pounds to the clinic faster? In order to facilitate the development of new screening methodologies many companies utilize existing technologies as platforms for developing new screening campaigns.
For example, Graffinity Pharmaceutical Design GmbH (Heidelberg)  uses Rapid Array Informed Structure Evolution (RAISE™), a surface plasmon resonance detection methodology, to identify novel target-specific compounds by flowing soluble proteins over gold surface immobilized fragments isolated from a combinatorial chemistry-derived library. 3-Dimensional Pharmaceuticals (now part of Johnson & Johnson) uses fluorescence-based thermal shift assays in a microplate, high throughput format to monitor ligand-induced stabilization of proteins. The technique has several advantages, namely that the general applicability of the thermal shift assay circumvents timely and costly development steps, and the assay is indiscriminant to any prior knowledge of protein function .
Finally, measuring amide hydrogen/deuterium (H/D) exchange in proteins, monitored by protein mass spectrometry, has been used to monitor ligand binding-induced shifts in protein stability [7-9]. The first technology SUPREX (stability of unpurified proteins from rates of H/D exchange) uses a fluorescence-based thermal shift assay, developed in a microplate, high throughput format, to monitor ligand-induced stabilization of proteins . Protein stability is assessed by following the extent of H/D exchange during a multi-point urea titration and establishing the midpoint for protein unfolding. When ligand binders are added protein stability is enhanced and a higher urea concentration is required to reach this midpoint. By choosing an appropriate single urea concentration (@3 M) the ability of individual ligands to influence protein stability can be measured, and this has been exploited as a high throughput screening technology. Briefly, test compounds (at 6 mM) are placed in microtiter plate wells, followed by deuterated exchange buffer that contains a constant urea concentration, and this mixture is allowed to equilibrate . Target protein is then added in small volumes (10 mL) to a final concentration of 1 mM and equilibrated for 30 min. Next, H/D exchange is quenched with trifluoracetic acid, the sample is concentrated and desalted using chromatography columns, and placed at —20 °C to prevent H/D backexchange. Finally, the samples are analyzed using MALDI-MS. A caveat is that the ligands must be in significant excess of both the protein concentration and the Kd of protein-ligand complex, which offers the possibility of compound solubility issues. For example, Powell and Fitzgerald alluded that ligand concentrations in excess of 100 mM may be required to measure 10 mM KD binding if a modest shift in stability toward unfolding is observed. Such high compound concentrations suggest solubility may be a limiting issue. The second technology PLIMSTEX (quantification of protein ligand interactions by mass spectrometry, titration and H/D exchange) monitors differences in H/D exchange of amide hydrogens of a target protein resulting from the interaction with a ligand by ESI-MS (see Chapter 11).
Notably, all of the above technologies function by observing quantitative functional changes or chemical modifications in the target protein, rather than the ligand. A disadvantage to this paradigm is that the ligands must be present in significant excess of both the protein concentration and the KD of the biological reaction. Considering the KD range of 500 nM to 5 mM as typical in early stage drug discovery, there is significant concern about compound solubility. Conversely, techniques that monitor ligands directly, rather than protein behavior, have the advantage of being performed at protein excess. Under these conditions, compound solubility typically is less of an issue because their concentrations can be held much lower, at least several-fold less than the KD and near the limits of MS detection. However, a major caveat to protein-excess screening paradigms is that protein consumption becomes a limiting factor. Hence, for these campaigns to be successful in early-stage drug discovery, constantly evolving strategies for re ducing target protein consumption must be implemented. Many companies have circumvented this obstacle by moving to much larger compound screening mixtures. With the advent of MS-based readout in affinity screening methodologies, the monoisotopic masses unique to each individual compound can be directly measured with MS, even in large mixes containing closely related monoisotopic redundant neighbors, allowing for target-specific ligands to be readily identified.
Several affinity screening methodologies that include MS-based readout and work under protein-excess conditions have been developed in the past decade . Some examples include affinity selection/mass spectrometry (ASMS; Abbott Labs ), size exclusion chromatography with LC-ESI-MS (see Chapter 2 and 3 [11-19]), the use of coupled or non-coupled pulsed ultra-filtration/mass spectrometry (summarized in this chapter [11, 20-23]), restricted access phase chromatography (see Chapter 5 [24, 25]), capillary electrophoresis [26, 27], target shift mass spectrometry , and multitarget affinity/specificity screening (MASS, see Chapter 10 [29, 30]).
Importantly, the central difficulty for high throughput affinity-based screening techniques is how to screen large compound collections in a realistic timeframe. Each of the above techniques has strengths and limitations with respect to assay development time, screening throughput, specialized protein requirements, and specialized library design requirements [11, 22, 31]. For example, for those techniques requiring the immobilization of reaction components (such as protein or compound tagging), there is the possibility for artifacts in protein character (alteration in conformation, inactivation of key residues) or limitations in library chemistry. Additionally, most affinity screening techniques coupled with MS become overwhelmed when hundreds of thousands of library compounds are screened per target, yet a consensus of operational and theoretical studies from HTS over the past ten years has indicated that screening is most effective by maximizing library size [32-34]. Hence, until we develop a more concrete understanding of small molecule structural diversity, and subsequently apply that knowledge to synthesizing small libraries that encompass the entire chemical effector space, our best chance of identifying a good starting point for medicinal chemistry optimization will increase only as the total number of compounds screened increases. Furthermore, as the library size and number of targets increases, a general concern about affinity-based screening is that the identification of a large number of non-selective, promiscuous, compounds can be overwhelming so that the best, selective compounds may be overlooked. Evidence for the above concerns is that most of these referenced techniques have been successful in screening only relatively small libraries, relatively small mixtures of compounds, and even fewer have reported the discovery of bona fide new lead(s).
To address these concerns, we at Abbott Laboratories developed a high throughput screening method that is efficient and robust enough to allow study of many targets against very large libraries on the basis of affinity. The method contains an adjustable selection stringency and a computational filter for removing promiscuous compounds that bind non-selectively to proteins in general. As discussed below, the method enabled the discovery of a novel compound series that binds specifically and inhibits the UDP-MurNAc-pentapeptide synthetase enzyme MurF, which catalyzes the final step in synthesis of the bacterial peptidoglycan cell wall precursor, addition of D-Alanine-D-Alanine to UDP-MurNac-tripeptide. Targeting the UDP-MurNAc-pentapeptide synthetic pathway has been a goal of antibacterial research for years . Two chemically related compounds were rapidly determined to be the most potent and selective ligands in a library of 123 405 compounds, screened in large pools of @2700 compounds per mixture with a stringency set by the protein concentration of 10 mM. The identification of this novel MurF inhibitor series led to a medicinal chemistry optimization effort described in detail elsewhere .
Was this article helpful?