Inhaled hydrogen sulfide protects against lipopolysaccharide-induced acute lung injury in mice

Background Local pulmonary and systemic infections can lead to acute lung injury (ALI). The resulting lung damage can evoke lung failure and multiple organ dysfunction associated with increased mortality. Hydrogen sulfide (H2S) appears to represent a new therapeutic approach to ALI. The gas has been shown to mediate potent anti-inflammatory and organ protective effects in vivo. This study was designed to define its potentially protective role in sepsis-induced lung injury. Methods C57BL/6 N mice received lipopolysaccharide (LPS) intranasally in the absence or presence of 80 parts per million H2S. After 6 h, acute lung injury was determined by comparative histology. Bronchoalveolar lavage (BAL) fluid was analyzed for total protein content and differential cell counting. BAL and serum were further analyzed for interleukin-1β, macrophage inflammatory protein-2, and/or myeloperoxidase glycoprotein levels by enzyme-linked immunosorbent assays. Differences between groups were analyzed by one way analysis of variance. Results Histological analysis revealed that LPS instillation led to increased alveolar wall thickening, cellular infiltration, and to an elevated ALI score. In the presence of H2S these changes were not observed despite LPS treatment. Moreover, neutrophil influx, and pro-inflammatory cytokine release were enhanced in BAL fluid of LPS-treated mice, but comparable to control levels in H2S treated mice. In addition, myeloperoxidase levels were increased in serum after LPS challenge and this was prevented by H2S inhalation. Conclusion Inhalation of hydrogen sulfide protects against LPS-induced acute lung injury by attenuating pro-inflammatory responses.


Background
Acute lung injury (ALI), and the acute respiratory distress syndrome (ARDS), are two major challenges in clinical practice and both are responsible for high rates of morbidity and mortality amongst intensive care patients [1][2][3]. A variety of stimuli can initiate ALI, such as mechanical ventilation, hyperoxia, ischemia/reperfusion, transfusion, or polytrauma [3]. Sepsis reflects one of the most important causes of ALI [4]. The underlying strong inflammatory response in sepsis-induced ALI is characterized by the transmigration of immune-competent cells (mostly neutrophils) into the lung interstitium and the alveolar space, and the release of numerous pro-inflammatory cytokines, e.g., interleukin-1β (IL-1β) and macrophage inflammatory protein-2 (MIP-2). As a consequence of the inflammatory process, alveolar structures change, endothelial and alveolar permeability increase and alveolar fluid clearance decreases, thus critically impairing lung function [5]. Currently available treatment options have failed to significantly decrease sepsis related mortality. Therefore, alternative strategies are urgently needed to improve supportive care.
In this respect, hydrogen sulfide (H 2 S) has come to be a focus of interest. H 2 S belongs to the group of gaseous transmitters, along with nitric oxide and carbon monoxide.
It is present in the blood and organs of humans and mammals in low micromolar extracellular concentrations. Endogenous H 2 S has been shown to be involved in a series of physiological processes, e.g., inflammation, vasodilatation, neuromodulation, pain perception, as well as in organ protective pathways (reviewed in [6]). Moreover, exogenous application of gaseous H 2 S and H 2 S donors in different animal disease models like ventilator-induced lung injury [7], ischemia/reperfusion injury [8], or oleic acid-induced ALI [9,10], succeeded in exerting organprotective effects. The underlying mechanism may be explained by the ability of H 2 S to inhibit the activation and transmigration of neutrophil cells and to attenuate the release of pro-inflammatory cytokines.
The aims of the present study were to mimic sepsisinduced ALI by intranasal administration of the endotoxin lipopolysaccharide (LPS) and to examine the role of continuously inhaled H 2 S in this injury model.

Animals and experimental setting
All animal experiments were performed in accordance with the guidelines of the local animal care commission (Ethics Committee University of Freiburg, Freiburg, Germany, permission No. G-07/25) and in conformance with the journals' requirements for human and animal trials. C57BL/6 N mice weighing 22.4 g (± 0.3 g) were randomly assigned into four experimental groups: (1) intranasal (i.n.) application of endotoxin-free phosphate buffered saline (70 μl) + spontaneous breathing of room air (PBS + air), (2) PBS treatment + spontaneous breathing of air supplemented with 80 parts per million (ppm) hydrogen sulfide (H 2 S, Air Liquide, Kornwestheim, Germany) (PBS + H 2 S), (3) i.n. application of lipopolysaccharide (70 μl dissolved in PBS, LPS; 0,25 ng E.coli 055:B5; Sigma-Aldrich Chemie GmbH, Munich, Germany) + spontaneous breathing of air (LPS + air), (4) LPS treatment + spontaneous inhalation of 80 ppm H 2 S (LPS + H 2 S). Spontaneous breathing of the respective gas mixture started 1 h prior to PBS or LPS instillation. Instillation of PBS or LPS was conducted under short isoflurane anesthesia. Afterwards, mice were again placed in a sealed plexiglass chamber and subjected to either breathe room air or H 2 S for another 6 h. H 2 S concentration was continuously monitored using a portable gas monitor (MX6 iBrid, Industrial Scientific Corporation, Oakdale, PA).

Tissue sampling and bronchoalveolar lavage
At the end of each experiment all mice were sacrificed by an intraperitoneal, overdosed injection of ketamine (180 mg/kg) and acepromazine (1.8 mg/kg) and additional bleeding. Bronchoalveolar lavage fluid, lung tissue samples, and blood samples were gained and analyzed as described recently [11].

Histological examination
The left lung was prepared, conserved, and cut into 12 μm thick cryosections for hematoxylin and eosin (H + E) staining as previously described [11]. From each lung, four representative photos were taken (magnification x200). Five high power fields (HPF) were randomly assigned to each photo. Subsequently, alveolar wall thickness and cellular infiltration were analyzed by Axiovision software (AxioVS40LE, Zeiss, Jena, Germany). For each HPF, the degree of lung damage was determined using a modified ALI score as described earlier [11]: In brief, (a) thickness of the alveolar walls, (b) infiltration or aggregation of inflammatory cells, and (c) hemorrhage were rated in a blinded fashion. Each item was graded according to the following point scale: 0: minimal damage, 1: mild damage, 2: moderate damage, 3: severe damage, 4: maximal damage. The degree of ALI was assessed by the sum of scores for each HPF, ranging from 0 to 12. The average of the sum of each field score per lung was compared among groups.

Statistical analysis
Experiments were performed with 6-8 mice per group. Group size was defined on the basis of expected neutrophil cell numbers in BAL fluid by power calculations prior to the study. Graphs represent means ± standard error of means (SEM). Data were further analyzed for normal variation prior to one way analysis of variance (ANOVA), followed by the Student-Newman-Keuls posthoc test. In cases of a failed normality test, Kruskal-Wallis ANOVA on Ranks was performed followed by Dunn`s posthoc test. P < 0.05 was considered significant.

Results and discussion
In this study, we clearly demonstrate, that inhalation of H 2 S in low dose prevents the development of acute lung injury. Furthermore, we show that H 2 S treatment substantially reduces local as well as systemic inflammation.

Hydrogen sulfide prevents lung damage in LPS-induced ALI
In the presented model of pulmonary sepsis H + E staining of lung cryosections showed that as compared to control conditions (PBS + air, PBS + H 2 S, Figure 1A + B, respectively), LPS treatment clearly stimulated the formation of lung edema and the influx of immune-competent cells (LPS + air, Figure 1C). In sharp contrast, exposure to 80 ppm H 2 S markedly reduced LPS-induced lung damage (LPS + H 2 S, Figure 1D), reflected by decreased edema formation and cellular infiltration into the lung tissue. These findings were confirmed by quantitative analysis: H 2 S inhalation prevented alveolar wall thickening ( Figure 2A) and cellular infiltration ( Figure 2B), that was otherwise observed in LPS-treated animals kept in room air. In addition, a reduction of lung damage to control levels by H 2 S inhalation was also detected by rating an overall ALI score ( Figure 2C), strongly suggesting a lung-protective role for H 2 S inhalation in LPS-induced ALI.
These results add important information to the role of exogenous H 2 S in sepsis. Conflicting data exist in models of cecal ligation and puncture demonstrating aggravation [12][13][14][15][16][17][18][19][20] as well as substantial reduction of the resulting lung injury in response to application of H 2 S donors [21,22]. It is likely that the route of administration, dosage, timing, and the purity of H 2 S donors may be accountable for the inconsistent data. This study is the first to show that inhalation of H 2 S clearly prevents lung damage due to local LPS-induced injury, underlining the therapeutic potential of this gas under septic conditions. Our findings are further supported by a recent publication by Tokuda et al. [23]. Here, in contrast to our study, LPS was applied systemically. The study found inhalation of 80 ppm H 2 S substantially increased survival. Although the authors did not analyze histopathological changes in the lung, the data suggest an organ-protective effect of H 2 S inhalation. Moreover, we have shown recently that inhalation of H 2 S in low dosage (80 ppm) ameliorated lung pathology in ventilator-induced lung injury [7]. In a related model, Francis et al. demonstrated, that inhalation of 60 ppm was sufficient to exert a series of lung protective effects [24]. These data suggest that application of lower concentrations as we used can achieve lung protection. Although we did not detect any toxic side effects of H 2 S inhalation in our model (data not shown), future studies in this direction are needed to minimize H 2 S exposure.

Hydrogen sulfide mediates lung protection by inhibition of the inflammatory response
We next investigated whether the observed H 2 S-mediated lung-protection was attributed to inhibition of LPSinduced inflammation. The development of lung damage in pulmonary sepsis is known to be critically dependent on the initiation of an inflammatory response, mainly characterized by neutrophil transmigration and activation and pro-inflammatory cytokine release [21,25]. Neutrophil sequestration into the bronchoalveolar space was markedly increased by LPS application as compared to the control ( Figure 3A). H 2 S administration in LPS treated mice substantially reduced neutrophil numbers to control levels. Our data are in line with recent publications, where pulmonary neutrophil activity was found to be largely decreased in mice or rats, that had been protected from LPS-induced systemic inflammation both by inhaled H 2 S [23] as well as by application of the slowly releasing H 2 S donors S-diclofenac and GYY4137 [25,26]. The findings of these trials clearly support the notion that exogenous H 2 S can inhibit pro-inflammatory processes. In combination with neutrophil transmigration, the release of proinflammatory cytokines, e.g., IL-1β and MIP-2, is known to aggravate lung injury [21,25]. In our study, quantitative analysis of IL-1β in the BAL revealed that it was nearly absent in both control groups (PBS + air and PBS + H 2 S), whereas LPS instillation alone led to a vast increase of IL-1β readings ( Figure 3B). In sharp contrast, H 2 S inhalation reduced IL-1β to control levels. Likewise, LPS treatment increased MIP-2 protein that was partially prevented by H 2 S ( Figure 3C). The attenuation of neutrophil transmigration and pro-inflammatory cytokine release by administration of H 2 S has also been shown in other models of ALI, e.g., ventilator-induced lung injury [7], oleic acidinduced lung injury [9,10], caerulein-induced acute pancreatitis [27], or myocardial ischemia/reperfusion injury [28], strongly supporting our findings that gaseous H 2 S substantially inhibits pulmonary inflammation and thereby limits LPS-induced lung damage. Future studies using different ALI models might discover the regulatory role of H 2 S in each single aspect of human ALI / ARDS in order to define H 2 S`therapeutic potential.
We finally questioned whether intranasal application of LPS would also induce systemic inflammation and whether H 2 S would exert any systemic effects. As a marker for neutrophil activity we determined the concentration of the MPO glycoprotein in serum [23,26]. In the present study, LPS treatment clearly increased serum MPO release. H 2 S inhalation tended to reduce MPO levels, irrespective of the mode of additional PBS-or LPS-treatment ( Figure 4A). Similar results were obtained by analyzing serum IL-1β levels ( Figure 4B). Our findings compliment the results of two previous studies, where LPS-induced systemic inflammation (e.g., plasma nitrite/nitrate levels, proinflammatory cytokine release) was also clearly prevented by H 2 S inhalation [23,29]. However, on the basis of our results, we cannot clearly decipher, whether i.n. LPS directly induced a systemic inflammation or whether the inflammatory response observed resulted from lung injury. Therefore two scenarios concerning the role of H 2 S appear possible: (1) H 2 S inhalation may directly inhibit lung and systemic inflammation, or (2) H 2 S inhalation may primarily inhibit lung inflammatory responses, secondary preventing a systemic inflammation. Either way, our data suggest that H 2 S inhalation has the potential to inhibit both local as well as systemic inflammatory responses to septic insults.

Conclusions
In our model, inhalation of hydrogen sulfide substantially reduced LPS-induced acute lung injury. The observed protection appears to be mediated by the anti-inflammatory effects of H 2 S, i.e., inhibition of neutrophil transmigration and pro-inflammatory cytokine release. Therefore, H 2 S application displays organ protective properties.