The rapid development of antibiotic resistance of pathogenic bacteria requires searching for adjuvants of existing drugs to improve their therapeutic efficacy and targets to create novel antibiotics. Research into the influence of cysteine and its derivatives on the virulence of bacteria and their susceptibility to antibiotics and oxidative stress is one of the actively developing areas [1]. Endogenous H₂S, formed during L-cysteine degradation, reduced the sensitivity of bacteria to antibiotics and oxidative stress, while mutations and inhibitors of H₂S-producing enzymes increased the efficacy of antimicrobial drugs [2, 3, 4]. High intracellular cysteine levels are potentially harmful, since cysteine is capable of autoxidation yielding ROS and can potentiate the Fenton reaction giving rise to toxic hydroxyl radicals by maintaining the Fe²⁺ pool [5, 6]. To explain the interplay between L-cysteine metabolism, H₂S production and oxidative stress, a model has been proposed where 3-mercaptopyruvate sulfotransferase (3MST) protects Escherichia coli against oxidative stress by catalyzing cysteine degradation to form H₂S, which binds free iron required for the Fenton reaction [3]. However, instead of enhancing antibiotic tolerance, exogenous and endogenous H₂S in Mycobacterium tuberculosis and exogenous H₂S in Acinetobacter baumannii sensitized them to antibiotics by triggering a pro-oxidative redox imbalance [7, 8, 9]. Endogenous H₂S in mycobacteria has been shown to be an effector molecule that maintains bioenergetic homeostasis by stimulating respiration, plays a key role in central metabolism, regulates redox homeostasis, and increases the susceptibility of M. tuberculosis to antituberculosis drugs clofazimine and rifampicin through its pro-oxidant function [9]. Host-derived H₂S has also been shown to stimulate M. tuberculosis respiration, primarily through cytochrome bd oxidase, and regulate genes involved in sulfur and copper metabolism, as well as the Dos regulon [8, 10]. The effect of H₂S on antibiotic sensitivity apparently depends on the bacterial species and can be determined by the features of endogenous H₂S production or the nature of its effect on metabolic pathways. We have previously shown that antibiotic exposure of E. coli and Bacillus subtilis is accompanied by changes in low molecular weight thiol levels and H₂S production [11, 12, 13, 14]. It still remains unknown whether mycobacteria are capable of producing H₂S in response to antibiotics.

Addition of cystine to E. coli growing in minimal medium with sulfate has been shown to cause cytoplasmic cysteine overload and stimulate cysteine export, its incorporation into glutathione (GSH) and degradation to form H₂S [6, 13, 15, 16, 17]. Multiple bacteria possess cysteine-inducible cysteine-specific efflux pumps for its export [18, 19, 20]. However, no such transporters have been detected in M. tuberculosis and the ability of mycobacteria to export cysteine into the medium remains unknown, although recently the involvement of the Rv0191 efflux pump in the export of low mlecular weight (LMW) thiols has been suggested [21]. L-cysteine supplementation has been shown to stimulate H₂S production in Bacillus anthracis, Pseudomonas aeruginosa, and Staphylococcus aureus [2, 3]. Various E. coli enzymes are known to possess in vitro cysteine desulfhydrase activity (TnaA, CysK, CysM, MalY, MetC, 3MST, IscS, CyuA); some of them may specifically cope with cysteine toxicity [2, 22, 23, 24, 25]. Addition of cysteine to M. tuberculosis was also found to increase H₂S production in a cysteine desulfhydrase (Cds1)-dependent manner, suggesting that this pathway could be used to get rid of excess cysteine [9].

Recently, in a series of studies, we have shown that a transient increase in intracellular cysteine concentration and activation of cysteine homeostasis mechanisms occur not only when cysteine/cystine is added to the growth medium, but also as a consequence of abrupt protein synthesis inhibition during various stresses in E. coli and B. subtilis [11, 12, 26, 27]. These stresses involve nutrient starvation (depletion of glucose, phosphate, or nitrogen) and exposure to antibiotics such as chloramphenicol, tetracycline, and ciprofloxacin. Excess cysteine formed in E. coli under these conditions was mainly incorporated into glutathione, and was also exported into the medium and partially degraded to form H₂S. Sulfide release from cells into the medium under stress can be recorded using selective sulfide or platinum electrodes as abrupt changes in the redox potential (Eh jumps). In our earlier studies, Eh jumps suppressed by the thiol reagent N-ethylmaleimide (NEM) occurred upon exposure to various stressors (starvation, elevated temperature, ultraviolet irradiation, treatment with acetate and antibiotics, etc.) on Gram-negative and Gram-positive bacteria, indicating the universal nature of the observed phenomenon [28].

Like other mycobacteria, Mycobacterium smegmatis contains no glutathione and uses mycothiol as its major LMW thiol, whose concentration in cells is comparable to that of glutathione in E. coli [29, 30, 31]. Mycothiol in mycobacteria performs many of the functions of glutathione in Gram-negative bacteria, including the maintenance of the intracellular redox balance, as well as protection against reactive oxygen and nitrogen species, alkylating agents, and antibiotics. Mycothiol (MSH) levels may be influenced by environmental conditions, such as oxidative or osmotic stress or nutrient starvation [30]. In response to the increased need for antioxidant molecules under hostile conditions such as oxidative stress, M. tuberculosis actively upregulates CysM-dependent L-cysteine biosynthesis through the transcription factors AosR and SigH, turn enhancing mycothiol and ergothioneine production [21]. The expression of MSH biosynthetic genes was also elevated by acidic conditions [32]. It was shown that MSH turnover occurs in M. smegmatis and that MSH can be a cysteine source [30]. Since MSH is autoxidized much more slowly than cysteine, it has been suggested that cysteine incorporation into MSH is a way to conserve cysteine and limit its intracellular concentration. However, no literature data on how the emergence of excess cysteine affects MSH levels are available.

In this study, we examined how cystine supplementation and treatment with chloramphenicol or ciprofloxacin affect changes in cysteine, mycothiol, and H₂S levels in M. smegmatis. We also monitored the effects of these treatments on bacterial growth and respiration, which may be important for explaining the effects of H₂S on antibiotic tolerance. M. smegmatis is frequently used as a model organism to study M. tuberculosis due to its genetic similarity, conserved metabolic pathways (including biosynthesis and turnover of LMW thiols), and lower pathogenicity. However, despite the similarity and widespread use of M. smegmatis to investigate the mechanisms of drug resistance in tuberculosis, these bacteria have some differences in gene expression, metabolic pathways, and cell structure. In particular, M. tuberculosis was shown to produce significantly more H₂S compared to M. smegmatis [9]. Moreover, these bacteria differ significantly in growth rate, which may play an important role in altering thiol levels during antibiotic-mediated growth inhibition. This suggests that some caution is needed when generalizing the results obtained in M. smegmatis to M. tuberculosis.

In this work, we have shown that, just like in E. coli and B. subtilis, the main cause of endogenous H₂S production in M. smegmatis is an increase in intracellular cysteine concentration, which may result from over import of exogenous cysteine/cystine or from dramatic inhibition of protein synthesis by chloramphenicol treatment. In both situations, the increase in intracellular cysteine levels induced mechanisms of its homeostasis. In E. coli, when protein synthesis is inhibited by antibiotics or as a result of nutrient starvation, most of the excess cysteine is incorporated into glutathione, while the remaining portion is exported into the medium and degraded to form H₂S [11, 12, 26]. All these mechanisms of cysteine homeostasis are maximally activated during intracellular cysteine overload as a result of excess cystine import when it is suddenly added to a medium with sulfate as the only source of sulfur [6, 13, 16, 17]. During E. coli growth in cysteine-containing Luria-Bertani (LB) medium (Miller), H₂S production by cells is carried out without any external influences. Under these conditions, the onset of H₂S release coincides with the slowdown of growth and respiration, which may also indicate the occurrence of excess cysteine at this stage of culture growth [27, 38].

In B. subtilis, which does not contain GSH, as well as in the glutathione-deprived gshA mutant of E. coli, the main pathways to restore cysteine homeostasis are its export and degradation to form H₂S, which is more intense and prolonged than in the wt strain of E. coli [12, 13, 14, 26]. It appears that bacillithiol, the major LMW thiol in B. subtilis as well as other Firmicutes [42], does not function as a cysteine buffer, unlike GSH in E. coli.

Mycothiol is the unique protective thiol of Actinobacteria [30] and the most abundant LMW thiol in M. smegmatis [29]. In this work, we have shown that an increase in intracellular cysteine is accompanied by an increase in mycothiol levels, indicating its role as a cysteine buffer in M. smegmatis. We also observed cysteine release from cells and H₂S production upon exposure to chloramphenicol, indicating that M. smegmatis possesses all the mechanisms of cysteine homeostasis characteristic of E. coli: incorporation of excess cysteine into buffer molecules, release into the medium, and desulfurization.

Measurements showed that M. smegmatis growing under normal conditions contained a low cysteine level (0.115 $\pm$ 0.008 µM/OD₆₀₀), which is close to the cysteine level in E. coli cells (0.13 $\pm$ 0.02 µM/OD₆₀₀) [12] and corresponds to a concentration of about 0.1 mM. The need to maintain a low concentration of free cysteine in the cytoplasm is related to its ability to generate ROS upon heavy metal-catalyzed autoxidation and to reduce Fe³⁺ to Fe²⁺, which potentiates the Fenton reaction yielding toxic hydroxyl radicals [5, 6]. In GSH, $\gamma$-glutamyl and glycine residues reduce the rate of Cu-catalyzed autoxidation from 8- to 26-fold, and in MSH, acetyl and GlcN-Ins residues make Cu-catalyzed autoxidation of MSH about 30-fold slower than cysteine and 7-fold slower than GSH [30]. It makes MSH in M. smegmatis, like GSH in E. coli, a suitable buffer for safely storing cysteine when its excess occurs. MSH levels in M. smegmatis have been shown to range from 16 nmol/10⁹ cells (~4.5 mM) in the early exponential phase to 26 nmol/10⁹ cells (~7.5 mM) in the late stationary phase [29], being comparable to GSH concentrations in E. coli. Unfortunately, using our method, we were unable to determine changes in the other major LMW thiol ergothioneine (ERG) in M. smegmatis, whose content is much lower than that of MSH and which is preferentially present as a thione rather than a thiol at physiological pH [37, 43]. MSH and ERG were shown to contribute to protection of mycobacteria against reactive oxygen and nitrogen species, alkylating agents and antibiotics, and also play a regulatory role and be involved in adaptation to low pH and other stresses, including through S-thiolation of proteins [31, 32].

E. coli can utilize cysteine-specific efflux pumps (EamA, EamB, Bcr) to export cysteine into the medium [18, 19, 20]. Thus, constitutive expression of the cysteine exporter EamA in E. coli has been shown to prevent H₂S formation by reducing the intracellular concentration of cysteine [26, 44]. However, no such transporters have been identified in M. tuberculosis. Nevertheless, we observed cysteine release from M. smegmatis cells and its accumulation in the medium upon exposure to chloramphenicol, which may indicate that an unidentified cysteine export system is present in this bacterium. In E. coli, AlaE was found to be the primary exporter of excessive intracellular cysteine, although this protein was previously identified as an alanine exporter [17]. In contrast to E. coli, where ciprofloxacin (10 µg/mL) caused a 3-fold accumulation of cysteine in the medium compared to the control [45], M. smegmatis did not increase extracellular cysteine levels under these conditions. Chloramphenicol-induced cysteine release in M. smegmatis was also 2.6 times lower than that in E. coli. This may be explained by the slower inhibition of growth and hence protein synthesis in M. smegmatis compared to E. coli, where the higher initial growth rate (0.68 $\pm$ 0.01 h^-1 compared to 0.25 $\pm$ 0.01 h^-1 in M. smegmatis) and its rapid decline immediately after antibiotic exposure result in a greater excess of intracellular cysteine. The action of other antibiotics on mycobacteria can also be accompanied by the release of LMW thiols. Accumulation of extracellular thiols has been previously observed when M. tuberculosis was treated with bacitracin. Mutants unable to produce extracellular thiols showed increased sensitivity to this antibiotic, supporting their role in detoxification [46].

M. tuberculosis encodes multiple enzymes that may produce H₂S. Cysteine desulfurization involving Cds1 has been shown to be the major, although not the only, source of endogenous H₂S in M. tuberculosis [9]. Cds1 is also present in M. smegmatis [9], but the involvement of this or other enzymes in H₂S generation upon addition of cystine and chloramphenicol needs to be further studied. M. smegmatis has been reported to produce little H₂S [9]. However, our experiments showed that sudden addition of cystine to M. smegmatis growing in sulfate medium causes the release of 0.3 µM sulfide, which is 4.5 times lower than that of E. coli (1.4 µM) [13], but is well detectable by the methods we used. The difference in H₂S production may be due to variations in the culture medium composition. We found that excluding glycerol from the medium intensified H₂S generation when cells were treated with cystine and chloramphenicol, which may be due to changes in the cell wall composition and permeability [40, 41]. The amount of sulfide released by M. smegmatis under the action of chloramphenicol is significantly lower than that of E. coli (10–15 nM and 180 nM, respectively). The reason, as in the case of cysteine export, may be a smaller excess of intracellular cysteine that occurs in M. smegmatis during rapid inhibition of protein synthesis.

Antibiotic-induced H₂S formation appears to result from disruption of cysteine homeostasis due to protein synthesis inhibition. Under these conditions, growth inhibition in M. smegmatis was accompanied by a decrease in the rate of respiration and glucose consumption, as observed previously in E. coli [11, 12, 13]. However, although both bacteria released H₂S when cystine was added, the physiological response was opposite. Reversible inhibition of growth and respiration was observed in the case of E. coli [13], whereas temporary stimulation of respiration and activation of glucose consumption occurred in M. smegmatis. These findings can be attributed to the peculiarities of the effect of H₂S on cytochrome oxidases: in E. coli, low micromolar concentrations of H₂S inhibit cytochrome bo oxidase [16, 47], while in M. tuberculosis, H₂S stimulates respiration via cytochrome bc₁/aa₃ and, primarily, via cytochrome bd [9]. Another evidence of the participation of H₂S in regulation of energy processes, including the membrane potential, in M. smegmatis may be the abrupt acceleration of K⁺ entry into cells, whose kinetics coincide with sulfide production upon addition of cystine and antibiotics. The ability of H₂S to interact with K⁺ ion channels has been previously shown [48]. The special role of exogenous and endogenous H₂S in regulating central metabolism and accelerating respiration and growth in mycobacteria may be among the reasons for the opposite effect of H₂S on antibiotic susceptibility in M. tuberculosis and E. coli [2, 8, 9].

We have previously shown that the absence of one of the mechanisms of cysteine homeostasis leads to increased activation of the remaining mechanisms when excess cysteine appears in the cytoplasm. In particular, the lack of glutathione in the gshA mutant of E. coli causes more intense production of H₂S and cysteine export when cells are treated with ciprofloxacin, whereas the mstA mutant, which does not produce H₂S under these conditions, synthesizes more glutathione than the wt strain does [13]. The gshA mutant exhibited an increased sensitivity to ciprofloxacin in minimal M9 medium, especially with fractional addition of cystine during cultivation. In the present work, we showed that with increasing concentration of intracellular cysteine in M. smegmatis, the level of mycothiol rises, similar to that of GSH in E. coli. We speculate that, similar to E. coli, a decrease in MSH levels due to mutations or inhibitors of the enzymes involved in its synthesis will intensify H₂S production under stress conditions. Since exogenous and endogenous H₂S have been shown to enhance the killing of M. tuberculosis by the anti-TB antibiotics clofazimine and rifampin [8, 9], it is conceivable that inhibition of MSH synthesis may increase the susceptibility of mycobacteria to antibiotics via the same mechanism. Mycothiol is essential for M. tuberculosis survival and intracellular levels of this thiol are associated with changes in resistance to antibiotics and oxidative stress [39]. M. smegmatis devoid of MSH exhibited an increased susceptibility to several antibiotics, such as erythromycin, azithromycin, rifamycin S, penicillin G, and vancomycin [30, 49], suggesting that MSH is involved in multiple detoxification mechanisms, one of which may be the maintenance of cysteine homeostasis. Although caution is needed when extrapolating the results obtained with M. smegmatis to M. tuberculosis and further more in-depth studies are needed, our findings may be promising in searching for ways to improve the efficacy of anti-TB drugs.

In this study, we showed that, similar to E. coli and B. subtilis, addition of cystine and chloramphenicol to growing M. smegmatis induces intracellular cysteine accumulation and activates homeostatic mechanisms such as cysteine export, its degradation to H₂S, and incorporation into mycothiol as a cysteine buffer. Ciprofloxacin also increased intracellular cysteine concentration and sulfide production but did not induce cysteine release. Our findings indicate that activation of cysteine homeostasis mechanisms may be part of a universal stress response across bacterial species. The molecular mechanisms underlying these processes in mycobacteria require further investigation, including key gene knockout experiments.

The datasets used or analyzed during the current study are available from the corresponding author on reasonable request.

Conceptualization, GS and OO; Investigation, AT, LS, TK, and VU; Formal analysis, OO, GS, AT, LS, TK, and VU; Methodology, OO, GS, AT, LS, TK, and VU; Data curation, GS and OO; Writing–original draft preparation, GS; Writing–review and editing, OO and GS. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This study does not involve human or animal subjects; therefore, according to Russian regulations and policies, ethical committee approval is not necessary.

In this study, we used the equipment of Centre of collective usage of scientific instruments of Perm Federal Research Centre.

This work was supported by assignments # 124020500028-4.

The authors declare no conflict of interest.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/FBL44441.