Medicinal Chemistry Newbies: Leave behind all your preconceived notions [Example 1]

Presumably, most of you might have encountered people saying “not to have a Michael-acceptor moiety” in your molecules.  This is due to the fact that nucleophilic species present in proteins (mostly cysteine residue) would bind to the Michael-acceptor moiety of drug thus making it unavailable for the required purpose.

 

Now the question is, Should we always be terrified of incorporating such Michael-acceptor moieties in our molecules? Answer is both Yes and No  

Recently, there was an article in J.Med.Chem. 2010, 53(7), 2719-2740 describing several moieties (substructures) featuring in molecules which are frequent hit compounds in many biochemical high throughput screens.  In this article there are a number of Michael-acceptor substructures which were described as ‘pan assay interference compounds’ (PAINS).  Any protein reactive compound comes under PAINS.  Authors go to an extent saying that such compounds are polluting scientific literature and causing nuisance.

   

On the other hand, in a review article published in Curr.Med.Chem2009;16(13):1596-629, rhodanines were called as ‘privileged scaffolds’.  Success stories of Epalrestat an aldose reductase inhibitor and glitazones (type II diabetes) were shown as examples.

 

Moving forward, efforts from Prof. Klein’s research group (Heidelberg University) provide some more insights into the biological activities for rhodanine and realated five-membered analogues.

http://pubs.acs.org/doi/abs/10.1021/jm201243p

From their own results and the literature precedents, they conclude that the so called ‘promiscuous or frequent hitters’ property of rhodanines and thiohydantoins with a exocyclic sulfur atom is not related to ‘unspecific’ properties such as aggregation or reactivity, but rather to electronic and hydrogen-bonding properties.

Justification: There are significant differences in the H-bond acceptor properties of C=S and C=O, with the sulfur having a more diffuse lone pair electron density distribution around the sulfur and frequently interacting with three and four hydrogen-bond-donor groups.  [energetics of H-bond to sulfur in C=S is inferior to those of oxygen in C=O].  However, in the context of protein-ligand-interactions, the desolvation of the ligand has to be considered (breaking of hydrogen bonds to water prior to protein binding), and the free energy penalty of breaking a ‘strong’ C=O hydrogen bond to water may result in the overall ligand binding process being more favorable for the C=S group.  In water, the stricter geometric requirements of the C=O group as hydrogen bond acceptor can more easily be satisfied than in protein-ligand complexes.  When a ligand has to ‘adjust’ to a geometry that is predefined by proteinaceous receptor, the more geometrically tolerant C=S group, which is also able to form a larger number of hydrogen-bonds has more flexibility to adjust its intermolecular interactions to the given requirements.  Thus, this property of C=S makes rhodanine scaffolds prone to bind to a large number of biological targets.  FINE TUNING might give EXTREMELY GOOD COMPOUNDS?

Now the conclusion to all of you is ‘to be open-minded’.  As newbies, one has to learn the concepts but do not have ‘preconceived notions’ that all the concepts should be taken for granted.  As you gain experience, you will be in a position to distinguish things clearly.  But for now, be open-minded and do not trash any result or publication just based on ‘mis-conceptions’ you heard of.   Concepts are correct but may not be applicable for every therapeutic area or biological target.      
**********************************************************************************