Combinatorial Chemistry & High Throughput Screening - Volume 12, Issue 10, 2009

The Human Genome Project [1] identified approximately 30,000 genes and subsequent to the completion of this project a subset of these genes, approximately 3,000-10,000, have been identified as critical in the pathogenesis of disease [2, 3]. In a comprehensive review of pharmaceutical industry portfolios Drews and colleagues suggested that there are approximately 500 drug targets currently, but they proposed that there may be as many as 5,000-10,000 therapeutic targets [4, 5]. One caveat in their estimate is that many of these therapeutic targets are currently under-exploited due to the long standing notion that they are intractable to oral, bioavailable small molecules and/or represent non-druggable targets [4, 6, 7]. While many of the currenty marketed drugs continue to be developed for G-protein coupled receptors (GPCR's), nuclear receptors, ion channels and enzymatic target space there is an emerging need to expand the druggable chemical landscape, which will include target space historically perceived as undruggable. This is a lofty goal, however many groups have recently documented their success in therapeutically targeting these historically refractory complex chemical spaces including; protein-protein interactions, protein-DNA interactions and/or macromolecular complexes through peptidomimetic and/or small molecule approaches. These seminal findings have opened up new and promising avenues for drug discovery initiatives as several groups continue to push the boundaries of what is considered a druggable target. In this issue of Combinatorial Chemistry and High Throughput Screening, we provide an overview of the utility and potential of virtual screening efforts which when partnered with structural biology methods including X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy and/or rigorous homology modeling offer a cost effective, streamlined approach for identifying hits and validating leads prior to preclinical and clinical evaluation. Although this approach still faces a number of challenges it is an alternative to traditional drug discovery and development initiatives. Virtual screening, which is commonly defined as structure based virtual screening (SBVS), or ligand based virtual screening (LBVS) approaches are often used alone or in parallel with high throughput screening (HTS) for hit identification and subsequent hit-through-lead optimization strategies. The virtual or in silico approach used is highly dependent upon the structural and/or functional information available for the target and/or benchmark therapeutics. In this issue we have provided an overview of these in silico screening approaches highlighting several success stories where these approaches were used alone or were partnered with novel combinatorial chemistry/peptidomimetic approaches that are important for translating biologically active peptides into small molecule therapeutics. Alone these virtual approaches have their own merit in discovery pipelines; however as many groups continue to study this expanded drug discovery target space it is likely that a comprehensive approach involving all of these tools will be needed to overcome the many barriers for these historically refractory targets. In support of the recent success in these approaches we have provided numerous case studies documenting the efforts of several groups to overcome these obstacles and identify small molecule modulators that selectively target protein-protein interaction interfaces (PPI2), transcription factor- DNA (TF-DNA) interfaces and protein-membrane interfaces. As noted in these examples the ability to selectively target and inhibit the interaction interfaces of PPI2 and/or TF-DNA complexes represents a novel strategy aimed at reprogramming specific gene pathways that are deregulated in many diseases of significant unmet need including; inflammatory disease, cancer and HIV. Although the number of documented success stories continues to rise, the significance of these in silico approaches to targeting novel target space has yet to be fully realized. Small molecule inhibition of these critical interaction interfaces provides a promising paradigm shift in small molecule therapy through unique mechanism of pathway specific regulation.

Cyclic peptides have been often utilized as metabolically stable, conformationally restricted mimics of different kinds of biologically active peptides, including peptide antibiotics, endogenous opioid peptides, integrin inhibitors, peptide hormones, anticancer peptides, and so on. And in particular, cyclic compounds which can mimic important secondary structure elements such as β-turns are of outstanding importance. Since greater chemical and structural diversity are primary features to pursue for finding novel leads for pharmacological and biotechnological applications, we explored the potential utility of the retro-inverso modification. We introduced this modification into the sequence of 13- membered cyclotetrapeptides, which can be regarded as easily available, conformationally stable analogs of cyclotetrapeptides composed of all α-residues. In this paper we describe the synthesis of a selected mini-library of partially modified retro-inverso cyclic peptides as conformationally homogeneous scaffolds for medicinal chemistry applications. The different compounds have been obtained by simple scramble of the same residues. Finally, we discuss the conformational features of such molecules as turn mimics. The comparison suggests that the retro-inverso modification allows a higher degree of three-dimensional diversity than normal peptides.

Advances in protein modeling algorithms and state-of-the-art sequence similarity comparison and fold recognition methods, in combination with growing protein structure information, are facilitating “genome-to-drug lead” approaches in which chemicals are virtually screened against computationally-predicted protein targets. Although the quality of predicted protein structures by homology modeling methods, and thus their applicability to drug discovery initiatives, predominantly depends on the sequence similarity between the protein of known structure and the protein target to be modeled, recent research underscores that this approach can be used to significant advantage in the identification and optimization of lead compounds, as well as for the identification and validation of drug targets. Rational structure-based drug design cycles begin with an iterative procedure that is dependent on the initial determination of the structure of the target protein, followed by the prediction of ligands for the target protein from molecular modeling computation. The structure determination of all proteins encoded by vast genome sequencing efforts appears to be an unrealistic goal with current technologies. Therefore, other approaches based on the development of technology useful for accurately predicting and modeling the structures of proteins have become exceedingly important in certain structurebased drug design efforts. This review provides an overview of the recent method advancements in protein structure prediction by homology modeling and includes an assessment of the application of homology modeling to pharmaceutically relevant questions. In addition, examples of successful applications of homology modeling approaches to genome-to-drug lead investigations are described.

Urokinase receptor (uPAR) is a widely recognized target for potential treatment of cancer. The development of uPAR inhibitors has been going on for over a decade. Despite the identification and validation of many highly potent hits using screening or medicinal approaches, none of them has been moved further along the drug discovery pipeline. The development of uPAR inhibitors exemplifies several challenges now faced by drug discovery. These include 1) hydrophobicity and thus poor bioavailability of the inhibitors from screening approaches; 2) specificity of the inhibitor, where a peptidyl inhibitor causes conformational change of the receptor; 3) species specificity, where some inhibitors developed based on the human receptor do not inhibit the murine receptor and thus cannot be validated in mouse models. The recently determined crystal structures of uPAR in complex with its ligand or inhibitor not only provide the structural insight to understand these challenges but also offer a potential solution for further inhibitor development and thus illustrate the importance of structural information in facilitating drug discovery.

Low molecular weight inhibitors of Protein-Protein Interactions (PPI's) have been identified for a number of different systems indicating the viability of the computer-aided drug design approaches. However, pathways undertaken by researchers in pharmaceutical companies or in academic laboratories are not always clearly defined and the protocols that allow the identification of lead compounds often remain blurry. We will enumerate in this review the main approaches carried out to identify and validate PPI's inhibitors. Emphasis will be placed, in a first part, on issues of particular significance to PPI's such as the problem of identification and validation of interacting sites and on the methods that allow assessing the 3D structure of a targeted complex. On the second part of this review, we will define approaches that allow a rapid identification of hits capable of inhibiting PPI's. We will highlight the problem of the scoring functions and demonstrate that the majority of the functions available to researchers are not especially relevant for PPI's. We will define why consensus scoring can be an alternative and we will propose GFscore, a non linear ranked-by number consensus scoring function as a solution to this problem. We will finally discuss the actual challenges that still remain, particularly the problem of the treatment of the receptor flexibility and of the water molecules at the interface of the Protein-Protein complexes.

Recent advances in combinatorial chemistry (CC) and High throughput screening (HTS) approaches for use in drug discovery have made it possible to synthesize and/or screen large repositories of chemically diverse scaffolds in search of small molecules that disrupt or regulate macromolecular function. Although successful in the discovery of novel therapeutics this approach is both costly and time consuming. In silico computer aided drug discovery (CADD) approaches including; structure based virtual screening (SBVS) or high throughput docking (HTD) and/or ligand based virtual screening (LBVS) are areas experiencing renewed interest both in the pharmaceutical industry and academia. The emerging success of these approaches alone or partnered with HTS platforms in search of, and/or optimization of, novel therapeutic compounds represents a potential approach for the identification of therapies that target novel space. Here we will discuss how LBVS has been and continues to be partnered with HTS in early stage compound identification and/or triage. We will also provide a significant overview of how SBVS when partnered with LBVS can overcome the limitations inherent to each approach when used alone. We will discuss this partnered approach in the context of both traditional drug discovery targets and provide thoughts on its applicability to study novel chemical space including protein-protein and/or other historical intractable interfaces.

Today, computational methods are commonly used in all areas of health science research. Among these methods, virtual ligand screening has become an established technique for hit discovery and optimization. In this review, we first introduce structure-based virtual ligand screening and briefly comment on compound collections and target preparations. We also provide the readers with a list of resources, from chemoinformatics packages to compound collections, which could be helpful to implement a structure-based virtual screening platform. Then we discuss seventeen recent success stories obtained with various receptor-based in silico methods, performed on experimental structures (Xray crystallography, 12 cases) or homology models (5 cases) and concerning different target classes, from the design of catalytic site inhibitors to drug-like compounds impeding macromolecular interactions. In light of these results, some suggestions are made about areas that present opportunities for improvements.

Alan Rigby is currently an Assistant Professor of Medicine at Harvard Medical School and the Interim Director of the Drug Discovery and Target Validation Program in the Center for Vascular Biology at Beth Israel Deaconess Medical Center. His laboratory is located in the Center for Vascular Biology Research, the Division of Molecular and Vascular Medicine in the Department of Medicine at Beth Israel Deaconess Medical Center. Dr. Rigby is a structural and computational biologist with expertise in the application of structural biology, proteomics, and computational approaches to novel drug discovery initiatives. His lab is actively engaged in partnering these approaches for the exploration of novel target and chemical space in a search for novel therapeutics for the treatment of prostate and breast cancer, rheumatoid arthritis as well as other inflammatory diseases. His laboratory is actively funded by the NIH, Prostate Cancer Foundation, Center for D-receptor Activation Research, National Cancer Institute and the Harvard Clinical and Translational Science Center. Dr Rigby is a former member of the Heme SBIR/STTR study section at the National Institutes of Health, he has reviewed research programs for the French and Portuguese government and has presided over the Proteomics panel for Genome Quebec in 2007 and was member in a recent CFI competition (2009). Dr Rigby is also a member of the award committee for Genome Quebec's PRIVAC initiatives and a committee member for the Canadian Government's National Centers of Excellence (NCEs) Panel. As the Founder of a niche Biotechnology Company, Dr Rigby has commercialization and business experience that stems from his role as a panelist, moderator and scientific consultant for TiE Boston functioning as a liaison between academics and New England based venture capitalists. Dr Rigby holds a Honors BSc. in Biochemistry from the University of Western Ontario (1991) and a Ph.D. in Biochemistry (1995) from the same university. He is the North American Editor for Current Computer Aided Drug Discovery and is on the Editorial Advisory Board for Future Medicinal Chemistry. He has published more than 38 scientific papers, a book chapter and currently holds 6 patents with 4 additional patents pending.