Compound aggregation was recently discovered to be one of the main causes for promiscuous enzyme inhibition [12
]. Under certain conditions, above certain concentrations, some compounds self-associate to form an aggregate structure, which, at 50–400nm in size, can be visualized by transmission electron microscopy (TEM) [31
]. Evidence suggests that enzymes are sequestered on the surface of the aggregate particles, where their function is non-specifically inhibited [30
]. Prevention of compound aggregation through addition of nonionic detergents, such as Triton X-100, effectively relieves enzyme inhibition by this mechanism [30
Whether or not a compound is prone to aggregation is dependent upon properties of the compound itself, the assay conditions [11
] and the protein target [34
]. For this reason, and because compounds that aggregate are structurally diverse [31
], interference due to aggregate inhibition must be empirically determined for a given assay [5
]. Generally, though, compounds tend to aggregate at micromolar concentrations, and a compound that aggregates at a higher concentration may have legitimate biological activity at lower concentrations [12
]. Compound interference by aggregation is relatively easy to identify with a little work, largely because aggregation-based inhibition has hallmark characteristics, as shown in . While the exact mechanism of how compound aggregates inhibit enzymes is unclear [35
], it has been found that addition of 0.01–0.1% Triton X-100 to assay reagents generally allows for significant relief of aggregation-based inhibition () [12
], thus making it possible to design a biochemical assay that is less sensitive to this form of inhibition. Significantly, Jadhav et al.
(2010) found that a screen run without detergent resulted in 15-times more actives than a screen run in the presence of detergent [11
]. A protocol to detect aggregation-based inhibition is published and available [36
In an effort to estimate the prevalence of aggregation-based inhibition for a typical HTS involving a biochemical assay, investigators have tested various small molecule libraries for enzyme inhibition sensitive to Triton X-100 [12
]. Interestingly, though the same enzyme (AmpC β-lactamase) was used in two of the assays screened, the percentage of compounds that appeared to be sensitive to detergent-dependent inhibition varied between the two screens, highlighting that compound aggregation is conditional and assay-specific. In a 96-well format assay, Feng et al.
(2005) found that 19% of the 1,030 drug-like compounds tested demonstrated detergent-dependent inhibition when screened at 30μM [32
]. For a 1536-well assay format, performed against a titration (3nM-30μM) of each of ~70K compounds (quantitative HTS-qHTS), Feng et al.
(2007) found that 95% of the actives identified in the screen were detergent-sensitive inhibitors, and consisted of 1.7% of the total library screened (PubChem AIDs 584, 585) [12
]. A screen of ~200K compounds targeting the cysteine protease cruzain (PubChem AID 2249) revealed that approximately 1.9% of the library were detergent-sensitive inhibitors, indicating that the prevalence of this type of assay interference is not library specific [11
]. In addition, compounds found to be aggregation-based inhibitors of AmpC were not necessarily found to inhibit cruzain, and vice versa, again underscoring the context-dependence of this phenomenon [11
]. Results from these screens indicate that aggregation-based inhibition has the potential to significantly inflate the number of apparent actives identified from a screen.