Lck and >1,300-fold versus Lyn). In addition, these investigators found that several of their less potent Src kinase inhibitors (IC50 values 1-3 |iM range) exhibit even better selectivity profiles than 28.
McMurray, Budde, and coworkers likewise examined the effectiveness of cyclic peptide inhibitors on the Src kinase. These investigators first examined the affinity of a series of cyclic peptide substrates for Src (i.e., the ability of these substrates to block the phosphorylation of poly Glu4Tyr) (McMurray et al. 1998). One of the lead compounds, cyclo[Asp-Asn-Glu-Tyr-Ala-Phe-Phe7-Gln-D-Phe-Pro) displays a Ki of 150 |iM, which is nearly identical to its Michaelis constant as a substrate (140 |iM). Insertion of an arginine residue at position 7 resulted in a dramatic loss in enzyme affinity, whereas a glutamic acid residue at this site is well tolerated. On the basis of this observation, these investigators concluded that residues at this site are positioned within a positively charged region of the enzyme. Indeed, when Phe7 was subsequently replaced with a series of 14 different analogs, the lead inhibitors contained a negatively charged residue at this position [4-car-boxyphenylalanine (Ki=0.85 |iM) and phosphotyrosine (Ki=1.1 |iM)j (Wang et al. 2000). The carboxylphenylalanine-containing cyclic peptide displays an impressive selectivity profile in favor of Src (>100-fold against Yes; >300-fold versus Lck; >1,000-fold versus PKA; 1,200-fold versus FGF receptor; 1,800-fold versus Abl; >2,000-fold versus CSK).
Watterson and his group described a different strategy for topologically biasing an active site-directed peptide (Lukas et al. 1999). Myosin light chain kinase (MLCK) was the target in this particular case. These investigators first identified a nonapeptide sequence, Arg-Lys-Lys-Tyr-Lys-Tyr-Arg-Arg-Lys-amide that exhibits both a remarkable affinity (IC50=50 nM) and selectivity (~4,000-fold versus CaM kinase II) for MLCK. Based on the screening results of closely related peptides, in combination with molecular modeling, it was proposed that the peptide might bind to the active site region in an extended conformation. The structural constraints inherent within 4-aminocyclohexanecarboxylic acid were used to promote this desired conformation by inserting the residue at specific sites within the peptide sequence.
The most potent of these derivatives (29) exhibits an IC50 (40 nM) similar to that of the parent nonapeptide. Although improved potency and selectivity were not observed versus the already formidable peptide parent, the results in this study suggest that conformational constraints could serve as scaffolds upon which an array of functionality can be appended.
Was this article helpful?