0.05 in the t test (n = 3) compared with ABA-non-treated WT. (F) Gene expression with or without 10-µM ABA treatment in WT and the mutant lines. Transient gene expression analysis was carried out using the construct of ARK fused to the rice actin promoter (proAct) or that without ARK (Vec). Error bars indicate standard error of the mean. *P < 0.05, **P < 0.01 in the t test (n = 4) compared with the values of WT (Vec, -ABA).

Figure 7

Working models showing ARK-mediated ABA responses in P. patens. ABA stimulates ARK autophosphorylation and inhibits PP2C-A, causing activation of SnRK2 regulating downstream factors, in which phosphorylation of SnRK2 by ARK precedes negative regulation by PP2C-A. ABA-stimulated activation of both ARK and SnRK2 occurs without PP2C-A, suggesting the presence of an unidentified negative regulator (NR) and PP2C-A-independent mechanisms for activation of ARK (dashed arrow). Details of these models are shown in Supplemental Figure S12.