INTRODUCTION

Recent studies have shown that hydrogen sulfide (H2S) in bacteria is an effective protector against oxidative stress and antibiotics [1]. The main source of H2S in E. coli cells is cysteine, which is prone to desulfidation with the participation of genes aspC and mstA. We have found a close relationship between the generation of hydrogen sulfide and the expression of genes involved in the biosynthesis and transport of cysteine and controlled by the regulatory protein CysB [2]. In particular, we showed that when growing bacteria on the LB medium, the level of H2S generation increases significantly as a result of constitutive expression of gene tcyP, which controls the transport of cystine/cysteine from periplasm to cytoplasm. These data indicate the important role of the exogenous cysteine contained in the LB medium in H2S production.

One of the objectives of this work was to elucidate the role of endogenously synthesized cysteine in H2S production. To this end, we inactivated genes cysK and cysM, which control the cysteine synthesis de novo [3], and studied the effect of the deletions obtained on H2S generation. Unexpectedly, it turned out that bacteria defective for genes cysK and cysM not only retain the ability to produce H2S but also, on the contrary, are characterized by its increased generation. We have shown that deletions of any of the known genes controlling the desulfohydrogenation of cysteine (tnaA, metC, and malY) [4], as well as new, recently discovered genes yciW [5, 6], cyuA, cyuP, and cyuR [79] involved in cysteine degradation, do not prevent H2S formation by ΔcysK ΔcysM mutants. Also, we showed that the cysteine degradation and the H2S formation in the ΔcysK ΔcysM mutant are suppressed by inactivation of genes cysB and cysE. Thus, a new role of the transcription factor CysB in the degradation of cysteine and the formation of H2S has been revealed.

MATERIALS AND METHODS

Bacterial strains and plasmids. The bacterial strains of Escherichia coli used in this work and their genotypes are presented in Table 1. Deletion mutants were obtained by growing phage P1 on strains from the Keio collection [10] containing deletions of genes cysK, cysM, cysB, cysE, tnaA, metC, malY, yciW, cyuA, cyuP, and cyuR and their subsequent transduction into the genome of the E. coli strain MG1655. The preparation of strains containing the genes tcyP and tcyJ under the control of the constitutive promoter Ptet is described in [2].

Table 1.   Genotype of E. coli strains used in this work

Media and culture conditions. The glucose-free LB medium was used as a complete nutrient medium for growing bacteria [11]. When necessary, 10 μg/mL chloramphenicol and 40 μg/mL kanamycin were added to the medium.

Determining the hydrogen sulfide production levels. The strains were grown in a complete LB medium for 18 h in test tubes under the caps of which a filter paper strip moistened with a 2% solution of lead acetate was attached. The level of H2S production by the strains was evaluated visually by the intensity of blackening of the filter paper as a result of the formation of the PbS complex.

Isolation of total RNA. Total RNA was isolated from a culture of E. coli cells with an optical density of 0.4–0.6. RNA purification was carried out using the RNeasy Mini Kit (Qiagen) in accordance with the manufacturer’s instructions. The resulting RNA preparations were treated with DNase I. The quality of total RNA was analyzed in a 1% agarose gel supplemented with formamide. The amount was determined spectrophotometrically by the absorption value at a wavelength of 260 nm.

PCR with product detection in real time. The E. coli cell cultures of the strains analyzed were grown to OD600 = 0.4–0.6, and then total RNA was isolated. Before carrying out the reverse transcription reaction, RNA samples were treated with DNase I (Thermo). The reverse transcription reaction was carried out in the presence of oligonucleotides specific to the genes under study using the SuperScript III First-Strand Synthesis Kit for RT-PCR (Invitrogen). Further, 1 μL of the volume of the entire reverse transcription reaction was used as a template for real-time detection PCR. Expression levels of genes def and rpoA were used for normalization. The analysis was carried out using a set of reagents for real-time PCR from the company Syntol. Amplification was carried out on a DTlite device (DNA-Technology). The reaction products were analyzed by electrophoresis in a 2% agarose gel to confirm that the products obtained had the expected size. Each reaction was set three times, where an average value was taken as a result. The level of transcription was determined from the values of the threshold cycle, taking into account that the concentration of the original specific DNA fragments increases approximately as 2N, where N is the number of cycles.

RESULTS

Detection of New Cysteine-Degrading Activity in E. coli Cells

According to our data [1], the main role in the H2S generation in E. coli cells is played by the enzyme 3‑mercaptopyruvate sulfotransferase (3-MST) encoded by gene mstA [12]. This enzyme catalyzes the conversion of 3-mercaptopyruvate into pyruvate and H2S (Fig. 1). In turn, 3-mercaptopyruvate is formed during the transamination of L-cysteine by the cysteine aminotransferase enzyme (gene aspC) [13]. As for the synthesis of L-cysteine, it is formed by the condensation of H2S and O-acetylserine in reactions catalyzed by two cysteine synthases under the control of genes cysK and cysM [3].

Fig. 1.
figure 1

The scheme of H2S generation from cysteine with the participation of cysteine aminotransferase enzymes (gene aspC) and 3-mercaptopyruvate sulfotransferase (gene mstA) and cysteine synthesis from O-acetyl-L-serine and H2S under the control of cysteine synthases (genes cysK and cysM). In addition, an alternative pathway for the synthesis of H2S from exogenous sulfate is provided, which is not implemented when growing bacteria on LB medium. The localization of genes tcyP and tcyJ controlling the transport of cystine from periplasm to cytoplasm is shown.

To determine the contribution of the endogenously synthesized cysteine to the H2S generation, deletions of genes cysK and cysM encoding the synthesis of two cysteine synthases in E. coli cells were obtained. It is known that inactivation of gene cysK causes auxotrophy of bacteria for cysteine owing to disruption in the cysteine synthesis de novo [3]; growth of such bacteria in a complete LB medium is provided by cystine/cysteine contained in it. In addition, it was shown that both cysteine synthases simultaneously possess desulfhydrase activity, that is, the ability to degrade cysteine to produce H2S [4]. On this basis, it could be expected that inactivation of genes cysK and cysM would lead to a loss of the ability of bacteria to produce H2S. Figure 2 shows the results of determining the level of H2S production in single mutants ΔcysK and ΔcysM and a double mutant ΔcysK ΔcysM. As controls in these experiments, we used the previously described strain containing the mstA gene deletion with disrupted synthesis of the 3-MST enzyme and characterized by a reduced ability to produce H2S [1] and the Ptet-mstA mutant with constitutive expression of gene mstA with a high level of H2S generation [2].

Fig. 2.
figure 2

Generation of H2S when growing bacteria on LB medium with strains of different genotypes: (1) wt; (2) Δmst; (3) ΔcysK; (4) ΔcysM; (5) ΔcysK ΔcysM; (6) Ptet-mstA; (7) ΔcysK ΔcysM ΔmstA; (8) ΔcysK ΔcysM + cysteine (500 μmol).

As follows from the data presented in Fig. 2, the ΔcysM mutant does not lose the ability to generate H2S (Fig. 2, 4), and the ΔcysK mutant produces it even in larger quantities (Fig. 2, 3) than the wild-type strain. The level of H2S generation in the double mutant cysK and cysM is even higher and close to that of the strain with constitutive expression of gene mstA (Ptet-mstA) (Figs. 2, 5 and 6). One possible reason for the increasing capacity of strains with inactivated cysK and cysM genes to generate H2S might be an increase in the efficiency of transport of cysteine or its derivatives comprising the LB medium into the cell and their subsequent degradation involving the mstA gene product. However, insertion of the mstA gene deletion into the genome of the ΔcysK ΔcysM double mutant resulted in a slight decrease in the ability of such a strain to generate H2S (Fig. 2, 7). At the same time, the addition of exogenous cysteine leads to an even greater increase in the level of H2S production (Fig. 2, 8).

Thus, the obtained data make it possible to conclude that the mstA-independent cysteine-degrading activity leading to efficient generation of H2S is revealed against the background of ΔcysK ΔcysM mutations in E. coli cells.

Generation of Hydrogen Sulfide by ΔcysK ΔcysM Mutants Does Not Depend on the Activity of the Known Desulfhydrase Genes

From published data, it is known that, in addition to cysK- and cysM-encoded cysteine synthases, at least three more proteins possess desulfhydrase activity: tryptophanase (gene tnaA), cystathionine-β-lyase (gene metC), and maltose transport protein (gene malY) [4]. In addition, recently, there have been reports about identification of new genes in E. coli involved in cysteine degradation. One of the genes, yciW, contains a binding motif for the CysB protein in the regulatory region, and its inactivation leads to the accumulation of intracellular cysteine content [5, 6]. The other two genes, cyuA (yhaM) and cyuP (yhaO), controlling the desulfidation of cysteine and its transport, respectively, form an operon that is prone to negative regulation by the CyuR (DecR) protein [79]. To determine the possible role of these desulfhydrases in the production of H2S by the ΔcysK ΔcysM mutant, the genome of the latter was introduced with deletions of the corresponding genes, and the level of H2S generation in the obtained triple mutants was checked (Fig. 3).

Fig. 3.
figure 3

Inactivation of genes tnaA, metC, malY, yciW, cyuA, cyuP, and cyuR does not affect the level of H2S generation by the ΔcysK ΔcysM mutant.

As shown in Fig. 3, none of the additional deletions affected the ability of the ΔcysK ΔcysM strain to produce H2S. Thus, the obtained data allow us to conclude that, against the background of ΔcysK ΔcysM mutations, a new unknown cysteine-degrading activity starts functioning in E. coli cells, leading to efficient generation of H2S.

The New Cysteine-Degrading Activity Depends on the Transcriptional Regulator CysB

Since genes cysK and cysM are part of the cysB regulon, the effect of the inactivation of genes cysB and cysE on the generation of H2S by the ΔcysK ΔcysM mutant was investigated. Gene cysB encodes the synthesis of a transcription factor regulating the expression of a large group of genes involved in the metabolism of cysteine and sulfates and forming the cysB regulon [14]. The CysB protein, as a result of allosteric interaction with N-acetyl-L-serine, becomes active and activates or represses the transcription of target genes [15]. N-acetyl-L-serine is formed in the cell spontaneously from O-acetyl-L-serine, which is the product of a serine-acetyltransferase reaction under the control of gene cysE. Since the transcription of both genes cysK and cysM is under the positive control of the CysB protein, inactivation of genes cysB and cysE results in auxotrophy of the bacteria for cysteine. The results of determining the ability of the cysK and cysM mutants to produce H2S after the insertion of deletions of genes cysB and cysE into their genome are presented in Fig. 4.

Fig. 4.
figure 4

The effect of inactivation of genes cysB and cysE on H2S production by ΔcysK ΔcysM mutants. The level of H2S generation is shown when growing bacteria on LB medium with strains of different genotypes: (1) wt; (2) ΔcysK ΔcysM; (3) ΔcysK ΔcysM ΔcysB; (4) ΔcysK ΔcysM ΔcysE; (5) ΔcysK ΔcysM ΔcysB + 5 mM O-acetyl-L-serine; (6) ΔcysK ΔcysM ΔcysE + 5 mM O-acetyl-L-serine.

As follows from the data presented in Fig. 4 (3 and 4), inactivation of genes cysB and cysE almost completely suppresses the generation of H2S in the ΔcysK ΔcysM strains. The addition of exogenous O-acetyl-L-serine restores the ability of ΔcysK ΔcysM bacteria to generate H2S against the background of the ΔcysE mutation (Fig. 4, 6), but not in the case of inactivation of gene cysB (Fig. 4, 5). The data obtained make it possible to make the following assumptions about the nature of the cysteine-degrading activity that we found. The negative effect of the cysB and cysE mutations on H2S production may be due to the fact that expression of the gene encoding an unknown enzyme with desulfhydrase properties is under positive control of the active form of the CysB protein. Another explanation for the negative effect of the cysB and cysE mutations on H2S production is based on the assumption that inactivation of these genes blocks the transport of cysteine, the main H2S precursor, into the cell.

Effect of Cysteine Transporters on H2S Production by ΔcysK ΔcysM Mutants

To study the effect of the TcyP and TcyJ transporters providing transport of cystine/cysteine from periplasm to cytoplasm of the cell [16, 17] on the ability of the ΔcysK ΔcysM strains to generate H2S, deletions of genes tcyP and tcyJ were introduced into the chromosome of these strains. Since both of these genes are under the control of the CysB protein [19], the ΔcysK ΔcysM mutants were used to construct additional strains containing deletions of genes tcyP and tcyJ against the background of the inactivated cysB gene. In addition, strains with constitutive Cys-B-independent expression of genes tcyP and tcyJ were obtained by placing them under the control of the strong Ptet promoter. In the obtained isogenic strains, the level of H2S generation was determined when growing bacteria on a standard LB medium (Fig. 5).

Fig. 5.
figure 5

The effect of cystine/cysteine transporters on H2S production by ΔcysK ΔcysM mutants. The levels of H2S generation are shown when growing bacteria on LB medium with strains of different genotypes: (1) ΔcysK ΔcysM; (2) ΔcysK ΔcysM ΔtcyP; (3) ΔcysK ΔcysM Ptet-tcyP; (4) ΔcysK ΔcysM ΔtcyJ; (5) ΔcysK ΔcysM Ptet-tcyJ; (6) ΔcysK ΔcysM ΔcysB; (7) ΔcysK ΔcysM ΔcysB Ptet-tcyP; (8) ΔcysK ΔcysM ΔcysB Ptet-tcyJ.

As expected, inactivation of the tcyP and tcyJ transporters results in a significant suppression of H2S production because of a decrease in the flow of exogenous cysteine entering the cell (Figs. 5, 2 and 4), whereas the increase in expression of both transporters under the control of constitutive promoters causes a significant increase in its generation (Figs. 5, 3 and 5). As noted above, inactivation of the cysB gene leads to the suppression of H2S production, which may be a result of a decrease in the cysteine flux to the cell owing to the low activity of the TcyP and TcyJ transporters controlled by the CysB protein. Earlier, we showed that expression of the tcyP gene under the control of the Ptet promoter restores mstA-dependent generation of H2S against the background of the cysB gene deletion [2]. However, as shown in Fig. 5 (7 and 8), the ΔcysK ΔcysM mutants containing copies of the tcyP and tcyJ genes under the control of the constitutive cysB-independent Ptet promoter show no amplification of H2S production compared to the control strain defective for cysB. It follows that the transport of cystine/cysteine into the cell is not a bottleneck for the manifestation and implementation of the new cysteine-degrading activity. On this basis, it can be assumed that the observed dependence of the new cysteine-degrading activity on cysB is most likely due to the fact that the transcription of the gene coding for this activity requires activation with the participation of the regulatory protein CysB.

Inactivation of Genes cysK and cysM Leads to Partial Activation of CysB Regulon Genes and Oxidative Stress Protection Genes

Since the strain defective for cysK cysM should accumulate not only H2S but also O-acetylserine (Fig. 1), which serves as the precursor of the inductor of the CysB regulon genes, we should expect an increase in their transcription level. By real-time PCR, we compared the level of transcription of several genes that make up the CysB regulon, as well as some genes involved in protecting the cells from oxidative stress, in the ΔcysK ΔcysM mutant and the wild-type strain. As follows from Fig. 6, the ΔcysK ΔcysM mutant exhibits an increased level of transcription of genes cysP, tau, and tcyP, which are under positive control of the regulatory protein CysB.

Fig. 6.
figure 6

Increase in the relative level of transcription of several genes in cells of the ΔcysK ΔcysM mutant.

It should be noted that, in addition to the CysB regulon genes, the ΔcysK ΔcysM mutant shows an approximately 2- to 3-fold increase in the expression of genes katG, sodA, arcA, and cydD, indicating an increased level of formation of reactive oxygen species in cells of the ΔcysK ΔcysM mutants (Fig. 6).

It is known that the expression of gene sodA controlling the synthesis of manganese-dependent superoxide dismutase is activated in response to the formation of the superoxide anion in the cell [19, 20], while the induction of gene katG encoding catalase indicates an increase in the level of hydrogen peroxide [21]. These data indicate that the disruption of cysteine resynthesis as a result of the inactivation of the cysK cysM genes provokes a state of oxidative stress. The arcA gene product is the global regulator responsible for switching the cellular metabolism of bacteria as they move from aerobic to anaerobic growth conditions [22, 23]. Approximately under the same physiological conditions, the expression of terminal cytochrome oxidase bd-I is activated, a component of which is the cydD gene product [24, 25]. The increase in the expression level of genes arcA and cydD in the ΔcysK ΔcysM mutants indicates the important role of cysteine synthases in maintaining the redox balance of the cell and requires further research.

DISCUSSION

It is generally believed that the reduced low-molecular-weight sulfur-containing metabolites (cysteine, glutathione, etc.), along with catalases and dismutases, are the main antagonists of reactive oxygen species (ROS). However, these tools have by now received a serious addition in the form of H2S, which is formed in cells of microorganisms and eukaryotes in the processes of transulfurization and degradation of cysteine. Earlier, we showed that the degradation of cysteine and the formation of H2S under the control of aspC and mstA genes of E. coli are effective conditions for inhibiting the most toxic form of ROS, the hydroxyl radical of the Fenton reaction [2]. These ideas about the priority of H2S as an effective protector against oxidative stress are fully supported by the data of this work. We have studied the paradoxical effect of deletions of genes cysK and cysM encoding cysteine synthase enzymes: such cysteine auxotrophs actively degrade cysteine to form H2S. It is noteworthy that the generation of H2S by the ΔcysK ΔcysM mutants is carried out without the participation of the canonical enzymes AspC and MstA. We have shown that deletions of any of the known genes controlling the degradation of exogenous cysteine, including the genes cysK, cysM, tnaA, metC, and malY, as well as the new genes yciW, cyuA, cyuP, and cyuR, do not deprive the cysteine auxotrophs ΔcysK ΔcysM of the ability to degrade cysteine and produce H2S. Thus, the data testify to the existence of a new cysteine-degrading activity leading to intense production of H2S. The nature of this activity remains unclear, but the search for an individual desulfhydrase may not be productive, since the total level of H2S production may be the result of many processes. Since in the growth of the ΔcysK ΔcysM mutants on the LB medium the only source of H2S is cysteine of the medium, it can be assumed that the high level of H2S generation is due to the absence of cysteine resynthesis, which is normally carried out by CysK and CysM cysteine synthases.

An unexpected result of this work is the discovery of the role of the regulatory protein CysB in the cysteine degradation and H2S production. It turned out that the inactivation of gene cysB, which encodes the CysB transcriptional regulator, or gene cysE, whose product O-acetylserine is necessary for the activation of protein CysB, completely suppresses the H2S production in the ΔcysK ΔcysM mutant. Thus, we discovered a new role of the CysB transcription factor as a positive regulator of cysteine degradation with the formation of H2S. At the same time, the activity of the CysB factor in the cysteine auxotroph leads to an increase in the level of transcription of several genes that make up the CysB-regulon (cysP, tau, tcyP). An important characteristic of the ΔcysK ΔcysM mutants is an increase in the level of transcription of the genes that protect against oxidative stress (katG, sodA, arcA, and cydD). This indicates the involvement of cysteine synthases, especially CysK, in maintaining the cell redox balance, which was not described previously: their inactivation provokes oxidative stress. The paradox of the situation is that the exogenous cysteine, which is necessary to maintain the growth of auxotrophs ΔcysK ΔcysM and synthesis of the antioxidant glutathione, undergoes intense degradation with the formation of H2S. From this, it follows that H2S is a more preferred antioxidant than cysteine and its derivatives and corresponds to our notions about the decisive role of H2S in protecting cells from oxidative stress [2] and antibiotics [1, 26].