Current Topics in Medicinal Chemistry - Volume 12, Issue 17, 2012

Transient interactions of endogenous and exogenous small molecules with flexible binding sites in proteins or macromolecular assemblies play a critical role in all biological processes. Current advances in high-resolution protein structure determination, database development, and docking methodology make it possible to design three-dimensional models for prediction of such interactions with increasing accuracy and specificity. Using the data collected in the Pocketome encyclopedia, we here provide an overview of two types of the three-dimensional ligand activity models, pocketbased and ligand property-based, for two important classes of proteins, nuclear and G-protein coupled receptors. For half the targets, the pocket models discriminate actives from property matched decoys with acceptable accuracy (the area under ROC curve, AUC, exceeding 84%) and for about one fifth of the targets with high accuracy (AUC > 95%). The 3D ligand property field models performed better than 95% in half of the cases. The high performance models can already become a basis of activity predictions for new chemicals. Family-wide benchmarking of the models highlights strengths of both approaches and helps identify their inherent bottlenecks and challenges.

More needs to be done by the private sector to optimize the drug discovery and development pipeline. In addition, significant efforts should also be focused on the understanding of mechanism of diseases, on the characterization of unexplored biochemical pathways and on the validation of new protein targets. Chemical genomics, which uses chemical probes to help understand the complexity of biological systems at the gene and protein levels, has proven in recent years to be an important tool. Experimental and computational chemical genomic screenings have been used by the private sector and recently also by academia and non-profit institutions for drug repurposing or repositioning to find new indications for known drugs. A detailed overview of the current initiatives in drug repurposing, initiated by the major governmental funding agencies around the world is reported. The push towards greater efficiency is encouraging drug repurposing and other techniques in chemical genomics. Finding the best ways to improve translational research and accelerate the regulation of clinical phases means being able to launch safer drugs into the market faster.

A significant portion of recent drug development efforts has been focused on protein kinases. More than a hundred different compounds are currently under clinical trials and nearly 30% of all the scientific articles in drug discovery are on protein kinase inhibitors. Protein kinases are very flexible targets and undergo significant conformational changes upon activation and during the catalytic cycle. This flexibility can be exploited in drug discovery. Some of the inactive states that emerge during the conformational changes are targeted by various inhibitors with a significant gain in selectivity. Here, we review the recent advances being made in understanding the details and the mechanism of these conformational changes thanks to the progress in molecular dynamics and free energy algorithms as well as to the availability of specialized computer hardware.

The legacy of the advances made in high-throughput screening (HTS) in the 1990's is a large source of public data from which models can be derived using QSAR methods. This paper will examine the integrity of these public data sources and the implications for model building.

The knowledge about to which biomolecules a small molecule binds is highly valuable in the drug development process. Although analytical methods to dissect ligand-binding proteome have made substantial progress in the past decades, it is generally too costly, if not infeasible, to know where a small molecule binds at very high resolution. Computational prediction of binding partners of small chemical molecules has become a useful approach to evaluate their potential therapeutic applications or adverse effects. In this article two computational approaches that were adopted to perform target identification, namely, molecular docking and pharmacophore fitting, are reviewed. Both approaches enable the identification of key interactions between the biomolecules and the small molecules. Databases that can be used to further improve the implementation and the computational methods and to benchmark their performances are also included.

Over the last 50 years, sequencing, structural biology and bioinformatics have completely revolutionised biomolecular science, with millions of sequences and tens of thousands of three dimensional structures becoming available. The bioinformatics of enzymes is well served by, mostly free, online databases. BRENDA describes the chemistry, substrate specificity, kinetics, preparation and biological sources of enzymes, while KEGG is valuable for understanding enzymes and metabolic pathways. EzCatDB, SFLD and MACiE are key repositories for data on the chemical mechanisms by which enzymes operate. At the current rate of genome sequencing and manual annotation, human curation will never finish the functional annotation of the ever-expanding list of known enzymes. Hence there is an increasing need for automated annotation, though it is not yet widespread for enzyme data. In contrast, functional ontologies such as the Gene Ontology already profit from automation. Despite our growing understanding of enzyme structure and dynamics, we are only beginning to be able to design novel enzymes. One can now begin to trace the functional evolution of enzymes using phylogenetics. The ability of enzymes to perform secondary functions, albeit relatively inefficiently, gives clues as to how enzyme function evolves. Substrate promiscuity in enzymes is one example of imperfect specificity in protein-ligand interactions. Similarly, most drugs bind to more than one protein target. This may sometimes result in helpful polypharmacology as a drug modulates plural targets, but also often leads to adverse side-effects. Many chemoinformatics approaches can be used to model the interactions between druglike molecules and proteins in silico. We can even use quantum chemical techniques like DFT and QM/MM to compute the structural and energetic course of enzyme catalysed chemical reaction mechanisms, including a full description of bond making and breaking.

In silico medicinal chemistry investigates molecular systems that are too large to be tackled by medium to high level ab initio quantum chemistry. Only atomistic force fields can deliver rapid computation of energy required in sampling the many conformational and orientational degrees of freedom of a ligand within a protein pocket. However, the predictive reliability of a force field critically depends on the quality and realism of its energy function. Particularly, the electrostatic component of this energy needs to be as accurate as possible because druglike ligands and proteins are polar molecules, whose interaction does not just depend on shape. Surprisingly, the challenging problem of energy accuracy receives much less attention than it deserves. Docking results in the literature are still dependent on atomic point charges, which are inherently inaccurate at short and medium range. This has been known for decades but improved and more accurate methods have not (yet) found their way in mainstream in silico medicinal chemistry. Moreover, often the “details” of the electrostatic energy are poorly and not at all reported, as if they do not matter. This article attempts to inspire future docking algorithms with ideas from an approach called Quantum Chemical Topology (QCT). The way this method partitions energy and treats the electrostatic interaction should inject more realism into the current paradigm. The gap between the medicinal chemistry “world view” and that of physical and computational chemistry needs to narrow en route to reach the currently elusive goal to make docking work for the right reasons. We discuss in detail a path to make electrostatics drastically more realistic, based on novel ideas, some partially implemented.

Fragment-based strategy in drug design involves the initial discovery of low-molecular mass molecules. Owing to their small-size, fragments are molecular tools to probe specific sub-pockets within a protein active site. Once their interaction within the enzyme cavity is clearly understood and experimentally validated, they represent a unique opportunity to design potent and efficient larger compounds. Computer-aided methods can essentially support the identification of suitable fragments. In this review, available tools for computational drug design are discussed in the frame of fragmentbased approaches. We analyze and review (i) available commercial fragment libraries with respect to their properties and size, (ii) computational methods for the construction of such a library, (iii) the different strategies and software packages for the selection of the fragments with predicted affinity to a given target, and (iv) tools for the in silico linkage of fragments into an actual high-affinity lead structure candidate.