High-throughput screening (HTS) is the dominant technique in early drug discovery and has had an important impact on human health. A key challenge in HTS is the occurrence of false-positive inhibitors, which can dominate hit-lists. Multiple mechanisms have been proposed to explain these artifacts, including chemical reactivity, oxidation, assay interference, and colloidal aggregation. In an earlier screen of the MLSMR, we developed a direct screen for aggregation-based inhibition, using ?-lactamase as a reporter enzyme. Two startling results emerged from this study. First, of the 1274 inhibitors found, 96% were aggregators and only 1% acted covalently. No true, competitive inhibitors were found by HTS. Second, a docking screen of the same library did find two novel inhibitors. These had Ki values of about 40 ?M and bound specifically to ?- lactamase by x-ray crystallography. Here we ask how representative these results might be. Is covalent inhibition really only a minor issue in enzyme HTS, far overwhelmed by aggregation? Can docking reliably prioritize compounds for experimental testing? We investigate a second enzyme for promiscuous and covalent inhibition, and as a target for docking.
The specific aims are: 1. To determine the ratio of promiscuous and covalent inhibitors in an HTS screen of the cysteine protease cruzain. As a cysteine protease, cruzain will be much more sensitive to reactive and compounds than ?-lactamase. It thus provides a robust test of the provocative hypothesis that covalent artifacts are rare in screening. We will use the same protocol we developed for ?-lactamase: compounds that inhibit cruzain in the absence but not the presence of detergent are aggregators. The aggregators that are found will be compared to those found in the ?-lactamase screen, which will itself be interesting for this widely used collection. Even more interesting will be the ratio of aggregators to all other mechanisms of inhibition. We will re-test characteristic inhibitors in low throughput secondary assays here at UCSF, determining their mechanisms of action using DLS, enzyme counter-screens, mass spectroscopy, and x-ray crystallography. 2. To compare docking hit rates with HTS hit rates. The HTS assay affords us an opportunity to, at no extra cost, compare high-throughput with virtual screening. We will prospectively dock the same MLSMR library (120,000 compounds) against the x-ray structure of cruzain, and compare the docking hits to those subsequently found by HTS. As with the screen against ?-lactamase, we will be at liberty to test high-scoring docking hits at higher concentrations of inhibitor and lower concentrations of substrates, allowing for mid- micromolar inhibitors that HTS might miss. We will ask: which hits does docking capture, what are the false positives and the false negatives for both techniques? Here too, we will determine x-ray structures of enzyme complex structures for specific inhibitors. We will pursue the SAR of interesting hits with the synthetic chemistry group at the NCGC. The most widely used technique to find new candidate drugs is high-throughput screening (HTS). A major challenge for HTS is the occurrence of false-positive inhibitors, which can dominate hit-lists. We test a rapid method to detect these artifacts, which should significantly advance this critical technique. ? ? ? ? ?
