UCSF

Early drug discovery is plagued by nonspecific molecules, which cannot be developed into drugs. These molecules appear active in initial studies, but upon further investigation are not interesting leads. One of the leading sources of these "false positives" is the formation of large aggregates of organic small molecules, which have the unusual property of promiscuously binding and inhibiting enzymes. Such molecules are common among screening libraries, biological reagents, and even known drugs. Despite their prevalence, there are many elementary properties of these particles that remain unknown and the extent of their repercussions beyond drug screening is not yet clear.

This work focuses on expanding our understanding of the fundamental characteristics and behavior of nonspecific small molecule aggregates. First, to determine whether aggregates may persist in biological contexts we investigated the behavior of these particles in high protein milieus. Second, we explored a number of elementary questions such as the concentration of particles in solution, whether aggregates were in equilibrium with free small molecule monomers, and whether aggregates exhibited micelle-like characteristics such as a "critical aggregation concentration." Finally, we used hydrogen deuterium exchange mass spectrometry and proteolysis to dissect the molecular principles underlying aggregate-based inhibition.

The key results of this work are as follows. Aggregates are not disrupted by milligram per milliliter concentrations of protein; however, they do exhibit a critical concentration and will rapidly dissociate if diluted below this threshold. Above this threshold, aggregates were detected at femtomolar concentrations. The stoichiometric inhibition of nanomolar enzyme concentrations by femtomolar aggregate concentrations suggests a tight binding interaction between enzyme and aggregate, which is consistent with observations that the off-rate from the aggregate was imperceptible within the timescale of several experiments. Lastly, it appears that aggregates inhibit enzyme by partial unfolding.