Evaluation of in vitro PXR-based assays and in silico modeling approaches for understanding the binding of a structurally diverse set of drugs to PXR.

The pregnane X-receptor (PXR) is a promiscuous nuclear receptor primarily responsible for the induction of genes from the cytochrome P450 3A family. In this study, we used a previously described PXR/SRC tethered protein to establish two in vitro assays for identifying PXR ligands: automated ligand identification system (ALIS) and temperature-dependent circular dichroism (TdCD). Kd values determined by ALIS and TdCD showed good correlations with the EC50 values determined by a PXR luciferase reporter-gene assay for 37 marketed drugs. The same set of compounds was modeled into the PXR ligand-binding domain that takes into consideration the structural variations of five published X-ray structures of PXR-ligand complexes. Major findings from our in silico analysis are as follows. First, the primary determinants for non-binders of PXR are molecular size and shape of the compounds. Low molecular weight (MW<300) compounds were in general found to be non-binders, and those molecules that do not match the shape of the PXR ligand-binding site may also act as a non-binder. Secondly, the favorable hydrophobic interactions, mostly through aromatic π-π interactions, and the presence of suitable hydrogen bond(s) between the compounds and PXR are attributes of strong binders. Thirdly, the structures of the PXR binding domain possess the flexibility that accommodates structurally diverse compounds, while some of the strong binders may also adapt flexible conformations for fitting into the binding site. The results from this study provide a molecular basis for future efforts in reducing/abolishing the PXR-dependent CYP3A4 induction liability.

Construction and characterization of a fully active PXR/SRC-1 tethered protein with increased stability.

The nuclear xenobiotic receptor PXR is a ligand-inducible transcription factor regulating drug-metabolizing enzymes and transporters and a master switch mediating potentially adverse drug-drug interactions. In addition to binding a coactivator protein such as SRC-1, the C-terminal ligand-binding domain (LBD) is solely responsible for ligand recognition and thus the ligand-dependent downstream effects. In an effort to facilitate structural studies of PXR to understand and abolish the interactions between PXR and its ligands, several recombinant PXR/SRC-1 constructs were designed and evaluated for expression, stability and activity. Expression strategies employing either dual expression or translationally coupled bicistronic expression were found to be unsuitable for producing stable PXR in a stochiometric complex with a peptide derived from SRC-1 (SRC-1p). A single polypeptide chain encompassing PXR and SRC-1p tethered with a peptidyl linker was designed to promote intramolecular complex formation. This tethered protein was overexpressed as a soluble protein and required no additional SRC-1p for further stabilization. X-ray crystal structures in the presence and absence of the known PXR agonist SR-12813 were determined to high resolution. In addition, a circular dichroism-based binding assay was developed to allow rapid evaluation of PXR ligand affinity, making this tethered protein a convenient and effective reagent for the rational attenuation of drug-induced PXR-mediated metabolism.

Key steps in the structure-based optimization of the hepatitis C virus NS3/4A protease inhibitor SCH503034.

The structures of both native and S139A holo-HCV NS3/4A protease domain were solved to high resolution. Subsequently, structures were determined for a series of ketoamide inhibitors in complex with the protease. The changes in the inhibitor potency were correlated with changes in the buried surface area upon binding the inhibitor to the active site. The largest contributions to the binding energy arise from the hydrophobic interactions of the P1 and P2 groups as they bind to the S1 and S2 pockets. This correlation of the changes in potency with increased buried surface area contributed directly to the design of a potent tripeptide inhibitor of the HCV NS3/4A protease, which is currently in clinical trials.

Discovery of the HCV NS3/4A protease inhibitor (1R,5S)-N-[3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]-3- [2(S)-[[[(1,1-dimethylethyl)amino]carbonyl]amino]-3,3-dimethyl-1-oxobutyl]- 6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)-carboxamide (Sch 503034) II. Key steps in structure-based optimization.

The structures of both the native holo-HCV NS3/4A protease domain and the protease domain with a serine 139 to alanine (S139A) mutation were solved to high resolution. Subsequently, structures were determined for a series of ketoamide inhibitors in complex with the protease. The changes in the inhibitor potency were correlated with changes in the buried surface area upon binding the inhibitor to the active site. The largest contribution to the binding energy arises from the hydrophobic interactions of the P1 and P2 groups as they bind to the S1 and S2 pockets [the numbering of the subsites is as defined in Berger, A.; Schechter, I. Philos. Trans. R. Soc. London, Ser. B 1970, 257, 249-264]. This correlation of the changes in potency with increased buried surface area contributed directly to the design of a potent tripeptide inhibitor of the HCV NS3/4A protease that is currently in clinical trials.

Protease domain of human ADAM33 produced by Drosophila S2 cells.

Human ADAM33 is a multiple-domain, type-I transmembrane zinc metalloprotease recently implicated in asthma susceptibility [Nature 418 (2002) 426]. To provide an active protease for functional studies, expression of a recombinant ADAM33 zymogen (pro-catalytic domains, pro-CAT) was attempted in several insect cells. The pro-CAT was cloned into baculovirus under the regulation of the polyhedron promoter and using either the honeybee mellitin or ADAM33 signal sequence. Sf9 or Hi5 cells infected with these recombinant viruses expressed the majority of the protein unprocessed and as inclusion bodies ( approximately 10 mg/L). On the other hand, similar constructs could be expressed, processed, and secreted by Drosophila S2 cells using a variety of constitutive (actin, pAc5.1) or inducible (metallothionein, PMT) promoters and leader sequences (e.g., native and BiP). Higher expression level of 10-fold was observed for the inducible system resulting in an average yield of 20 mg/L after purification. The majority of the catalytic domain purified from the Drosophila conditioned media remained associated with the pro-domain after several chromatography steps. An induction cocktail containing cadmium chloride and zinc chloride was subsequently developed for the PMT system as an alternative to using cupric sulfate or cadmium chloride as single inducers. The novel induction cocktail resulted in an increased ratio of secreted catalytic to pro-domain, and yielded milligram amounts of highly purified protease. The availability of this modified expression system facilitated purification of the wild type and several glycosylation mutants, one of which (N231Q) crystallized recently for X-ray structure determination [J. Mol. Biol. 335 (2003) 129].

Crystallization of glycosylated human BACE protease domain expressed in Trichoplusia ni.

Human beta-amyloid precursor protein cleaving enzyme (beta-secretase, or BACE) belongs to the aspartyl protease family, and is responsible for generating the N-terminus of beta-amyloid peptide (Abeta). BACE is a type I transmembrane glycoprotein with pre-, pro- and catalytic domains, a short transmembrane helix and a cytoplasmic region. In this study, a truncated form was engineered to produce the authentic catalytic domain of BACE in Trichoplusia ni (High 5) cells. The glycosylated BACE zymogen (proBACE) was secreted into the conditioned medium for facile purification by metal chelate and gel filtration chromatographies. The mature catalytic domain was obtained by a trans cleavage event under acidic conditions and crystallized in the absence of a bound inhibitor. A complete 3.4 A data set was collected on a single orthorhombic crystal with unit cell parameters a=74 A, b=130 A, c=134A. Successful molecular replacement shows two BACE molecules in the asymmetric unit.

Crystal structure of the catalytic domain of human ADAM33.

Adam33 is a putative asthma susceptibility gene encoding for a membrane-anchored metalloprotease belonging to the ADAM family. The ADAMs (a disintegrin and metalloprotease) are a family of glycoproteins implicated in cell-cell interactions, cell fusion, and cell signaling. We have determined the crystal structure of the Adam33 catalytic domain in complex with the inhibitor marimastat and the inhibitor-free form. The structures reveal the polypeptide fold and active site environment resembling that of other metalloproteases. The substrate-binding site contains unique features that allow the structure-based design of specific inhibitors of this enzyme.