Crystal Structure Analysis of Deamino-Oxytocin: Conformational Flexibility and Receptor Binding
THE NEUROHYPOPHYSEAL HORMONE oxytocin (Fig. 1) elicits smooth muscle contraction, causing milk ejection and uterine contractions in mammals. The synthesis of oxytocin (1) led to a systematic study of the relation of primary structure to biological activity, and more recent studies have highlighted the additive nature of modifications that favor certain pharmacological effects. This has resulted in the design of highly selective, long-acting superagonists and antagonists of therapeutic potential (2, 3). One synthetic analogue of particular interest is deamino-oxytocin (1-mercapto-propionateoxytocin), which was the first to be found more active in most tests than the natural hormone (4).
Spectroscopic studies such as nuclear magnetic resonance (NMR), laser Raman, and circular dichroism have shown that oxytocin can exist in several conformations although certain well-defined intramolecular hydrogen bonds characterize most conformers in solution (5, 6). In view of this inherent flexibility, it is necessary to examine the conformation and dynamics of the hormone and its analogues not only in aqueous conditions, but also in other environments that are models for the hormone in its complex with the receptor. Although crystals of oxytocin were first reported in 1952 (7), the crystal structure has proved elusive. In 1965, crystal data for deamino-oxytocin were reported by Low and Chen (8) and an active 6-selenodeamino-oxytocin analogue was later purified and crystallized (9). Little further progress has been reported although other crystals of oxytocin have been described and a structure analysis has been completed for two COOH-terminal peptides (10, 11).
Here we present the three-dimensional structures of two crystal forms of deamino-oxytocin and a related crystal form of 6-selenodeamino-oxytocin defined by x-ray analyses at resolutions between 1.09 and 2.1 . We compare the crystal structures to conformations proposed from NMR and other spectroscopic studies in dimethylsulfoxide (DMSO) and water. We show that even in the crystals the deamino-oxytocin is flexible and that there are at least two well-defined conformers. We suggest that the partially hydrophobic environment in the crystal may have features in common with the receptor, and that the flexibility seen in the deamino analogues may be required for full agonist rather than antagonist activity.
In view of the lack of success encountered earlier (8, 9), we synthesized fresh materials, grew new crystals, and re-collected x-ray data for deamino-oxytocin (wet form, space group P2(1)), air-dried deamino-oxytocin (space group C2) and 6-selenodeamino-oxytocin (space group C2). The similarity in cell parameters of the air-dried C2 crystals contrasts with a large angle change reported for dried deamino-oxytocin crystals previously. The sulfur positions in the C2 crystal form were located with the use of selenium isomorphous difference, selenium anomalous difference, and sulfur anomalous difference measurements. The selenium single isomorphous replacement with anomalous differences (SIRAS) technique alone did not provide an interpretable map; but, when supplemented with higher quality anomalous differences due to sulfur in the native structure, a best-phased Fourier map at 2.1 resolution provided a rough model of the structure. Restrained refinement (12) with anisotropic thermal ellipsoids for nonhydrogens and...