Verapamil and its metabolite norverapamil inhibit the Mycobacterium tuberculosis MmpS5L5 efflux pump to increase bedaquiline activity

Abstract

Bedaquiline is the cornerstone of a new regimen for the treatment of drug-resistant tuberculosis. However, its clinical use is threatened by the emergence of bedaquiline-resistant strains of Mycobacterium tuberculosis. Bedaquiline targets mycobacterial ATP synthase but the predominant route to clinical bedaquiline resistance is via upregulation of the MmpS5L5 efflux pump due to mutations that inactivate the transcriptional repressor Rv0678. Here, we show that the MmpS5L5 efflux pump reduces susceptibility to bedaquiline as well as its new, more potent derivative TBAJ-876 and other antimicrobial substrates, including clofazimine and the DprE1 inhibitors PBTZ-169 and OPC-167832. Furthermore, the increased resistance of Rv0678 mutants stems entirely from increased MmpS5L5 expression. These results highlight the potential of a pharmacological MmpS5L5 inhibitor to increase drug efficacy. Verapamil, primarily used as a calcium channel inhibitor, is known to inhibit diverse efflux pumps and to potentiate bedaquiline and clofazimine activity in M. tuberculosis. Here, we show that verapamil potentiates the activity of multiple diverse MmpS5L5 substrates. Using biochemical approaches, we demonstrate that verapamil does not exert this effect by acting as a disruptor of the protonmotive force used to power MmpS5L5, as previously proposed, suggesting that verapamil inhibits the function of the MmpS5L5 pump. Finally, norverapamil, the major verapamil metabolite, which has greatly reduced calcium channel activity, has equal potency in reducing resistance to MmpS5L5 substrates. Our findings highlight verapamil's potential for enhancing bedaquiline TB treatment, for preventing acquired resistance to bedaquiline and other MmpS5L5 substrates, while also providing the impetus to identify additional MmpS5L5 inhibitors.

Keywords: MmpS5L5 drug efflux pump; Mycobacterium tuberculosis; TBAJ-587 and TBAJ-876; bedaquiline; verapamil and norverapamil.

Fig. 1.

Mutations in Rv0678 confer resistance to structurally and functionally diverse antitubercular drugs. (A) Fold change in MIC between wild-type Mycobacterium tuberculosis and isogenic Rv0678 G65GfsX10 strains against the stated drugs. Results are representative of at least three independent experiments. (B) Fold change in MIC between wild-type M. tuberculosis and isogenic ∆S5L5::loxP and ∆S5L5::loxP + pMINTF3 M. tuberculosis MmpS5L5 strains against the stated drugs. Symbols represent the results of technical replicates (N = 3 to 6). Results are representative of at least three independent experiments. The drug MICs for the tested M. tuberculosis strains are shown in SI Appendix, Table S2.

Fig. 2.

CRISPRi knockdown of MmpS5 or MmpL5 increases drug susceptibility in wild-type and Rv0678 mutant strains of M. tuberculosis. (A) Fold change in MICs upon transcriptional knockdown of MmpS5 or MmpL5 in wild-type M. tuberculosis. Fold change in MIC is expressed relative to the wild-type parent strain containing a nontargeting CRISPRi plasmid pLJR965. (B) Fold change in MICs upon transcriptional knockdown of MmpS5 or MmpL5 in an M. tuberculosis Rv0678 background. Fold change in MIC is expressed relative to the Rv0678 mutant strain containing a nontargeting CRISPRi plasmid pLJR965. Red dotted lines indicate the fold change of the wild-type MIC relative to the M. tuberculosis Rv0678 mutant. Symbols represent technical replicates (N = 2). Drug MICs of the wild-type and Rv0678 mutant are given in SI Appendix, Table S2. The differences in MIC fold changes across the tested drugs as compared to those in Fig. 1 are due to methodological differences in the MIC assays between the two laboratories performing them.

Fig. 3.

Verapamil and its major metabolite potentiate the activity of MmpL5 substrates in wild-type and MmpS5L5 overexpressing strains (Rv0678 mutant and complemented MmpS5L5 knockout). (A) Isobolograms showing the geometric mean MIC (three technical replicates) of bedaquiline, clofazimine, PBTZ-169, OPC-167832, TBAJ-876, and TBAJ-587 as a function of verapamil concentration. (B) Isobolograms showing the geometric mean MIC (three technical replicates) of bedaquiline, TBAJ-876, and TBAJ-587 as a function of norverapamil concentration. Because the norverapamil MIC for M. tuberculosis was lower than that of verapamil for the knockout and its complement (100 µM vs. >100 µM), we could not assess its effect beyond 50 µM in these experiments. Note that the Y-axis scales differ between graphs.

Fig. 4.

Verapamil does not function as a disruptor of the protonmotive force in mycobacteria. (A) NADH-driven proton translocation in IMVs of Mycobacterium smegmatis (0.25 mg/mL). Quenching of acridine orange fluorescence in IMVs was initiated with 0.5 mM NADH, and at the indicated time points, either DMSO (control), verapamil (1 to 500 µM), or the protonophore CCCP at 10 µM were added to collapse the proton gradient (reversal of acridine orange fluorescence). IMVs were prepared from M. smegmatis cells grown under aerobic conditions in Hartman’s de Bont (HdB) minimal medium supplemented with 0.2% glucose (w/v). Traces were normalized to a starting value of approximately 1,000 relative fluorescence units (RFU), and measurements were made with a Cary Eclipse Fluorescence spectrophotometer. The excitation and emission wavelengths were 493 and 530 nm, respectively, and experiments are representative of a technical triplicate. (B) Intracellular pH [(7-¹⁴C)benzoic acid, 10 µM final concentration] and (C) membrane potential [(³H)methyltriphenylphosphonium iodide, 5 nM final concentration] of glucose-energized cell suspensions (10 mM glucose, OD600 = 1.0) in the presence of either DMSO, valinomycin, nigericin, CCCP, or verapamil as indicated. Error bars indicate SD from three biological replicates. (D–F) The effect of CCCP and verapamil on the oxygen consumption rates (OCR) of either (D) glucose-energized (10 mM final) (E) succinate-energized (10 mM final) cell suspensions of M. smegmatis (OD600 = 1.0) or (F) NADH-energized (0.5 mM final) IMVs of M. smegmatis. OCR was measured using a Oroboros Oxygraph-2k and inhibitors were added once the OCR reached steady state (~3 min after electron donor addition). SRC; spare respiratory capacity. Error bars for OCR measurements represents SEM of three biological replicates and three technical replicates for IMVs.