ROLE OF PHOSPHORYLATED AMINOACYL RESIDUES IN GENERATING ATYPICAL CONSENSUS SEQUENCES WHICH ARE RECOGNIZED BY CASEIN KINASE-2 BUT NOT BY CASEIN KINASE-1

Casein kinase-2 (CK-2) is a ubiquitous Ser/Thr specific protein kinase that recognizes phosphorylatable residues located upstream of acidic determinants, its consensus sequence being Ser(Thr)-Xaa-Xaa-Acidic. Here we show that the phosphotetrapeptide AcSer(P)-Ser(P)-Ser-Ser(P), which is devoid of the canonical consensus sequence, is nevertheless phosphorylated by CK-2 with rates comparable to that of typical peptide substrates Ser-Glu-Glu-Glu-Glu-Glu and Arg-Arg-Arg-Glu-Glu-Glu-Thr-Glu-Glu-Glu routinely employed for assaying CK-2 activity. The phosphopeptide AcSer(P)-Ser-Ser(P) [but not Ac-Ser-Ser(P)-Ser(P) or AcSer(P)-Ser(P)-Ser] is also phosphorylated albeit less efficiently than AcSer(P)Ser( P)-Ser-Ser(P). Further N-terminal elongation with additional phosphoseryl residues to give the peptides AcSer(P)-Ser(P)-Ser(P)-Ser-Ser(P) and AcSer(P)-Ser(P)-Ser(P)-Ser(P)-Ser-Ser(P) does not improve but rather slightly decreases the phosphorylation efficiency by CK-2. These two peptides are conversely excellent substrates for CK-1, which does not appreciably phosphorylate either AcSer(P)-Ser-Ser(P) or AcSer-(P)-Ser(P)-Ser-Ser(P). Either individual or multiple replacement of the phosphorylated residues with glutamic acid in the peptide AcSer(P)-Ser(P)-Ser-Ser(P) drastically reduces the phosphorylation efficiency by CK-2, the phosphoseryl residue at position -2 playing an especially crucial role which cannot be surrogated by glutamyl residues. Phosphoserine is also more effective than glutamic acid at positions +1 and +2 of peptides that fulfill the consensus sequence Ser-Xaa-Xaa-Glu, the K(cat) value of Ser-Ser(P)-Ser(P)-Glu-Glu being 3-fold higher than that of Ser-Glu-Glu-Glu-Glu which, in tum, is a better substrate than Ser-Ala-Ala-Glu-Glu. Such a superiority of phosphoserine over glutamic acid is conversely not so evident at the canonical +3 position, Ser-Ala-Ala-Glu-Glu and Ser-Ala-Ala-Ser(P)-Ser(P) being comparable substrates. These data suggest that partial phosphorylation of clustered hydroxylic residues might generate phosphoacceptor sites for CK-2 which are not predictable on the basis of its canonical consensus sequence. In particular, a very effective atypical consensus sequence is Ser(P)-Ser(P)-Ser-Ser(P). Such a sequence is not recognized by casein kinase-1 (CK-1) whose specificity is determined by phosphorylated residue(s) located at more remote positions on the N-terminal side, notably -3 and -4.