Trading a little aromaticity for a little 3D

We are all very aware of the negative impact of aromatic groups on physchem properties, especially on solubility. The negative impact on the latter comes from the strong crystal packing that you can find between aromatic groups. With this in mind, replacing aromatics by their saturated analogs has often been used as a strategy to improve solubility.

The following examples showcase how one could add a little 3D in heteroaromatics while partially maintaining some of the aromatic attributes. Unfortunately, I could not find solubility data supporting the impact, these are just transforms to keep in mind and to eventually try out.

The first one uses 3,3-disubstituted 3H-indoles as exemplified in this recent CDK4/CDK6 program (EuropJMedChem2018). The authors were able to replace the benzimidazole by a disubstituted 3H indole and maintain potency. 

Although there is limited data (other then SAR), their lead compound was used QD for 4 weeks in a xenograft model, suggesting, the PK was good enough and no major stability (metabolic or chemical) was detected.

Looking at the following set of measured pKa, the sp2 nitrogen maintains a certain level of basicity (CanJChem1993).

Even if it turns out to have limited impact on physprops, bringing the sp3 carbon in the ring allows interesting vectors to explore as illustrated below.

The second one highlights an interesting use of a benzodiazepine as a quinazoline bioisostere (JMedChem2019).

I’m not sure this one is as general as the previous one but it certainly works in this case (both compounds are equipotent), maintaining roughly the same angles (as seen in the protein co-structure). The seven-membered ring brings a nice little pucker to the flat aromatic structure and could impact solubility (data is given in this case but it is hard to really understand the impact in this case with di-basic compounds).

The last example I wanted to highlight is the rarely used benzothiophene oxide. I have to admit I had never considered this as something to suggest, but it is slowly growing on me. First of all, there is precedent –PD144795 by Parke Davis scientists (JMedChem1995) and although there is no PK data shared in the article, the authors have used this molecule in vivo and claim it was well absorbed and not readily metabolized.

You’ll notice that the molecule is chiral. That’s the really cool part: it turns our that the activity comes only from one enantiomer. A quick look in the CSD and the PDB comes up with a handful of hits but it is clear the oxygen sits- as expected- out of plane. The 5xdg highlights a nice H-bond interaction with a water molecule.

One additional piece of data can be found with the impact of the oxidation on measured logD on one matched pair (measured logD 4.4->1.9 ; JACS2010) depicted on the right.

So, as often when contemplating sulfoxides, this will have a drastic impact on your physprops and could ruin your permeability. However, you have to admit that the pretty boring and lipophilic benzothiophene (a very scientific way of describing this) suddenly became full of possibilities … add the chirality and they become so much more then the more widely used pyridine oxides… could these be prodrugs as in Sulindac? … I’ll keep you updated if I find additional data.

#drug #drugdiscovery #medicinal #medicinalchemistry #Drug Hunter #CSD