Examination of structural characteristics of the potent oxytocin antagonists (dPen1 Pen6)-OT and (dPen1 Pen6, 5-tBuPro7)-OT by NMR, raman, CD spectroscopy and molecular modeling

The synthesis and biological evaluation of perricillamine(6)-5-tert-butylproline(7)-oxytocin analogs and comparison with their proline(7)-oxytocin counterparts has led to the discovery of two potent oxytocin (OT) antagonists: [dPen(1),Pen(6)]-oxytocin (1, pA(2) = 8.22, EC50 - 6.0 nm) and [dPen(1),Pen(6),5-tBuPro(7)]-oxytocin (2, pA(2) = 8.19, EC50 = 6.5 nM). In an attempt to understand the conformational requirements for their biological activity, spectroscopic analyses of 1 and 2 were perfonned using H-1 NMR, laser Raman and CD techniques. In H2O, oxytocin analogs 1 and 2 exhibited cis-isomer populations of 7% and 35%, respectively. Measurement of the amide proton temperature coefficients revealed solvent shielded hydrogens for Gin(4) and Pen6 in the major trans-conformer of 1 as well as for Gln(4) in the minor cis-conformer of 2. Few long-distance NOES were observed, suggesting conformational averaging for analogs 1 and 2 in water; moreover, a lower barrier (16.6 +/- 0.2 kcal/mol) for isomerization of the amide N-terminal to 5-tBuPro(7) relative to OT was calculated from measuring the coalescence temperature of the GlY(9) backbone Nit signals in the NMR spectra of 2. Observed bands in the Raman spectra of 1 and 2 correspond to C-beta-S-S-C-beta dihedral angles of +110-115 degrees and 90 degrees, respectively. In water, acetonitrile and methanol, the CD spectra for 1 xhibited a positive maximum around 236-239 nm; in trifluoroethanol, the spectra shifted and a negative maximum was observed at 240 nm. The CID spectra of 2 were unaffected by solvent changes and exhibited a negative maximum at 236-239 nm. The CID and Raman data both suggested that a conformation having a right-handed screw sense about the disuffide and a chi(CS-SC) dihedral angle value close to 115 degrees was favored for analog 1 in water, methanol and acetonitrile, but not trifluoroethanol, where a +/- 90 degrees angle was favored. Analog 2 was more resilient to conformational change about the disuffide, and adopted a preferred disuffide geometry corresponding to a +/- 90 degrees chi(CS-SC) dihedral angle. Monte Carlo conformational analysis of analogs 1 and 2 using distance restraints derived from NMR spectroscopy revealed two prominent conformational minima for analog 1 with disultide geometries around + 114 degrees and + 116 degrees. Similar analysis of analog 2 revealed one conformational minimum with a disuffide geometry around + 104'. In sum, the conformation about the disuffide in [dPen(1),Pen(6)]-OT (1) was shown to be contingent on environment and in TFE, adopted a geometry similar to that of [dPen(1), Pen(6),5- tBuPro(7)]-OT (2) which appeared to be stabilized by hydrophobic interactions between the 5-tBuPro(7) (5R)-tert-butyl group, the Len(8) isopropyl sidechain and the Pen(6)beta-methyl substituents. In light of the conformational rigidity of 2 about the disultide bond, and the similar geometry adopted by 1 in TFE, a S-S dihedral angle close to +110 degrees may be a prerequisite for their binding at the receptor. Copyright (c) 2005 European Peptide Society and John Wiley & Sons, Ltd.