• 2018-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2019-12
  • 2020-01
  • 2020-02
  • 2020-03
  • 2020-04
  • 2020-05
  • 2020-06
  • 2020-07
  • 2020-08
  • 2020-09
  • 2020-10
  • 2020-11
  • 2020-12
  • 2021-01
  • 2021-02
  • 2021-03
  • High throughput screening HTS of the AstraZeneca compound co


    High throughput screening (HTS) of the AstraZeneca CB-839 collection was conducted and identified oxadiazole amide () originating from a kinase directed screening library (although inactive against a panel of kinases). This compound had moderate potency as a DGAT-1 inhibitor (IC=0.52μM) and was selective against the related DGAT-2 enzyme (zero inhibition at 10μM). The log was measured at 3.0 resulting in an acceptable ligand lipophilicity efficiency (LLE) start point of 3.2. Consequently, improvement in potency and LLE were key aims throughout the optimisation campaign. Initial chemistry efforts were focused on understanding the key features necessary for potency within the core of the molecule. The oxadiazole was systematically replaced by numerous heteroaryl and phenyl ring systems () but all resulted in significant potency loss (>10-fold). Similarly, methylation of either the amide or aniline nitrogen resulted in loss of potency. The pharmacophore was not tolerant to changes anywhere but the terminal fragments. The only encouraging alteration to the amide-oxadiazole-aniline core was conversion of the aniline (IC=0.30μM, LLE=3.4) to an amide (IC=5.2μM, LLE=4.3) as shown in . Although potency was reduced, the bis-amide compound showed improved LLE and both these sub-series were progressed to lead expansion activities. The compounds described in this Letter were prepared using similar methods, which allowed diversification of R1 and R2 (). Starting with anilines , reaction with methyl chlorooxoacetate followed by hydrazine afforded acyl-hydrazines which were condensed with various isothiocyanates. Ring closure to form oxadiazoles was effected using polymer-supported carbodiimide (PS-CDI) The amide-oxadiazole-aniline sub-series was initially explored as a two-dimensional array, varying R1 and R2. Principal component analysis, using common descriptors such as log, polar surface area (PSA) and hydrogen bond counts, was used to guide design of a diverse set of compounds but all significant structural deviations away from resulted in a loss of potency (data not shown). Consequently, more conservative SAR exploration was conducted around the morpholino aryl side chain R1 (). Addition of a single F () or Me () group on the arylamide-3-position maintained potency but with lower LLE. However, at the arylamide-2-position F () resulted in a loss of potency. Lowering lipophilicity by conversion of the aryl group to a pyridine could be tolerated with the 3-aza improving potency whereas the 2-aza compound showed a significant decrease. Changing the morpholine to more lipophilic groups, for example piperidine or thiomorpholine , also improved potency. Although potent, exhibited high in vitro clearance in human microsomes (Cl=258μL/min/mg) and low oral exposure in rat (=0.02μM from 2mg/kg dose) thus was not progressed. Oxidation of to sulfoxide reduced potency. Piperazines and were both tolerated and gave improved LLE. Low aqueous solubility, typically less than 10μM was a problematic feature for all these analogues which were generally synthesized in a crystalline form. Basic compound did show higher solubility of 91μM but high clearance in rat of greater than 100% liver blood flow precluded further optimisation of such compounds. Increasing flexibility through introducing sp character was used as a tactic to improve solubility and explore SAR (). Unfortunately, solubility remained low although carbon, oxygen, nitrogen or sulfur could be used as a linker with a suitably large substituent to achieve good potency. Notably, increasingly lipophilic ethers such as compounds –, gave increasing potency suggesting a lipophilic binding interaction was responsible. Reduced potency of pyridine is consistent with this hypothesis. Lipophilicity was generally too high to be accurately measured, only and returning log values of 3.2 and 3.1 respectively, thus further exploration to achieve our dual potency and LLE aims was not justified