Biochemical characterization of the tobacco 42-kD protein kinase activated by osmotic stress

Abstract

In tobacco (Nicotiana tabacum), hyperosmotic stress induces rapid activation of a 42-kD protein kinase, referred to as Nicotiana tabacum osmotic stress-activated protein kinase (NtOSAK). cDNA encoding the kinase was cloned and, based on the predicted amino acid sequence, the enzyme was assigned to the SNF1-related protein kinase type 2 (SnRK2) family. The identity of the enzyme was confirmed by immunoprecipitation of the active kinase from tobacco cells subjected to osmotic stress using antibodies raised against a peptide corresponding to the C-terminal sequence of the kinase predicted from the cloned cDNA. A detailed biochemical characterization of NtOSAK purified from stressed tobacco cells was performed. Our results show that NtOSAK is a calcium-independent Ser/Thr protein kinase. The sequence of putative phosphorylation sites recognized by NtOSAK, predicted by the computer program PREDIKIN, resembled the substrate consensus sequence defined for animal and yeast (Saccharomyces cerevisiae) AMPK/SNF1 kinases. Our experimental data confirmed these results, as various targets for AMPK/SNF1 kinases were also efficiently phosphorylated by NtOSAK. A range of protein kinase inhibitors was tested as potential modulators of NtOSAK, but only staurosporine, a rather nonspecific protein kinase inhibitor, was found to abolish the enzyme activity. In phosphorylation reactions, NtOSAK exhibited a preference for Mg(2+) over Mn(2+) ions and an inability to use GTP instead of ATP as a phosphate donor. The enzyme activity was not modulated by 5'-AMP. To our knowledge, these results represent the first detailed biochemical characterization of a kinase of the SnRK2 family.

Figure 1.

NtOSAK belongs to the SnRK2 subfamily. A, Comparison of the predicted amino acid sequence of NtOSAK with sequences of other protein kinases. The sequences of five protein kinases exhibiting sequence similarity are aligned with that of NtOSAK. References for the sequences of the listed kinases are as follows: CPPK1-protein kinase 1 from Craterostigma plant (P. Heino, M. Nylander, T. Palva, and D. Bartels, unpublished data; Swiss-Prot accession no. O65765); soybean protein kinases SPK-3 (Swiss-Prot accession no. Q43466) and SPK-4 (Swiss-Prot accession no. Q39893; Yoon et al., 1997); and Arabidopsis protein kinases ASK1 (Swiss-Prot accession no. P43291) and ASK2 (Swiss-Prot accession no. P43292; Park et al., 1993). The conserved Ser/Thr protein kinase subdomains (Hanks et al., 1988) are indicated by roman numerals. Gaps are marked with dashes and were introduced to maximize alignment. The peptide sequences obtained by microsequencing are shown above the NtOSAK sequence. The peptide sequences used for designing PCR primers are underlined. B, Phylogenetic tree of plant SNF1-related protein kinases that are closely related to NtOSAK. The tree is a consensus of 100 maximum parsimony “bootstrap” trees. Percentage values on each branch represent the corresponding bootstrap probabilities. The multiple alignment was created using the CLUSTALW computer program (Thompson et al., 1994). Columns with gaps and places where alignment seemed doubtful were excluded. The multiple alignment after edition contains 282 columns. The consensus maximum parsimony tree was calculated using the PHYLIP package (Felsenstein, 1989).

Figure 2.

Antibodies specific to the C terminus of the protein kinase encoded by the cloned cDNA (anti-NtOSAK) immunoprecipitate the 42-kD OSAK. A, Western-blot analysis. Protein extracts from BY-2 cells untreated (lane 1) and treated for 5 min with 250 mm NaCl (lane 2) were probed with anti-NtOSAK antibodies. B, Immunoprecipitation followed by immunocomplex kinase assay. Protein extracts (50 μg) from untreated (lane 1) and from cells subjected for 5 min to 250 mm NaCl (lane 2) were immunoprecipitated with anti-NtOSAK antibodies, and the resulting immunocomplexes were analyzed by in-gel kinase assay with MBP as substrate. Lanes 3 and 4 represent controls where the preimmune serum was used instead of the antibodies for immunoprecipitation; lane 3, protein extracts (50 μg) from untreated, and lane 4, from cells subjected for 5 min to 250 mm NaCl. Molecular mass markers are given in kilodaltons at left.

Figure 3.

Immunological analysis of the 42-kD NtOSAK purified from tobacco cells. Proteins eluted from the MonoQ column were subjected to immunoprecipitation with anti-NtOSAK activity. Activity of the preparation before immunoprecipitation (lane 1) and after immunoprecipitation; immunocomplex (lane 3) and remaining supernatant (lane 2) were analyzed using in-gel kinase assay with MBP. Molecular mass markers are given in kilodaltons at left.

Figure 4.

Dependence of Mg²⁺ and Mn²⁺ concentration on casein phosphorylation by NtOSAK. Phosphorylation of casein was performed at various Mg²⁺ (as MgCl₂) and Mn²⁺ (as MnCl₂) concentrations as indicated.

Figure 5.

Effect of temperature on the phosphorylation rate catalyzed by NtOSAK. Phosphorylation of casein was performed at each temperature indicated. Results shown are the average of three independent assays plotting the initial rates against incubation temperature.

Figure 6.

Inhibition of NtOSAK activity by staurosporine. The effect of staurosporine was estimated using casein phosphorylation in the presence of various concentrations of the inhibitor, as indicated. The activity measured in the absence of staurosporine was taken as 100%.