Combinatorial Chemistry & High Throughput Screening - Volume 5, Issue 7, 2002

During the last 12 years, antibody combinatorial libraries have provided a new approach for the construction and production of reagents and drugs based on the human monoclonal antibodies. Studies employing antibodies or antibody mimics have become an important part of the explosive growth of proteomics. This places tremendous emphasis on the new approaches for faster library screening, improved methods of selection and evaluation of novel applications. The phage display system, together with its variants of ribosome and bacterial display, is the most extensively used method for the rapid screening of large antibody libraries. However, in the last two years the need to improve selection methods together with a complex patent situation regarding the phage display system, has also directed research towards the possibility of performing antibody selection by colony filter screening. Here, we summarise the results obtained by these different methods of selection comparing their efficacy and advantages.

We outline a joint academic / industrial (CNRS / AVENTIS) functional genomics project aiming at the discovery of new anti-bacterial gene targets. Starting from all publicly available bacterial genomes, a subset of the most evolutionary conserved protein-coding genes has been identified. We retained genes with clear homolog in E. coli and at least one gram-positive bacterium among B.subtilis, M. tuberculosis, L. lactis or S. pyogenes. This subset was further reduced to genes encoding non-membrane proteins of unknown or hypothetical functions. The 221 E. coli Open Reading Frames (ORFs) identified through this comprehensive bioinformatic analysis are now submitted to a systematic 3-D structure determination protocol including cloning, protein expression and purification, crystallisation and X-ray diffraction. Our strategy was designed to focus on promising wide-spectrum targets as well as original biochemical pathways. Bioinformatics is used throughout all phases of project, including the initial large-scale comparative genomics analyses, the purification / expression and crystallisation stages for the detection of helpful sequence-specific features (e.g. cofactor binding motifs, non-structured N- or C- term extremities, etc...), and finally for the interpretation of the structures in conjunction with multiple sequence alignments for the identification of key residues, interaction areas on molecular surfaces, and overall function predictions.

Flow-injection NMR is a very attractive technique for identifying weak but specific ligandmacromolecule interactions. The low-sensitivity of this method is, however, a major concern which hinders its widespread application to large-scale screening. Focussing on RNA-targeting ligands, we have shown in this report that it is possible to design a peptide library that will simultaneously satisfy a number of physicochemical restraints and sample the sequence space in an “optimal” fashion. This library, which contains less than 200 peptides, covers most of the allowable tripeptide sequences in a non-redundant fashion. In addition, for almost any allowable tetrapeptide sequences, the library contains a close neighbor that will differ at the most by a single conservative replacement. A subset of this library was actually synthesised and used in a preliminary screen against human tRNALys3, the primer of HIV-1 reverse transcriptase.

This paper describes the applications which are used to automate the analysis of encoded combinatorial libraries. Commercial packages from MDL, Oracle and Agilent are linked with application software written in C / C++, in Microsoft Access and in ChemStation macro language. Encoding correspondence lists for each of up to three synthetic steps are conveniently associated with building block lists using the first application, CodeGen.The second application Decode allows the user to identify the individual beads picked onto a 96-well plate and the pool number for each bead. The decoding chromatography data for each well is then loaded into the program. The chromatography data is used to identify the tags used in the synthesis. Along with the building block information from ISIS / Host, the building block used in each step of the synthesis can be identified. A third routine, Code-to-Structure, takes the coded library building blocks and creates the connection table in ISIS for each structure found by the decode program.For quality control of encoded library synthesis, the decoded structures on a set of beads is compared to the LC / UV / MS data for the ligand cleaved from the same bead. CAPTURE, a GlaxoSmithKline proprietary application, is used to display and analyze the decoded structures and associated mass spectral data. This application uses simple isotopic composition and electrospray ionization rule sets to predict mass spectra and judge the concordance of a structure- mass spectrum data set. An ancillary program, EIC, is used to extract predicted single ion chromatograms from the full scan LC / MS data.

A library of 72 compounds related to N- [4-(benzyloxy) benzoyl]alanine (I) was synthesized, prepared and screened for α-glucosidase inhibitory activity. Four compounds showed potent inhibition, six compounds moderate inhibition, and 16 were weak inhibitors. One compound, N- [4-(benzyloxy) benzoyl] serine, was found to be a potent inhibitor of α-glucosidase with 100% inhibition at 1μM. This inhibitor was at least five times more potent than the lead compound I.

Peptides that consist of 19 residues with random sequences (X19) were considered to deliver antigenic stimuli to CD4T cells. When IL-4, IL-7, IL-9, IL-15 and agonistic Ab to CD29 were co-cultured with single peripheral CD4T cells in the presence of X19 and feeder cells, T cells exhibited clonal expansion. These T cell clones showed heterogeneous proliferation patterns against KGXXXXXXXXXGK-based and KGXXXXXXXXXGKGKK-based combinatorial peptide libraries. Pattern-match search on one of the T cell clones resulted in peptide ligand candidates, one of which induced proliferation, as did protein molecules carrying the corresponding sequence. Combinatorial chemistry was useful in determining not only peptide ligands but also peptide superagonists. For this purpose, use of reverse-phase hydrophobic interaction chromatography and mass spectrometry analysis was efficient. Detailed methods are described in the paper.

A simple and fast one-pot method using microwave irradiation for the synthesis of a number of small libraries of diverse heterocycles is described in this paper. The two-step one-pot method includes the formation of alkylaminopropenones or alkylaminopropenoates in 5 min at 180 °C and a subsequent treatment with dinucleophiles for 3 to 5 min at 150 °C to 180 °C to form a variety of biologically interesting heterocycles in a cascade-type reaction. The combination of combinatorial chemistry and microwave-assisted synthesis was found to be very efficient.

A general procedure for the synthesis of a large variety of compounds comprising an α, β-unsaturated carbonyl functionality was developed. The use of one-pot cascade synthesis with (triphenylphosphora-nylidene)ethenone as a versatile reagent for various formations including heterocycles of different ring sizes and unsaturated amides in combination with microwave dielectric heating is described. The method was used to synthesize a small library of unsaturated amides.

Aqueous solubility is a critical physicochemical property and must be addressed early during drug discovery research. Due to the difficulty in accurately predicting aqueous solubility in silico, high throughput experimental determination of aqueous solubility is in great demand. This study evaluates a method using a multi-wavelength UV plate reader and disposable 96-well UV plates for fast solubility determination. It was demonstrated that this method has the sensitivity and reproducibility to effectively determine solubility as low as 1 μM. Excellent correlation (R>0.97) was observed between the solubility determined using the UV reader method and the HPLC method over the range of 1-1000 μM for a diverse set of pharmaceutical compounds. In addition to excellent sensitivity and reproducibility, the UV plate reader method also offers the flexibility of being able to determine thermodynamic solubility in the presence or absence of dimethyl sulfoxide, which is a solvent widely used for combinatorial compounds during high throughput screening.