Arabidopsis receptor of activated C kinase1 phosphorylation by WITH NO LYSINE8 KINASE

Abstract

Receptor of activated C kinase1 (RACK1) is a versatile scaffold protein that binds to numerous proteins to regulate diverse cellular pathways in mammals. In Arabidopsis (Arabidopsis thaliana), RACK1 has been shown to regulate plant hormone signaling, stress responses, and multiple processes of growth and development. However, little is known about the molecular mechanism underlying these regulations. Here, we show that an atypical serine (Ser)/threonine (Thr) protein kinase, WITH NO LYSINE8 (WNK8), phosphorylates RACK1. WNK8 physically interacted with and phosphorylated RACK1 proteins at two residues: Ser-122 and Thr-162. Genetic epistasis analysis of rack1 wnk8 double mutants indicated that RACK1 acts downstream of WNK8 in the glucose responsiveness and flowering pathways. The phosphorylation-dead form, RACK1A(S122A/T162A), but not the phosphomimetic form, RACK1A(S122D/T162E), rescued the rack1a null mutant, implying that phosphorylation at Ser-122 and Thr-162 negatively regulates RACK1A function. The transcript of RACK1A(S122D/T162E) accumulated at similar levels as those of RACK1(S122A/T162A). However, although the steady-state level of the RACK1A(S122A/T162A) protein was similar to wild-type RACK1A protein, the RACK1A(S122D/T162E) protein was nearly undetectable, suggesting that phosphorylation affects the stability of RACK1A proteins. Taken together, these results suggest that RACK1 is phosphorylated by WNK8 and that phosphorylation negatively regulates RACK1 function by influencing its protein stability.

Figure 1.

WNK8 interacts with RACK1. A, Y2H assays. Yeasts expressing Gal4-AD and LexA-BD fusion proteins were grown on Synthetic Defined (SD)-Leu-Trp medium to validate efficient transformation or SD-Leu-Trp-His-Ura + 3-amino-1,2,4-triazole (3-AT) selection medium to test interactions of the indicated protein pairs. B, BiFC assay. Carboxyl terminal domain of yellow fluorescent protein (cYFP)-tagged WNK8 was expressed with amino-terminal domain of YFP (nYFP)-tagged RACK1A, RACK1B, or RACK1C in tobacco leaf epidermal cells. mCherry-mitochondrial marker was coinfected as the expression control. Complementation of split YFP and fluorescence of mCherry are shown. cYFP-tagged WNK8 expressed with nYFP-tagged P31 and nYFP-tagged RACK1A, RACK1B, and RACK1C expressed with cYFP-tagged P31 were used as negative controls.

Figure 2.

WNK8 phosphorylates RACK1. A, Recombinant GST or GST-tagged RACK1A, RACK1B, or RACK1C protein was incubated with [γ-³²P]ATP and GST-WNK8 for 6 h at room temperature. Phosphorylated proteins were then separated with SDS-PAGE and detected with a phosphoimage analyzer. Bottom shows Coomassie Blue staining of the gel. B, Quantification of phosphorylation level of RACK1 proteins.

Figure 3.

RACK1 phosphorylation sites. A, Liquid chromatography-tandem mass spectometry (LC-MS/MS) spectra of phosphorylated peptides of RACK1C. GST-RACK1C protein was phosphorylated by WNK8 and digested with trypsin. The digested sample was supplied to LC-MS/MS analysis. Peaks corresponding to phosphorylated RACK1 peptide were obtained from MS/MS analysis. B, Phosphorylated peptides identified by MS. RACK1C phosphorylated by WNK8 was used for MS analysis. Amino acid sequences and positions of two phosphorylated peptides are shown. pS, Phosphorylated Ser; pT, phosphorylated Thr. C, Phosphorylation sites on Arabidopsis RACK1A structure (Protein Data Bank ID code 3DM0; Ullah et al., 2008). Each WD40 repeat is drawn with different colors. Two phosphorylation sites identified by MS are drawn with space fill model. Note that Thr-162 of RACK1A corresponds to Thr-161 of RACK1C.

Figure 4.

Ser-122 is required and sufficient for the phosphorylation of RACK1. A, Recombinant GST-tagged RACK1A and the point mutants were incubated with [γ-³²P]ATP and GST-WNK8 for 6 h at room temperature. Proteins were then separated by SDS-PAGE: S122D, RACK1A^S122D; T162E, RACK1A^T162E; and S122D/T162E, RACK1A^S122D/T162E. Note that the Asp substitution at Ser-122 induced a partial cleavage of RACK1A protein, which resulted in a cleaved protein that was approximately 5 kD smaller than the native RACK1A protein. Neither the intact protein nor the cleaved protein was phosphorylated by WNK8. B, Partial amino acid sequence indicating the positions of Ser-122 and Thr-162. Thr-161 of RACK1C corresponds to Thr-161 of RACK1B and Thr-162 of RACK1A.

Figure 5.

Genetic complementation of rack1a mutants. A, qRT-PCR analysis of RACK1A transcripts. Total RNAs were isolated from rosette leaves of 4-week-old plants. Amplification of ACTIN8 was used as a control. Shown are means of three replicates ± se. SA/TA, RACK1A^S122A/T162A; SD/TE, RACK1A^S122D/T162E. B, Seven-week-old transgenic plants. Plants were grown under a 10-h-light/14-h-dark photoperiod. C, Rosette leaf production, days to flowering, and Glc sensitivity. Shown are means of a minimum 10 plants ± se for rosette leaf production and flowering assays. The percentages of green seedlings were recorded 10 d after seeds had been transferred to germination conditions with 6% Glc. At 1% Glc, the percentages of green seedlings for all genotypes were 100%. Shown are means of three biological replicates ± se. *, Significant difference from rack1a-2 mutant, P < 0.05.

Figure 6.

Immunoblot analysis of RACK1A protein. A, RACK1A protein in wild-type (Col-0) and transgenic plants. Total proteins were extracted from leaves of 6-week-old plants and loaded to an SDS-PAGE gel. Anti-RACK1A peptide antibodies were used for immunoblot analysis. SA/TA, RACK1A^S122A/T162A; SD/TE, RACK1A^S122D/T162E. B, The stability of RACK1A protein. Total proteins were extracted from 1-week-old Arabidopsis seedlings of Col-0 and P_RACK1A::RACK1A^S122A/T162A-YFP transgenic plants (in rack1a-2 mutant background) grown in one-half-strength Murashige and Skoog liquid medium. Anti-RACK1A peptide antibodies were used for immunoblot analysis. The same membrane was blotted with anti-Arabidopsis GTP-binding protein α subunit1 (GPA1) antibodies as a loading control. Lanes 1 and 2, Col-0 without cycloheximide (CHX) treatment; lanes 3 to 5, Col-0 treated with 70 µM CHX for 6 h; lane 6, P_RACK1A::RACK1A^S122A/T162A-YFP without CHX treatment; and lanes 7 to 9, P_RACK1A::RACK1A^S122A/T162A-YFP treated with CHX for 6 h.

Figure 7.

rack1a wnk8 double mutants. A, Five-week-old rackla wnk8 double mutants. Plants were grown under a 14-h-light/10-h-dark photoperiod. B, Glc sensitivity phenotype determined by the percentage of green seedlings. Seeds from the Col-0 wild type and mutants were germinated on one-half-strength Murashige and Skoog medium with 1% or 6% Glc. The percentages of green seedlings were recorded 10 d after seeds had been transferred to germination conditions with 6% Glc. At 1% Glc, the percentages of green seedlings for all genotypes were 100%. Shown are means of three biological replicates ± se. C, Flowering phenotype determined by the number of days to flowering. Shown are means of a minimum of 10 plants ± se. D, Flowering phenotype determined by the number of rosette leaves upon bolting. Shown are means of a minimum of 10 plants ± se. *, Significant difference from Col-0, P < 0.05.