Abstact

Starting from dihydropyridopyrimidine benzyl ester 1, we pursued a macrocyclization strategy by linking its two aryl rings, hypothesizing that decreasing the conformational flexibility of the ester side chain would increase bromodomain and extra-terminal (BET) protein affinity and selectivity. We prepared 14 analogs and tested them in our fluorescent polarization (FP) assay for BRDT-1 and BRD4-1 affinity. Based on their Ki values, we selected compounds 6b (BRDT-1 Ki = 1.05 μM and BRD4-1 Ki = 0.68 μM) and 6d (BRDT-1 Ki = 0.86 μM and BRD4-1 Ki = 0.70 μM) for further testing. Differential Scanning fluorimetry (DSF) experiments with the BD1 and BD2 proteins of BRD4 and BRDT showed that the most significant increases in the melting temperatures occurred for BRDT-2 for both compounds (13 °C for 6b and 8.9 °C for 6d). Preferential binding to the second bromodomain of BRDT-2 was further confirmed by protein-observed fluorine NMR with the tandem bromodomain of BRDT. A BROMOscan showed that both compounds are pan-BET-BD2 selective (KD = 33-160 nM). A bromoMAX assay with 32 bromodomains verified BET bromodomain selectivity for 6d. The co-crystal structure between macrocyclic analogs 6b and 6d and BRD4-1 shows that the two molecules adopt almost identical conformations despite different spacer lengths. We posit that the increased BD2 selectivity could result from pi-stacking (and additional H-bonds) between the inhibitors and a His residue that is conserved across BET-BD-2 but is absent in BET-BD-1. Compound 6d inhibited MM.1S cancer cell growth with an IC50 of 2.6 μM. The study exemplifies how constraining conformational flexibility can impart target selectivity. The results indicate that the macrocyclization strategy achieved an increase in pan-BD1-affinity and comparable pan-BD2-affinity compared to lead compound 1.