Intravenous infusion of H2-saline suppresses oxidative stress and elevates antioxidant potential in Thoroughbred horses after racing exercise

Upon intensive, exhaustive exercise, exercise-induced reactive oxygen species may exceed the antioxidant defence threshold, consequently resulting in muscular damage or late-onset chronic inflammation. Recently, the therapeutic antioxidant and anti-inflammatory effects of molecular hydrogen (H2) for human rheumatoid arthritis have been demonstrated. However, it is also important to clarify the effects of administrating H2 in large animals other than humans, as H2 is thought to reach the target organ by passive diffusion upon delivery from the blood flow, indicating that the distance from the administration point to the target is critical. However, data on the effects of H2 on oxidative stress in real-life exhaustive exercise in large animals are currently lacking. We here investigated 13 Thoroughbred horses administered intravenous 2-L saline with or without 0.6-ppm H2 (placebo, N = 6; H2, N = 7) before participating in a high-intensity simulation race. Intravenous H2-saline significantly suppressed oxidative stress immediately, 3 h, and 24 h after the race, although the antioxidant capability was not affected throughout the study. The serum creatine kinase, lactate, and uric acid levels were increased in both groups. Taken together, these results indicate that intravenous H2-saline can significantly and specifically suppress oxidative stress induced after exhaustive racing in Thoroughbred horses.

race compared to in the placebo group. On the other hand, the mean values of the d-ROMs, which is an indirect method to reflect the free radicals in the serum, were 149 ± 24.6 U.CARR in the placebo group and 154 ± 24.8 U.CARR in the H 2 group at baseline, and no significant increases were observed during the study in either group (Fig. 1b).
The BAP test was also performed to measure the reducing potential in the serum. The mean values of the BAP test, which reflects the reduction potential in the serum, were 3850 ± 251 μ mol/L in the placebo group and 3560 ± 1100 μ mol/L in the H 2 group at baseline. The relative ratios of the scores at each time point after the race to the baseline are shown in Fig. 2. Although there was no significant difference  between the H 2 and placebo groups, the ratio tended to be elevated in the H 2 group (36.8 ± 20.6%), as compared to in the placebo group (8.50 ± 12.6%; p = 0.0212) immediately after the race.
Biomarkers for muscular damage and anaerobic metabolism during the race. As biomarkers for exercise-induced muscle damages, CK, AST, and LDH were measured (Table 1). In both groups, the CK levels were significantly increased 3 h after the race, but were restored to near the baseline values within 24 h after the race. There were no significant differences between the placebo and H 2 groups. In both groups, the AST and LDH levels did not significantly change during the 24-h period after the race.
Next, to estimate the alterations induced by anaerobic metabolism, the serum lactate and UA levels were measured ( Table 2). The serum lactate levels were significantly increased in both groups 1 h after the race. They reached their maximum levels (28.8 and 38.1-fold increases in the placebo and H 2 groups, respectively) immediately after the race and reduced rapidly within 3 h, remaining 1.7 and 1.8-fold  elevated compared to the baseline in the placebo and H 2 groups, respectively, after 24 h. However, there were no significant differences in the lactate levels between the groups at any time point. On the other hand, the UA level was significantly increased immediately after the race only in the H 2 group; at 1 h after the race, significant increases in the UA levels were observed in both groups. However, there were no significant differences between the groups. At 3 h after the race, the UA levels gradually decreased and were restored to near the baseline values within 24 h after the race.

Discussion
In this study, intravenous infusion of H 2 -saline showed significant antioxidant effects in Thoroughbred horses after high-intensity racing exercise, as demonstrated by the prevention of increases in the formation of 8-OHdG in the H 2 group compared to in the placebo group. This result strongly indicates protective effects of H 2 against exercise-induced ROS-mediated detrimental tissue damage in racing horses. However, it should be noted that another assay to measure oxidative stress, namely d-ROM, showed no significant differences between the two groups. We speculate that this discrepancy between the 8-OHdG and d-ROM assays is likely caused by the indirect and non-specific metabolites detected by d-ROMs. For example, superoxide and H 2 O 2 , two major ROS, are balanced by the corresponding scavengers, superoxide dismutase and catalase, respectively. The oxidative stress measured by d-ROM may reflect that of the ROS after they are reduced by their specific scavengers. On the other hand, serum 8-OHdG is a direct and reliable marker for elevated oxidative stress, as it reflects oxidised DNA. Further, while 8-OHdG reflects the intracellular oxidation or oxidants that reach the DNA, d-ROM may reflect the extracellular oxidants, which may be scavenged in the serum and may not reach the intracellular components, such as the mitochondrial or genomic DNA. Therefore, we believe that the results obtained by serum 8-OHdG are reliable and accurate.
On the other hand, although a tendency of an elevation of the antioxidant potential in the serum measured by the BAP test was observed immediately after the race in the H 2 -saline infused group, no significant differences in the elevation of antioxidant potential was observed in either group throughout the study. Recently, it was demonstrated that, in humans in rest position, intravenously infused H 2 emerged relatively slowly from the skin, and that it took more than 30 min after administration for it to be excreted 36 . In the present study, although there is a possibility that the H 2 molecules administered 2 h before the race may have been partially retained in the body of the horses during and after the race, the H 2 would have been discharged or consumed early during the race due to the high intensity of horse races, and therefore, elevation of the BAP value may not have been observed throughout the race except for immediately after the race. Alternatively, the discrepancy between the BAP and 8-OHdG values seems to suggest that the serum BAP value does not reflect the intracellular antioxidant potential, unlike 8-OHdG. These results suggest that the intravenously administered H 2 molecules have reached the muscles cells and sustained the anti-oxidant potential even after the race. However, there remain the possibilities that the H 2 molecules in blood were insufficient to reach the muscle cells for the racing exercise and instead, other unknown mechanisms which had worked on the cells or molecules in the circulating blood, may have indirectly suppressed 8-OHdG in the muscle cells.
There are numerous previous studies regarding muscular damage due to anaerobic exercise [37][38][39] . In the present study, although the AST and LDH levels did not significantly change in either group, the CK level was elevated equally in both groups upon racing, indicating the occurrence of muscular damage during horse racing (Table 1). Additionally, these results also indicate that infused H 2 did not completely protect the muscular cells from damage caused by the exhaustive racing exercise. In addition, neutrophils, which may infiltrate into the muscle tissue upon strenuous muscle contractions, can damage the muscle cells via nicotinamide adenine dinucleotide phosphate oxidase-generated superoxide, thus resulting in lysis of the muscular cells 40,41 . The finding that muscle cell damage was not influenced by the infused H 2 -saline, however, suggests that the occurrence of muscular damages upon exercise in the present study was independent from the exercise-induced ROS generation, as indicated by the suppression of the elevated oxidative stress in H 2 -infused horses. Apart from chemical stress, including ROS, the muscle cell membranes are thought to be exposed to mechanical stress due to strenuous loading and intensive contraction of the muscle fibres during high-intensity exercise 14 . When mechanical stress overwhelms the muscular resistance, the sarcomere units, which are assembled from myofibrils and connective fibrils and fixed by structural proteins such as actin, dystrophin, and titin, become overstretched and disrupted, resulting in membrane failure and leakage of the sarcoplasmic components. In the present study, we observed a significant influx of CK into the serum after the race, regardless of the pre-treatment. In this case, the muscle damage seems to originate from mechanical stress; in spite of the increased oxidative stress and absence of responsive antioxidant potentials, as indicated by the serum 8-OHdG and BAP test results, the recovery from muscular damage, demonstrated by the CK values, was completed within 24 h after the race in both groups. In other words, the muscle damage seemed to be transient, suggesting prompt recovery of the muscle membrane and absence of ischemic tissue necrosis.
The resistance to the prolonged muscle damages observed in this study regardless of the administration of H 2 could be explained by the adaptive response to the habitual endurance training of the horses. It has been demonstrated that endurance training induces mitochondrial biogenesis, represented by muscle fibre switching, which can improve energy metabolism and increase the resistance to fatigue as well as enhance the antioxidant potential [42][43][44] . It should be noted that the Thoroughbred horses included in the present study undergo daily training; this endurance training is thought to induce, at least in part, metabolic and antioxidant changes of the cellular defence mechanisms, which are believed to be modulated by PGC-1α 1 . It has been reported that PGC-1α is up-regulated in Thoroughbred horses undergoing high-intensity training for 18 weeks 45 . In addition, it is known that the muscles of Thoroughbred horses have a high proportion of type IIa fibres, which is consistent with the responsiveness to daily training 46 . In this study, the muscle fibres, particularly the type IIa fibres, may have been accustomed by the habitual endurance training, thus resulting in resistance to the increase of d-ROMs upon racing and in rapid improvement of the CK levels after the race.
Highly intensive exercises, which exhaust the anaerobic energy supply, require aerobic generation of ATP. Bioenergy in such strenuous exercises, including horse racing and human sprint athletics, is therefore provided by mixed and interlinked metabolic pathways composed of both anaerobic and aerobic reactions 47 . The present data obtained from Thoroughbred horses clearly demonstrated the influences of anaerobic metabolism as well as the presence of muscular damages. In the present study, a rapid and significant increase of serum lactate was observed (Table 1); the serum lactate levels peaked immediately after the race, clearly demonstrating the breakdown of free ATP and phosphocreatine, followed by the activation of anaerobic glycolysis. The notion of depletion of the anaerobic supply of ATP during racing is also consistent with the subsequent peak observed in the serum UA at 1 h after the race in both groups, which indicated insufficient re-phosphorylation of AMP during the race. Although a significant increase of UA immediately after the race compared to the baseline was observed only in the H 2 group, no difference in the manner of energy supply was observed between the placebo and H 2 groups. It should be noted here that, under such exhaustive circumstances, in which ischemia and hypoxia in the muscle are also induced, xanthine oxidase, which is converted from xanthine dehydrogenase, could generate superoxide via conversion of hypoxanthine to xanthine, and of xanthine to UA 48 . Further, apart from electron leakage through the mitochondrial membrane, another source of ROS may have appeared and contributed to the continuous oxidative state observed in the horses, which in turn may have been scavenged in the H 2 group. The elevations of the UA and lactate levels in both groups are suggestive of post-exercise fatigue 8,13 . Although H 2 may not improve such primary symptoms, the suppression of oxidative stress suggests a reduction of the secondary deleterious effects of racing exercise.
Finally, as mentioned, the horses used in the present study seemed to be equipped with adaptive antioxidant responses in the serum, as demonstrated by the suppression of the elevation of d-ROMs after high-intensity, exhaustive racing in both groups (Fig. 1b). The inveterate oxidative stress associated with high-intensity racing may induce late-onset or masked inflammation, which in turn is associated with increased risks of injuries, morbidity, and early mortality; this prolonged inflammation is accompanied and modulated by the elevated ROS via coupling with pro-inflammatory cytokines and NF-κ B positive feedback loops 1,31 . Importantly, it should be noted that, unlike in the placebo group, significant suppression of the serum 8-OHdG was observed in the H 2 group even 24 h after the race (Fig. 1a). This finding suggests that the administration of H 2 may be applicable to many other situations in which oxidative stress is excessive, regardless of the status of the antioxidant defence. The possibility that oral daily consumption of water containing 5-7 ppm of H 2 may counteract such oxidative risks has been hypothesised 31,49 . In addition, as demonstrated herein, it seems important to prepare for unexpected or dangerous oxidative stress increases that may overwhelm the effects of hydrogen-enriched water. Our findings suggest that one potential solution is infusion of H 2 -saline, and recent progress in the treatment of rheumatoid arthritis in humans using H 2 -saline supports this conclusion 50 .
There were some limitations to this study. Most importantly, collecting real-life data from habitually trained horses before and after racing under serious, strenuous conditions is an extensive task, and therefore, the number of horses studied was relatively small (N = 13) and a cross-over study design was not possible. However, although the present study was focused on the acute oxidative or antioxidative responses of Thoroughbred horses after racing, and although the data were restricted to within 24 h of the race, the remarkable reduction of serum 8-OHdG observed in the H 2 group indicates a preventable potential of pre-treatment with H 2 -saline against the prolonged and detrimental effects resulting from exhaustive exercise. To confirm our findings, the effects of H 2 -saline injections on anaerobic metabolism should be further investigated, both in large-scale animal studies and in humans, in the future.

Conclusions
Intravenous infusion of H 2 -saline significantly and integrally suppressed exercise-induced oxidative stress in Thoroughbred horses after exhaustive racing. Further, pre-exercise treatment also resulted in enhanced antioxidant potential. On the other hand, the transient muscle damages induced by the racing exercise, during which the anaerobic energy supply is depleted, were not affected by H 2 -saline injection. Taken together, the results of the present study clearly demonstrated that the direct injury caused by rapid and strong muscle contractions occurred independently from the generation of ROS. The antioxidant capabilities of H 2 during exercise, in which inveterate oxidative stress significantly elevates the 8-OHdG levels, may aid the beneficial properties of exercise by enhancing the antioxidant potentials as well as the anti-inflammatory effects.
Scientific RepoRts | 5:15514 | DOi: 10.1038/srep15514 Methods Subjects. Thirteen retired Thoroughbred racehorses, belonging to the Horseracing School of Japan Racing Association and regularly used by training jockeys, were randomly divided into a placebo group (N = 6; age: 4-11 years; 4 males, 2 females; body weight: 496 ± 35.1 kg) and H 2 group (N = 7; age: 3-8 years; 4 males, 3 females; body weight: 487 ± 35.7 kg). There were no significant differences in the baseline characteristics between the two groups. All of the horses are trained 6 days a week, with each training sessions including 20 min of walking, 1,200 m of trotting, and 1,000 m of canter, followed by 2,000-3,000 m of gallop, and ending with cool down trotting and walking for at least 15 min.
Preparation of H 2 -saline. To prepare the H 2 -saline, 1 L of saline in a soft plastic bag was placed for 12 h in a bath to circulate the saline. An electrolysis instrument (hydrogen circulator; Ecomo International Co. Ltd., Iizuka, Fukuoka-city, Japan) was used to generate 1.6 ppm H 2 . Two bags were placed in the infusion apparatus (INFUSTAT, Meiyu Co. Ltd., Yachio-city, Japan) and the H 2 concentration was confirmed by using the methylene blue-platinum colloid regent-based titration method 51 . Just before the infusion, approximately 0.6 ppm of H 2 was dissolved in the saline. Placebo saline was prepared in the same water bath without the addition of H 2 .
Study design. This study was approved by and conducted in accordance with the guideline of the Animal Welfare and Experiment Management Committee of Miho Training Center, Japan Racing Association. The horses took part in a simulation race in which they ran 1,000 m in a dirt track. Immediately after removing the soft bags from the hydrogen circulator, 2 L of H 2 -saline was infused into the jugular vein of each horse over 5 min, 2 h before the race. Blood samples were collected from the jugular vein just before the infusion (baseline), and immediately (7-10 min), 1 h, 3 h, and 24 h after racing. All blood samples were put on ice immediately after collection, and the serum was separated by centrifugation and frozen at − 80 °C. The blood lactate concentrations were analysed immediately after collection. The jockeys were not informed whether the infusion contained H 2 or not, and were instructed to ride the horses as if they were taking part in a real, regular race.

Biochemical analyses.
To estimate the oxidative potentials in the serum samples, d-ROMs were measured as previously described 34 . This method is an indirect detection method of free radicals, including ROS, in which N,N-Diethyl-para-phenylenediamine reacting with hydrogen peroxide in the serum is measured by the change in the colour as the absorbance at 505 nm, using the Free Radical Elective Evaluator (WISMERLL Co. Ltd. Hongo, Tokyo, Japan). The results are presented as Carratelli units (U.CARR) based on the calculated absorption. Using the same apparatus, the BAP test was also performed to evaluate the reducing potential of the serum, in which the reduction of FeCl 3 is detected as the disappearance of reddish colour 35 . Serum 8-OHdG, a marker for oxidative stress that reflects 8-hydroxyguanosine in the DNA, was measured using enzyme-linked immunosorbent assay (ELISA), as previously described 52 . The assay was performed using the highly sensitive ELISA kit for 8-OHdG (JaICA, NIKKEN SEIL Co., Ltd., Shizuoka, Japan). Serum CK, AST, LDH, and UA were measured at the Miho Training Center of the Japan Racing Association. Blood lactate was analysed using Lactate Pro2 (ARKRAY, Inc., Kyoto, Japan). Statistical Analysis. The data are shown as the mean ± standard deviation. Dunnett's test was used for analysis of transition data using continuous measurements and for the baseline pre-dose absolute value in each group. For these analyses, p values < 0.05 were considered significant. Relative changes between the H 2 and placebo groups were compared using Student's t-test, with the level of significance set at p < 0.0125 after Bonferroni correction. All statistical analyses were performed using JMP v.6.0.3a software (SAS Institute, Cary, NC).