Current Computer - Aided Drug Design - Volume 3, Issue 1, 2007

Protein toxins play key roles in many infectious diseases of humans which are caused by bacteria. In some cases the toxin alone is directly responsible for the majority of the symptoms of the disease (e.g. tetanus, anthrax, diphtheria). In others the toxin is one of an arsenal of virulence factors which allow the bacterium to cause disease. Antibiotics are currently the mainstay for the treatment of bacterial infections. However, increasing levels of antibiotic resistance and the indiscreet nature of antibiotic therapy are limitations. Prior to the availability of antibiotics, antisera against toxins were often used to treat bacterial disease. Nowadays, animal-sourced products, such as antisera, are generally not acceptable for use in humans. Against the background there is an increasing interest in the development of low molecular weight inhibitors of toxins for the treatment of disease. For some toxins, like anthrax toxin, botulinum toxin and shigella toxin, low molecular weight inhibitors demonstrate proof of principle of this concept. For most other toxins the design and development of inhibitors is now a very real prospect; the crystal structures of many toxins are available, and in most cases the identity of the substrate or receptor is known. This article describes in detail the rational design of bacterial toxin inhibitors.

The intake of medicines in form of aerosols is becoming increasingly popular, especially in the treatment of different lung diseases and allergies. In addition, there is a great interest to utilize the inhalation pathway for systemic therapy. Hence, determination of the required local distribution of inhaled therapeutic aerosols within the respiratory system is a key issue of modern aerosol drug design. In general, deposition characteristics of inhaled particles depend on the properties of the aerosols, the breathing mode and the geometry of the airways. All three parameters must be analyzed for the optimal design of therapeutic aerosols. A recommended way of drug inhalation may differ for various illnesses and patients. There are two different modeling directions for the description of deposition characteristics of inhaled drugs in the respiratory system. One way is the application of lung deposition models for the determination of total, regional and airway generation-specific deposition, and the other way is the usage of computational fluid dynamics techniques for the characterization of local deposition patterns, which, at present, cannot be applied to the whole respiratory system. This computational fluid dynamics approaches will be analyzed in another study. This work describes the general background of aerosol drug delivery optimization, summarizes previous important studies in the field, and provides a comprehensive discussion about numerical lung modeling and the salient features of the newest models and techniques. In the last part, the stochastic lung deposition model is applied to determine the optimal particle size and breathing technique for bronchial and pulmonary drug delivery.

The continued implementation of NMR-based approaches in hit-through-lead drug discovery in academic and corporate settings is founded upon NMR applications that assess structure activity relationships. A very recent application of NMR spectroscopy to these discovery initiatives involves fraganomics, in which NMR is used to iteratively “guide” the assembly of several weakly interacting fragments or small molecules through chemical links. Moreover, several groups have recently reported the potential of integrating NMR spectroscopy with in silico, virtual screens of large chemical repositories possessing diverse collections of small molecules. Importantly an improved understanding of the intermolecular forces that mediate protein-protein/ protein-ligand interactions has been integral to improving these virtual screening approaches, resulting in the identification of novel ligands for several therapeutic targets. Recent success of these structure-based discovery initiatives in targeting protein-protein interactions that are responsible for the non-covalent assembly and/or regulation of macromolecular complexes and are a critical paradigm in many disease pathologies will be discussed. The atomic details of these requisite interactions are the cornerstone of NMR and crystallographic “structure-guided”, drug discovery initiatives aimed at disrupting complex formation. This review will predominantly focus on the recent advances in structure based computational screening approaches, highlighting the successful integration of in silico virtual screens with NMR-based techniques. The application of this powerful, combinatorial approach for the evaluation of well-characterized target space as well as its application to unique chemical space such as the protein-protein interaction inhibition (2P2I) that has recently been shown to be tractable to small molecule intervention will be discussed.

Protein-protein interactions play a crucial role in numerous vital cell functions. However proteinprotein interactions are also responsible for pathological formation of protein aggregates, which determine the development of several diseases. The key role of protein-protein interactions for manifestation of numerous cell functions attracts much attention to protein complexes as perspective drug targets. So design or discovery of small molecules that would regulate protein-protein interactions represents great pharmacological interest. The recent progress in understanding of mechanism protein-protein interaction, including role of flexibility of protein-protein interfaces, thermodynamic of complex formation, discovery of small molecules modifying protein-protein interactions, the advantages and limitation of protein-protein inhibitors as potential drugs are discussed in this review.

The dominant paradigm in drug discovery emphasizes techniques that generate large amounts of data. What was possible by simple inspection in the past, nowadays cannot be effectively achieved without the aid of informatics techniques. In this context substructural analysis techniques are increasing their role in the organization and management of information generated. Advances in the field of substructure analysis have expanded the applicability of substructural analysis in multiple fronts in early lead discovery and optimization. It can be applied beyond the management of information, including compound library design and virtual screening to structure activity relationships. The relationships between chemical substructures and drug-like properties also aid in developing more robust rationales for fragment-based approaches for lead discovery, predictive toxicology, and elucidation of pharmacokinetic properties.A review of recent developments in substructure analysis in a broad range of areas in drug discovery is presented. The focus is on the application of substructural analysis in computational chemistry for drug design and the methods used to identify substructures in a chemical database, as well as their relation to fragment-based drug discovery. The discussion shows the benefits of substructural analysis to the drug discovery process and gives impetus to further advancement of substructure analysis techniques.

Fragment based screening has become a valuable tool to complement traditional lead finding methods like high throughput screening in drug discovery. Fragments are low molecular mass compounds and are usually screened using high sensitivity biophysical methods which are suitable for the detection of weakly binding ligands. Because fragments have a low affinity, efficient methods to improve their affinity are required. Structure based methods, i.e. methods which make use of a three dimensional structure of the protein have been applied in most of the cases for fragment optimization programs which are reported in the literature. De novo design, combinatorial docking and interactive optimization fell in this category and belong to the computer-aided drug design field. While de novo design is a computational method where a ligand is build completely de novo, combinatorial docking is applied to evaluate easily accessible or physically existing compound libraries around a previously identified core and interactive optimization alternates computational, biological and structural experiments to progress towards a drug. The principles, advantages, drawbacks of the different methods are being discussed together with examples of applications taken from the literature. At the end of the article we define a new metric to express the efficiency of optimization and show that small molecular molecules, i.e. fragments with a molecular mass below 250 Da, tend to be more easily optimized than larger molecules, thus reinforcing the interest of the fragment approach in the drug discovery process.