Protein kinase casein kinase CK comprises a
Protein kinase casein kinase 1 (CK1) comprises a family of highly related, constitutively active serine/threonine protein kinases (reviewed by ). CK1 is involved in controlling a wide variety of different cellular events including protein turnover , , nuclear import  and the cellular response to DNA damage , . CK1 phosphorylates substrates which are generally acidic on the N-terminal side of the target residue and can utilise phosphorylated NS3694 synthesis as recognition determinants , . Murine p53 is phosphorylated at serines 4, 6 and 9 by protein kinase CK1 in vitro and in cultured cells , , but it has not yet been determined whether CK1 can phosphorylate p53 from other species. The involvement of CK1 in cellular events which influence p53 function suggests that phosphorylation of p53 by CK1 may be an important regulatory route by which p53 can sense changes in the environment.
In order to explore the phosphorylation of human p53 by CK1, we have utilised a series of well-characterised GST-p53 fusion proteins as substrates. The spacing between serine 15 and threonine 18 suggested that phosphorylation of serine 15 might generate a recognition determinant through which threonine 18 could become a CK1 substrate , , the phosphorylation of which would be regulated through the serine 15 modification. In the present paper, we show that, in vitro, this is indeed the case. The data therefore provide a potential mechanism by which DNA damage-induced phosphorylation of serine 15 may nucleate additional and physiologically important modification of the p53 protein.
Materials and methods
Results A series of GST-human p53 fusion proteins (comprising the first 42 amino acids of p53), in which known or potential phosphorylation sites were substituted with alanine residues, was used to investigate the phosphorylation of the N-terminus by DNA-PK and CK1. Fig. 1A shows that most of these proteins were substrates for DNA-PK. We have shown previously that the upper band in this experiment contains 2 mol phosphate per mol of protein while the lower band represents 1 mol phosphate per mol of protein . Mutation of serine 15 or 37 led to loss of the upper band, confirming that these residues are DNA-PK target residues in vitro. When both serines 15 and 37 were mutated, the fusion protein was only a very poor DNA-PK substrate. GST was not phosphorylated at all by DNA-PK. These data confirm that DNA-PK phosphorylates serines 15 and 37 in human p53 , . The human p53 proteins were also phosphorylated in vitro by CK1 (Fig. 1B). Substitution of serine 37 reduced the level of phosphorylation while the S20A mutant could not be phosphorylated by CK1, suggesting that these residues are either in vitro CK1 targets or are required for interaction with the protein kinase. Strikingly, serines 6 and 9 of human p53, residues which are key targets of CK1 in murine p53, are not phosphorylated by CK1. GST-murine p53 proteins were also examined as controls. As expected, the 267 protein (comprising amino acids 1–64 of murine p53) was phosphorylated by CK1 whereas the 380 protein (in which serines 4, 6 and 9 are substituted by alanine) was not a CK1 substrate. In general, the human p53 proteins were much weaker substrates than the murine p53 protein. To determine whether serine 15 phosphorylation could influence the ability of CK1 to phosphorylate p53, the GST-p53 fusion proteins (both murine and human) were phosphorylated stoichiometrically using recombinant DNA-PK and excess unlabelled ATP as phosphate donor. The proteins were bound to glutathione-Sepharose beads and washed prior to the addition of CK1 and [γ-32P]ATP. Under the conditions used, the rate of murine p53 phosphorylation by CK1 was stimulated by up to 10-fold following prior phosphorylation of serine 15 (Fig. 2A). CK1-mediated phosphorylation of the 380 protein was also observed, only after prior phosphorylation of serine 15, indicating that the residue(s) phosphorylated by CK1 was neither serine 4, 6 or 9. WT human p53 and the panel of phosphorylation site mutants were similarly examined (Fig. 2B). p53 which had previously been phosphorylated at serine 15 (hereafter designated p53-15P) was phosphorylated by CK1 at least 50-fold more effectively as compared with the unphosphorylated protein. The GST moiety was not phosphorylated by CK1 under these conditions. The role of serine 15 in this process was underscored by the observation that CK1-dependent phosphorylation of the 15A and 15A/37A mutants, but not the 37A alone mutant, was barely detectable and was only slightly higher as compared with the mock DNA-PK-phosphorylated protein. After phosphorylation by DNA-PK, there were some differences in the levels of phosphorylation of the WT and several of the mutant proteins. However, these differences could also be discerned at the basal level (i.e. when not phosphorylated by DNA-PK) and may reflect conformational differences or a partial requirement of specific serine residues to permit recognition or binding to the kinase. A convenient approach for determining the effect of serine 15 phosphorylation on subsequent phosphorylation by CK1 was to measure the ratio of CK1-mediated phosphorylation of p53-15P compared to the mock-phosphorylated proteins. These ratios are expressed as ‘fold increase’ under the autoradiograph in Fig. 2B. These data revealed that the 6A/9A, 37A and 20A mutants showed increases comparable to WT p53 in their ability to be phosphorylated by CK1 following prior phosphorylation of serine 15. Strikingly, however, the ability of DNA-PK phosphorylation to stimulate CK1-mediated phosphorylation of the 18A mutant was reduced by approximately 7-fold. This is consistent with phosphorylation of threonine 18 in p53-15P by CK1. To determine whether threonine 18 was indeed phosphorylated under these conditions, phosphoamino acid analysis of the GST-WT human p53 protein was carried out. The data (Fig. 2C) indicated that most of the radioactivity in the protein was present as phosphothreonine. There is only one threonine residue in the N-terminal 42 amino acids of p53 and this is located at position 18 (Fig. 4).