Abstact

Rifampin (RMP), a very potent inhibitor of the Mycobacterium tuberculosis (MTB) RNA polymerase (RNAP), remains a keystone in the treatment of tuberculosis since its introduction in 1965. However, rifamycins suffer from serious drawbacks, including 3- to 9-month treatment times, Cyp450 induction (particularly problematic for HIV-MTB coinfection), and resistant mutations within RNAP that yield RIF-resistant (RIFR) MTB strains. There is a clear and pressing need for improved TB therapies. We have utilized a structure-based drug design approach to synthesize and test novel benzoxazinorifamycins (bxRIF), congeners of the clinical candidate rifalazil. Our goal is to gain binding interactions that will compensate for the loss of RIF-binding affinity to the (RIFR) MTB RNAP and couple those with substitutions that we have previously found that essentially eliminate Cyp450 induction. Herein, we report a systematic exploration of 42 substituted bxRIFs that have yielded an analogue (27a) that has an excellent in vitro activity (MTB RNAP inhibition, MIC, MBC), enhanced (∼30-fold > RMP) activity against RIFR MTB RNAP, negligible hPXR activation, good mouse pharmacokinetics, and excellent activity with no observable adverse effects in an acute mouse TB model. In a time-kill study, 27a has a 7 day MBC that is ∼10-fold more potent than RMP. These results suggest that 27a may exhibit a faster kill rate than RMP, which could possibly reduce the clinical treatment time. Our synthetic protocol enabled the synthesis of ∼2 g of 27a at >95% purity in 3 months, demonstrating the feasibility of scale-up synthesis of bxRIFs for preclinical and clinical studies.