Combinatorial Chemistry & High Throughput Screening - Volume 4, Issue 8, 2001

The complexation of p-tert-butylphenyl p-tert-butylbenzoate, N- (p-tert-butylphenyl)-p-tert-butylbenzamide and a bisadamantyl-phosphate derivative with a beta-cyclodextrin derivative formed by two cyclodextrin units linked by a disulfide bridge on one of the C6 atoms have been studied by computational methods (free energy perturbation (FEP) and Molecular Mechanics / Poisson Bolzmann Surface Area (MM / PBSA)). The calculated relative free energies of the amide and ester are in good agreement with experiment only for MM / PBSA and not for FEP. Only MM / PBSA was applied to the bisadamantyl-phosphate complex and its calculated association free energy was calculated to be similar to that of the ester, which is consistent with the experimental tendencies.

An overview of the simplified linear interaction energy (LIE) method for calculation of ligand binding free energies is given. This method is based on force field estimations of the receptor-ligand interactions and thermal conformational sampling. A notable feature is that the binding energetics can be predicted by considering only the intermolecular interactions between the ligand and receptor. The approximations behind this approach are examined and different parametrizations of the model are discussed. In general, LIE type of methods appears particularly useful for computational drug lead optimization.

The complexes of human neutrophil elastase with a series of 40 N3-substituted trifluoromethylketone-based pyridone inhibitors have been modelled. The series spans three orders of magnitude in inhibition constants despite the fact that it was originally developed in an attempt to improve the oral activity of a lead compound. Ligand-receptor interaction energies calculated using molecular mechanics did not correlate well with the experimental activities. A good correlation with activity was found, however, when a COMBINE analysis of the same data was carried out, which allowed a quantitative interpretation of the modelled complexes. The essence of this method is to partition the ligand-receptor interaction energies into individual residue-based van der Waals and electrostatic contributions, and to subject the resulting energy matrix to partial least squares analysis. Incorporation of two additional descriptors representing the electrostatic energy contributions to the partial desolvation of both the receptor and the ligands improved the QSAR model, as did the replacement of the distance-dependent electrostatic contributions with solvent-screened electrostatic interactions calculated by numerically solving the Poisson-Boltzmann equation. The model was validated both internally (cross-validation) and externally, using a set of twelve 6-phenyl-pyridopyrimidine analogs. The analysis reveals the subtle interplay of binding forces which occurs within the enzyme active site and provides objective information that can be interpreted in the light of the receptor structure. This information, gained from a series of real compounds, can be easily translated into 3D real or virtual database queries in the search for more active derivatives.

This review describes computational procedures for deriving the amino acid sequences that are compatible with a given protein backbone structure. Such procedures can be used to gain insight into the constraints imposed by the 3D structure of the protein sequence, or to design proteins that are likely to adopt a given backbone conformation. We start by presenting a short overview of the various types of approaches to protein design developed over more than a decade. This is followed by a more detailed presentation of a recently developed sequence selection procedure DESIGNER. This latter presentation illustrates the basic principles underlying this type of procedures, described what they may teach us when applied to small proteins, and highlights issues that need to be addressed in order to go forward.

A new computational approach (PEP) is presented for the structure-based design of linear polymeric ligands consisting of any type of amino acid. Ligands are grown from a seed by iteratively adding amino acids to the growing construct. The search in chemical space is performed by a build-up approach which employs all of the residues of a user-defined library. At every growing step, a genetic algorithm is used for conformational optimization of the last added monomer inside the binding site of a rigid target protein. The binding energy with electrostatic solvation is evaluated to select sequences for further growing. PEP is tested on the peptide substrate binding site of the insulin receptor tyrosine kinase and farnesyltransferase. In both test cases, the peptides designed by PEP correspond to the sequence motifs of known substrates. For tyrosine kinase, tyrosine residues are suggested at position P and P+2. While the tyrosine at P is in agreement with the experimental data, the one at P+2 is a prediction which awaits experimental validation. For farnesyltransferase, it is shown that electrostatic solvation is necessary for the correct design of peptidic inhibitors.

The genetic information that determines the structure and function of living organisms is encoded in the nucleotide sequence of double-stranded DNA molecules. Despite an apparent structural homogeneity displayed by DNA, subtle local variations in structure and dynamics are functionally significant. Short sequences exhibit specificity for regulatory and catalytic proteins, which mediate fundamental processes necessary to the survival of the cell. However, the molecular basis for specific recognition is still incompletely understood. The “indirect readout” mechanism suggests that the relative propensity of DNA to undergo structural deformations induced by the protein contributes to specific protein-DNA recognition. Although the hypothesis was originally formulated to explain recognition of specific nucleic acid sequences by DNA-binding proteins, it may have particular application to the recognition of DNA damage, because damaged sites in DNA have different equilibrium structure and dynamics from undamaged DNA. In this work, we review the approaches that we took to investigate the questions of sequence- and damage-dependent structure and dynamics of DNA.We describe a statistical thermodynamic model that relates DNA configurational flexibility to sequence-specific protein-DNA binding. The model provides a theoretical basis for interpreting experimental measurements of DNA dynamics. We describe results from MCSCF calculations of the excited states of 2-aminopurine (2AP), which provide the theoretical basis for the intramolecular mechanism of quenching as well as the effect of environment on this process. We then describe our investigations of the effect of stacking, base pairing, and base dynamics on the fluorescence of 2-AP in model systems, which allow us to develop the relationships between steady-state and time-resolved fluorescence parameters on the one hand and local structural and dynamic properties of DNA on the other hand. Finally, we describe the application of the experimental approach to study the conformational heterogeneity of DNA abasic sites, a commonly occurring type of DNA damage. We demonstrate the power of the experimental algorithm to characterize the physical differences between undamaged and damaged DNA, as well as the effects of nucleic acid sequence in both of these contexts. Thus, the work described herein comprises a combination of theoretical and experimental approaches to the problem of sequence- and damage-dependent DNA deformation.

We have recently developed a theoretical means of studying the mechanical and interaction properties of nucleic acids as a function of their base sequence. This approach, termed ADAPT, can be used to obtain the physical properties of millions of base sequences with only modest computational expense. ADAPT is based on a multi-copy algorithm using special nucleotides (“lexides”) containing all four standard bases whose contribution to the energy of the molecule can be varied. We present here a deeper study of the energy minima which occur in the multi-dimensional space defined by these variable sequences. We also present an extension of the approach termed “gene threading” which enables us to scan genomic sequence data in an attempt to locate preferential binding sites. This technique is illustrated for the case of TATA-box protein binding. ADAPT enables us to demonstrate that, for this protein, DNA deformation alone explains a large part of the experimentally observed consensus binding sequence.

Recent tools that analyze microarray expression data have exploited correlation-based approaches such as clustering analysis. We describe a new method for assessing the importance of genes for sample classification based on expression data. Our approach combines a genetic algorithm (GA) and the k-nearest neighbor (KNN) method to identify genes that jointly can discriminate between two types of samples (e.g. normal vs. tumor). First, many such subsets of differentially expressed genes are obtained independently using the GA. Then, the overall frequency with which genes were selected is used to deduce the relative importance of genes for sample classification. Sample heterogeneity is accommodated; that is, the method should be robust against the existence of distinct subtypes. We applied GA / KNN to expression data from normal versus tumor tissue from human colon. Two distinct clusters were observed when the 50 most frequently selected genes were used to classify all of the samples in the data sets studied and the majority of samples were classified correctly. Identification of a set of differentially expressed genes could aid in tumor diagnosis and could also serve to identify disease subtypes that may benefit from distinct clinical approaches to treatment.