Combinatorial Chemistry & High Throughput Screening - Volume 8, Issue 6, 2005

While innovation, scientific expertise, business flair and leadership vision clearly are important factors in the success of a pharmaceutical company as it strive to create new medicines, a pharmaceutical company's most valuable asset may prove to be its compound screening libraries. Maintaining and growing these libraries in order to facilitate timely and effective high throughput screening campaigns is a challenging responsibility, This challenge generally falls upon "Material Management" or HTS groups and their work is crucial to screening success. Challenges that "Material Management" or HTS groups must address include determining criteria for deciding which compound should be included in the library. Selection can be based on combinatorial chemistry libraries, medicinal chemistry intermediates, and even natural products extracts, purified natural products, etc... The goal is to provide high quality, chemically diverse, reproducible chemical matter for biological screening and ultimately lead matter for medicinal chemistry. Structural integrity has to be assessed, including solution state, hydration level, and concentration of screening banks. Decisions have to be made in order to determine when to make such assessments. Structural evaluation can be performed at submission, at regular intervals throughout the compound library's live cycle, or exclusively after activity was detected by HTS. They must also address at which stage of HTS a compound should be considered for such assessment. At the primary level? After activity confirmation? Or after dose response determination (i.e. IC50)? In this Issue of Combinatorial Chemistry & High Throughput Screening, we have an exciting collection of contributions illustrating ways to generate, assess, QC and interpret data used to address the issues listed above. Edward Kerns (Wyeth) and colleagues extensively reviewed the recent literature describing the use of the most up to date analytical techniques (such as LC-MS, ELSD, CLND, etc...) for assessing compound integrity in an HTS environment. Bernard Choi (Merck) and coworkers focused their paper on an approach to rapidly process and interpret LC-MS data, based on a computer application developed in-house to process LC-MS report files containing both spectral and chromatographic data. Mark Strege (Lilly) and colleagues discuss the added value of the use of MS-MS, ELSD and CLND coupled with the bioassay-guided fractionation, or "biofractionation" technique where an activity profile correlates with a chromatographic profile. Richard Ellson (Labcyte) and coworkers discuss a new acoustic method of assessing water content in DMSO solutions (a major contributor to sample instability in solution) in storage plates. Michelle Kelly (Pfizer) and colleagues discuss the adverse effects of water uptake and freeze-thaw cycles on compound solubility in DMSO and demonstrate how one can re-solubilize a precipitated DMSO screening stock solution by low energy sonication. Ulrich Schopfer (Novartis) and coworkers describe a unique strategy used for addressing compound stability and solubility issues for HTS compound storage. Unsuitable compounds being excluded upfront, storage occurs at 4°C/20% relative humidity in a DMSO/water mixture, thus avoiding freeze-thaw cycles, and compounds are resolubilized at regular intervals. Ramesh Padmanabha (Bristol-Myers Squibb) and colleagues concentrated their effort on the process of HTS with an emphasis on quality control, reducing the variability of all the processes that have an impact on the final result. George Harrigan and Gilles Goetz (Pfizer) reviewed the current trends in Natural Products screening, including chemical screening, virtual screening , NP based library design. It has been a privilege reading those excellent pieces of work dedicated to help us find out "what is in our wells" in a preview atmosphere, but it is now time for prime-time release, therefore, enjoy.

Integrity profiling of HTS hits is valuable for verification of the hit identity and purity. This provides early discovery researchers with more confident SAR theories. Methodology for integrity profiling of HTS hits must be high throughput, consume little material, and selectively provide structure-based data. Analytical techniques that can be utilized for integrity profiling methods are reviewed for their appropriateness in sample preparation, component separation, detection, purity quantitation, identity confirmation, and follow-up.

An approach to rapidly process and interpret high-throughput liquid chromatography mass spectrometry data is presented. This approach applies an in-house developed computer application to process LC-MS report files containing spectral and chromatographic data from four different detectors (i.e. electrospray positive ionization, electrospray negative ionization mass spectrometry, UV absorption, and evaporative light scattering detection). Properties characteristic of detection and chromatographic retention are extracted and populated into a database. Approaches to applying this analytical information database for quality control analysis of ca. 400,000 samples are presented. Compound quality assessment methods employing average purity and detection data fields are compared to methods employing multiple quality control criteria (e.g. detection, purity, retention, and signal to noise). Structural similarity searches were applied with the analytical information database to identify compounds that may be undetectable by electrospray mass spectrometry. In addition, an approach to applying the database to aid in the selection of analytical detection and chromatography conditions for rapid analytical method development is also discussed.

In this investigation the utility of evaporative light scattering detection (ELSD) combined with HPLC-MS was demonstrated as a key component of a bioassay-guided fractionation, or "biofractionation" technique, for the evaluation of high throughput screen actives. ELSD provided on-line analyte mass information that was critical for the classification of the samples. Chemiluminescent nitrogen detection (CLND) was also evaluated for sample concentration estimation for nitrogen-containing compounds, and accurate mass LC-MS-MS analysis was employed for rapid structural confirmation and elucidation of components previously identified as active via biofractionation.

Compounds used in high throughput screening (HTS) are typically dissolved in DMSO. These solutions are stored automation-friendly racks of wells or tubes. DMSO is hygroscopic and quickly absorbs water from the atmosphere. When present in DMSO compound solutions, water can accelerate degradation and precipitation. Understanding DMSO hydration in an HTS compound library can improve storage and screening methods by managing the impact of water on compound stability. A non-destructive, acoustic method compatible with HTS has been developed to measure water content in DMSO solutions. Performance of this acoustic method was compared with an optical technique and found to be in good agreement. The accuracy and precision of acoustic measurements was shown to be under 3% over the tested range of DMSO solutions (0% to 35% water by volume) and insensitive to the presence of HTS compounds at typical storage concentrations. Time course studies of hydration for wells in 384-well and 1536-well microplates were performed. Well geometry, fluid volume, well position and atmospheric conditions were all factors in hydration rate. High rates of hydration were seen in lower-volume fills, higher-density multi-well plates and when there was a large differential between the humidity of the lab and the water content of the DMSO. For example, a 1536-well microplate filled with 2μL of 100% DMSO exposed for one hour to a laboratory environment with ∼40% relative humidity will absorb over 6% water by volume. Understanding DMSO hydration rates as well as the ability to reverse library hydration are important steps towards managing stability and availability of compound libraries.

Dissolution of organic compounds in DMSO in HTS plate or tube format is a difficult problem as users move to higher compression plate formats. Precipitation of compounds from DMSO screening stocks is a recognized problem in the HTS materials management process. The adverse effect of freeze thaw cycles on DMSO stock solutions stored in plate format as a result of cherry picking operations has led to the gradual replacement of plate-based storage with tube-based storage so as to minimize the number of freeze thaw cycles. Compound solubility in DMSO is markedly decreased by uptake of small quantities of water. We attribute this effect to the non ideal properties of DMSO water mixtures such that cavity formation in solvent, a necessary step in dissolution, is more difficult in wet DMSO than in dry DMSO or in pure water. We report here that efficient compound dissolution is possible even in 384 well format by the use of in-well plate-based sonication. Surprisingly, compounds precipitated from DMSO stocks either by water uptake or repeated freeze thaw cycles can be re-dissolved by low energy sonication. Finally, we demonstrate that precipitation of compound from DMSO stock solutions is synergistically enhanced by water uptake into DMSO compound stock solutions as well as by increasing the number of freeze thaw cycles.

As HTS technologies come of age, pharmaceutical companies are focusing increasingly on the quality of their screening collections. Storage conditions and their influence on compound stability and solubility are debated intensely. At Novartis, a strategy was developed that is different to most other companies: (1) compounds unsuitable for storage in solution are excluded by computational methods; (2) compounds are stored at 4°C/20% relative humidity in a DMSO/water mixture to avoid freeze-thaw cycles and water uptake and to allow rapid plate replication; (3) resolubilisation of compounds at regular intervals.

Changes in all aspects of HTS from compound management through to evaluation of hits and leads, strengthened by infrastructure improvements, in both automation and informatics, have made possible increased analysis and implementation of process and quality control throughout HTS. This paper focuses on the process of HTS with an emphasis on quality control, reducing the variability of all the processes that have an impact on the final result, and argue that by increasing the quality of the entire process that data mining of primary screening data is in fact possible and will reduce cycle times to medicinal chemistry.

Due to pressure from combinatorial chemistry and the streamlining of the drug discovery process through automated high-throughput screening technologies, pharmaceutically based natural products programs are under increasing scrutiny. However by taking advantages of technologies originally developed for high-throughput screening and combinatorial chemistry and applying them to processes considered as bottlenecks in classical natural products chemistry (purification, structure elucidation, sample availability) it is our opinion that natural products can still contribute to the effective discovery of novel bioactive and pharmaceutically relevant metabolites. We describe here several such strategies that if universally implemented, will demonstrate i) whether chemical diversity is truly being accessed, ii) that novel metabolites can be formatted in a manner appropriate for modern screening paradigms, and iii) that natural products can be rapidly identified not only for novelty and pharmaceutical relevance but to assess their true biological origin.

Gilles Goetz obtained his Masters (Biochemistry 1991) and his PhD degrees (Organic Chemistry 1995) from the University Louis Pasteur in Strasbourg (Fr) working under the supervision of Prof. J-P Lepoittevin studying the molecular processes of Poison Ivy contact dermatitis. He then studied Marine Natural Products as a post-doctoral fellow from 1996 to 1998 at the University of Hawaii (US) working with late Prof. Paul J. Scheuer. From 1998 to 2000 he was research assistant (post-doc) for Prof. Raphaelle Tabacchi at the University of Neuchatel (CH) studying Fungal and Terrestrial Natural Products. From 1996 to 2002 he isolated, purified and determined the structures of numerous biologically active natural products, from marine, terrestrial and fungal sources utilizing MS and advanced NMR techniques. Since 2000 he has been a principal research scientist at Pfizer (legacy Searle-Monsanto, Pharmacia) in St Louis, Missouri. As a member of the analytical chemistry team supporting High Throughput Screening, he contributes to provide bona fide leads to numerous research projects in several therapeutic areas. His research interests include: natural products chemistry; chemical ecology; combinatorial chemistry, computational chemistry; chemo-informatics; advanced mass spectrometric methodology for the analysis of complex organic samples. SELECTED PUBLICATIONS 1. Luesch, H.; Horgen, F.D.; Goetz, G.; Harrigan, G.G. The cyanobacterial origin of potent anticancer agents originally isolated from sea hares and other marine macro organisms. Curr. Med. Chem. 2002, 9, 1241-1253. 2. Kimura, J.; Takada, Y.; Inayoshi, T.; Nakao, Y.; Goetz, G.; Yoshida, W.Y.; Scheuer, P.J. Kulokekahilide-1, a cytotoxic depsipeptide from the cephalaspidean mollusk Philinopsis speciosa. J. Org. Chem. 2002, 67, 1760-1767 3. Goetz, G.H.; Harrigan, G.G.; Likos, J. Ugibohlin: A new dibromo-seco-isophakellin from Axinella carteri. J Nat. Prod. 2001, 64, 1581 -1582 4. Becerro, M.A.; Goetz, G. Paul, V.J.; Scheuer, P.J. Chemical defenses of the mollusk Elysia rufescens and its host alga Bryopsis sp. J. Chem. Ecol. 2001, 27, 2287-2299 5. Goetz, G.; Meschkat, E.; Lepoittevin, J.P. Synthesis of a water soluble analog of 6-methyl-3-n-alkyl catechol labeled with carbon 13: NMR approach to the reactivity of poison ivy/oak sensitizers toward proteins. Bioorg. Chem. Letters, 1999, 9, 1141-1146.

An oligonucleotide-based mutagenesis method is presented where, contrary to most classical mutagenic approaches, preselection of the variants is performed at the oligonucleotide level to avoid cloning of non-desired sequences. The method relies on the generation of differentially phosphate-protected oligonucleotides. Protection of the phosphates is accomplished by substoichiometric incorporation of an Fmoc-protected and n-propyl-protected trinucleotide phosphoramidite during ordinary oligonucleotide assembly. Instead of the alkali-labile ß-cyanoethyl group introduced in ordinary assembly, the trinucleotide introduces the alkali-stable n-propyl group. As a result, single mutants carry three ionic phosphates less than the wild-type sequence, double mutants carry six ionic phosphates less and so on. This difference in ionic ratio enables separation of the variants by conventional polyacrilamide gel electrophoresis. In the exemplified library described herein, two sub-populations containing mainly triple and quadruple mutants were selected out of five possible sub-populations.

Selections from phage-displayed combinatorial peptide libraries are an effective strategy for identifying peptide ligands to target proteins. Existing protocols for constructing phage-displayed libraries utilize either ligation into double-stranded phage DNA or Kunkel mutagenesis with single-stranded phagemid DNA. Although the Kunkel approach rapidly provides library sizes of up to 1011, as many as 20% of the phagemids may be non-recombinant. With several modifications to current Kunkel protocols, we have generated peptide libraries with sizes of up to 1011 clones and recombination frequencies approaching 100%. The production of phage libraries, as opposed to phagemid libraries, simplifies selection experiments by eliminating the need for helper phage. Our approach relies upon the presence of an amber stop codon in the coding region of gene III of bacteriophage M13. Oligonucleotides containing randomized stretches of DNA are annealed to the phage genome such that the randomized region forms a heteroduplex with the stop codon. The oligonucleotide is then enzymatically extended to generate covalently-closed, circular DNA, which is electroporated into a non-suppressor strain of Escherichia coli. If the amber stop codon is present in the DNA molecule, protein III is not synthesized and the phage cannot propagate itself. This method is customizable for the display of either random or focused peptide libraries. To date, we have constructed 22 different libraries ranging from 8-20 amino acids in length, utilizing complete or reduced codon sets.