Hydrogen-Rich Water Ameliorates Total Body Irradiation-Induced Hematopoietic Stem Cell Injury by Reducing Hydroxyl Radical

We examined whether consumption of hydrogen-rich water (HW) could ameliorate hematopoietic stem cell (HSC) injury in mice with total body irradiation (TBI). The results indicated that HW alleviated TBI-induced HSC injury with respect to cell number alteration and to the self-renewal and differentiation of HSCs. HW specifically decreased hydroxyl radical (∙OH) levels in the c-kit+ cells of 4 Gy irradiated mice. Proliferative bone marrow cells (BMCs) increased and apoptotic c-kit+ cells decreased in irradiated mice uptaken with HW. In addition, the mean fluorescence intensity (MFI) of γ-H2AX and percentage of 8-oxoguanine positive cells significantly decreased in HW-treated c-kit+ cells, indicating that HW can alleviate TBI-induced DNA damage and oxidative DNA damage in c-kit+ cells. Finally, the cell cycle (P21), cell apoptosis (BCL-XL and BAK), and oxidative stress (NRF2, HO-1, NQO1, SOD, and GPX1) proteins were significantly altered by HW in irradiated mouse c-kit+ cells. Collectively, the present results suggest that HW protects against TBI-induced HSC injury.


Introduction
Total body irradiation (TBI) may induce injury in many tissues and organs. Direct action refers to the effects of TBI on bioactive macromolecules such as proteins and nucleic acids. Ionization, excitation, chemical bond rupture, and changes in molecular structure all occur during this process, leading to abnormal function and metabolic disorders. Indirect action occurs via the generation of free radicals by water radiolysis [1]. The hydroxyl radical ( • OH), the free radical component that is one of the strongest oxidant species and that reacts with nucleic acids, lipids, and proteins, accounts for approximately 75% of lethal DNA damage caused by ionizing radiation [2]. Therefore, high concentrations of • OH radical scavengers may be promising radioprotective agents.
It is well known that molecular hydrogen (H 2 ) acts as an antioxidant by efficiently reducing • OH. Studies have shown that H 2 plays an important role in animal disease models, such as focal ischemia and reperfusion in rats [3], depressive-like behavior, and obstructive jaundice in mice [4]. In addition, it has been reported that H 2 protects mice from TBI-induced injury in many tissues and organs, including bone marrow cells [5][6][7], heart [8], testis [6,7], intestine [5], lung [9], and skin [10][11][12]. In the present study, we explore the radioprotective effect of HW on the injury of hematopoietic stem cells (HSCs).

Preparation of HW.
The preparation of HW was based on a previous study with slight modification. H2 gas was generated from a hydrogen gas generator (SHC-300, Saikesaisi HW Energy, Shandong, China) and bubbled into 500 mL of sterile water at a rate of 150 mL/min for 20 min. The concentration of H 2 in the water was detected with a dissolved hydrogen meter (Trustlex ENH-1000, Japan). HW was freshly prepared each day to ensure that an H 2 concentration of more than 0.8 ppm was maintained.

TBI and HW Administration.
Mice were divided into four groups for a 30-day survival experiment; the groups included 6.8 Gy TBI + vehicle (normal water), 6.8 Gy TBI + HW, 7.2 Gy TBI + vehicle, and 7.2 Gy TBI + HW. For the other experiments, mice were also divided into 4 groups, namely, vehicle, HW alone, 4 Gy TBI, and TBI + HW. The mice with TBI were irradiated with -rays at a dosage rate of 0.99 Gy/min. Mice received 0.5 mL HW by gavage for 10 min before TBI and up to the 7 days following TBI. The control mice with vehicle received normal water for the same frequency and volume as those in the mice with HW. Mice were finally euthanized on the 15th day after TBI.

Peripheral Blood Cell and Bone Marrow Cell Counts.
Blood was obtained from mice via the orbital sinus and collected in micropipettes coated with EDTA. K 3 , white blood cells (WBCs), the percentages of lymphocytes (LY%), and neutrophil granulocytes (NE%) were counted and calculated. Bone marrow cells were isolated from tibias and femurs as previously reported [16], and cell numbers were analyzed with a hematology analyzer (Nihon Kohden, Japan).

Flow Cytometry Analysis.
Bone marrow cells were suspended in phosphate-buffered saline (PBS), and cells were filtered and counted prior to antibody staining. To analyze the B cells, T cells, and myeloid cells in peripheral blood, 50 L of peripheral blood was harvested and stained with B220, CD3, CD11b, and Gr1 antibodies at room temperature. The red blood cells in the blood samples were removed using the BD FACS6 lysing solution. For HSC analysis, 5 × 10 6 bone marrow cells were first stained with biotin-labeled Ter119, B220, Gr1, CD11b, CD4, and CD8 antibodies and subsequently stained with streptavidin, c-kit, sca1, and CD34 antibodies. To assess ROS levels, 1 × 10 6 bone marrow cells were stained with anti-c-kit antibody and then incubated with 2,7-dichlorodihydrofluorescein diacetate (DCFDA, Beyotime Biotechnology, Nanjing, China, 10 M), MitoSox (Life Technologies, Grand Island, NY, USA; 10 M), and dihydroethidium (DHE, Beyotime Biotechnology, Nanjing, China, 5 M) in a 37 ∘ C water bath. For cell cycle analysis, bone marrow cells were first stained with a c-kit antibody and fixed and permeabilized with Cytofix/Cytoperm buffer (BD Biosciences, USA) before being stained with anti-Ki67 antibody and PI. For analysis of -H2AX, NRF2, HO-1, NQO1, BCL-XL, and BAK, bone marrow cells were first processed as described for ki67 staining and then staining -H2AX, NRF2, HO-1, NQO1, BCL-XL, and BAK antibodies at room temperature. For apoptosis analysis, bone marrow cells were also stained with c-kit and annexin V and PI according to the instructions for a BD apoptosis kit. Data acquisition was conducted on a BD Accuri C6 and analyzed using BD Accuri C6 software (BD Bioscience, San Jose, CA, USA).

Colony-Forming Units of Granulocyte Macrophage Cells
(CFU-GM) Assay. A total of 1 × 10 4 bone marrow cells in the nonirradiated control mice or 1 × 10 5 bone marrow cells in the 4-Gy TBI mice were cultured in M3534 methylcellulose medium (Stem Cell Technologies, Vancouver, Canada) for 5 days. CFU-GM colonies with more than 30 cells were counted according to the kit instructions. The results were expressed as the number of CFU-GM per 10 5 bone marrow cells.

2.9.
Isolation of c-kit + Cells. Bone marrow cells were stained with c-kit-APC antibody for 30 min on ice, washed with PBS, and resuspended the pellet with anti-APC microbeads (Miltenyi Biotec, Germany) for 15 min. c-kit + cells were sorted with a LS column in the magnetic separation filed. For analysis of ROS level and DNA damage, 1 × 10 6 c-kit positive cells were cultured in 1 mL hydrogen-rich Stem Span5 serumfree expansion medium (SFEM) 10 min before irradiation at 4 Gy. The cells were harvested for -H2AX and ROS analysis 0.5 hours and 18 hours after irradiation, respectively. For analysis of ROS level, SFEM medium (both hydrogen-rich medium and normal medium) renewed half of the liquid volume every 3 hours.

Analysis of Hydroxyl Radicals ( • OH) and of SOD and GPX Enzyme
Activities. Hydroxyl radicals and SOD and GPX enzyme activity were detected using detection kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China; and Beyotime Biotechnology, Nanjing, China) according to the manufacturer's instructions. A Fenton reaction occurred in the process of • OH detection following the manufacturer's instruction, and absorbance value was measured after chromogenic reaction happened. Level of • OH was evaluated and calculated by the formula following the manufacturer's instruction: B is the blank tube, S is the standard tube, C is the control tube, and T is the testing tube.

Statistical
Analysis. Data are shown as means ± SEM, and an unpaired -test (two-tails) was used for the majority of comparisons, along with Welch's correction -test when the variances were not equal. Comparisons of overall survival were performed using log-rank test. Statistical analyses were performed using GraphPad Prism 5 software. A < 0.05 represented statistical significance.

HW Improves the Survival of Lethally Irradiated Mice.
To test whether HW affected the survival of mice after TBI, we fed mice with 0.5 mL of HW 10 min before 6.8 Gy or 7.2 Gy TBI and then kept HW consumption daily for 7 days after irradiation. As shown in Figure 1, all mice irradiated at 6.8 Gy or 7.2 Gy died within 27 days or 15 days following TBI. However, approximately 67% of mice exposed to 6.8 Gy and 40% of mice exposed to 7.2 Gy were alive 30 days after TBI under HW consumption. These findings suggest that HW significantly increases the survival of irradiated mice, at least 6.8 Gy and 7.2 Gy.

HW Alleviates Myelosuppression and Promotes Myeloid
Skewing Recovery in Irradiated Mice. It has been well established that TBI can induce myelosuppression, a condition in which bone marrow activity decreased, resulting in a significant decline of peripheral blood cells [17,18]. Wang and colleagues showed that lymphoid-biased HSCs were more sensitive to radiation-induced differentiation than myeloidbiased HSCs, resulting in myeloid skewing in irradiated mice [19]. Thus, to determine if HW consumption affected radiation-caused myelosuppression, we analyzed the number alteration of peripheral blood cells and the percentages of B cells, T cells, and myeloid cells. As illustrated in Figure 2, the irradiated mice exposed to 4 Gy TBI exhibited a significant decrease of WBCs and lymphocyte percentage (LY%) in peripheral blood 15 days following irradiation compared to the unirradiated controls. Moreover, the percentages of B cells and T cells, as detected by flow cytometry, were also declined. Conversely, there was an increase in both neutrophilic granulocyte percentage (NE%) and myeloid  cell number in irradiated mice compared to unirradiated mice (Figures 2(c) and 2(f)). These findings indicated that TBI could result in myelosuppression and myeloid skewing. Irradiated mice with HW uptaken showed an increase of WBC counts, LY%, and B cell percentages and a decrease of NE% and myeloid cell percentage in the peripheral blood (Figures 2(c) and 2(f)). No alteration of T cell numbers was found in mice with TBI + HW. These results suggest that HW consumption improves mice recovery from TBIinduced myelosuppression and myeloid skewing.

HW Increases Number of Bone Marrow Cells (BMCs) of Irradiated Mice.
To determine whether HW consumption affected BMCs, we analyzed number alteration of BMCs per femur and the percentages of c-kit + cells (Lineage − ckit + BMCs), HPCs (Lineage − sca1 − c-kit + BMCs), LSKs (Lineage − sca1 + c-kit + BMCs), CD34 − LSK, and CD34 + LSK cells. As shown in Figure 3, 4 Gy TBI caused a decreased number of BMCs, a decrease of c-kit + cells, HPCs, and LSKs, CD34 + LSK frequency, and an increase of CD34 − LSK percentage in mice at day 15 after irradiation compared to unirradiated mice. However, HW consumption reduced or inhibited these effects caused by TBI, that is, BMCs (Figure 3(a)), c-kit + cells (Figure 3 (Figure 3(e)), and CD34 + LSK frequency (Figure 3(f)) in mice with TBI. These results suggested that HW could also improve mice recovery from TBI-induced alterations in BMCs.

HW Increases BMC Self-Renewal Ability in Irradiated
Mice. TBI causes an injury in self-renewal ability of HSCs [20,21]. 4 Gy TBI significantly decreased CFU-GM number compared to the unirradiated mice: HW rescued such decline (Figure 4(a)). The competitive bone marrow transplantation assay is a gold standard for evaluating the self-renewal and regeneration capability of HSC [16]. Thus, we also transplanted lethally irradiated (CD45.1 + ) recipient mice with bone marrow cells from unirradiated donor mice or from donor mice with 4 Gy or 4 Gy + HW. The donor-derived cells (CD45.2 + ) in the peripheral blood of recipient mice were measured by FACS 16 weeks after transplantation. The chimerism of donor cells in irradiated mice (about 4%) was significantly lower than that in unirradiated mice (about 25%), while the donor cells in mice with 4 Gy + HW showed increased chimerism (about 15%) (Figures 4(b) and 4(c)). These results suggested that HW consumption could improve BMC self-renewal ability of irradiated mice.

HW Decreases • OH Levels in the c-kit + Cells of Irradiated
Mice. To determine whether HW consumption affected ROS levels in irradiated c-kit + cells, we treated c-kit + cells with 10 M DCFDA, 10 M MitoSox, and 5 M DHE to detect total cellular ROS, mitochondria-derived ROS (superoxide), and superoxide free radicals, respectively. The • OH level was measured with a detection kit and evaluated by the ability to produce • OH as described in the Materials and Methods. All of three ROS types increased in c-kit + cells of mice with 4 Gy TBI in comparison to unirradiated mice, while HW consumption decreased the total ROS (Figures 5(a) and 5(e)) and • OH levels ( Figure 5(b)) and showed no effects on the mitochondria-derived ROS levels ( Figures 5(c) and 5(f)) or the superoxide free radical levels (Figures 5(d) and 5(g)) in irradiated mice. These results indicated that HW consumption decreases • OH levels in c-kit + cells of irradiated mouse.   c-kit + cells after irradiation. As expected, HW rescued TBIinduced cell cycle arrest and decreased the percentage of early apoptosis. These results showed that HW consumption indeed affected TBI-mediated cell cycle progression and early apoptosis of c-kit + cells in irradiated mice.

HW Decreases DNA Damage and Oxidative DNA Damage
in c-kit + Cells of Irradiated Mice. As reported previously, TBI can induce persistent oxidative stress in HSCs which may cause sustained DNA damage and oxidative DNA damage [22]. 8-oxoguanine (8-oxoG) is a naturally abundant base and impactful oxidative DNA lesions with a well-characterized mutagenic potential [23]; it is easily generated in DNA by reactive oxygen species induced by TBI [24]. Our results showed that MFI of -H2AX and percentage of 8-oxoG positive cells increased significantly in irradiated c-kit + cells compared to control group, and HW consumption downregulated expression of -H2AX and 8-oxoG (Figure 7). These results showed that HW consumption decreased DNA damage and oxidative DNA damage in c-kit + cells 15 days after 4 Gy TBI.

HW Upregulates Expression Antioxidation Proteins.
Nuclear factor erythroid 2-related factor 2 (NRF2) is a cellular sensor of oxidative stress [25]. In response to oxidative stress, NRF2 dissociates from kelch like ECH-associated protein 1 (KEAP1) and translocates into the nucleus which in turns regulates the transcription of heme oxygenase-1(HO-1) and 8 Oxidative Medicine and Cellular Longevity     NAD (P)H: quinine oxidoreductase 1 (NQO1) [26,27]. Moreover, NRF2 is found to promote the survival of irradiated cells, including BM cells, as a result of ROS scavenging [28]. To explore whether HW consumption affects expression antioxidation proteins, we detected expression of NRF2, HO-1, and NQO1 by FACS and SOD, GPX1 enzyme activity. As shown in Figure 8, HW consumption upregulated expression of NRF2 targeted proteins and enzyme activity of SOD and GPX1. These results illustrated that HW decreases ROS level through upregulating expression antioxidation proteins 15 days after 4 Gy TBI.

HW Affects the Expression of Proteins
Related to Proliferation and Apoptosis. As shown in Figure 9, a downregulation of P21 (also known as cdkn1a or Cip1), a cell cycle inhibitor, is also involved in cell apoptosis and transcriptional regulation after DNA damage [29][30][31] (Figures 9(a) and 9(b)) and BAK, a proapoptotic protein, were observed at protein levels in HWtreated irradiated mice c-kit + cells (Figures 9(e) and 9(f)). In contrast, the protein levels of the antiapoptotic protein BCL-XL were upregulated in c-kit + cells of mice bearing TBI + HW (Figures 9(c) and 9(d)). These results revealed the alteration of these proteins expressions may be potential mechanisms involved in HW radioprotection in c-kit cells.

HW Decreases ROS Level and DNA Damage in c-kit + Cells with 4 Gy IR In Vitro.
To explore the direct radioprotective effect of HW, c-kit + cells were cultured with hydrogenrich SFEM medium 10 min before 4 Gy IR, and then we detected ROS level and DNA damage 18 hours and 0.5 hours after IR, respectively. As shown in Figure 10, when cultured with hydrogen-rich medium, DCF MFI and -H2AX foci in irradiated c-kit positive cells decreased significantly compared to cells cultured with normal medium. These results suggested that HW alleviated ROS production and DNA damage in vitro.

Discussion
It has been reported that H 2 protects mice from TBIinduced bone marrow injury; however, existing research in the literature has focused on cultured hematopoietic cells or WBCs in peripheral blood, colony-forming units in the spleen, or the number of bone marrow cells in mice. The effect of H 2 on TBI-induced HSC injury and its underlying mechanisms are poorly understood. Guo et al. [32] and Chuai et al. [6] reported that mice with HW consumption 5 min before TBI alleviated radiationinduced hematopoietic system injury. Liu and colleagues reported that the hydrogen concentrations in the blood and tissues reached a maximum at 5 minutes after the oral administration of HW and then decreased slowly within 30 mins [33]. It has been widely accepted that with oral intake drugs or any other agents, drugs were absorbed into blood through gastrointestinal venous and then transported to all the tissues and organs in the body. In the publication of Liu et al., with oral administration of HW with concentration of 1.25 ppm, H2 blood levels did not increase up to an administered dose of 2.5 ppm; however, H2 levels in liver, kidney, heart, spleen, pancreas, intestine, muscle, and brain increased significantly when rats were treated with 1.25 ppm. So it is probably that H2 increase in bone marrow cells after mice were treated with 1.25 ppm HW (concentration of H2 in our experiment is about 1 ppm). Based on those previous studies, we administrated HW to mice 10 min (exactly, 5-10 min each mice per group) before TBI and 7 days after TBI. To confirm whether H2 reached bone marrow 10 mins after oral administration of HW, we measured ROS level in bone marrow cells after 4 Gy TBI (data not shown). In brief, mice were divided into 4 groups, control, 4 Gy, 4 Gy + amifostine, and 4 Gy + HW, 4 mice per group, and two independent experiments were executed. Amifostine is a ROS scavenger and radioprotective drug that has been approved by the US Food and Drug Administration (FDA), and mice received 200 mg/kg amifostine 30 mins before TBI in 4 Gy + amifostine group and 0.5 mL HW 10 mins before TBI in 4 Gy + HW group. Mice were sacrificed immediately after 4 Gy TBI and bone marrow cells were harvested; then we measured ROS level with flow cytometry. When treating irradiated mice with amifostine or HW, ROS decreased significantly which show indirect evidence that H2 reach to bone marrow 10 mins after HW consumption.
Our results show that the number of WBCs in peripheral blood and the number of BMCs per femur 15 days after irradiation are increased when irradiated mice are supplied with HW. These results are consistent with other studies [6]. Our present findings demonstrate that HW consumption can counteract imbalances in myeloid-lymphoid differentiation, increase the percentages of c-kit + cells, LSKs, HPCs, and CD34 + LSKs, and decrease the CD34 − LSKs frequency in TBI mice. Notably, self-renewal and reconstitution were significantly improved by HW.
To explore the underlying mechanisms, we measured the ROS levels, cell proliferation, cell apoptosis, DNA damage, oxidative DNA damage, and the expression of related proteins. HW consumption decreases total ROS level and selectively reduces the • OH level in irradiated c-kit + cells. In contrast, no significant alteration of mitochondria-derived ROS (superoxide) and superoxide free radicals were observed between c-kit + cells with TBI and c-kit + cells with TBI + HW. These data confirm that H 2 selectively reduces • OH level, this result is consistent with the results obtained by other researchers [3,34]. In addition, HW results in an increased percentage of proliferative cells, a decreased proportion of early apoptotic cells and the frequency of DNA damage, and oxidative DNA damage in irradiated c-kit + cells by regulating related proteins. Consistent with the experiments in vivo, when culturing c-kit + cells with hydrogen-rich SFEM medium in vitro, ROS level and -H2AX foci decreased significantly compared to these cells cultured with normal SFEM medium. The variation in cell proliferation, cell apoptosis, and DNA damage may contribute to the increased cell numbers and frequencies observed in HW-treated BMCs.
It has been reported that ROS in bone marrow cells increased significantly at 7,9,15,22,30, and 56 days after mice received TBI [22,[35][36][37][38][39]. However, it is a very short period that ROS persist after TBI, how should these radicals be generated 15 days after TBI? In our experiments, there are almost no LSKs and HPCs 4 h after TBI, then survival HSCs replenish themselves and differentiate into blood cells of all hematopoietic lineages, LSKs and HPCs appear about 10 days after TBI with high level ROS production, and the reason why ROS is produced in newly generated hematopoietic stem cells deserved to be further explored. According to the published research, NRF2 and targeted genes and proteins play an important role in ROS production in irradiated hematopoietic stem cell [28]. Our unpublished works focus on the effect of the systemic (circulatory) environment on TBI-induced bone marrow cells injury, indicating that components in irradiated mice circulatory environment may regulate expression of NRF2 and targeted genes and proteins. So next we detected expression of antioxidative proteins in c-kit + cells.
H 2 has been reported to elevate the production of SOD and GSH in the thymus [40] and testis [41]. Consistent with these studies, we demonstrate that H 2 enhances the activity of SOD and GPX enzyme. A number of in vitro and in vivo experiments have found that H 2 as a potent antioxidant activates NRF2 and its downstream signaling molecules [42][43][44][45][46][47][48][49]. Our results show that H 2 elevates the expression of NRF2 and targeted protein (HO-1 and NQO1) in c-kit + cells of TBI mice. However, the mechanism by which H 2 activates NRF2 and its related signaling pathways is unknown and warrants further exploration. In addition, our results show that H 2 upregulates the expression of the antiapoptotic protein BCL-XL, consistently with prior studies [9]. H 2 downregulates the expression of the cell cycle inhibitor P21 and the proapoptotic protein BAK in irradiated c-kit + cells. These findings suggest that H 2 may regulate many other signaling pathways to protect BMCs from TBI-induced injury.
Taken together, our present study preliminarily reveals the protective effect of H 2 by HW consumption on TBIinduced HSC injury with alteration of the related proteins. Future studies are carried on to further study the molecule mechanism by which H 2 mediated radioprotection of HSCs. H 2 , as a low-toxicity agent, shows promising radioprotective effects on TBI-induced injury in many organs and tissues. Future studies are needed to determine how to counteract issues related to solubility and stability.
Saijun Fan designed the research; Junling Zhang, Xiaolei Xue, Xiaodan Han, and Yuan Li performed the experiments with the help from Lu Lu; Deguan Li, Junling Zhang, and Xiaolei Xue analyzed and interpreted the data; and Junling Zhang wrote the manuscript with contributions from Saijun Fan. Junling Zhang and Xiaolei Xue contributed equally to this work.