A host-directed oxadiazole compound potentiates antituberculosis treatment via zinc poisoning in human macrophages and in a mouse model of infection

Abstract

Antituberculosis drugs, mostly developed over 60 years ago, combined with a poorly effective vaccine, have failed to eradicate tuberculosis. More worryingly, multiresistant strains of Mycobacterium tuberculosis (MTB) are constantly emerging. Innovative strategies are thus urgently needed to improve tuberculosis treatment. Recently, host-directed therapy has emerged as a promising strategy to be used in adjunct with existing or future antibiotics, by improving innate immunity or limiting immunopathology. Here, using high-content imaging, we identified novel 1,2,4-oxadiazole-based compounds, which allow human macrophages to control MTB replication. Genome-wide gene expression analysis revealed that these molecules induced zinc remobilization inside cells, resulting in bacterial zinc intoxication. More importantly, we also demonstrated that, upon treatment with these novel compounds, MTB became even more sensitive to antituberculosis drugs, in vitro and in vivo, in a mouse model of tuberculosis. Manipulation of heavy metal homeostasis holds thus great promise to be exploited to develop host-directed therapeutic interventions.

Fig 1. Identification of new compounds inhibiting the intracellular growth of MTB.

(A) Human monocyte-derived macrophages were infected with GFP-expressing MTB (GFP-MTB) and incubated with epigenetics-related compounds. After 96-h treatment, cells were labeled with Hoechst 33342 and HCS CellMask Blue. Fluorescence was analyzed by the Opera Phenix Plus High-Content Screening System. Quantification of GFP staining and enumeration of cells were performed using Columbus image analysis software. Only compounds that were not toxic (cell viability >75%, in blue in the upper panel) were kept for further analysis. The number of live cells and the areas of intracellular bacteria (px²: pixels²), lower panel) were expressed as the percentage in compound-treated cells compared to cells incubated with DMSO. (B) Representative confocal images of macrophages infected with MTB (green) and treated with MC3465. Hoechst 33342 and HCS CellMask (blue) were used to delimit nuclei and cytoplasm shapes, respectively. Scale bar: 10 μm. (C) MTB-infected macrophages were treated with MC3465 (1 and 10 μM). The number of intracellular bacteria was enumerated at 0, 24, and 96 h postinfection. One representative experiment (of at least 3) is shown. (D) Macrophages were infected with clinical strains of MTB, namely, GC1237, CDC1551, and Myc5750 and were treated with MC3465 (10 μM). The number of intracellular bacteria was enumerated at 96 h posttreatment. (E) Macrophages were infected with L. monocytogenes or S. Typhimurium. Intracellular bacteria were enumerated at 0 and 24 h postinfection. (F) Growth of MTB in culture liquid medium in the presence of MC3465 (10 and 50 μM). Data are representative of 2 independent experiments. (G) MTB-infected macrophages, MTB separated from macrophages using Transwell inserts, and MTB alone were treated with MC3465 (10 μM) or DMSO for 96 h. Cells were lysed and the number of intracellular bacteria was enumerated. Panels D, F, and G: Data are representative of 2 independent experiments. Error bars represent the mean ± SD. One-way ANOVA test (1C) and t test (1D, 1G) were used. ns, not significant, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. The data underlying the graphs shown in the figure can be found in S1 Data.

Fig 2. Differentially expressed genes upon MC3465 treatment.

Naive and MTB-infected macrophages derived from 4 individual donors were treated with MC3465 (10 μM) for 4 h and 24 h. Differentially expressed genes were identified by mRNAseq. (A) Volcano plot showing differentially expressed genes due to MC3465 treatment in MTB-infected cells (p-value <0.05, Log Fold Change (FC) <−0.5 and >0.5). (B) Heatmap showing expression of genes differentially expressed in naive (UNF) and MTB-infected cells (INF) upon MC3465 treatment. Genes related to cellular zinc ion homeostasis and type I IFN signaling pathway were represented at 4 and 24 h, respectively. Genes that were not differentially expressed were represented by a grey square.

Fig 3. Free zinc is released upon MC3465 treatment and accumulates within the mycobacterial phagosome.

(A) Macrophages were infected with MTB during 2 h and incubated with the zinc-chelating agent TPEN (2.5 μM). After 1 h, cells were treated with MC3465 for 4 h. Cells were fixed and stained with the free zinc-specific fluorescent probe FluoZin 3-AM. Scale bar, 10 μm. (B) Average FluoZin 3-AM signal intensity, for at least 400 cells per condition. (C, D) Macrophages were infected with DsRed-MTB and incubated with TPEN, ZnSO₄ (100 μM), or MC3465 (10 μM) during 4 h. Cells were then fixed and stained with FluoZin 3-AM. Data are representative of 2 independent experiments. (C) Representative image of FluoZin 3-AM association with MTB. (D) Quantification of FluoZin 3-AM association with DsRed-MTB. At least 100 phagosomes were counted per condition. (E) MTB-infected cells were incubated with TPEN (2.5 μM). After 1 h, cells were treated with MC3465 (10 μM). The number of intracellular bacteria was enumerated at 96 h posttreatment. Data are representative of 3 independent experiments. One-way ANOVA test was used. Error bars represent the mean ± SD. ns, not significant, * p < 0.05, *** p < 0.001, **** p <0.0001. The data underlying the graphs shown in the figure can be found in S2 Data.

Fig 4. The mycobacterial P_1B-ATPase metal exporter, CtpC, along with the chaperone-like protein PacL1, reverses zinc intoxication mediated by MC3465.

(A) RT-qPCR analysis of ctpC expression upon incubation with MC3465 (10 μM) or ZnSO₄ (100 μM). After 4, 24, and 96 h treatment, total cellular RNA was extracted and analyzed by RT-qPCR. Data are normalized relative to ftsZ gene. (B) Growth of the triple ctp mutant, expressing P_pacL1-driven CtpC and anhydrotetracycline (Atc)-inducible PacL1, was observed in liquid culture medium supplemented with ZnSO₄ (200 μM) and Atc (200 ng/ml). (C) MTB-infected macrophages were treated with either DMSO or MC3465 (10 μM) in the presence or absence of Atc (200 ng/ml). The number of intracellular bacteria was enumerated at 96 h posttreatment. Data are representative of 2 independent experiments. Error bars represent the mean ± SD. One-way ANOVA test was used. ns, not significant, *p < 0.05, **p < 0.01. The data underlying the graphs shown in the figure can be found in S3 Data.

Fig 5. Structure–activity relationship analysis of MC3465 and identification of MC3466 as a more effective analog.

(A) Macrophages were infected with GFP-MTB and incubated with 29 MC3465 analogs (10 μM). Images were acquired by automated confocal microscopy followed by image analysis 96 h postinfection, as described in Fig 1A and 1B. The spot area per cell was expressed as the percentage of GFP area in compound-treated cells compared to cells incubated with DMSO. (B) Structures of MC3465 and MC3466. (C) MTB-infected macrophages were treated with MC3465 or MC3466 (10 μM). After 96 h, the number of intracellular bacteria was enumerated. One representative experiment (of at least 3) is shown. (D) MTB growth in liquid culture medium in the presence of the analog MC3466 at different concentrations, determined by OD₆₀₀. Data are representative of 3 independent experiments. (E) The dose–response curves (DRCs) for MC3465 and MC3466 were obtained by automated confocal microscopy followed by image analysis. The ratio of GFP area in compound-treated macrophages, compared to cells incubated with DMSO, was normalized with the negative control DMSO (0% inhibition) and the positive control RIF (100% inhibition). (F) MTB-infected cells were treated with different concentrations of MC3465 or MC3465 (10 μM). After 96 h, bacteria were enumerated by CFU and the IC50 of each compound was determined. (G-I) MTB-infected macrophages were incubated with TPEN (2.5 μM). After 1 h, cells were treated with MC3465 or MC3466 (10 μM). (G) Representative image of FluoZin 3-AM staining at 4 h posttreatment. Scale bar, 20 μm. (H) Average FluoZin 3-AM signal intensity for at least 400 cells per condition. (I) The number of intracellular bacteria was enumerated at 96 h posttreatment. Panels E and F: Data are representative of 2 independent experiments. One-way ANOVA test was used. Error bars represent the mean ± SD. ns, not significant, ** p < 0.01, *** p < 0.001, **** p <0.0001. The data underlying the graphs shown in the figure can be found in S4 Data.

Fig 6. MC3466 potentialized the efficacy of rifampicin and bedaquiline in MTB-infected macrophages.

(A) Macrophages were infected with the drug-susceptible MTB H37Rv and treated with various concentrations of RIF in association or not with MC3466. After 96 h, cells were lysed and the number of intracellular bacteria enumerated. (B) As in (A) except that RIF has been replaced by BDQ. (C, D) Heatmaps representing variation in drug combination, using the Bliss independence model, ranging from antagonism (blue) to synergy (red). Cells were infected with GFP-MTB and were treated with a range of concentrations of MC3466 and RIF or BDQ. Images were acquired by automated confocal microscopy followed by image analysis 96 h posttreatment, as described in Fig 1. Panels A and B: One representative experiment (of at least 3) is shown. Panels C and D: One representative experiment (of 2) is shown. One-way ANOVA test was used. Error bars represent the mean ± SD. ns, not significant, *p < 0.05, **p < 0.01, ****p < 0.0001. The data underlying the graphs shown in the figure can be found in S5 Data.

Fig 7. Synergy between RIF and MC3466 in MTB-infected mice.

(A) C57BL/6J mice were infected by aerosol with MTB. After 1 wk, mice were treated for 2 wk, 6 d a week with MC3466 with or without RIF. Lungs were harvested, and the number of bacteria was estimated by CFU. (B) Representative hematoxylin and eosin stains of lungs, 2 wk after treatment. Scale bar: 1 mm. (C) Heatmap representing the histological scores upon RIF and MC3466 treatment. Lung sections are graded according to the percentage of abnormal area, presence of interstitial syndrome, alteration of bronchiolar epithelium, severity of inflammation, and consolidation of lung parenchyma. (D) Quantification of IL-1β in lung lysates from MTB-infected mice treated with MC3466 for 2 wk by ELISA. One representative experiment (of 2) is shown. Error bars represent the mean ± SD. One-way ANOVA test was used. **p < 0.01, ****p < 0.0001. The data underlying the graphs shown in the figure can be found in S6 Data.