Hydrogen gas inhalation protects against cutaneous ischaemia/reperfusion injury in a mouse model of pressure ulcer

Abstract Pressure ulcer formation depends on various factors among which repetitive ischaemia/reperfusion(I/R) injury plays a vital role. Molecular hydrogen (H2) was reported to have protective effects on I/R injuries of various internal organs. In this study, we investigated the effects of H2 inhalation on pressure ulcer and the underlying mechanisms. H2 inhalation significantly reduced wound area, 8‐oxo‐dG level (oxidative DNA damage) and cell apoptosis rates in skin lesions. H2 remarkably decreased ROS accumulation and enhanced antioxidant enzymes activities by up‐regulating expression of Nrf2 and its downstream components in wound tissue and/or H2O2‐treated endothelia. Meanwhile, H2 inhibited the overexpression of MCP‐1, E‐selectin, P‐selectin and ICAM‐1 in oxidant‐induced endothelia and reduced inflammatory cells infiltration and proinflammatory cytokines (TNF‐α, IL‐1, IL‐6 and IL‐8) production in the wound. Furthermore, H2 promoted the expression of pro‐healing factors (IL‐22, TGF‐β, VEGF and IGF1) and inhibited the production of MMP9 in wound tissue in parallel with acceleration of cutaneous collagen synthesis. Taken together, these data indicated that H2 inhalation suppressed the formation of pressure ulcer in a mouse model. Molecular hydrogen has potentials as a novel and alternative therapy for severe pressure ulcer. The therapeutic effects of molecular hydrogen might be related to its antioxidant, anti‐inflammatory, pro‐healing actions.

prominent bony protrusions. After a change of body position, reperfusion of blood to the ischaemic skin initiates a series of harmful events because of a large increase in reactive oxygen species (ROS). 6 Excessive ROS cause direct damage on lipids, proteins and nucleic acids, leading to cell apoptosis and tissue injury. Meanwhile, these free radicals induce the development of inflammatory responses such as endothelial dysfunction, neutrophil and macrophage infiltration, production of proinflammatory cytokines and, thereafter, tissues necrosis. 6,7 Consistently, several studies demonstrated that skin ulcers induced by cutaneous I/R were inhibited by treatment with antioxidants such as vitamin E, melatonin and deferoxamine. [8][9][10] Therefore, antioxidant administration may be a good therapeutic strategy for promoting healing of pressure ulcers.
In the past decade, H 2 as a novel medical gas has gained wide attention. In 2007, Ohsawa and colleagues found that H 2 affords neuroprotection against brain I/R injury by selectively neutralizing hydroxyl radicals and peroxynitrite. 11 Later studies demonstrated similar protective effects of hydrogen on I/R injuries of other organs such as liver, heart and intestines. [12][13][14] The potential mechanisms might be involved in its antioxidant and anti-inflammatory and antiapoptotic properties. 15 These biological effects further enhanced the potential of H 2 in clinical application of various organ system diseases. 15,16 Moreover, molecular hydrogen displayed high safety in vivo even at high concentration and high pressure, which could effectively reach target tissues and cells by gaseous diffusion but has no effect on physiological variables such as pH, oxygen saturation and blood pressure. 17,18 Taken together, we hypothesized that H 2 could act as an effective treatment for pressure ulcer induced by repeated cutaneous I/R injury.
A recent study reported that hydrogen-water intake promoted wound size reduction and early recovery in 22 elderly in-patients with severe pressure ulcer. 19 However, there has been no experimental evidence of the beneficial effects of molecular hydrogen on I/R-induced pressure ulcer using animal models. This study aimed to determine the possible protective effects of H 2 on pressure ulcer and the underlying mechanisms.

| I/R cycles and analysis
The cutaneous I/R model was established according to previously published reports. [20][21][22] Briefly, all mice were anaesthetized, and their backs were shaved and cleaned with 75% ethanol. The dorsal skin was gently pulled up and placed between two round ferrite magnetic plates that had a 12-mm diameter and 5-mm thickness (NeoMag Co, Ichikawa, Japan). A single I/R cycle was initiated with a 12-hour period of magnet placement, and followed by a release or rest period of 12 hours. After three I/R cycles, all of the mice developed two circular ulcers separated by a bridge of normal skin. For analysis, each wound site was digitally photographed after wounding, and wound areas were measured on photographs using Image J (version 1.48, NIH, Bethesda, MD) as previously described. 20,23 To assess the effects of hydrogen gas on wound healing, the mice were housed in a specific airtight device producing air mixture including 2% or 75% H 2 and 21%O 2 for one week (6 hours per day) before the beginning of I/R cycles. H 2 treatment was continuously performed until the wounds completely healed. The hydrogen-producing device was provided by Shanghai Asclepius Meditec Co. Ltd (Shanghai, China).
Wound sites were digitally photographed at various time-points after wounding, and wound areas were measured on the images using ImageJ software version 1.46r (NIH, Bethesda, MD).

| Histological and immunohistochemical examinations
The wounds were harvested with a 5-mm rim of unwounded skin tissue from sacrificed mice. Skin samples were fixed in 10% paraformaldehyde and embedded in paraffin. Sections (

| CCK8 assay
The rate of cell proliferation was detected using CCK8 assay kit (Beyotime, Shanghai, China). After H 2 preincubation and 0.25 mmol/ L H 2 O 2 treatment, 10 4 of HUEVC cells were placed into 96-well plate (200 lL per well) and 20 lL CCK8 was added into each well for 24 hours. The optical density (OD) was read at 450 nm wavelength by ELISA microplate reader (Thermo Multiskan MK3).  (Table S1) were synthesized by Shanghai Sangon Biotech Co., Ltd (Shanghai, China). Relative expression of PCR products was determined using the 2 ÀDDCT method and calculated relative to the control group.

| Western blot
For protein blotting, total proteins were extracted from homogenized skin tissue using as previously described. 24 Twenty micrograms of proteins was loaded into 8% SDS-PAGE and then transferred onto a polyvinylidene fluoride membrane. After blocking and washing, the membranes were incubated with the indicated primary Abs. The membranes were then incubated with horseradish peroxidaselabelled secondary antibody and developed with the ImmobilonTW Western Chemiluminescent HRP Substrate (Millipore, Billerica, MA).
The blots were assessed by Image J software (NIH, Bethesda, MD).
All the antibodies (Abs) and their sources in the study are listed in Table S2.

| ELISA
Supernatants of mouse wound homogenates or cell culture medium were used for ELISA assay for SOD, GPx, CAT (BioVision), IL-1b, TNF-a, IL-6, IL-8 and IL-22 (eBioscience) according to the manufacturers' recommendation. Total protein in the supernatant was detected with a commercial kit (BCA Protein Assay kit; Pierce, Rockford, IL). The data were expressed as anti-oxidative enzyme (unit/ mL) or cytokine (pg/mL)/total protein (mg/mL) for each sample.

| Statistical analysis
Results were expressed as means AE SEM. Unpaired, two-sided Student's t test was performed determine the statistical differences between the sample means using GraphPad Prism 6.0 (La Jolla, CA). P < .05 were considered statistically significant.

R-induced pressure ulcer formation
To assess the preventive effects of H 2 on the development of cutaneous pressure ulcers after I/R injury in vivo, a decubitus ulcer-like mouse model was constructed as previously described. 20,23 In order to reach fully saturation in cutaneous tissue, 2% or 75% H 2 was inhaled by mice in experimental groups for one week (6 hours per day) before the beginning of I/R cycles according to the discovery of Scottish physiologist John Scott Haldane about the human body and the nature of gases. 25 Wound areas in H 2 -treated mice were significantly smaller than those in control mice especially in early stage after I/R cycles ( Figure 1). Inhalation of 75% H 2 significantly shortened the wound closure time in pressure ulcer mice. Furthermore, we also observed cleaner wounds and less scratching behaviour in H 2 -treated mice than in control group. These results suggested that H 2 inhalation especially at high concentration protected the formation of cutaneous ulcers after I/R cycles.

| H 2 inhalation altered the histopathological characteristics of pressure ulcer skin after cutaneous I/R
Haematoxylin and eosin (HE) staining showed that H 2 inhalation reduced inflammatory cell infiltration and tissue necrosis in skin wound caused by I/R cycles ( Figure S1). Compared with the control group, H 2 -treatment groups (especially 75%H 2 ) displayed notable acceleration and enhancement in dermal collagen synthesis of skin wound ( Figure 2). These results suggested that H 2 inhalation alleviated the inflammatory response and promoted the wound healing in pressure ulcer.

| H 2 inhalation alleviated oxidative DNA damage and suppressed apoptosis in skin tissues after cutaneous I/R
We further evaluated the levels of oxidative DNA damage (8-oxo-dG) and cell apoptosis after cutaneous I/R injury. Immunohistochemical (IHC) analysis showed that three groups had similar levels of

| H 2 inhalation reduced ROS accumulation and up-regulated antioxidant enzyme activities in skin tissue after cutaneous I/R
As ROS is essential mediators of reperfusion induced tissue damage, [26][27][28][29] we assessed ROS levels and the activities of major

| H 2 inhalation inhibited cutaneous I/R-induced inflammation
As proinflammatory cytokines are important for tissue damage during I/R injury, the levels of proinflammatory cytokines in the wounded skin were assessed. At day 6 post-I/R, H 2 treatment significantly reduced the relative mRNA and protein expression of TNF-a, IL-1b, IL-6 and IL-8 in wounded tissue, and their expression was lowest in pressure ulcer mice treated with 75% H 2 ( Figure 5A and B). IHC further confirmed that the protein levels of TNF-a, IL-1b, IL-6 and IL-8 in cutaneous I/R injured skin tissue of mice inhaled H2 were substantially lower than those of control mice ( Figure 5C).
Meanwhile, H 2 treatment significantly up-regulated the expression of IL-22, an essential pro-healing cytokine involved in repair events on different models of epithelial regeneration, 30,31 at both mRNA and protein levels and 75% H 2 treated group displayed the highest level of IL-22 ( Figure 5), which was mainly expressed in the dermal layer of perilesional skin tissues in mice treated by H 2 ( Figure 5C).
These results indicated that hydrogen inhalation reversed proinflammatory effects of skin wound environment after cutaneous I/R.

| H 2 inhalation modulated the expression of healing-associated molecules in vivo after cutaneous I/R
The metalloprotease MMP9 and pro-healing factors TGF-b, VEGF and IGF-1 play crucial roles during proliferative and remodelling phases of wound healing. To investigate how H 2 improves wound healing in pressure ulcer mice, its effects on expression of aforemen-  HUVECs showed a 2.5-folds increase in the apoptosis rate, which was markedly reduced by H 2 pretreatment (especially 75%H 2 ) (P < .01) ( Figure S3B and C).  Figure S4A-D). Quantitative PCR assay demonstrated that H 2 significantly up-regulated the expression of NRF2 and its target genes HO-1, AKR1C1 and NQO1 in HUVECs exposed to H 2 O 2 in a concentration-dependent manner ( Figure S4E-H). Together, these results suggested that H 2 might maintain intracellular homoeostasis of oxidation-reduction system in vascular endothelial cells through the activation of Nrf2/ARE pathway.

| H 2 inhibited the expression of several chemokine and adhesion molecules in vascular endothelial cells stimulated by H 2 O 2
Endothelial damage is a critical event in the early phase of inflammatory responses induced by cutaneous I/R, which could drive the up-

| DISCUSSION
In the present study, we first demonstrated the protective effects of H 2 inhalation on pressure ulcers in a murine model. H 2 treatment significantly reduced the wound area in a concentration-dependent F I G U R E 4 Hydrogen activated the NRF2-ARE pathway in wounded tissue. A, The mRNA levels of NRF2, HO-1, NQO1 and AKR1C1 were analysed by quantitative RT-PCR assay. B, Immunoblots of NRF2, HO-1, NQO1, AKR1C1 and b-actin. *P < .05; **P < .01 compared to control manner in the early phase after cutaneous I/R. The improvement in skin ulcer recovery induced by H 2 was paralleled by significant reductions in oxidative DNA damage, cell apoptosis and acceleration of cutaneous collagen synthesis. This result was consistent with a clinical study that hydrogen-water intake (0.8-1.3 ppm, 600 mL per day) via tube-feeding significantly reduced the wound size in 22 hospitalized patients with pressure ulcer. 19 Theoretically, hydrogen administration through respiratory route displays higher blood concentration and more accessibility for cutaneous lesions than those through gastrointestinal absorption. 15 Therefore, H 2 inhalation has a greater potential to be applied in the prevention and treatment of patients with pressure ulcer.  We also determined the mechanism underlying the antioxidant property of H 2 against cutaneous I/R injury, which was closely associated with the activation of Nrf2/ARE pathway. The transcription factor Nrf2 is essential for regulating the adaptive response to exogenous and endogenous oxidative stresses. 33,34 Under moderate oxidative stress, Nrf2 translocates to the nucleus where it binds to ARE and induces the transcription of downstream antioxidant genes.
In our study, H 2 treatment significantly up-regulated the expression of Nrf2 and its downstream targets such as HO-1, NQO1 and AKR1C1 in wound tissue and HUEVCs exposed to H 2 O 2 . Similar phenomena were also observed in previous studies of other organ I/R injury or other inflammatory diseases. [35][36][37][38] Molecular hydrogen could attenuate intestinal injury in wild-type but not NRF2-knockout mice with severe sepsis by regulating HO-1 expression. 37 Therefore, hydrogen might activate the Nrf2/ARE pathway to restore the homoeostasis of cutaneous oxidation-reduction system against I/R injury.
The protective effects of H 2 were also attributed to its potent anti-inflammatory activity. After I/R cycles, plenty of macrophages and neutrophils were accumulated in the treated tissue accompanied by significant production of proinflammatory cytokines, which caused skin injury and necrosis in the wound. 21 In our study, H 2 inhalation attenuated inflammatory cells infiltration and cutaneous F I G U R E 6 Hydrogen modulated the expression of healing-associated molecules in skin wounds. A, mRNA expression of MMP9, TGF-b, VEGF and IGF-1 was analysed by quantitative RT-PCR. B, Immunoblot assessment of MMP9, TGF-b, VEGF and IGF-1 protein levels in wounded skin tissue. *P < .05; **P < .01 compared to control F I G U R E 7 Hydrogen decreased the expression of several molecules associated with leucocyte-endothelium in H 2 O 2treated HUEVCs. Quantitative RT-PCR assay of MCP-1 (A), P-selectin (B), Eselectin (C) and ICAM-1(D) in HUEVCs. **P < .01, *P < .05, compared to the control group; ## P < .01, # P < .05, compared to H 2 O 2 -treated group necrosis in the wound induced by I/R injury. Proinflammatory cytokines such as TNF-a, IL-1, IL-6 and IL-8 were decreased while IL-22 was increased in the wounded skin of H 2 -treated mice. Distinct from proinflammatory cytokines, IL-22 mediates a crosstalk between immune system and cutaneous cells (such as fibroblasts and keratinocytes) and plays a pro-healing role in wound repairment. 30,39 Our data demonstrated that H 2 reversed the excessive inflammatory response induced by cutaneous I/R injury.
The anti-inflammatory property of hydrogen might be partially attributed to its protection on vascular endothelial cells against oxidative stress. During I/R injury, endothelial damage initiated leucocyte adhesion, recruitment and infiltration by up-regulating the expression of chemokines and adhesion molecules. 6,7,32 Our in vitro assay showed that H 2 treatment significantly alleviated the oxidative injury of endothelial cells and reduced the overexpression of MCP-1, E-selectin, P-selectin and ICAM-1 in HUEVCs exposed to H 2 O 2 . MCP-1 is a critical molecule for chemotaxis and activation of macrophage, which is a significant source of proinflammatory cytokines and contributes to I/R injury of skin and other organs. 21,[40][41][42] The adhesion molecules studied here were mainly responsible for leucocyte rolling, localization and adhesion to the endothelium. 43,44 Hence, hydrogen might modulate those molecules associated with leucocyte-endothelium interaction to inhibit subsequent inflammatory reaction caused by cutaneous I/R injury.
In pressure ulcer, the wound sites were constantly in the dynamic pathological alterations of inflammatory injury and tissue repair. Outcome of wound healing process was mainly determined by the presence and concentration of the healing-associated factors such as MMPs, TGF-b, VEGF and IGF1. [45][46][47] Overproduction or high activity of MMP9 and suppressed expression of TGF-b were identified as indicators of poor healing in skin samples of chronic ulcers. 46,48 In our study, H 2 promoted the expression of pro-healing factors (TGF-b, VEGF and IGF1) and inhibited the production of MMP9 in wound tissue of pressure ulcer, accompanied by acceleration of dermal collagen synthesis. Hence, hydrogen had a wound healing promoting effect against cutaneous I/R injury.
Taken together, the present results indicated that hydrogen suppressed the formation of decubitus ulcers by its antioxidant, antiinflammatory, pro-healing activities against cutaneous I/R injury. Similar therapeutic effects of hydrogen were also reported in common senile diseases concomitant with decubitus, such as cerebral or myocardial infarctions, COPD, diabetes, hyperlipaemia, malignant tumours. 11,36,[49][50][51][52][53][54] Furthermore, H 2 had no cytotoxicity in vivo in human body even at a high concentration. 17 Therefore, hydrogen gas has a great potential for preventing and/or treating pressure ulcer.
There were still some limitations in our study. Firstly, we did not detect the cutaneous hydrogen concentration, which might provide direct evidence to support the dosage-dependent protection of hydrogen gas against cutaneous I/R injury. In addition, the exact mechanisms by which hydrogen modulates oxidative stress, inflammation and wound repair in pressure ulcer were still unclear. More experiments are needed to work out these problems before the clinical trial of H 2 in pressure ulcers.

ACKNOWLEDG EMENTS
We extend our sincerest gratitude to the staff of Shanghai Asclepius Meditec Co., Ltd. for their vital technical support in the hydrogenoxygen nebulizer device. We also thank Prof. Zhimin Kang of Shanghai Huikang Hydrogen Medical Research Center for technical assistance in study design.

CONFLI CT OF INTEREST
The authors have no conflicts of interest to report.