Automation Highlights from the Literature

ACCELERATING DRUG DISCOVERY BY INTEGRATIVE IMPLEMENTATION OF LABORATORY AUTOMATION IN THE WORK FLOW

Over the last decades most pharmaceutical companies heavily invested in robotics and laboratory automation equipment for combinatorial chemistry in order to accelerate the production of large numbers of new organic compounds to be tested for their biological properties. However, optimal integration of automated techniques into the general drug discovery process is of utmost importance to achieve this goal. In a recent publication, M. Nettekoven and A. W. Thomas (Current Medicinal Chemistry, 2002, 9, 2179) describe their experience at Roche with “Accelerating Drug Discovery by Integrative Implementation of Laboratory Automation in the Work Flow”. The authors examine the different stages of the workflow for preparation, purification, and quality assessment of the produced compounds and the laboratory automation equipment used for these different tasks. Besides using standardized devices capable of performing complex chemistry transformations, data handling and tracking for a large number of compounds is extremely important to realize the full value of automation in drug discovery. Several case studies of successful and accelerated drug discovery taken from Roche (A2a antagonists), Aventis (TNF-α release inhibitors), and Glaxo SmithKline (enzyme inhibitor) are discussed.

SYSTEM FOR THE MULTI-PARALLEL SYNTHESIS OF POLYMER BEADS

Resins for solid phase synthesis have been used pretty much unchanged since their introduction by Merrifield in 1963. The aim of a project at the Combinatorial Centre of Excellence at the University Southampton was to develop libraries of new resins with enhanced reaction profiles, improved reaction kinetics, low swelling properties, and good loadings. In order to accelerate the preparation of new polymer resins using standard suspension polymerization protocols, a new system for the multi-parallel synthesis of polymer beads was developed (M. Bradley, Current Medicinal Chemistry, 2002, 9, 2173). This instrument enabled the production of six samples of resin on a 10–20 g scale per sample (allowing up to 18 resin samples per week to be prepared and processed). Overall this has allowed several hundred resin samples to be prepared which make up the most comprehensive set of resins ever made and fully analyzed. Clearly studies of this type are fundamental in allowing improvements in the resins currently being used while enabling broader applications for all types of supported chemistry.

A SIMPLE AND EFFECTIVE METHOD FOR PRODUCING NONRANDOM PEPTIDE LIBRARIES USING COTTON AS A CARRIER IN CONTINUOUS FLOW PEPTIDE SYNTHESIZERS

Since Merrifield's pioneering work on solid-phase peptide synthesis, numerous methods have been developed to accelerate the synthesis of peptide libraries. Different carrier materials and methodologies have been introduced to create high-yield peptide libraries while minimizing the number of coupling steps and increasing the speed of synthesis. A new method for generating nonrandom pep-tide libraries on cotton was developed recently by J. E. S. Wikberg et al. (J. Comb. Chem., 2003, 5, 1). Disks of cotton fabric were chemically modified to enable peptide synthesis. Incorporation of a 6-aminocaproic acid residue handle on the cellulose turned out to be advantageous. Disks were labeled with silver ink, stacked one on top of another in a continuous flow peptide synthesizer column, and simultaneously subjected to automated synthesis procedures. Depending on the sequences to be synthesized, the automatic synthesis procedure was stopped, and the disks were removed from the column, sorted and reapplied to subsequent synthesis steps. In this way, individual peptides could be easily prepared in milligram quantities on each of the cotton disks.

A NEW APPROACH TO RAPID PARALLEL DEVELOPMENT OF FOUR NEUROKININ ANTAGONISTS

Optimal integration of automated systems into the workflow is not only very important in drug discovery but also in the further development and optimization of the synthesis of new potential drug candidates. J. D. Moseley, W. O. Moss, and J. S. Parker et al. describe in a series of papers (Organic Process Research & Development, 2003, 7, 53, 58, 67) how they used automation (Zymark robot) and factorial experimental design in combination with classical manual optimization to rapidly develop and scale-up the synthesis of four neurokinin antagonists at AstraZeneca in a parallel manner. By using the Zymark robot extensively for factorial experimental design studies, they were able to evaluate very quickly the robustness of the single reaction steps prior to performing them in larger scale. This approach enabled them to rapidly deliver the necessary amounts of bulk drug for further toxicological and clinical investigations. In one specific example it took them only six months from the start of lab work until delivery of 1 kg drug substance, which required a 22-step synthesis.

FLUORESCENCE SPECTROSCOPY AND MULTIVARIATE SPECTRAL DESCRIPTOR ANALYSIS FOR HIGH-THROUGHPUT MULTIPARAMETER OPTIMIZATION OF POLYMERIZATION CONDITIONS OF COMBINATORIAL 96-MICROREACTOR ARRAYS

Selection of optimum process conditions in combinatorial microreactors is essential if the combinatorial synthesis process is to be correlated with the synthesis process on a more conventional scale. A new methodology for the high-throughput multiparameter optimization of polymerization reaction conditions in arrays of microreactors was developed by R. A. Potyrailo et al. from General Electric Company (J. Comb. Chem., 2003, 5, 8). Their strategy is based on the application of nondestructive spectroscopic techniques to measure chemical properties of polymers directly in individual microreactors followed by the multivariate spectral descriptor analysis for the rapid determination of the optimal process conditions. During the screening of new polymerization catalysts the system was optimized for the best processing parameters using a set of input variables that included reactant parameters (relative amounts of starting components and catalyst loading) and processing variables (reaction time, reaction temperature, and inert gas flow). The measured output parameters were the chemical properties of materials and reproducibility of the material formation in replicate polymerizations in microreactors. An automated acquisition of fluorescence spectra setup with a fiber-optic probe was used to collect the data.

Although the methodology described in this paper was implemented for high-throughput optimization of polymerization conditions, it is more general and has further potential for a variety of applications in which optimization of process parameters can be studied in situ or off-line using spectroscopic and other tools.

INTEGRATED DRUG DISCOVERY TECHNOLOGY IN A TEST TUBE

Combinatorial chemistry and high-throughput screening have each proved to be useful techniques for discovering potential drug candidates. The Evolutionary Chemistry^™ technology integrates these two central drug discovery technologies into a process that improves the odds of identifying novel drug candidates and increases the productivity of the drug discovery process.

The evolutionary chemistry process is a new technology that allows the synthesis of libraries that can be chemically and structurally diverse. Libraries can contain 5 to 50 million compounds built around virtually any chemical variant on multiple chemical backbones. At the heart of this new technology lies a set of highly engineered and chemically modified RNA molecules that are capable of catalyzing a diverse array of carbon-carbon and carbon-nitrogen bond-forming reactions (e.g. cycloadditions, substitutions, condensations). Beside the synthesis of big libraries, the ability of selecting the best compounds from the millions of molecules that are assembled by the RNA catalysts is equally important. Different sets of compounds can be removed by using different methods such as incubating in a cytochrome P450 preparation or in the presence of human serum albumin or passing the mixture over an immobilized artificial membrane. The selective pressures that are applied during the evolutionary chemistry experiment are those that are most likely to contribute to identifying optimized lead compounds with increased chances of succeeding in preclinical and clinical studies. A first test of the evolutionary chemistry technology was done with the development of several monobactam inhibitors of penicillin-binding protein 2a (PBP2a). Within one year several inhibitors of PBP2a with IC50 values ranging from 0.3 to 0.7 ig/ml could be found. (Current Drug Discovery, 2002, 7, 21)

HIGH-THROUGHPUT PURIFICATION OF COMBINATORIAL ARRAYS

Over the last decade strategies in combinatorial chemistry for synthesis of new biologically active compounds have significantly changed. Most combinatorial chemistry departments in the pharmaceutical industry now focus most of their efforts in the parallel synthesis of large numbers of individual compounds rather than preparing mixtures of several compounds in one pot. With the massive increase of the number of compounds synthesized, establishing high-throughput purification systems becomes crucial for the overall drug discovery process. In a recent paper, C. Edwards (Biotage) and D. J. Hunter (GlaxoSmithKline) describe the performance and reproducibility of the Biotage Parallex, a high-throughput purification system (J. Comb. Chem., 2003, 5, 61). The results indicate that parallel purification is a robust technique for purifying large numbers of compounds. Results from one of the first libraries to be purified on the Biotage Parallex are presented and discussed. Since fractionation by UV can often result in a large number of fractions, threshold trigger versus yield and number of fractions was also investigated. This approach was used to purify an array of 4320 compounds, produced by an 11-step solid-phase synthesis in Irori MicroKans. Ninety-three percent of the compounds were successfully processed, with >90 % having purity >95 %. Using the methods described, at least 200 samples can be purified in 10 hours.

AUTOMATED MASS SPECTROMETRIC SEQUENCE DETERMINATION OF CYCLIC PEPTIDE LIBRARY MEMBERS

Cyclic peptides are an important class of potential antimicrobial therapeutic agents and are therefore the target of intensive combinatorial synthesis efforts. M. Reza Ghadiri et al. (Scripps Research Institute) recently reported a new sequencing protocol for rapid identification of the members of a cyclic peptide library from a split- and-pool synthesis (J. Comb. Chem., 2003, 5, 33). By using sonic spray ionization CID ion trap mass spectrometry and automated computer analysis of spectra, there is no requirement for chemical encoding, which simplifies library production and purification. Validation of the method was carried out using 11 individually synthesized cyclic D, L-hexapeptides and 30 cyclic D, L-octapeptides of known sequence. In these validation tests a high overall accuracy was achieved.

PROTEIN BIOCHIPS FOR AFFINITY-MALDI-TOF MASS SPECTROMETRY

A protein chip for affinity MALDI MS that might be revolutionizing proteomics research has recently been described by G. E. M. Tovar (Bioforum International, 2001, 5, 235). Particles with a high affinity surface for proteins are used for separation and enrichment process during the sample preparation for MALDI-TOF MS. Any non-specific binding of proteins to the chip surface can be prevented by a specifically designed SAM. A break-through of chip-based Affinity MALDI mass spectrometry can be awaited, when combining micro-structuring of protein-chips by spotting-techniques, e.g. by the inkjet principle or μ-contact printing with the MALDI-microprobe-technique. Parallel analysis of a large number of different binding processes can thus be analyzed within a short time. The screening of libraries and discovering of unknown protein-protein interactions will thus be enabled.

PHOTOBONDED BIOARRAYS

Bioarrays and biochips are emerging analytical tools for fast molecular screening and efficient prognostic and diagnostic medical testing. Thus molecular engineering and covalent immobilization of biomolecules on material surfaces retaining their biological activity is a major challenge in bioarray and biochip engineering. The development of photobonded bioarrays has been reported by H. Gao et. al (Bioforum International, 2001, 6, 288). A photolinker polymer enables the photoimmobilization of biomolecules or mixtures of biomolecules (proteins, nucleic acids, oligosaccharides). Therefore, specific pretreatment of the surface or the biomolecules is not required. The polymer is applicable with different sensor systems such as bioarrays and microfluidic systems, optical sensors (fluorescence, refractometry), SAW or electrochemical biotransducer systems. Immunocompetent and catalytically active biosensor surfaces can be generated with retained biological functions.

A NOVEL BIOCHIP WITH ELECTRONIC EVALUATION

Biochips are presently generally evaluated by optical procedures. This classical method could in the future have to compete against a novel new method. R. Thewes recently reported about a biochip that is equipped with integrated evaluation electronics (New Drugs, 2002, 6, 50).

The medium-density DNA sensor array has 16×8 positions and is based on CMOS (complementary metal oxide semiconductor). Redox-cycling is used as sensor principle. Each of the sensors is composed of interdigitated gold electrodes. Single strand “catcher” molecules are immobilized on the gold surface. Any complementary sequences that are present in an analytical sample with denatured DNA molecules hybridize at the individual sensor position. On applying an oxidation and a reduction potential a substrate such as para-amino phenylphosphate is converted to an electrochemically redox-active component. Electric currents of different signs are generated at both electrodes.

For the use of the chip within a wide area of biological applications, the circuits of each single sensor cover a range of sensor currents from 1 pA to 100 nA. This chip will offer the benefit of significantly quicker, simpler, and more cost-favorable diagnostics.

OVERCOMING THE BOTTLENECKS: LARGE SCALE SNP DETECTION

The identification and analysis of Single Nucleotide Polymorphisms (SNPs) have become important tools for unraveling the genetic basis for a variety of human diseases. It is estimated that each person's DNA contains three million SNPs of which 200,000–300,000 may be relevant. In order to position new compounds into the market, it is assumed that future efforts will require the detection of 30 billion SNP genotypes. Detection of SNPs on a HT scale has thus become a priority for pharmaceutical companies. Proteodyne recently introduced a suitable instrument matching these requirements. The Molecular BioCube is a multi-functional machine equipped with an ultra fast servo-driven four-axis Cartesian robot. The system is completed with multiple tools e.g. plate gripper tool or a 96-channel pipetting tool for liquid handling. A fully automated plate sealer and a temperature controlled centrifuge with four plate positions or a temperature controlled carousel for continuous random access storage of up to 256 microtiter plates have also been integrated. A complex system of two BioCube modules is able to handle up to 90,000 SNP reactions per day.

The BioCube systems are fully automated robotic platforms designed to overcome the greatest bottleneck in biotechnological processes. Proteodyne offers highly versatile system solutions, which can compress the time to discovery, reduce the overall costs, and improve the data integrity. (G. Weidner; New Drugs, 2002, 10, 24)