Starting from a potent pantolactone ketoamide cathepsin K inhibitor discovered from structural screening, conversion of the lactone scaffold to a pyrrolidine scaffold allowed exploration of the S(3) subsite of cathepsin K. Manipulation of P3 and P1' groups afforded potent inhibitors with drug-like properties.

An orally bioavailable series of ketoamide-based cathepsin K inhibitors with good pharmacokinetic properties has been identified. Starting from a potent inhibitor endowed with poor drug properties, conformational constraint of the P(2)-P(3) linker and modifications to P(1') elements led to an enhancement in potency, solubility, clearance, and bioavailability. These optimized inhibitors attenuated bone resorption in a rat TPTX hypocalcemic bone resorption model.

Conversion of the proline-derived cyanamide lead to an acyclic cyanamide capable of forming an additional hydrogen bond with cathepsin K resulted in a large increase in inhibitory activity. An X-ray structure of a co-crystal of a cyanamide with cathepsin K confirmed the enzyme interaction. Furthermore, a representative acyclic cyanamide inhibitor 6r was able to attenuate bone resorption in the rat calvarial model.

Starting from a PDE IV inhibitor hit derived from high throughput screening of the compound collection, a key pyrrolidine cyanamide pharmacophore was identified. Modifications of the pyrrolidine ring produced enhancements in cathepsin K inhibition. An X-ray co-crystal structure of a cyanamide with cathepsin K confirmed the mode of inhibition.

Starting from potent aldehyde inhibitors with poor drug properties, derivatization to semicarbazones led to the identification of a series of semicarbazone-based cathepsin K inhibitors with greater solubility and better pharmacokinetic profiles than their parent aldehydes. Furthermore, a representative semicarbazone inhibitor attenuated bone resorption in an ex vivo rat calvarial bone resorption model. However, based on enzyme inhibition comparisons at neutral pH, semicarbazone hydrolysis rates, and 13C NMR experiments, these semicarbazones probably function as prodrugs of aldehydes.

The synthesis and biological activity of a series of aldehyde inhibitors of cathepsin K are reported. Exploration of the properties of the S2 and S3 subsites with a series of carbamate derivatized norleucine aldehydes substituted at the P2 and P3 positions afforded analogs with cathepsin K IC50s between 600 nM and 130 pM.

A series of ketoamides were synthesized and evaluated for inhibitory activity against cathepsin K. Exploration of the interactions between achiral P-2 substituents and the cysteine protease based on molecular modelling suggestions resulted in potent cathepsin K inhibitors that demonstrated high selectivity versus cathepsins B, H, and L. Subsequent modifications of the P-3, p(1), and P-1' moieties afforded orally bioavailable inhibitors. (C) 2004 Elsevier Ltd. All rights reserved.

The synthesis and biological activity of a series of aldehyde inhibitors of cathepsin K are reported. Exploration of the properties of the S(1) subsite with a series of alpha-amino aldehyde derivatives substituted at the P(1) position afforded compounds with cathepsin K IC(50)s between 52 microM and 15 nM.

Osteoclast-mediated bone matrix resorption has been attributed to cathepsin K, a cysteine protease of the papain family that is abundantly and selectively expressed in osteoclast. Inhibition of cathepsin K could potentially be an effective method to prevent osteoporosis. Structure-activity studies on a series of reversible ketoamides based inhibitors of cathepsin K have led to identification of potent and selective compounds. Crystallographic studies have given insights into the mode of binding of these inhibitors. A series of ketoamides with varying P1 moieties were first synthesized to find an optimum group that would fit into the S1 subsite of the cysteine protease, cathepsin K. With a desired P1 group in place a variety of heterocyclic analogues in the P' region were synthesized to study their steric and electronic effects. In the process of exploring these P' heterocyclic variations, excellent selectivity was gained over other highly homologous cysteine proteases, including cathepsins L, S, and V. The favorable pharmacokinetic properties of some of these cathepsin K inhibitors in rats make them suitable for evaluation in rodent osteoporosis models. A representative cathepsin K inhibitor was shown to attenuate PTH-stimulated hypercalcemia in the TPTX rat model. These inhibitors provide a viable lead series in the discovery of new therapies for the prevention and treatment of osteoporosis

Mouse melanocortin receptors, MC1-R, MC3-R, MC4-R, and MC5-R, when expressed in HEK293 cells and stimulated with either alpha -melanocyte-stimulating hormone (alpha -MSH) or desacetyl-alpha -MSH, mediate increases in intracellular free calcium concentration ([Ca2+](i)) with EC50 values between 0.3 and 4.3 nM. The increase in [Ca2+](i) is cholera toxin sensitive and pertussis toxin insensitive. The mechanism involves calcium mobilization from intracellular stores without a transient rise in inositol trisphosphate. Mouse agouti protein (55 nM) is a competitive antagonist of alpha -MSH (6-fold) and desacetyl-alpha -MSH (8-fold), coupling the mMC1-R to increased [Ca2+](i). Agouti protein (55 nM) significantly increased the EC50 for alpha -MSH (3-fold), and 550 nM agouti protein significantly increased the EC50 for desacetyl-alpha -MSH (4-fold), coupling the mMC4-R to a rise in [Ca2+](i). However, agouti protein antagonism of the MC4-R may not be competitive since there was a trend for the maximum response to also increase. There was no significant antagonism of the MC3-R and MC5-R by agouti protein (55 nM). Understanding the physiological relevance of the transduction of a calcium signal by melanocortin peptides may be important for future development of therapeutic targeting of the melanocortin receptors.

Stromal-cell-derived factor-1 (SDF-1alpha) is an 8-kDa chemokine that is constitutively expressed in bone-marrow-derived stromal cells and has been identified as a ligand for the CXCR4 receptor. We produced the chemokine recombinantly as methionine-SDF-1alpha in Escherichia coli without the leader peptide sequence. The protein was denatured, refolded, and further purified by reversed-phase HPLC. SDF-1alpha was shown to be >95% pure as judged by SDS-PAGE. The final yield of purified and refolded SDF-1alpha was 1-2 mg per gram of wet cell paste. The refolded protein is a ligand for the CXCR4 receptor and has been used to block HIV-mediated cell fusion and downmodulates the CXCR4 receptor. Our ability to purify hundreds of milligrams of refolded protein allowed us to conduct detailed studies of the biophysical properties of the protein. We have used a combination of biophysical techniques to study the solution properties of SDF-1alpha. The average mass of SDF-1alpha, as determined by static light scattering, gave us the first indications that the chemokine may self-associate. Further investigation with sedimentation velocity ultracentrifugation confirmed the existence of two species. The measured s(20, W) values defined two masses corresponding to monomer and dimer. Finally, sedimentation equilibrium ultracentrifugation and dynamic light scattering yielded a composite value of 150 +/- 30 microM for the dimerization constant. We conclude that SDF-1alpha exists in a monomer-dimer equilibrium.

Receptor activator of NF-kappaB ligand (RANKL) is a type II transmembrane protein found on osteoblasts which functions as a major determinant of osteoclast differentiation and activation. RANKL mediates bone homeostasis through binding to the cognate ligand on osteoclasts, RANK, and a soluble decoy receptor, osteoprotegerin (OPG). We designed a construct encoding the extracellular domain of human RANKL that conformed to reports of native processing. To encourage folding and posttranslational modification of a normally membrane-inserted moiety, we expressed the RANKL truncate as a secreted protein using the signal sequence from OPG in a Trichoplusia ni cell line using a baculovirus expression vector. RANKL was purified by a three-step process including an OPG-Fc affinity column. SDS-PAGE and mass spectral analysis indicated that the protein was >99% pure and glycosylated. Circular dichroism spectra revealed that the protein exhibited structural elements similar to tumor necrosis factor-alpha. By BIAcore analysis, RANKL bound to OPG with an affinity of 6.7 nM. Sedimentation equilibrium analytical ultracentrifugation analyses established that our protein existed as a trimer. We conclude that our expressed human RANKL truncate is folded, is functional, and exhibits self-association consistent with other family members.

Cyclosporin A is a cyclic peptide believed to exist as multiple conformers in aqueous solution. Two major conformations, distinguished by a single cis-trans isomerization and the presence of four either intramolecular or intermolecular hydrogen bonds, have been confirmed depending on whether CsA is characterized in organic solvents or bound in aqueous complex with cyclophilin. The relationship between CsA conformation and its ability to penetrate biological membranes is currently unknown. Using Caco-2 cell monolayers, we documented a remarkable increase (more than 2 orders of magnitude) in the membrane permeation of the peptide as temperature was increased from 5 to 37 degrees C. The solubility of CsA was 72 microM at 5 degrees C, but decreased by more than an order of magnitude at 37 degrees C. Moreover, CsA partitioned into non-hydrogen bond donating solvents linearly as a function of increasing temperature, suggestive of a significant conformational change. However, while NMR spectra of CsA confirmed the previously predicted presence of multiple conformers in aqueous solution, the equilibrium between the two major species was not affected by changes in temperature. These NMR data indicated that the observed temperature-dependent changes in the membrane permeability of CsA do not originate from changes in the peptide backbone conformation. Sedimentation equilibrium analysis revealed that CsA behaves in a highly nonideal manner over the temperature range tested. We interpret this behavior as a change in the hydration state with a smaller (or weaker) hydration shell surrounding the peptide at higher temperatures. Such a change would result in lower peptide desolvation energy, thereby promoting partitioning into cellular membranes. We contend that changes in membrane penetration result from alterations in the hydration state of CsA and are not related to the interconversion of the defined conformations.

Desacetyl-alpha-MSH is more abundant than alpha-MSH in the brain, the fetus, human blood, and amniotic fluid, but there is little information on its ability to interact with melanocortin receptors. The aim of this study is to compare and contrast the ability of desacetyl-alpha-MSH and alpha-MSH to couple melanocortin receptors stably expressed in HEK293 cells, to the protein kinase A (PKA) signaling pathway. Desacetyl-alpha-MSH activated mouse MC1, MC3, MC4 and MC5 receptors with EC50s = 0.13, 0.96, 0.53, and 0.84 nM, and alpha-MSH activated these receptors with EC50s = 0.17, 0.88, 1.05, and 1.34 nM, respectively. Mouse agouti protein competitively antagonized alpha-MSH and desacetyl-alpha-MSH coupling to the MC1-R similarly. In contrast, mouse agouti protein antagonized desacetyl-alpha-MSH much more effectively and potently than alpha-MSH coupling the MC4-R to the PKA signaling pathway. Furthermore, mouse agouti protein (10 nM) significantly reduced (1.4-fold) the maximum response of mMC4-R to desacetyl-alpha-MSH and 100 nM mouse agouti significantly increased (4.8-fold) the EC50. Minimal antagonism of alpha-MSH coupling mMC4-R to the PKA signaling pathway was observed with 10 nM mouse agouti, whereas both 50 and 100 nM mouse agouti appeared to reduce the maximum reponse (1.1- and 1.3-fold, respectively) and increase the EC50 (2.5- and 3.4-fold respectively). Mouse agouti protein did not significantly antagonize either alpha-MSH or desacetyl-alpha-MSH coupling mouse MC3 and MC5 receptors. Understanding the similarities and differences in activation of melanocortin receptors by desacetyl-alpha-MSH and alpha-MSH will contribute to delineating the functional roles for these endogenous melanocortin peptides.

Publisher Summary Nuclear receptors have been historically associated with the steroid hormone receptors, for example, estrogen and glucocorticoid receptors, by virtue of DNA binding domain sequence homology comprising two zinc finger motifs. Many of these are orphan receptors, having no defined ligand. The nuclear receptors present tempting targets in the pursuit of a systems based research approach as so many at present have been cloned. However, when recombinant forms of a receptor are available before its cognate ligand has been identified, confusion arises on weather an orphan receptor is active for use in vitro assays. Researchers now have access to unparalleled amounts of DNA sequence and genetic data. Families of homologous gene products can be studied with the intent of connecting specific proteins to various disease conditions. Specifically, to apply this type of strategy in studies, the problem has been approached with two premises in the chapter. First, the recombinant constructs of orphan nuclear receptors are engineered to contain domains with hypothetical functional homology to receptors with known activities and ligands. In particular, a lot of knowledge has been gained concerning retinoid X receptor a (RXRα) and the domains necessary for DNA binding, retinoid binding, and self/hetero-association. For the purpose of this study, PPARα, PPARδ, PPARγ, RXRα, and LXRα constructs were created to contain the putative ligand binding domains (LBD). The amino acid residues within this conserved contiguous region have been shown to be both necessary and sufficient to demonstrate ligand binding competence for RXRα and other nuclear receptors. Structurally, the LBDs are composed primarily of multiple α-helicies. Second, each nuclear receptor is characterized using a variety of biophysical techniques. This chapter has presented data that suggests that the usefulness of a recombinant protein can be determined before an appropriate ligand is available. These results are an encouraging start in the attempt to predict the binding competency of recombinant orphan nuclear receptors.

The interactions between (R)-rolipram and purified human recombinant low-Km, cAMP-specific phosphodiesterase (HSPDE4B2B) constructs were investigated using biochemical, kinetic, and biophysical approaches. The full-length protein (amino acids 1-564) and an N-terminal truncated protein (amino acids 81-564) exhibited high-affinity (R)-rolipram binding, whereas an N-terminal and C-terminal truncated protein (amino acids 152-528) lacked high-affinity (R)-rolipram binding. The 152-528 and 81-564 proteins had similar Km's and kcat/Km's and differed less than 4-fold compared with the 1-564 protein. (R)-Rolipram inhibition plots were biphasic for the 1-564 and 81-564 proteins and fit to two states, a high-affinity (Ki = 5-10 nM) state and a low-affinity (Ki = 200-400 nM) state, whereas the 152-528 protein fit to a single state (Ki = 350-400 nM). The stoichiometry for high-affinity binding using a filter binding assay was found to be <1 mol of (R)-rolipram per mole of 1-564 or 81-564 protein. Titration microcalorimetric studies revealed both a high-affinity state with a stoichiometry of 0.3 mol of (R)-rolipram per mole of protein and a low-affinity state with a stoichiometry of 0.6 mol of (R)-rolipram per mole of protein for the 81-564 protein. A single low-affinity state with a stoichiometry of 0.9 mol of (R)-rolipram per mole of protein was seen using the 152-528 protein. The data indicate that purified HSPDE4B2B 1-564 and 81-564 proteins contain a single binding site for (R)-rolipram and suggest that the proteins exist in two different states distinguishable by their affinity for (R)-rolipram. Furthermore, the high-affinity binding state of the protein requires amino acid residues at the N-terminus (81-151) of the protein and catalytic domain (152-528), whereas the low-affinity binding state only requires residues in the catalytic domain (152-528). Phosphorylation at residues 487 and 489 of the 81-564 protein does not appear to alter the substrate kinetics or the stoichiometry and binding affinity of (R)-rolipram.

We have overexpressed in a baculovirus expression system, and purified to >95% homogeneity, milligram quantities of a human recombinant rolipram-sensitive cAMP phosphodiesterase, HSPDE4B2B (amino acid residues 81–564). The protein expression levels were approximately 8 mg of HSPDE4B2B (81–564) per liter of Sf9 cells. TheKmof the purified enzyme for cAMP was 4 μmand theKifor the Type 4 phosphodiesterase-specific inhibitor (R)-rolipram was 0.6 μm. The specific activity of the purified protein was 40 μmol/min/mg protein. A nonequilibrium filter binding assay revealed a high-affinity (R)-rolipram binding site on the purified enzyme with aKdof 1.5 nmand a stoichiometry of 0.05–0.3 mol of (R)-rolipram per mol of HSPDE4B2B (81–564). Equilibrium dialysis experiments revealed a single binding constant of 140 nmwith a stoichiometry of 0.75 mol of (R)-rolipram per mol of HSPDE4B2B (81–564). Size exclusion chromatography and analytical ultracentrifugation experiments suggest that the protein exists in multiple association states larger than a monomer. Proteolysis experiments revealed a 43-kDa fragment that contained catalytic and rolipram-inhibitable activities, but the fragment showed no high-affinity (R)-rolipram binding. Based on the proteolytic cleavage studies a 43-kDa protein was constructed, expressed, and purified. This protein, HSPDE4B2B (152–528), hadKmandVmaxsimilar to those of the HSPDE4B2B (81–564) protein, but did not exhibit high-affinity (R)-rolipram binding. The protein did show low-affinity (R)-rolipram binding using the equilibrium binding assay. These results show that a low-affinity binding site for (R)-rolipram is solely contained within the catalytic domain of HSPDE4B2B, whereas high-affinity (R)-rolipram binding requires residues within the catalytic domain and residues flanking N- and/or C-terminal to the catalytic region.

Interleukin-5 (IL-5), a disulfide-linked homodimer, can be induced to fold as a biological active monomer by extending the loop between its third and fourth helices (Dickason, R. R., and Huston, D. P. (1996) Nature 379, 652–655). We have designed eight monomeric IL-5 proteins to optimize biological activity and stability of the monomer. This was achieved by (i) inserting the joining loop at three different positions, (ii) by introducing an additional intramolecular disulfide bridge onto these backbones, and (iii) by creating circular permutations to fix the position of the carboxyl-terminal helix relative to the three other helices. The proteins dimerize with K d values ranging from 20 to 200 μm and are therefore monomeric at the picomolar concentrations where they are biologically active. Introduction of a second disulfide confers increased stability, but this increased rigidity results in lower activity of the protein. Contrary to wild type IL-5, mutation of the βc contact residue on the first helix, Glu12, to Lys, into the circularly permutated constructs, did not abolish TF-1 proliferative and eosinophil activation activities. These results indicate that activation of the IL-5 receptor complex is not mediated solely by Glu12 on the first helix, and alternative mechanisms are discussed.