Hydrogen protects lung from hypoxia/re-oxygenation injury by reducing hydroxyl radical production and inhibiting inflammatory responses

Here we investigated whether hydrogen can protect the lung from chronic injury induced by hypoxia/re-oxygenation (H/R). We developed a mouse model in which H/R exposure triggered clinically typical lung injury, involving increased alveolar wall thickening, infiltration by neutrophils, consolidation, alveolar hemorrhage, increased levels of inflammatory factors and recruitment of M1 macrophages. All these processes were attenuated in the presence of H2. We found that H/R-induced injury in our mouse model was associated with production of hydroxyl radicals as well as increased levels of colony-stimulating factors and circulating leukocytes. H2 attenuated H/R-induced production of hydroxyl radicals, up-regulation of colony-stimulating factors, and recruitment of neutrophils and M1 macrophages to lung tissues. However, H2 did not substantially affect the H/R-induced increase in erythropoietin or pulmonary artery remodeling. Our results suggest that H2 ameliorates H/R-induced lung injury by inhibiting hydroxyl radical production and inflammation in lungs. It may also prevent colony-stimulating factors from mobilizing progenitors in response to H/R-induced injury.

. Hydrogen inhalation reduced H/R-induced lung injury. Mice were exposed for 4 weeks (8 h per day) to hypoxia, hypoxia with 4% H 2 or normoxia with 4% H 2 and then housed in normoxia ( animals experienced a smaller increase in body weight (5.65 ± 0.22 vs. 2.52 ± 0.30 g, p < 0.001; Fig. 1B). H 2 significantly attenuated H/R-induced infiltration of inflammatory cells and alveolar wall thickening, and it significantly decreased lung injury score (2.11 ± 0.38, p < 0.001 vs. H/R). H 2 also led to a larger increase in body weight in hypoxic mice (3.86 ± 0.39 g, p = 0.019 vs. H/R), although the weight of hypoxic mice was still lower than that of normoxia-treated mice (p = 0.003).
These results suggest that H 2 attenuates lung injury, although it does not substantially affect pulmonary artery remodeling induced by H/R. H 2 inhalation reduces H/R-induced production of hydroxyl radicals. To begin to understand how H 2 inhalation can protect lung from H/R-induced injury, tissue from exposed animals was sectioned and stained for hydroxyl radicals (Fig. 3A), levels of which were then quantified by chromatometry (Fig. 3B). H/R triggered a significant increase in hydroxyl radicals in lung tissue (0.72 ± 0.07 vs. 0.46 ± 0.02 μM/mg, p < 0.001 vs. normoxia), which was markedly lower in the presence of H 2 (0.48 ± 0.02 μM/mg, p < 0.001 vs. H/R, p = 0.79 vs. normoxia).
Since the primary source of inflammatory factors in the lung is thought to be pulmonary M1 macrophages, lung tissue from H/R-exposed animals was stained for M1 macrophages. These cells were abundant in lung tissue from animals exposed only to H/R, but they were rare in tissue from animals exposed to H/R in the presence of H 2 and in tissue from normoxia-treated animals (Fig. 5A,B).
These results suggest that inhaled H 2 inhibits H/R induction of the CSF system but not of the EPO system.

H 2 protects CD133 + progenitors from H/R-induced injury.
Because G-CSF can mobilize progenitors from bone marrow, which home to sites of injury to mount anti-inflammatory and repair responses 32 , we wanted to test whether H/R mobilizes progenitors that home to lung tissue. Therefore, we investigated pulmonary and circulating CD133 + progenitors, which are pluripotential cells that can be mobilized by G-CSF 33 . These cells were abundant in lung tissues after H/R exposure, while they were rare in mice exposed to hypoxia in the presence of H 2 and in normoxia-treated mice (Fig. 7A). Flow cytometry showed significantly higher percentages of pulmonary CD133 + progenitors in H/R-exposed mice (9.73 ± 1.95%) than in normoxia-treated mice (3.03 ± 1.06%, p = 0.007) and mice exposed to H/R in the presence of H 2 (4.83 ± 1.87%, p = 0.024 vs. H/R, p > 0.05 vs. normoxia; Fig. 7B). In contrast, percentages of circulating CD133 + cells were similar among the various animal groups (Fig. 7C).
Next we tested the effects of H/R and H 2 on progenitors. Cultures of mouse CD133 + progenitors were exposed for 8 h at 37 °C to hypoxia (10% O 2 , 5% CO 2 , 85% N 2 ) or hypoxia in the presence of H 2 (10% O 2 , 5% CO 2 , 4% H 2 , 81% N 2 ). Control cultures were exposed to normoxia (5% CO 2 , 95% air) or normoxia in the presence of H 2 (5% CO 2 , 21% O 2 , 4% H 2 , 70% N 2 ). Levels of hydroxyl radicals were higher in cells exposed to hypoxia than in cells treated with normoxia, and most of the radicals colocalized with DAPI in the nucleus (Fig. 7D). H 2 reduced the levels of hydroxyl radicals, especially those in the nucleus.
Our results suggest that H/R induces the homing of CD133 + progenitors to lung tissue, presumably to repair damage. H 2 helps protect CD133 + progenitors from H/R injury by inactivating hydroxyl radicals.

Discussion
In this study, we found that mice exposed to chronic H/R exhibited significant lung injury, which was significantly improved by 4% H 2 inhalation. H 2 treatment inhibited the generation of hydroxyl radicals and down-regulated GM-CSF and G-CSF, which may attenuate infiltration by neutrophils and M1 macrophages, as well as release of proinflammatory factors. H 2 may also protect the progenitor cells by inactivating hydroxyl radicals (Fig. 8). Our results demonstrate that molecular hydrogen is effective at protecting lung from H/R-induced injury.
Our results showed that H/R triggered massive generation of hydroxyl radicals, similar to previous reports 34 , and that H 2 inhalation reduced these levels, likely reflecting the established ability of H 2 to inactivate hydroxyl radicals 22,35 . The decrease in hydroxyl radicals was associated with reduction of GM-CSF and G-CSF, followed by down-regulation of systemic and pulmonary inflammatory responses. These results suggest that H/R-induced production of hydroxyl radicals is important for up-regulating CSFs, which can be abolished by H 2 inhalation.
CSFs are important mediators of both systemic and tissue inflammation, but their role in H/R-induced lung injury has been unclear. GM-CSF and G-CSF can be secreted from various cell types, including epithelial cells, fibroblasts 36 , T cells, B cells 37 , neutrophils and macrophages 38 under the stimulation of IL-1β and TNF-α 12,15 . Studies have shown that CSFs, especially GM-CSF, aggravate inflammation 39 by enhancing proinflammatory cytokine production 40 and mobilizing leukocytes, promoting their survival, proliferation 41 , differentiation 42 , and stimulating their activation [43][44][45] and migration. Recent studies show that GM-CSF blockade has a therapeutic effect against cardiac inflammation during Kawasaki disease or aortic aneurysm formation 46 , as well as against lung diseases such as COPD 47 , interstitial lung disease 48 , allergy 49 and asthma 50 . Furthermore, GM-CSF promotes macrophage polarization toward M1-skewed cells 51 and stimulates macrophage plasminogen activator activity 52 , which may induce up-regulation of pro-inflammatory factors. Both M1 macrophages and neutrophils may aggravate lung and systemic inflammation by releasing proinflammatory factors and elastase. In the present study, we found that H/R increased serum levels of both GM-CSF and G-CSF, which was followed by increases in leukocytes and pro-inflammatory factors in circulation and lung tissues. In addition, M1 macrophages accumulated in the lung tissues. These findings suggest that CSFs are key mediators of H/R-induced lung injury.
G-CSF not only promotes inflammation, but also mobilizes stem cells derived from bone marrow for repair 53 . To assess the role of these cells in H/R-induced lung injury, we analyzed CD133 + cells, a homogeneous population with multi-proliferative potential previously shown to play a role in lung repair 54,55 . We found that H/R triggered accumulation of CD133 + progenitors, although this did not seem to prevent lung tissue from H/R injury. In our experiments with CD133 + progenitor cultures, hypoxia significantly increased hydroxyl radicals. It is possible that these radicals cause oxidative injury in the progenitors, damaging their ability to repair tissue. This implies that H 2 improves progenitor function by inactivating hydroxyl radicals in those cells. This may be another reason for the protective effects of H 2 .
In our experiments, chronic H/R increased the level of EPO in serum, which increases red blood cell count and blood viscosity, thereby aiding pulmonary artery remodeling in response to chronic H/R 56 . However, H 2 inhalation in our mouse model of H/R-induced injury did not influence the EPO-red blood cell system, nor did it significantly affect pulmonary artery remodeling or right ventricular hypertrophy. These results suggest hydroxyl radicals activate the CSF system but not the EPO-red blood cell system.
In conclusion, our study shows that H 2 protects lung from H/R-induced injury, at least in part by scavenging hydroxyl radicals, thereby inhibiting lung inflammation induced by G-CSF and GM-CSF. This helps protect CD133 + progenitors from oxidative damage. Because inhalation of 1-4% H 2 appears to show no cytotoxicity 35 , consistent with what we found in this study, our results justify further work toward developing H 2 as a treatment against H/R-induced lung injury, such as during acute exacerbation and remission of asthma, bronchiectasis and early COPD.

Materials and Methods
Animal. Male C57BL/6 mice 8 weeks old were provided by the Sichuan Provincial Experimental Animal Center. The H/R model was established as follows. Mice were placed in a closed container with inlet and outlet ports on the two sides. Sodium lime was placed at the bottom of the container to absorb CO 2 . Animals were Scientific REPORTS | (2018) 8:8004 | DOI:10.1038/s41598-018-26335-2 exposed to hypoxia (10% O 2 , 90% N 2 ), hypoxia in the presence of H 2 (10% O 2 , 4% H 2 , 86% N 2 ), or normoxia (21% O 2 , 79% N 2 ) in the presence of H 2 (21% O 2 , 4% H 2 , 75% N 2 ) (8 h per day). Then they were exposed to air for re-oxygenation (16 h per day) for 4 weeks. Control mice were exposed to normoxia (21% O 2 , 79% N 2 ) for 4 weeks. Mice had free access to water and conventional laboratory diet throughout the exposure.
Hydrogen (purity >99.9%) was generated using a hydrogen generator (Beijing Zhongxing Huili Technology Development, Beijing, China). During experiments, the O 2 and CO 2 concentrations were monitored using a Philips Airway Gases monitor, and H 2 concentration was monitored using a STP1000 Multiplexed Gas Analyzer (T & P Union (Beijing), Beijing, China).
All animal experimental protocols were approved by the Ethics Committee for Animal Experiments of Sichuan University and were performed in accordance with the Guide for the Care and Use of Laboratory Animals prepared by the Institutional Animal Care and Use Committee of Sichuan University.
Blood cell count. Blood samples were taken from the apical artery, and blood cells were counted using an automatic blood analyzer (Sysmex-XE5000, Toagosei Co., Ltd, Yokohama, Japan).
Lung histological injury. The lung was harvested and fixed overnight in 4% paraformaldehyde at 4 °C.
Paraffin-embedded sections were stained with hematoxylin and eosin, and examined under a light microscope by pathologist blinded to the experimental groups. Severity of lung injury was scored using a 5-point scale 57 (0 = normal histology, 5 = most severe injury) that took into account the parameters of alveolar congestion, hemorrhage, neutrophil accumulation in the airspace or vessel wall, alveolar wall thickness, and hyaline membrane formation. We chose 10 images for each animal randomly, and parameter values were averaged for the 10 images.
The average values of all parameters were then added together to generate a total lung injury score. Scoring was performed by two pathologists in a blinded fashion, and the two scores for a given animal were averaged to obtain the final score for that animal.

Measurement of factors.
Blood samples from the apical artery were centrifuged at 3000 rpm for 10 min at 4 °C to obtain serum. To obtain lung tissues, lung was harvested on ice immediately after euthanasia and weighed. An aliquot of tissue (50 mg) was cut into 1-mm 3 pieces, added to 500 μl PBS and homogenized. Samples were left on ice for 5 min and centrifuged at 3500 g for 20 min at 4 °C. Levels of EPO, G-CSF, GM-CSF, IL-1β and TNF-α in serum and lung tissue supernatant were assayed using commercial ELISA kits (Neobioscience Technology, Beijing, China).

Evaluation of vascular remodeling and right ventricle hypertrophy. Pulmonary artery remodeling
was assessed in terms of percent medial thickness, which was calculated according to the equation: (medial wall thickness × 2)/vessel diameter ×100% 58 . Only vessels with a circular appearance and external diameter between 50 and 100 μm were used. Lung sections were examined by an investigator blinded to experimental treatment using an Olympus-BHS microscope.
Right ventricular hypertrophy was quantified by calculating the ratio of right ventricle to left ventricle plus septum weight [RV/(LV + S)]. The ventricles and septum were collected, and the wet and dry ventricle and septal weights were obtained by drying for 24 h at 60 °C.
Hydroxyl radical stains and measurements. Frozen sections (5 μm) of the right lung were used.
To quantify hydroxyl radicals, lung tissue (100 mg) was ground up in liquid nitrogen, added to 500 ul PBS and centrifuged at 10 000 g for 20 min at 4 °C. Supernatants were transferred to Eppendorf tubes and analyzed as quickly as possible using a commercial colorimetric kit (Genmed, Scientifics, USA).
Immunofluorescence. week H/R, the right lung was harvested and cut into frozen sections (5 μm), which were fixed with 4% paraformaldehyde, washed with PBS, and blocked for 1 h at room temperature with 1% bovine serum albumin (Sigma, USA) in PBS.
For detecting M1 macrophage in lung, sections were incubated at 4 °C overnight with rabbit anti-mouse CD86 antibody (1:100; Genetex, USA) and rat anti-mouse MOMA-2 antibody (1:50; Genetex, USA). Then the sections were washed in PBS and incubated for 1 h at room temperature with secondary antibodies conjugated to Alexa Fluor 488 or 555 (1:200; Biofroxx, Germany). As a negative control, sections were incubated with PBS instead of primary antibody. Sections were stained with DAPI (Invitrogen) to label nuclei. Images were then visualized using a fluorescence microscope (LSM 510 Meta, Carl Zeiss) equipped with a 20×/0.75D objective.
Flow cytometric analysis of CD133 + progenitors. Blood was diluted 1:1 with Hank's solution, layered onto the top of a mononuclear cell separation solution, and centrifuged at 1500 rpm for 35 min at 4 °C. The white mononuclear cell loop was obtained, washed with PBS, and incubated at 37 °C for 1 h with anti-CD133 antibody conjugated to AF555 (Bioss). Then cells were washed twice, re-suspended in 500 μl PBS, and analyzed by flow cytometry on an Esp Elite device (Beckman Coulter, Chicago, IL, USA). PBS served as a negative control.
For determining pulmonary CD133 + cells, harvested lung tissues were placed in 2 ml PBS and 1 mg/ml Liberase TM (Thermo Fisher Scientific, USA). Tissues were minced with scissors, and digested for 30 min at 37 °C before filtration through a 70-μm cell strainer and red blood cell lysis. Samples were then filtered through a 40-μm filter and resuspended, after which CD133 + cells were analyzed as described above.
Cell culture. Bone marrow was collected as described 59 by flushing the femurs and tibias of 2-month-old C57BL/6 mice with complete DMEM-LG medium (Thermo Fisher Scientific, USA). Cells were cultured for 24 h in a Petri dish (Thermo Fisher Scientific), then non-adherent cells were removed by washing with PBS. Adherent cells were further cultured in complete medium and retrieved by trypsinization with 0.25% trypsin (Thermo Fisher Scientific) for 5 min at 37 °C. Treated adherent cells were cultured and passaged three times. Third-passage CD133 + cells were retrieved using immunomagnetic microbeads, and further cultured in complete medium. Cultured cells were retrieved, and their morphology and ability to differentiate into osteoblasts and adipocytes were examined. CD133 + cells were split into four Petri dishes (2 × 10 5 cells/dish) and cultivated in standard medium composed of MEM alpha (20% fetal bovine serum(FBS) +1% Penicillin-Streptomycin (P/S) (Thermo Fisher Scientific) for two days, and then incubated for 8 h at 37 •C in a hypoxic atmosphere (10% O 2 , 5% CO 2 , 85% N 2 ) or a hypoxic atmosphere containing H 2 (10% O 2 , 5% CO 2 , 4% H 2 , 81% N 2 ). Control cultures were incubated in a normoxic atmosphere (5% CO 2 , 95% air) or a normoxic atmosphere containing H 2 (5% CO 2 , 21% O 2 , 4% H 2 , 70% N 2 ). After the 8 h incubation, hydroxyphenyl fluorescein solution (1:500) and Hoechst (1:1000; Thermo Fisher Scientific) were added to cultures, which were returned to their incubators for another 30 min. Then cultures were washed with PBS, centrifuged under normoxic conditions and analyzed using confocal microscopy. Statistical analysis. Statistical analysis was performed using SPSS 18.0 (IBM, Chicago, USA). Results were reported as mean ± SEM. Differences between more than two groups were assessed using one-way ANOVA. The threshold for significance in all statistical tests was p < 0.05.