Sequence specificity of C-terminal Src kinase (CSK) - A comparison with Src-related kinases c-Fgr and Lyn

An eicosapeptide encompassing the C-terminal tail of c-Src (Tyr527) which is conserved in most Src-related protein kinases, is phosphorylated by C-terminal Src kinase (CSK) and by the two Src-related protein kinases c-Fgr and Lyn, with similar kinetic constants. Two related peptides reproducing the C-terminal segments of c-Src mutants defective in CSK phosphorylation [MacAuley, A., Okada, M., Nada, S., Nakagawa, H. & Cooper, J. A. (1993) Oncogene 8, 117-124] AFLED<(SC)under bar>T (G) under bar TEP (L) under bar YQ (R) under bar GENL (mutant number 28) and AFLED (N) under bar FT (G) under bar T (K) under bar PQY (H) under bar GENL (mutant number 29), proved a better and a much worse substrates, respectively than the wild-type peptide, with either CSK or the two Src kinases. By changing individual residues in the best peptide substrate, it was shown that the main element responsible for its improved phosphorylation is leucine at position -1 (instead of glutamine), while lysine at position -3 (instead of glutamate) has a detrimental effect, possibly accounting for the negligible phosphorylation of peptide derived from mutant number 29. By contrast to most peptide substrates, including the Src C-terminal peptides, which exhibit relatively high K-m values, a polyoma-virus-middle-T-antigen-(mT)-derived peptide with tyrosine embedded in a highly hydrophobic sequence (EEEPQFEEIPIYLELLP) exhibits with CSK a quite low K-m value (63 mu M). Consistent with this, the optimal sequence selected by CSK in an oriented peptide library is XXXIYMFFE. This is different from sequences selected by Lyn (DEEIYEELX) and c-Fgr (XEEIYGIFF), although they all share a high selection for a hydrophobic residue at n-1. In sharp contrast, TPKIIB/p38(syk), related to the catalytic domain of p72(syk), selects acidic residues at nearly all positions, n-1 included. These data support the notion that the features determining the specific phosphorylation of the C-terminal tyrosine residue of Src do not reside in the primary structure surrounding the target tyrosine. They also show that this site does not entirely fulfil the optimal consensus sequence recognized by CSK, disclosing the possibility that as yet unrecognized CSK targets structurally unrelated to the C-terminal tyrosine residue of Src kinases may exist.