Corticotropin releasing factor (CRF) agonists with reduced amide bonds andSer(7) substitutions

Citation
L. Cervini et al., Corticotropin releasing factor (CRF) agonists with reduced amide bonds andSer(7) substitutions, J MED CHEM, 42(4), 1999, pp. 761-768
Citations number
51
Categorie Soggetti
Chemistry & Analysis
Journal title
JOURNAL OF MEDICINAL CHEMISTRY
ISSN journal
00222623 → ACNP
Volume
42
Issue
4
Year of publication
1999
Pages
761 - 768
Database
ISI
SICI code
0022-2623(19990225)42:4<761:CRF(AW>2.0.ZU;2-R
Abstract
Strategies to generate competitive antagonists of bioactive peptides includ e several possible structural modifications such as the introduction of D-r esidues and of reduced amide bonds, the substitution of amino acid side cha ins, dimerization of fragments, and deletion of part of the sequence, among others. Whereas we have identified the two most likely residues responsibl e for receptor activation in corticotropin releasing factor (CRF) (Ser(7) a nd Leu(8))(1) and generated potent antagonists by deleting residues 1-8,(2, 3) the question remained as to whether we could generate CRF antagonists wi th enhanced affinity after reduction of amide bonds at the N-terminus of CR F or through subtle alteration of those residues' side chains. Reduced amid e bond replacements (psi[CH2NH]) between residues 6-9 in oCRF((5-41)) (11, 12, 15) analogues consistently yielded potencies of <1% that of oCRF. Excep t for the (10)psi(11) and (12)psi(13) analogues 19 and 20, reduced amide bo nd replacements were generally well-tolerated in the longer hCRF((4-41)) an alogues, with the (7)psi(8)-, (8)psi(9)-, and (9)psi(10)-modified peptides (13, 14, 18) yielding potencies that were 2-4 times that of hCRF. Although somewhat promising as agonists, they were, however, 3-7 times less potent t han the parent [D-pro(4)Nle(21,38)]-hCRF((4-41)) (2) Since O-alkylation of Tyr(3) in vasopressin yields an antagonist, and since Ser(7) is one of the eight fully conserved residues in the CRF family (inclusive of sauvagine, u rocortins, and urotensins) and likely to be critical for receptor binding, we synthesized cyclo(30-33)[Ser(OMe)(7),D-Phe(12),Nle(21),GlU(30),Lys(33),N le(38)]Ac-hCRF((7-41)) (22), which was found to exhibit full efficacy and 4 0% of the potency of cyclo(30-33)[D-Phe(12),Nle(21),Glu(30),Lys(33),Nle(38) ]Ac-hCRF((7-41)) (5). Other substitutions at position 7 included aminoglyci ne (23, 24) and alkylated and/or acylated [alpha or alpha'-methyl (25-28), alpha'-formyl (29, 30), alpha'-formyl, alpha'-methyl(31), alpha'-acetyl (32 ), alpha'-acetyl, alpha'-methyl(33)], D- or L-aminoglycines. All analogues were active although less potent than the parent compound 2, and all elicit ed maximal ACTH response as compared to hCRF. The most potent analogue in t his series (33) had the bulkiest side chain, Agl(Me, Ac), and was 60% and 8 0% as potent as the Ser(7) analogue 5 and the Ala(7) analogue 6, respective ly. In conclusion, we found that neither reduction of the individual amide linkages between residues 6-11 and 12-13 nor introduction of a carbamide mo iety in lieu of the side chain of Ser(7) led to CRF antagonists.