Combinatorial Chemistry & High Throughput Screening - Volume 13, Issue 5, 2010

First and foremost, I extend my sincerest appreciation to Richard B. van Breeman, the Editor-in-Chief for the past dozen years, for his hard work and dedication in heralding Combinatorial Chemistry & High Throughput Screening to a highly respected place in the annals of publications related to biomolecular screening. As he aptly said in his editorial on the tenth anniversary of this publication, CCHTS occupies ‘a unique position in the peer reviewed literature by focusing on the publication of review articles and original research papers in combinatorial chemistry, high throughput screening, and the interface of these related fields. … no other journal specializes in this combination of topics'. The tireless dedication of the Regional Editors and the Editorial Advisory Board Members has contributed immensely in the global reach of CCHTS. It is Richard van Breeman who guided the journal during its formative years, and it is his leadership that made the journal what it is today. Richard has passed on the baton to me, and I am incredibly honored to carry the mantle forward to steward CCHTS to greater heights, and I would say, from a solid foundation laid by him and the founding editor, John Pezzuto. I have been a student of high throughput screening from its inception, which include my involvement in the development of the first microplate-based enzyme- and cytotoxicity assays, design of the first Zymark robotic screening platform for new lead discovery, institution of the first principles of chemical library screening for new probe discovery, and was one of the founding members, as a representative of FMC corporation, of the Society for Biomolecular Sciences. It is my sincere hope that my past experience has prepared me well to lead CCHTS to the pinnacle of journals dealing with drug discovery. Towards achieving this goal, we have expanded the publication areas of CCHTS into four clearly defined sections: Combinatorial Chemistry (with more emphasis on medicinal chemistry), High Throughput Screening, Chemoinformatics (new) and Laboratory Automation and Compound Management (new), each headed by eminently qualified, recognized experts in the respective fields. The Regional Editors from the Americas, Europe and Asia will work very closely with the Section Editors in expanding the global reach of the journal. The Section Editors will have the ultimate responsibility for their respective sections. According to Thomas Webb, Section Editor for Combinatorial/Medicinal Chemistry, CCHTS will have a much more substantial focus on the integration of medicinal chemistry into combinatorial chemistry. A major recent trend in combinatorial chemistry is the development of multidisciplinary approaches to the design and synthesis of protein target-oriented libraries, also called focused libraries. Combinatorial chemistry itself has technically evolved from a focus on large libraries prepared via solid-phase synthesis to smaller elegant libraries prepared primarily via solution-phase parallel synthesis. Compound libraries are valuable for their information content. The factors that influence the quality of the information content include relevance (target focus), structural breadth (relative diversity), applicability (physical properties), fidelity (purity) and practicality (synthetic tractability). In order to design compound libraries that are best suited to chemical biology and high throughput screening the goal should be to design and apply libraries with the highest relevance for the specific project at hand. Integration of computational and chemoinformatic methods and medicinal chemistry expertise has become essential in the library design process in order to design libraries with the greatest possible information content. The integration of these methods into the design and synthesis of high quality novel target-oriented libraries is still limited primarily only by the imagination, creativity and skill of the chemists who are involved in this part of the process. The past decade has taught us that high throughput screening technologies are a key component of effective drug discovery, but that they are put to the best advantage when combined with informatics-based technologies that establish sampling studies that optimize the likelihood of elucidating important knowledge, according to Gerry Lushington, Section Editor for Chemoinformatics. For example, combinatorial chemistry is unlikely to benefit discovery efforts if the resulting products are either indistinguishable from chemotypes for which comprehensive SAR is already available, or if the products have little chance of demonstrating pharmacologically relevant activity. Similarly, high throughput screens are most effective when they either sample diverse biologically relevant chemical space that has not yet been explored for a given target, or when they have been trained to divulge full SAR for chemical space regions that have yielded preliminarily promising but inadequately characterized bioactivity profiles. Informatics tools provide an excellent, well-validated basis for intelligent specification of chemical synthesis efforts and screening campaigns and thus play important roles in many of the steps underlying modern discovery campaigns. Careful application of existing informatics techniques and development of new algorithms that extend the accuracy and scope of in silico analysis are thus considered to be a key technology within the field of high throughput drug discovery applications, and hence the new section on Chemoinformatics.

The identification of clones expressing high levels of recombinant protein in Pichia pastoris is usually dependent upon SDS-PAGE, Western blotting, or bioactivity-based assays that are labor and time-consuming. We describe a rapid method that images green fluorescence protein (GFP) of individual P. pastoris clones transformed with vectors that express the proteins as GFP C- terminal fusion. In this report we have used the system to monitor expression of three proteins from Venturia inaequalis. Culture plates containing individual colonies were imaged on a Fuji LAS-3000 system and the intensity of fluorescence of GFP [Mean Gray Value (MGV)] of each colony recorded. Two common variables, the time course of expression and induction temperature were also optimized using this method. The results show that colonies with high levels of GFP fluorescence can be successfully used to identify, at an early stage, colonies expressing high levels of recombinant proteins. This correlation can be used to monitor the conditions for optimization of the expression and accumulation of extracellular recombinant protein in medium and to identify fractions containing GFP-tagged recombinant proteins during protein purification.

DNA is a valid drug target for development of target-based therapeutics against cancer. Screening DNA-targeted anticancer drugs is a key process for the research and development of new anticancer drugs. The traditional anticancer drug screening methods, including animal experiments and cell-based screening assays, have unavoidable drawbacks. In this contribution, the new instrument-based screening assay for DNA-targeted anticancer drugs in vitro using resonance light scattering (RLS) technique was proposed. The experiments suggested that the increment of RLS intensity was directly proportional to the antitumor effect of anticancer drugs. Therefore, it was intuitive to obtain the sequence of the antitumor activity of four anticancer drugs without data processing as follows: mitoxantrone (MIT) > pirarubicin (PIR) > daunorubicin (DAU) > doxorubicin (DOX) by RLS screening spectra. Moreover, the apparent equilibrium constant (K) was 1.23 x 104, 2.22 x 104, 4.66 x 104 L/mol for DOX, DAU, and PIR, respectively. The inhibitory concentration 50 (IC50) was 0.148, 0.102, 0.025, 0.013 μmol/L for DOX, DAU, PIR, MIT, respectively. Therefore, the antitumor effect of four drugs was as follows: MIT > PIR > DAU > DOX, which was in good agreement with the result obtained from RLS screening assays. The mechanism between DNA and anthracycline drugs was investigated using UV-vis spectroscopy, fluorescence spectroscopy, and electrophoresis experiments. The proposed assay is a rapid, intuitive, and easy-to-conduct bioassay with good accuracy and reproducibility.

A library of 117 ligands was combined with three transition metals Ru, Rh and Ir and screened with three different operating conditions for the asymmetric H-transfer reduction of acetophenone into phenylethanol. The combinatorial approach was based on evolution of a first library containing 60 ligands. For the evolution, operators such as replication, regression, cross-over and mutation were used. The study was performed with a XYZ robot and fast chiral GC analysis. Over only 4 generations, the average targeted criterion, enantioselectivity, was increased from 20% to ca. 80% for the 4th generation. The best results provided enantiomeric excess up to 93%.

Pharmaceutical companies are being forced by market competition to find new ways of novel drug target screening instead of traditional methods. The completion of human genome sequencing has provided a flood of new information that might help identify potential drug targets. Finding promising novel drug targets from this flood of information remains challenging. For de novo target screening, the interactions between a drug and cellular components must be comprehensively characterized for better understanding of the pharmacological activities of the drug. The multidisciplinary chemistry-based functional proteomics can be used to elucidate the interactions, because it provides a method to focus initial new drug target identification towards proteins that are more easily validated and most likely to be effective, thereby creating a higher potential for success. This review covers major chemistry-based functional proteomic approaches and highlights their recent advances in applications for novel drug target screening.

In the one-bead-one-compound (OBOC) combinatorial method, compounds are constructed on bead resin via split-mix library synthesis such that multiple copies of the same compound are displayed on each bead. These libraries are rapidly screened with enzyme-linked colorimetric, fluorescent, radiometric, or whole-cell binding assays. While fluorescence-based probes are powerful tools in OBOC screening, their utility is greatly limited by the intrinsic fluorescence of many commonly used solid supports, (e.g., TentaGel), residual coupling reagents, and library compounds. To overcome this problem, we topologically partitioned TentaGel resin with a thin Fmoc-protected outer layer and an unprotected inner core. The inner core was derivatized with 3-nitro-tyrosine, followed by random peptide library construction. Spectral scans from a confocal microscope showed a dramatic decrease in the autofluorescence of blank beads and OBOC peptide libraries across a broad range of the optical spectrum. The quenching capacity of 3-nitro-tyrosine was also visualized in fluorescent micrographs. Using biotin/streptavidin as a model ligand/receptor system, we demonstrated a marked increase in visibility of three commercially available fluorescent probes binding to quenched beads, and increased feasibility of using a robust and efficient fluorescence-based, bead sorting platform known as COPAS™. These data show that using 3-nitro-tyrosine as an internal quencher greatly enhances the compatibility of fluorescence-based applications and OBOC combinatorial screening.

High throughput screening (HTS) remains a very costly process notwithstanding many recent technological advances in the field of biotechnology. In this study we consider the application of machine learning methods for predicting experimental HTS measurements. Such a virtual HTS analysis can be based on the results of real HTS campaigns carried out with similar compounds libraries and similar drug targets. In this way, we analyzed Test assay from McMaster University Data Mining and Docking Competition [1] using binary decision trees, neural networks, support vector machines (SVM), linear discriminant analysis, k-nearest neighbors and partial least squares. First, we studied separately the sets of molecular and atomic descriptors in order to establish which of them provides a better prediction. Then, the comparison of the six considered machine learning methods was made in terms of false positives and false negatives, method's sensitivity and enrichment factor. Finally, a variable selection procedure allowing one to improve the method's sensitivity was implemented and applied in the framework of polynomial SVM.

Virtual screening (VS) is presently a key component in the process of drug design and development. VS is normally regarded as the selection of likely drug candidates from large libraries of chemical structures by using computational methodologies. However, the generic definition of VS is significantly wider and may encompass many different methods. This review tries to present a comprehensive overview of the virtual screening process and of its importance in the present drug discovery and development paradigm. Following a focused contextualization on the subject, an introduction to the general types of virtual screening methodologies is presented. The main stages of a virtual screening campaign, including its strengths and limitations, are the subject of particular attention in this review. This analysis is complemented with a careful selection of VS success stories. Finally, a reflection on the future challenges of this promising methodology is drawn.