Mechanism of the dual action self-potentiating antitubercular drug morphazinamide

Abstract

Pyrazinamide (PZA) is a cornerstone of first-line antitubercular drug therapy and is unique in its ability to kill nongrowing populations of Mycobacterium tuberculosis through disruption of coenzyme A (CoA) metabolism. Unlike other drugs, PZA action is conditional and requires potentiation by host-relevant environmental stressors, such as low pH and nutrient limitation. Despite its pivotal role in tuberculosis therapy, the durability of this crucial drug is challenged by the emergent spread of drug resistance. To advance drug discovery efforts, we characterized the activity of a more potent PZA analog, morphazinamide (MZA). Here, we demonstrate that like PZA, MZA acts in part through impairment of CoA metabolism. Unexpectedly, we find that, in contrast to PZA, MZA does not require potentiation and maintains bactericidal activity against PZA-resistant strains due to an additional mechanism involving aldehyde release. Further, we find that the principal mechanism for resistance to the aldehyde component is through promoter mutations that increase expression of the mycothiol oxidoreductase MscR. Our findings reveal a dual-action synergistic mechanism of MZA that results in a faster kill rate and a higher barrier to resistance. These observations provide new insights for the discovery of improved therapeutic approaches for addressing the growing problem of drug-resistant tuberculosis.

Keywords: coenzyme A; drug discovery; pyrazinamide; thiol stress; tuberculosis.

Fig. 1.

Impact of MZA on PZA susceptible and resistant M. tuberculosis. A) Structure of MZA and potentiation of PZA followed by conversion to POA by PncA_Mtb. Dose response curves comparing MZA and PZA susceptibility for M. tuberculosis strains (B) H37Ra at pH 7.2 and (C) H37Rv at pH 5.8. D) Assessment of pantothenate antagonism of PZA and POA action on H37Ra at pH 5.8 and 7.2, respectively. E) Assessment of pantothenate antagonism of MZA action on H37Ra at pH 7.2. Determination of MZA susceptibility of PZA-resistant strains (F) H37Rv ΔsigE, H37Rv P_clpC1::himar1, H37Rv panD::himar1 and (G) H37Rv ΔpncA compared with parental H37Rv. Dose-dependent kill curves of H37Rv exposed to MZA at (H) pH of 7.2 and (I) pH 5.8 over a 14-day period. J) Kill curves performed with H37Rv under starvation conditions following a one-time treatment with PZA, MZA, or DMSO over a 60-day period at pH of 7.2. Kill curves performed with (K) H37Rv and (L) H37Rv ΔpncA (PZA resistant) incubated at neutral pH under starvation conditions with MZA replenished every 3 days. M) Time course monitoring intracellular and (N) extracellular abundance of MZA to PZA and POA from H37Ra exposed to 250 µM MZA. All assays were performed in biological triplicate with the mean displayed and error bars indicative of SD.

Fig. 2.

Transcriptional profiling of M. tuberculosis exposed to MZA, PZA, and POA. Volcano plots showing significantly DEGs in presence of (A) PZA, (B) POA, and (C) MZA. Cells were treated with 200 µM MZA, POA, or PZA at pH 5.8 for 24 h prior to RNA purification and sequencing. The MZA, POA, and PZA transcriptional profiles share many key features, including up-regulation in CoA and lipid metabolism pathways. Transcriptional changes associated with exposure of an H37Rv ΔpncA to (D) PZA and (E) MZA. Cells were treated with 200 µM MZA or PZA at pH 5.8 for 24 h prior to RNA purification and sequencing. F) Venn diagram showing up-regulated genes in MZA-, POA-, and PZA-treated cultures. G) Venn diagram showing down-regulated genes in MZA-, POA-, and PZA-treated cultures. H) GO term analysis of common differentially regulated genes (right column) with CoA synthesis, mono-carboxylic acid metabolism, stress response, and lipid modification in dark gray (>2-log fold increase) and lipid synthesis, cell wall assembly, and lipid transport in white (>2-log fold decrease). RNA-seq was performed in biological triplicate. RNA-seq data are available in Supplementary Dataset 1 and primary sequencing data is available at NCBI SRA under BioProject PRJNA1104292.

Fig. 3.

Metabolism of MZA results in release of PZA and drives thiol stress. A) mscR promoter region highlighting MZA-resistance mutations in H37Rv ΔpncA and BCG backgrounds, respectively. Inhibition curve of (B) wild-type BCG and (C) mscR promoter mutant exposed to MZA and methenamine (MET). D) Comparison of GFP expression between wild-type mscR promoter and mscR promoter variant GFP reporter constructs in BCG. All samples were normalized to the OD₆₀₀, and significance was determined using a two-tailed Student's t test using n = 6. E) Inhibition curve comparing susceptibility of H37Rv and H37Rv ΔmscR to MZA. F) BCG wild type and (G) BCG mscR promoter mutant exposed to 0.9 mM MZA or MET in the absence or presence of cystine. H) DiaMOND assay showing exposure of M. tuberculosis H37Ra to a geometric series of PZA and methenamine (MET) alone and in combination. All assays were performed in biological triplicate unless indicated otherwise with mean displayed and error bars indicative of SD.

Fig. 4.

Genome-wide fitness profiling of M. tuberculosis transposon insertion exposed to PZA, POA, and MZA. Saturated library of Tn-insertion mutants treated with (A) 233 µM PZA, (B) 105 µM POA, and (C) MZA 112 µM MZA at pH 6 to achieve a 50% reduction in growth rate. Each point corresponds to a specific TA site harboring a transposon insertion within the genome. The plot shows every transposon insertion identified. Highly depleted or enriched TA sites are highlighted with genes of interest noted. D) Venn diagram showing genes enriched for (left) and depleted (right) shared by all three treatments. All assays were performed in biological triplicate with mean displayed and error bars indicative of SD. Tn-seq data are available in Supplementary Dataset 2, and primary sequencing data are available at NCBI SRA under BioProject PRJNA1104292.

Fig. 5.

MZA efficacy against PZA-susceptible and PZA-resistant M. tuberculosis strains in resting and activated macrophages. A) CFU comparison between RAW 264.7 macrophages infected with H37Rv and treated with either MZA, PZA, or DMSO. B) H37Rv-infected macrophages were activated using IFN-γ. C) Same experimental design as in (A) but using the PZA-resistant H37Rv ΔpncA strain to test efficacy of MZA against PZA-resistant strains. D) RAW 264.7 macrophages activated with IFN-γ and infected with PZA-resistant H37Rv ΔpncA. Cells were treated with 0.9 mM MZA or PZA with daily media exchange. All assays were performed in biological triplicate with mean displayed and error bars indicative of SD. Significance was determined via Kruskal–Wallis test combined with post hoc Dunn test with P < 0.01 (_**), P < 0.001 (_***).

Fig. 6.

Dual action mechanism of MZA. MZA releases aldehyde and PZA in M. tuberculosis. PZA is converted to POA which interferes with CoA metabolism. Aldehyde release drives thiol stress resulting, in part, in synergistic disruption of CoA metabolism with PZA and rapid bacterial cell death.