Chemogenomics strategies for G-protein coupled receptor hit finding.

Targeting protein superfamilies via chemogenomics is based on a similarity clustering of gene sequences and molecular structures of ligands. Both target and ligand clusters are linked by generating binding affinity profiles of chemotypes vs a target panel. The application of this multidimensional similarity paradigm will be described in the context of Lead Generation to identify novel chemical hit classes for G-protein coupled receptors.

Library design practices for success in lead generation with small molecule libraries.

The generation of novel structures amenable to rapid and efficient lead optimization comprises an emerging strategy for success in modern drug discovery. Small molecule libraries of sufficient size and diversity to increase the chances of discovery of novel structures make the high throughput synthesis approach the method of choice for lead generation. Despite an industry trend for smaller, more focused libraries, the need to generate novel lead structures makes larger libraries a necessary strategy. For libraries of a several thousand or more members, solid phase synthesis approaches are the most suitable. While the technology and chemistry necessary for small molecule library synthesis continue to advance, success in lead generation requires rigorous consideration in the library design process to ensure the synthesis of molecules possessing the proper characteristics for subsequent lead optimization. Without proper selection of library templates and building blocks, solid phase synthesis methods often generate molecules which are too heavy, too lipophilic and too complex to be useful for lead optimization. The appropriate filtering of virtual library designs with multiple computational tools allows the generation of information-rich libraries within a drug-like molecular property space. An understanding of the hit-to-lead process provides a practical guide to molecular design characteristics. Examples of leads generated from library approaches also provide a benchmarking of successes as well as aspects for continued development of library design practices.

Discovery of an orally active non-peptide fibrinogen receptor antagonist based on the hydantoin scaffold.

Antagonists of the platelet fibrinogen receptor (GP IIb/IIIa receptor) are expected to be a promising new class of antithrombotic agents. The binding of fibrinogen to the fibrinogen receptor depends on an Arg-Gly-Asp-Ser (RGDS) tetrapeptide recognition motif. Structural modifications of the RGDS lead have led to the discovery of a non-peptide RGD mimetic GP IIb/IIIa antagonist 44 (S 1197). Compound 44 inhibited, in a dose dependent and reversible manner, human and dog platelet aggregation as well as 125I-fibrinogen binding to ADP-activated human gel filtered platelets and isolated GP IIb/IIIa with K(i) values of 9 nM and 0.17 nM, respectively. A pharmacophore mapping procedure with QXP and a 3D-QSAR analysis applying the GRID/GOLPE methodology yielded a stable, rather predictive model and revealed structural features which are important for binding. Hydrophobic substitutions both at the hydantoin nucleus and at the C-terminus increase the affinity toward the fibrinogen receptor. The crystalline ethyl ester prodrug 48 (HMR 1794) is an orally active antithrombotic agent which is a promising drug candidate for the treatment of thrombotic diseases in humans.

VolSurf: a new tool for the pharmacokinetic optimization of lead compounds.

A method for the modeling and prediction of pharmacokinetic properties based on computed molecular interaction fields and multivariate statistics has been investigated in different experimental datasets. The program VolSurf was used to correlate 3D molecular structures with physico-chemical and pharmacokinetic properties. In membrane partitioning, VolSurf produced a two-component model explaining 94% of the total variation with a predictive q(2) of 0.90. This result was achieved without conformational sampling and without any quantum-chemical calculation. For the prediction of blood-brain barrier penetration the VolSurf model was able to predict the BBB profile for most of the drugs in the external prediction set. In Caco-2 and MDCK permeation experiments, VolSurf was used with success to establish statistical models and to predict the behaviour of new compounds. The method thus appears as a valuable new property filter in virtual screening and as a novel tool in optimizing the pharmacokinetic profile of pharmaceutically relevant compounds.

Three-dimensional structure of moenomycin A--a potent inhibitor of penicillin-binding protein 1b.

The first three-dimensional structure of moenomycin A in aqueous solution based on NMR-derived distance constraints and molecular dynamics simulations is presented. The antibiotic moenomycin A was obtained from the FlavomycinR complex by ultrafiltration, chromatography on a DEAE-cellulose ion exchanger and reverse-phase chromatography in 98% purity. In contrast to the previously reported behaviour, the compound gave rise to well-resolved NMR spectra in standard solvents. Using several two-dimensional experiments, a complete assignment of proton and carbon chemical shifts was achieved in (CD3)2SO, CD3OD and H2O/D2O (9:1, pH 7.3). A total of 175 interproton distances were determined from 600-MHz rotating-frame NOE (ROE) spectra and were used as restraints in molecular dynamics calculations. These restraints included 84 ROEs between protons of the moenocinol part leading to a very well-defined structure of the lipid part of the molecule. The relative orientation of the subunits was determined by 66 ROEs among different sugar rings and between the sugar rings and moenocinol. As a result of the molecular dynamics calculation, rings D, E, and F as well as the moenocinol part are very well-defined (average rms deviation over all heavy atoms 0.48 A) whereas rings A, B and C display a higher degree of conformational flexibility which might be an artefact due to the lower number of ROEs in this part of the molecule. A three-dimensional pharmacophore hypothesis comprising functional groups of rings E and F and the carboxyl group of glyceric acid is presented on the basis of the degradation and derivatization studies of Welzel and coworkers [Welzel, P., Kunisch, F., Kruggel, F., Stein, H., Scherkenbeck, J., Hiltmann, A., Duddeck, H., Müller, D., Maggio, J. E., Fehlhaber, H.-W., Seibert, G., van Heijenoort, Y. & van Heijenoort, J. (1987) Moenomycin A: Minimum structural requirements for biological activity, Tetrahedron 43, 585-598; Welzel, P. (1993) Transglycosylase inhibition, in Antibiotics and antiviral compounds--Chemical synthesis and modifications (Krohn, K., Kirst, H. A. & Maag, H., eds) pp. 373-378, VCH Weinheim].

3D structure of ramoplanin: a potent inhibitor of bacterial cell wall synthesis.

The 3D structure of ramoplanin was studied by NMR spectroscopy in aqueous solution. A total of 320 interproton distances were determined from a NOESY spectrum and were used as restraints in distance geometry calculations. A structural refinement was carried out by molecular dynamics calculations in a solvent box. The structure of ramoplanin is characterized by two antiparallel beta-strands which are formed by the residues 2-7 and 10-14, respectively. The beta-strands are connected by six intramolecular hydrogen bonds and a reverse beta-turn which is formed by Thr8 and Phe9 (in positions i+1 and i+2, respectively). Residues 2 and 14 are connected by a loop consisting of Leu15, Ala16, Chp17, and the side chain of Asn2. Although residues 14-17 show the formation of a beta-turn, only the N-terminal end of the turn is directly connected to one of the beta-strands (Gly14), whereas the C-terminal end (Chp17) is linked via the side chain of Asn2. The 3D conformation of ramoplanin is also stabilized by a hydrophobic cluster of the aromatic sidechains of the residues 3, 9, and 17. This hydrophobic collapse leads to a U-shaped topology of the beta-shee: with the beta-turn at one end and the loop at the other end. The structure found for ramoplanin differs corsiderably from the published structure of ramoplanose which might be due to a smaller number of NOE distance restraints used in the previous study.

Transport of the antibacterial agent oxazolidin-2-one and derivatives across intestinal (Caco-2) and renal (MDCK) epithelial cell lines.

The transepithelial passage of the orally bioavailable antibacterial agent oxazolidin-2-one (OXa) and 10 derivatives has been studied with human intestinal (Caco-2) and canine renal (MDCK) cell lines grown on polycarbonate filters. The transepithelial passage was assayed in the apical-to-basolateral (AP-to-BL) direction and in the opposite direction (BL to AP) in both cell lines. The observed passage rates of OXa were similar in both directions in the two cell lines, suggesting passive diffusion. This was further confirmed by the fact that transport kinetics were linear as a function of initial concentration. The rates of AP-to-BL passage of OXa and seven of the derivatives in both cell lines were linearly related to lipophilicity, whether expressed as high-passage liquid chromatography retention time or as the logarithm of the n-octanol-water partition coefficient (log P). These data suggest that the lipophilicity of OXa is important for its observed bioavailability after oral administration. Interestingly, three of the derivatives exhibited a higher passage rate than predicted by lipophilicity. Further studies indicated that this transport was saturable, similar in the two directions, and not affected by energy depletion, suggesting the presence of an additional carrier-mediated facilitated-transport mechanism.

New evidence for a membrane-bound pathway in hormone receptor binding.

The fully active cholecystokinin analog (Thr,Nle)-CCK-9 was lipo-derivatized by N-terminal grafting of a dimyristoylglycerol moiety to induce tight interdigitation with cell membrane bilayers. While the parent CCK peptide was shown to interact only transiently with small unilamellar phospholipid vesicles, the lipo-CCK peptide, although self-aggregating into vesicles, inserts rapidly and quantitatively into phospholipid bilayers. Fluorescence and, even more so, NMR data are supportive for a chain reversal of the CCK moiety of the lipo derivative with embedment of the C-terminus into hydrophobic compartments of the bilayer. MD simulations allowed for a proposal of the folded form of CCK in bilayers with a helical array parallel to the interface and an amphipathic display of the side chains. In this model, the phenylalanine aromatic ring is heading the peptide molecule and may thus play a decisive role in the lateral penetration of the receptor at the water/lipid interface. In fact, despite the membrane-bound state, its binding affinity for rat pancreatic acini is comparable to that of the CCK peptide when tested after a 3-h equilibration period but 5-6-fold lower at 45 min, suggesting that the association rate is significantly lower than that of the unmodified CCK peptide. This can rationally be attributed to the tight interdigitation of the double-tailed lipo moiety with the membrane bilayer. Moreover, an escape of the lipopeptide into the extracellular aqueous phase is energetically highly unfavored; therefore, the receptor can only be reached by a membrane-bound two-dimensional migration. The observed difference in amplification between binding and amylase secretion may result from inadequate occupation of low-affinity CCK receptors, which leads then to poor couplings to G-proteins. Nevertheless the data confirm that lateral penetration of receptor structures is possible, and thus, preadsorption of peptide (neuro)hormones at the cell membrane bilayer may indeed represent the first step in the receptor recognition process.

Synthetic immunogens. The effect of the conformational space on biological and immunological responses to dimeric hormone constructs.

Chimeras of the double chain bis-cystinyl hinge fragment 225-232/225'-232' of the human IgG1 and of peptides related to human little-gastrin were synthesized, whereby the fully bioactive gastrin sequences 2-17 and 5-17 were amide-bond-linked N- and N- or C-terminally, respectively, to the hinge peptide. All the dimeric constructs proved to be efficient immunogens; however, both the configuration of the constructs and the length of the haptenic gastrin molecule were found to drastically affect the specificity of the antibody response and, thus, the type of dominant immune epitope expressed. The different degree of accessibility of the gastrin chains in the dimers is similarly reflected by their binding affinities to gastrin receptors and their bioactivities in vivo. Molecular dynamics simulations of the chimeric compounds clearly revealed that the conformational space of the gastrin peptide chains 2-17 and 5-17 is strongly restricted upon linkage to the hinge peptide. Only in the gastrin-(2-17) construct does sufficient free conformational space seem to be retained, at least for one of the two gastrin chains, in order to allow folding into the bioactive structure. This also agrees with the observation that the dimeric gastrin-(2-17) behaves like a gastrin monomer in terms of receptor binding affinity and biopotency in vivo; but it could additionally explain why an antibody response of gastrin receptor-like specificity could only be induced with this construct. The experimental data may therefore suggest a high degree of parallelism between the mechanism of recognition of the gastrin peptides in the dimeric constructs as hormonal ligands by the gastrin receptors and as haptens by the immune competent cells.