Ioana Popa-Burke

A Chemogenomic Analysis of the Human Proteome: Application to Enzyme Families

Sequence-based phylogenies (SBP) are well-established tools for describing relationships between proteins. They have been used extensively to predict the behavior and sensitivity toward inhibitors of enzymes within a family. The utility of this approach diminishes when comparing proteins with little sequence homology. Even within an enzyme family, SBPs must be complemented by an orthogonal method that is independent of sequence to better predict enzymatic behavior. A chemogenomic approach is demonstrated here that uses the inhibition profile of a 130,000 diverse molecule library to uncover relationships within a set of enzymes. The profile is used to construct a semimetric additive distance matrix. This matrix, in turn, defines a sequence-independent phylogeny (SIP). The method was applied to 97 enzymes (kinases, proteases, and phosphatases). SIP does not use structural information from the molecules used for establishing the profile, thus providing a more heuristic method than the current approaches, which require knowledge of the specific inhibitors structure. Within enzyme families, SIP shows a good overall correlation with SBP. More interestingly, SIP uncovers distances within families that are not recognizable by sequence-based methods. In addition, SIP allows the determination of distance between enzymes with no sequence homology, thus uncovering novel relationships not predicted by SBP. This chemogenomic approach, used in conjunction with SBP, should prove to be a powerful tool for choosing target combinations for drug discovery programs as well as for guiding the selection of profiling and liability targets. (Journal of Biomolecular Screening 2007:972-982)

Establishing Quality Assurance Criteria for Serial Dilution Operations on Liquid-Handling Equipment

Since the advent of high-throughput screening (HTS) in the early 1990s, parallel multichannel liquid handlers have become a mainstay in every drug discovery setting. Although several peer-reviewed publications have discussed methods and criteria for stamping multiwell copies, there is very little information about establishing a standard operating procedure (SOP) for standard (microliter-level) serial dilutions of compounds used in dose-response experiments. The authors discuss the 4 main criteria any serial dilution process must pass (accuracy, precision, fold dilution, and outliers) and the process for establishing thresholds for all of these values in a compound management or biological screening laboratory. The thresholds need to be both low enough to be acceptable from a biological potency variability perspective and high enough to allow the instruments to pass the quality assurance (QA) analysis on a regular basis. In this article, the authors suggest suitable thresholds arrived at by a variety of methods, including trend analysis of QA data, survey questionnaire from the main stakeholders (screening scientists, chemists), and published criteria for single-shot stamping. A mathematical analysis of the effect of threshold values on estimated XC50s was performed to ensure that the variability introduced by the serial dilution step is within acceptable overall variability limits.

Discovery of Highly Selective Inhibitors of p38α

The p38 MAP kinases are a family of serine/threonine protein kinases that play a key role in cellular pathways leading to pro-inflammatory responses. We have developed and implemented a method for rapidly identifying and optimizing potent and selective p38alpha inhibitors, which is amenable to other targets and target classes. A diverse library of druggable, purified and quantitated molecules was assembled and standardized enzymatic assays were performed in a microfluidic format that provided very accurate and precise inhibition data allowing for development of SAR directly from the primary HTS. All compounds were screened against a collection of more than 60 enzymes (kinases, proteases and phosphatases), allowing for removal of promiscuous and non-selective inhibitors very early in the discovery process. Follow-up enzymological studies included measurement of concentration of compound in buffer, yielding accurate determination of K(i) and IC50 values, as well as mechanism of action. In addition, active compounds were screened against less desirable properties such as inhibition of the enzyme activity by aggregation, irreversible binding, and time-dependence. Screening of an 88,634-compound library through the above-described process led to the rapid identification of multiple scaffolds (>5 active compounds per scaffold) of potential drug leads for p38alpha that are highly selective against all other enzymes tested, including the three other p38 isoforms. Potency and selectivity data allowed prioritization of the identified scaffolds for optimization. Herein we present results around our 3-thio-1,2,4-triazole lead series of p38- selective inhibitors, including identification, SAR, synthesis, selectivity profile, enzymatic and cellular data in their progression towards drug candidates.

The Effect of Initial Purity on the Stability of Solutions in Storage

Many modern compound-screening technologies are highly miniaturized, resulting in longer-lasting solution stocks in compound management laboratories. As the ages of some stocks stretch into years, it becomes increasingly important to ensure that the DMSO solutions remain of high quality. It can be a burden to check the quality of a large library of compound solutions continuously, and so a study was devised to link the effects of initial compound purity and physicochemical properties of the compounds with the current purity of DMSO solutions. Approximately 5000 compounds with initial purity of at least 80% were examined. Storage conditions were held or observed to be relatively constant and so were eliminated as potential predictors. This allowed the evaluation of the effects of other factors on the stability of solutions, such as initial purity, number of freeze-thaw cycles, age of the solution, and multiple calculated physicochemical parameters. Of all the factors investigated, initial purity was the only one that had a clear effect on stability. None of the other parameters investigated (physicochemical properties, number of freeze-thaw cycles, age of solutions) had a statistically significant effect on stability.

Compound Precipitation in High-Concentration DMSO Solutions

Screening compounds precipitate out of high-concentration DMSO solutions and cause issues in the liquid-handling equipment used to process the samples as well as false-positives and false-negatives in the biological data. This report examines the extent of the precipitation issue, correlation with physico-chemical properties of compounds, and measured concentrations. Possible ways to allow recovery of the precipitated solids are also described. This study also investigated whether particulates were present from the time of initial solubilization in DMSO or whether precipitation occurs during freeze-thaw cycling.

Analysis of Compound Weighing Precision in Drug Discovery

Early drug discovery laboratories often call for the precise weighing of 1- to 5-mg solids into 4- to 5-g glass vials. For the balance used in this study (Mettler Toledo XP205), the manufacturer rates its accuracy at ±0.01 mg over the working range of 1 mg to 220 g and its precision or repeatability at 0.015 mg for 10-g weights. The manufacturer ratings were confirmed using standard steel weights, but these calibrators do not well represent the weighing precision of drug compound. For example, when pre-taring a 4- to 5-g vial on the balance and then weighing 1- to 5-mg calibration weights, although no bias was observed, precision dropped appreciably. When measuring solid sample in the range of 1 to 5 mg, deviation of the measured weight from the actual (true) weight was even worse, in the range of ±20% to 50%. Balance settings and environmental factors exert a strong influence on weighing precision. Although most environmental factors, such as air draughts, temperature, vibrations, and levelness, can be optimized to the extent practical in laboratory settings, problems due to static electricity are often overlooked. By controlling static electricity, we demonstrate how we optimized the process to where measurements were within ±10% of actual weight when weighing solid sample in the range of 2 to 5 mg and ±20% when weighing 1 mg into a 4- to 5-g vial. Our weighing process and method to calculate actual weight are given in detail.

Nuisance Compounds, PAINS Filters, and Dark Chemical Matter in the GSK HTS Collection:

High-throughput screening (HTS) hits include compounds with undesirable properties. Many filters have been described to identify such hits. Notably, pan-assay interference compounds (PAINS) has been adopted by the community as the standard term to refer to such filters, and very useful guidelines have been adopted by the American Chemical Society (ACS) and subsequently triggered a healthy scientific debate about the pitfalls of draconian use of filters. Using an inhibitory frequency index, we have analyzed in detail the promiscuity profile of the whole GlaxoSmithKline (GSK) HTS collection comprising more than 2 million unique compounds that have been tested in hundreds of screening assays. We provide a comprehensive analysis of many previously published filters and newly described classes of nuisance structures that may serve as a useful source of empirical information to guide the design or growth of HTS collections and hit triaging strategies.