Targeted disruption of the pp60 src (Src) gene has implicated this tyrosine kinase in osteoclast-mediated bone resorption and as a therapeutic target for the treatment of osteoporosis and other bone-related diseases. Herein we describe the discovery of a nonpeptide inhibitor (AP22408) of Src that demonstrates in vivo antiresorptive activity. Based on a cocrystal structure of the noncatalytic Src homology 2 (SH2) domain of Src complexed with citrate [in the phosphotyrosine (pTyr) binding pocket], we designed 3,4-diphosphonophenylalanine (Dpp) as a pTyr mimic. In addition to its design to bind Src SH2, the Dpp moiety exhibits bone-targeting properties that confer osteoclast selectivity, hence minimizing possible undesired effects on other cells that have Src-dependent activities. The chemical structure AP22408 also illustrates a bicyclic template to replace the post-pTyr sequence of cognate Src SH2 phosphopeptides such as Ac-pTyr-Glu-Glu-Ile (1). An x-ray structure of AP22408 complexed with Lck (S164C) SH2 confirmed molecular interactions of both the Dpp and bicyclic template of AP22408 as predicted from molecular modeling. Relative to the cognate phosphopeptide, AP22408 exhibits significantly increased Src SH2 binding affinity (IC50 ‫؍‬ 0.30 M for AP22408 and 5.5 M for 1). Furthermore, AP22408 inhibits rabbit osteoclast-mediated resorption of dentine in a cellular assay, exhibits bone-targeting properties based on a hydroxyapatite adsorption assay, and demonstrates in vivo antiresorptive activity in a parathyroid hormone-induced rat model.

Using structure-based design and protein mutagenesis we have remodeled the FKBP12 ligand binding site to include a sizable, hydrophobic specificity pocket. This mutant (F36V-FKBP) is capable of binding, with low or subnanomolar affinities, novel synthetic ligands possessing designed substituents that sterically prevent binding to the wild-type protein. Using binding and structural analysis of bumped compounds, we show here that the pocket is highly promiscuous-capable of binding a range of hydrophobic alkyl and aryl moieties with comparable affinity. Ligand affinity therefore appears largely insensitive to the degree of occupancy or quality of packing of the pocket. NMR spectroscopic analysis indicates that similar ligands can adopt radically different binding modes, thus complicating the interpretation of structure-activity relationships.

Metallothionein is a cysteine-rich metal-binding protein whose biosynthesis is closely regulated by the level of exposure of an organism to zinc, copper, cadmium, and other metal salts. The metallothionein from Callinectes sapidus is known to bind six divalent metal ions in two separate metal-binding clusters. Heteronuclear 1H-113Cd and homonuclear 1H-1H NMR correlation experiments have been used to establish that the two clusters reside in two distinct protein domains. The three-dimensional solution structure of the metallothionein has been determined using the distance and angle constraints derived from these two-dimensional NMR data sets and a distance geometry/simulated annealing protocol. There are no interdomain short distance (< or = 4.5 A) constraints observed in this protein, enabling the calculation of structures for the N-terminal, beta domain and the C-terminal, alpha domain separately. A total of 18 structures were obtained for each domain. The structures are based on a total of 364 experimental NMR restraints consisting of 277 approximate interproton distance restraints, 12 chi 1 and 51 phi angular restraints, and 24 metal-to-cysteine connectivities obtained from 1H-113Cd correlation experiments. The only element of regular secondary structure in either of the two domains is a short segment of helix in the C-terminal alpha domain between Lys42 and Thr48. The folding of the polypeptide backbone chain in each domain, however, gives rise to several type I beta turns. There are no type II beta turns.

3D solution structural calculations for yeast silver(1)-substituted metallothionein (MT) and native copper(I) MT were completed using experimentally determined NOE and dihedral angle constraints, in conjunction with experimentally derived metal-to-Cys connectivities for AgMT which were assumed identical for CuMT. For the first 40 residues in both structures, the polypeptide backbone wraps around the metal cluster in two large parallel loops separated by a deep cleft containing the metal cluster. Minor differences between the two structures include differences in hydrogen bonds and the orientation of the N-terminus with the overall protein volume conserved to within 6.5%.