The Detection, Prevalence and Properties of Aggregate-Based Small Molecule Inhibition
In the early phases of drug discovery, high-throughput screening (HTS) has emerged as the dominant technique used to discover potential therapeutics. A critical weakness of this technique is sensitivity to artifactual, nonspecific inhibition. One source of nonspecific inhibition is the aggregation of organic molecules into colloidal particles. These aggregates have a unique property--they nonspecifically associate with proteins and inhibit enzyme function. If uncontrolled for, this inhibition is difficult to distinguish from classical mechanisms of small-molecule inhibition. In the context of high-throughput screening, this poses a large problem, as aggregate-forming molecules inherently lack the specificity desired in therapeutics or chemical tools. Thus, they cloud screening results by obscuring potentially desirable specific inhibitors.
Much of the work contained in this dissertation concerns the development of a high-throughput methodology to detect small-molecule aggregates. This was undertaken to provide a robust method for identification of aggregates and to probe the prevalence of this phenomenon. Two technologies were evaluated. First, an enzymatic assay was developed that detects detergent-sensitive inhibition, a conspicuous characteristic of aggregate-based inhibition. Second, a light scattering approach was employed to detect aggregate particles in solution. Subsequently, the more robust detergent-dependant inhibition assay was applied to a large library of small molecules in order to measure the prevalence of aggregate-formation. Finally, to complement the methods developed to detect aggregates, the ability of aggregates to inhibit a non-enzymatic reaction, amyloid fibrillization, was studied using an in vitro fluorescence assay and electron microscopy.
Several conclusions emerge from this dissertation. First, screening for detergent-sensitive inhibition is a robust and efficient method for identifying aggregate-based enzyme inhibition. Second, aggregate formation is common; nearly 2% of a 70,563 molecule library formed aggregates below 30 μM. Third, whereas there are many mechanisms of nonspecific inhibition, aggregate formation may be the most common. Fourth, aggregate formation can be accentuated or reduced in mixtures of molecules, suggesting caution in mixture-based HTS. Finally, beyond enzymes, aggregates also inhibit the formation of amyloid fibrils, and should be considered as a source of artifact in screens for small molecule inhibitors amyloid formation.