BI p38 inhibitors: breaking from the urea

Boehringer Ingelheim has a new series of p38 kinase inhibitors (DOI: 10.1021/jm070415w), where they seek to do something different from their famous diaryl urea inhibitors. These diaryl urea inhibitors have the single oxygen of a morpholino ring bonding to the NH group of the backbone residue (Met in this case). BI decided to use a computer program called LigBuilder to find a suitable link between the newly envisaged backbone-amide bonding pyridine and a sulfonamide which bound to the Phe DFG out residue, thus blocking one conformation of the kinase.




LigBuilder helped them find a 5,6 fused heterocyle ring which would substitute in part for the urea moiety; the best activity came from a benzothiophene ring. An interesting physical-organic issue came up at this point. They had assumed that the sulfur of the benzothiophene would prefer to be trans to the oxygen of the succeeding amide. However, in the protein, things changed and the sulfur prefered to be cis to the oxygen, because the trans form had electrostatic repulsion with a Glu bound to the amide NH. This was confirmed by the crystal structure. It's important to note that it was essentially through trial and error that they discovered this; they designed and synthesized two compounds based on the intuitive hypothesis, and when there was a radical drop in potency, surmised that perhaps something funny was going on.



Why they chose to add an ortho -NH2 to the backbone binding pyridine at this late stage is something that puzzles me. The ortho amino pyridine/pyrimidine moiety is well-known to form a set of correlated hydrogen bonds to the backbone carbonyl oxygen and the NH and is present in hundreds of inhibitors, and it is not surprising that this moiety increased potency. Standard stuff like inhibition of LPS-induced TNF-alpha production was noted. In the end, they had a good viable inhibitor.


X=S, R=NH2 etc.