The mechanism of peptide-enhanced neurophysin self-association was investigated to address questions raised by the crystal structure of a neurophysin-dipeptide complex. The dependence on protein concentration of the binding of a broad range of peptides to the principal hormone-binding site confirmed that occupancy of this site alone, and not a site that bridges the monomer-monomer interface, is the trigger for enhanced dimerization. For the binding of most peptides to the principal hormone-binding site on bovine neurophysin I, the affinity of each dimer site was at least 10 times that of monomer under the conditions used. No interactions between the two sites of the dimer were evident. Fluorescence polarization studies of pressure-induced dimer dissociation indicated that the volume change for this reaction was almost 4 times greater in the liganded than in the unliganded state, pointing to a significant alteration of the monomer-monomer interface upon peptide binding. Novel conformational changes in the vicinity of the single neurophysin tyrosine, Tyr-49, induced by pressures lower than required for subunit dissociation, were also observed. The bovine neurophysin I dimer therefore appears to represent an allosteric system in which there is thermodynamic and functional communication between each binding site and the monomer-monomer interface, but no communication across the interface to the binding site of the other subunit. A model for the peptide-enhanced dimerization is proposed in which intersubunit contacts between monomers reduce the large unfavorable free energy associated with binding-induced intrasubunit conformational change. Structural origins of the lack of communication across the interface are suggested on the basis of the low volume change associated with dimerization in the unliganded state and monomer-monomer contacts in the crystal structure. Potential roles for the peptide alpha-amino group and position 2 phenyl ring in triggering conformational change are discussed.

Neurophysin is an allosteric protein in which peptide binding and self-association are positively linked. Reaction of neurophysin with succinic anhydride led to a large decrease in peptide affinity assignable to succinylation of a serine or threonine hydroxyl group. To identify the residue involved, acetimidated protein was reacted with [14C]succinic anhydride and the active and inactive components were separated by affinity chromatography. Performic acid oxidation and tryptic and Asp-N mapping of the two components, followed by automated Edman degradation, allowed identification of the critical residue as Ser-56. This residue is not a direct participant in peptide binding and is distant from the subunit interface of the dimer, but it is immediately adjacent to the site of one of the known mutations associated with familial diabetes insipidus. Examination in solution of the peptide affinity of neurophysin succinylated at Ser-56 indicated a binding affinity approximately 1/20th that of the native protein or of protein succinylated at other residues, and a loss of the normal dependence of binding affinity on protein concentration. Under the same buffer conditions, loss of the concentration dependence of binding, in addition to the previously demonstrated loss of binding affinity, also accompanied excision of residues 1-8, an effect attributed to the loss of binding site residue Arg-8. However, in contrast to the effects of succinylation on native neurophysin, only minor effects of succinylation on the binding affinity of the des-1-8 protein were observed.(ABSTRACT TRUNCATED AT 250 WORDS)

Circular dichroism was used to compare the environment of peptides bound to native and des 1‐8 neurophysin in order to further elucidate the role of the neurophysin 1‐8 sequence in peptide‐binding. A very large positive ellipticity (∼6000 degcm2dmol−1), shown earlier to be induced in tyrosine at position 2 of peptides bound to the native protein, was determined by the present study to be paralleled by similar induced changes in tyrosine at peptide position 1. Deletion of the neurophysin 1‐8 sequence led to loss of half of the induced optical activity at peptide positions 1 and 2 and changes in binding‐induced optical activity in the protein, the latter partially assignable to protein disulfides. In the rnononitrated native and des 1‐8 proteins, the optical activity of neurophysin Tyr‐49, a residue at the peptide‐binding site, was reduced by 80% in complexes of the des 1‐8 protein relative to those of the native protein. The results suggest a role for neurophysin Arg‐8 in modulating the optical activity at the binding site by directly placing a charge proximal to the binding site and/or by altering binding site conformation. The data provide the first unambiguous evidence of a difference in the environment of bound peptide between the native and des 1‐8 proteins.

Binding and spectroscopic properties of ostrich neurophysins were examined with emphasis on the behavior of Tyr‐35, a residue that provides a potential probe of the monomer‐monomer interface and of allosteric interrelationshíps between this region and the binding site. Mesotocin‐associated ostrich neurophysin was found to bind oxytocin and related peptides with affinities comparable to the mammalian proteins, but induced a significantly different optical activity in bound peptides than the mammalian proteins. Gel‐filtration studies indicated higher dimerization constants for the ostrich neurophysins than for the bovine neurophysins. Consistent with this, Tyr‐35 was found to be largely buried, as monitored by tyrosine titration and lack of reactivity towards tetranitromethane under non‐denaturing conditions. Reaction of Tyr‐35 of the mesotocin‐associated protein with tetranitromethane under denaturing conditions, followed by refolding, allowed isolation of an active product with an altered interface region as partially evidenced by its titration properties and consistent with its markedly altered CD spectrum. Comparison of the CD spectra of the modified and native proteins and analysis of pH effects indicated the contribution of Tyr‐35 to an unusual 237 nm band in the mesotocin‐associated protein. Small shifts in the 350 nm CD band of nitrated Tyr‐35 on binding peptide and apparent effects of nitration on the induced optical activity in bound peptide provided evidence of at least weak structural communication between Tyr‐35 and the binding site. However, no significant effect of nitration on binding affinity was observed, suggesting that, in the mesotocin‐associated protein, the region around residue 35 is not a stringent modulator of the thermodynamic behavior of the binding site.