Down the Rabbit Hole Pitfalls When Dealing with Low-Affinity Binders

1 Down the Rabbit Hole Pitfalls When Dealing with Low-Affinity Binders 

Many types of reactive molecules are well known to medicinal chemists acyl halides, aldehydes, aliphatic esters, aliphatic ketones, alkyl halides, anhydrides, alpha-halocarbonyl compounds, aziridines, 1,2-dicarbonyl compounds, epoxides, halopyrimidines, heteroatom-heteroatom single bonds, imines, Michael acceptors and (l-heterosubstituted carbonyl compounds, perhalo ketones, phosphonate esters, thioesters, sulfonate esters, and sulfonyl halides, to name a few [14]. This is not to say that these functionalities are not useful - some even appear in approved drugs -but all of these can react covalently with proteins, and thus should be regarded with suspicion. However, molecules can react covalently with proteins even if they do not contain functionalities that raise alarm. Jonathan Baell has referred to these as pan assay interference compounds, or PAINS, and has published a list of moieties to watch out for, as well as strategies to detect them [15, 16]. 

Even less obvious are molecules that may not react with proteins directly but that act as oxidizers, for example by generating hydrogen peroxide, which can in turn 

The degree to which this is a problem can be appreciated by a screen of 70,563 molecules to discover inhibitors of the enzyme AmpC fi-lactamase [25], Of the 1,274 hits, 1,213 turned out to be aggregators - more than 95% Even worse, these compounds often display structure-activity relationships (SAR), and the effect can persist even at fairly low concentrations. Recently, a series of cruzain inhibitors with IC50 values as low as 200 nM were reported, but follow-up studies determined 

---