Functionality map analysis of the active site cleft of human thrombin

Quantum and molecular dynamics study for binding of macrocyclic inhibitors to human alpha-thrombin

Molecular dynamics simulations followed by quantum mechanical calculation and Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA) analysis have been carried out to study binding of proline- and pyrazinone-based macrocyclic inhibitors (L86 and T76) to human alpha-thrombin. Detailed binding interaction energies between these inhibitors and individual protein fragments are calculated using DFT method based on a new quantum mechanical approach for computing protein-ligand interaction energy. The analysis of detailed interaction energies provides insight on the protein-ligand binding mechanism. Study shows that T76 and L86 bind to thrombin in a very similar "inhibition mode" except that T76 has relatively weaker binding interaction with Glu(217). The analysis from quantum calculation of binding interaction is consistent with the MM-PBSA calculation of binding free energy, and the calculated free energies for L86/T76-thrombin binding agree well with the experimental data.

Computer-aided design of chiral ligands. Part 2. Functionality mapping as a method to identify stereocontrol elements for asymmetric reactions

A computational method to determine the energetically favorable positions of functional groups with respect to the transition states of stereoselective reactions based on force field energy minimization is presented. The parameters of this functionality mapping, the characteristics of the target transition states, and the features of the probe structures are outlined. Our method was found to reproduce the positions of the stereodiscriminating fragments for some known chiral ligands including the Masamune dimethylborolane, dimenthylborane, the Corey stien reagent, the Roush allylboronate tartrates, and the secondary amine Diels-Alder catalysts described by MacMillan. Functionality mapping can be used to better understand the specific interactions in the transition states leading to the products by providing a quantitative measure of the stabilization/destabilization afforded by the different ligand components via nonbonded interactions. The method can determine if a chiral ligand imparts the observed selectivity by stabilizing one reaction pathway, by destabilizing a reaction pathway, or by a combination of both. Orientational as well as positional information about potential functional groups is readily obtained. In addition to its utility as an analytical tool, functionality mapping can be used to explore starting points for the design of new chiral ligands.

Evaluation of designed ligands by a multiple screening method: application to glycogen phosphorylase inhibitors constructed with a variety of approaches

Glycogen phosphorylase (GP) is an important enzyme that regulates blood glucose level and a key therapeutic target for the treatment of type II diabetes. In this study, a number of potential GP inhibitors are designed with a variety of computational approaches. They include the applications of MCSS, LUDI and CoMFA to identify additional fragments that can be attached to existing lead molecules; the use of 2D and 3D similarity-based QSAR models (HQSAR and SMGNN) and of the LUDI program to identify novel molecules that may bind to the glucose binding site. The designed ligands are evaluated by a multiple screening method, which is a combination of commercial and in-house ligand-receptor binding affinity prediction programs used in a previous study (So and Karplus, J. Comp.-Aid. Mol. Des., 13 (1999), 243-258). Each method is used at an appropriate point in the screening, as determined by both the accuracy of the calculations and the computational cost. A comparison of the strengths and weaknesses of the ligand design approaches is made.

De novo design of molecular architectures by evolutionary assembly of drug-derived building blocks

An evolutionary algorithm was developed for fragment-based de novo design of molecules (TOPAS, TOPology-Assigning System). This stochastic method aims at generating a novel molecular structure mimicking a template structure. A set of approximately 25,000 fragment structures serves as the building block supply, which were obtained by a straightforward fragmentation procedure applied to 36,000 known drugs. Eleven reaction schemes were implemented for both fragmentation and building block assembly. This combination of drug-derived building blocks and a restricted set of reaction schemes proved to be a key for the automatic development of novel, synthetically tractable structures. In a cyclic optimization process, molecular architectures were generated from a parent structure by virtual synthesis, and the best structure of a generation was selected as the parent for the subsequent TOPAS cycle. Similarity measures were used to define 'fitness', based on 2D-structural similarity or topological pharmacophore distance between the template molecule and the variants. The concept of varying library 'diversity' during a design process was consequently implemented by using adaptive variant distributions. The efficiency of the design algorithm was demonstrated for the de novo construction of potential thrombin inhibitors mimicking peptide and non-peptide template structures.

Design of dimerization inhibitors of HIV-1 aspartic proteinase: a computer-based combinatorial approach

Inhibition of dimerization to the active form of the HIV-1 aspartic proteinase (HIV-1 PR) may be a way to decrease the probability of escape mutations for this viral protein. The Multiple Copy Simultaneous Search (MCSS) methodology was used to generate functionality maps for the dimerization interface of HIV-1 PR. The positions of the MCSS minima of 19 organic fragments, once postprocessed to take into account solvation effects, are in good agreement with experimental data on peptides that bind to the interface. The MCSS minima combined with an approach for computational combinatorial ligand design yielded a set of modified HIV-1 PR C-terminal peptides that are similar to known nanomolar inhibitors of HIV-1 PR dimerization. A number of N-substituted 2,5-diketopiperazines are predicted to be potential dimerization inhibitors of HIV-1 PR.

Comparison of two implementations of the incremental construction algorithm in flexible docking of thrombin inhibitors

A set of 32 known thrombin inhibitors representing different chemical classes has been used to evaluate the performance of two implementations of incremental construction algorithms for flexible molecular docking: DOCK 4.0 and FlexX 1.5. Both docking tools are able to dock 10-35% of our test set within 2 A of their known, bound conformations using default sampling and scoring parameters. Although flexible docking with DOCK or FlexX is not able to reconstruct all native complexes, it does offer a significant improvement over rigid body docking of single, rule-based conformations, which is still often used for docking of large databases. Docking of sets of multiple conformers of each inhibitor, obtained with a novel protocol for diverse conformer generation and selection, yielded results comparable to those obtained by flexible docking. Chemical scoring, which is an empirically modified force field scoring method implemented in DOCK 4.0, outperforms both interaction energy scoring by DOCK and the Böhm scoring function used by FlexX in rigid and flexible docking of thrombin inhibitors. Our results indicate that for reliable docking of flexible ligands the selection of anchor fragments, conformational sampling and currently available scoring methods still require improvement.

Artificial neural networks for computer-based molecular design

The theory of artificial neural networks is briefly reviewed focusing on supervised and unsupervised techniques which have great impact on current chemical applications. An introduction to molecular descriptors and representation schemes is given. In addition, worked examples of recent advances in this field are highlighted and pioneering publications are discussed. Applications of several types of artificial neural networks to compound classification, modelling of structure-activity relationships, biological target identification, and feature extraction from biopolymers are presented and compared to other techniques. Advantages and limitations of neural networks for computer-aided molecular design and sequence analysis are discussed.

Towards structural models of molecular recognition in olfactory receptors

The G protein coupled receptors (GPCR) are an important class of proteins that act as signal transducers through the cytoplasmic membrane. Understanding the structure and activation mechanism of these proteins is crucial for understanding many different aspects of cellular signalling. The olfactory receptors correspond to the largest family of GPCRs. Very little is known about how the structures of the receptors govern the specificity of interaction which enables identification of particular odorant molecules. In this paper, we review recent developments in two areas of molecular modelling: methods for modelling the configuration of trans-membrane helices and methods for automatic docking of ligands into receptor structures. We then show how a subset of these methods can be combined to construct a model of a rat odorant receptor interacting with lyral for which experimental data are available. This modelling can help us make progress towards elucidating the specificity of interactions between receptors and odorant molecules.

Modern computational chemistry and drug discovery: structure generating programs

During 1996 and 1997, the first reports were disclosed of active enzyme inhibitors based entirely on novel structures created by de novo methods. De novo methods have also been used to modify and significantly improve the binding affinity of an HIV protease inhibitor. Work continues in the improvement of methods for the de novo design of compounds which fit and chemically complement a binding site. De novo algorithms that generate only synthetically feasible structures have also been reported. In addition, methods are being developed for the automatic computer generation of virtual molecular libraries which can be searched to identify molecules to match a pharmacophore or fit into a binding site.

Active-site-directed 3D database searching: pharmacophore extraction and validation of hits

Two new computational tools, PRO_PHARMEX and PRO_SCOPE, for use in active-site-directed searching of 3D databases are described. PRO_PHARMEX is a flexible, graphics-based program facilitating the extraction of pharmacophores from the active site of a target macromolecule. These pharmacophores can then be used to search a variety of databases for novel lead compounds. Such searches can often generate many 'hits' of varying quality. To aid the user in setting priorities for purchase, synthesis or testing, PRO_SCOPE can be used to dock molecules rapidly back into the active site and to assign them a score using an empirical scoring function correlated to the free energy of binding. To illustrate how these tools can add value to existing 3D database software, their use in the design of novel thrombin inhibitors is described.

Computational combinatorial ligand design: application to human alpha-thrombin

A new method is presented for computer-aided ligand design by combinatorial selection of fragments that bind favorably to a macromolecular target of known three-dimensional structure. Firstly, the multiple-copy simultaneous-search procedure (MCSS) is used to exhaustively search for optimal positions and orientations of functional groups on the surface of the macromolecule (enzyme or receptor fragment). The MCSS minima are then sorted according to an approximated binding free energy, whose solvation component is expressed as a sum of separate electrostatic and nonpolar contributions. The electrostatic solvation energy is calculated by the numerical solution of the linearized Poisson-Boltzmann equation, while the nonpolar contribution to the binding free energy is assumed to be proportional to the loss in solvent-accessible surface area. The program developed for computational combinatorial ligand design (CCLD) allows the fast and automatic generation of a multitude of highly diverse compounds, by connecting in a combinatorial fashion the functional groups in their minimized positions. The fragments are linked as two atoms may be either fused, or connected by a covalent bond or a small linker unit. To avoid the combinatorial explosion problem, pruning of the growing ligand is performed according to the average value of the approximated binding free energy of its fragments. The method is illustrated here by constructing candidate ligands for the active site of human alpha-thrombin. The MCSS minima with favorable binding free energy reproduce the interaction patterns of known inhibitors. Starting from these fragments, CCLD generates a set of compounds that are closely related to high-affinity thrombin inhibitors. In addition, putative ligands with novel binding motifs are suggested. Probable implications of the MCSS-CCLD approach for the evolving scenario of drug discovery are discussed.