SOLID-PHASE SYNTHESIS OF ADENOSINE PHOSPHOPEPTIDES AS POTENTIAL BISUBSTRATE INHIBITORS OF PROTEIN-KINASES

A protocol has been developed far the synthesis of bisubstrate analogs of protein kinases containing adenosine 5'-phosphates (AP(n), n = 2-4 ) and substrate peptides attached covalently through the hydroxyl grou p of serine. The procedure involves the coupling of the terminal phosp hate group of the phosphoadenosine to phosphorylated substrate peptide (s) after activation with 1,1'-carbonyldiimidazole. The peptide substr ates were synthesized by standard Fmoc chemistry and phosphorylated at the target serine using either di-tert-butyl or diallyl N,N-diisoprop ylphosphoramidite, followed by oxidation with tert-butyl hydroperoxide . Allyl protection was preferred in the solid phase strategy since the se groups could be selectively removed and the activated AP(n) moietie s coupled to the phosphopeptides while still bound to the resin for a yield of approximate to 40-80%. A less satisfactory approach was initi ally attempted using di-tert-butyl N,N-diisopropylphosphoramidite as t he phosphite transfer reagent to yield free phosphopeptides which were subsequently coupled in a solution phase reaction (less than or equal to 2-3% yield). The high efficiency of the solid phase method affords a general approach for synthesizing peptide-nucleoside analogs as wel l as potentially other types of covalent peptide-ligand complexes thro ugh the formation of phosphodiester linkages. In this study, a series of bisubstrate inhibitor candidates were synthesized using this new ap proach based on kemptide (LRRASLG), a substrate of cAMP dependent prot ein kinase (cAPK), and several extended analogs. The bisubstrate analo gs containing ADP, ATP, and AP(4) were tested as inhibitors of the cat alytic subunit of cAPK. IC50 values were determined and the best inhib itor in this series was the adenosine 5'-tetraphosphate peptide kempti de-AP(4) (IC50 = 68 mu M). The efficiency of inhibition was found to d ecrease rapidly with shorter phosphate chains, yielding IC50 values of 226 and 935 mu M for kemptide-ATP and kemptide-ADP, respectively. All three inhibitors in this series acted competitively with respect to A TP but not with respect to the peptide substrate, suggesting that larg er peptides may be required to generate bisubstrate inhibitors for cAP K.