A total of 60 adult male Wistar rats (age, 10-12 weeks; weight, 400-450 g) were purchased from the Experimental Animal Center of Shandong University (Jinan, China). All animals were housed together in a room maintained at 40-60% relative humidity and 23±2˚C, with 12-h light/dark cycles and ad libitum access to food and water. All animal experiments were approved by the Animal Care and Use Committee of the Qilu Hospital of Shandong University (Jinan, China; approval no. KYLL-2020-ZM-122), and adhered to the Care and Use of Laboratory Animals guidelines and to the Animal Research: Reporting of In Vivo Experiments guidelines. All animal experiments took place at Shandong Provincial Engineering Laboratory for Emergency and Critical Care Medicine, Qilu Hospital of Shandong University. All animals were anesthetized with phenobarbital sodium and then sacrificed with carbon dioxide release devices.

The 10-min asphyxial CA/CPR model was used in this study. Rats were randomly assigned to the following four groups (n=15 per group): i) Sham + normoxia (anesthetized with 4% isoflurane mixed with room air and normoxia inhalation by ventilator for 2 h); ii) Sham + H₂ (anesthetized with 4% isoflurane mixed with room air and H₂ inhalation by ventilator for 2 h); iii) CPR + normoxia (anesthetized with 4% isoflurane mixed with room air, CA/CPR treatment and normoxia inhalation by ventilator for 2 h after ROSC); and iv) CPR + H₂ (anesthetized with 4% isoflurane mixed with room air, CA/CPR treatment and H₂ inhalation using a ventilator for 2 h after ROSC). Rats were anesthetized with 4% isoflurane mixed with room air and were under tracheal intubation with connection to a ventilator. Intravascular catheters were placed into the right femoral artery and right vein for blood pressure monitoring and drug administration, respectively. After stabilization for 20 min, the ventilator was disconnected to induce CA. Circulatory arrest was determined by cessation of the arterial pulse and a mean arterial pressure (MAP) of <20 mmHg. After 10 min of asphyxia, CPR was performed, with inhalation of air with a ventilator and intravenous administration of epinephrine (0.01 mg/kg). In addition, the rate of artificial chest compressions was ~200 per min. Epinephrine (0.02 mg/kg) was administered at 2-min intervals until ROSC was achieved. ROSC was defined by an MAP of >60 mmHg that lasted for at least 10 min. A total of 3 rats with ROSC failure within 5 min or those that could not be disengaged from the ventilator after observation for 1 h were excluded from the study. The core temperature of each rat was maintained at 37.0±0.5˚C. After ROSC, gas inhalation was continued for 2 h (Fig. 1A): Premixed gas (1.3% H₂ and 26% O₂) was used in the H₂ therapy groups, and normoxia (26% O₂) was used as the control (27). Following this, the rats were euthanized by exposure to a gradually increasing concentration of CO₂ (the flow rate was 50% of the chamber volume/min).

H9C2 cells were purchased from the American Type Culture Collection and cultured in Dulbecco's modified Eagle's medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) and 100 U/ml penicillin + 100 mg/ml streptomycin in a cell incubator (95% air and 5% CO₂) at 37˚C. In the cell hypoxia/reoxygenation (H/R) model, H9C2 cells were exposed to hypoxia (5% CO₂ and 1% O₂ at 37˚C) for 24 h and reoxygenation (95% air and 5% CO₂) for 4 or 12 h.

In the H₂ treatment group, H₂-rich medium was used to culture the H92C cells instead of normal DMEM. H₂ was diluted into cell culture medium to produce an H₂-rich culture medium (0.6 mmol/l) (28). The H₂-rich medium was freshly prepared for each experiment. DMEM was used as the vehicle.

Part of the left ventricle (LV) was obtained from the rats and fixed quickly in 2% glutaraldehyde and 2% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at room temperature for 2 h. After 3 washes in phosphate-buffered saline, tissues were fixed and stained using 1% osmium tetroxide on 4˚C for 2 h and washed 3 times in phosphate-buffered saline. After ethanol dehydration, samples were embed in LX112 resin (Ladd Research Industries) at room temperature for 2 h. Ultrathin sections (~70 nm) were obtained with a MT-X ultramicrotome (Leica EM UC7; Leica Microsystems, Inc.) and observed using an electron microscope (H-7650; Hitachi, Ltd.). Images of sections were assessed using ImageJ (V1.8.0.112; National Institutes of Health).

The TEM images were processed using ImageJ V1.8.0.112. The outline of each mitochondrion was precisely drawn using a Surface Pro 6 tablet equipped with a touch pen (Microsoft Corporation), and filled with a solid bright color. The image with color-filled mitochondria was converted into a binary black and white image using the ‘color threshold’ command, and areas of mitochondria were then generated using the ‘analyze particles’ command. Briefly, this command recognized the black objects (mitochondria) in the binary images and outlined them such that the area of each outlined object was automatically computed. Fractional area was calculated as total mitochondrial area divided by image area for the cardiomyocytes. Mitochondrial number and average mitochondrial area (an indication of mitochondrial size) were also measured (29).

Samples (left ventricle tissues from rats and H9C2 cells) were lysed with RIPA buffer (Beyotime Institute of Biotechnology) for 30 min on ice. After centrifugation at 12,000 rpm for 10 min at 4˚C, the supernatant was transferred to a new tube. PMSF (1:100) was added, and a BCA kit (Beyotime Institute of Biotechnology) was used to determine protein concentration. Proteins (20 µg) were separated using 10% SDS-PAGE and transferred onto PVDF membranes. After blocking with 5% milk in PBS at room temperature for 2 h, the membranes were incubated with primary antibodies at 4˚C overnight. The membranes were washed three times with TBS-Tween-20 (TBS-T) and incubated with secondary antibody (dissolved in 1% BSA; 1:5,000) at room temperature for 1 h. After being washed with TBS-T three times, the membranes were exposed to ECL substrate (Beyotime Institute of Biotechnology; catalog no. P0018FS) and detected by the chemiluminescence method in an AI600 gel imaging system (GE Healthcare). The following primary antibodies were used: Anti-Beclin-1 (catalog no. ab223348; 1:1,000), anti-LC3B (catalog no. ab51520; 1:1,000), anti-p62 (catalog no. ab109012; 1:1,000) and anti-β-actin (catalog no. ab8226; 1:1,000) (all Abcam). The secondary antibodies were goat anti-rabbit IgG H&L (HRP) (catalog no. ab205718; 1:5,000) and goat anti-mouse IgG H&L (HRP) (catalog no. ab6789; 1:5,000) (both Abcam).

Rats tissues were fixed with 4% polyformaldehyde at room temperature for 24 h, then dehydrated with alcohol and washed with xylene. The tissues were embedded in a wax block and sliced to a 5-µm thickness. Dewaxing agent was used to dewax the sections at 55˚C for 1 h, and different concentrations of alcohol (100, 95, 90, 80, 70, 60 and 50%, for 30 min each) were used to wash the sections. According to the instructions of the SABC-POD kit (Boster Biological Technology; catalog no. SA1028), sections were incubated with 3% H₂O₂ at room temperature for 5 min, and infiltrated into 0.01 M Citrate Antigen Retrieval solution (Wuhan Servicebio Technology Co., Ltd.; catalog no. G1201-1L) at 100˚C for 20 min. To cool them down to room temperature, sections were washed three times with PBS (phosphate-buffered saline). After blocking with 5% BSA (Wuhan Servicebio Technology Co., Ltd.; catalog no. SW3015) at 37˚C for 30 min, sections were incubated with primary antibodies at 4˚C overnight. Next, the sections were washed three times with PBS, and incubated with anti-rabbit IgG H&L (HRP; 1:1,000; cat. no. ab205718; Abcam) and goat anti-mouse IgG H&L (HRP; 1:1,000; cat. no. ab6789; Abcam) at 37˚C for 30 min. After washing three times with PBS, the sections were incubated with SABC at room temperature for 20 min. Finally, sections were incubated with DAB substrate (Boster Biological Technology; catalog no. AR1022) and washed with ddH₂O. Sections were redyed with hematoxylin for 30 sec, dehydrated with inversed different concentrations of alcohol and sealed with Permount^™ Mounting Medium (Sangon Biotech, Co., Ltd.; catalog no. E675007). Immunohistochemical staining was performed with anti-LC3B (1:1,000). Images of sections were captured (IX53; Olympus Corporation) and assessed (ImageJ V1.8.0.112) according to the percentage of stained cells (total original magnification, x40; area, 250x250 µm).

H9C2 cells were incubated with the aforementioned primary antibodies to evaluate the levels of Beclin-1 (1:100) and LC3B (1:100). Samples were incubated with DAPI after washing three times with PBS at room temperature, and then sealed with coverslips immediately. Images of cells were taken using confocal laser scanning fluorescence microscopy (SP8; Leica Microsystems GmbH).

The changing fluorescent punctae of LC3 in H9C2 cells were observed with a tandem red fluorescent protein (RFP)-green fluorescent protein (GFP)-LC3 construct (Ad-RFP-GFP-LC3). Ad-RFP-GFP-LC3 adenovirus was purchased from ViGene Biosciences (Charles River Laboratories, Inc.). H9C2 cells (American Type Culture Collection; cat. no. CRL-1446) were transfected with Ad-RFP-GFP-LC3 at 50 MOI. In brief, H9C2 cells were inoculated into a 24-well plate at a density of 1x10⁵ cells/well. A total of 250 µl of DMEM containing 1% FBS (both Thermo Fisher Scientific, Inc.), 100 µl adenovirus (10⁶ PFU/ml; ViGene Biosciences; Charles River Laboratories, Inc.) and 100 µl Lipofectamine 2000^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was added to each well. After mixing, the cells were cultured for 6 h in an incubator containing 5% CO₂ at 37˚C. After 6 h, the adenovirus was moved off and the transfected cells were cultured for 48 h continuously. The green and red fluorescence intensities were assessed under laser scanning fluorescence microscopy (SP8; Leica Microsystems GmbH). Images of sections were assessed (ImageJ software V1.8.0.112; National Institutes of Health) according to the numbers of red and yellow dots in each cell (total original magnification, x63).

Under isoflurane anesthesia, the spontaneous breathing of the rats was maintained. The two-dimensional images of the LV [left ventricular fraction shortening (LVFS) and left ventricular ejection fraction (LVEF)] were collected using ultrasonic cardiogram equipment (Vevo2100; VisualSonics, Inc.) on the short-axis and long-axis section of the parasternal papillary muscles. The two-dimension-guided M-mode or B-mode ultrasonic cardiogram of 10 cardiac cycles was obtained. Images were assessed using Vevo2100 software.

Blood was collected at sacrifice and then the serum was isolated using centrifugation (1,000 g; 4˚C for 15 min). The serum concentrations of CKMB (Cloud-Clone Corp.; catalog no. SEA479Ra) and cTnT (Cloud-Clone Corp.; catalog no. SED232Ra) were measured by ELISA kit following the manufacturer's instructions.

The statistical significance was performed using GraphPad Prism 8 (GraphPad Software, Inc.) Differences between groups were estimated using one-way ANOVA followed by Tukey's post hoc test. Comparisons across two variables were used a two-way ANOVA followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference. All data are expressed as the mean ± standard error.

As shown in Fig. 1B and C, there were no statistical differences in MAP and heart rate during post-resuscitation care whether using inhalation of H₂ gas or not. The survival rate of the rats was recorded continuously for 7 days after CA/CPR (Fig. 1D). In the Sham groups, the survival rate was 100% whether using inhalation of H₂ gas or not. However, compared with that in the normoxia groups, the survival rate at 7 days after ROSC was significantly higher following inhalation of H₂ gas during post-resuscitation care.

Ultrasonic cardiogram detection was used to further evaluate the effect of H₂ on the cardiac function of rats after ROSC. As shown in Fig. 1E, cardiac function was evaluated by echocardiography at 4 and 72 h post-ROSC. Echocardiograms were analyzed by the LV trace of M mode images using VevoStrain software. Compared with the Sham group, the LVFS and LVEF were significantly decreased within 4 h following ROSC. However, the LVFS and LVEF significantly increased after ROSC in the CPR + H₂ group (Fig. 1F and G). However, there was no significant difference in levels of LVFS and LVEF between CPR and CPR + H₂ groups at 72 h. Moreover, compared with inhalation normoxia, H₂ therapy also markedly decreased the levels of myocardial injury biomarker CKMB and cTnT in serum after CPR at 4 and 72 h (Fig. 1H and I). Thus, inhalation of H₂ after ROSC significantly improved the cardiac function of rats.

To investigate the potentially cardioprotective mechanism of the inhalation of H₂ after ROSC, the cross-section and TEM images of LV cardiomyocytes were observed. As shown in Fig. 2A and B, TEM analysis revealed the presence of extensive mitochondrial abnormalities, such as swelling, disorganization and loss of cristae, and the relative mitochondrial mass were significantly deceased in post-resuscitation rat LV cardiomyocytes. However, inhalation of H₂ gas after ROSC significantly decreased the number of abnormal mitochondria and increased the relative mitochondrial mass in rat LVs, and cardiac tissue seemed to have returned to its original morphology at 72 h.

Moreover, autophagic lysosomal structures were common in post-resuscitation rat LV cardiomyocytes, suggesting the activation of autophagic cell death mechanisms. In the CPR groups, a significant decrease in autophagic vesicles was observed at 4 and 72 h in post-resuscitation rats after inhalation of H₂ gas compared with normoxia (Fig. 2A and C).

The study next investigated how H₂ affects autophagy in the heart. As show in Fig. 2D-G, western blot analysis revealed increased expression levels of the autophagy promotor protein Beclin1 in normoxic post-resuscitation rat hearts. The expression levels of LC3B were also higher in the CPR + Normoxia groups at 4 and 72 h compared with that in the Sham-Normoxia group, and the ratio of LC3BII/I was also significantly higher, indicating an enrichment of LC3BII. However, the expression levels of p62 were significantly lower in the same groups. There results suggested excessive autophagy activation. However, compared with the CPR + Normoxia groups, H₂ treatment suppressed autophagy activation with significantly lower Beclin-1 and LC3B levels, and a higher p62 protein level. Furthermore, immunohistochemical staining revealed that LC3B levels were significantly higher in cardiomyocytes from rats after ROSC, while H₂ therapy significantly decreased the LC3B protein levels (Fig. 2H and I).

All H/R-induced injury experiments were performed on the rat heart embryonic H9C2 cell line. The present study data (Figs. 1 and 2) had shown that H₂-treatment protected cardiomyocytes from CA/resuscitation by inhibiting autophagy activation. To clarify the role of H₂ therapy in CPR-induced myocardial I/R injury, H9C2 cells were subjected to 24 h of hypoxia followed by 4 or 12 h of reoxygenation. According to a previous study protocol, H₂-rich culture medium was used as H₂ treatment for cells in vitro (28).

The expression levels of autophagic markers Beclin-1, LC3B and p62 were measured by western blot analysis in H/R-treated H9C2 cells. After H/R, H9C2 cells showed an increase in expression levels of autophagic proteins Beclin-1, LC3B and p62, compared with the control (Fig. 3A and B). H₂ significantly decreased Beclin-1 and LC3B expression in the H/R-treated cardiomyocytes, suggesting the inhibition of autophagy. However, the p62 expression was increased significantly in these H/R + H₂ groups.

Consistent with the western blot analysis, the immunofluorescence staining also indicated that the expression of Beclin-1 and LC3B was increased in H/R-treated cells, while H₂ administration significantly decreased the expression of these proteins (Fig. 3C-F). These data support the proposal that H₂ protects against H/R-induced injury by decreasing H/R-mediated autophagy.

In order to study the formation and process of autophagy, H9C2 cells were transfected with Ad-monomeric RFP (mRFP)-GFP-LC3 and observed at different time points after H/R. As shown in Fig. 4, the generation of both GFP- and mRFP-positive autophagosomes was significantly increased in the H9C2 cells after H/R, while H₂ treatment resulted in a significant decrease in both autophagosome types.