H2-Rich Saline and Intestinal Barrier DysfunctionScientific Research

BACKGROUND

The gut is considered to be the motor of multiple organ dysfunction syndrome in critical illness ⁽¹⁾. This hypothesis is supported by the phenomenon of bacterial translocation, which is the term given to the concept by which intestinal bacteria or toxic mediators escape from the gut to remote organs ⁽²⁾. A previous study reported that septic morbidity and mortality were significantly higher in multiple organ failure patients with bacterial translocation than in those without it ⁽³⁾. Thus, the management of bacterial translocation is an important strategy.

Although the mechanism behind bacterial translocation is undoubtedly complex, intestinal microbiome and epithelial integrity seem to play key roles. Under stress conditions, excessive production of reactive oxygen species, inducible nitric oxide synthase (iNOS), and inflammatory cytokines induces the disruption of microbial composition, which is known as dysbiosis ⁽⁴⁾. These inflammatory responses also lead to barrier dysfunction and hyperpermeability and subsequently cause the passage of potentially pathogenic bacteria ⁽⁵⁾. On the basis of this mechanism, several ways to modulate the intraluminal environment to prevent bacterial translocation, for example, with selective digestive decontamination ⁽⁶⁾ or probiotics ⁽⁷⁾, have been evaluated. However, their benefit in the clinical situation remains controversial.

Over the past few years, many researchers have shown an interest in molecular hydrogen as a novel therapeutic application in various diseases ⁽⁸⁾. Anaerobic bacteria in the body produce a large amount of hydrogen, which has been regarded as a physiological inert gas because of its low solubility. Since Ohsawa et al. ⁽⁹⁾ first reported the activities of hydrogen as a hydroxyl radical scavenger in 2007, it has been widely confirmed that hydrogen also contains anti-inflammatory, anti-apoptotic, and cytoprotective properties ⁽¹⁰⁾. Moreover, hydrogen has advantages due to its safety, availability, and variety of methods of application.

In this study, we mainly focused on the effects of hydrogen in the intestine and postulated that the administration of hydrogen could regulate intestinal dysbiosis in a murine model of sepsis. We further investigated the impact of hydrogen treatment as a countermeasure against bacterial translocation.

MATERIALS AND METHODS

Animal sepsis model

Specific pathogen-free 6-week-old male C57/BL6 mice weighing 20 g to 25 g were obtained from SLC (Japan SLC, Inc., Hamamatsu, Japan). All mice were housed in laboratory cages in a light/dark room at 22°C to 25°C and were allowed free access to standard chow and water for 7 days until the experiments began. All the animal experiments were approved by the Animal Care and Use Committee of Osaka University Graduate School of Medicine. We performed a cecal ligation and puncture (CLP) procedure as a sepsis model. Briefly, we anesthetized the mice with intraperitoneal injection of medetomidine (0.3 mg/kg) and midazolam (4.0 mg/kg) and provided analgesia with butorphanol (5.0 mg/kg). We exposed the cecum with a 1-cm abdominal midline incision and then ligated 1 cm distal from the cecum top and made a single puncture with a 21-gauge needle for moderate CLP (30% 7-day survival). We returned the cecum to the abdomen and closed the incision. Immediately thereafter, all mice were resuscitated with saline (50 mL/kg body weight) by subcutaneous injection.

Experimental protocols

Experimental protocols in this study were divided into the sham (sham) group, saline treatment (control) group, and hydrogen-rich saline treatment (H2) group. The sham group received anesthesia without laparotomy. In the control group, 15 mL/kg of normal saline was gavaged daily for 7 days following CLP. In the H2 group, the same amount of hydrogen-rich saline was gavaged daily for 7 days following CLP. In both the treatment groups, the gavage was administered as soon as possible after the CLP procedure. Hydrogen-rich saline was prepared as saturated hydrogen saline containing 7 ppm of dissolved hydrogen according to the manufacturer’s methods (MiZ Co, Ltd, Kanagawa, Japan) ⁽¹¹⁾.

Determination of bacterial translocation

Bacterial translocation was examined in the mesenteric lymph nodes (MLN), lungs, and blood. Tissue samples were aseptically removed at 24 h after CLP and weighed and homogenized in PBS to achieve a 50-mg/mL concentration. Blood was also collected at the same time, and serum was obtained following centrifugation at 3,000 × g for 3 min. Samples were serially diluted 10-fold and were plated on tryptic soy agar (TSA) plates with 5% sheep blood and on MacConkey agar plates to grow total and gram-negative bacteria, respectively. Plates were cultured anaerobically in a 37°C incubator for 24 h and colonies were counted. Bacterial numbers are expressed as colony forming units per gram of MLN and lung tissue and per milliliter of blood, respectively.

Determination of microbiome by 16S rRNA sequencing

Fecal samples were collected on day 0 (before CLP) and days 1, 3, and 7 after CLP to determine the microbiome. DNA was extracted from fecal samples using a PowerSoil DNA Extraction Kit (MO BIO Laboratories, Carlsbad, Calif). Polymerase chain reaction (PCR) was performed using a primer set (784F: 5′-AGGATTAGATACCCTGGTA-3′ and 1061R: 5′-CRRCACGAGCTGACGAC-3′) targeting the V5-V6 region of the 16S rRNA genes with KAPA HiFi HotStart Ready Mix (KAPA Biosystems, Woburn, Mass) ⁽¹²⁾. DNA libraries were prepared using an Ion Fragment Library Kit (Life Technologies, Gaithersburg, Md) according to the manufacturer’s instructions. Sequencing was performed using two 318 chips and Ion PGM Sequencing Hi-Q Kits (Life Technologies) on an Ion PGM sequencer (Life Technologies). The resulting sequences were analyzed with the quantitative insights into the microbial ecology pipeline ⁽¹³⁾.

Quantitative analyses of enterobacteriaceae

Each fecal sample for nucleic acid extraction was weighed and suspended in nine volumes of PBS(−) to make a fecal homogenate (100 mg feces/mL). Bacterial DNA was extracted as described previously ⁽¹⁴⁾. Briefly, glass beads (0.3 g; diameter, 0.1 mm; BioSpec Products Inc, Bartlesville, Okla), 300 μL Tris-SDS solution, and 500 μL TE-saturated phenol were added to 200 μL of the fecal homogenate or bacterial culture, and the mixture was vortexed vigorously for 30 s using a FastPrep-24 homogenizer (M.P. Biomedicals, Santa Ana, Calif) at 5.0 power level for 30 s. After centrifugation at 20,000 × g for 5 min at 4°C, 400 μL of the supernatant was collected, and an equal volume of phenol-chloroform-isoamyl alcohol (25:24:1) was added to the supernatant. After further centrifugation at 20,000 × g for 5 min at 4°C, 250 μL of the supernatant was collected and subjected to isopropanol precipitation. Finally, the DNA was suspended in 200 μL of TE buffer and stored at −30°C. Real-time quantitative PCR (qPCR) was performed with GoTaq qPCR Master Mix (Promega, Tokyo, Japan) to quantify bacterial rRNA gene abundance with an ABI PRISM 7900HT sequence detection system (Applied Biosystems, Foster City, Calif). A primer set was used that is specific to Enterobacteriaceae (En-lsu-3F, 5′-TGCCGTAACTTCGGGAGAAGGCA-3′ and En-lsu-3′R, 5′-TCAAGGACCAGTGTTCAGTGTC-3′) ⁽¹⁵⁾. The primers were added at a concentration of 1 μM in each reaction. The amplification program consisted of one cycle at 95°C for 5 min, followed by 45 cycles at 94°C for 20 s, 55°C for 20 s, and 72°C for 50 s. The fluorescent products were detected in the last step of each cycle. A melting curve analysis was performed after amplification to distinguish the targeted PCR products from the non-targeted ones. The melting curve was obtained by slow heating at temperatures from 60°C to 95°C at a rate of 0.2°C/s with continuous fluorescence collection. qPCR amplification and detection were performed in 384-well optical plates with an ABI PRISM 7900HT sequence detection system (Applied Biosystems). Standard curves were generated by using quantification cycle values of DNA extracted from Escherichia coli Japan collection of microorganisms 1649 whose bacterial counts were determined microscopically with the 4′,6-diamidino-2-phenyllindole staining method as previously described ⁽¹⁶⁾. The quantification cycle values in the linear range of the assay were applied to the analytical curve generated in the same experiment to obtain the corresponding bacterial counts in each nucleic acid sample, and this count was converted to the count per sample.

Intestinal permeability

To determine intestinal permeability, the appearance in blood of 4.4 kDa fluorescein isothiocyanate-labeled dextran (FITC-dextran) (Sigma-Aldrich Co, St. Louis, Mo) was measured. The mice were gavaged with 0.2 mL of 25 mg/mL FITC-dextran in PBS at 21 h after the CLP procedure. After 3 h, a blood sample was taken from the mice by cardiac puncture. The blood was centrifuged at 4°C, 3,000 × g for 10 min, and the plasma was measured with an SH9000Lab Fluorescence Microplate Reader (Corona Electric Co, Ltd, Ibaraki, Japan) with an excitation wavelength of 480 nm and emission wavelength of 520 nm. The concentration of FITC-dextran in the plasma was determined by serial dilution of FITC-dextran as the standard.

Histologic analysis

The mice were decapitated at 24 h after CLP and then perfused transcardially with PBS followed by 4% paraformaldehyde in 0.1 M phosphate buffer (PB). The small intestine (terminal ileum) was dissected, immersed in the same fixative, and cryoprotected in a series of sucrose solutions (15%, 20%, and 25% sucrose in 0.1 M PB) at 4°C for 3 days. After the specimens were frozen in optimal cutting temperature compound (Sakura Finetechnical Co, Ltd, Osaka, Japan), they were sliced into 8-μm-thick segments by cryostat (CM3050S; Leica Microsystems, Wetzlar, Germany), and the cryosections were stained with hematoxylin and eosin. Histological damage in the intestine was quantitatively assessed with the Chiu histological injury scoring system for intestinal ischemia ⁽¹⁷⁾. The grades are as follows: 0, normal mucosa; 1, slight subepithelial detachment; 2, moderate subepithelial detachment; 3, large subepithelial detachment; 4, denuded villi; 5, ulceration.

Evaluation of oxidative stress

To determine oxidative stress, we measured tissue malondialdehyde (MDA) levels at 6 h after CLP. The MDA levels were assayed for products of lipid peroxidation observed by measurement of thiobarbituric acid-reactive substance levels. Tissue samples were frozen immediately at −80°C and divided into 50-μg samples. Divided samples were homogenized in RIPA buffer (Wako Pure Chemical Industries Ltd, Osaka, Japan) to prevent sample oxidation. All the samples were centrifuged (10,000 × g, 10 min, 4°C) to collect the supernatant and assessed using an OxiSelect TBARS Assay Kit (Cell Biolabs Inc, San Diego, Calif) according to the manufacturer’s instructions. The absorbance at 532 nm was measured using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Yokohama, Japan). MDA concentrations are expressed as nanomoles per milligram (nmols/mg) of protein.

messenger RNA (mRNA) expression of inflammatory mediators and tight junction proteins in the intestine by RT-PCR

To evaluate mRNA expressions levels of inflammatory mediators including iNOS, tumor necrosis factor α (TNF-α), interleukin 6 (IL-6), and interleukin 1β (IL-1β), and tight junction proteins including zonula occludens-1 and occludin in the small intestine (terminal ileum), tissue samples were obtained at 6 and 24 h after CLP, respectively. Total RNA was extracted and reverse transcribed to complementary DNA using a High-Capacity complementary DNA Reverse Transcription Kit (Life Technologies) according to the manufacturer’s protocol. Reverse-transcriptase PCR (RT-PCR) was performed with Fast SYBR Green Master Mix on a StepOne Plus real-time PCR cycler (Applied Biosystems). Each of the specific primers used are summarized in Table 1. PCR products were amplified (95°C for 3 s, 60°C for 30 s, 45 cycles) and detected on the Step One Plus (Applied Biosystems). The levels of mRNA expression are expressed relative to the β-actin levels.

Statistical analysis

Data are expressed as the means ± standard deviation. Differences between the experimental groups were determined using Tukey test for parametric multiple comparisons and Steel-Dwass test for nonparametric multiple comparisons. The survival rate was analyzed with Kaplan–Meier analysis, and intergroup differences were compared by the log-rank test. Statistical analyses were performed using Graph Pad Prism 7.0 (Graph Pad Software Inc, La Jolla, Calif), and P < 0.05 was considered significant.

RESULTS

Hydrogen-rich saline improves survival

Three groups of mice (n = 10 in the sham and n = 26 each in the control and H2 groups) were monitored for 1 week after CLP to determine whether the treatment improved survival. Survival curves are shown in Figure 1. The survival rates during the 7-day study periods were 100% in the sham group, 31% in the control group, and 69% in the H2 group. The survival rate of the H2 group was significantly higher than that of the control group (P < 0.01).

Hydrogen-rich saline prevents bacterial translocation

In the analysis of the MLN, blood, and lung cultivations, colonies on the TSA and MacConkey agar plates were counted at 24 h after CLP to determine if bacterial translocation had occurred (n = 3 in the sham and n = 6 each in the control and H2 groups). No colonies were observed in the sham group. In the control group, colonies had developed on the TSA and MacConkey agar plates in both the MLN and blood, whereas they were rarely detected in the lungs. The number of colonies present on the MacConkey agar plates in both the MLN and blood were significantly decreased in the H2 group compared with those in the control group (P < 0.05) (Fig. 2).

Hydrogen-rich saline regulates alterations of the intestinal microbiome

The abundant taxa from the fecal samples determined by 16S rRNA analysis are shown in Figure 3A (n = 8 each in the control and H2 groups). Microbial compositions before CLP were almost similar between the two groups and were dominated by the Muribaculaceae (S24–7) group or the families Clostridiales and Lactobacillaceae. In contrast, 1 day after CLP, the microbial composition had changed remarkably in the control group, in which an especially dynamic increase of the family Enterobacteriaceae was observed. In the H2 group, the overgrowth of Enterobacteriaceae was significantly suppressed. Furthermore, an increase in the relative abundance of Lachnospiraceae was observed in the control group 7 days after CLP but not in the H2 group. In the quantitative analysis, the counts of Enterobacteriaceae had increased by about 10⁵ on day 1 in the control group, whereas they were significantly suppressed in the H2 group (Fig. 3B). The counts of Enterobacteriaceae had normalized at 7 days after CLP in each group.

Hydrogen-rich saline attenuates intestinal hyperpermeability

We assessed intestinal permeability by measuring the appearance of FITC-dextran in plasma at 24 h after CLP (n = 12 each in the sham, control, and H2 groups). Increased levels of plasma FITC-dextran were observed in the control group compared with the sham group but were significantly ameliorated in the H2 group (P < 0.05) (Fig. 4). In the analysis of mRNA expression of tight junction proteins at 24 h after CLP, there were no significant differences between the control and H2 groups in zonula occludens-1 (4.30 ± 0.72 vs. 3.25 ± 0.37, P = 0.05) and occludin (0.75 ± 0.17 vs. 0.71 ± 0.10, P = 0.97).

Hydrogen-rich saline ameliorates intestinal morphologic damage

Histological findings of intestinal mucosal injury are shown in Figure 5A. Features including shortening or loss of villi were observed in the control group, whereas they were ameliorated in the H2 group. The histological injury score was significantly lower in the H2 group compared with that in the control group (n = 6 each in the sham, control, and H2 groups, P < 0.05) (Fig. 5B).

Hydrogen-rich saline reduces oxidative stress

The tissue levels of MDA at 6 h after CLP were measured for the analysis of oxidative stress (n = 5 in the sham and n = 6 each in the control and H2 groups). The levels of MDA in the H2 group were significantly lower than those in the control group (P < 0.05) (Fig. 6).

Hydrogen-rich saline reduces inflammatory responses in the intestinal tissue

The mRNA expression of pro-inflammatory mediators in the intestinal tissue at 6 h after CLP was determined by quantitative RT-PCR (n = 6 each in the sham, control, and H2 groups) (Fig. 7). The levels of TNF-α, IL-1β, and IL-6 were significantly elevated in the control group compared with those in the sham group. The level of iNOS also tended to be higher in the control group. The mRNA expression of these inflammatory mediators was significantly suppressed in the H2 group (P < 0.05).

DISCUSSION

In this study, we showed the impact of hydrogen-rich saline on intestinal microbiome, permeability, and survival in a murine model of sepsis.

The gut is the target organ of injury in critical illness including trauma, burns, ischemia-reperfusion injury, and sepsis ⁽¹⁸⁾. When the gut is damaged after a severe insult, not only intestinal bacteria but also toxic mediators are allowed to cross the mucosa and spread to the systemic circulation via mesenteric lymph and the bloodstream, which is known as bacterial translocation ⁽²⁾. The most serious concern is that bacterial translocation is associated with the development of multiple organ dysfunction syndrome. Therefore, prevention of bacterial translocation should be a part of the therapeutic strategy in critical illness.

In the present study, we focused on the biological effects of molecular hydrogen as a therapeutic application. Hydrogen exerts a strong anti-oxidative effect on specific reactive oxygen species such as •OH and peroxynitrite (ONOO−) ⁽⁹⁾. Several experimental studies on sepsis proved that hydrogen has potential anti-inflammatory effects ⁽¹⁹⁾. Xie et al. ⁽²⁰⁾ also reported that hydrogen gas improved survival by reducing the release of high-morbidity group box-1 and oxidative release in CLP mice. In this study, we hypothesized that oral administration of hydrogen would have beneficial effects on the intestine, particularly the microbiome, in septic mice. Regarding the intestinal effects of hydrogen, Shigeta et al. ⁽²¹⁾ previously indicated that luminal administration of hydrogen solution could attenuate intestinal ischemia-reperfusion injury in a rat model. We also inferred that hydrogen saline effectively reached the terminal ileum after CLP based on a previous study that showed an increased concentration of hydrogen in the colon 30 min after oral administration in mice with dextran sodium sulfate-induced colitis ⁽²²⁾. It should be noted in our experimental protocol that we gavaged a high concentration of hydrogen (7 ppm) saline to mice based on a study that reported the concentration-dependent effects of hydrogen ⁽²³⁾. As a result, we showed oral administration of hydrogen-rich saline could prevent the translocation of coliform bacteria to the MLN and blood by cultivation on MacConkey agar plates.

Deitch ⁽²⁴⁾ suggested two possible factors as the mechanism of bacterial translocation, including intestinal microbiome and barrier function. The microbiome is made up of about 40 trillion microorganisms and plays a crucial role in the maintenance of health ⁽²⁵⁾. Recently, it has become widely known that the balance of the bacterial community is disrupted in critical illness, and this has become a notable subject of research ⁽²⁶⁾. Shimizu et al. ⁽²⁷⁾ previously reported that alteration of the intestinal composition of the microbiome is associated with morbidity and mortality in patients with systemic inflammatory response syndrome. Ojima et al. ⁽²⁸⁾ recently showed a relation between bacterial imbalance and intensive care unit mortality by using new technologies of 16S rRNA metagenomic sequencing. In the present study, we investigated the serial change of the microbiome and evaluated the effects of hydrogen by using the 16S rRNA analysis. The most remarkable finding was the dramatic overgrowth of facultative anaerobic Enterobacteriaceae at 24 h after CLP in the control group but which was significantly suppressed in the H2 group. Generally, the family Enterobacteriaceae contains some potentially pathogenic bacteria such as Escherichia, Klebsiella, Proteus, and Citrobacter. Although the reason for the increase of Enterobacteriaceae remains unclear, inflammatory mediators such as iNOS may have a key role ⁽²⁹⁾. Winter et al. ⁽³⁰⁾ showed that mice deficient in iNOS did not support the growth of Enterobacteriaceae in the inflamed gut. The authors also suggested that iNOS provides a growth advantage to facultative anaerobic bacteria such as Enterobacteriaceae for anaerobic respiration during inflammation. We further demonstrated that mRNA expressions of iNOS, TNF-α, IL-1β, and IL-6 in the intestinal tissue were increased in the control group but were significantly reduced in the H2 group. A previous study showed that the plasma levels of cytokines were decreased by the inhibition of the nod-like receptor protein-3 (NLRP3) inflammasome following the administration of hydrogen-rich saline, which was paralleled by a decrease in NF-κB activity in an experimental model of acute pancreatitis ⁽³¹⁾. These data suggested that hydrogen would regulate the overgrowth of Enterobacteriaceae associated with the decrease of excessive inflammatory mediators. However, the cause–effect relationship between them was not clearly elucidated in our study.

We also showed that oral administration of hydrogen-rich saline could ameliorate intestinal hyperpermeability. Spitz et al. ⁽³²⁾ previously reported that enteropathogenic E coli adherence directly disrupts barrier function, allowing the bacteria to pass across the mucosa via an intracellular route. These data suggest that the beneficial effects of hydrogen against bacterial translocation may essentially be associated with the suppression of the overgrowth of Enterobacteriaceae.

In addition, we found an increase in the relative abundance of Lachnospiraceae instead of the normalization of Enterobacteriaceae at 7 days after CLP in the control group, but no such relative increase of Lachnospiraceae was found in the H2 group. Lachnospiraceae belongs to the phylum Firmicutes, which includes potentially protective bacteria. Although the cause of the shifts in bacterial abundance is not known, it may help to explain the host immune response to outcompete with the pathogenic bacteria to maintain the balance of the gut microbiome ⁽³³⁾.

There are some limitations in this study. First, the CLP model was not provided antibiotics so as to avoid their influence on the intestinal microbiome ⁽³⁴⁾, thus limiting causal interference to the mechanism of hydrogen in our study. Second, neither the most suitable amount of hydrogen nor the timing of administration was well established. Third, the survival benefit observed with hydrogen may be associated with other factors such as systemic responses other than the regulation of bacterial translocation. Further studies are needed to delineate the mechanism behind the cause and effect.

In conclusion, this is the first study, to our knowledge, to show the effectiveness of hydrogen-rich saline in preventing intestinal dysbiosis, hyperpermeability, and bacterial translocation in a murine model of sepsis, suggesting that hydrogen-rich saline could potentially become a new therapeutic strategy in critical illness.