Uncovering the roles of Mycobacterium tuberculosis melH in redox and bioenergetic homeostasis: implications for antitubercular therapy

ABSTRACT

Mycobacterium tuberculosis (Mtb), the pathogenic bacterium that causes tuberculosis, has evolved sophisticated defense mechanisms to counteract the cytotoxicity of reactive oxygen species (ROS) generated within host macrophages during infection. The melH gene in Mtb and Mycobacterium marinum (Mm) plays a crucial role in defense mechanisms against ROS generated during infection. We demonstrate that melH encodes an epoxide hydrolase and contributes to ROS detoxification. Deletion of melH in Mm resulted in a mutant with increased sensitivity to oxidative stress, increased accumulation of aldehyde species, and decreased production of mycothiol and ergothioneine. This heightened vulnerability is attributed to the increased expression of whiB3, a universal stress sensor. The absence of melH also resulted in reduced intracellular levels of NAD⁺, NADH, and ATP. Bacterial growth was impaired, even in the absence of external stressors, and the impairment was carbon source dependent. Initial MelH substrate specificity studies demonstrate a preference for epoxides with a single aromatic substituent. Taken together, these results highlight the role of melH in mycobacterial bioenergetic metabolism and provide new insights into the complex interplay between redox homeostasis and generation of reactive aldehyde species in mycobacteria.

IMPORTANCE

This study unveils the pivotal role played by the melH gene in Mycobacterium tuberculosis and in Mycobacterium marinum in combatting the detrimental impact of oxidative conditions during infection. This investigation revealed notable alterations in the level of cytokinin-associated aldehyde, para-hydroxybenzaldehyde, as well as the redox buffer ergothioneine, upon deletion of melH. Moreover, changes in crucial cofactors responsible for electron transfer highlighted melH’s crucial function in maintaining a delicate equilibrium of redox and bioenergetic processes. MelH prefers epoxide small substrates with a phenyl substituted substrate. These findings collectively emphasize the potential of melH as an attractive target for the development of novel antitubercular therapies that sensitize mycobacteria to host stress, offering new avenues for combating tuberculosis.

INTRODUCTION

Mycobacterium tuberculosis (Mtb) causes tuberculosis (TB), a significant global health threat that, in 2022, was the second leading cause of death from an infectious agent after coronavirus disease 2019 (1). Within macrophages, the Mtb pathogen encounters formidable host cell defense responses, including nutrient limitation and redox stress. Despite these challenges, Mtb has developed mechanisms to evade macrophage antibacterial responses, adjusting its redox homeostasis and metabolic processes (2). The pathogen modulates its nutritional behavior and metabolic fluxes in response to different carbon sources during infection and growth (3). However, the precise role of Mtb’s metabolic flexibility in maintaining redox homeostasis remains unclear. Therefore, gaining a comprehensive understanding of the adaptive mechanisms employed by Mtb to survive in the human host holds the potential to significantly contribute to the identification of innovative therapeutic strategies.

In the Mtb genome, at least eight potential epoxide hydrolases (EHs) are identified, characterized by the presence of the αβ hydrolase domain (4). These EHs play a crucial role in converting epoxides to trans-dihydrodiols, believed to be essential for detoxification reactions necessary to withstand the hostile environment within host macrophages (5). Some EHs have been identified as potential therapeutic targets due to their involvement in detoxification processes (6–8). Previous evidence suggests that the mel2 locus significantly contributes to oxidative stress resistance in Mtb-infected macrophages, although biochemical data regarding its function remain unclear (7). In this study, our focus was on characterizing epoxide hydrolase B (EphB or MelH) encoded by the melH (Rv1938) gene within the mel2 locus. We successfully produced soluble MelH and demonstrated its epoxide hydrolase activity in vitro using a synthetic fluorescent substrate (PHOME) and screened additional epoxide substrate candidates to assess substrate specificity.

Several studies have unveiled Mtb’s sophisticated mechanisms for continuously monitoring and orchestrating appropriate responses against host-generated stresses. Mtb activates various transcriptional regulators in response to adverse conditions, exemplified by the control of various regulators to counteract oxidative stress. One such regulator is whiB3, a redox-sensing transcription factor encoding a 4Fe-4S redox sensor that is sensitive to reactive oxygen species (ROS) (9). WhiB3 plays a pivotal role in maintaining intracellular redox homeostasis, ensuring both metabolic and cellular integrity (10). Mce3R, a TetR-type transcriptional repressor, controls the expression of mel2 genes, including melH (11). A recent report identified an interrelationship between cholesterol, pH, and potassium levels dependent on Mce3R, emphasizing the crucial role of this regulon in host survival and its importance in responding to environmental stresses within the infected macrophage (12).

Previous studies have demonstrated that whiB3 protects against the acidic pH encountered inside cells by modulating the mycothiol redox system (13). The whiB3 repressor regulates the biosynthesis of both mycothiol (MSH) and ergothioneine (EGT) that serve as major redox buffers against various stressors (14) in mycobacteria. These redox systems contribute to mycobacterial survival strategies within the host (14).

In this study, we combined metabolomic, bioenergetic, biochemical, and transcriptomic approaches to evaluate how melH deletion impacts redox balance and bioenergetic homeostasis, which in turn moderate survival. Our investigation elucidated the critical importance of melH in regulating whiB3, maintaining MSH and EGT levels, and aldehyde accumulation, suggesting a link between all three in protecting against imbalanced redox stresses and in maintaining bioenergetic homeostasis.

RESULTS

melH-encoded protein has epoxide hydrolase activity

melH (Rv1938) encodes a putative EphB (EphB or MelH) with an alpha/beta hydrolase domain and a cap domain (8). Recombinant Mtb MelH was produced and purified (Fig. S1A), and its identity was confirmed by mass spectrometry. We showed that a MelH hydrolyzed the artificial fluorogenic epoxide substrate PHOME (Fig. S1B). The enzyme also demonstrated epoxide hydrolase activity similar to that of the C-terminal domains of a human soluble epoxide hydrolase but lacked phosphatase activity (Fig. S1C and D) (15).

melH affects Mtb and Mm susceptibility to oxidative stress

Previous work has suggested that mel2 locus is involved in the regulation of redox homeostasis (7, 16, 17). Through in silico analysis, it was discovered that among the six genes (melF–melK) in the mel2 locus of Mycobacterium marinum (Mm) and Mtb, melF, melG, and melH closely resemble homologs of bioluminescence-related genes, luxA, luxG, and luxH, respectively (16). The results of the study highlight the potential significance of melF, melG, and melH genes in the resistance of ROS-mediated oxidative stress. Here, we investigated how individual mel2 gene mutations affected Mtb’s response to oxidative stress. Under normoxic conditions, melF:Tn, melG:Tn, and melH:Tn transposon mutants did not show increased ROS levels relative to the wild type (WT) in cultures grown with glycerol as sole carbon source. However, in response to potent oxidants, the melH mutant exhibited a significantly greater increase in ROS levels than the melF and melG mutants and WT Mtb CDC1551 (Fig. 1A). The melH mutant also showed reduced survival rates compared to the WT and the melF and melG mutants under oxidative stress (Fig. 1B). These findings suggest that melH plays a critical role in regulating ROS-mediated oxidative stress in Mtb.

Fig 1.

melH deletion exacerbates effect of and increases susceptibility to oxidative stress. (A) Mean fluorescence intensity of CellROX Green (a ROS-sensitive dye) in WT Mtb (CDC1551) and melF/G/H transposon mutants treated with CHP or H₂O₂. (B) Percentage of survival of WT and mutant Mtb strains determined by measuring bacterial CFU counts after a 30-min treatment with CHP or H₂O₂ in Middlebrook 7H9 medium containing glycerol as a single carbon source. (C) Mean fluorescence intensity of CellROX Green in WT Mtb the ∆melH mutant and the ∆melH complemented strain (melH::mel2) in 7H9 medium containing glycerol as a single carbon source. (D) Percentage of survival of Mtb determined by measuring bacterial CFU counts after a 30-min treatment with CHP or H₂O₂ in 7H9 medium containing glycerol as a single carbon source. (E) Mean fluorescence intensity of CellROX Green in WT Mm, the ∆melH mutant and the ∆melH complemented strain (melH::mel2) in 7H9 medium containing glycerol as a single carbon source. (F) Percentage of survival of Mm determined by measuring bacterial CFU counts after a 30-min treatment with CHP or H₂O₂ in 7H9 medium containing glycerol as a single carbon source. Error bars indicate standard deviations of three replicate experiments. P values were determined by one-way analysis of variance using GraphPad Prism. *P < 0.05, **P < 0.01, ***P < 0.001. CHP, cumene hydroperoxide; ns, not significant; NT, no treatment.

Mm has recently been proven to be an applicable model for studying TB. Most importantly, the complete mel2 locus, exhibiting gene arrangement and transcription orientation identical to the Mtb, has been identified in Mm (Fig. S2A) (16). Using National Center for Biotechnology Information Protein BLAST, we performed a sequence alignment between the amino acid sequences of MelH from Mtb and Mm. The result showed that the two MelH sequences share a high degree of similarity (Fig. S2B). Specificially, there was 87% amino acid identity, and 94% of the aligned positions showed positive matches, suggesting identity of physicochemical properties and function of MelH between Mtb and Mm.

To avoid any polar effects of transposition and to better understand the biochemical role of melH in Mtb’s defensive regulation of oxidative stress, we generated in frame ΔmelH knockout mutants and ΔmelH complemented strains of Mtb and Mm by using specialized transduction in the clinical isolate Mtb CDC1551 and in Mm ymm1, and we validated the transformants by PCR (Fig. S2C through F) and quantitative real-time PCR (Fig. S2G). First, we characterized the phenotypes of the ΔmelH Mtb and Mm mutant and complemented strains (Fig. 1C through F). Compared with ROS levels in WT Mtb, the levels in ΔmelH Mtb were three- to fourfold and two- to threefold higher in response to cumene hydroperoxide (CHP) and H₂O₂ treatment, respectively (Fig. 1C; Fig. S3A). Moreover, the survival rate of ΔmelH was lower than the rates of the WT and complemented strains upon treatment with CHP or H₂O₂ (Fig. 1D; Fig. S3B). We observed an identical trend in Mm (Fig. 1E and F). These results suggest that Mtb and Mm need melH to respond to ROS and to maintain redox balance, and utilize melH identically.

∆melH mutant exhibits altered growth profiles even in the absence of oxidative stress

Mtb utilizes metabolic pathways that include the tricarboxylic acid (TCA) cycle, pentose phosphate pathway, Embden–Meyerhof–Parnas pathway, methyl citrate cycle, and the B12-dependent methylmalonyl pathway. These pathways equip Mtb to exploit a diverse array of carbon sources, including carbohydrates, sugars, fatty acids, amino acids, and sterols (18). While fatty acids from lipid droplets serve as the primary carbon source in vivo, host cells harbor various soluble nutrients that can function as alternative carbon sources (3). Our primary focus lies in comprehending how the melH mutant influences the modulation of metabolite flux within this network under distinct nutritional conditions. To understand the impact of different carbon sources on the metabolism of both WT Mm (ymm1), ΔmelH Mm, and ΔmelH complemented Mm, we cultured the mycobacteria in a carbon-defined minimal medium and examined their growth rates in the presence of diverse carbon sources. The growth curve was plotted on a logarithmic scale (Fig. 2; Fig. S4). The time period during which the strain exhibited logarithmic growth was used to calculate doubling times for graphing and comparison. The slope of the straight line fitted to the linear part of the curve is interpreted as the doubling time. We then compared the doubling times for different carbon sources used as sole carbon sources (Fig. S5). We observed altered growth phenotypes for both Mm and Mtb ΔmelH strains when they were grown with glycerol, propionate, or cholesterol as the sole carbon source (Fig. S4A through C; Fig. S5A and B). Interestingly, the growth rates of the two strains were similar when they were cultured with even-chained or long-chain fatty acids as the sole carbon source (Fig. 2D, E, G and K; Fig. S4D and E). Also, the two strains exhibited similar growth rates when cultured with pyruvate, fumaric acid, succinic acid, or malic acid as the sole carbon source (Fig. 2F and H through J). These results suggest that Mm and Mtb adapt their carbon flux in response to the environment, and the carbon source-dependent growth-rate decrease observed in ΔmelH emphasizes the importance of melH in central carbon metabolism. Given the identical phenotypes of melH in both Mtb and Mm, we performed subsequent experiments in Mm strains to facilitate experimental throughput.

Fig 2.

Effect of melH deletion on Mm growth rate depends on carbon source. Growth rates of WT Mm (black lines), the ∆melH Mm mutant (blue lines), and ΔmelH Mm complemented (melH::mel2) (red lines) in minimal medium containing either (A) glycerol, (B) propionate, (C) cholesterol, (D) oleic acid, (E) acetate, (F) pyruvate, (G) palmitic acid, (H) fumaric acid, (I) succinic acid, (J) malic acid, or (K) butyric acid as a single carbon source. Error bars indicate standard deviations of three replicate experiments.

melH deletion modulates bioenergetic functions of Mm

In order to elucidate the molecular mechanisms underlying the carbon source-dependent growth-rate defect and ROS-mediated oxidative damage observed in ΔmelH Mm, we conducted untargeted metabolomics analyses of the WT, ΔmelH, and ΔmelH complemented strains grown in broth supplemented with glycerol, propionate, cholesterol, pyruvate, or oleic acid as the sole carbon source. The experimental design from the sample collection through data analysis is depicted in Fig. S6A. The metabolites were extracted, separated by high-performance liquid chromatography, and subjected to electrospray ionization time-of-flight mass spectrometry, and the metabolomes were analyzed with MetaboScape. After evaluating heatmaps of differential abundance, we conducted a hierarchical clustering of m/z and retention time pairs, which revealed diverging sample clusters, with three groups mapped adjacent to each other (glycerol, propionate, and cholesterol) and clearly separated from the oleic acid group (Fig. S7A and B). Interestingly, in all groups of metabolomes, the differential abundance of metabolites between the WT and the mutant was minor. Of all the groups, the glycerol group showed the highest differential abundance between the WT and the mutant.

To gain a deeper understanding of the mechanisms linked to melH deficiency in Mm, we first selected the annotated and statistically significant metabolites observed in the tandem mass spectrometric confirmation experiments. Then we considered metabolites with an adjusted P value of <0.05 and a fold change of ≥|2.0| between WT and mutant for each carbon source to be differentially abundant metabolites (Fig. 3A). The annotated and statistically significant metabolites were subjected to metabolic pathway analysis (19) to assess changes in the abundance of metabolites from various biochemical pathways in ΔmelH Mm compared to WT. The results showed the most statistically perturbed metabolic pathways to be nicotinamide metabolism, nucleotide metabolism, TCA cycle, and several amino acid biosynthesis pathways (Table 1; Fig. S8A).

Fig 3.

Liquid chromatography–mass spectrometry analysis of ΔmelH Mm metabolites demonstrates increased pHBA levels and decreased EGT levels. (A) Volcano plots showing fold change versus significance for metabolites extracted from WT and ΔmelH Mm strains cultured with glycerol, propionate, cholesterol, oleic acid, or pyruvate as the sole carbon source. Fold change of metabolite abundance in ΔmelH Mm relative to the abundance in the WT. Red and blue dots indicate metabolites with ≥2-fold increase or decrease, respectively, on the x-axis and a corrected P value of <0.05 (−log¹⁰ >1.3), on the y-axis; gray dots indicate metabolites with a <2-fold change and/or a P value of ≥0.05. (B and C) Chemical structures of (B) pHBA and (C) EGT. (D) Corresponding pHBA levels of WT Mm (n = 5, blue dots), the ∆melH Mm mutant (n = 5, black squares), and ΔmelH complemented (melH::mel2) Mm (n = 5, red triangles). P values were determined by one-way analysis of variance using GraphPad Prism: *P < 0.05, **P < 0.01, ***P < 0.001. (E) Fold change of EGT levels in ΔmelH Mm relative to the level in the WT. (F) Corresponding EGT levels of WT Mm (n = 5, blue dots), the ∆melH Mm mutant (n = 5, black squares), and ΔmelH complemented (melH::mel2) Mm (n = 5, red triangle). P values were determined by one-way analysis of variance using GraphPad Prism: *P < 0.05, **P < 0.01, ***P < 0.001. (G) Fold change of EGT levels in ΔmelH Mm relative to the level in the WT. pHBA, para-hydroxybenzaldehyde

TABLE 1.

MetPA analysis of metabolite pathways altered in △melH compared to WT Mm^a

Kyoto Encyclopedia of Genes and Genomes pathway nameb	Totalc	Hits	P valued	Impact
Pyrimidine metabolism	39	17	2.57E-08	0.511
Purine metabolism	65	22	5.29E-08	0.367
Glyoxylate and dicarboxylate metabolism	32	13	3.40E-06	0.362
Pyruvate metabolism	22	10	1.50E-05	0.485
Alanine, aspartate and glutamate metabolism	28	11	2.98E-05	0.763
Citrate cycle (TCA cycle)	20	9	4.59E-05	0.456
Glycolysis/gluconeogenesis	26	10	8.70E-05	0.39
Arginine biosynthesis	14	7	1.51E-04	0.406
Nicotinate and nicotinamide metabolism	15	7	2.60E-04	0.461
Arginine and proline metabolism	38	11	6.91E-04	0.384
Pentose phosphate pathway	22	8	7.22E-04	0.296
Histidine metabolism	16	6	0.00291	0.344
Glycerolipid metabolism	16	6	0.00291	0.386
Glycine, serine, and threonine metabolism	33	9	0.0034	0.128
Riboflavin metabolism	4	3	0.0035	0.5
Glutathione metabolism	28	8	0.00418	0.086
Glycerophospholipid metabolism	36	9	0.00639	0.532
Butanoate metabolism	15	5	0.01163	0
Beta-alanine metabolism	21	6	0.01298	0.056
Nitrogen metabolism	6	3	0.01508	0
D-Glutamine and D-glutamate metabolism	6	3	0.01508	0.5
Tryptophan metabolism	41	8	0.042441	0.05583

^a Metabolic pathway analysis (MetPA), combining pathway enrichment analysis (P values) and pathway topology analysis (pathway impact) across all five carbon sources tested to identify which specific metabolic pathways are significantly altered.

^b Metabolites identified to have significant abundance differences between WT and ΔmelH Mm were analyzed with MetaboAnalyst v.5.0 (https://www.metaboanalyst.ca/MetaboAnalyst/) to elucidate which pathways were changed upon mutation.

^c The total/maximum importance of each pathway = 1; pathway impact value = cumulative % from the matched metabolic nodes.

^d Pathways with the most significant change (P < 0.05) are presented.

We found lower abundance of NAD metabolite in the ΔmelH mutant (Fig. S8B). NAD acts as a key cofactor in many amino acid metabolism pathways, glycolysis, and the TCA cycle (20–23). Independent validation showed significantly lower intracellular concentrations of NAD⁺ and NADH (~50%) in ΔmelH Mm compared to WT across most carbon sources (Fig. 4A). When cholesterol was the exclusive carbon source, a minimal difference between WT and mutant was exhibited in contrast to other carbon sources. Because the NAD⁺:NADH ratio serves as a critical indicator of cellular redox status, metabolic activity and cell health (24), we next asked whether melH deletion affected the NAD⁺:NADH ratio. Interestingly, the NAD⁺:NADH ratios in the WT and mutant did not differ significantly under any of the carbon source conditions, suggesting that melH deletion did not drive the cells into a more reduced state in the absence of oxidative stress (Fig. 4A).

Fig 4.

Deletion of melH reduces intracellular ATP levels and NAD⁺ and NADH concentrations in Mm. (A) Intracellular NAD⁺ and NADH concentrations in WT, ΔmelH, and ΔmelH complemented (melH::mel2) strains of Mm, as measured via recycling assays, along with calculated NAD⁺/NADH ratios. (B) Bacterial membrane potentials, measured with a BacLight Membrane Potential kit, for WT, ΔmelH, and melH::mel2 Mm cultured with succinic acid, fumaric acid, malic acid, oleic acid, pyruvate, acetate, propionate, glycerol, cholesterol, butyric acid, or palmitic acid as the sole carbon source. (C) Intracellular ATP levels, as measured by means of the luciferase/luciferin system, in WT, ΔmelH, and melH::mel2 Mm cultured with succinic acid, fumaric acid, malic acid, oleic acid, pyruvate, acetate, propionate, glycerol, cholesterol, butyric acid, or palmitic acid as the sole carbon source. P values were determined by one-way analysis of variance using GraphPad Prism. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Error bars indicate standard deviations of three replicate experiments. DiOC₂, 3,3′-diethyloxacarbocyanine iodide.

Because electron transfer reactions are the main function of NAD (21), we next determined whether the lower intracellular NAD⁺ and NADH levels in the mutant might be due to increased proton conductance and resulting membrane depolarization, as seen with classical uncouplers. We used the fluorescent dye 3,3′-diethyloxacarbocyanine iodide (3) to measure the membrane potential of Mm cells. ΔmelH showed no changes in membrane potential compared to WT under the tested carbon source conditions, although, as expected, the membrane uncoupler carbonyl cyanide m-chlorophenyl hydrazone (CCCP) did reduce the membrane potential compared to untreated mycobacteria (Fig. 4B). Given this technique’s limit of detection for membrane changes, we cannot rule out a membrane potential change smaller than that caused by CCCP.

In mycobacteria, ATP synthesis depends primarily on the generation of proton motive force through the electron transport chain. Inhibition of the chain alters oxidative phosphorylation and ATP production (25). Given the lower metabolite abundance of ATP observed in the ΔmelH mutant (Fig. S8C), we directly assessed the bioenergetic status of the ΔmelH Mm mutant by measuring intracellular ATP levels using a luciferase system. Independent validation showed that ΔmelH had significantly lower intracellular ATP levels than WT (Fig. 4C). Overall, the intracellular ATP level in the mutant was three- to fourfold lower than that in WT under most of the carbon source conditions. The difference in ATP level exceeded fourfold when acetate, propionate, or glycerol was the sole carbon source. In contrast, a less than twofold difference was observed when oleic, butyric, or palmitic acid was the sole carbon source, consistent with lack of growth phenotype in even carbon-chain fatty acids. Although the NAD⁺:NADH ratio is not significantly altered, the reduced levels of NAD⁺ and NADH can still impact metabolic flux through redox reactions, potentially leading to decreased ATP levels. In sum, our results demonstrate that melH deletion lowered redox cofactor and energy storage levels in Mm.

Metabolomic studies of ΔmelH Mm show aldehyde accumulation and EGT depletion

Using the established cutoffs (Fig. 3A), we further identified specific metabolites that were differentially abundant in the ΔmelH mutant relative to WT across all tested carbon sources (Fig. S9A and B). We observed increased levels of para-hydroxybenzaldehyde (pHBA) (Fig. 3B, D and E), isonicotinic acid, and quinolinic acid (Fig. S9C), and decreased levels of EGT (Fig. 3C, F and G), L-acetylcarnitine, and tetrahydrofurfuryl butyrate (Fig. S9C) in ΔmelH relative to WT across all media.

The increase in pHBA observed in ΔmelH compared to WT varies, depending on the carbon source, with an approximately twofold increase in cholesterol or pyruvate medium and approximately a fourfold elevation in glycerol or oleic acid medium (Fig. 3D and E). The highest increase was observed in propionate medium, with an approximately sixfold increase in the mutant compared to WT. Additionally, in ΔmelH, we observed significant reductions in the levels of EGT in comparison to WT (Fig. 3F and G). Specifically, EGT levels were decreased by approximately 10-fold when cholesterol was used as sole carbon source. When glycerol served as sole carbon source, EGT levels were reduced by about threefold. Similarly, in the presence of propionate, oleic acid, or pyruvate as sole carbon source, ΔmelH showed a decrease in EGT levels, with an approximately twofold reduction in each case.

We also looked for MSH, a redox buffer distinct from EGT with a different reduction potential (26). We were only able to detect MSH in glycerol and pyruvate carbon sources (Fig. S9D); it was absent in the remainder of our culture conditions. MSH is susceptible to facile oxidation under cell lysis conditions, whereas EGT is resistant to auto-oxidation and predominantly exists in its thione form rather than the thiol form and has exceptional stability (27). These differences in EGT and MSH likely explain our inability to detect or quantify MSH metabolite in our metabolomics data.

Aldehydes accumulate in ∆melH

Darwin and coworkers have reported that adenine-based cytokinins and their degradation products, in particular pHBA, accumulate in an Rv1205 mutant that sensitizes mycobacterial cells to NO (28–31). Given the accumulation of pHBA in ΔmelH Mm, we examined the overall intracellular aldehyde levels in ΔmelH Mm. Because aldehydes are inherently reactive and exist in different hydration states, their detection by mass spectrometry can be challenging. Therefore, we used a fluorgenic substrate that reacts with a broad range of aldehydes (32, 33). We measured intracellular aldehyde concentrations in Mm grown in glycerol, propionate, cholesterol, oleic acid, or pyruvate as the sole carbon source (Fig. 5A). To ensure our assay was specific for aldehydes and not other reactive electrophiles, we tested glycidyl ether, an epoxide, as a control, and confirmed that epoxide does not react with the fluorogenic reagent.

Fig 5.

Disruption of melH causes intracellular aldehyde accumulation. (A) Workflow of aldehyde measurement. (B) Aldehyde concentrations in WT, ΔmelH, and ΔmelH complemented (melH::mel2) strains of Mm cultured with glycerol, propionate, cholesterol, oleic acid, or pyruvate as the sole carbon source. Aldehyde concentrations were determined with a fluorometric assay kit. After the mycobacteria were incubated with the aldehyde detection reagent for 30 min, the absorbance of the supernatants at 405 nm was measured using a plate reader. (C) Aldehyde concentrations in WT and ΔmelH Mm lysates incubated with MelH recombinant protein or denatured MelH recombinant protein for 1 h. (D) Aldehyde concentrations in enzyme-denatured WT and ΔmelH Mm lysates incubated with MelH recombinant protein or denatured MelH recombinant protein for 1 h. P values were determined by one-way analysis of variance using GraphPad Prism. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. Error bars indicate mean ± standard deviation (n = 3 independent experiments).

ΔmelH showed significantly higher aldehyde concentrations when grown in glycerol, propionate, or cholesterol compared to the WT Mm and the ΔmelH complemented strains. There were slightly higher, statistically insignificant concentrations of aldehyde in the mutant when grown in pyruvate, and no difference was observed when Mm was grown in oleic acid (Fig. 5B). These results suggest that melH mutation alters aldehyde metabolism.

To explore further the role of MelH with respect to intracellular aldehyde concentrations, we determined whether exogenous recombinant MelH protein could prevent aldehyde generation in WT or ΔmelH Mm cultured with propionate as the sole carbon source. We found that exogenous recombinant MelH reduced aldehyde concentrations in cell lysates of both strains, whereas heat-denatured recombinant MelH had no effect (Fig. 5C).

In a subsequent experiment, we sought to investigate whether the accumulated aldehydes observed in the ΔmelH Mm mutant were due to aldehyde-containing compounds serving as direct substrates for MelH or if MelH might function to eliminate an aldehyde precursor present in cell lysates. To address this question, we heat denatured cell lysates and then added purified MelH protein to these cell lysates. If the accumulated aldehyde levels were solely due to the absence of catalytically active MelH that formed aldehyde-containing compounds as its enzymatic product, then the addition of active MelH to the denatured lysate should have reduced the aldehyde concentrations. However, we observed that the aldehyde levels remained unchanged after the addition of MelH to denatured cell lysates (Fig. 5D). This result strongly suggested that MelH’s direct substrate is not an aldehyde. Rather aldehydes are formed enzymatically from an unknown catalytic activity present in cell lysate, and this enzymatic pathway utilizes as its substrate a metabolite that is depleted by addition of a MelH epoxide hydrolyase catalyst to the cell lysate. This metabolite may be an epoxide or a precursor metabolite. In the absence of functional MelH in the ΔmelH Mm mutant, these epoxides or precursors accumulate, and upon cell lysis, the aldehydes continue to be formed unless MelH is exogenously added to remove the epoxide/precursor. These results suggest a possible connection between MelH epoxide hydrolase activity and detoxification of intracellular aldehyde accumulation.

Aldehyde accumulation minimally sensitizes ∆melH to NO

In addition to enduring the presence of ROS, Mtb has developed the capability to resist host-generated antimicrobial molecules, including nitric oxide (NO). The precise mechanism of NO-mediated toxicity to Mtb remains unknown, but Mtb relies on a Pup–proteasome system (PPS) to withstand NO (34). Recent research has uncovered that pHBA accumulates significantly in a PPS mutant. Notably, pHBA is sufficient to render Mtb sensitive to NO (29).

We next tested if aldehyde accumulation due to melH deletion correlated with increased susceptibility of the ΔmelH Mm mutant to NO. Mm cells were exposed to acidified nitrite, a source of NO (35, 36). We observed that the nitrosative stress-induced changes in nitrite concentration between the WT and the ΔmelH strain were slightly higher in the mutant; however, the differences were not statistically significant (Fig. S10A). We also determined whether melH deletion reduced mycobacterial survival in Mm compared to WT upon induction of RNS. Acidified NaNO₂ caused cell death at 30 min post-exposure as compared to unstressed bacteria, but the ΔmelH survival rate was only slightly lower than the WT survival rate (Fig. S10B). Complementation eliminated the slight reduction in ΔmelH survival in response to acidified NaNO₂. We also tested the effects of S-nitroso-N-acetylpenicillamine (SNAP) on the survival of the three strains and found that ΔmelH was insignificantly more sensitive to SNAP than WT or the complemented strain. These results demonstrate that melH is not likely to be a major contributor to NO resistance in Mm despite the accumulation of pHBA.

MelH prefers epoxide substrates with a single aromatic substitutent

We investigated the epoxide hydrolase activity of MelH on a set of potential epoxide substrates building on the previous work of James and coworkers (8). They found through activity assays and a liganded crystal structure that MelH has a preference for substrates with phenyl moieties. Consistent with their reports, UniProt bioinformatics similarity searches suggested the closest homologies were with enzymes that utilize aromatic substrates. Therefore, we selected the following five commercially available substrates for our investigation: styrene oxide (Fig. S11 and S12), 1,4-naphthoquinone 2,3-epoxide (Fig. S13), (E)-1,3-diphenyl-2,3-epoxypropan-1-one (Fig. S14), 2-biphenylyl glycidyl ether (Fig. S15), and vitamin K₁ 2,3-epoxide (Fig. S16). We tested each epoxide as an inhibitor and as a substrate of MelH. Inhibition was analyzed as percent reduction of turnover of the fluorescence substrate epoxy fluor 7 in the presence of a fixed concentration of epoxide. Substrate turnover was analyzed by analytical thin-layer chromatography (TLC) and ¹H-nuclear magnetic resonance (NMR) spectroscopy (Table 2). Styrene oxide (Fig. S11) and vitamin K₁ 2,3-epoxide are completely converted to the corresponding diol. 1,4-Naphthoquinone 2,3-epoxide was only partially hydrolyzed after 1 h of incubation with the enzyme, and the remaining two substrates were not hydrolyzed. Active substrates were poor inhibitors, presumably because they were rapidly converted to their corresponding product diols which do not bind as tightly to the enzyme. Interestingly, the three non-reactive or low-reactivity substrates, (E)-1,3-diphenyl-2,3-epoxypropan-1-one, 2-biphenylyl glycidyl ether, and 1,4-naphthoquinone 2,3-epoxide inhibited MelH activity effectively (Table 2), suggesting that they bind non-productively to MelH due to their aromatic nature.

TABLE 2.

Evaluation of MelH potential epoxide substrates

Epoxide tested	Structure	Activity observed by TLC and NMRa	Percent inhibition (%) observed at 25 μMb
Vitamin K1 2,3-epoxide		+++	25
2-Biphenylyl glycidyl ether		–	94
Styrene oxide		+++	64
1,4-Naphthoquinone 2,3-epoxide		+	99
(E)-1,3-Diphenyl-2,3-epoxypropan-1-one		–	98

^a Each substrate (50 μM) was incubated with 150 ng of MelH at pH 7.5, 37°C for 1 h. Reactions were quenched, and the extracted organic layer was analyzed by thin-layer chromatography and ¹H-NMR spectroscopy. +++ reaction, 100% complete; ++ reaction, 70% complete; + reaction, 30% complete in 1 h; – reaction, 0% complete.

^b The initial velocity of epoxy fluor substrate turnover was measured in the presence of 25 μM of the indicated epoxide at pH 7.5, at 37°C for 10 min, and the rate of turnover compared to the negative control without addition of a second epoxide.

whiB3 regulates MSH and EGT production in response to oxidative stress

Mtb whiB3, an intracellular redox sensor, regulates EGT production in a carbon source-dependent manner (14). EGT and MSH are the major redox buffer present in Mtb. Steyn and coworkers established that EGT and MSH are both required for maintaining redox balance and bioenergetic homeostasis, which influence drug susceptibility and pathogenicity (14). Our metabolomics analysis revealed that melH significantly decreased EGT and MSH levels (Fig. 3F; Fig. S9D). Therefore, we compared the transcript levels for MSH and EGT biosynthesis genes MMAR_5212 (egtD), MMAR_5215 (egtA), and MMAR_0812 (mshA), as well as whiB3 (MMAR_1132) in the ΔmelH Mm mutant relative to WT (37, 38) using quantitative real-time PCR.

Regardless of carbon source, deletion of melH resulted in an approximately twofold upregulation of whiB3 transcription compared with that in WT under normoxic conditions (Fig. 6A). When treated with oxidative stressors, expression of whiB3 was upregulated 6- to 10-fold in ΔmelH compared with that in the WT in glycerol, propionate, cholesterol, pyruvate, or oleic acid culture conditions.

Fig 6.

melH deletion affects expression of whiB3 and EGT and MSH biosynthesis genes (egtA, egtD, and mshA) in Mm in response to oxidative stress. Comparison of (A) whiB3, (B) mshA, (C) egtA, and (D) egtD expressions in Mm melH mutant relative to Mm WT under conditions of NT, CHP, or H₂O₂. Total RNA was extracted from the bacteria, and intracellular gene expression was measured by quantitative real-time PCR. P values were determined by one-way analysis of variance using GraphPad Prism. *P < 0.05, **P < 0.01. Error bars indicate standard deviations of three replicate experiments. NT, no treatment.

In contrast, exposure of Mm to CHP or H₂O₂ significantly reduced expression of mshA, egtA, and egtD in the mutant compared to that in WT (Fig. 6B through D, respectively). Notably, when glycerol or propionate was the sole carbon source, the reduction in egtA expression in the CHP- or H₂O₂-treated mutant was more pronounced than the reduction observed under the other carbon source conditions. We observed significantly lower mshA expression in the ΔmelH mutant than in the WT, particularly under oxidative stress conditions. Taken together, these results establish that melH deletion results in induction of whiB3 expression and suggests that increased whiB3 expression downregulates the expression of EGT and MSH biosynthesis genes.

ΔmelH Mm failure to maintain intracellular redox homeostasis is independent of carbon source

We examined the effect of melH deletion on redox homeostasis in Mm grown with different carbon sources and found that melH deletion resulted in increased ROS accumulation in response to oxidative stress regardless of carbon source. Specifically, ΔmelH exhibited an approximately 1.5- to 3.0-fold increase in ROS production in response to CHP or H₂O₂ treatment across all carbon sources. Interestingly, the results do not show a significant change in the ROS production in ΔmelH compared to the WT and the complemented strains under non-stressing growth conditions (Fig. 7A).

Fig 7.

Absence of melH in Mm results in heightened vulnerability to oxidative stress in response to carbon catabolism. (A) Mean fluorescence intensity of CellROX Green (a ROS-sensitive dye) CHP- or H₂O₂-treated WT, ∆melH, and ∆melH complemented (melH::mel2) strains of Mm cultured in 7H9 medium with glycerol, propionate, cholesterol, oleic acid, or pyruvate as the sole carbon source. (B) Percentage of survival of WT, ∆melH, and melH::mel2 strains of Mm determined by measuring bacterial CFU counts after a 30-min treatment with CHP or H₂O₂ in 7H9 medium containing glycerol, propionate, cholesterol, oleic acid, or pyruvate as a single carbon source. P values were determined by one-way analysis of variance using GraphPad Prism. *P < 0.05, **P < 0.01. Error bars indicate standard deviations of three replicate experiments. ns, not significant.

To determine the impact of melH-mediated ROS toxicity on cellular survival, we measured CFUs after treatment of Mm cells with CHP or H₂O₂. We found that the survival rate of ΔmelH Mm was significantly lower than the WT survival rate and that complementation of the mutant with the mel2 operon restored intracellular survival to near WT levels across all carbon source growth conditions (Fig. 7B). Altogether, the high expression of whiB3 exhibited under CHP- or H₂O₂-induced oxidative stress conditions in ΔmelH is consistent with the increased ROS levels and the decreased survival of the ΔmelH mutant. The heightened accumulation of ROS in the mutant strain has a detrimental effect on its survival rate when exposed to an external oxidative stress.

DISCUSSION

In this study, our investigations unveiled that melH deletion disrupts coordination between ROS detoxification and thiol homeostasis. Our collective observations suggest an intricate narrative: melH deletion leads to heightened epoxide and aldehyde levels and accumulation of purine and quinolinic acid metabolites. Subsequently, increased levels of WhiB3 sensitize mycobacteria to oxidative stress through reduction in cellular levels of thiol buffers, EGT and MSH. Inability to clear ROS further disrupts the delicate balance of cellular redox homeostasis, overwhelming the antioxidant defense mechanisms of the mutant (Fig. 8).

Fig 8.

Model for metabolic consequences of melH deficiency. In Mtb and Mm, melH encodes MelH (epoxide hydrolase b), which catalyzes hydrolysis of epoxides and aids mycobacterial survival within host macrophages. In the absence of functional MelH in the ΔmelH mutant, there is an accumulation of epoxides or their precursors, leading to the buildup of aldehydes. The deletion of melH results in elevated expression of whiB3 and decreased production of the redox buffer EGT. In addition, deletion of melH causes a decrease in intracellular levels of NAD⁺, NADH, and ATP. All of these changes affect bioenergetic homeostasis and bacterial growth and ultimately sensitize the mutant to oxidative stress.

The natural substrate for MelH remains to be identified. The structure of Mtb MelH has been solved, revealing a relatively small and hydrophobic active site in a classic αβ hydrolase fold and substrate selectivity is altered compared to mammalian epoxide hydrolases (6, 8). Our metabolomics analysis did not reveal an accumulated epoxide substrate; therefore, we investigated several epoxides as possible substrates with varying aromatic moieties and found that MelH efficiently catalyzes hydrolysis of styrene oxide and vitamin K₁ 2,3-epoxide into the corresponding diol and does not catalyze hydrolysis of epoxides containing a biphenyl or two aromatic substituents (Table 2). Considering all the data in combination, the natural substrates of MelH are most likely aromatic species generated through metabolite breakdown and can be hindered epoxides like vitamin K.

It is important to note that both epoxides and aldehydes can be volatile, are reactive electrophiles, and exist in hydrated forms in the case of aldehydes. These factors may explain why we did not observe directly the accumulation of epoxides or aldehyde species other than pHBA by mass spectrometry. We did observe related metabolites, for example, an increased abundance of quinolinic acid (QA) in ΔmelH Mm (Fig. S9C). QA is a product of 2-amino-3-carboxymuconate-semialdehyde, an unstable compound that can be non-enzymatically transformed to QA (39). Hence, it is possible that 2-amino-3-carboxymuconate-semialdehyde is also one of the accumulated aldehydes in ΔmelH Mm.

In addition, we found an accumulation of other metabolites in the cytokinin pathway, specifically N-isopentenyladenine and adenine in ΔmelH Mm. This suggests that melH might be involved in cytokinin metabolism in Mtb. While cytokinins are well-established adenine-based signaling molecules in plants, their function in Mtb remains unknown (40).

Darwin and coworkers investigated the effects of accumulation of pHBA, a major product of cytokinin breakdown, on gene expression in mycobacteria (29–31, 40). In their RNA-Seq comparisons of WT Mtb treated with pHBA versus untreated WT, pHBA exhibited no influence on the expression of whiB3, egtA, egtD, and mshA genes. Thus, formation of pHBA is not directly responsible for whiB3 upregulation. The inducer of whiB3 upregulation may be an unidentified epoxide or another accumulating species that acts through one of the many known regulators of whiB3, e.g., PhoPR, RegX3, or GlnR (41, 42).

Additional contributors to the melH oxidative stress phenotype may be cross-talk with sigma factors (SigH and SigE), two-component systems (SenX-RegX and DosR/S/T), or serine-threonine kinases (PknG), all of which are reported to be key mediators of the oxidative stress response in Mtb (43). Another possibility is that the accumulation of QA and isonicotinic acid contributes to the growth and oxidative stress-resistant phenotype we observed in the ΔmelH mutant. Previous studies have shown that QA can form complexes with Fe (II), inducing ROS formation, especially hydroxyl radical (•OH), which, in turn, is responsible for DNA chain breakdown and lipid peroxidation (44).

Furthermore, QA is a product of tryptophan degradation, which is crucial to produce NAD⁺ (45). Thus there is a potential connection between melH deletion and perturbations in tryptophan metabolism. The perturbations in tryptophan metabolites correlate with the reduced levels of NAD⁺ and NADH observed upon melH deletion. Reduction of NAD⁺/NADH levels can still impact metabolic flux through redox reactions, despite maintenance of the NAD⁺:NADH ratio. A recent study by the Schnappinger lab determined that bacteriostatic levels of NAD⁺ depletion cause compensatory remodeling of NAD-dependent metabolic pathways without impacting NADH:NAD⁺ ratios (23). On the other hand, bactericidal levels of NAD⁺ depletion led to a disruption of NADH:NAD⁺ ratios and inhibition of oxygen respiration. All of which may lead to decreased ATP levels.

Our CFU results indicated that the changes in NAD levels occuring upon melH deletion are not sufficient to elicit alterations in bacterial survival. However, under oxidative stress conditions, these changes are adequate to induce shifts in bacterial physiology, resulting in reduced survival. These data point toward a role for melH deletion causing NAD(H) depletion, further contributing to an inability to maintain redox homeostasis. Mycobacterium smegmatis ΔatpD exhibits reduced ATP levels, accompanied by elevated levels of ROS and significant alterations in NAD levels, indicating an imbalanced redox state within the cell (46). This study demonstrated a threefold upregulation of whiB3 in M. smegmatis ΔatpD compared to the WT. WhiB3 serves as a crucial redox regulator, playing a pivotal role in dissipating reductive stress through the biosynthesis of lipids, including PAT/DAT, SL-1, and PDIM, as well as TAG (9).

In future studies, it would be intriguing to investigate the underlying mechanisms responsible for triggering whiB3 in the melH mutant. This could be accomplished by generating a melH and whiB3 double knockout strain and examining its phenotype through untargeted metabolomic and lipidomic analyses. This approach will provide valuable insights into the intricate interplay between melH and whiB3, shedding light on their cooperative roles in maintaining cellular redox homeostasis.

In summary, the findings of this study provide insight into the role of melH and its potential interactions with other pathways in mycobacterial metabolism and highlight the importance of further investigations of MelH function for understanding mycobacterial intracellular survival.

MATERIALS AND METHODS

Bacterial strains and culture conditions

The melF:Tn (NR-13643), melG:Tn (NR-13645), and melH:Tn (NR-18002) Mtb CDC1551 strains were obtained from BEI Resources (www.beiresources.org), and the ymm1 Mm strain was kindly provided by Dr. Jeffrey D. Cirillo. Mtb strains were grown at 37°C, while Mm strains were grown at 30°C, in Middlebrook 7H9 medium (Difco) supplemented with 0.2% glycerol or 0.1-mM cholesterol, 0.5% bovine serum albumin, 0.08% NaCl, and 0.05% (vol/vol) tyloxapol (Sigma-Aldrich) and were filter-sterilized. Alternatively, cultures were grown on Middlebrook 7H11 agar supplemented with 10% oleate-albumin-dextrose-NaCl (OADC) and 0.5% glycerol. If necessary, cultures were supplemented with sodium acetate, succinic acid, fumaric acid, butyric acid, malic acid, or palmitic acid to a final concentration of 50 µM as sole carbon source. The melF:Tn, melG:Tn, and melH:Tn CDC1551 strains were grown in the presence of 30-µg/mL kanamycin, whereas the ΔmelH Mtb CDC1551 and ΔmelH Mm ymm1 strains were grown in the presence of 50-µg/mL hygromycin B. All complemented strains were grown in the presence of 50-µg/mL hygromycin and 50-µg/mL kanamycin. For determination of the effect of the carbon source on melH function, Mtb and Mm were cultured in Middlebrook 7H9 medium supplemented with 10% OADC, 0.05% tyloxapol, and one of the following carbon sources: glycerol, sodium propionate, sodium pyruvate, cholesterol, oleic acid, sodium acetate, succinic acid, fumaric acid, butyric acid, malic acid, or palmitic acid to a final concentration of 50 µM.

Cloning of ΔmelH and ΔmelH complemented strains of Mtb

ΔmelH Mtb was generated through recombinase-mediated recombination of a hygromycin-resistance cassette PCR product flanked by sequences homologous to the ~545-bp regions upstream and downstream of the melH (Rv1938) gene. Mtb CDC1551 cells were transformed with an episomal plasmid containing the RecET recombinase (pNit-recET-sacB-kanR) and were induced to express the recombinase by addition of isovaleronitrile (final concentration, 1 µM) for 24 h followed by addition of glycine (final concentration, 0.2 M) for 16 h. The electrocompetent cells prepared from the induced culture were transformed with 1 µg of PCR product and were recovered for 24 h in 7H9 medium supplemented with 1-mM L-arginine. The transformants were selected on Middlebrook 7H10 agar containing 0.5% glycerol, 10% OADC, 50-µg/mL hygromycin, and 1-mM L-arginine and were validated by PCR. The validated strains were streaked on 7H11 agar supplemented with 0.5% glycerol, 10% OADC, 1-mM L-arginine, and 8.5% sucrose to select for clones that had lost the pNit-recET-sacB-kanR plasmid. The sucrose-resistant colonies were patched first onto 7H11 agar supplemented with 0.5% glycerol, 10% OADC, 1-mM L-arginine, and 30-µg/mL kanamycin and then onto the same agar without kanamycin. The kanamycin-sensitive strains were grown in 7H9 medium supplemented with 50-µg/mL hygromycin and 1-mM L-arginine to obtain the final ΔmelH Mtb strain. For complementation of ΔmelH strain, the target gene with a native promoter corresponding to 200 bp upstream of the first gene in the corresponding putative operon was cloned into a pMV306 integrating plasmid. The resulting construct was electroporated into the ΔmelH mutant, and kanamycin-resistant transformants were selected.

Cloning of ΔmelH and ΔmelH complemented strains in Mm

The predicted melH (MMAR_2866) gene and its flanking genomic regions, as annotated in the MycoBrowser portal, were amplified by PCR using Mm genomic DNA. The resulting upstream and downstream amplicons were fused by PCR with the primers listed in Table S1, and the fused PCR product was purified and digested with the HindIII restriction enzyme (New England BioLabs, Ipswich, MA, USA) according to the manufacturer’s instructions. The HindIII-linearized p1NIL plasmid (Addgene plasmid number 20187) was ligated with the HindIII-digested PCR product, and the ligation mixture was introduced into Escherichia coli DH5α. Colonies bearing the p1NIL plasmid with the melH flanking regions from the Mm genome were selected on Lysogeny Broth (LB).

Central metabolite extraction

Mycobacteria were grown to an OD₆₀₀ of 1.0. Approximately 1 × 10⁸ cells were collected by filtration on 0.22-µm nylon/polyvinylidene fluoride filters (Millipore GVWP02500) and transferred to 7H11 supplemented with 0.4% glycerol, sodium propionate, cholesterol, sodium pyruvate, or oleic acid as the sole carbon source. Plates were incubated for 6 days at 30°C or 37°C for bacterial replication and biomass production. The polar metabolites were extracted in prechilled (−40°C) 2:2:1 methanol/acetonitrile/water, and cells were lysed six times by bead-beating for 30s with incubation on ice for 30s between pulses. Soluble extracts were filtered (Spin-X filter tubes) at 5,000 × g for 5 min at 4°C and then stored at −80°C for liquid chromatography–mass spectrometry. The bacterial biomass of each sample was determined by measuring the residual protein content in the metabolite extracts.

Metabolome analysis by liquid chromatography–mass spectrometry

The sample extract was fractionated by an Agilent 1290 Infinity II system equipped with (i) a Zorbax Eclipse Plus C18 column (1.8 µm, 2.1 by 50 mm) or (ii) an InfinityLab Poroshell 120 HILIC-Z column (2.7 µM, 2.1 by 100 mm), or (iii) a HiPlex H column (4.6 × 250 mm, Agilent). For (i), reversed-phase separation was achieved using a gradient from solvent A (95% water, 5% acetonitrile, and 0.1% formic acid) to solvent B (acetonitrile and 0.1% formic acid) as follows: 100% A held for 1.2 min, 100% B for 20.8 min, 100% B held for 1 min, 100% A for 0.1 min, and re-equilibration for 3 min (total run time, 26 min/sample). The flow rate was maintained at 200 µL/min for the duration of the run; the column was held at 35°C; and samples were held at 4°C. For (ii) HILIC mode, separation was achieved for positive mode using a gradient from solvent A (10-mM ammonium acetate in water) to solvent B (90% acetonitrile, 10% water, and 10-mM ammonium acetate) as follows: 98% B held for 0.8 min, 75% B for 19.2 min, 0% B for 3 min and held for 1 min, 98% B for 0.1 min, and re-equilibration for 3 min (total run time, 27 min/sample). For negative mode, separation was achieved using a gradient from solvent A (20-mM ammonium acetate in water, pH 9) to solvent B (acetonitrile) as follows: 100% B held for 0.8 min, 70% B for 17.2 min, 40% B for 4 min, 0% B for 1 min and held for 1 min, 100% B for 0.1 min, and re-equilibration for 3 min (total run time, 27 min/sample). The flow rate was maintained at 250 µL/min for the duration of the run; the column was held at 35°C; and samples were held at 4°C. For (iii) HIPLEX mode, isocratic separation was achieved via flushing solvent at 200 µL/min (0.01% formic acid, 20% acetonitrile in water) for 20 min. The column was held at 50°C. The column eluate was infused into a Bruker Impact II QTOF system with an electrospray ion source. Data were collected in both (i) positive and (ii) negative ion mode, with the following settings: (i) capillary voltage of 4,500 V; endplate offset of 500 V; nebulizer gas pressure of 1.8 bar; dry gas flow rate of 8.0 L/min; dry temperature of 220°C; MS spectra acquisition rate of 8 Hz; and m/z range of 50–1,500 Da); (ii) capillary voltage of 4,200 V; endplate offset of 500 V; nebulizer gas pressure of 2.0 bar; dry gas flow rate of 8.0 L/min; dry temperature of 210°C; spectra acquisition rate of 8 Hz; and m/z range of 50–1,500 Da).

Analysis of metabolomics data

The data were analyzed using the Bruker Compass MetaboScape 2022b software (v.9.0.1, Bruker Daltonics). Ion chromatograms were aligned, and high-resolution mass was re-calibrated using sodium formate clusters as reference mass. The metabolites were identified by comparing the accurate m/z (mass tolerance <5 ppm) with libraries including HMDB library, Mtb LipidDB library, and MycoMass library, and the MS/MS spectra with libraries including MetaboBASE Personal Library (Bruker), MSDIAL library, and HMDB library. Then the average peak intensities and the standard deviations from three biological replicates were calculated. Pathway map was plotted using Mtb H37Rv metabolic map diagram (Biocyc.org). MetaboAnalyst (v.5.0) was used to support metabolic pathway analysis (integrating pathway enrichment analysis and pathway topology analysis).

Determination of Mtb and Mm susceptibility to oxidants

WT Mtb (CDC1551); melF:Tn, melG:Tn, melH:Tn, and ΔmelH Mtb; WT ymm1 and ΔmelH Mm; and their respective complemented strains were exposed to CHP or H₂O₂. The strains were grown to 0.6–0.8 OD₆₀₀ in 7H9 medium supplemented with 0.4% glycerol or sodium propionate. For each strain, 2 × 10⁶ cells were transferred into each well of a 96-well plate. H₂O₂ (10 mM) or CHP (5 mM) was added to the wells, and the plates were incubated at 37°C or 30°C for 30 min. Control cells (1 × 10⁶) were not treated with either oxidant. The treated cells were subsequently plated on 7H11 agar plates supplemented with 10% OADC. Colonies were enumerated after 3 weeks of incubation at 37°C or 30°C. The survival of the mycobacterial strains was expressed as percentage of survival relative to mycobacterial strains with no treatment.

Measurement of endogenous ROS in Mtb and Mm

Exponentially growing WT Mtb (CDC1551); melF:Tn, melG:Tn, melH:Tn, and ΔmelH Mtb; WT ymm1 and ΔmelH Mm; and their respective complemented strains cultured in 7H9 medium supplemented with 0.4% glycerol or sodium propionate were treated with CellROX Green (Life Technologies; final concentration, 5 µM) at 37°C or 30°C. The cells were pelleted, and the supernatant was discarded. The cells were washed with 7H9 medium to remove any extracellular CellROX Green. The washed cells were re-suspended in medium and analyzed on a plate reader at excitation (Ex)/emission (Em) wavelengths of 485/520 nm.

Measurement of change in membrane potential

The changes in membrane potential of WT ymm1 Mm, ΔmelH Mm, and the ΔmelH complemented strain were determined using a BacLight Bacterial Membrane Potential Kit (ThermoFisher Scientific) according to the manufacturer’s instructions. In short, 1-mL aliquots of culture at an OD₆₀₀ of 0.8 were treated with 3,3′-diethyloxacarbocyanine iodide (3 mM) and then CCCP (20 mM). Dimethyl sulfoxide-treated and -untreated bacilli were used as controls. Aliquots were incubated for 30 min before analysis on a plate reader.

Measurement of total ATP content

Whole-cell pellet samples were collected from Mm cell cultures; aliquots of the suspensions were removed from the samples; and the sample was mixed with 3 mL of boiling Tris–EDTA (100-mM Tris, 4-mM EDTA, pH 7.75). Then the bacterial cells were lysed for 2 min with glass beads, heated at 100°C for 5 min, and cooled on ice. Cellular debris was removed by centrifugation. Supernatants were collected; an equal volume of luciferase reagent (ATP Bioluminescence Assay Kit HS II; Roche, Mannheim, Germany) was added to each supernatant; and luminescence was measured. ATP reaction mixing and ATP measurement were performed with an ATP Colorimetric/Fluorometric Assay Kit (BioVision Research Products, Milpitas, CA) according to the manufacturer’s protocol.

NADH and NAD⁺ determination

Mm Cells were rapidly harvested (two 2-mL samples) and re-suspended in 0.2-M HCl (for NAD⁺ determination) or 0.2-M NaOH (for NADH determination). The tubes were centrifuged at 12,535 × g for 1 min. The supernatant was removed, and the pellets were suspended in 300 mL of 0.2-M HCl (for NAD⁺ extraction) or 0.2-M NaOH (for NADH extraction). The resulting suspensions were placed in a 50°C water bath for 10 min and then on ice to cool them to 0°C. The extracts were then neutralized with 300 mL of 0.1-M NaOH (for NAD⁺ extraction) or 300 mL of 0.1-M HCl (for NADH extraction) added dropwise while vortexing. Cellular debris was removed by centrifuging at 12,535 × g for 5 min. Supernatants were transferred to new tubes, and intracellular NADH and NAD⁺ concentrations were measured by means of a very sensitive cycling assay (47). The assay was performed with a reagent mixture consisting of equal volumes of 1.0-M bicine buffer (pH 8.0), absolute ethanol, 40-mM EDTA (pH 8.0), and 4.2-mM 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide and twice the volume of 16.6-mM phenazine ethosulfate, which had previously been incubated for 10 min at 30°C. The following volumes were added to 1-mL cuvettes: 50 µL of neutralized extract, 0.3 mL of water, and 0.6 mL of the above-described reagent mixture. The reaction was started by adding 50 µL of yeast alcohol dehydrogenase II [500 or 100U/mL in 0.1-M bicine buffer (pH 8.0)]. The absorbance at 570nm was recorded for 10min at 30°C.

Quantitative real-time PCR

Total RNA was extracted from Mm cells using the TRIzol Reagent (Takara, Japan). cDNA was synthesized using a PrimeScript reverse transcriptase reagent kit (Takara). mRNA expression was examined by quantitative real-time PCR using iTaq Universal SYBR Green Supermix (Bio-Rad) and a LightCycler480 detection system (Roche). The relative expression of indicated genes was analyzed by the 2^−ΔΔCt method. For comparisons between WT and ΔmelH Mm, the induction ratio for each gene was normalized to Mm 16s rRNA expression. The primer sequences used for PCR are listed in Table S1.

Nitrite measurement

The Mm cells were treated with acidified NaNO₂ and SNAP, and 30 min later, the Griess reagent system was used to determine levels of nitric oxide production. NO₂⁻ was determined by using NaNO₂ as a standard. Briefly, 50 µL of each experimental sample was added to wells, and then 50 µL of 1% sulfanilamide in 5% phosphoric acid was added to each experimental sample and to wells containing the dilution series for the nitrite standard reference curve. Then 50 μL of 0.1% aqueous N-1-napthylethylenediamine dihydrochloride was added to each well. Absorbance was measured at 550 nm immediately after addition of the reagents to the samples. The cells were subsequently plated on 7H11 agar supplemented with 10% OADC. Colonies were enumerated after 3 weeks of incubation at 30°C. The survival of mycobacterial strains was expressed as percentage of survival relative to mycobacterial strains with no treatment.

Aldehyde measurement

Aldehyde reaction mixing and aldehyde concentration measurements were performed with a Fluorometric Aldehyde Assay Kit (cat. no. MAK141, MilliporeSigma), which uses a fluorogenic dye that reacts with aldehydes to generate a fluorometric product, according to the manufacturer’s protocol.

MelH substrate assays

One hundred fifty nanograms of MelH recombinant proteins was used in each reaction. The reaction mixture contained 50 µM of vitamin K₁ 2,3-epoxide, 2-biphenylyl glycidyl ether, styrene oxide, 1,4-naphthoquinone 2,3-epoxide, or (E)-1,3-diphenyl-2,3-epoxypropan-1-one as the substrate. The reaction volume was adjusted with 50-mM Tris–HCl, pH 7.5, to 250 µL, and the reactions were incubated at 37°C for 60 min. Reactions were stopped by adding 500 µL of ethyl acetate and vortexing. The samples were briefly centrifuged, and the upper organic phase was transferred to a clean tube, dried under a stream of nitrogen, and re-suspended in 20 µL of CHCl₃:CH₃OH (2:1). Half of the suspension was analyzed by TLC in n-hexane:diethyl ether:formic acid (70:30:2). The remaining samples were dried under a stream of nitrogen, washed with CHCl₃, re-suspended in CHCl₃, and analyzed by ¹H-NMR spectroscopy. The initial velocity of epoxy fluor 7 substrate (Cayman, USA) turnover was measured in the presence of 25 µM of the indicated epoxide at pH 7.5, at 37°C, for 10 min. The rate of turnover was then compared to the negative control without the addition of a second epoxide. Epoxy fluor 7 (50 µM) was incubated with 150 ng of MelH in 50-mM Tris–HCl, pH 7.5. The fluorescence intensity was monitored by a plate reader (Ex/Em: 330/465 nm).

Statistics

Statistical computations were performed with GraphPad Prism (v.9.0). Pairwise comparisons were performed using Student’s t-test for normally distributed data. Multiple comparisons were performed using the one-way analysis of variance module of GraphPad.

DATA AVAILABILITY

Metabolomics data sets from this study are deposited in the Global Natural Products Social Molecular Networking (http://gnps. ucsd.edu) under reference MassIVE #MSV000092681. The data that support the findings of this study are available either within this article and its supplemental information files or upon reasonable request to the corresponding author.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/msphere.00061-24.

Supplemental Information. msphere.00061-24-s0001.pdf.

Supplemental figures and table.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.