Discovery of a cofactor-independent inhibitor of Mycobacterium tuberculosis InhA

Abstract

New antitubercular agents are needed to combat the spread of multidrug- and extensively drug-resistant strains of Mycobacterium tuberculosis. The frontline antitubercular drug isoniazid (INH) targets the mycobacterial enoyl-ACP reductase, InhA. Resistance to INH is predominantly through mutations affecting the prodrug-activating enzyme KatG. Here, we report the identification of the diazaborines as a new class of direct InhA inhibitors. The lead compound, AN12855, exhibited in vitro bactericidal activity against replicating bacteria and was active against several drug-resistant clinical isolates. Biophysical and structural investigations revealed that AN12855 binds to and inhibits the substrate-binding site of InhA in a cofactor-independent manner. AN12855 showed good drug exposure after i.v. and oral delivery, with 53% oral bioavailability. Delivered orally, AN12855 exhibited dose-dependent efficacy in both an acute and chronic murine model of tuberculosis infection that was comparable with INH. Combined, AN12855 is a promising candidate for the development of new antitubercular agents.

Figure 1.. Complex crystal structures of oxaborole and diazaborine inhibitors with M. tuberculosis InhA.

(A) AN2918 (yellow) forms a ternary complex with NAD⁺ (cyan) and InhA (blue). Enzyme residues at the bottom of the deep pocket are disordered. (B) AN3438 (gray) forms a ternary complex with NAD⁺ (cyan) and InhA (blue), and a relayed hydrogen-bonding network between the oxime and Glu219 and Arg195 is highlighted. (C) AN12855 (green) forms a binary complex with InhA (blue) occupying both the NAD⁺ and the deep substrate pocket. For all images, hydrogen bonds formed between the inhibitors and InhA are highlighted in yellow dash lines. Enzyme residues involved in hydrogen bonds are shown in blue sticks and other key binding residues in blue lines.

Figure 2.. Thermodynamic analysis of the interaction between diazaborines and M. tuberculosis InhA.

Binding of (A) AN12855 and (B, C) AN12908 with M. tuberculosis InhA as measured by ITC. For each compound, the interaction with InhA was measured in either (A, B) the absence of cofactor or (C) the presence of NAD⁺. InhA, compounds, and cofactors were added as described in the Materials and Methods section.

Figure 3.. In vitro kill kinetics of AN12541 and AN12855 against M. tuberculosis.

In vitro kill kinetics of AN12541 and AN12855 against M. tuberculosis under replicating conditions (A) AN12541 and AN12855 (B) INH. Limit of detection is marked by dashed lines.

Figure 4.. AN12855 is efficacious in acute and chronic murine models of TB infection.

(A, B) In vivo efficacy in a murine GKO (C57BL/6-Ifng^tm1Ts) model of acute TB. Compounds were dosed orally daily for 9 d after 14 d of infection (start) with a low-dose aerosol of M. tuberculosis Erdman. Mean (A) lung and (B) spleen log₁₀ CFUs were determined from five mice at the start of treatment and 1 d following the last day of dosing. (B, C) In vivo efficacy in a murine BALB/c model of chronic TB infection. Compounds were dosed orally 5 d a week for 8 weeks after infection with M. tuberculosis Erdman with a low-dose aerosol 27 d prior (start). Mean (B) lung and (C) spleen log₁₀ CFUs were determined from five to six mice at the start of treatment and following 2, 4, and 8 weeks of treatment. For (A–D), statistical analysis was performed as described in the Materials and Methods section; *P < 0.05.

Figure S1.. AN12855 is efficacious in acute murine models of TB infection.

(A, B) In vivo efficacy in a murine GKO (C57BL/6-Ifng^tm1Ts) model of acute TB with higher doses. Compounds were dosed orally daily for 9 d after 14 d of infection (start) with a low-dose aerosol of M. tuberculosis Erdman. Mean (A) lung and (B) spleen CFU were determined from five mice at the end.