Somatostatin Receptor 1 Selective Analogues: 3. Dicyclic Peptides Jean E. Rivier,* ,† Dean A. Kirby, † Judit Erchegyi, † Beatrice Waser, ‡ Ve ´ronique Eltschinger, ‡ Renzo Cescato, ‡ and Jean Claude Reubi ‡ The Clayton Foundation Laboratories for Peptide Biology, The Salk Institute, 10010 N. Torrey Pines Road, La Jolla, California 92037, and Division of Cell Biology and Experimental Cancer Research, Institute of Pathology, University of Berne, Berne, Switzerland Received June 18, 2004 The binding affinity of short chain somatostatin (SRIF) analogues at the five human SRIF receptors (sst) was determined to identify sterically constrained somatostatin receptor subtype 1 (sst 1 ) selective scaffolds. Des-AA 1,2,4,13 -[D-Trp 8 ]SRIF (2) retained high binding affinity at all receptors but sst 1 , Des-AA 1,2,4,5 -[D-Trp 8 ]SRIF (3) at sst 4 and sst 5 , and Des-AA 1,2,4,5,13 -[D-Trp 8 ]- SRIF (4) at sst 2 and sst 4 (AA ) amino acid). Des-AA 1,2,4,12,13 -[D-Trp 8 ]SRIF (6) was potent and sst 4 -selective (>25-fold); Des-AA 1,2,5,12,13 -[D-Trp 8 ]SRIF (7) and Des-AA 1,2,4,5,12,13 -[D-Trp 8 ]-SRIF (9, ODT-8) were most potent at sst 4 and moderately potent at all other receptors. Dicyclic SRIF agonists of the sst 1 -selective Des-AA 1,5 -[Tyr 2 ,D-Trp 8 ,IAmp 9 ]SRIF, (14, sst 1 IC 50 ) 14 nM) were prepared in which a lactam bridge introduced additional conformational constraint (IAmp ) 4-(N-isopropyl)-aminomethylphenylalanine). Cyclo(7-12)Des-AA 1,5 -[Tyr 2 ,Glu 7 ,D-Trp 8 ,IAmp 9 ,- hhLys 12 ]SRIF (31) (sst 1 IC 50 ) 16 nM) and cyclo(7-12) Des-AA 1,2,5 -[Glu 7 ,D-Trp 8 ,IAmp 9 ,m-I- Tyr 11 ,hhLys 12 ]SRIF (45) (sst 1 IC 50 ) 6.1 nM) had equal or improved affinities over that of the parent 14. Binding affinity was decreased in all other cases with alternate bridging constraints such as cyclo (6-11), cyclo (6-12), and cyclo (7-11). Compound 45 is an agonist (EC 50 ) 8.8 nM) in the adenylate cyclase assay. Introduction The cyclic peptide hormone somatostatin (SRIF), first isolated from ovine hypothalamus in 1973, 1 character- ized, 2 and synthesized 3 in our laboratories, has since been found to modulate numerous actions in the body that are mediated by at least five SRIF receptors (sst 1-5 ). 4-10 Among its most important roles, SRIF is an inhibitor of growth hormone (GH), 2 glucagon, and insulin secretions. 11 The aim of this study was to use the available structure-activity relationships (SAR) of sst 1 -selective analogues for the design of sst 1 -selective constrained analogues amenable to NMR investigation for the determination of the sst 1 pharmacophore 12 and to generate compounds with high in vivo potency and either agonist or antagonist activity for mechanistic studies. An updated review of sst 1 receptor localization and function as well as a rationale for the need of sst 1 - selective ligands is given in the preceding paper. 13 The early observation that shortened chain analogues of SRIF, both peptidic as well as nonpeptide mimet- ics, 14-17 retained significant biological activities 18,19 ultimately led to the development of several drugs and drug candidates. These include octreotide 20 and deriva- tives such as Octreoscan, vapreotide, and lanreotide and, more recently, SOM230 21 and KE108. 22 Several reviews describe the use of these analogues in the management of numerous pathological conditions and cancer treatments. 14,17,23-27 Because these early ana- logues had been tested for their relative potencies in inhibiting GH, insulin, and glucagon secretions and shown in some cases to be selective, 19,28-31 we measured their binding affinity for the five sst (Table 1). In addition, highly constrained analogues that are preferred as their solution conformations will help define the structure of the pharmacophore using NMR and computer modeling techniques. 32-39 In our quest for novel scaffolds that would lead to ligands with high- binding affinity and sst selectivity, we synthesized a series of shortened, presumably more structurally stable, SRIF analogues (Table 1). This design strategy for a constrained backbone motif has precedence in the design of ODT-8 (Figure 1B), 18,19 octreotide (Figure 1C), 20 and CH-275 (Figure 1D) 40 as compared to SRIF (schemati- cally depicted in Figure 1A). The ODT-8 scaffold origi- nated from the observation that, in an alanine scan of SRIF, substitutions at positions 4, 5, 10, 12, and 13 resulted in analogues essentially equipotent to SRIF whereas substitutions at positions 6, 7, 8, 9, and 11 resulted in significantly less potent analogues (100-fold loss). 18,19 This suggested that double deletions such as Des-AA 4,13 and Des-AA 4,5,12,13 may conserve the struc- tural features of the postulated -turn responsible for activity. This was indeed the case as shown by us 18,19 and confirmed later by others with the octreotide scaffold (Figure 1C) 20 and the work by the Merck group headed by R. Hirschmann, which made the seminal observation that the π-stacking of the phenylalanine side chains at positions 6 and 11 of SRIF may have a structure-stabilizing role 32,33 critical for the propitious alignment of the side chains of residues 7-10 in a bioactive conformation. The scaffold of CH-275 was first identified by Sarantakis et al. 41 who observed that Des- Ala 1 ,Gly 2 ,Asn 5 -SRIF, {H-cyclo[Cys 3 -Lys 4 -Phe 6 -Phe 7 -Trp 8 - * Author to whom correspondence should be addressed. Phone: (858) 453-4100. Fax: (858) 552-1546. E-mail: Jrivier@salk.edu † The Salk Institute. ‡ University of Berne. 515 J. Med. Chem. 2005, 48, 515-522 10.1021/jm049519m CCC: $30.25 © 2005 American Chemical Society Published on Web 12/23/2004