Combinatorial Chemistry & High Throughput Screening - Volume 13, Issue 3, 2010

Fingerprint representations of molecular structure and properties were among the first computational tools for similarity searching and are widely applied to this date, which is in part due to their computational efficiency and ease-ofuse. Moreover, despite their simplicity, 2D molecular fingerprints have been surprisingly successful in identifying novel active compounds. However, properly applying 2D fingerprint representations and similarity metrics in the search for active molecules is much more complicated than one might assume. For example, no generally applicable similarity functions and threshold values exist that reliably indicate a level of molecular similarity that results in similar biological activity. Furthermore, fingerprint search calculations are known to be biased by molecular size and complexity effects, which often lead to a significant increase in false-positive rates. In this contribution, we will describe known caveats associated with fingerprint similarity searching and strategies to overcome these difficulties including the design of complexity-independent fingerprint representations and similarity metrics.

The miniaturisation of high throughput screening (HTS) assays has been a widely debated and researched strategy with the aims of reducing screening costs and increasing speed while not compromising data quality. In this context, liquid handling technologies continue to be improved. A new and promising development is the emergence of nanolitre dispensers, which are capable of direct compound transfer to assay microplates. In this study, we investigated the miniaturisation of a HTS kinase assay and compared the real life performance of current state-of-the-art air displacement transfer technology (MiniTrak V System) and a capillary based nanolitre dispenser (CyBi-HummingWell). The robustness and effectiveness of the miniaturised assay formats were compared by testing staurosporine to generate dose-response curves and 340 previously identified active compounds. Using the MiniTrak device, assay miniaturisation was achieved from 18 μL to 6 μL in 384-well and 1536-well plate formats. Utilising the nanolitre dispenser, miniaturisation was performed down to 5 μL in 1536-well plates. The Z' factors obtained for each assay format were consistently above 0.5. The data presented here describe the reproducibility of the results obtained with the two transfer technologies and highlight possible issues for hit identification.

High throughput immunoprecipitation studies of transcription factors and splicing factors have revolutionized the fields of transcription and splicing. Recent location studies on Nova1/2 and Fox2 have identified a set of cellular targets of these splicing factors. One problem with identifying binding sites for splicing factors arises from the transient role of RNA in gene expression. The primary role of most splicing factors is to bind pre-mRNA co-transcriptionally and participate in the extremely rapid process of splice site selection and catalysis. Pre-mRNA is a labile species with a steady state level that is three orders of magnitude less abundant than mRNA. As many splicing factors also bind mRNA to some degree, these substrates tend to dominate the output of location studies. Here we present an in vitro method for screening RNA protein interactions that circumvents these problems. We screen approximately 4000 alternatively spliced exons and the entire hepatitis C genome for binding of ASF/SF2, the only splicing factor demonstrated to function as an oncogene. From the pre-mRNA sequences returned in this screen we discovered physiologically relevant ASF recognition element motifs. ASF binds two motifs: a C-rich and a purine-rich motif. Comparisons with similar data derived from the hnRNP protein PTB reveal little overlap between strong PTB and ASF/SF2 sites. We illustrate how this method could be employed to screen disease alleles with the set of small molecules that have been shown to alter splicing in search for therapies for splicing diseases.

The capacity to rapidly and efficiently elucidate a reliable set of disease specific biomarkers is paramount to enable a future of personalized medicine. High throughput screening methods applied to human clinical samples for the discovery of diagnostic, prognostic, and therapeutic targets address this need. Although the ability to analyze either thousands of markers from one sample or one marker from thousands of samples is the current state of high throughput screening, it would be ideal to have the ability to analyze thousands of markers from thousands of samples to expedite the early discovery phase of biomarkers and their validation. This review summarizes the current state of high throughput screening of tissue specimens and discusses its applications. In addition, the rationale, difficulties, strategies, and development of new technologies to address the need for improved high throughput capabilities are discussed.

Using computer-assisted combinatorial chemistry techniques, we have designed a virtual library of antiinfluenza agents, analogs of inhibitor A-315675, containing a novel pyrrolidine core, which effectively inhibits both wild type and common oseltamivir-resistant mutant forms of the neuraminidase (NA) subtype N1 of avian influenza virus H5N1. A target-specific Potential of Mean Force (PMF) scoring function parameterized on a training set of 13 known pyrrolidine-based inhibitors of NA and validated on 3 others was used to predict the N1 inhibition constants for the focused library of A-315675 analogs. Nine virtual hits (best pyrrolidine inhibitors designed in the present study) are predicted to exhibit inhibition constants in the low picomolar range, up to 200 fold lower than the parent inhibitor A- 315675 while displaying favorable predicted ADME-related properties. Proposed small highly-focused combinatorial subsets composed of R-groups most frequently occurring in the 200 most active analogs can be useful as a guide for synthetic and medicinal chemists who are developing a new generation of drugs against the avian influenza virus H5N1 by focusing their attention on this small portion of the chemical space.

The aim of this study was to screen for acetylcholinesterase (AChE) inhibitors from a large chemical library of commercially available compounds. For this purpose, the Ellman's reaction based assay was miniaturized into 384-well plate format, and two modifications of the kinetic protocol were studied with the aim of developing a rapid screening platform that could ensure high efficiency in finding true hits. It was proven that when starting the kinetic reaction by addition of the substrate, better assay performance was achieved and more practical benefits obtained. Using the optimized automated protocol, a chemical library of 56,320 compounds was screened. A total of 350 positive hits were identified and their IC50 calculated. Three highly active compounds were identified with IC50 values close or even lower to physostigmine (< 0.1 μM). The activity towards butyrylcholinesterase (BChE) of these three most active hits was also evaluated. The most active hit (IC50(AChE) = 0.019 μM), was identified as a new inhibitor, belonging to ChemDiv chemical library: (N-[3-(3,5-dimethyl-1-piperidinyl)propyl]-5-ethyl-2-methyl-8-oxo-thieno[2',3':4,5]pyrrolo[1,2-d] [1,2,4] triazine- 7(8H)-acetamid), with no other biological activities reported until now. The interactions of this hit with both cholinesterases were further analyzed using computational docking studies. To our knowledge, this is the largest published screening campaign of commercially available compounds that has focused on finding new AChE inhibitors. The miniaturized 384-well plate format of the Ellman's method was proven to be robust and to perform reliably.

The growing drug resistance of Plasmodia spp. to current antimalarial agents in the quinine and artemisinin families further asserts the need for novel drug classes to combat malaria infection. One approach to the discovery of new antimalarials is the screening of natural product extracts for activity against the formation of hemozoin, a biomineral essential to parasite survival. By mimicking the in vivo lipid-water interface at which native hemozoin is found, hemozoin can be synthesized outside the parasite. In this work, a variety of lipophilic mediators was used to determine the optimal platform for in vitro hemozoin formation and then tested for efficacy in preliminary screens containing crude natural product extracts. The complete optimization and validation of a NP-40 detergent-mediated assay provide a screening template with an expedited 4-hour incubation time and identical IC50 values to those measured from the parasite's native lipid component.

In the family of biologically active heterocyclic templates, quinoxalines, quinoxalinones, and quinoxalinols have received much attention owing to a wide range of their biological activities. As a consequence, approaches to the combinatorial synthesis of these compounds were surveyed, and this review discusses the solid-phase, solution-phase, solvent-free, and microwave-assisted parallel synthesis of quinoxalines, quinoxalinones, and quinoxalinols.