Similarly towards the acylsulfamoyl benzoxaboroles the boron

Similarly, towards the 5-acylsulfamoyl benzoxaboroles, the boronate intermediate was prepared according to , while the boronate , an intermediate towards 4-acylsulfamoyl benzoxaboroles, was prepared according to . The sulfonamides and used in the synthesis were made from the starting sulfonyl chlorides and , respectively. With these key intermediates in hand, we set out to investigate the P1–P3/P2–P4 macrocyclic series and also explore the impact of regioisomers of acylsulfamoyl benzoxaboroles. The P1–P3 macrocyclic inhibitors – were synthesized according to . Initially one of our targeted compounds was the hypothetical P1–P3 macrocyclic inhibitor in which 6-acylsulfamoyl benzoxaborole is used to replace the cyclopropyl acylsulfonamide in danoprevir (structure shown in ). During the synthesis, we found P4-Boc group did not survive the RG7227 synthesis treatment in the final step and thus it was replaced with a chemically more-stable cyclopentyl carbamate. The P1–P3 macrocyclic acid , derived from danoprevir, was made according to a published procedure. Acid was converted to the corresponding acylsulfonamides using boronate or in the presence of HATU and DIEA. The removal of pinacol/acetate groups and spontaneous formation of benzoxaborole ring was accomplished by acid treatment in the presence of isobutyl boronic acid and 1N HCl to afford the desired product and , respectively. The P2–P4 macrocyclic inhibitors – were synthesized according to . The starting macrocyclic acid , derived from MK-7009, was made according to a reported procedure. Acid was converted to the corresponding acylsulfonamides by reaction with boronate , or as described previously, followed by acid treatment to give isomeric benzoxaboroles –. These compounds were evaluated for enzymatic potency in FRET assay with NS3/4A 1a protease domain. The cellular activity was determined using 1a and 1b HCV replicon assays. As shown in , P1–P3 macrocyclic compounds – and P2–P4 macrocyclic compounds – inhibited NS3 1a enzyme with IC values in the sub-nanomolar range (IC=0.3–0.8nM). They were equipotent against NS3 1a enzyme, compared to danoprevir. These benzoxaboroles exhibited low nanomolar potency against replicon 1b (EC=8.0–20nM), approaching that of danoprevir (EC=1.1nM). However, a higher shift between the enzyme potency IC and replicon activity EC (especially for replicon 1a) was observed for these benzoxaborole inhibitors than that observed for danoprevir, which could be attributed to their poor cell membrane permeability. Interestingly, the regioisomers of acylsulfamoyl benzoxaboroles appear to have a minimal influence on the inhibitor activity. For the P1–P3 macrocyclic series, 6-acylsulfamoyl benzoxaborole is equipotent in the enzyme and replicon 1a and 1b assay, compared to 5-acylsulfamoyl analog . For the P2–P4 macrocyclic series, very little difference in potency was observed with 6-, 5- or 4-acylsulfamoyl benzoxaboroles (–) in the enzyme and replicon assay. These results demonstrate that regioisomers of acylsulfamoyl benzoxaboroles are well tolerated, consistent with the shallow enzyme binding pocket that is able to incorporate a variety of diverse inhibitors. Our results in acylsulfamoyl benzoxaboroles show that the P1–P3 macrocyclic inhibitors are equipotent in the enzyme and replicon assay compared to the P2–P4 macrocyclic inhibitors, suggesting that these two macrocyclization strategies are equally effective in enhancing inhibitor activity. Further in vivo PK evaluation was undertaken to prioritize these two potent sub-series, which is shown in . Despite increased solubility relative to danoprevir, both the P1–P3 macrocyclic inhibitor and the P2–P4 macrocyclic inhibitor displayed undetectable to low oral exposures in rats. Portal vein sampling evidenced very low absorption of and . By contrast, danoprevir exhibited calculated 17.4% absorption and 20% oral bioavailability in rats. These results suggest that the poor oral bioavailability of the benzoxaborole-containing compounds is caused by their limited absorption. We believe that the introduction of an unsubstituted benzoxaborole moiety results in unbalanced physicochemical properties, such as high molecular weight (MW) and high polar surface area (PSA), of the final molecule (). Further rebalancing of MW and PSA with carefully redesigned benzoxaborole moieties could improve the permeability of the final inhibitors, contributing to simultaneous increase of potency in the replicon assay as well as improvement of bioavailability in the in vivo PK assays. In fact, we applied this strategy to another benzoxaborole-based series, resulting in more drug-like properties in these molecules (manuscript in preparation).