Covalently cyclized agonist and antagonist analogues of bombesin and related peptides

During a search for possible cyclization points in shortened, potent bombesin agonists and antagonists, it was found that the joining of amino acid residues in positions 6 and 14 by various means resulted in retention of significant binding affinity for rat pancreatic acini and murine Swiss 3T3 cells. In one series of analogues, Cys residues in these positions were used for bridging via a disulfide bond. (D)-C-Q-W-A-V-GH-L-C-NH₂ retained significant binding affinity for rat pancreatic acini cells and was a full amylase releasing agonist (EC₅₀ 187 nM). Potency was markedly increased by substituting D-Ala for Gly (EC₅₀ 67 nM compared to 10 nM for its linear counterpart) and was decreased by substituting L-Cys for D-Cys in this analogue (EC₅₀ 214 nM), thus strongly suggesting stabilization of peptide folding by the D residues. Elimination of the COOH-terminal amino acid produces competitive antagonists in the linear analogues; however, (D)-C-Q-W-A-V-G-H-C-NH₂ was devoid of activity. Likewise, cyclization to position 13 with the 14 amino acids intact to give (D)-C-Q-W-A-V-G-H-C-L-NH₂ resulted in an almost inactive peptide. On the other hand, as in the linear series, the reduced peptide bond analogue, (D)-C-Q-W-A-V-(D)-A-H-L-Ψ(CH₂NH)-C-NH₂, was a receptor antagonist (IC₅₀ 5.7 mM), albeit much weaker than the corresponding linear analogues, but with no residual agonist activity. Direct head-to-tail cyclization was also tried. Both cyclo[(o)-F-Q-W-A-V-G-H-L-L] (EC₅₀346 nM) and the shorter cyclo[Q-W-A-V-G-H-L-L] (EC₅₀ 1236 nM) were full agonists. Elimination of the COOH-terminal residue in cyclo[(D)-p-Cl-F-Q-W-A-V-(o)-A-H-L] produced an agonist (EC₅₀ 716 nM) rather than an antagonist. These results provide support for the proposal that both bombesin agonists and antagonists adopt a folded conformation at their receptor (s). Furthermore, the retention of appreciable potencies using several cyclization strategies and chain lengths suggests that further optimization of these structures both in terms of potency and ring size is possible. Since these peptides have increased conformational restriction, they should begin to serve as useful substrates for NMR and molecular modeling studies aimed at comparing the obviously subtle differences between agonist and antagonist structures.