Regulatory interaction of PRL1 WD protein with Arabidopsis SNF1-like protein kinases

Abstract

Mutation of the PRL1 gene, encoding a regulatory WD protein, results in glucose hypersensitivity and derepression of glucose-regulated genes in Arabidopsis. The yeast SNF1 protein kinase, a key regulator of glucose signaling, and Arabidopsis SNF1 homologs AKIN10 and AKIN11, which can complement the Deltasnf1 mutation, were found to interact with an N-terminal domain of the PRL1 protein in the two-hybrid system and in vitro. AKIN10 and AKIN11 suppress the yeast Deltasnf4 mutation and interact with the SNF4p-activating subunit of SNF1. PRL1 and SNF4 bind independently to adjacent C-terminal domains of AKIN10 and AKIN11, and these protein interactions are negatively regulated by glucose in yeast. AKIN10 and AKIN11, purified in fusion with glutathione S-transferase, undergo autophosphorylation and phosphorylate a peptide of sucrose phosphate synthase in vitro. The sucrose phosphate synthase-peptide kinase activity of AKIN complexes detected by immunoprecipitation is stimulated by sucrose in light-grown Arabidopsis plants. In comparison with wild type, the activation level of AKIN immunocomplexes is higher in the prl1 mutant, suggesting that PRL1 is a negative regulator of Arabidopsis SNF1 homologs. This conclusion is supported by the observation that PRL1 is an inhibitor of AKIN10 and AKIN11 in vitro.

Figure 1

AKIN10 and AKIN11 complement the snf1Δ10 deficiency, suppress the snf4Δ2 mutation, and interact with SNF4p and PRL1 in yeast. (A) Unlike controls with the empty vector pNEV, yeast strains MCY1846 snf1Δ10 (Left) and MCY1853 snf4Δ2 (Right), carrying pNEVakin10 and pNEVakin11, grew equally well on glucose and glycerol. (B) Interaction of GBD-SNF1p with GAD-PRL1 and GAD-SNF4p. Yeast carrying pAS1snf1 was transformed with pACT2prl1, pACT2prl1_1–620, pACT2snf4, and pACT2 as control and grew on SD medium in the presence of 50 mM 3-aminotriazole. (C) Mapping of PRL1 domain involved in binding of AKIN10 and AKIN11. GBD-AKIN10 and GBD-AKIN11 were combined with GAD-PRL1 and GAD fusions of PRL1 peptides located between positions 1 and 35, 1 and 195, and 37 and 195. LacZ-filter lift assays were performed with yeast colonies grown on SC medium with 2% glucose for 3 days at 30°C. In open bars symbolizing the GAD-PRL1 constructs (open bars), numbers indicate the boundaries of deletions and WD-40 repeats. 5-Bromo-4-chloro-3-indolyl β-d-galactoside staining after 4 h is marked by +++, whereas weaker staining observed after 16 h is labeled by +. (D) Glucose regulation of protein interactions of AKIN10 and AKIN11 with SNF4p and PRL1 in yeast. GBD-AKIN10 and GBD-AKIN11 were combined with either GAD-SNF4 or GAD-PRL1 in yeast Y190 grown either in SC medium with 2% glucose (open bars) or in SC-Gal/Gly/EtOH medium with 0.05% glucose (solid bars). β-Galactosidase activities were determined by using an o-nitrophenylgalactosidase assay (15). As control, interaction of GBD-PRL1 with the PRL1-binding protein GAD-PIPC was assayed similarly. The standard deviation of each assay was less than 10% of the maximum value measured. (E) Interaction of PRL1 with GST-AKIN10 and GST-AKIN11 in vitro. ³⁵S-labeled PRL1 was incubated with immobilized GST, GST-AKIN10, GST-AKIN11, and control glutathione-Sepharose 4B matrix. The supernatant (Left) and matrix-bound (Right) fractions were separated by SDS/PAGE to detect the labeled PRL1 protein by autoradiography.

Figure 2

Amino acid sequence comparison of AKIN10, AKIN11, AKIN12, and NPK5. Arrows mark boundaries of the kinase catalytic domain, and numbers indicate the conserved subdomains (6). An arrow marks sequences that show homology with the SNF4p-binding domain of SNF1p. A conserved threonine residue required for SNF1 kinase activation is framed.

Figure 3

Autophosphorylation, substrate specificity, and inhibition of AKIN10 and AKIN11 by PRL1 in vitro. (A Left) GST-AKIN10 and GST-AKIN11 on glutathione-Sepharose were incubated with [γ-³²P]ATP. After elution and SDS/PAGE separation, the autophosphorylation was detected by autoradiography. (Right) Phosphorylation of the TRX-KD substrate by immobilized GST-AKIN10 and GST-AKIN11 was detected by SDS/PAGE followed by autoradiography. (B) Immobilized GST-AKIN10 (1 μg) was preincubated for 30 min with either 1 μg TRX-PRL1 (●) or 1 μg His₆-thioredoxin (□) protein before the addition of the TRX-KD substrate and [γ-³²P]ATP. Equal amounts of samples withdrawn at different time points were separated by SDS/PAGE, and the phosphorylated TRX-KD bands were excised to measure the incorporated radioactivity. (C) PRL1-inhibition assay with GST-AKIN11 (as described in B). Phosphate incorporation per picomole of TRX-KD substrate is plotted against the reaction time.

Figure 4

Enhancement of AKIN kinase activation by the prl1 mutation and sucrose. (A) Protein extracts were prepared from shoots and roots of wild-type (wt) and prl1 mutant (prl) seedlings grown in the presence of either 0.1% (−) or 3% sucrose (+) in the light. The amount of kinase catalytic subunits was adjusted to equal in the samples by Western blot titration with the anti-NPK5 IgG followed by enhanced chemiluminescence detection. (B and C) Titration of kinase assay with AKIN immunocomplexes. Protein extract (75, 125, and 175 μg) from shoots (in B) and roots (in C) of wild-type (solid bars) and prl1 mutant (open bars) seedlings treated with 0.1 or 3% sucrose were immunoprecipitated with 10 μg anti-NPK5 IgG, bound to protein A-Sepharose, and used in kinase assays to measure AKIN activities. Units show pmol phosphate incorporation × 10⁻⁶ per pmol TRX-KD substrate per min.