Dulbecco’s phosphate-buffered saline (D-PBS) with magnesium but without calcium, Hank’s balanced salt solution with calcium and magnesium but without phenol red and sodium hydrogen carbonate (HBSS), NaSH, diamide, N-formyl-l-methionyl-l-leucyl-l-phenylalanine (fMLP), 6-aminonicotinamide (6-AN), and fibrinogen from human plasma were purchased from Sigma (Steinheim, Germany). Dextran T-500 was from Pharmacosmos (Holbæk, Denmark). The acetoxymethyl ester (AM)-conjugated carboxy-2',7'-dichlorodihydrofluorescein diacetate (H₂DCF-DA), fura-2 AM, and the goat anti-mouse IgG secondary antibody, Alexa Fluor™ 488, were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Purified mouse anti-5-LOX monoclonal antibodies (mAbs) were from BD Biosciences (Franklin Lakes, NJ, USA). The GSH/GSSG-Glo™ and RealTime-Glo™ Extracellular ATP Assay kits were from Promega Corp. (Madison, WI, USA).

Bacteria (S. typhimurium strain IE 147) were obtained from the collection of the N.F. Gamaleya National Research Center for Epidemiology and Microbiology (Moscow, Russia). Bacteria were grown in Luria–Bertani broth to a concentration of 1 × 10⁹ colony-forming units (CFU) per milliliter. In this study, non-opsonized bacteria were used.

Human PMNLs were isolated from freshly collected citrate-anticoagulated blood obtained from healthy adult volunteers of both sexes. Leukocyte-rich plasma was obtained from donated blood by sedimentation in the presence of dextran T-500. Granulocytes were obtained as described (17). Cell viability was examined using the trypan blue exclusion method. PMNLs (96%–97% purity and 98%–99% viability) were stored at room temperature in calcium-free D-PBS containing 1 mg/ml glucose until use.

PMNLs [(1.3–1.6) × 10⁷/6 ml HBSS with 10 mM HEPES] were placed at 37°C in a CO₂ incubator for 10 min, and then bacteria or reagents were added. Intact or heat-inactivated S. typhimurium bacteria were used. Heat inactivation was performed by incubation in a water bath at 70°C for 1 h (18). S. typhimurium and the indicated reagents were added for 30 min, followed by exposure to 0.1 μM fMLP for 10 min. The incubations were stopped by adding an equal volume of methanol (−18°C) with 90 ng PGB2 as an internal standard. The major 5-LOX metabolites—5S,12R-dihydroxy-6,14-cis-8,10-trans-eicosatetraenoic acid (LTB₄), iso-LTB₄ (5S,12SR-all-trans-diHETE), ω-OH-LTB₄, ω-COOH-LTB₄, and 5S-hydroxy-6-trans-8,11,14-cis-eicosatetraenoic acid (5-HETE)—were identified as previously described (19).

PMNLs (10⁶/ml HBSS/HEPES) were incubated in fibrinogen-coated confocal dishes at 37°C with 5% CO₂ according to the experimental design. After the incubation period, the cells were fixed with 2.5% paraformaldehyde solution followed by acetone permeabilization. The samples were then incubated overnight at 4°C with mouse anti-5-LOX mAb [1:100 in 1% bovine serum albumin (BSA)/PBS]. After washing three times, the samples were stained with the goat anti-mouse Alexa Fluor™ 488 (1:100 in 1% BSA/PBS) for 6 h at 4°C. The cell nuclei were stained with 0.5 μg/ml Hoechst 33342 (Thermo Fisher Scientific, Waltham, MA, USA). Image acquisition was performed using a fluorescence microscope, Olympus IX 83 (Tokyo, Japan), equipped with ×60 oil immersion objective. At least eight random pictures were captured for each sample.

Intracellular ROS accumulation was quantified by measuring the green fluorescence of 2',7'-dichlorofluorescein (DCF). Loading was performed according to the manufacturer’s instruction. Briefly, the neutrophils were incubated in D-PBS supplemented with 5 mM H₂DCF-DA for 60 min at room temperature followed by washing with PBS, suspended in D-PBS, and then stored at room temperature in the dark until use. Before the experimental treatment, the cells were equilibrated for 5 min in HBSS/HEPES in fibrinogen-coated wells of a 96-well plate (4 × 10⁵ cells/well) at 37°C and 5% CO₂. The fluorescence intensity at excitation and emission wavelengths of 488 and 525 nm, respectively, was measured using a CLARIOstar multi-mode microplate reader (BMG Labtech, Cary, NC, USA). MARS data analysis software package from BMG Labtech was used to process the data obtained.

Quantitative assessment of the ratio of reduced to oxidized GSH was performed using a commercial luminescence-based system, GSH/GSSG-Glo™ assay. Briefly, two sets of PMNLs (one for total GSH and another for GSSG measurement) in HBSS/HEPES were treated according to the experimental protocol in a white 96-well plate (10⁵ cells/well). The cell lysis and all subsequent manipulations were carried out in strict accordance with the manufacturer’s instructions. Luminescence measurements were made using a CLARIOstar microplate reader. To convert the luminescence intensity values (relative light units, RLU) into GSH and GSSG concentrations, a GSH standard curve (0–16 µM) was used.

Changes in the intracellular calcium concentration ([Ca²⁺]_i were detected with ratiometric calcium-sensitive fluorescent dye fura-2 AM. All procedures were performed according to the manufacturer’s instructions, with minor modifications. Briefly, isolated PMNLs (10⁷ cells/ml) were incubated with 1 µM fura-2 AM in D-PBS for 30 min at 37°C. Loaded cells were washed once with PBS and resuspended in D-PBS. Immediately before the experimental procedure, labeled cells were seeded in fibrinogen-coated black 96-well F-bottom plates containing warmed HBSS/HEPES medium, equilibrated for 5 min, and treated according to the experimental design at 37°C in 5% CO₂. Stimuli were added using reagent injectors integrated into the reader platform. Changes in the fluorescence emitted at 510 nm were measured when excited at both 380 nm (for Ca²⁺-free dye) and 335 nm (for Ca²⁺-bound dye) every 0.6 s. Manipulations were performed on a CLARIOstar microplate reader. MARS data analysis software package was used to process the data obtained. [Ca²⁺]_i shifts were assessed based on the changes in the ratio of fluorescence intensities produced by excitation at two wavelengths. Data were quantified using areas under the kinetic curves (AUCs) above the baseline.

An ATP detection component from the RealTime-Glo™ Extracellular ATP Assay kit was used. In accordance with the manufacturer’s protocol, the lyophilized enzyme/substrate mixture (ATP assay substrate) was reconstituted by HBSS/HEPES to obtain the ATP detection reagent. Just before the experiment, the PMNLs were seeded in fibrinogen-coated solid white 384-well F-bottom plates (5 × 10⁴ cells/well) and pre-incubated for 5 min at 37°C with 5% CO₂. S. typhimurium alone or in combination with NaSH and/or diamide was added for 20 min, followed by treatment with fMLP. The PMNLs incubated without the addition of stimuli were used for data normalization. For total ATP assessment, digitonin (40 µg/ml final concentration) and the ATP detection reagent were added either immediately before or 3 min after the addition of fMLP. After 3 min orbital shaking, the luminescence intensity was measured on a CLARIOstar microplate reader (BMG Labtech, Ortenberg, Germany). MARS data analysis software package from BMG Labtech was used to process the data obtained.

Graphing and statistical analysis were performed using GraphPad Prism software version 10.3.1 for Windows. Results are presented as the mean ± SEM. Differences with a p-value <0.05 were considered statistically significant. Two-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons test was used to quantify the 5-LOX product synthesis and ROS assessment. Repeated measures (RM) one-way ANOVA and Tukey’s multiple comparisons test were used to quantify the GSH/GSSG ratio and the calcium flux and total ATP, respectively.

Gut microbes may synthesize H₂S (20). Bacteria use H₂S in response to host-induced stress factors such as oxidative stress (21). The sulfur compound H₂S is a highly reactive molecule that can suppress the accumulation of both ROS and reactive nitrogen species (RNS) in inflammatory conditions (22). In the experimental model of neutrophil interaction with the bacteria S. typhimurium, the exogenous H₂S donor NaSH enhanced leukotriene synthesis depending on the bacterial load ( Figure 1 ). The bacterial load is represented by the multiplicity of infection (MOI) value, i.e., the ratio between the number of bacteria and PMNLs. The effect was observed after 30 min pre-incubation of the neutrophils and bacteria with NaSH, followed by the addition of 0.1 µM fMLP. The effect was significant at medium values of MOI ( Figures 1A–D ).

NaSH had a moderate antioxidant effect, suppressing the intracellular ROS accumulation in neutrophils including in cells interacting with bacteria and/or stimulated by formyl peptide ( Figure 1E ).

What is the interplay between hydrogen sulfide and oxidative stress?

Oxidative stress is an imbalanced condition caused by the excess production of ROS and the lack of antioxidants (23). Generation of ROS by the catalytically active NADPH oxidase complex is one of the main mechanisms of pathogen degradation by neutrophils (24). Bacteria use various strategies to evade restriction by human neutrophils. Intracellular Salmonella, under oxidative stress, initiates the synthesis of the antioxidant H₂S (9).

H₂S can scavenge ROS and increase the GSH level in cells (25). In addition, viable bacteria may inhibit the assembly of the phagocyte NADPH oxidase complex, thus preventing synthesis of the microbicide ROS in neutrophils.

The addition of nanomolar amounts of phorbol-12-myristate-13-acetate (PMA) to PMNLs at the pre-incubation with bacteria stage caused a moderate increase in the level of intracellular ROS and blocked the synthesis of leukotrienes in response to the addition of fMLP, regardless of the presence of the antioxidant NaSH ( Figures 2A, B ). Under the influence of PMA, a depletion of the intracellular pool of reduced GSH was also observed, which was insensitive to NaSH addition ( Figure 2C ). Moreover, an artificially induced oxidative burst superimposed on the period of PMNL–bacteria interaction suppressed the fMLP-induced calcium influx ( Figure 2D ).

ROS-producing neutrophils efficiently kill bacteria (26, 27). It can be assumed that ROS do not contribute to the increase in leukotriene synthesis in neutrophils as there is no need to attract more neutrophils.

S. typhimurium expresses proteins that disrupt the neutrophil NADPH oxidase assembly and alter the ROS production by neutrophils (28). In addition, antioxidant GSH is used by the majority of Gram-negative bacteria, including Salmonella (29, 30). Oxidation of GSH and low-molecular-weight (LMW) thiols was observed when Pseudomonas aeruginosa is phagocytosed by human neutrophils (31). For protection from irreversible oxidative damage by neutrophil oxidants, bacterial LMW thiols form disulfides with each other and with protein cysteines (so-called S-thiolation) (32, 33). Therefore, oxidative stress can induce the oxidation of thiol groups and the formation of disulfide bonds, supporting easy thiol-based redox switches.

To simulate such a situation, i.e., oxidative stress without ROS, diamide was used (15), under the influence of which a significant decrease in the GSH/GSSG ratio was observed, which was not compensated by the H₂S donor compound, NaSH ( Figure 2C ). Under diamide-induced disulfide stress, the formation of leukotrienes was slightly increased ( Figure 2E ). At the concentration used, diamide did not exert an independent effect on the accumulation of intracellular ROS and did not suppress the antioxidant effect of NaSH (data not shown). However, in combination with NaSH, it synergistically supported leukotriene synthesis during the interaction of neutrophils with the bacteria S. typhimurium ( Figure 2E ).

NaSH and diamide demonstrated the ability to influence the Ca²⁺ influx during the interaction of neutrophils with the bacteria S. typhimurium ( Figures 2F, G ). Figure 2G presents the Ca²⁺ flux data both during neutrophil–bacteria interaction and at fMLP addition. NaSH and diamide had virtually no effect on fMLP-induced Ca²⁺ response, but caused a significant increase in Ca²⁺ influx when added during PMNL pre-incubation with bacteria ( Figure 2G ). These data are presented in Figure 2F as AUCs.

An increase in Ca²⁺ initiates the translocation of 5-LOX into the nuclear membrane (34), which is necessary for 5-LOX activity (35). The stimulation of neutrophils with bacteria, NaSH, and diamide supported the translocation of 5-LOX into the nuclei ( Figure 2H ). In combination with the ROS-producing compound PMA, NaSH did not increase leukotriene synthesis ( Figure 2B ).

In combination with the sulfhydryl (SH)-targeting agent diamide, NaSH increased leukotriene synthesis ( Figure 2E ), i.e., the thiol-mediated signaling events differed from that of ROS-dependent signaling.

The intracellular ATP concentration is known as an activation factor of 5-LOX (36, 37). Glycolysis plays a major role in ATP production in granulocytes (26, 38). Upon activation, neutrophils switch from glycolysis to the pentose phosphate pathway (PPP) to increase NADPH production; however, this comes at the expense of ATP (39). Stimulation with 10⁻⁷ M fMLP resulted in the rapid activation of PPP, which was complete at approximately 5 min. When a neutrophil is activated, the hexose monophosphate shunt starts to produce NADPH through the oxidation of glucose-6-phosphate to ribulose-5-phosphate. Oxidative stress addresses more glucose toward the PPP with increased NADPH production.

The neutrophil ATP levels slightly increased after cell incubation with bacteria, but decreased sharply after the addition of fMLP ( Figure 3A ). In experimental conditions corresponding to Figure 2E , when NaSH and diamide synergistically increased leukotriene synthesis, the ATP level in neutrophil, detected just before fMLP addition, significantly increased ( Figure 3A ).

H₂S can modulate the activity of the key enzyme of PPP, glucose-6-phosphate dehydrogenase (G6PD) (40), the enzyme responsible for NADPH accumulation. PPP stimulates NADPH recycling for antioxidant protection. In resting cells, G6PD is inhibited by NADPH (41). The addition of GSH-oxidizing agents such as diamide to HeLa cells resulted in an immediate decrease in the intracellular NADPH dependent on the availability of glucose in the culture medium (42). Increasing both the level of oxidized glutathione (GSSG) and the NADP⁺/NADPH ratio restored the activity of NADPH-inhibited G6PD (43). Indeed, diamide fuels the PPP pathway. Furthermore, diamide in combination with NaSH synergistically increased leukotriene synthesis during the interaction of neutrophils with the bacteria S. typhimurium ( Figure 2E ).

Using inhibitors of glycolysis and PPP, we attempted to find functional evidence for the modulation of leukotriene synthesis by the glycolytic and PPP pathways. Without NaSH or diamide, leukotriene synthesis was suppressed by the inhibition of glycolysis with 2-deoxy-d-glucose (2-DG) and was slightly increased by the inhibition of PPP with 6-AN ( Figure 3 ).

Diamide is known to stimulate PPP-dependent NADPH production (44). Treatment of cells with diamide (not more than 100 µM) increased protein glutathiolation and influenced cellular bioenergetics, increasing the glycolytic flux (45). This activity can contribute to increased ATP levels in the presence of diamide ( Figure 3A ). Diamide decreased the dependence of leukotriene synthesis on glycolysis inhibition by 2-DG ( Figure 3D ). The slight inhibition of PPP in favor of glycolysis supported leukotriene synthesis ( Figures 3B, E ).

It was reported that H₂S elevated G6PD activity (40), the key enzyme in PPP. At the same time, NaSH suppressed ROS generation ( Figure 1 ), a pathway utilizing NADPH. Reduction in the levels of ROS allows cells to maintain a larger pool of reducing equivalents, in particular NADPH, which eliminates the need for the active turnover of PPP. It was found that, in the presence of NaSH, leukotriene synthesis was not sensitive to the G6PD inhibitor 6-AN and to the inhibitor of glycolysis 2-DG at low bacterial loads ( Figures 3B, D ). At high bacterial loads, the inhibition of glycolysis decreased leukotriene synthesis; however, the inhibition of G6PD with 6-AN increased leukotriene synthesis in the presence of NaSH ( Figure 3E ).

Inhibition of non-oxidative PPP (non-oxPPP) with the transketolase inhibitor OT suppressed leukotriene synthesis at 4 mM, but was not affected at 3 mM ( Figure 3C ). Leukotriene synthesis was sensitive to the inhibition of glycolysis by 2-DG at high bacterial loads ( Figure 3E ). Glycolysis is the main source of ATP in neutrophils (38, 46), and at high bacterial loads, a slight inhibition of oxPPP (in favor of glycolysis) increased leukotriene synthesis ( Figure 3E ).

It can be concluded that redox processes induced by the GSH-supporting compound NaSH and disulfide stress induced by diamide influence the 5-LOX activity also through the energy metabolism in neutrophils.

LTB₄ ω-hydroxylation stimulates its transformation into ω-hydroxy-LTB₄, an endogenous inhibitor of LTB₄ chemotactic activity (47). When LTB4 is transformed, the signals for attracting neutrophils are attenuated. The increasing concentration gradient of LTB₄ near the microbial cluster is ensured, among other things, by blocking the ω-OH and ω-COOH transformation of LTB₄, i.e., self-amplification near the microbial cluster, and a decrease in the concentration of LTB₄ as the number of bacteria decreases. Such self-reinforcement and self-limitation mechanisms form the self-organized swarming behavior of neutrophils.

Previously, we have shown that LTB₄ ω-hydroxylation is inhibited at high bacterial loads (19). In this study, it was observed that NaSH supported the ω-hydroxylation of LTB₄ at low and medium bacterial loads ( Figure 4 ). Moreover, it was found that, in the presence of heat-inactivated bacteria, LTB₄ ω-hydroxylation was not inhibited with increasing bacterial loads ( Supplementary Figure S1 ).

The ROS-producing oxidative stress induced by PMA did not change this dependence and supported the ω-hydroxylation of LTB₄ at low and medium bacterial loads ( Figure 4 ). The disulfide stress induced by diamide decreased the ω-hydroxylation of LTB₄ ( Figure 4 ).

When the bacterial load increased, ω-hydroxylation was inhibited in all treatments ( Figure 4 ). Interestingly, the same shift in metabolic profile occurred when an NADPH oxidase inhibitor (diphenyleneiodonium, DPI) was added to the cells. The NADPH oxidase inhibitor DPI suppressed the ω-hydroxylation of LTB₄ regardless of the bacterial load ( Figure 4 ). This indicates that self-limitation of swarming, based on the ω-hydroxylation of LTB₄, will not work during this treatment. A recently published paper has shown that neutrophils from patients with chronic granulomatous disease (i.e., with dysfunctional NADPH oxidase) formed a swarm that grew continuously and disproportionately to the point of invasion (48).

Neutrophil swarming is controlled by LTB₄ and its interaction with its receptor, BLT1 (49). The disulfide stress induced by diamide, in combination with NaSH, resulted in a significant enhancement in LTB₄ synthesis, in parallel with the decreased ω-hydroxylation of LTB₄ ( Figure 2 ), which may accelerate the accumulation of neutrophils and the formation of clusters around the invading microorganism.