Current Topics in Medicinal Chemistry - Volume 7, Issue 15, 2007

Melanin-concentrating hormone (MCH) is a neuropeptide that plays a role in multiple physiological processes, including the regulation of feeding behavior and energy balance. Since the discovery of the first MCH receptor in 1999, referred to as MCHr1, reports of mice deficient in both the MCH peptide as well as the receptor have been disclosed. Genetically altered mice lacking the gene encoding MCH are hypophagic, lean, and maintain elevated metabolic rates while those lacking the gene encoding MCHr1 maintain elevated metabolic rates yet remain lean despite hyperphagia on a normal diet. While some differences are apparent in the two phenotypes, the possible result of other functional peptides being derived from the prepro MCH gene, both results support a role for MCHr1 antagonism in the treatment of obesity. This and other data have since fueled an aggressive effort by the pharmaceutical industry to identify small molecule antagonists that are suitable for the mentioned indication. Since the first disclosure of Takeda's T-226296, which appeared in print in 2002, a large number of diverse small molecule antagonists has been reported. Many of these have demonstrated dose dependent weight loss and/or food intake in various animal models. Interestingly, some reported compounds cause weight loss while not affecting food intake, which raises questions regarding the role of increased energy expenditure. Finally, MCHr1 has been implicated in behavioral roles, and significant effort has been expended in deriving utility from MCHr1 antagonists as anti-depressants and anxiolytics as well. This collection of accounts will focus on the original indication for MCHr1 antagonists. Seminal work preceding the development of small molecule antagonists was provided by the identification of receptor selective peptide agonists and antagonists. These peptides were then used to delineate a number of functions that were dependent on receptor activation or antagonism. In the first contribution, Maria Bednarek, of Merck Research Laboratories, provides a synopsis of efforts to identify these peptides. Additionally, a description of the SAR is included, along with the results of several in vitro and in vivo studies. This paved the way for small molecule efforts at the same company, and Robert DeVita describes some of these efforts in the subsequent manuscript. By first identifying novel 2-aminoquinoline hits from a high-throughout screen (HTS), DeVita's group was able to optimize the hit compounds into potent leads that were suitable for mechanism of action studies in rats. A similar hit-to-lead story emerged from the Schering Plough group, which has published a multitude of high quality manuscripts describing their efforts in the MCHr1 field. A portion of their strategy relied on the identification of suitable replacements for a biphenyl aniline, a moiety with known mutagenic properties. As described by Mark McBriar, this led to the identification of a highly potent and novel class of bicyclohexyl ureas. In order to optimize these compounds for brain penetration and, ultimately, in vivo efficacy, the Schering group relied on a medium throughput ex vivo receptor occupancy assay. This method allowed for the identification of compounds that delivered a considerable reduction in food intake when dosed in diet-induced obese mice (DIO), and the authors were able to clearly demonstrate the correlation between ex vivo binding and longer term in vivo effects. In the fourth contribution, Don Hertzog and Dave Whitty from Glaxo Smithkline provide an overview of their efforts to develop a thieno[3,2-d]pyrimidinone class of MCHr1 antagonists. They describe the systematic optimization of an HTS lead, leading to the identification of a compound class with excellent receptor affinity as well as brain penetration in mice. Additionally, members of this class are among the most efficacious of any MCHR1 antagonists reported, with up to 17% weight loss observed upon once daily oral dosing in DIO mice......

The melanin-concentrating hormone receptor 1 (MCH-1R) has been recognized as a receptor which mediates effects of the endogenous melanin-concentrating hormone (MCH) on appetite and body weight gain in rodents. In the last several years, a number of hMCH analogs have been designed which were potent and selective ligands for hMCH-1R. These peptidic agonists and antagonists have served as research tools in animal studies that showed a key role of the MCH-1R in the development of obesity and proved that MCH-1R antagonism can produce anti-obesity effects in rodents.

Structure-activity relationships of a 4-aminoquinoline MCH-1R antagonist lead series were explored by synthesis of analogs with modifications at the 2-, 4- and 6-positions of the original HTS hit. Improvements to the original screening lead were made by addition of lipophilic groups at the 2-position and biphenyl, cyclohexyl phenyl and hydrocinnamyl carboxamides at the 6-position. Viable modifications of the 4-amino group were limited and did not allow further optimization of the physical-chemical properties of this class of compounds. Transposition of the 4-amino group to the 2-position of the quinoline core structure provided the 2-aminoquinoline lead class with similar MCH1R binding affinity. A series of 2-aminoquinoline compounds was prepared and evaluated in MCH-1R binding and functional antagonist assays. Small dialkyl, methylalkyl, methylcycloalkyl and cyclic amines along with 3-substituted pyrrolidines were tolerated at the quinoline 2-position. The in vivo efficacy of compound A was explored and compared to that of a related inactive compound B to determine their effects on food intake and body weight in rodents. The biological activities of this matched active - inactive pair provide in vivo proof of concept in rodents that antagonism of MCH1R by a 2-aminoquinoline MCH1R antagonist which led to a reduction of food intake in an acute feeding assay paradigm.

Melanin concentrating hormone (MCH) is a cyclic, nonadecapeptide expressed in the CNS of all vertebrates that regulates feeding behavior and energy homeostasis. The MCH-1 receptor (MCH-R1) has been identified as a key target in MCH regulation, as small molecule antagonists of MCH-R1 have demonstrated activity in vivo. Herein, we chronicle our efforts to optimize a hit identified via high throughput screening of our proprietary compound library. Several challenges such as selectivity over other receptors, toxicity of a potential metabolite and determining receptor occupancy via a medium throughput assay will be reviewed.

The pursuit of MCH R1 antagonists for the treatment of obesity has become an active area of research for many pharmaceutical companies. The evidence supporting the use of MCH R1 antagonists for the treatment of obesity is ample, and the recent demonstration of MCH R1 antagonists’ efficacy in animal models of obesity has served to augment earlier studies involving MCH peptide and transgenic animals. We report herein our search for MCH R1 antagonists from the discovery of a biphenyl amide by high throughput screening, through the optimization of the biphenyl amide to a series of constrained aryl-substituted thienopyrimidinones, and extending the application of the thienopyrimidinone substructure to other series of MCH R1 antagonists. Importantly, these MCH R1 antagonists have demonstrated efficacy in animal models of obesity through once-daily oral administration at low doses.

The discovery of small molecule melanin concentrating hormone receptor (MCHr1) antagonists as novel therapeutic agents for the treatment of obesity has been actively pursued across the pharmaceutical industry. While multiple chemotypes of small molecule MCHr1 antagonists have been identified and shown to deliver weight loss in animal models of obesity, many of these lead compounds have been found to cross-react with the hERG channel and/or demonstrate deleterious effects on cardiovascular hemodynamic parameters. This review describes an approach to rapidly identifying safer MCHr1 antagonists by placing assays to assess cardiovascular safety early in the lead optimization compound prioritization process. Ultimately, despite putting significant effort toward the discovery of a MCHr1 antagonist for the treatment of obesity, we were unable to deliver a candidate compound that attained an acceptable therapeutic index (TI = 30-100) in our in vivo models. Our inability to identify a compound with an acceptable therapeutic index was driven by two primary factors: 1) high levels of sustained drug exposure in the brain was required to achieve efficacy; and 2) many small molecule MCHR1 receptor antagonists suffer from receptor cross-reactivity that leads to cardiovascular toxicity at low multiples of their therapeutic plasma concentration.

Computational chemistry has become a pervasive tool for the medicinal chemistry work because the entire process of drug discovery is becoming without question information intensive. The challenge for the chemist today is to navigate through the deluge of information, select compounds for synthesis and create the most informative series of analogs that can be made to optimize the chemical series. Even when knowledgeable scientists may be available to determine the priority of compounds for synthesis, the decision process could be laborious, particularly when large datasets have to be considered. Predominantly, expertise is scarce or unevenly distributed, particularly at the on-set of new projects. The way in which this data intensive paradigm for drug discovery presents a challenge for traditional chemoinformatics has been given considerable attention [1] and present a significant opportunity for research [2]. In very broad terms, computational chemistry techniques are used for modeling or data mining and management. There is some convergence between the two. In recent past, two areas in modeling have received the most attention. One is the problem of virtual screening, which is easy to understand, generates considerable interest. The other is the modeling of metabolic and pharmacokinetic processes. Accurate prediction of binding affinity and binding mode are crucial in drug design, as decisions as to which compounds or compound libraries to evaluate next as a lead evolves towards a drug candidate. Some reviews provide a good overview of the state of the art in these techniques, indicating their limitations and applicability [3-7]. The different algorithms and scoring functions to rank docked compounds have also been extensively reviewed. [8-10], and docking techniques are generating a list of successes that can be impressive depending on the target class [11]. However, efforts to compare the outcomes of the different techniques are sparse. [12-15]. Another area of modeling that has elicited considerable interest in the last few years is the prediction of toxicological, pharmacokinetic or other properties related to preclinical development. [16-19]. Also metabolism with its multiple implications for drug-drug interactions receives continuous attention by the computational teams. [20, 21]. The shift observed in the way in which early drug discovery is carried out has spurred significant activity in this area. Predictions are used to tailor libraries and prioritize compounds. From the calculation of octanol/water partition coefficients to define pharmacokinetic characteristics to modeling of complex biological processes predictions are becoming ubiquitous and means to prioritize directions. Data management and mining issues are likely to increase in complexity for the chemist and new generations of tools are needed to aid in what was done manually in the past. Aligning experimental and in-silico techniques has been an on-going goal [22], but systems biology and high content screening are likely to increase the amount of information handled by even the simplest of projects [23,24]. and seems the most effective way to go in some therapeutic indications [25]. The integration of chemical data with bioinformatics is likely to add yet another dimension to the complexity of the data sets to be handled [26] and the data to be analyzed [27,28]. The breadth of the developments in computational techniques applied to drug discovery is extremely wide and it would not be possible to cover them in a single issue. The areas we chose to cover in this issue are only a small fraction of the innovative computational techniques recently put forward to aid in drug design. We decided to provide a sampling of the breadth in computational applications covering some areas less frequently addressed in review form. We start with coverage of sources of information for the medicinal chemist. Ertl and Jelfs provide an overview of tools available on the internet that can be of use to medicinal chemists, while Southan, Varkonyi and Muresan contrast public and commercial sources of chemical information. Villar and Hansen review how computational techniques are applied to an alternative paradigm for drug discovery, namely fragment based drug design.....

The drug discovery process is supported by a multitude of freely available tools on the Internet. This paper summarizes some of the databases and tools that are of particular interest to medicinal chemistry. These include numerous data collections that provide access to valuable chemical data resources, allowing complex queries of compound structures, associated physicochemical properties and biological activities to be performed and, in many cases, providing links to commercial chemical suppliers. Further applications are available for searching protein-ligand complexes and identifying important binding interactions that occur. This is particularly useful for understanding the molecular recognition of ligands in the lead optimization process. The Internet also provides access to databases detailing metabolic pathways and transformations which can provide insight into disease mechanism, identify new targets entities or the potential off-target effects of a drug candidate. Furthermore, sophisticated online cheminformatics tools are available for processing chemical structures, predicting properties, and generating 2D or 3D structure representations - often required prior to more advanced analyses. The Internet provides a wealth of valuable resources that, if fully exploited, can greatly benefit the drug discovery community. In this paper, we provide an overview of some of the more important of these and, in particular, the freely accessible resources that are currently available.

The last two years have seen a dramatic expansion in public cheminformatics, as exemplified by the approximate five-fold growth of PubChem from over 50 contributing data sources. Consequently, medicinal chemists who were hitherto limited to commercial databases now also have access to public sources that they can download and/or query directly over the Web. The range of public sources, particularly where they link out to structured bioinformatic and biological data, already offer utilities that have no commercial equivalent. This work reviews compound content comparisons between selected public and commercial databases that capture bioactive content. We focused particularly on those that specify relationships between compounds and their protein targets. Our stringent filtering produced lower unique compound numbers than those reported for individual databases and thereby facilitated standardised comparisons of content. The resultant matrix shows the pairwise comparison of each database and selected subsets. Overall, this showed an unexpected degree of non-overlap, thereby emphasising the complementarity gained from combining public and commercial sources. This conclusion is supported by a Venn-type analysis of GVKBIO, WOMBAT (both commercial) and PubChem (public). These databases show not only overlap but also unique bioactive content in each case because of their different strategies for source selection and data collection.

Fragment based drug discovery is gaining acceptance as a complement to other more established techniques to identify leads and optimize drug candidates. In this review we illustrate areas where fragment based drug discovery has had an impact and point to some examples that show how fragment based analysis is being applied to new arenas. The traditional uses of computational methods in fragment based for lead discovery and optimization and for risk assessment are briefly summarized. The application of fragment analysis for the definition of bioisosteric replacements are discussed together with techniques to characterize the diversity of chemical libraries based on fragment distribution.

We review recent advances in computer modeling of molecular shape in drug discovery. We summarize the ways of representing shape computationally, discuss the various means of aligning molecules and shapes, consider the various ways of scoring similarity of shapes, and describe the ways in which these shapes can be used to construct molecular descriptors. Finally, we evaluate the success of these methods to date, suggest when they are best applied, and provide our recommendations for the direction of future work.

Membrane bound P-glycoprotein (Pgp) acts as an active transport pump. It plays a major role as a cause of multidrug resistance (MDR) and acts as a component of the blood-brain barrier. Pgp transports a wide variety of structurally unrelated compound from the cell interior into the extracellular space. Recent molecular modeling efforts, mostly in homology modeling and QSAR studies, have brought some understanding to the interactions between the protein and the drugs at the atomic level. We review the recent developments from the point of view of methodology.

Many efforts are currently being made to connect small molecules to target proteins by extracting pharmacological data from bibliographic sources and storing them in annotated chemical libraries. Here, small molecules are further connected to biological pathways, with particular focus to pathways involving members of the nuclear receptor family. The results bring to light the relative importance for molecules on gaining selectivity at the target level, when the target has an intrinsic promiscuity at the pathway level, and highlight the implications for drug discovery to address current challenges related to poor drug efficacy and toxicity. Details on the main limitations encountered during the molecule-to-target-to-pathway annotation process are also discussed.

Nanobiology is a fast-emerging discipline that brings the tools of nanotechnology to the biological sciences. The introduction of new techniques may accelerate the development of highly specific biomedical treatments, increase their efficiency, and minimize their side effects. Introducing foreign bodies into the complex machinery of the human body is, however, a great and humbling challenge, as past experience has shown. In order for nanobiology to reach its full potential, we must devise a means to alter the properties of nanoparticles, as expressed in the human body, in a predictable manner. Computer-aided methods are the natural option to speed up the development of these technologies. Yet, the procedures for annotation and simulation of nanoparticle properties must be developed and their limitations understood before computational methods can be fully exploited. In this review we will compare the state of development of nanoscale simulations in the biological sciences to that of the computer-aided drug design efforts in the past, tracing a historical parallel between both disciplines. From this comparison, lessons can be learned and bottlenecks identified, helping to speed up the development of computer-aided nanobiodevice design tools.

Full text available