Current Topics in Medicinal Chemistry - Volume 4, Issue 6, 2004

HIT-TO-LEAD: DRIVING FORCES FOR THE MEDICINAL CHEMIST Drug discovery is a succession of design, synthesis and selection steps. The understanding at a molecular level of the biological processes underlying many diseases has lead to a miniaturization of the bioassays and a dramatic increase in throughput, at the initial step of selection. During the shift from animal testing to target-based assays much relevant information was lost for the medicinal chemist. Although activity data was easier to use and allowed quantitative-structure-activity relationships, they were also less relevant and even sometimes misleading. In the meantime, the increase in throughput of initial testing has also created a unprecedented demand for new compounds. To palliate this apparent shortage, combinatorial chemistry has been a the major source of leads. Mechanically, numbers had become a driving force not only for the screener but also for the chemist at the detriment of compound quality and druglikeness. Thus in the late '80s, better understanding of diseases has paradoxically disabled drug discovery. In the mid 90s it has been realized that numbers and pure diversity are not the only keys of success. The linear process whereby drug discovery started with the selection of compounds binding to the chosen target driven by affinity before animal testing has not been successful. The drug discovery process is now seen less linearly than previously. Discovery teams are now trying to optimize all possible inputs in concerted efforts: quality and (directed) diversity of compounds, as well as quality and diversity of information to drive changes in compound structure. The early identification of the most suitable active compound is now a common matter for chemists, molecular biologist and pharmacologists, who work in cross-disciplinary teams. One of the main tasks of hit-to-lead is to bridge the gap between on-target activity and efficacy in a whole organism. As exemplified by the meeting of the Medicinal Chemistry Division of the American Chemical Society in New Orleans (March 2003), hit-to-lead activities are clearly identified as a critical point in the drug discovery process. An important session of the meeting focused on of lead generation following high throughput screening. This issue tends to overview several key aspects of the chemical sciences underlying the success of hit optimization. Another strategy, described by Roger Crossley, relies on the design of targeted libraries using structural information of the target and statistical analysis of ligand binding data. Researchers from Biofocus apply the concept of informed diversity to the discovery of leads in the field of GPCRs. In another paper, Dragos Horvath from Cerep, describes the neighborhood behavior of bioactive compounds and rationalize the concept of similarity that underlies the work of every medicinal chemist in the course of hit optimization. Finally, Frank Landolt from Devgen develops the impact of early patenting of targets, screening methods and compounds in the drug discovery process, and especially at the hit-to-lead step. The techniques and concepts described in this issue, as well as related ones that could not been covered here, will surely provide valid hits from screening and reduce significantly hit-to-lead and hopefully hit-to-drug attrition.

Nicotinic Acetylcholine Receptors: Ligand Design Issues Nearly 100 years have passed since Langley suggested the existence of nicotinic acetylcholine receptors (nAChRs) [1]. Only in the past decade has enough evidence accumulated to demonstrate, beyond the anecdotal, the potential of nAChR-based therapeutics to address significant unmet medical need. Areas of intense academic and pharmaceutical industry focus include neurodegenerative diseases and neuropsychiatric disorders. Not only do nAChRs continue to provide challenging research opportunities at the molecular and pharmacological level, but they have also become an emerging theme in therapeutics where modulation of complex temporal phenomena takes precedence over indiscriminate activation or inhibition sought by traditional pharmacotherapies. In this special section, S. Breining reviews recent publications describing synthesis of nAChR ligands based on variants of natural products and / or first-principles design. As expected, this review highlights recent focus on the synthesis of ligands containing the aza[2.2.1]bicycloheptane system of epibatidine. Other highlights include numerous stereospecific routes, nicotinoid synthesis using the Hamaguchi-Ibata reaction and intramolecular Heck arylation. R. Glennon and coworkers, discuss current thinking in nAChR molecular recognition in light of established pharmacophore models and emerging SAR. Vector pharmacophore models proposed by Tønder (essentially pharmacophore models that include site points) are envisaged to hold great promise in explaining many of the contradictions in nicotinic SARs. Glennon suggests that the vector pharmacophore could be refined by structure-affinity data from different structural series. A contribution by T. Grutter encapsulates contributions of the Changeux group in using the Molluskan acetylcholine binding protein structure for developing nAChR binding-domain models. Results of docking experiments may explain why different allosteric states exhibit distinct ligand affinities; this work also indicates that a specific E-loop residue is critical in determining subtype specificity in the desensitized state. So-called ‘gain of function’ and ‘loss of function’ mutants are also discussed as tools for a new generation of research on the biophysical mechanisms of allosterism. In closing, I would like to express my appreciation for the authors that contributed to this issue. Their contributions exemplify what an exciting time it is in the world of nicotinic acetylcholine receptor research and its role in health and disease. REFERENCE [1] Langley, J.N. J. Physiol. 1905, 33, 374.

The goal of this paper is to review the variety of approaches adopted to improve lead generation, and make the process easier for the chemist, faster and more likely to succeed in later phases of drug development. Our analysis shows that successful lead generation requires not only an accurate definition of the needs (to define the most relevant assay protocols and readouts), but most of all a good hit as a starting point. It also appears that teams where techniques are combined are more successful in that difficult game.

The screening of large libraries in order to obtain hits for receptors of interest has been the mainstay of drug research for some time now. It is increasingly being recognised that this is a relatively inefficient way to achieve this end and the screening of libraries either designed or selected to hit particular targets is rapidly becoming the method of choice. The advantages in terms of success rate to achieve viable lead series are magnified by the cost and time savings achieved by screening more carefully selected groups of compounds. A number of approaches have been used for the design and production of such libraries or methods for selection of such focused sets from larger diverse collections. These range from combinatorially produced ligand-mimetic approaches through pharmacophore-based design to those methods based on statistical selection techniques. Most recently, progress in chemogenomic approaches has thrown new light on the relationship between receptor sequence and compounds that interact at particular receptors and this is also having an impact on the design of targeted libraries.

This paper reviews the main efforts undertaken up to date in order to understand, rationalize and apply the similarity principle (similar compounds => similar properties) as a computational tool in modern drug discovery. The best suited mathematical expression of this classical working hypothesis of medicinal chemistry needs to be carefully chosen (out of the virtually infinite possible implementations in terms of molecular descriptors and molecular similarity metrics), in order to achieve an optimal validation of the hypothesis that molecules that are neighbors in the Structural Space will also display similar properties. This overview will show why no single “absolute” measure of molecular similarity can be conceived, and why molecular similarity scores should be considered tunable tools that need to be adapted to each problem to solve.

The extraordinary pharmacology of nicotine and epibatidine have indicated the potential for nicotinic acetylcholine receptor (nAChR) ligands to serve as a new therapeutic class for a host of CNS disorders. Many such ligands are natural products, or analogs thereof, which represent a significant challenge to the synthetic chemist. Synthesis of such molecules often serves as a showcase to demonstrate the potential of newly developed methodology. This synthetic challenge coupled with the promise of pharmacological activity in compounds possessing the nicotinic pharmacophore has stimulated a great deal of synthetic activity over the last five years. The present report provides an overview of novel synthetic methodology occurring during this period directed toward the synthesis of compounds with presumed affinity for the neuronal nAChR. Syntheses chosen for review here represent the major efforts toward molecules such as epibatidine analogs, anatoxin-a, nicotine and related alkaloids, conformationally constrained nicotine derivatives, cytisine and methyllycaconitine (MLA).

Several pharmacophore models were previously formulated to account for the actions of nicotinic acetylcholinergic (nACh) agents. Most of these models were developed without express consideration of specific radioligand binding data because such data were not available at the time the models were described. In this review, the ability of these models to account for the binding of nicotinic agents at α4β2 nACh receptors (or rat brain receptors for which α4β2 receptors are the major component) is assessed. It seems that none of the early models can adequately explain the binding of these agents as a group. Furthermore, different series of nicotinic agents behave differently depending upon the nature of terminal amine substituents and the spacer that separates the amine from the pyridine ring. A region of bulk tolerance has been identified that accommodates substituents on some nicotinic ligands, but not the same substituents at seemingly corresponding locations of others. The concept of multiple modes of binding has been previously raised and, clearly, cannot yet be discarded. Nevertheless, new vector models seemingly provide a better picture of nACh receptor binding and account for many of the shortcomings associated with the earlier models.

The atomic determination of the acetylcholine binding protein (AChBP), a molluscan cholinergic protein, homologous to the amino-terminal extracellular domain of nicotinic receptors (nAChRs), offers opportunities for the modeling of the acetylcholine binding site and its ligands. Recently, we constructed three-dimensional models of the Nterminal part of nAChR and docked in the putative ligand-binding pocket, different agonists (acetylcholine, nicotine and epibatidine) and antagonist (snake α-bungarotoxin). These hypothetical docking models offer a structural basis for rational design of drugs differentially binding to resting and active (or desensitized) conformations of the receptor site. These models thus pave the way to investigate, at the molecular level, the exciting challenge of the fast ion channel gating mechanisms by nicotinic agonists.