2.2. Hydrogen Concentration and pH of MHW

The dissolved hydrogen concentration in MHW was measured using a specific MH electrode, a DM-10B2 meter (ABLE Corporation, Tokyo, Japan). In brief, a 2-week supply of MHW and DGW was prepared every 3 to 4 days. When measuring the MH concentration in the MHW treated with MH supplier sticks and DGW, we first adjusted independently the standardized MH water in the absence of MH supplier sticks by bubbling 0.03 MPa of hydrogen gas into drinking water at 25 ℃ for 30 min. The MHW MH concentration before and after the addition bubbling of hydrogen gas was measured and standardized as 0% and 100%, respectively. Subsequently, the amounts of dissolved MH in the stored MHW and freshly prepared MHW with the sticks and DGW were calculated by measuring the MH percentages, based on the saturated MH concentration (1.56 ppm to 100% at 25 °C) in MHW at 25 °C. The pH was also measured using a pH meter (F-71SHoriba; Kyoto, Japan).

2.3. Blood Glucose and Body Weight (BW) Measurements

The right submandibular vein of non-fasting DIO mice with no diet limitation was punctured using a Goldenrod^TM animal lancet (point length, 5 mm; MEDIpoint Inc., Mineola, NY, USA) during the evening once per week, and the blood glucose level of oozing blood was measured using a Glutest sensor (Panasonic Healthcare Holdings Co., Ltd. Tokyo, Japan). BW was measured at the same time.

2.4. Measurement of Cardiac Function by Echocardiography

After 14 days of MHW intake, transthoracic echocardiography was performed using an Aloka IPC-1530 ultrasound monitor (Hitachi Healthcare, Tokyo, Japan) as previously reported [28]. The mice were lightly anesthetized (1.5% isoflurane) for the duration of the recordings. Heart rate was monitored simultaneously by electrocardiography.

Left ventricular (LV) end-diastolic diameter (LVEDD) and end-systolic diameter (LVESD) were used to calculate fractional shortening (FS) and the ejection fraction (EF) using the following formulas [28]:

FS(%) = [(LVEDD−LVESD)/LVEDD] × 100%,

EF(%) = [(LVEDD−LVESD)³]/LVEDD³] × 100%

Further, the LV mass index was calculated from the LV septal wall thickness (SWT) and LV posterior wall thickness (PWT) at end-diastole and then normalized by BW according to the following formula [28]:

LV mass (mg/g) = [(LVEDD + SWT + PWT)³ − LVEDD³] × 1.055/BW

Here, 1.055 (in mg/mm³) is the density of the myocardium. All echocardiography-derived values were obtained by averaging two or three measurements for each mouse.

2.5. Sample Preparation for Immunohistochemical Analysis, Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) and Endothelial Progenitor Cell (EPC) Colony Formation Assay

As individual experiments, the mice were anesthetized by intraperitoneal administration of pentobarbital sodium (60–70 mg/kg BW; Somnopentyl; Kyouritu Seiyaku Corporation, Tokyo, Japan), and blood was then aspirated by heart puncture using a 1 mL heparinized syringe. Immediately after excision, the BAT, WAT, and heart of each mouse were weighed and each of the BAT or WAT was divided for immunohistochemical analysis and qRT-PCR. As a separate experiment, the aspirated blood and the tibia and femur, excised from left side limb, were applied for endothelial progenitor cell (EPC) colony formation assay.

2.6. Immunohistochemical Analysis

After fixation with 4% paraformaldehyde overnight, the BAT, WAT, and heart were collected, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) for histologic analysis. Deparaffinized tissue samples were blocked with avidin and biotin solution for 15 min using an Avidin/Biotin Blocking Kit (SP-2001; Vector Laboratories, Burlingame, CA, USA) as described in the supplemental protocol. Subsequently, the samples were incubated with isolectin GS1-B4 biotin conjugate (Vector Laboratories, Burlingame, CA, USA) diluted (1:100) with PBST (0.2% Tween 20, 0.1% CaCl₂, and 0.1% MgCl₂ in PBS) at RT for 2 h, washed, and then incubated with streptavidin-Alexa Fluor 488 diluted (1:500) with PBST at RT for 1 h. The stained samples were mounted with glycerol mounting medium with 1, 4-Diazabicyclo[2.2.2] octane (DABCO )(Sigma-Aldrich, St. Louis, MO, USA). Separately, the deparaffinized samples were also stained with anti-alpha smooth muscle actin-Cy3 conjugate (Cy3-αSMA; Sigma-Aldrich), blocked with 10% normal goat serum in PBST without Ca²⁺ or Mg²⁺ at RT for 30 min, incubated with Cy3-αSMA diluted (1:200) with 5% normal goat serum in PBST without Ca²⁺ or Mg²⁺ at RT for 2 h, washed with PBS, and mounted with glycerol mounting medium with DABCO^TM. Independently, the deparaffinized samples were stained with anti-8-hydroxy-2′-deoxyguanosine (8-OHdG) monoclonal antibody (Japan Institute for the Control of Aging, Shizuoka, Japan) following the same blocking protocol. The samples were incubated with anti-8-OHdG antibody 5 μg/mL with 5% normal goat serum in PBST without Ca²⁺ or Mg²⁺ at 4 °C overnight, washed with PBS, stained with DAKO EnVision™+ System, Peroxidase (DAKO EnVision™+ System; HRP, DAKO Japan, Tokyo, Japan), and DAB (3,3′-Diaminobenzidine). After sequential H-staining, the samples were mounted with glycerol mounting medium with DABCO. The stained tissue samples were observed under a fluorescence microscope (FSX-100; Olympus Corporation, Shinjuku, Tokyo, Japan), and analyzed using the cellSens software (Olympus Corporation, Shinjuku, Tokyo, Japan).

2.7. qRT-PCR

Parts of the BAT and WAT were incubated in 1 mL of RNAlater™ Stabilization Solution (Thermo Fisher Scientific, Waltham, MA, USA) at 4 °C overnight, and then stored at −80 °C until use. Total RNA was isolated using QIAzol Lysis Reagent (QIAGEN, Hilden, Germany). Contaminated genomic DNA was digested with DNase I (Thermo Fisher Scientific, Waltham, MA, USA) at 37 °C for 15 min. DNase-I-treated total RNA was purified by phenol extraction and ethanol precipitation. Then, 2 µg of purified total RNA was reverse-transcribed into complementary DNA (cDNA) with the Applied Biosystems™ High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). The cDNA mixture was diluted 10-fold after first-strand cDNA synthesis and then amplified by qRT-PCR using the ABI PRISM^® 7700 Sequence Detection System with TaqMan probes (Thermo Fisher Scientific) according to the manufacturer’s protocol. Relative mRNA levels were calculated using the ∆∆Ct method and normalized against mouse 18S rRNA. All primers and probes are listed in Table S1.

2.8. Isolation of Peripheral Blood Mononuclear Cells (PBMCs) and Bone Marrow (BM)-Derived c-Kit + /Sca-1 + Lineage-Negative Cells (BM-KSL Cells)

The aspirated blood was collected into a 1.5 mL tube per mouse. Then, the femur and tibia from the left limb of each mouse were dissected and transferred to a 15 mL conical tube containing Hanks’ balanced salt solution supplemented with 1% fetal bovine serum (FBS). PBMCs and BM-KSL cells, which are stem cell populations of EPCs, were isolated as previously reported [29].

Lineage-negative BM cells (BM-Lin-) were incubated with saturating concentrations of directly labeled anti-c-Kit (dilution, 1:25; BD Biosciences, Franklin Lakes, NJ, USA) and anti-Sca-1 antibodies (at 1:25 dilution) (BD Biosciences) for 30 min on ice. Afterward, BM-KSL cells were isolated by live sterile cell sorting using a BD FACSAria™ Fusion flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA).

2.9. EPC Colony Formation Assay

As previously described [29], PBMCs or isolated BM-KSL cells were cultured in medium containing methylcellulose (MethoCult™ SF M3236; StemCell Technologies, Vancouver, BC, Canada) with 20 ng/mL of stem cell factor (Peprotech Inc., Rocky Hill, NJ, USA), 50 ng/mL of vascular endothelial growth factor (Peprotech Inc.), 20 ng/mL of interleukin-3 (Peprotech Inc.), 50 ng/mL of basic fibroblast growth factor (Peprotech Inc.), 50 ng/mL of epidermal growth factor receptor (Peprotech Inc.), 50 ng/mL of insulin-like growth factor-1 (Peprotech Inc.), 2 U/mL of heparin (Ajinomoto Co., Inc., Tokyo, Japan), and 10% FBS on a 35-mm Falcon™ Primaria™ Cell Culture Dish (Corning Incorporated, corning, NY, USA) for 8 days. PBMCs were seeded at 5 × 10⁵ cells per dish and BM-KSL cells at 500 cells per dish. The colony-forming units (CFUs) of EPCs were identified as small-EPC-CFUs (primitive EPC-CFUs; pEPC-CFUs) or large-EPC-CFUs (definitive EPC-CFUs; dEPC-CFUs) by visual inspection with an inverted microscope at 40× magnification as previously reported [29]. The differentiation ratio was calculated as the percentage of dEPC-CFUs of the total EPC-CFUs.

3.2. MHW Had no Effect on BW or Postprandial Blood Sugar (PPBS) of DIO Mice

The BW of mice in the MHW and DGW groups significantly increased on days 7 and 14 vs. day 0 (before drinking MHW or DGW), although no significant difference was observed between groups at any time point (Table S3). However, the postprandial blood sugar (PPBS) of DIO mice in the MHW group decreased on day 14 vs. day 0, whereas there was no significant difference in BW and PPBS between groups on any day. These findings indicate that MHW had no significant effect on BW or PPBS during the observation period, although PPBS was often lower in the MHW group compared with the DGW group (Figure 2, Table S3).

3.3. MHW Morphometrically Changed the Anti-Metabolic Phenotype of Adipose Tissue in DIO Mice

After continuous MHW supplementation for 2 weeks, there was no change in the WAT mass of DIO mice (Figure 3a,b), although the BAT mass increased by 1.3-fold (Figure 3c,d). The percentage of WAT weight per kg BW was 8.32% ± 0.52% in the DGW group and 8.00% ± 0.62% in the MHW group, whereas the percentage of BAT was 0.40% ± 0.05% in the DGW group and 0.53% ± 0.07% in the MHW group. The circumferential lengths of WAT adipocytes significantly shortened in the MHW group compared with the DGW group (221.9 ± 33.5 vs. 294.7 ± 46.5 μm, respectively) (Figure 3e,f), indicating that MHW intake was associated with a decrease in the size of WAT adipocytes in DIO mice. In other words, MHW intake was associated with the transition from unilocular (hypertrophic) to multilocular (hyperplasia) WAT adipocytes in DIO mice. Similarly, it changed the phenotype of BAT adipocytes from hypertrophic to hyperplastic (Figure 3g). Notably, 8-OHdG+ cells indicating ROS-induced DNA damage decreased (17.4 ± 6.6/mm² vs. 43.5 ± 10.5/mm²) in the MHW group compared with that in the DGW group (Figure 3h,i, Table S4), whereas 8-OHdG- cells without the damage increased (11.6 ± 3.4/mm² vs. 3.75 ± 2.5/mm²). These findings indicate that MHW intake alleviated the oxidative stress in WAT in DIO mice.

3.4. MHW Intake Upregulated the Expression of Thermogenic- and Angiogenic-Related Genes in the BAT of DIO Mice

MHW intake was associated with 1.98-fold increase in eNos expression in the WAT of DIO mice. Conversely, it significantly increased the mRNA levels of thermogenic-related (eNos, Ucp-1) and angiogenic-related (eNos, Vegf) genes in the BAT of DIO mice compared with DGW by 1.63-, 1.64-, and 1.57-fold for Ucp-1, eNos, and Vegf, respectively (Figure 4). MHW intake tended to upregulate the expression of the thermogenic gene iNos and downregulate that of the inflammatory cytokine genes Tnf-α and Il-1β in both the WAT and BAT, although there were no significant changes (Figure S1, Table S5). These findings indicate that MHW intake accelerates thermogenesis and angiogenesis, resulting in the anticipated energy expenditure as well as NO production in adipose tissues.

3.5. MHW Intake Alleviated Metabolic Cardiomyopathy in DIO Mice

MHW intake contributed to the morphometric and functional improvements in metabolic cardiomyopathy in DIO mice.

• (1)
  Morphometric improvement in cardiac hypertrophy

Assessing H&E-stained cardiomyocytes at a magnification of 40×, the average width of cardiomyocytes in the MHW group was significantly narrowed by 0.78-fold compared with that in the DGW group (11.8 ± 0.58 vs. 15.1 ± 0.69 µm, respectively) (Figure 5a,b).

• (2)
  MHW intake promoted eNOS phosphorylation in cardiomyocytes of DIO mice

The LV area occupied by cardiomyocytes with phosphorylated Ser1177 of eNOS was significantly extended by 1.7-fold in the MHW group compared with the DGW group (50.5% ± 5.1% vs. 29.2% ± 5.1% per section, respectively) (Figure 5c,d).

• (3)
  Cardiac arteriolar and capillary dilatation in DIO mice in the MHW group

No significant difference was observed in the counts of cardiac capillaries stained with isolectin B4-Alexa Fluor 488 between the MHW and DGW groups (220.7 ± 60.2 vs. 215.3 ± 54.6/mm², respectively) (Figure 5e,f). Conversely, the average short-axis diameter was significantly increased by 1.3-fold in the MHW group compared with the DGW group (4.2 ± 0.52 vs. 3.3 ± 0.65 µm, respectively) (Figure 5e,g).

Further, the average short-axis diameter and count of arterioles stained with α-SMA antibody were calculated. These features of the cardiac arterioles were similar to those of the capillaries. No significant difference was observed in the arteriolar counts between the MHW and DGW groups (52.0 ± 14.0 vs. 54.8 ± 7.6/mm², respectively) (Figure 5h,i). Conversely, the average short-axis diameter was significantly increased by 1.4-fold in the MHW group compared with that in the DGW group (8.0 ± 1.67 vs. 5.6 ± 0.92 µm, respectively) (Figure 5h,j).

• (4)
  MHW intake alleviated cardiac oxidative stress in DIO mice in the MHW group

Similar to the features observed in the case of WAT, 8-OHdG+ cells tended to decrease in the MHW group compared with the DGW group (1977 ± 361/mm² vs. 3021 ± 1280/mm²), although the difference was not statistically significant (Figure 5k,l, Table S6). Notably, 8-OHdG- cells increased in the MHW group compared with the DGW group (620 ± 307/mm² vs. 124 ± 119/mm²), indicating that MHW intake alleviated the cardiac oxidative stress in DIO mice.

• (5)
  MHW intake improved cardiac function in DIO mice

Cardiac function was assessed by echocardiographic measurement of the LV parameters, namely, SWT, LVEDD, LVESD, and PWT, in M-mode (Figure 6a,b). There was no significant difference in the heart rate between the two groups (Figure 6c, Table S7). Notably, the LV mass per g BW was decreased by 0.8-fold in the MHW group compared with that in the DGW group (3.64% ± 0.54% vs. 4.57% ± 0.20%, respectively) (Figure 6d, Table S7). The EF% significantly recovered in the MHW group compared with the DGW group (40.1% ± 5.54% vs. 22.8% ± 2.53%, respectively) (Figure 6e, Table S7), as did EF% (90.3% ± 4.1% vs. 75.3% ± 4.4% respectively) (Figure 6f).

4.1. BAT Activation and WAT Browning

As reported in some previous studies, the BAT or WAT of DIO mice behaves as a critical source of inflammation, giving rise to diabetic cardiomyopathy accompanied by obesity [30,31,32]. MHW intake has been reported to alleviate inflammation by reducing the protein levels of TNF-α and IL-1β in the rat injured lung through the inhibition of NF-kβ [33].

BW was similarly increased in both the DGW and MHW groups for the entire 2-week observation period, with % ratio of BAT mass to BW significantly increased in the MHW group and no change in the % ratio of WAT mass (Figure 2a, Figure 3).

While there was no change in PPBS in the DGW group over the 2-week observation period, this parameter decreased in the MHW group at day 14 compared with pretreatment at day 0 (Figure 2b).

In addition, MHW intake for 14 days did not significantly decrease the expression levels of the proinflammatory cytokine genes Tnf-α and Il-1β in the BAT or WAT compared with the DGW group, although the gene expression levels tended to decrease, which may have become apparent due to the longer treatment period.

Kamimura et al. [21] reported that compared with degassed water, MHW intake even for 74 weeks had no effect on the BW of DIO mice. Moreover, the MHW intake improved hyperglycemia, hyperinsulinemia, and the plasma triglyceride level with decreasing BW in diabetic db/db mice with genetically engineered leptin receptor deficiency at 3 months [7]. Taking these references into account, in the present study, a more prolonged MHW intake period might be necessary to improve hyperglycemia levels.

Additionally, MHW intake was reported to upregulate the gene expression of ghrelin [34], which is an orexigenic (appetite-stimulating) and adipogenic peptide hormone secreted in the stomach. Moreover, a high-fat diet increases ghrelin-expressing cells in the gastric mucosa of mice, which contributes to obesity [35]. As noted in previous studies, the possible increase in ghrelin by MHW intake and feeding of a high-fat diet might extinguish the beneficial effect of MHW on BW or WAT mass.

Conversely, the expression of genes (Pgc-1a, Fgf21) associated with lipid metabolism for energy expenditure in the liver of db/db mice was definitively promoted by day 14 after the initiation of MHW intake [21]. Alternatively, MHW intake restored the excess amount of lipids stored in the liver of DIO mice by day 14 [7]. Collectively, these findings and the results of the present study indicate that MHW intake could restore lipid metabolism toward energy expenditure more favorably than glucose metabolism.

Notably, MHW intake improved the ROS-induced DNA damage in WAT in DIO mice, indicating that MHW exerted an antioxidative effect (Figure 3h,i). MHW intake was also associated with upregulated expression of the thermogenic genes Ucp-1, Vegf, and eNos, which promoted adipose tissue remodeling and subsequent BAT activation or WAT browning in DIO mice [36,37,38].

Further, in the present study, eNos expression was increased, which promoted the expression of Ucp-1 and Vegf in the WAT of DIO mice in the MHW group (Figure 4), indicating that MHW intake accelerated WAT browning and BAT activation to evade CVD in DIO mice [39,40,41].

4.2. Cardiovascular Protective Effects

Emerging evidence proves that BAT activation or WAT browning in obesity is beneficial to cardiovascular health and decreases the risk of CVD [16,42].

With regard to the underlying molecular mechanisms, Thoonen et al. [41] reported that UCP-1 in BAT can reverse catecholamine-induced cardiomyopathy. BAT transplantation isolated from wild-type mice recovers catecholamine-induced cardiomyopathy, which leads to fibrosis in Ucp-1-/- mice. Moreover, overexpression of Ucp-1, which is a specific protein uncoupler of oxidative phosphorylation capable of collapsing the proton electrochemical gradient, also prevented the effects of hyperglycemia [43]. In addition, ablation of Ucp-1 expression in BAT promoted ROS generation [44].

More recently, WAT (cardiac VAT) in obese rats with disturbed Ucp-1 gene expression was reported to activate hypertrophic cardiomyocytes with impaired viability and fibrosis in a myocardial organ culture system [45].

Considering the previous studies, the upregulation of Ucp-1 expression in BAT and WAT browning of DIO mice via MHW intake is considered to induce protective effects against diabetic cardiomyopathy via the upregulation of Ucp-1 expression in BAT, thereby probably decreasing ROS generation and activating eNOS in BAT (Figure 4). As a result, this mechanism conceivably repairs cardiac function in DIO mice treated with MHW. Of note, in the present study, MHW intake promoted eNOS activation in the myocardium (eNOS-Ser 1177 phosphorylation) (Figure 5c,d) and adipose tissue, especially BAT (eNos expression) (Figure 4) in DIO mice. However, future studies are required to determine whether MHW can boost eNOS phosphorylation in the myocardium directly or through Ucp-1 upregulation in BAT or WAT browning.

The activation (phosphorylation) of eNOS-Ser 1177 in cardiomyocytes in DIO mice in the MHW group (Figure 5e–j) promoted NO generation and inversely inhibited ROS production (Figure 5k,l), which enhanced NO bioactivity to improve cardiac function with vascular dilation as cardiovascular protective effects [46,47,48]. In fact, echocardiography revealed LV systolic function in DIO mice in the MHW group with a decrease in LV mass, indicating recovery from cardiac hypertrophy (Figure 6).

4.3. Activation of EPC Bioactivity

In the present study, MHW intake promoted the mobilization of EPCs into the circulation, as well as expansion and differentiation in BM (Figure 7).

Previous studies have reported the dysfunction of BM-derived EPCs in MetS in both humans and mice [16,49,50].

In the present study, MHW intake promoted eNOS activation in the adipose tissues and myocardium in DIO mice, indicating NO generation. Accordingly, exogenous NO generation may restore EPC dysfunction [51]. Further, the antioxidant effect of MHW in DIO mice might induce endogenous NO production via eNOS activation of EPCs, thereby autonomously improving the bioactivities of EPCs.

4.4. The Unlikely Alkaline Effect of MHW

The MHW utilized in this study also contained Mg(OH)₂ produced by the reaction of H₂O with Mg in the MH supplier sticks, leading to an increase in pH to around 10, thereby producing an alkaline solution. Hence, future studies are required to investigate the synergistic effects of hydrogen molecules and alkaline ions in MHW. In addition, Jackson et al. [52] recently reported that hydrogen-rich water improved fatty liver in a high-fat diet-induced model of nonalcoholic liver disease but not alkaline-electrolyzed water at pH 11, which was likely due to H₂ in hydrogen-rich water. Previous studies demonstrated, as the beneficial mechanism, that H₂, but not electrolyzed water, upregulated peroxisomal fatty acyl-CoA oxidase, a rate-limiting enzyme for beta-oxidation of very long chain fatty acids in peroxisomes, contributing to lipid metabolism.

Of note is that, in the present study, no difference was observed in the pH values of DGW and MHW at both pre- and post-supplementation. Accordingly, those findings are concluded to have originated via MHW intake.

We gratefully thank for their outstanding technical supports the Teaching and Research Support Center in Tokai University of Medicine, and Yoko Maniwa as the secretary assistant. Moreover, MH supplier sticks (Suisosui) were kindly supplied by FDR Friendear Inc. The authors would like to thank Enago (www.enago.jp, 8 April 2021) for the English language review.