Inhibiting Mycobacterium tuberculosis CoaBC by targeting an allosteric site

Abstract

Coenzyme A (CoA) is a fundamental co-factor for all life, involved in numerous metabolic pathways and cellular processes, and its biosynthetic pathway has raised substantial interest as a drug target against multiple pathogens including Mycobacterium tuberculosis. The biosynthesis of CoA is performed in five steps, with the second and third steps being catalysed in the vast majority of prokaryotes, including M. tuberculosis, by a single bifunctional protein, CoaBC. Depletion of CoaBC was found to be bactericidal in M. tuberculosis. Here we report the first structure of a full-length CoaBC, from the model organism Mycobacterium smegmatis, describe how it is organised as a dodecamer and regulated by CoA thioesters. A high-throughput biochemical screen focusing on CoaB identified two inhibitors with different chemical scaffolds. Hit expansion led to the discovery of potent and selective inhibitors of M. tuberculosis CoaB, which we show to bind to a cryptic allosteric site within CoaB.

Targeting Mycobacterium tuberculosis CoaBC through Chemical Inhibition of 4′-Phosphopantothenoyl-l-cysteine Synthetase (CoaB) Activity

“…To assess the mode of inhibition of CoaBC by 1f , competition experiments were carried out between the compound and the three substrates of CoaB (PPA, CTP, and l -cysteine) using the EnzChek coupled enzyme assay, as previously described. 41 We first confirmed that compound 1f was inactive on the coupled reporter enzymes at 100 μM. The IC 50 for this compound using the EnzChek assay was determined to be 24.3 μM against the Mtb CoaBC, comparable to that observed in the high-throughput screen ( Figure 3 a).…”

“…High-throughput screening of 215 000 small molecules from the DDU compound library was carried out to identify inhibitors of Mtb CoaB activity using an adaptation of the BIOLMOL Green end-point pyrophosphate quantification assay, as previously described. 41 This compound library predominantly comprises commercially available molecules representing a wide range of chemical space that possessing the favorable physicochemical and molecular properties required for a potential preclinical drug candidate, with a small subset of the library (<0.5%) comprising proprietary compounds with known phenotypic activity against a number of various pathogens. The primary screen, which had an overall hit rate of 0.6%, led to the identification of compound 1a , with an IC 50 of 9.9 μM against Mtb CoaB.…”

“…This is in contrast to our previous observation with allosteric inhibitors of CoaBC, which bound preferentially in the presence of all three substrates. 41 We therefore decided to employ a different approach to further validate these findings. Using differential scanning fluorimetry (DSF), the melting temperature of CoaBC in the absence of natural ligands was determined to be 46 °C ( Table 2 ; Figure S4 ).…”

“…A crystal structure of CoaB with 1f bound could not be obtained, but the observed mode of inhibition together with known movements that occur in the protein after PPA binding ( Figure S5 ), with a flexible loop covering the PPA binding site upon binding of this substrate, 52 suggests a different binding mode for 1f with distinct properties from the previously reported allosteric inhibitors. 41 …”

“…We recently reported the first crystal structure of the mycobacterial CoaBC together with the first potent and selective inhibitors of Mtb CoaB. 41 However, these potent inhibitors of CoaB activity displayed limited whole-cell activity against Mtb due to impaired permeability/efflux and intracellular metabolism. 41 In this study, we confirm the druggability of Mtb CoaBC by identifying a new CoaB inhibitor that confers whole-cell activity against Mtb and providing evidence linking the growth inhibitory effects of this compound on Mtb to inhibition of CoaBC.…”