Activating Mitochondrial Cleanup Protects the Brain and Spinal Cord from Oxygen Toxicity Caused by High-Pressure Oxygen

Transmission electron microscopy revealed normal mitochondrial morphology in hippocampal neurons from the NC (normobaric normoxia) group, with intact cristae and uniform matrix density. Notably, to distinguish the effects of hyperbaric pressure itself from those of hyperoxia, we included a hyperbaric normoxia (HNO, 0.6 MPa nitrogen-oxygen mixture with normobaric oxygen partial pressure) control group. The HNO group also exhibited normal mitochondrial ultrastructure. In contrast, rats exposed to HBO (hyperbaric hyperoxia, 0.6 MPa 100% O₂) exhibited marked mitochondrial abnormalities, including swelling, disorganization of cristae, and matrix vacuolization (Figure 1a,b). To assess the functional consequences of HBO-induced mitochondrial damage, we measured ΔΨm using JC-1 staining. The red/green fluorescence ratio was significantly reduced in hippocampal tissues from HBO-exposed rats compared to the NC and HNO groups (Figure 1c,d), indicating mitochondrial depolarization.

Notably, autophagosome-like double-membrane structures surrounding mitochondria were occasionally observed in the HBO group (Figure 1a, arrow), suggesting attempted sequestration of damaged mitochondria. Such structures were rarely detected in NC or HNO groups, indicating that HBO stress triggers mitochondrial damage and initiates autophagic responses. To further characterize the autophagic response, we examined the expression of mitophagy-related proteins by Western blotting. The microtubule-associated protein 1 light chain 3 (LC3)-II/LC3-I ratio, an indicator of autophagosome formation, showed no significant change in the HBO group compared to the NC group immediately after exposure (Figure 1e,g). Consistently, P62 levels remained unchanged (Figure 1e,f), suggesting that autophagic flux had not yet been completed at this early time point. Additionally, translocase of the outer mitochondrial membrane 20 (TOMM20), a mitochondrial outer membrane protein that reflects mitochondrial mass and is typically degraded during mitophagy, showed no significant reduction (Figure 1e,h), further indicating that clearance of damaged mitochondria had not yet occurred.

To elucidate the molecular mechanisms underlying HBO-induced neurotoxicity, we performed RNA sequencing on hippocampal tissues from the NC and HBO groups. Differential expression analysis identified a substantial number of genes significantly altered following HBO exposure (Figure 2a). Hierarchical clustering analysis revealed distinct separation between the NC and HBO groups (Figure 2b).

The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of differentially expressed genes revealed significant enrichment of multiple signaling pathways (Figure 2c). Notably, the MAPK signaling pathway was among the most significantly enriched pathways, along with the tumor necrosis factor (TNF) signaling pathway, nuclear factor kappa B (NF-κB) signaling pathway, and forkhead box O (FoxO) signaling pathway. Additionally, pathways associated with inflammatory responses, including the interleukin-17 signaling pathway, chemokine signaling pathway, Toll-like receptor signaling pathway, and C-type lectin receptor signaling pathway, were also significantly enriched. Other enriched pathways included those related to infectious diseases, immune-related conditions, and metabolic or cardiovascular processes. The enrichment of the apoptosis pathway further supported the involvement of cell death programs in HBO-induced neurotoxicity.

Gene ontology (GO) enrichment analysis of biological processes revealed that the downregulated genes were predominantly associated with regulation of blood pressure, regulation of blood circulation, regulation of hormone secretion, hormone transport, and positive regulation of secretion (Figure 2d). Terms related to multicellular organism processes and multi-organism reproductive processes were also enriched, reflecting the systemic nature of the response to hyperbaric stress. GO molecular function analysis identified enrichment of calcium-dependent phospholipid binding, ligand-activated transcription factor activity, nuclear receptor activity, G protein-coupled peptide receptor activity, and peptide receptor activity (Figure 2d). These findings suggest that calcium signaling and transcriptional regulation play important roles in HBO-induced neuronal responses.

Collectively, these transcriptomic data demonstrate that HBO exposure triggers robust activation of MAPK, TNF, and NF-κB signaling pathways, upregulates pro-apoptotic gene expression, and induces a transcriptional program associated with inflammation and cellular stress, providing a molecular basis for further investigation into the mechanisms underlying HBO-induced neurotoxicity.

Based on the in vivo observations of mitochondrial damage and autophagic membrane recruitment, along with transcriptomic enrichment of apoptosis and MAPK pathways, we next examined the direct cellular effects of HBO using HT22 hippocampal neurons. Cells were exposed to NC, HNO (0.6 MPa nitrogen–oxygen mixture with normobaric oxygen partial pressure), or HBO (0.6 MPa 100% O₂) for 1 h, and mitochondrial function was assessed at the indicated time points after decompression.

Mitochondrial membrane potential (ΔΨm) was evaluated using TMRE staining, a membrane potential-driven cationic probe that accumulates in healthy mitochondria. Flow cytometry analysis revealed that HBO exposure caused a significant decrease in the percentage of TMRE-positive cells, from 50% in the NC group to 39.2% in the HBO group, indicating marked mitochondrial depolarization (Figure 3a). In contrast, the HNO group showed no significant change in ΔΨm compared to NC controls, confirming that the observed mitochondrial dysfunction is specifically attributable to hyperbaric hyperoxia rather than pressure alone. We also assessed total and mitochondrial ROS levels immediately after decompression and removal from the chamber using DCFH-DA and MitoSOX Red staining in living cells, respectively. Flow cytometry analysis revealed a significant increase in both DCFH-DA-positive cells and MitoSOX Red-positive cells in the HBO group compared to NC controls (Figure 3b,c), indicating elevated total ROS production with mitochondria as a major source. No significant changes in ROS levels were observed in the HNO group.

We then assessed apoptosis in the three groups following HBO exposure using Annexin V-PE staining. At 0 h post-exposure, no significant differences were observed among the three groups (Figure 3d). However, at 6 h post-exposure, the HBO group exhibited a marked increase in apoptosis, with the percentage of annexin V-PE-positive cells rising from 21.8% in the NC group to 55.2% in the HBO group. In contrast, the HNO group showed no significant change (20.5%) compared to NC controls (Figure 3e). Collectively, these findings demonstrate that HBO exposure—but not HNO—induces mitochondrial depolarization, elevates mitochondrial ROS production, and triggers apoptosis in HT22 cells at 6 h post-exposure, consistent with the in vivo observations.

Given the observation that autophagosome-like membranes were recruited to damaged mitochondria following HBO exposure in vivo (Figure 1a), we next investigated whether this morphological event translates into functional mitophagy and whether it is sufficient to clear damaged mitochondria. To preliminarily explore the time course of mitophagy induction and dynamically monitor mitophagic flux, we utilized HeLa cells stably expressing parkin and the pH-sensitive mitophagy probe Mito-Keima [20,21]. Cells were exposed to HBO for 1 h, and Mito-Keima fluorescence was observed by confocal microscopy at various time points after decompression. As shown in Figure 4a, mitophagy was detectable at 3 h post-exposure and became markedly increased at 6 h, indicating that mitophagy is progressively induced following HBO exposure.

To further characterize the autophagic response in hippocampal neurons, HT22 cells were exposed to HBO for 1 h, and cell lysates were collected at 6 h post-exposure for Western blot analysis. Compared to the NC group, the HBO group exhibited a significant increase in the LC3-II/LC3-I ratio and a decrease in P62 expression, indicating activation of autophagic flux. However, TOMM20 showed no significant reduction in the HBO group (Figure 4c–f). These findings suggest that although autophagy is induced by 6 h post-exposure, the clearance of damaged mitochondria lags behind the injury process. The time required for mitophagic induction and completion allows damaged mitochondria to accumulate in the interim, and by the time autophagic flux is fully engaged, the window for effective clearance may have already passed.

We next examined how modulation of mitophagy affects HBO-induced apoptosis. HT22 cells were pretreated with SAR405, a Vps34 inhibitor that blocks autophagosome formation, or everolimus, an mTOR inhibitor and autophagy activator, for 3 h, followed by HBO exposure for 1 h. Protein expression was assessed at 6 h post-exposure by Western blotting (Figure 4b). Compared to the NC group, the HBO group showed decreased P62 expression, an increased LC3-II/LC3-I ratio, and an elevated cleaved caspase-3/caspase-3 ratio, with no significant change in TOMM20. Notably, SAR405 pretreatment further increased P62 expression and the cleaved caspase-3/caspase-3 ratio, while TOMM20 expression remained unchanged (Figure 4c–g). This indicates that SAR405 blocked autophagic flux and exacerbated apoptosis. The LC3-II/LC3-I ratio did not show inhibition when compared to the HBO group, which may be attributed to the combined effects of increased apoptosis and autophagic blockade. In contrast, everolimus pretreatment significantly decreased P62 and TOMM20 expression while further increasing the LC3-II/LC3-I ratio, indicating enhanced autophagic flux and successful clearance of damaged mitochondria, accompanied by a marked reduction in the cleaved caspase-3/caspase-3 ratio (Figure 4c–g). Of note, CCK-8 assay confirmed that treatment with SAR405 or everolimus alone at the indicated concentrations for 10 h had no cytotoxicity (Figure S1). Collectively, these results demonstrate that although HBO exposure triggers mitophagy, the endogenous response is initiated with a temporal delay, allowing damaged mitochondria to accumulate before clearance mechanisms become fully operational. Pre-activation of mitophagy with everolimus effectively eliminates damaged mitochondria in a timely manner and protects against apoptosis, whereas blockade of autophagic flux exacerbates cell death.

We next investigated whether everolimus exerts similar protective effects in vivo. Rats were pretreated with everolimus (2.5 mg/kg) or vehicle by oral gavage once daily for 4 consecutive days, followed by HBO exposure on day 5. Seizure latency was recorded during exposure, and hippocampal tissues were collected immediately after decompression (0 h) for analysis (Figure 5a). Notably, everolimus pretreatment, to some extent, prolonged seizure latency compared to vehicle-treated controls (Figure 5b). Immunofluorescence staining for lysosome-associated membrane protein 1 (LAMP1), a lysosomal marker, and TOMM20 revealed that the HBO group exhibited limited colocalization at 0 h, indicating that endogenous mitophagic flux was not yet fully engaged; conversely, everolimus pretreatment markedly increased LAMP1/TOMM20 colocalization (Figure 5c,d), demonstrating that everolimus establishes a baseline of activated mitophagy, enabling immediate delivery of damaged mitochondria to lysosomes for degradation upon HBO exposure.

To assess whether everolimus preserves mitochondrial function and suppresses MAPK signaling, we measured ΔΨm using JC-1 staining and examined p-p38 and c-Fos expression by Western blotting. As shown in Figure 5e,f, the HBO-induced reduction in the red/green fluorescence ratio was significantly attenuated by everolimus pretreatment, indicating protection against mitochondrial depolarization. Additionally, HBO exposure significantly increased p-p38 and c-Fos levels, and both increases were significantly reduced by everolimus pretreatment (Figure 5g–i).

Male Sprague Dawley rats weighing 200–250 g (approximately 8 weeks of age) were purchased from Bikai Keyi Biotechnology Co., Ltd. (Shanghai, China). All animal procedures were approved by the Institutional Animal Care and Use Committee and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Animals were housed under a 12 h light/dark cycle with free access to food and water.

To distinguish the effects of hyperbaric pressure from those of hyperoxia, a total of 18 rats were randomly assigned into three groups (n = 6 per group): normobaric normoxia (NC) group, hyperbaric normoxia (HNO) group, and hyperbaric oxygen (HBO) group. Hippocampal tissues from the NC and HBO groups were collected for RNA sequencing analysis, as the HNO group did not exhibit mitochondrial damage. To evaluate the protective effect of everolimus pretreatment, another 18 rats were randomly divided into three groups (n = 6 per group): NC group, vehicle-treated HBO group, and everolimus-treated HBO group. Seizure latency was recorded in the two HBO-exposed groups. For everolimus pretreatment, rats in the everolimus-treated HBO group received intragastric administration of 2.5 mg/kg everolimus (Macklin (Shanghai, China), 159351-69-6) suspended in 0.5% sodium carboxymethyl cellulose (CMC-Na) once daily for four consecutive days. Rats in the vehicle-treated HBO group received an equal volume of 0.5% CMC-Na solution following the same schedule. The dose of 2.5 mg/kg was selected based on previous studies demonstrating effective autophagy activation and neuroprotection in vivo [23].

Based on preliminary experiments in our laboratory, rats were placed in the animal hyperbaric chamber (customized by Hongyuan Oxygen Industrial (Yantai, China)) and exposed to 100% oxygen at 0.6 MPa (6 ATA) for 30 min (HBO). During the exposure, animals were continuously monitored for convulsions, consciousness, and respiration and were removed from the chamber by decompression at 0.1 MPa/min either immediately upon prolonged loss of consciousness or after completion of the fixed 30 min exposure, whichever occurred first. For hyperbaric normoxia (HNO) controls, rats were exposed to a nitrogen–oxygen mixture at 0.6 MPa with a normobaric oxygen partial pressure. The normobaric normoxia (NC) group was kept at room air under normal atmospheric pressure.

The procedure was performed as previously described [24]. Hippocampal tissues were fixed in 2.5% glutaraldehyde, post-fixed in 1% osmium tetroxide, dehydrated through a graded ethanol series, and embedded in epoxy resin. Ultrathin sections (60 nm) were stained with uranyl acetate and lead citrate, then examined under a transmission electron microscope.

Total RNA was extracted from hippocampal tissues of the NC and HBO groups using TRIzol reagent (Invitrogen (Carlsbad, CA, USA), 15596026). RNA purity and concentration were assessed using a NanoDrop spectrophotometer (Wilmington, DE, USA). Libraries were constructed and sequenced on an Illumina NovaSeq 6000 platform (San Diego, CA, USA). Differential expression analysis was performed using DESeq2, with genes showing adjusted p-value < 0.05 considered significantly differentially expressed. KEGG pathway enrichment and Gene Ontology (GO) analysis were conducted using clusterProfiler (v4.14.6).

HT22 mouse hippocampal neurons were cultured in a specialized medium optimized for this cell line (ServiceBio (Wuhan, China), GZ20011) consisting of DMEM (ServiceBio, G4511) supplemented with 10% fetal bovine serum (ServiceBio, G8002) and 1% penicillin–streptomycin (ServiceBio, G4003). HeLa cells stably expressing parkin and Mito-Keima were generated by retroviral transduction as previously described [20,21] and were a generous gift from Prof. Jiahong Lu at the University of Macau. Cells were maintained at 37 °C in a humidified atmosphere containing 5% CO₂.

HT22 cells or HeLa cells were placed in the cell hyperbaric chamber (customized by Hongyuan Oxygen Industrial) and exposed to 100% oxygen at 0.6 MPa for 1 h. After decompression, cells were collected at various time points (0 h, 3 h, or 6 h) for further analysis. HNO and NC groups were treated similarly as described above.

Total ROS levels were assessed using DCFH-DA (Beyotime (Shanghai, China), S0034S). According to the manufacturer’s instructions, HT22 cells were incubated with 10 μM DCFH-DA at 37 °C for 30 min, and fluorescence was analyzed by flow cytometry. Mitochondrial ROS were measured using MitoSOX Red (Beyotime, S0061S). Cells were incubated with 5 μM MitoSOX Red at 37 °C for 10 min following the manufacturer’s protocol, and fluorescence was analyzed by flow cytometry.

According to the manufacturer’s instructions, HT22 cells were harvested at 0 h or 6 h after HBO exposure, washed with PBS, and stained using the Annexin V-PE Apoptosis Detection Kit (Beyotime, C1065S). Samples were incubated at room temperature for 15 min in the dark, and fluorescence was analyzed by flow cytometry.

ΔΨm was assessed using JC-1 staining (Beyotime, C2003S) for isolated mitochondria and TMRE staining (Macklin (Shanghai, China), 115532-52-0) for cultured cells. For JC-1, mitochondria were isolated from hippocampal tissues and stained according to the manufacturer’s instructions. Fluorescence was observed under a fluorescence microscope, and the red/green fluorescence ratio was calculated to represent ΔΨm. For TMRE, HT22 cells were stained following previously described methods [25], and fluorescence intensity was analyzed by flow cytometry.

Western blotting was performed according to the method described previously [26]. In brief, hippocampal tissues or HT22 cells were lysed in RIPA buffer containing protease and phosphatase inhibitors. Protein concentrations were determined using the BCA assay (Thermo Scientific (Waltham, MA, USA), 23225). Equal amounts of protein were separated by SDS-PAGE and transferred to PVDF membranes. Membranes were blocked with 5% non-fat milk and incubated overnight at 4 °C with primary antibodies against LC3B (AiFang biological, SHDW251215), p62 (Abclone (Wuhan, China), A19700), TOMM20 (Abclone, A19403), caspase-3 (Abclone, A19654), p-p38 (Abclone, AP0526), p38 (Abclone, A4771), c-Fos (Abclone, A24620), and β-actin (Abclone, AC026). After incubation with HRP-conjugated secondary antibodies (CST (Danvers, MA, USA), 7074P2), signals were visualized using enhanced chemiluminescence (MCE (South Brunswick Township, NJ, USA), HY-K1005) and quantified using ImageJ software (v1.54f).

Immunofluorescence was performed according to the method described previously [26]. For colocalization analysis, hippocampal sections were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and blocked with 10% normal goat serum. Samples were incubated with primary antibodies against LAMP1 (ServiceBio (Wuhan, China), GB112949↗) and TOMM20 (Selleckchem (Houston, TX, USA), F0513) overnight at 4 °C, followed by incubation with Alexa Fluor-conjugated secondary antibodies. Nuclei were counterstained with DAPI. Images were acquired using a confocal laser scanning microscope (Zeiss (Oberkochen, Germany), LSM900), and colocalization was quantified using ImageJ software.

All data are presented as mean ± standard error of the mean (SEM) from at least three independent experiments. Statistical analyses were performed using GraphPad Prism 8.0. Comparisons between two groups were conducted using a two-tailed Student’s t-test. Comparisons among multiple groups were performed using one-way ANOVA followed by Tukey’s post hoc test. A p-value < 0.05 was considered statistically significant.