H2 Gas Protects Against Intestinal Injury in Severe SepsisScientific Research

INTRODUCTION

Sepsis is a systemic inflammatory response to infection caused by the over release of inflammatory cytokines and an imbalance between pro-inflammatory and anti-inflammatory responses ⁽¹⁾. Severe sepsis accompanied by septic shock, which is characterized by the development of multiple-organ dysfunction syndrome (MODS) and/or tissue hypoperfusion, is the 10th leading cause of death in intensive care units in the United States ⁽²⁾. Unfortunately, there are no effective therapies for this disease to date. The small intestine, which plays an important role in the pathogenesis of sepsis, has been described as the “motor” of multiple organ failure, as it plays a driving role in sepsis ⁽³⁾. In addition, Moore et al. ⁽⁴⁾ demonstrated that severe sepsis or shock can induce intestinal reperfusion, which subsequently leads to the production of additional pro-inflammatory mediators such as interlukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), and high-mobility group box 1 protein (HMGB-1), resulting in amplification of the systemic inflammatory response syndrome ⁽³⁾. This response is often followed by systemic inflammation, organ dysfunction, and organ failure.

Hydrogen gas (H₂) is a selective antioxidant that has been extensively used as a curative therapy for more than 70 types of diseases ⁽⁴⁾. Our group previously reported that consumption of H₂ either by inhalation or by drinking H₂-containing water has protective and anti-inflammatory effects in many disorders such as sepsis, MODS, stroke and diabetic peripheral neuropathy via anti-oxidative stress, anti-inflammatory, and anti-apoptotic mechanisms ^(5–8). In addition, we also showed that H₂ plays a role in protecting against sepsis-induced intestinal injury both in vivo and in vitro^{(9, 10)}. However, the specific mechanisms underlying its functions remain unknown.

Nuclear factor-erythroid 2-related factor 2 (Nrf2), a transcription factor belonging to the CNC “cap “n” collar” (CNC) subfamily of basic region-leucine zipper transcription factors ⁽¹¹⁾, is a redox-sensitive master switch that can regulate the expression of antioxidant and protective enzymes. Under conditions of oxidative stress, Nrf2 translocates into the nucleus and binds to the cis-acting enhancer antioxidant response element (ARE) sequence to activate its downstream molecules such as heme oxygenase-1 (HO-1), a phase II antioxidant enzyme ⁽¹²⁾. HO-1, which is modulated by Nrf2, is crucial for regulating inflammatory responses and oxidative stress in a variety of diseases ⁽¹³⁾.

We used a cecal ligation and puncture (CLP) model in wild-type (WT) and Nrf2 knockout mice (Nrf2-KO) to investigate if Nrf2 plays a key role in the protective effects of H₂ on severe sepsis-induced intestinal injury and to determine the underlying mechanisms.

MATERIALS AND METHODS

Animals

This protocol has been approved by the Animal Experimental Ethics Committee of Tianjin Medical University General Hospital, Tianjin, China. Male WT and Nrf2-KO Institute for Cancer Research (ICR) mice (6–8 weeks old weighing 20 g–25 g) were purchased from the Better Biotechnology Company (Nanjing, China). The mice were housed in cages (five mice per cage), and maintained in an environmentally controlled room (22°C–25°C) under a regular 12 h/12 h light/dark cycle with food and water ad libitum.

CLP model

Severe sepsis models of CLP and CLP+H₂ were induced by CLP, as previously described ⁽¹⁴⁾. In brief, mice were deeply anesthetized by intraperitoneal injection of 2% pentobarbital sodium solution (50 mg/kg). Then the cecum was exposed via a 1-cm-long abdominal median incision and the ileocecal valve was ligated on the distal three-quarters of the cecum, which was punctured by a 20-gauge needle in a sterile surgical environment. A small amount of feces was squeezed from the punctured hole, and the cecum was returned into the abdomen before closing the incision with a sterile 3-0 silk suture. The mice in the sham and sham+H₂ groups only underwent laparotomies without ligation and puncture. All of the mice were given a hypodermic injection of 1 mL saline solution immediately after the operation.

H₂ treatment

As previously described ⁽¹⁵⁾, the animals in the sham + H₂ and CLP + H₂ groups were placed in a sealed plastic box with inflow and outflow outlets. H₂ gas (mixed with air) was supplied by a TF-1 gas flowmeter (YUTAKA Engineering Corp, Tokyo, Japan) and delivered through a tube at a rate of 4 L/min. The concentration of H₂ in the box was continuously monitored by a HY-ALERTA handheld detector (Model 500; H₂ Scan, Valencia, Calif) and was maintained at 2% during the treatment period, in which 2% H₂ was inhaled for 60 min each time, both 1 and 6 h after the CLP or sham operations. Animals in the sham and CLP groups were placed in the same box as that used for the sham + H₂ and CLP + H₂ groups; those animals inhaled air instead of H₂.

Reagents

Enzyme-linked immunosorbent assay (ELISA) kits for TNF-α, IL-6, and IL-10 were obtained from R&D Systems Inc (Minneapolis, Minn) and the HMGB1 ELISA kit was purchased from IBL (Hamburg, Germany). The commercial kits used to measure catalase (CAT), superoxide dismutase (SOD), and malondialdehyde (MDA), and 8-iso-PGF2α levels were purchased from Cayman Chemical Company (Ann Arbor, Mich). The RNAprep pure tissue kit was obtained from TIANGEN (Beijing, China), the All-in-One First-Strand cDNA Synthesis Kit was purchased from GeneCopoeia Inc (Rockville, Md), and the RealMasterMix (SYBR Green) was obtained from TIANGEN. The rabbit anti-HO-1 and rabbit anti-HMGB1 antibodies were purchased from Abcam (Cambridge, UK), the mouse anti-β-actin antibody was from Sigma-Aldrich (St. Louis, Mo), the goat-antirabbit and goat-antimouse antibodies were obtained from Sigma-Aldrich, and the goat-antirabbit IgG/FITC was purchased from Beijing Zhong Shan Golden Bridge Bio-Technology Co (Beijing, China). The BCA Protein Assay Kit was purchased from Well Biochemical (Nanjing, China).

Experimental protocols

The goal of Experiment 1 was to determine the effects of 2% H₂ treatment on the 7-day survival rate of WT, and Nrf2-KO, septic mice. Male WT mice and Nrf2-KO mice were randomly assigned into four groups for WT mice and four groups for Nrf2-KO mice (n = 20 per group): sham, sham+H₂, CLP, and CLP+H₂. All of the mice were fed with a standard laboratory diet and water ad libitum, and the survival rate in each group was observed within 7 days after the operation (sham or CLP).

The goal of Experiment 2 was to determine the effects of 2% H₂ treatment on CLP-induced intestinal injury in WT mice and Nrf2-KO mice. WT and Nrf2-KO mice were randomly divided into the aforementioned four groups (n = 6 per group), and animals were sacrificed 24 h after the CLP or sham operation. The peritoneal lavage fluid (PLF) from each mouse was collected to measure the colony-forming units (CFUs), which allows determination of the bacterial load. Moreover, the pro-inflammatory cytokines (TNF-α, IL-6, and HMGB1), anti-inflammatory cytokine (IL-10), antioxidant enzymes (SOD and CAT), and oxidative products (MDA and 8-iso-Prostaglandin F2α [8-iso-PGF2α]) were detected in the serum and intestinal tissue (which was harvested 2 cm from the ileocecal valve) at 24 h after surgery.

The goal of Experiment 3 was to investigate the effects of 2% H₂ treatment on regulation of the Nrf2/HO-1-HMGB1 axis in Nrf2-KO and WT mice. WT and Nrf2-KO mice were randomly divided into aforementioned four groups (n = 6 per group). HO-1 and HMGB1 were detected at 24 h after the CLP or sham operation by Western blotting and quantitative PCR (qPCR). Moreover, changes in HO-1 and HMGB1 levels were observed by immunohistochemistry and immunofluorescence, respectively.

Determination of CFU

Mice were sacrificed at 24 h after the CLP or sham surgery. In short, 12 μL PLF from each mouse was obtained and gradient diluted with sterile saline (1:10 dilution) on ice, and 12 μL of the dilution from each mouse was placed on agar plates and incubated overnight at 37°C. Then the number of colonies was counted and data are shown as CFU/mL, similar to a previous study ⁽¹⁶⁾.

Detection of inflammatory cytokines

ELISA kits were used to measure the levels of TNF-α, IL-6, IL-10 (R&D Systems), and HMGB1 (IBL) in the serum and intestinal tissue 24 h after the CLP and sham surgery using a microplate reader (CA 94089, Molecular Devices, Sunnyvale, Calif), according to the manufacturer’s instructions with minor modifications.

Detection of antioxidant enzymes activities

The serum and intestinal tissue were obtained to evaluate the activities of antioxidant enzymes (CAT and SOD) using commercial kits (Cayman Chemical Company) according to the manufacturer’s instructions. The protein concentration of the intestinal tissue was measured using a standard commercial kit (Bio-Rad Laboratories, Hercules, Calif).

Detection of oxidative products

MDA and 8-iso-PGF2α (two types of oxidative products) in the serum and intestinal tissue were determined with commercial kits (Cayman Chemical Company), according to the manufacturer’s instructions.

Western blotting

The protein of the intestinal tissue was extracted at 24 h after the CLP or sham operation and was quantified using the BCA Protein Assay Kit (Well Biochemical). Equal amounts (50 μg) of protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and were transferred onto polyvinylidene fluoride membranes (EMD Millipore, Billerica, Mass). The membranes were blocked in 5% fat-free milk for 1 h before being incubated at 4°C overnight with mildly consistent agitation with the following primary antibodies: HO-1 (1:1,000 dilution), HMGB1 (1:1,000 dilution), and β-actin (1:2,000 dilution). The membranes were washed three times (5 min for each time) with Tris-buffered saline/0.05% Tween and were incubated in specific secondary antibodies for 1 h. Then the membranes were washed again and treated with a prepared chemiluminescent horseradish peroxidase substrate (EMD Millipore). The blots were visualized by Quantity One software, version 4.5.2 (Bio-Rad) and the integrated optical density was analyzed using a Gel-Pro analyzer (Media Cybernetic Inc, Rockville, Md); 100% of protein level changes refer to control levels (sham group) for the purpose of comparison to experimental conditions.

qPCR

Total RNA from each group was extracted with an RNAprep pure tissue kit, and the mRNA was quantified with a NanoDrop 2000. A total of 500 ng mRNA in each group was reverse-transcribed into cDNA by the All-in-one First-Strand cDNA synthesis kit (Cat. No. AORT-0050, GeneCopoeia), and PCR amplification was performed with RealMasterMix (SYBR Green) using the iQ^TM5 system. Relative mRNA levels were calculated by the 2^−ΔΔCt method. The primer sequences of different genes, which have been used in our previous studies, are as follows: HO-1, Forward: 5′-ACAGATGGCGTCACTTCG-3′, Reverse: 5′-TGAGGACCCACTGGAGGA-3′; HMGB1, Forward: 5′-TCACAGCCATTGCAGTAC-3′, Reverse: 5′-ATAACGAGCCTTGTCAGC-3′; GAPDH, Forward: 5′-CATCACTGCCACCCAGAAGAC-3′, Reverse: 5′-CCAGTGAGCTTCCCGTTCAG-3′ ⁽⁹⁾.

Immunohistochemical staining

The intestinal tissues from different groups were fixed in 4% paraformaldehyde for more than 6 h, embedded in paraffin, and then cut into 5-μm slices for immunohistochemical staining. After dewaxing, hydration, and antigen retrieval, the sections were blocked in 5% goat serum and incubated overnight at 4°C with anti-HO-1 rabbit monoclonal antibody (1:200). The next day, all of the slices were washed with PBS three times (5 min for each time), incubated with secondary antibody (1:200) at room temperature for 2 h, and stained with diaminobenzidine. Hematoxylin was used to stain the nuclei for 5 min. HO-1 was observed using an Olympus eclipse 80i microscope (Olympus, Tokyo, Japan) and the images were analyzed using the Image Pro Plus 6.1 system.

Immunofluorescence staining

The frozen slides of the intestinal tissue were fixed in 4% paraformaldehyde for 30 min and incubated with 0.3% Triton (Solarbio, Beijing, China) for 15 min. After washing thrre times (5 min for each time) with PBS, all of the slides were blocked in 5% goat serum for 1 h, followed by incubation with anti-HMGB1 primary antibody (1:400) at 4°C overnight. The next day, the slides were incubated with goat antirabbit-IgG/FITC secondary antibody (1:200) at room temperature for 1 h in the dark. Then the samples were stained with 4′,6′-diamidino-2-phenylindole for 5 min and washed three times (5 min for each time) with PBS. The intracellular expression of HMGB1 was visualized under a fluorescence microscope (Leica, Wetzlar, Germany) and the images were analyzed with the Image Pro Plus 6.1 system.

Statistical analysis

The survival rates are expressed as percentage (%), and the other data are reported as means ± standard deviation (SDs). Long-rank (Mantel–Cox) Test was used to analyze the difference of survival rates among groups, and unpaired t test (if the values were approximately normally distributed [Gaussian distribution]) or Mann–Whitney test (if the values were not normally distributed) were used to analyze the differences between the sham and CLP groups, and between the CLP+H₂ and CLP groups. Two-way analysis of variance was used to analyze the interaction among all of the groups. P values less than 0.05 were considered statistically significant, and the significance test was two-tailed. Statistical analyses were conducted using the GraphPad Prism software (version 5.0) and SPSS statistical software (version 21.0).

RESULTS

Survival rate

The 7-day survival rates of WT and Nrf2-KO mice were 100% in the sham and sham+H₂ groups, but were 0% in the CLP group. However, inhalation of 2% H₂ improved the survival rate in WT mice with sepsis (P < 0.05 vs. CLP group in WT mice), but not in Nrf2-KO mice (P > 0.05 vs. CLP group in Nrf2-KO mice) with sepsis (Fig. 1).

Bacterial load in the PLF

As shown in Fig. 2, CFUs (which can reflect bacterial load) in the PLF at 24 h after the sham or CLP operation were dramatically increased in the CLP group in both WT and Nrf2-KO mice (P < 0.05 vs. sham group). Treatment with 2% H₂ significantly increased bacterial clearance (decreased CFUs) in WT mice with severe sepsis (P < 0.05 vs. CLP group in WT mice), but had no statistical effects in septic Nrf2-KO mice (P > 0.05 vs. CLP group in Nrf2-KO mice).

Inflammatory cytokines in serum and intestinal tissue

In the CLP group, the levels of early pro-inflammatory cytokines (TNF-α and IL-6) and late pro-inflammatory cytokines (HMGB1) in the serum and intestinal tissue were markedly increased compared with levels in the sham group in both WT and Nrf2-KO mice (P < 0.05 vs. sham group). However, 2% H₂ inhalation decreased early and late pro-inflammatory cytokine expression in CLP-induced WT mice (P < 0.05 vs. CLP group in WT mice), but not in Nrf2-KO septic mice (P > 0.05 vs. CLP group in Nrf2-KO mice). We also observed the significant elevation of anti-inflammatory cytokine (IL-10) in the serum and intestine tissue in the CLP and CLP+H₂ groups compared with the sham group in WT mice (P < 0.05 vs. sham group in WT mice), but not in Nrf2-KO mice (P > 0.05 vs. sham group in Nrf2-KO mice). In addition, the CLP+H₂ group had more IL-10 release (in serum and intestinal tissue) than the CLP group in WT mice (P < 0.05 vs. CLP group in WT mice). However, there were no large differences in IL-10 expression between the CLP and CLP+H₂ groups in Nrf2-KO mice (P > 0.05 vs. CLP group in Nrf2-KO mice) (Fig. 3).

Oxidation and anti-oxidation in serum and intestinal tissue

In the CLP and CLP+H₂ groups of WT mice, activities of the anti-oxidant enzymes SOD and CAT were diminished, but levels of oxidative products MDA and 8-iso-PGF2α were markedly increased in the serum and intestinal tissue compared with the sham groups (P < 0.05 vs. sham group in WT mice). Furthermore, 2% H₂ treatment increased the activities of SOD and CAT, but reduced the levels of MDA and iso-PGF2 in the serum and intestine of WT mice with severe sepsis (P < 0.05 vs. CLP group in WT mice). However, there were no statistical differences in anti-oxidant enzymes and oxidative products between the CLP and CLP+H₂ groups in the serum and intestine of Nrf2-KO mice (P > 0.05 vs. CLP group in Nrf2-KO mice) (Fig. 4).

HO-1 and HMGB1 expression

At 24 h after the CLP or sham operation, the mRNA and protein expression levels of HO-1 increased in the CLP+H₂ group compared with the CLP group in WT mice (P < 0.05 vs. CLP group in WT mice). Immunohistochemical staining showed the same results, namely, that under conditions of severe sepsis, 2% H₂ treatment increased the level of HO-1 in WT mice after the CLP operation (P < 0.05 vs. CLP group in WT mice). Figure 5D pathology shows alterations in the microvili structure and underlying lamina propria in the Nrf2-KO sham mice, and Nrf2-KO mice have a tendency to develop age-dependent autoimmune and inflammatory lesions in multiple tissues ^{(17, 18)}. However, as shown in Fig. 5, in Nrf2-KO mice, there were no statistical differences in HO-1 expression between the CLP+H₂ and CLP groups based on the qPCR, Western blotting, and immunohistochemistry data (P > 0.05 vs. CLP in Nrf2-KO mice).

In WT mice, there was a significant difference in expression between the CLP+H₂ and CLP groups (P < 0.05 vs. CLP group in WT mice). However, this was not the case in Nrf2-KO mice with regard to the CLP+H₂ and CLP groups (P > 0.05 vs. CLP group in Nrf2-KO mice) (Fig. 6).

DISCUSSION

CLP is a well-accepted and widely used clinically relevant model for sepsis studies ⁽¹⁹⁾, and low concentrations (2%) of H₂ gas inhalation reportedly protects against multiple diseases, especially sepsis ⁽⁸⁾. In this study, we not only demonstrated the anti-oxidative stress and anti-inflammatory effects of H₂ on intestinal injuries induced by severe sepsis (the 7-day mortality rate was 100%), but also show the key role of Nrf2 and its downstream molecules in these effects.

We successfully developed severe sepsis models and found that mice in the CLP group had a 0% survival rate and higher bacterial load compared with those in the sham group of both WT mice and Nrf2-KO mice. However, treatment with 2% H₂ for 60 min at 1 and 6 h after the CLP operation increased the survival rate and alleviate the bacterial loads in WT mice, but not in Nrf2-KO mice, compared with the CLP group. The excessive release of inflammatory cytokines and abundance of oxidative stress underlie the pathogeneses of sepsis ⁽²⁰⁾. To observe the effects of H₂ on intestinal injury induced by severe sepsis, we examined the levels of inflammatory factors, including pro-inflammatory cytokines TNF-α and IL-6 (two types of “early” pro-inflammatory cytokines in the pathogenesis of sepsis) and HMGB1 (a type of “late” pro-inflammatory cytokines), which play an important role in intestinal injury caused by severe sepsis ⁽⁹⁾ as well as the anti-inflammatory cytokines IL-10. Moreover, we also detected the levels of some oxidative stress factors, such as oxidative products (MDA, 8-iso-PGF2α) and antioxidant enzymes SOD and CAT, which can convert different kinds of reactive oxygen species (ROS) from one type to another. Similarly, compared with the CLP group, inhalation of 2% H₂ decreased levels of TNF-α, IL-6, HMGB1, MDA, and 8-iso-PGF2α, but increased levels of IL-10, SOD, and CAT in both the serum and intestinal tissue of WT, but not Nrf2-KO mice. These data demonstrate that 2% H₂ had no therapeutic effects on septic mice lacking Nrf2.

The Nrf2 transcription factor plays a crucial role in regulating an appropriate innate immune response and mediating antioxidant and anti-inflammatory responses ⁽²¹⁾. After severe septic injury, the intestinal barrier can break down, leading to a state of systemic inflammation, and oxidative stress ⁽²²⁾. However, at this time, Nrf2 can be activated and translocates from the cytoplasm into the nucleus, where it binds to the ARE gene to control the expression and activation of antioxidants, such as SOD and CAT, and phase II genes including glutathione S-transferase, NAD(P)H quinine oxidoreductase, and HO-1 ⁽⁶⁾. HO-1, which is downstream of Nrf2, can degrade the pro-inflammatory heme and produce the anti-inflammatory compounds CO and bilirubin, which can downregulate the inflammation ⁽²³⁾. Our group previously showed that H₂ (2% H₂ gas inhalation or 0.6 mmol/L H₂-rich medium) was able to significantly increase the levels of Nrf2 and HO-1 in the septic organs of mice (e.g., lung, intestine, brain, kidney) and in septic cell models (e.g., HUVEC, CACO2) ^{(15, 20, 24, 25)}. However, 2% hydrogen gas treatment did not increase the survival rate in Nrf2-deficient mice ⁽²⁰⁾.

In this study, we used qPCR, Western blot analysis, and immunohistochemistry to examine the expression of HO-1 in WT and Nrf2-KO mice. In WT mice, the data was in accordance with what we previously reported, in that the CLP+H₂ group expressed more HO-1 than the CLP group. However, in Nrf2-KO mice, a small amount of HO-1 was expressed in all four groups (sham, sham+H₂, CLP, and CLP+H₂), and there were no differences in expression between the CLP and CLP+H₂ groups. With regard to HMGB1, which is a type of “late” pro-inflammatory mediator and a diagnostic biomarker of bacteremia or severe sepsis ⁽²⁶⁾, we found that in WT mice, the CLP+H₂ group had lower expression levels of HMGB1 than the CLP group. However, in Nrf2-KO mice, even though there was a greater amount of HMGB1 mRNA and protein expression in the CLP group, 2% H₂ treatment had no positive effect in the CLP+H₂ group compared with the CLP group. Together, these data indicate that Nrf2 plays a very important role in protecting against oxidative stress and inflammation by stimulating the expression of HO-1 and HMGB1 via activation of Nrf2.

In conclusion, our current research study demonstrates that inhalation of 2% H₂ may be a promising therapeutic strategy for intestine injuries caused by severe sepsis through the regulation of HO-1 and HMGB1 release. In addition, Nrf2 plays a key role in the protective effects of H₂ against intestinal damages in severe sepsis.

Acknowledgment

The authors thank Letpub for its linguistic assistance during the preparation of this manuscript.