Materials and methods

PMN Purification

Blood was collected from buffy coats of healthy donors. Purified human granulocytes were obtained by a two-step gradient separation. Briefly, 25 ml of 1:4 PBS-diluted blood was layered upon 20 ml of Lympholyte-H (Cedarlane, Hornby, Ontario, Canada) and centrifuged at 390 g for 35 min. Plasma and platelets were discarded, while lymphocytes were collected for subsequent experiments in this step. The pellet of red cells and granulocytes was resuspended in cold lysis buffer (NH₄Cl 155 mM and KHCO₃ 10 mM) and centrifuged at 250 g for 10 min. Recovered cells were then resuspended in PBS and the granulocytes were isolated from red-cell ghosts by a Percoll gradient (Sigma-Aldrich, Milano, Italy). In detail, 35 ml of resuspended granulocytes were stratified on 15 ml of a 1:1.58 Percoll PBS-dilution and centrifuged at 350 g for 25 min. Purity of PMN cells was checked by anti-CD15 FITC mAb (Exalpha, Maynard, MA, USA) staining and flow cytometry. Only samples exceeding 95% purity were used for subsequent experiments. Contaminant lymphocytes were also monitored by staining PMN aliquots with anti-CD3 FITC mAb (Exalpha).

In some experiments, eosinophils were isolated from granulocytes preparations by immunomagnetic negative selection based on a anti-human CD16 antibodies cocktail (Milteny, Auburn, CA, USA). Purity of eosinophils was assessed by immunofluorescent double staining with anti-CD15 FITC and anti-CD16 PE mAbs (Exalpha) and flow cytometry.

Cell Cultures and Treatment

Purified human PMN or lymphocytes were cultured up to 24 h, at an optimal cell density of 1 10⁶ cell/ml, in serum-free RPMI medium (Cederlane). NaHS (Aldrich, cat. no. 161527) was immediately added to the cell cultures at concentrations ranging from 0.23 to 3.66 mM. Control cultures were treated as above in the absence of NaHS.

In some experiments, purified cell cultures were treated for 24 h with (i) the MEK-1/MEK-2 inhibitor U0126 (30 M), or (ii) the PI3 kinase inhibitor LY294002 (60 M), or (iii) the p38 MAPK inhibitor SB202190 (30 M). Caspase activation was inhibited by ZVAD-FMK (30 M). All inhibitors were from Calbiochem (Darmstadt, Germany), and were resuspended in DMSO stock solution. Control samples were therefore treated with DMSO alone at the same working dilutions. Cell cultures were treated with the appropriate inhibitor 30 min before addition of NaHS. The NaHS solution was prepared immediately before use at 1 M concentration in RPMI medium, and added to the cell cultures to the final concentration indicated.¹⁷

Assessment of Apoptosis

Cell culture viability was assessed by trypan blue exclusion. Apoptotic cells were analyzed by flow cytometry by staining unfixed cells with propidium iodide (PI) (2.5 g/ml in PBS). The percentage of dead/apoptotic cells was monitored by flow cytometry gating on PI-positive cells. Apoptotic cells stained dimly with PI, while necrotic cells stained bright, as previously described.¹⁸

Western Blot

Cultured cells (1 10⁶) were collected at specific time points, washed in PBS and centrifuged at 200 g for 10 min. Pellets were resuspended in Laemmli sample buffer (0.125 M Tris-HCl pH 6.8, SDS 8%, glycerol 48% w/v, 2-mercaptoethanol 20% v/v and blue bromophenol). The samples were heated at 95°C for 5 min run on 10% SDS-acrylamide gels and finally blotted onto nitrocellulose filters.

Blotted filters were blocked and incubated with specific primary antibodies following manufacturers' protocols. Rabbit polyclonal anti-phospho-p38, anti-Caspase-3, and anti-phospho-PDK-1 antibodies (Cell Signaling Technology, Beverly, MA, USA) were all used at the concentration of 1 l/ml. Mouse anti--actin (Sigma) was diluted 1:5000.

Filters were then washed and further incubated for 1 h at room temperature with 1:5000 peroxidase-conjugated anti-rabbit or with 1:2000 peroxidase-conjugated anti-mouse IgG (Pierce, Milwaukee, WI, USA) in the primary antibody working solution at room temperature. Specific reactions were revealed with the ECL Supersignal West Pico Chemiluminescent Substrate detection system (Pierce).

Bactericidal In Vitro Assay

Bacteria used in this study were a clone of Escherichia coli HB101 strain (Takara Bi Inc., Otsu, Japan), stably transformed for Kanamycin resistance. Bacteria were kept in log phase of growth in Luria broth (LB), in the presence of 30 g/ml Kanamycin (Sigma) and stored in 10% glycerol at -80°C. Stationary phase overnight cultures were transferred to 1 ml of RPMI medium (E. coli dilutions were 1:200, 1:20 000, 1:2 000 000) without antibiotic or serum and grown overnight at 37°C with or without 4 10⁶ of human granulocytes cultured for 24 h in the presence of 1.83 mM NaHS. E. coli concentration was determined in the supernatants as OD absorbance measured by a spectrophotometer (Anthos 2010; Duotech s.r.l., Milano, Italy) at 620 nm wavelength. To determine exactly the residual viability of bacteria, aliquots of E. coli/PMN supernatants were dispersed in LB-agar-kanamycin plates. The number of E. coli colonies was then counted after overnight growth at 37°C.

The number of E. coli remaining alive in 1 l of the supernatant of E. coli/PMN cultures has been reported as colony forming units (CFU)/l.

Transmission Electron Microscopy

Neutrophil cultures were pelleted at 200 g for 10 min, washed and fixed in 2.5% glutaraldehyde in 0.1 M PBS, pH 7.4. Pellets were then fixed in 2% osmic acid in 0.1 M PB, pH 7.4, and embedded in Durcupan (Fluka Chemie, Buchs, Switzerland). Semithin sections were stained with Toluidine blue–0.5% sodium carbonate, while ultra-thin sections were stained with uranyl acetate and lead citrate according to Reynolds¹⁹ and observed with a Philips 300 electron microscope (Philips, Eindhoven, The Netherlands).

Statistical Analysis

All values in the figures and text are expressed as meansstandard error of n observations (with n3). Data sets were examined by analysis of variance (ANOVA) and Dunnett's test (when required). A P-value less than 0.05 was considered as statistically significant.

Dose–response curves of cell viability were analyzed with the nonlinear least squares curvi-fitting program Prism (GraphPAD, San Diego, CA, USA). Percentages of living cells at different doses of NaHS were fitted with a sigmoid dose–response function, thus obtaining the dose that produced a 50% reduction of spontaneous cell death (EC₅₀).

Results

HS^- Selectively Promotes Neutrophils Viability

The effects of hydrogen sulfide on PMN viability were assessed in vitro on purified human PMN cultures, supplemented with increasing doses of NaHS (ranging from 0 to 3.66 mM). Neutrophil viability was quantified after 24 h cultures by PI staining and flow cytometry (Figure 1a). Whereas in control cultures the viability was around 50%, increasing doses of NaHS increased cell viability up to 80%.

The NaHS dose that produced a 50% reduction of spontaneous cell death (EC50s.e.) was 0.500.12 mM. Transmission electron microscopy (TEM) analysis of purified neutrophils cultured 24 h in the presence/absence of NaHS showed that the morphology of viable cells in the presence of HS^- was similar to that of freshly purified, healthy cells (T=0). On the contrary, most cells in the absence of HS^- died by apoptosis (Figure 1b). Time-course experiments performed with a fixed dose of NaHS (1.83 mM) showed that apoptosis of neutrophils was significantly inhibited between 18 and 36 h treatment (Figure 1c). As controls, we isolated from the same donors both eosinophils and PBL. Whereas NaHS at the same doses did not affect eosinophils' viability, it significantly enhanced PBL cell death. To exclude that small lymphocytes contaminants in purified PMN preparations might release factors potentially responsible for the observed pro-survival effect of HS^- on neutrophils, we co-cultured PMN and lymphocytes from the same donor at different ratios. Figure 2 shows that lymphocyte co-culture did not modify the effects of HS^- on PMNs.

Figure 2.

Effects of lymphocytes on PMN viability. Increasing numbers of autologous lymphocytes were cultured with neutrophils in the presence or absence of 1.83 mM NaHS. Bars indicate the percentage of living neutrophils detected by flow cytometry as PI-negative cells. Meanss.e. of three independent experiments.

 

To investigate the cell signalling pathway(s) involved in HS^- pro-survival effect on granulocytes, we next treated our cell cultures with various pharmacological inhibitors that selectively interfere with PI3 kinase, MAP kinase and caspase activities. Figure 3 shows that ZVAD is able to completely preserve the viability of granulocytes after 24 h. This result was somehow expected, because the death of granulocytes is essentially due to caspase-dependent apoptosis. Similarly, the inhibition of p38 MAPK by SB202190 significantly increased granulocytes survival, like ZVAD and HS^- alone. On the contrary, the PI3 kinase inhibitor LY294002 abrogated the pro-survival effect induced by HS^-, not affecting the viability of cells cultured in the absence of HS^-.

Figure 3.

Effect of different pharmacological inhibitors on the survival of purified human neutrophils cultured for 24 h in the presence or absence of 1.83 mM NaHS. Control samples were treated with DMSO alone. The PI3 kinase inhibitor LY294002 was used at a concentration of 60 M; the p38 MAP kinase inhibitor SB202190 was used at a concentration of 30 M, while the general caspase inhibitor ZVAD was used at a concentration of 30 M. Meanss.e. of three independent experiments. ^*P<0.01 LY294002/NaHS-treated cells vs NaHS-treated control cells. ^#P<0.01 SB202190 or ZVAD-treated cells vs untreated control (DMSO) cells.

 

On the basis of these functional data, we therefore analyzed by Western blot the cleavage of caspase-3 and the phosphorylation of p38 MAP kinase in our granulocytes cultures in the presence or absence of NaHS. Figure 4 shows that NaHS (1.83 mM) inhibits caspase-3 cleavage at 6 h as well as p38 MAP kinase phosphorylation as early as 1 h, and that this inhibitory effect persists after 6 h. Of note, -actin degradation, which is an early event in granulocytes apoptosis, is also inhibited by NaHS.

Figure 4.

Time course of caspase 3 (a) and p38 MAP kinase (c) phosphorylation in the presence or absence of NaHS (1.83 mM) in purified neutrophils cultures. (b) and (d) the densitometric histograms of the respective Western blots. (e) The degradation of -actin is inhibited by NaHS, while (f) no modulation of PDK-1 could be observed. Representative Western blots from three independent experiments are shown. Histograms are expressed as meanss.e. of three independent experiments. ^*P<0.01.

 

Bactericidal Effects of Granulocytes

To understand whether or not the pro-survival effect of NaHS on granulocytes could impair their bactericidal activity, we finally performed a set of experiments in which granulocytes were co-cultured with E. coli HB101 in the presence or absence of NaHS. Figure 5 shows that neutrophils cultured in the presence of NaHS (1.83 mM) were able to kill bacteria as efficiently as neutrophils in the absence of NaHS. After 24 h, as expected, neutrophils cultured in the presence of NaHS are much more efficient then those cultured in the absence of NaHS.

Figure 5.

Bactericidal properties of neutrophils in the presence or absence of NaHS. (a) The presence of NaHS (1.83 mM) does not affect the bactericidal properties of neutrophils. On the contrary, no residual colony can be observed in the presence of NaHS (C) owing to the higher number of surviving neutrophils. (b) The densitometric histograms of the culture supernatants of three independent experiments. Meanss.e. of three independent experiments.

Discussion

Hydrogen sulfide is now considered, together with CO and NO, a 'gasotransmitter' able to generate cellular responses upon specific induction. H₂S can be endogenously generated from cysteine or exogenously administered to cells, frequently as a component of more complex sulfurous compounds, such as sulfurous thermal waters.

Knowledge about the effects of such compounds is currently confined to clinical information about their efficacy in the prevention and resolution of bronchial and pulmonary diseases characterized by mucous hypersecretion, inflammatory infiltrate and impairment of mucociliary clearance. Hypersecretion of mucus, in particular, is the defining feature of chronic bronchitis and is associated with an inflammatory process involving epithelium, glands and central airways. Of note, it has recently been demonstrated that progression of chronic obstructive pulmonary disease (COPD) is significantly associated with increasing infiltration of the airways by polymorphonuclear neutrophils and lymphocytes,²⁰ which play a fundamental role against the development of microbial colonization and infection that worsen the progression of chronic bronchitis and COPD.

Given the lack of basic biological information on the effects of sulfur ions on the cellular subsets of the inflammatory infiltrate—as opposed to the abundance of clinical and epidemiological observations on the beneficial effects of sulfur compounds—we decided to study in vitro the effects of sulfur ions on purified polymorphonuclear cell cultures.

We demonstrate that HS^- ions have pro-survival effects on polymorphonuclear neutrophils in vitro, delaying the onset of apoptosis. Particularly under the stress condition of serum-deprivation, the prolonged survival of neutrophils in the presence of HS^- is ultimately due to the inhibition of Caspase-3 cleavage, which delays the onset of apoptosis.

In fact, pharmacological inhibition of caspases prolongs the life of neutrophils, as do HS^- ions. Frasch and co-workers¹⁶ described the spontaneous or Fas-induced neutrophil apoptosis as independent from p38 MAPK activity. While recent data from Alverado-Kristensson offer a deeper insight on the role of p38 in Fas-induced apoptosis, showing that p38 can directly phosphorylate and inhibit caspase 8 and 3,¹⁶ a general agreement exists on the p38 MAPK dependence of stress-induced apoptosis of human neutrophils. Moreover, Yang et al recently demonstrated that both endogenously generated and exogenous H₂S can inhibit cell proliferation.²¹

Accordingly, our data show that sulfur ions inhibit p38 MAPK phosphorylation, which is relevant in inducing apoptosis in neutrophils, as pharmacological inhibition of p38 by SB202190 efficiently inhibits apoptosis to the same extent as HS^- ions. Moreover, surviving neutrophils show efficient bactericidal properties, which adds clear physiological relevance to their prolonged survival.

The pro-survival effect of HS^- should be considered specific for neutrophils, as no effect was observed on eosinophils, and opposite effects were observed on lymphocytes of the same donors. These different biological effects might explain some contradictory data that can be found in the still scarce literature available on HS^-. The opposite effects of sulfur ions on the viability of neutrophils vs lymphocytes might represent the basis for the potentiation of the acute inflammatory and bactericidal response and de-potentiation of the chronic and inflammatory cellular response. Such an inbalance of cellular infiltrate of the respiratory airways during inflammation—together with the known direct antibacterial and fluidifying effects on mucus—might likely represent the underlying biological mechanism of sulfurous water compounds that are successfully used in the prevention and therapy of chronic inflammatory diseases of the respiratory tract.