Abstract
Background
Carbon monoxide poisoning (COP) induces systemic hypoxia and oxidative stress–related injury, leading to myocardial injury and persistent cardiac dysfunction. However, reliable biomarkers for monitoring long-term cardiac sequelae and therapeutic responses remain lacking. Extracellular vesicles (EVs), which reflect the molecular status of their cells of origin, may serve as candidate biomarkers for organ-specific injury. This study investigated whether cardiac EV proteins capture COP-induced myocardial and mitochondrial dysfunction and reflect the therapeutic effects of hyperbaric oxygen (HBO) therapy.
Methods
A rat model of COP was established with or without HBO treatment. Cardiac function was assessed by echocardiography, and myocardial injury was evaluated using histological, ultrastructural, and biochemical analyses. Cardiac-enriched EVs isolated from ex vivo whole-heart perfusate were used for global proteomic profiling. Candidate differentially abundant proteins were analyzed with emphasis on pathways related to mitochondrial dynamics, mitochondrial energy metabolism, calcium handling, and myocardial contractility. Key EV-associated and tissue proteins were further validated, and selected candidates were examined in serum-derived EVs as preliminary targeted circulating EV validation.
Results
COP induced significant cardiac dysfunction, as evidenced by reduced ejection fraction and fractional shortening, together with histological myocardial injury, all of which were attenuated by HBO treatment. Proteomic analysis demonstrated that COP reshaped the cardiac EV proteome in a manner consistent with mitochondrial abnormalities, altered calcium-handling protein profiles, and impaired myocardial contractile function. These EV proteomic alterations were enriched in pathways related to mitochondrial dynamics, calcium signaling, and cardiac contractile regulation. Specifically, COP was associated with dysregulation of mitochondrial dynamics regulators, including optic atrophy type 1 (Opa1) and mitochondrial fission protein 1 (FIS1), as well as calcium-handling proteins such as ryanodine receptor 2 (Ryr2) and phospholamban (Pln). Ultrastructural and biochemical analyses showed mitochondrial cristae disruption, altered mitochondrial fusion–fission protein profiles, mitophagy-related protein changes, and pyroptosis-associated signaling in cardiac tissue following COP, whereas HBO mitigated these abnormalities. Notably, EV-associated Opa1 and FIS1 were associated with COP-related alterations in mitochondrial dynamic balance, whereas EV-associated Ryr2 and Pln were associated with impaired myocardial contractile parameters. Additional analysis of EV proteins related to mitochondrial function and ATP energy production further supported COP-associated mitochondrial energy metabolism–related protein remodeling. Targeted analysis of serum-derived EVs further showed that selected calcium-handling proteins, including Ryr2 and Pln, were detectable in circulating EVs and exhibited COP-associated changes consistent with cardiac tissue alterations. These findings support selected cardiac-enriched EV proteins as candidate molecular readouts of COP-associated myocardial, mitochondrial, and contractile abnormalities.
Conclusions
Cardiac EV proteomic remodeling reflects COP-associated mitochondrial and contractile abnormalities and captures the therapeutic effects of HBO. These findings identify cardiac-enriched EV proteins as candidate molecular readouts of myocardial injury and treatment response, providing a cardiac-enriched EV discovery framework for future blood-based biomarker development. The observed alterations in mitochondrial dynamics–, mitochondrial energy metabolism–, and calcium-handling–related proteins provide hypothesis-generating insight into molecular pathways associated with COP-induced cardiac dysfunction. Further validation using circulating EV proteomics, biomarker classifier analyses, and dedicated redox proteomics will be required to establish clinical utility and redox-regulated mechanistic relevance.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s10020-026-01534-0.
Keywords: Carbon monoxide poisoning, Cardiac extracellular vesicles, Proteomics, Mitochondrial dynamics, Calcium signaling, Cardiac dysfunction, Hyperbaric oxygen therapy
Introduction
Carbon monoxide (CO) is a major air pollutant and an important indicator of ambient air quality. At elevated concentrations, it poses a substantial public health threat. Carbon monoxide poisoning (COP) is one of the most common and fatal forms of poisoning worldwide (Rose et al. 2017). In the United States alone, approximately 50,000 patients with COP require emergency department evaluation annually, and about 15,000 of these cases are intentional exposures, accounting for more than two-thirds of COP-related deaths (Hampson 2016). COP causes systemic hypoxia by impairing oxygen transport and utilization, primarily through the formation of carboxyhemoglobin (COHb), which reduces the oxygen-carrying capacity of hemoglobin. This hypoxic insult leads to marked injury in oxygen-dependent organs, particularly the heart (Jeon et al. 2018; Munoz Moreno et al. 2022). Cardiac complications associated with COP include impaired ventricular systolic function (Kalay et al. 2007), myocardial injury, and fibrosis (Cho et al. 2021), all of which may aggravate systemic ischemic-hypoxic damage. Accordingly, cardiac function is likely to be a major determinant of the severity and systemic consequences of COP.
Extracellular vesicles (EVs) are lipid bilayer-bound particles secreted by cells that carry diverse biomolecules, including proteins and nucleic acids (Niel et al. 2022). EVs mediate intercellular communication and participate in both physiological regulation and pathological processes (Hill 2019). Because their molecular cargo reflects the biological status of the cells from which they originate, EVs have emerged as promising biomarker carriers (Urabe et al. 2020). In addition, EVs isolated from biofluids are increasingly being recognized as informative analytes for liquid biopsy-based biomarker discovery and disease monitoring (Kosaka et al. 2019; Dong et al. 2025), and cardiomyocyte-derived circulating EVs may provide a liquid biopsy of the myocardium (Spanos et al. 2024). However, circulating EVs are heterogeneous and may originate from multiple organs and cell types, particularly in systemic toxic injuries such as COP. Direct analysis of circulating EV proteomes alone may therefore be confounded by EVs released from non-cardiac tissues, making it difficult to distinguish cardiac-derived alterations from systemic EV changes. In this context, cardiac-enriched EV discovery approaches may help identify myocardial injury–associated EV protein signatures before subsequent validation in circulating EVs. Despite growing interest in EV-based biomarker platforms, whether cardiac-enriched EV proteins reflect persistent cardiac complications and treatment-associated molecular changes following COP remains insufficiently characterized.
In clinical practice, COP severity is commonly assessed using COHb levels. However, COHb concentrations often correlate poorly with symptom severity and long-term outcomes (Kosaka et al. 2019). Several alternative prognostic indicators, including blood pH (Turner et al. 1999), the serum lactate levels (Lee et al. 2022), the Glasgow Coma Scale scores (Chi et al. 2022), and leukocyte counts (Choi 2001), have been proposed. Nevertheless, these markers mainly reflect acute systemic disturbance and do not adequately capture ongoing or organ-specific injury, particularly in the heart. This limitation supports the development of complementary molecular indicators that may help monitor persistent cardiac injury and treatment response following COP.
Hyperbaric oxygen (HBO) therapy is an established clinical intervention for moderate to severe COP. It involves administration of 100% oxygen at pressures above atmospheric pressure, thereby enhancing oxygen delivery to hypoxic tissues and accelerating CO dissociation from hemoglobin (Thom 2009). HBO has been shown to reduce neurological sequelae and improve short-term outcomes in patients with COP (Rose et al. 2017). However, its effects on long-term cardiac recovery, particularly with respect to mitochondrial homeostasis and intercellular stress signaling, remain insufficiently defined.
To address these gaps, the present study investigated cardiac-enriched EV proteins as candidate molecular readouts of chronic myocardial injury following COP. Using a validated rat model of COP in combination with long-term HBO therapy, we examined whether EVs isolated from ex vivo heart perfusate at Day 28 post-COP reflect persistent myocardial contractile dysfunction and mitochondrial dysregulation. Because COP is a systemic toxic insult, this cardiac-enriched discovery strategy was used to reduce potential confounding from circulating EVs released by non-cardiac tissues. We further evaluated whether selected candidate proteins could be detected in serum-derived EVs as preliminary targeted circulating EV validation. Together, these objectives provide a cardiac-enriched EV discovery framework for identifying candidate protein signatures associated with COP-induced cardiac sequelae and treatment response, while recognizing that comprehensive circulating EV proteomics and biomarker classifier analyses are required for future blood-based biomarker development.
Materials and methods
Animals
Eight-week-old male Wistar rats (BioLASCO company) were raised in the Chi-Mei Animal Center under a 12-h light–dark rhythm under controlled conditions of 25 ± 1 ℃ and stable humidity. Three rats were housed per cage. Rats had unlimited access to pellet chow and tap water. The Institutional Animal Care and Use Committee of Chi Mei Medical Center, Tainan, Taiwan (approval number: 110122316) sanctioned all animal procedures conducted in this study. The experimental design followed the ARRIVE (Kilkenny et al. 2010). An online randomization system (https://www.randomizer.org) was used to allocate the rats to four groups, and they were marked with a nontoxic marker pen at the base of the tail for identification. After COP induction, the animals were removed from the plexiglass chamber as soon as possible (Huang et al. 2021). Experimental animals that experienced severe hypoxia and prolonged coma (> 8 h) were humanely euthanized under deep anesthesia according to the approved institutional animal care protocol.
COP induction
Following the procedures previously established in our laboratory (Huang et al. 2021), the rats were kept in a plexiglass chamber, exposed to CO (1000 ppm for 20 min), and subsequently to 3000 ppm for another 80 min. A gas flow meter maintained a constant CO flow throughout the experiment, and CO levels were monitored with a CO detection device (TPI-708, Korea). Under these exposure conditions, rats become comatose and reach levels of acute severe CO poisoning (carboxyhemoglobin ≥ 40%; COP) (Huang et al. 2021). Rats rendered unconscious were subsequently extracted from within the box to recover sentience by drawing in pure air.
Hyperbaric oxygen (HBO) therapy
HBO therapy was performed in a hyperbaric chamber (Chun Hung International Co., Ltd, Taiwan). Rats received 100% oxygen at 3 atmospheres absolute (ATA) using the following protocol: 20 min of slow compression, three cycles of 20 min at 3 ATA with 10-min air breaks, followed by 20 min of slow decompression. Each session lasted approximately 1.5 h. HBO therapy was administered 5 days per week from Days 1 to 28 after COP. The control group received normobaric air (NBA; 21% O₂ at 1 ATA).
Study framework
Figure 1A illustrates the study design. A total of 57 rats were randomly assigned to four groups. The Sham + NBA group underwent room air exposure and received normobaric air (NBA) treatment (21% O₂ at 1 ATA) for 28 consecutive days. The Sham + HBO group was exposed to room air and treated with hyperbaric oxygen (HBO) at 3.0 ATA for one hour per day over 28 days. The COP + NBA group was exposed to CO and subsequently treated with NBA under the same conditions as the Sham + NBA group. The COP + HBO group was exposed to CO and received HBO therapy (pure oxygen at 3.0 ATA for one hour per day) for 28 consecutive days. Echocardiography (n = 6 per group) and measurements of cardiac injury markers (n = 6 per group) were conducted before COP induction and on Days 1, 7, 14, and 28 after COP; and cardiac tissue and hematoxylin and eosin (H&E) staining were executed on Day 28 after COP (n = 6 per group). On Day 28 post-COP, cardiac-enriched EVs were collected from ex vivo whole-heart perfusate using the Langendorff perfusion system (Claridge et al. 2021) and the size exclusion (SEC) method (Vanderboom et al. 2021). Cardiac EV characteristics were observed via nanoparticle tracking analysis (NTA) (n = 3 per group). Label-free quantification by nano-liquid chromatography-tandem mass spectrometry (nLC-MS/MS) (n = 5 per group) and western blotting (n = 6 per group) were employed to observe the proteomic changes in EVs after COP. In addition, selected candidate proteins identified from cardiac EV analysis, including Ryr2 and Pln, were further examined in serum-derived EVs as targeted preliminary validation of their detectability in circulating EVs.
Fig. 1.
Study design and cardiac injury assessment after COP. A Echocardiography and blood sampling (blue arrows) were performed before COP induction on Day − 1. COP was induced on Day 0, followed by NBA or HBO treatment. Echocardiography and blood sampling were repeated on Days 1, 7, 14, 21, and 28 after COP. After the final assessment on Day 28, animals were euthanized for cardiac tissue collection, cardiac-enriched EV isolation, histological analysis, proteomic profiling, and targeted validation. B–G Time-course measurements of LVIDd, LVIDs, EDV, ESV, FS, and EF before COP induction and on Days 1, 7, 14, 21, and 28 after COP. H Representative H&E-stained cardiac tissue sections on Day 28 after COP. I Quantification of cardiac damage scores on Day 28. J–K Serum troponin I and LDH levels before and after COP. n = 6 animals per group. *P < 0.05 for Sham + NBA versus COP + NBA; #P < 0.05 for COP + NBA versus COP + HBO. Abbreviations: LVIDd, left ventricular internal dimension in diastole; LVIDs, left ventricular internal dimension in systole; EDV, end-diastolic volume; ESV, end-systolic volume; FS, fractional shortening; EF, ejection fraction; LDH, lactate dehydrogenase
Sample sizes were determined based on previous studies (Huang et al. 2021). Rats that died immediately after COP and during the observation period were excluded from further analysis to minimize data skewing from susceptible individuals. In the COP + NBA group, approximately 10% of the animals were excluded from this study. To minimize the potential influence of circadian rhythm on the experimental outcomes, the timing of COP induction and each HBO treatment were standardized to occur between 9:00 AM and 11:00 AM in all experimental groups. Experiments were conducted in the designated research facility at the Animal Center of Chimei Hospital. A one-week acclimation period in a certified animal house preceded the commencement of the animal experiments. At each data measurement stage, random samples were measured to mitigate confounding factors. The analysis of the data was conducted by researchers blinded to the experimental groups. This study has not previously been registered in a research protocol.
Measurement of cardiac damage markers
From the tail vein of rats, one milliliter of blood was drawn and analyzed to establish the baseline concentrations of troponin I, and lactate dehydrogenase (LDH). The measurements were obtained using an immunoassay analyzer (ARCHITECT i2000SR, Abbott Laboratories). Blood samples of 1 mL were collected from the tail vein at specific time points (Days 1, 7, 14, 21, and 28) following COP induction for protein analysis.
H&E staining
On Day 28 after COP induction, rats were euthanized by an overdose of Zoletil (tiletamine–zolazepam, 100 mg/kg, intraperitoneal injection), in accordance with institutional animal care guidelines. Subsequently, the heart was prepared for paraffin embedding. Sagittal sections, each with a thickness of 5 microns, were sliced and then subjected to H&E staining for histopathological analysis. Pathological alterations were identified using a Carl Zeiss light microscope (Jena, Germany). Myocyte damage was graded according to predetermined criteria: 0 when no abnormalities were observed and 1, when isolated early signs of ischemia, such as fiber edema, waviness, or loss of cross-striations were observed; the degree of lesion was categorized using the following grading scale: 0 for no lesions, 1 for a few small lesions observed, none surpassing 0.25 square millimeters in area, 2 indicates the presence of several small lesions or a small number of moderately large lesions, none surpassing 6.25 square millimeters in size and 3 for numerous substantial lesions or larger ones (Chang et al. 2006; Okura et al. 1998).
Echocardiography
Echocardiography was performed before and after COP induction in anesthetized rats using a GE Vivid S6 ultrasound system, which features a 5.0–11.0 MHz sector with specific measurement capabilities. Measurements were obtained using 2D, harmonic, color, M-mode, and pulsed-wave Doppler echo techniques. Various cardiac parameters, including ventricular thickness, dimensions, volumes, ejection fraction, and stroke volume, were analyzed. A control group comprised of healthy rats underwent echocardiography to establish baseline values and temporal variability. Parameters measured included left ventricular internal dimension in systole (LVIDs), left ventricular internal dimension in diastole (LVIDd), left ventricular end-diastolic volume (EDV), end-systolic volume (ESV), left ventricular ejection fraction (EF), and fractional shortening (FS). For each rat, values were averaged across multiple cardiac cycles.
Heart perfusion buffer and serum collection
Cardiac perfusion buffer was collected as described by Bethany Claridge et al. (2021). On Day 28, rats were euthanized by an overdose of Zoletil, facilitating the collection of blood samples from the abdominal cavity. Next, the heart was removed and suspended in the Langendorff circulation system with Krebs–Henseleit buffer (KHB) for 10 min at 37 °C to flush out the blood. The heart was subsequently perfused with KHB containing collagenase II (Merck, SI-C6885, 3 mg/kg), after which the perfusion buffer was collected. Serum samples collected at the endpoint were used for serum-derived EV isolation and targeted validation of selected candidate proteins. These serum-derived EV analyses were performed to assess whether selected cardiac EV candidate proteins could also be detected in circulating EVs. Because the perfusate was collected from the whole heart, the resulting EV preparation was considered cardiac-enriched and may contain EVs released from multiple cardiac-associated cell types, including cardiomyocytes, endothelial cells, fibroblasts, and resident immune cells.
Isolation of cardiac-enriched EVs by size exclusion chromatography (SEC)
Cardiac-enriched EVs were isolated from whole-heart perfusion buffer by size exclusion chromatography (SEC). Briefly, cardiac perfusion buffer (100 mL) was concentrated to 500 μL using a 100 K Amicon® Ultra-15 centrifugal filter (Thermo, UFC9100), followed by EV isolation using a commercially available SEC column (IZON qEV Original, 70 nm) according to the manufacturer’s instructions. Specifically, the two to five fractions were collected, pooled using PBS, and filtered through a 0.2-μm filter. EV particles were collected using a qEV concentrate kit (IZON). The collected EV particles are stored at −80°C, preventing repeated freezing.
Nanoparticle tracking analysis (NTA)
The concentration of cardiac-enriched EVs and the distribution of their particle sizes were characterized using NTA (NanoSight LM10-HS, Malvern). The cardiac EV solution was diluted at least 1:1000 in 0.2 μm-filtered Phosphate buffered saline (PBS) to obtain particles within the desired reading range of 20–100 particles per image. After injecting the diluted sample into the flow cell using a syringe. At an ambient temperature of 21 °C, six regions per sample were randomly selected for 30-s imaging. The particle size and concentration were measured using NanoSight NTA 3.1 software (Malvern Panalytical, Inc., Malvern, UK).
Transmission electron microscopy (TEM), immunogold stain and mitochondria complexity index
Preparation involved coating a 400-mesh copper grid with Formvar, onto which five microliters of cardiac-enriched EVs were applied and left undisturbed for 5 min. The surplus liquid was removed, phosphotungstic acid was added for negative staining. For cardiac tissue, the tissue was swiftly immersed in a fixative for 30 min, sectioned into 1 mm × 1 mm × 2 mm cuboids, and fixed, dehydrated, resin embedded and cut into 70 nm thick sections with an ultramicrotome.
For immunogold staining, a nickel grid was used, anti-FIS1(ab71498; Abcam, UK) and anti-Opa1(ab42364; Abcam, UK) antibodies conjugated to gold nanoparticles were applied, the grids were fixed with 2.5% glutaraldehyde/PBS for 10 min, and the samples were washed, dried, and stained. The slices were viewed using a transmission electron microscope (JEM-1400, JEOL, Japan) at 120 keV equipped with a CCD camera system (Ultrascan 4000 4k × 4k CCD Camera System Model 895, Gatan, USA). The calculation of the mitochondrial complexity index (MCI) was performed as previously described (Williams et al. 2013). Briefly, for each mitochondrion, two orthogonal axes (major and minor) were defined, corresponding to the longest dimension and the maximal perpendicular width, respectively. The number of intersections between mitochondrial cristae and lines drawn along each axis was manually counted, and the MCI was calculated as the sum of cristae intersections along both axes. MCI values are expressed as the mean number of cristae intersections per mitochondrion.
For quantitative analysis of the mitochondrial integrity index (MCI), three animals per group were analyzed. TEM images were randomly selected at a fixed magnification from four fields of view in the apical region of the left ventricle from each animal. All intact and clearly identifiable mitochondria within these fields were included in the MCI calculation. For each mitochondrion, the intersection points of the cristae along the long and short axes were counted and combined to calculate the MCI value.
To quantify Opa1 and FIS1 immunogold labeling, three animals per group were analyzed. TEM images were randomly selected at a fixed magnification from four fields of view in the apical region of the left ventricle of each animal. Opa1-positive and FIS1-positive immunogold particles within all mitochondrial regions in each image were counted. Each data point represents the number of Opa1-positive or FIS1-positive immunogold particles across all mitochondrial regions within a single field of view.
Western blot and ELISA assay
Cardiac tissue, cardiac EV, and serum-derived EV proteins were extracted using RIPA buffer (50 mM Tris, 100 mM sodium deoxycholate, 150 mM NaCl, 1% SDS, 1% NP40). The BCA protein assay kit (ab102536; Abcam, UK) was employed to assess protein concentrations. 10% SDS‒PAGE was employed for sample separation, followed by blotting onto a polyvinylidene fluoride membrane. Overnight incubation at 4 °C was conducted with the membranes and the following primary antibodies: anti-FIS1 (1:1000, ab71498; Abcam, UK), anti-Opa1 (1:1000, ab42364; Abcam, UK), anti-Drp1 (1:1000, ab56788; Abcam, UK), anti-PINK1 (1:1000, ab186303, Abcam, UK), anti-Parkin (1:1000, ab77924, Abcam, UK), anti-NLRP3 (1:1000, ab263899, Abcam, UK), anti-OPTN (1:1000, ab213556, Abcam, UK), anti-LC3B-II (1:1000, ab62721; Abcam, UK), anti-ASC (1:1000, sc-514414, Santa Cruz, USA), anti-GSDMD-N (1:1000, #10137, Cell signaling, USA), IL-1β (1:1000, ab9787, Abcam, UK) and GAPDH (1:10,000, ab8245, Abcam, UK). After washing, the membranes were incubated with HRP-conjugated secondary antibodies at a dilution of 1:10,000 for 1 h at room temperature. Quantification of Western blot signals was performed by scanning with a Hewlett-Packard ScanJet scanner and using ImagePro Plus 6.0 imaging software. For all Western blot analyses, uncropped and unprocessed original blot images are provided in the final section of the Supplemental data. Commercially available ELISA kits were used to quantify Ryr2 (USED024Ra; USCN, China) and Pln (abx518708; Abbexa, UK) protein levels in both cardiac tissue and serum-derived EV. This targeted analysis was used as preliminary circulating EV validation of selected candidate proteins and was not intended to replace comprehensive serum EV proteomics or biomarker classifier analysis.
ATP production analysis
ATP content in cardiac tissue was measured using a commercial ATP assay kit (ab83355; Abcam, UK). Briefly, 10 mg of cardiac tissue from each group was collected, homogenized, and centrifuged to remove tissue and cellular debris. The resulting supernatant was transferred to a clean tube, and the assay enzymes were added sequentially under dark conditions, with incubation for 30 min. ATP levels were determined by measuring the absorbance at 570 nm. All procedures were conducted on ice to maintain sample integrity, and triplicate measurements were taken for each sample to ensure both accuracy and reproducibility.
Sample preparation for proteomics analysis
EV particles were collected using a qEV Concentration Kit (IZON) and lysed with RIPA buffer. In brief, 50 uL of RIPA buffer was added to the 2 uL EV particle pellet (~ 109 particles), and the complete breakdown of the EVs was confirmed after repeated shaking and sonication. Assessment of protein concentration was carried out using a BCA protein assay kit suitable for low concentrations. For label-free quantitative proteomics analysis, 30 μg of proteins precipitated from each sample was alkylated in solution with 10 mM iodoacetamide. Digestion of proteins was initiated at a trypsin-to-protein ratio of 1:50 (w:w), and desalting was subsequently performed using a StageTip C18 (Thermo Fisher Scientific, Germany). This global label-free proteomic workflow was designed to quantify EV protein abundance and was not intended to assess mitochondrial redox status, cysteine oxidation, S-nitrosylation, or other redox-sensitive post-translational modifications. Therefore, proteins related to mitochondrial function, ATP energy production, oxidoreductase-related function, and thioredoxin-related regulation were interpreted based on abundance changes rather than redox-specific modifications.
nLC-MS/MS and data analysis
Protein samples were prepared for digestion and analyzed using an Ultimate System 3000 nanoLC system (Thermo Fisher Scientific, Bremen, Germany) coupled with an Orbitrap Fusion Lumos Tribrid Orbitrap mass spectrometer (Thermo Fisher Scientific, San Jose, CA, USA). A C18 Acclaim PepMap NanoLC column (25 cm in length, 75 µm in inner diameter, packed with 2 μm particles, and with a pore size of 100 Å; Thermo Scientific, San Jose, CA, USA) was utilized for chromatography. The mobile phase consisted of (A) 0.1% formic acid in water and (B) 100% acetonitrile with 0.1% formic acid. The data-dependent acquisition mode was used, and the 15 most intense peptide ions with charge states above 2 were selected for fragmentation via higher-energy collisional dissociation (HCD) at a normalized collision energy (NCE) of 38%. An isolation width of 1.4 Da was used. MS1 and MS/MS spectra were acquired at resolutions of 120,000 and 60,000, respectively, using the Orbitrap. The raw MS/MS spectral data were subsequently subjected to a Mascot search against the rat sequence database in Proteome Discoverer 2.2 (Matrix Science, London, UK; version 2.3) using label-free quantification. Search settings included trypsin digestion, a maximum of 2 missed cleavages, and mass tolerances of 10 ppm for precursor ions and 0.02 Da for MS/MS ions. Carbamidomethyl cysteine was considered a common modification, while oxidized methionine and deamidation of asparagine and glutamine were regarded as variable modifications. A peptide mass tolerance with a false discovery rate (FDR) of less than 0.01 was applied to ensure the reliability of protein identification and minimize false-positive results.
Proteomic data analysis, mediation analysis, and statistics
Software Perseus (version 2.0.5.0) was used to convert proteomic data to a logarithmic scale, ensuring normality and handling missing values via imputation. Several bioinformatics tools were utilized for enrichment analysis of cardiac EV proteins. Using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) bioinformatics resource, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted on cardiac EV proteins. The cluster heatmaps illustrating the expression abundance of cardiac EV proteins were generated using Python. The volcano plot and GO enrichment Bubble Charts were generated in R version 4.2.1 using the ggplot2 package. A Venn diagram was constructed using FunRich software. Mediation analysis was conducted using the Pingouin package in Python, and the significance of indirect effects was assessed via bootstrap resampling. Because the present study was not designed as a causal intervention study, mediation analysis was used only to explore potential statistical associations among EV-associated mitochondrial proteins, calcium-handling proteins, and cardiac functional parameters, rather than to establish causality. Statistical analyses were conducted using GraphPad Prism v8, employing Tukey's multiple comparisons test in one-way ANOVA for echocardiography, cardiac damage marker measurements, NTA, and Western blot analyses. For nLC-MS/MS proteomic analysis, protein abundance comparisons between groups were performed using unpaired Student’s t tests based on log2-transformed protein abundances. Candidate differentially abundant proteins were initially prioritized using nominal p-value thresholds combined with fold-change filtering for exploratory pathway analysis. To address multiple-testing considerations, Benjamini–Hochberg false discovery rate (BH-FDR)-adjusted p-values were subsequently calculated for all 975 quantified proteins in each protein-level comparison and added to the Supplemental Tables 2–6.
In the COP + NBA versus Sham + NBA comparison at Day 28 post-COP, proteins with BH-FDR < 0.05 were further used for an additional GO biological process sensitivity analysis to evaluate whether the major pathway-level biological themes were retained under a stricter multiple-testing threshold. Therefore, the nominal p-value plus fold-change-based protein lists were interpreted as exploratory candidate protein sets, whereas the BH-FDR-based GO biological process analysis was used as a robustness assessment for the main pathway-level interpretation. A p-value of less than 0.05 was considered statistically significant for non-proteomic analyses and was used as a nominal threshold for exploratory proteomic candidate selection.
Results
COP causes heart dysfunction and tissue damage
An overview of the experimental workflow is shown in Fig. 1A. To evaluate whether cardiac-enriched EVs reflect long-term cardiac sequelae following COP, we first established a rat model of sustained cardiac dysfunction after COP induction. Longitudinal echocardiographic assessment revealed progressive left ventricular dilation, characterized by increased left ventricular internal diameters and volumes, accompanied by a persistent decline in ejection fraction (EF) and fractional shortening (FS), indicating impaired systolic contractility (Fig. 1B-G; Supplemental Fig. 1). Notably, HBO therapy significantly ameliorated COP-induced cardiac dysfunction (P < 0.05).
Histopathological analysis at Day 28 post-COP demonstrated pronounced myocardial injury, including inflammatory cell infiltration, ischemic alterations, and focal necrosis, which were quantitatively supported by myocardial damage scoring (Fig. 1H-I). These findings confirm the presence of sustained structural and functional cardiac injury following COP exposure.
We next evaluated the temporal profiles of conventional serum biomarkers of myocardial injury. Lactate dehydrogenase (LDH) and troponin I exhibited transient elevations during the acute phase after COP induction, but returned to baseline by Day 28 (Fig. 1J-K; P < 0.05), despite persistent cardiac dysfunction. These results indicate that conventional serum biomarkers are insufficient for monitoring medium- to long-term cardiac impairment after COP, underscoring the need for alternative biomarkers that reflect sustained myocardial dysfunction. Collectively, this model provides a robust platform for discovering cardiac-enriched EV protein signatures associated with long-term cardiac outcomes following COP.
Cardiac-enriched EV proteome on Day 28 post-COP is associated with mitochondrial dysregulation, contractile dysfunction, and recovery after HBO therapy
To characterize cardiac-enriched EV signatures while minimizing the confounding contribution of EVs from non-cardiac tissues in systemic COP, EVs were isolated from ex vivo perfused hearts using the Langendorff system. Nanoparticle tracking analysis demonstrated that EVs from all groups exhibited comparable size distributions, with mean diameters ranging from 143 to 158 nm and a predominance of small EVs (< 200 nm) (Fig. 2A; Supplemental Table 1). NTA analysis showed no statistically significant differences in EV particle concentration or size distribution among the Sham + NBA, COP + NBA, and COP + HBO groups at Day 28 post-COP. Although the COP + NBA group showed a numerical trend toward higher EV particle concentration, this difference did not reach statistical significance, and EV concentration in the COP + HBO group was comparable to that in the Sham + NBA group. Western blot analysis supported EV identity/enrichment by the presence of CD81 and TSG101 (Fig. 2B), indicating that COP did not alter the general EV characteristics of the cardiac-enriched EV preparations.
Fig. 2.
Characterization and proteomic profiling of cardiac-enriched EVs. A NTA size distribution profiles of cardiac-enriched EVs isolated from ex vivo whole-heart perfusate. B Western blot analysis of EV markers CD81 and TSG101. C Workflow of label-free nLC–MS/MS proteomic analysis. D Venn diagram of EV proteins identified across groups. E Volcano plot of EV protein abundance changes between COP + NBA and Sham + NBA groups. F Heatmap of 975 identified EV proteins across groups. Candidate differentially abundant proteins were selected using nominal P < 0.05 and fold change ≥ 1.5. EV size and concentration data are provided in Supplemental Table 1. Abbreviations: FC(CO/SH), fold change of COP + NBA/Sham + NBA; FC(COH/SH), fold change of COP + HBO/Sham + NBA
Cardiac-enriched EV proteomes from Sham + NBA, COP + NBA, and COP + HBO groups (n = 5 per group) were subsequently analyzed by nLC–MS/MS using label-free quantification. After stringent filtering (FDR < 0.01), 975 proteins were identified across all groups (Fig. 2C; Supplemental Table 2). Comparative analysis of the COP + NBA group versus the Sham + NBA group revealed 295 differentially abundant EV proteins, including 52 COP-upregulated and 243 COP-downregulated proteins, based on nominal P < 0.05 combined with a fold-change threshold of ≥ 1.5 (Fig. 2C, E; Supplemental Tables 3–4). Hierarchical clustering demonstrated a distinct proteomic profile in the COP + NBA group compared with Sham controls (Fig. 2F).
Venn diagram analysis further indicated substantial overlap among groups, with 738 proteins shared across all conditions (Fig. 2D), suggesting that COP primarily remodels EV cargo composition rather than introducing novel protein species. Notably, HBO treatment markedly reversed a subset of COP-induced proteomic alterations. A total of 261 proteins exhibited recovery toward control levels after HBO treatment, including 31 proteins reversed from COP-induced upregulation and 230 proteins reversed from COP-induced downregulation (Fig. 2C, F; Supplemental Tables 5–6). These protein lists were used for downstream enrichment analyses as exploratory candidate protein sets. To further address multiple-testing considerations and evaluate the robustness of COP-associated proteomic changes, BH-FDR-adjusted p-values were calculated for all 975 quantified EV proteins and added to the Supplemental Tables. In the COP + NBA group compared with the Sham + NBA group at Day 28 post-COP, 126 of the 975 quantified cardiac EV proteins remained significantly altered after protein-level BH-FDR correction (BH-FDR < 0.05; Supplemental Table 2). We then performed an additional GO biological process sensitivity analysis using these 126 protein-level BH-FDR-significant COP-altered proteins. This analysis identified GO biological process terms that remained significant after GO term-level adjustment, including heart process, regulation of heart contraction, striated muscle contraction, ATP transport, ATP metabolic process, mitochondrial transport, NADH metabolic process, positive regulation of intracellular transport, protein insertion into ER membrane, and protein localization to endoplasmic reticulum (Supplemental Fig. 2; Supplemental Table 7). These findings support the robustness of the major pathway-level interpretation, particularly for cardiac contraction, mitochondrial/ATP-related metabolism, and organelle-associated protein transport, while the broader nominal p-value-based GO/KEGG results should remain interpreted as exploratory.
To explore the biological significance of COP-altered and HBO-recovered EV candidate proteins identified using nominal p-value and fold-change thresholds, exploratory Gene Ontology enrichment analysis was performed. In the COP + NBA group, EV proteins were enriched in cytosolic, mitochondrial, and endoplasmic reticulum (ER) compartments, whereas these changes were largely normalized following HBO treatment (Fig. 3A). The 261 HBO-recovered proteins were predominantly localized to ribosomes, mitochondria, and the ER and were involved in biological processes including protein translation, ribosome biogenesis, myocardial contraction, and calcium ion release (Fig. 3B-D). Exploratory KEGG pathway analysis showed enrichment in myocardial energy metabolism pathways, such as oxidative phosphorylation and glucagon signaling, as well as pathways associated with cardiomyopathies (Fig. 3E). To further support the involvement of mitochondrial energy metabolism in COP-associated EV proteomic remodeling, we summarized EV proteins related to mitochondrial function and ATP energy production based on global protein abundance changes (Supplemental Table 8). These included proteins associated with the mitochondrial respiratory chain, such as Ndufv3, Ndufa10, and Mtnd5; ATP synthase subunits, including Atp5f1c, Atp5po, Atp5f1d, Atp5f1a, and Atp5pb; mitochondrial homeostasis-related proteins, including Timm13, Higd1a, Maip1, and Hspd1; oxidoreductase- and thioredoxin-related proteins, including Pyroxd2 and Tmx2; and mitochondrial dynamics–related proteins, including Fis1 and Opa1. Although EV-associated Opa1 did not remain significant after BH-FDR correction at the global proteomic screening level, Opa1 was retained as a biologically relevant targeted candidate because of its established role in mitochondrial fusion and cristae maintenance, its functional relationship with the FDR-supported mitochondrial dynamics protein FIS1, and its consistency with tissue-level mitochondrial abnormalities observed by ultrastructural and biochemical analyses.
Fig. 3.
Exploratory GO and KEGG enrichment analyses of cardiac-enriched EV candidate proteins. A GO cellular component enrichment analysis of EV proteins across Sham + NBA, COP + NBA, and COP + HBO groups. B–D GO cellular component, biological process, and molecular function enrichment analyses of HBO-recovered EV proteins. E KEGG pathway enrichment analysis of HBO-recovered EV proteins. For panels B–E, the x-axis indicates enrichment score and the y-axis indicates GO term or KEGG pathway. Dot size represents the number of enriched proteins, and dot color represents enrichment strength
These findings support the presence of mitochondrial energy metabolism–related and oxidoreductase/thioredoxin-related protein abundance changes in cardiac-enriched EVs after COP. However, because the present proteomic workflow was not designed for redox proteomics, these data should not be interpreted as direct evidence of mitochondrial redox status or redox-sensitive protein modifications.
Further stratification of the HBO-recovered proteins revealed that EV proteins with increased abundance after HBO treatment were primarily derived from mitochondria and cardiac structural components, whereas proteins with decreased abundance were mainly associated with ribosomes and the ER (Fig. 4A-B). GO biological process analysis demonstrated that upregulated proteins were enriched in ATP metabolism and myocardial contraction, whereas downregulated proteins were associated with ER calcium release, a key process in myocardial relaxation (Fig. 4C-D).
Fig. 4.
GO enrichment analysis of HBO-recovered EV proteins. A–B GO cellular component enrichment analysis of EV proteins with increased or decreased abundance after HBO treatment. C–D GO biological process enrichment analysis of EV proteins with increased or decreased abundance after HBO treatment. The x-axis indicates enrichment score and the y-axis indicates GO term. Dot size represents the number of enriched proteins, and dot color represents enrichment strength
To assess the relationship between selected EV proteins and cardiac functional parameters, we systematically annotated EV proteins related to mitochondrial function and cardiac contractility. Sixteen proteins were associated with mitochondrial energy metabolism, dynamics, or homeostasis, and eight proteins were related to cardiomyocyte structure or contractile regulation (Table 1). To prioritize EV proteins associated with cardiac functional impairment, correlation analysis identified ten EV proteins statistically associated with both EF and FS (P < 0.05), including proteins involved in mitochondrial dynamics (Opa1, FIS1), mitochondrial homeostasis (Aifm1, Timm13, Ak2, SLC25A46), electron transport (HIGD1A), cardiac structural organization (DMD), and calcium handling (Ryr2, Pln) (Supplemental Fig. 3).
Table 1.
Cardiac EV proteins related to mitochondrial function and cardiac contractility
| List of 24 proteins associated with myocardial contractility, endoplasmic reticulum calcium regulation, and ATP metabolism | ||||||
|---|---|---|---|---|---|---|
| Gene Symbol | Description | Functional Category | Correction with EF | Correction with FS | ||
| R-square | p-value | R-square | p-value | |||
| Opa1 | Dynamin-like 120 kDa protein, mitochondrial | Mitochondrial dynamics | 0.4758 | 0.0273 | 0.4853 | 0.0252 |
| FIS1 | Mitochondrial fission 1 protein | Mitochondrial dynamics | 0.8272 | 0.0003 | 0.7951 | 0.0005 |
| Idh3g | Isocitrate dehydrogenase subunit gamma 1, mitochondrial | TCA cycle | 0.3270 | 0.0842 | 0.3716 | 0.0613 |
| IDH3B | Isocitrate dehydrogenase subunit beta, mitochondrial | TCA cycle | 0.5101 | 0.0203 | 0.3574 | 0.0679 |
| HIGD1A | HIG1 hypoxia inducible domain family member 1A | Electron transport chain | 0.5527 | 0.0137 | 0.4449 | 0.0351 |
| COX7B | Cytochrome c oxidase subunit 7B, mitochondrial | Electron transport chain | 0.3436 | 0.0749 | 0.268 | 0.1254 |
| Ndufa10 | NADH dehydrogenase 1 alpha subcomplex subunit 10, mitochondrial | Electron transport chain | 0.2333 | 0.1573 | 0.143 | 0.2812 |
| Uqcc2 | Ubiquinol-cytochrome-c reductase complex assembly factor 2 | Electron transport chain | 0.0737 | 0.448 | 0.1073 | 0.3555 |
| Aifm1 | Apoptosis-inducing factor 1, mitochondrial | Mitochondrial homeostasis | 0.7597 | 0.001 | 0.7112 | 0.0022 |
| Timm13 | Mitochondrial import inner membrane translocase subunit Tim13 | Mitochondrial homeostasis | 0.8141 | 0.0004 | 0.7192 | 0.0019 |
| Ak2 | Adenylate kinase 2, mitochondrial | Mitochondrial homeostasis | 0.7763 | 0.0008 | 0.7524 | 0.0011 |
| Dhodh | Dihydroorotate dehydrogenase (quinone), mitochondrial | Mitochondrial homeostasis | 0.3755 | 0.0596 | 0.2958 | 0.1041 |
| SLC25A46 | Solute carrier family 25 member 46 | Mitochondrial homeostasis | 0.6975 | 0.0026 | 0.6741 | 0.0036 |
| TIMM44 | Translocase of inner mitochondrial membrane 44 | Mitochondrial homeostasis | 0.3094 | 0.095 | 0.2162 | 0.1757 |
| FASTKD2 | FAST kinase domain-containing protein 2, mitochondrial | Mitochondrial homeostasis | 0.3011 | 0.1004 | 0.2723 | 0.1218 |
| Coq6 | Ubiquinone biosynthesis monooxygenase C | Mitochondrial homeostasis | 0.3020 | 0.0999 | 0.2402 | 0.1504 |
| DMD | Dystrophin | Cardiac structural proteins | 0.5661 | 0.012 | 0.4609 | 0.0309 |
| MYH7 | Myosin Heavy Chain 7, Cardiac Muscle, Beta | Cardiac structural proteins | 0.3758 | 0.0595 | 0.2656 | 0.1274 |
| MYBPC3 | Myosin Binding Protein C, Cardiac Type | Cardiac structural proteins | 0.3692 | 0.0624 | 0.2619 | 0.1305 |
| MYH6 | Myosin Heavy Chain 6, Cardiac Muscle, Alpha | Cardiac structural proteins | 0.0840 | 0.4166 | 0.1204 | 0.326 |
| Ryr2 | Ryanodine Receptor 2 | Cardiac contractility | 0.7326 | 0.0016 | 0.8492 | 0.0002 |
| Pln | Phospholamban | Cardiac contractility | 0.6701 | 0.0038 | 0.7486 | 0.0012 |
| PRKCA | Protein kinase C alpha | Cardiac contractility | 0.3652 | 0.0642 | 0.3882 | 0.0543 |
| Ckm | Creatine Kinase, Muscle Type | Cardiac contractility | 0.0579 | 0.503 | 0.02453 | 0.6657 |
Bold indicates the 10 cardiac EVs significantly associated with both EF and FS (P < 0.05)
Given the central role of ER calcium handling in cardiac contractility, mediation analyses were performed to explore whether the associations between mitochondrial-related EV proteins and cardiac functional parameters were statistically linked to ryanodine receptor 2 (Ryr2) or phospholamban (Pln) expression. Proteins involved in mitochondrial dynamics and homeostasis showed a higher frequency of significant mediation effects than those in other functional categories (Supplemental Figs. 4–6). Bootstrap validation identified Opa1 as the only protein showing a robust statistical mediation pattern exceeding 50% in the association models involving Ryr2, Pln, EF, and FS (Fig. 5A-D; Table 2).
Fig. 5.
Mediation analysis of EV-associated Opa1, calcium-handling proteins, and cardiac functional parameters. A Opa1–Pln–EF mediation model. B Opa1–Pln–FS mediation model. C Opa1–Ryr2–EF mediation model. D Opa1–Ryr2–FS mediation model. Indirect effects were assessed using bootstrap resampling. Mediation analyses were exploratory and do not establish causality
Table 2.
Bootstrap-validated mediation effects of candidate proteins on cardiac function (EF and FS) via Ryr2- and Pln-dependent pathways
| Mediation analysis assessing the effects of six candidate proteins on EF and FS mediated by Ryr2 or Pln | ||||
|---|---|---|---|---|
| A to M to B | Indirect effect | Total effect | Mediation effect | p-value |
| Opa1_Ryr2_EF | −0.000000034 | −4.56E-08 | 74.5614% | 0.0108 |
| Opa1_Ryr2_FS | −6.88E-08 | −7.99E-08 | 86.1076% | 0.0096 |
| Opa1_Pln_EF | −2.77E-08 | −4.56E-08 | 60.7456% | 0.0664 |
| Opa1_Pln_FS | −5.36E-08 | −7.99E-08 | 67.0839% | 0.0464 |
| FIS1_Ryr2_EF | 8.56E-08 | 2.64E-07 | 32.4242% | 0.0636 |
| FIS1_Ryr2_FS | 1.91E-07 | 4.60E-07 | 41.5217% | 0.0432 |
| FIS1_Pln_EF | 6.88E-09 | 2.64E-07 | 2.6061% | 0.948 |
| FIS1_Pln_FS | 1.19E-07 | 4.60E-07 | 25.8696% | 0.5984 |
| Aifm1_Ryr2_EF | −4.56E-09 | −1.13E-08 | 40.3540% | 0.3012 |
| Aifm1_Ryr2_FS | −1.25E-08 | −1.89E-08 | 66.1376% | 0.3092 |
| Aifm1_Pln_EF | −3.63E-09 | −1.13E-08 | 32.1239% | 0.0308 |
| Aifm1_Pln_FS | −7.91E-09 | −1.89E-08 | 41.8519% | 0.0296 |
| Timm13_Ryr2_EF | −1.95E-08 | −5.87E-08 | 33.2198% | 0.4332 |
| Timm13_Ryr2_FS | −5.57E-08 | −9.58E-08 | 58.1420% | 0.3512 |
| Timm13_Pln_EF | −1.67E-08 | −5.87E-08 | 28.4497% | 0.0172 |
| Timm13_Pln_FS | −3.70E-08 | −9.58E-08 | 38.6221% | 0.02 |
| AK2_Ryr2_EF | −1.08E-08 | −2.94E-08 | 36.7347% | 0.2632 |
| AK2_Ryr2_FS | −3.00E-08 | −5.02E-08 | 59.7610% | 0.2404 |
| AK2_Pln_EF | −9.19E-09 | −2.94E-08 | 31.2585% | 0.1088 |
| AK2_Pln_FS | −2.02E-08 | −5.02E-08 | 40.2390% | 0.0908 |
| SLC25A46_Ryr2_EF | 4.75E-07 | 1.09E-06 | 43.5780% | 0.1424 |
| SLC25A46_Ryr2_FS | 9.89E-07 | 1.85E-06 | 53.4595% | 0.1384 |
| SLC25A46_Pln_EF | 4.05E-07 | 1.09E-06 | 37.1560% | 0.7208 |
| SLC25A46_Pln_FS | 1.45E-06 | 1.85E-06 | 78.3784% | 0.6504 |
Collectively, these results indicate that COP induces sustained remodeling of cardiac-enriched EV proteomes, characterized by mitochondrial and ER-associated protein alterations that are associated with impaired cardiac contractile parameters. HBO treatment partially reversed these changes. Together, the correlation and mediation analyses suggest that mitochondrial dynamics–related proteins, particularly Opa1, are associated with ER calcium-handling proteins and impaired cardiac contractile parameters following COP. However, these findings should be interpreted as hypothesis-generating associations rather than evidence of direct causality.
Cardiac-enriched EVs reflect sustained mitochondrial damage and altered mitochondrial dynamics following COP
Given that cardiac contractility is highly dependent on mitochondrial ATP production and tightly regulated by mitochondrial dynamics, we next examined whether the EV proteomic alterations observed at Day 28 post-COP were associated with sustained mitochondrial structural and functional abnormalities in cardiac tissue. Proteomic analysis revealed significant changes in EV proteins related to ATP metabolism, mitochondrial respiratory complexes, ATP synthase subunits, mitochondrial homeostasis, and fusion–fission regulation (Table 1), prompting further investigation of mitochondrial integrity at this time point.
Transmission electron microscopy demonstrated well-preserved mitochondrial morphology, characterized by intact double membranes and organized cristae, in the Sham + NBA group. In contrast, mitochondria from the COP + NBA group exhibited marked structural abnormalities, including disruption of the outer membrane and loss of cristae. These alterations were partially reversed following HBO treatment, with mitochondrial morphology resembling that of Sham controls (Fig. 6A; Supplemental Fig. 7). Quantitative analysis using the mitochondrial complexity index showed a significant reduction in mitochondrial cristae complexity after COP, which was restored by HBO treatment (Fig. 6B; P < 0.05).
Fig. 6.
Mitochondrial morphology and related protein changes in cardiac tissue and cardiac-enriched EVs on Day 28 after COP. A Representative TEM images of cardiac mitochondria. B Quantification of mitochondrial complexity index (MCI). C Representative immunogold staining images for Opa1 and FIS1. Yellow arrows indicate Opa1-positive particles; red arrows indicate FIS1-positive particles. D–E Quantification of Opa1-positive and FIS1-positive immunogold particles. F Cardiac ATP content. G, J–K Western blot analysis and quantification of mitochondrial fission/fusion-related proteins in cardiac tissue and cardiac-enriched EVs. H, L Western blot analysis and quantification of mitophagy-related proteins. I, M Western blot analysis and quantification of pyroptosis-related proteins. For panel B, n = 3 animals per group, with four fields of view analyzed per animal; each data point represents the MCI value of one intact and clearly identifiable mitochondrion within the analyzed fields. For panels D–E, n = 3 animals per group, with four fields of view analyzed per animal; each data point represents the number of Opa1-positive or FIS1-positive immunogold particles counted within mitochondrial regions in a single field of view. Data points with the same color indicate measurements obtained from the same animal. Statistical comparisons were performed using animal-level mean values. For panels F and J–M, n = 6 animals per group. *P < 0.05 for Sham + NBA versus COP + NBA; #P < 0.05 for COP + NBA versus COP + HBO
To further assess mitochondrial fusion–fission balance, immunogold labeling of Opa1 and FIS1 was performed in cardiac tissue. In the Sham + NBA group, Opa1 labeling was more prominent than FIS1 labeling, whereas the COP + NBA group showed a pattern of increased FIS1 labeling and reduced Opa1 labeling. HBO treatment partially restored the predominance of Opa1 over FIS1 labeling (Fig. 6C-E). These findings were corroborated by western blot analysis, which demonstrated upregulation of the fission-related proteins Drp1 and FIS1 and downregulation of Opa1 in cardiac tissue following COP, with partial reversal after HBO treatment (Fig. 6G, J; P < 0.05).
Consistent with impaired mitochondrial integrity, cardiac ATP levels were significantly reduced in the COP + NBA group and partially restored following HBO treatment (Fig. 6F; P < 0.05). Notably, proteomic profiling and immunoblotting of cardiac-enriched EVs revealed an inverse abundance pattern of mitochondrial dynamic proteins compared with cardiac tissue, with increased EV-associated Opa1 and decreased EV-associated FIS1 following COP (Fig. 6G, K; Table 3; P < 0.05). Because these EVs were isolated from whole-heart perfusate, this EV-associated Opa1 increase should be interpreted as a cardiac-enriched EV cargo alteration rather than evidence of increased intracellular Opa1 expression in cardiomyocytes.
Table 3.
Proteomic profiling of cardiac EVs-associated Opa1 and FIS1 following COP
| Protein Symbol | Aboundance Average | Fold Change (COP + NBA/Sham + NBA) | p-value (COP + NBA VS SHAM + NBA) | Fold Change (COP + HBO/Sham + NBA) | ||
|---|---|---|---|---|---|---|
| Sham + NBA | COP + NBA | COP + HBO | ||||
| Opa1 | 69,251,242.24 | 210,317,915.6 | 75,054,637.31 | 3.04 | 0.0398 | 1.08 |
| FIS1 | 73,297,087.88 | 29,997,520.73 | 76,705,741.9 | 0.41 | 0.0002 | 1.05 |
Abbreviations: NBA Normobaric air, COP Carbon monoxide poisoning
In parallel with mitochondrial structural and functional impairment, COP increased the abundance of mitophagy-related proteins, including PINK1, Parkin, OPTN, and LC3B-II (Fig. 6H, L; P < 0.05), suggesting changes in mitophagy-related mitochondrial quality-control signaling. Furthermore, increased expression of NLRP3, ASC, GSDMD-N, and IL-1β in the COP + NBA group suggested activation of inflammasome-associated pyroptotic signaling (Fig. 6I, M; P < 0.05).
Collectively, these results demonstrate that COP is associated with sustained mitochondrial structural damage, altered mitochondrial dynamics, and impaired energy metabolism in cardiac tissue at Day 28 post-COP. These tissue-level abnormalities were accompanied by changes in EV-associated mitochondrial proteins, suggesting that cardiac EV cargo may reflect COP-related mitochondrial alterations.
Targeted validation of Ryr2 and Pln alterations in cardiac tissue and serum-derived EVs following COP
Based on the mediation analyses suggesting statistical associations among mitochondrial-related EV proteins, Ryr2, Pln, and impaired cardiac functional parameters, we selected Ryr2 and Pln for targeted validation and examined whether alterations in these calcium-handling proteins
were detectable in cardiac tissue and serum-derived EVs at Day 28 post-COP. Protein expression levels of Ryr2 and Pln were first assessed in cardiac tissue (Fig. 7A-B; P < 0.05). Both proteins were significantly reduced in the COP + NBA group compared with Sham controls, and these reductions were partially attenuated following HBO treatment.
Fig. 7.
Targeted validation of Ryr2 and Pln protein levels in heart tissue and serum-derived EVs. A–B Ryr2 and Pln protein levels in cardiac tissue. C–D Ryr2 and Pln protein levels in serum-derived EVs. n = 6 animals per group. *P < 0.05 for Sham + NBA versus COP + NBA; #P < 0.05 for COP + NBA versus COP + HBO
Before assessing Ryr2 and Pln in serum-derived EVs, serum EV protein concentration was measured to evaluate sample yield and loading consistency. No significant differences were observed among groups (Sham + NBA vs. COP + NBA, P = 0.3078; COP + NBA vs. COP + HBO, P = 0.1639; Sham + NBA vs. COP + HBO, P = 0.913; Supplemental Fig. 8). We then evaluated whether these selected candidate proteins were detectable and altered in serum-derived EVs (Fig. 7C-D). Although HBO treatment did not result in statistically significant changes in serum EV-associated Ryr2 and Pln levels, COP was associated with reduced abundance of both proteins in serum-derived EVs compared with Sham animals.
Notably, the direction of change observed in serum-derived EVs was consistent with that detected in cardiac tissue, suggesting that selected EV-associated calcium-handling proteins may reflect COP-associated alterations in myocardial protein abundance. These findings provide preliminary targeted evidence that selected cardiac EV candidate proteins, particularly Ryr2 and Pln, can also be detected in circulating EVs and show changes consistent with cardiac tissue alterations after COP. However, because comprehensive serum EV proteomics was not performed, these data should be interpreted as targeted circulating EV validation rather than definitive blood-based biomarker confirmation. In addition, direct measurements of intracellular or sarcoplasmic reticulum calcium flux were not performed; therefore, these findings indicate altered calcium-handling protein abundance rather than direct evidence of calcium-handling dysfunction.
Discussion
COP remains a major public health concern, largely due to its detrimental effects on vital organs such as the heart and brain. Clinical and experimental studies have demonstrated that CO-induced myocardial injury is associated with subsequent neurological complications (Lin et al. 2018), as well as ventricular dysfunction, myocardial damage, and fibrosis, thereby exacerbating hypoxic injury in organs that rely on adequate systemic oxygen delivery (Cha et al. 2016). In the present study, although conventional serum biomarkers of acute myocardial injury, including LDH and troponin I, had returned to baseline by day 28 after COP, comprehensive histopathological and echocardiographic analyses revealed persistent myocardial damage and sustained impairment of cardiac contractile function. These findings indicate that conventional acute-phase biomarkers inadequately reflect long-term cardiac remodeling following COP, underscoring the need for more sensitive and mechanism-informed indicators to monitor chronic cardiac sequelae. In this context, the present study used ex vivo cardiac perfusion EVs as a cardiac-enriched discovery platform to identify myocardial injury–associated EV protein signatures. Because COP is a systemic toxic insult that may affect multiple organs and cell types, direct circulating EV proteomics may be confounded by EVs released from immune cells, vascular endothelium, liver, brain, or other injured tissues. Therefore, the ex vivo whole-heart perfusion approach was used to reduce systemic confounding and to better define cardiac-associated EV protein alterations before targeted evaluation of selected candidates in serum-derived EVs. This approach provides cardiac-enriched EV signatures relevant to myocardial injury, while recognizing that the EV preparations may include vesicles released from multiple cardiac-associated cell types rather than purified cardiomyocyte-derived EVs. Because EVs were collected at Day 28 post-COP, the EV proteomic signatures identified in this study should be interpreted as molecular readouts of persistent cardiac injury and recovery after COP rather than as acute or COP-specific diagnostic markers. Future studies comparing acute post-COP EV profiles with chronic-phase EV profiles, as well as EV signatures from other cardiac injury models such as myocardial infarction or ischemia/reperfusion, will be required to determine the temporal dynamics and disease specificity of these EV protein signatures.
EVs have been increasingly recognized as critical mediators and reporters of intercellular communication in both physiological and pathological contexts (Shah et al. 2018). However, whether cardiac EV cargo reflects the severity and persistence of COP-induced cardiac dysfunction remains largely unexplored. In this study, we identified ten cardiac EV proteins whose expression levels were strongly correlated with EF and FS, two key parameters of myocardial contractile performance. Among these, Ryr2 and Pln are central regulators of sarcoplasmic reticulum calcium handling, a process fundamental to excitation–contraction coupling. Ryr2 functions as the principal calcium-release channel in the sarcoplasmic reticulum, and its dysregulation predisposes to arrhythmias and impaired cardiac contractility (Bround et al. 2012). Pln modulates SERCA2 activity and plays an essential role in maintaining calcium cycling and diastolic function (Rodriguez and Kranias 2005; Hagemann and Xiao 2002), with pathogenic variants being closely linked to heart failure and contractile dysfunction (Haghighi et al. 2006). The altered abundance of these calcium-regulatory proteins in cardiac-enriched EVs observed in this study is consistent with the echocardiographic evidence of impaired systolic function, suggesting that EV-associated calcium-handling proteins may reflect cardiac tissue-level disturbances in myocardial contractile machinery following COP. From a translational perspective, blood-based EV biomarkers would be more directly applicable than EVs isolated from ex vivo cardiac perfusate. Nevertheless, the cardiac-enriched discovery approach used here allowed us to identify candidate EV proteins associated with myocardial dysfunction under reduced systemic confounding. Our targeted analysis of serum-derived EVs showed that Ryr2 and Pln were detectable in circulating EVs and exhibited changes directionally consistent with cardiac tissue alterations, providing preliminary evidence that selected cardiac EV candidates may be measurable in blood. However, these findings do not replace comprehensive serum or plasma EV proteomics or classifier-based validation.
From a clinical perspective, EV-based cardiac biomarkers may be most relevant for patients with moderate-to-severe COP, particularly those with myocardial injury, abnormal electrocardiographic or echocardiographic findings, persistent cardiopulmonary symptoms, older age, or pre-existing cardiovascular risk factors. This is important because COHb levels are useful for confirming exposure but do not reliably predict clinical severity or long-term outcomes, and acute cardiac injury markers may normalize despite persistent myocardial remodeling. Myocardial injury after moderate-to-severe COP has been associated with increased long-term mortality, and cardiac magnetic resonance studies have shown persistent myocardial injury or fibrosis after acute COP (Cho et al. 2021; Henry et al. 2006). In addition, population-based studies indicate that CO poisoning is associated with increased subsequent cardiovascular risk, including arrhythmia, coronary artery disease, and congestive heart failure (Huang et al. 2021; Lee et al. 2015). Thus, EV-associated cardiac proteins may complement current clinical tools, including COHb, troponin, electrocardiography, and echocardiography, by identifying patients who may benefit from longitudinal cardiac surveillance after discharge. Because our study evaluated EV protein changes at Day 28 post-COP, these candidate biomarkers are best interpreted as molecular readouts for subacute-to-long-term monitoring rather than acute triage markers. Future studies with serial circulating EV sampling will be required to evaluate their utility for acute risk stratification, recovery monitoring, and assessment of HBO treatment response.
In parallel, we observed that EV-associated levels of the mitochondrial dynamics–related proteins Opa1 and FIS1 were also strongly correlated with EF and FS. Mitochondrial fission and fusion are essential adaptive processes that preserve mitochondrial integrity under cellular stress. Fusion facilitates the complementation of partially damaged mitochondria, whereas fission supports mitochondrial quality control, biogenesis, the elimination of dysfunctional organelles, and the initiation of apoptosis under severe stress conditions (Youle and Bliek 2012). Importantly, the balance between fission and fusion is tightly coupled to mitochondrial ATP production efficiency, with fusion generally promoting, and excessive fission impairing, ATP generation (Rambold et al. 2011; Gomes et al. 2011; Jheng et al. 2012). In the present study, transmission electron microscopy revealed marked disruption or loss of mitochondrial cristae following COP, structures that are critical for oxidative phosphorylation and ATP synthesis. This structural abnormality is consistent with the observed reduction in Opa1 levels, a protein essential for cristae maintenance (Varanita et al. 2015), as well as increased mitochondrial fission detected by immunogold labeling. Given that approximately 95% of myocardial ATP is generated through mitochondrial oxidative phosphorylation under physiological conditions (Doenst et al. 2013), disruption of mitochondrial ultrastructure and dynamics may contribute to impaired cardiac contractile function. In addition to mitochondrial dynamics–related proteins, our supplementary analysis (Supplemental Table 8) identified multiple EV proteins associated with mitochondrial respiratory chain function, ATP synthase activity, mitochondrial homeostasis, and oxidoreductase/thioredoxin-related regulation. These protein abundance changes provide additional support for the interpretation that COP is associated with mitochondrial energy metabolism disturbance and oxidoreductase/thioredoxin-related protein remodeling. Importantly, these pathway-level interpretations were further supported by independent experimental observations, including mitochondrial ultrastructural changes, ATP measurements, western blot validation of mitochondrial dynamics–related proteins, and echocardiographic evidence of contractile dysfunction. Nevertheless, these results were derived from global protein abundance profiling and do not directly measure mitochondrial redox status or redox-sensitive post-translational modifications. Therefore, the observed oxidoreductase- and thioredoxin-related protein changes should be interpreted as supportive abundance-based evidence of redox-related protein remodeling rather than direct redox proteomics evidence.
Mediation analysis further suggested a potential statistical association among mitochondrial fusion-related proteins, endoplasmic reticulum (ER) calcium-handling proteins, and cardiac contractile performance. Mitochondria and the ER form specialized contact sites known as mitochondria-associated ER membranes, which play critical roles in calcium signaling, autophagy, and lipid metabolism (Yang et al. 2020; Degechisa et al. 2022). Previous studies have demonstrated that reduced Opa1 expression is associated with diminished calcium content in both mitochondria and the sarcoplasmic reticulum (Hinton et al. 2024), highlighting the importance of mitochondrial–ER crosstalk in maintaining calcium homeostasis. Our findings suggest that COP may be associated with alterations in inter-organelle communication, calcium-handling protein abundance, and sustained impairment of myocardial contractility. While prior investigations of CO toxicity have largely focused on direct mitochondrial injury (Alonso et al. 2003) or documented cardiac functional outcomes (Henry et al. 2006; Suner and Jay 2008), the molecular pathways underlying long-term cardiac dysfunction have remained insufficiently defined. The present study suggests a potential association among COP-induced alterations in mitochondrial dynamics, ER calcium-handling proteins, and chronic myocardial contractile dysfunction. However, direct causal links among these processes remain to be experimentally validated.
Emerging evidence indicates that EVs can carry mitochondrial components, including mitochondrial DNA, mitochondrial proteins, and even intact mitochondria, thereby linking EV biology to mitochondrial communication and quality-control pathways (Wu et al. 2024). In addition, MDVs have been increasingly recognized as a mitochondrial quality-control mechanism involving selective cargo sorting and inter-organelle or extracellular trafficking of mitochondrial components (Konig and McBride 2024; Picca et al. 2023). D’Acunzo and colleagues demonstrated the presence of Opa1 within mitochondria-derived vesicles (MDVs) (D’Acunzo et al. 2021), and Todkar et al. further showed that Opa1- and Snx9-dependent MDVs are involved in directing mitochondrial proteins toward multivesicular bodies for subsequent EV release (Todkar et al. 2021). In the present study, we observed an inverse abundance pattern of Opa1, with reduced Opa1 levels in cardiac tissue but increased Opa1 levels in cardiac-enriched EVs following COP. This finding is particularly intriguing because Opa1 is essential for mitochondrial inner membrane fusion and cristae maintenance; therefore, reduced tissue Opa1 is consistent with the mitochondrial cristae disruption and impaired ATP production observed after COP. In contrast, increased EV-associated Opa1 may reflect the selective export of Opa1-containing mitochondrial components through MDV-related EV pathways during sustained mitochondrial stress. Therefore, the increased Opa1 abundance detected in EVs should not be interpreted as increased cellular Opa1 expression, but rather as a potential redistribution or selective packaging of Opa1-containing mitochondrial cargo into cardiac-enriched EVs. Rather than representing simple passive leakage of damaged mitochondrial proteins, this inverse tissue–EV pattern may indicate a stress-responsive mitochondrial quality-control process in which cardiac-associated cells remove or redistribute mitochondrial components via EV-mediated export. This interpretation is also consistent with the concept that MDV or mitochondrial EV release may limit the intracellular accumulation or extracellular uncontrolled release of mitochondrial damage-associated molecular patterns under stress conditions (Picca et al. 2023; Heyn et al. 2023; Iorio et al. 2024).
The coexistence of increased mitophagy-related protein abundance and increased EV-associated mitochondrial proteins is not necessarily contradictory. Rather, mitophagy and mitochondria-derived vesicle (MDV)/EV-mediated mitochondrial cargo export may represent complementary mitochondrial quality-control pathways operating at different levels of mitochondrial stress or damage. MDVs have been proposed as an early and cargo-selective quality-control mechanism capable of removing specific damaged mitochondrial proteins or subcompartments before whole-organelle degradation becomes necessary (Peng et al. 2022), whereas mitophagy is generally involved in the clearance of more extensively damaged mitochondria. In addition, MDV biogenesis and cargo sorting are regulated processes involving mitochondrial trafficking and fission-related proteins (Konig et al. 2021), and experimental studies suggest functional overlap between MDV-associated pathways and mitophagy-related mitochondrial quality-control mechanisms (Towers et al. 2021). Accordingly, the increased abundance of PINK1, Parkin, OPTN, and LC3B-II observed in the present study may reflect changes in mitophagy-related mitochondrial quality-control signaling rather than definitive evidence of complete mitophagic flux, whereas increased EV-associated Opa1 may reflect stress-associated selective export of mitochondrial components through MDV-related EV pathways during sustained mitochondrial dysfunction.
Although EV-mediated mitochondrial transfer has been proposed to improve mitochondrial function in recipient cells (Ikeda et al. 2021), the functional consequences of increased mitochondrial protein release via EVs in the context of COP remain to be elucidated. Notably, we also observed a reduction in EV-associated FIS1, a key mediator of mitochondrial fission. Given the tight regulation required to maintain mitochondrial dynamic equilibrium (Youle and Bliek 2012), these findings raise the possibility that EV secretion may be associated with intracellular remodeling of mitochondrial dynamics, potentially through differential representation of fusion- and fission-associated proteins in EV cargo. However, whether EV-mediated export of mitochondrial proteins directly contributes to mitochondrial fragmentation remains unknown. Collectively, our data suggest that EV cargo composition may serve as an informative readout of intracellular mitochondrial status under COP-induced stress.
Hyperbaric oxygen (HBO) therapy, which delivers high concentrations of oxygen under elevated atmospheric pressure (Edwards 2010), is widely recognized as an effective treatment for moderate- to severe COP, particularly in reducing long-term neurological complications (Casillas et al. 2019). Accumulating evidence indicates that HBO therapy can also restore mitochondrial energy metabolism. For example, Hu et al. reported that HBO therapy significantly increased ATP and NAD⁺ levels in a rat model of cerebral embolism (Hu et al. 2017). In the present study, HBO therapy markedly ameliorated COP-induced cardiac mitochondrial damage, improved mitochondrial morphology, and partially restored ATP synthetic capacity. Moreover, HBO treatment corrected COP-associated abnormalities in cardiac structural proteins and calcium-regulatory proteins, including those identified within cardiac EVs. These findings extend previous observations by showing that HBO treatment is associated with improved mitochondrial morphology, partial restoration of ATP production, and normalization of EV-associated mitochondrial and calcium-handling protein signatures. However, whether these molecular changes directly mediate the cardioprotective effects of HBO, or whether they involve redox-regulated mitochondrial protein modifications, requires further mechanistic validation using targeted functional assays and dedicated redox proteomics approaches.
This study represents, to our knowledge, the first comprehensive characterization of cardiac-enriched EVs isolated 28 days after COP, with the aim of evaluating their utility as indicators of long-term cardiac sequelae and therapeutic response to HBO. Several limitations should be acknowledged.
First, EVs were isolated from whole-heart ex vivo perfusate as a cardiac-enriched discovery approach to reduce confounding from non-cardiac circulating EVs in systemic COP. However, this approach limits direct extrapolation to blood-based biomarker applications and does not allow definitive attribution of EVs to cardiomyocytes or specific cardiac cell types. Because cardiomyocyte-derived EV purity was not assessed using markers such as POPDC2, CHRNE, or TNNT2, these EV preparations should be interpreted as cardiac-enriched mixed cardiac-cell EV isolates rather than purified cardiomyocyte-derived EVs. Future studies using cardiomyocyte-specific EV lineage-tracing models, such as cardiomyocyte-specific tdTomato or mitoDendra2 transgenic mice, will be valuable for definitively tracking cardiomyocyte-derived EVs and validating cardiomyocyte-specific mitochondrial-contractile mechanisms. Although targeted measurement of Ryr2 and Pln in serum-derived EVs provided preliminary circulating EV validation, comprehensive serum or plasma EV proteomics and biomarker classifier analysis will be required in future studies to establish their clinical utility.
Second, correlation and mediation analyses identified associations among EV-associated mitochondrial dynamics proteins, calcium-handling proteins, and cardiac functional parameters, but do not establish causality. Because gain- or loss-of-function experiments targeting Opa1, FIS1, Ryr2, or Pln and direct calcium flux measurements were not performed, the proposed relationship among mitochondrial dynamics, calcium-handling protein alterations, and contractile dysfunction should be interpreted as a hypothesis-generating candidate mechanism requiring further validation. Although increased levels of mitophagy-related proteins were observed, mitophagic flux and mitochondria–lysosome co-localization were not directly assessed; therefore, these findings should be interpreted as mitophagy-related protein changes rather than direct evidence of enhanced mitophagy.
Third, the present global proteomic workflow was not designed to assess mitochondrial redox status or redox-sensitive post-translational modifications. Although the supplementary mitochondrial protein analysis supports mitochondrial energy metabolism–related and oxidoreductase/thioredoxin-related protein abundance changes, these findings should not be interpreted as direct evidence of redox proteomic alterations. Dedicated redox proteomics approaches will be required in future studies to determine whether redox-regulated mitochondrial protein modifications contribute to COP-induced cardiac dysfunction and the therapeutic response to HBO.
Finally, the initial protein candidate selection for GO and KEGG enrichment analyses was based on nominal p-values combined with fold-change filtering because of the limited sample size and exploratory nature of this EV proteomic discovery study. We acknowledge that this approach may increase the possibility of false-positive protein candidates. To address this limitation, BH-FDR-adjusted p-values were recalculated for all 975 quantified cardiac EV proteins and added to the Supplemental Tables 2–6. In the COP + NBA versus Sham + NBA comparison at Day 28 post-COP, 126 proteins remained significantly altered after protein-level BH-FDR correction. In addition, an FDR-based GO biological process sensitivity analysis was performed using these 126 protein-level BH-FDR-significant COP-altered proteins. This stricter analysis identified GO biological process terms that remained significant after GO term-level adjustment, particularly those related to cardiac contraction, ATP metabolism, mitochondrial transport, NADH metabolism, intracellular transport, and ER-associated protein localization. These findings support the robustness of the main pathway-level interpretation, especially for COP-associated cardiac contractile dysfunction and mitochondrial/ATP-related metabolic remodeling. Nevertheless, the broader nominal p-value-based GO/KEGG results, including additional mitochondrial quality-control and calcium-related terms, should still be interpreted as exploratory and hypothesis-generating. Future studies with larger sample sizes and independent validation cohorts will be required to confirm these protein candidates and pathway-level findings.
Conclusions
This study demonstrates that cardiac extracellular vesicles (EVs) reflect COP-associated alterations in mitochondrial dynamics–related, mitochondrial energy metabolism–related, and calcium-handling proteins during chronic cardiac injury following COP. These EV-associated protein signatures were partially restored by HBO therapy and were accompanied by improved mitochondrial morphology, ATP production, and cardiac functional parameters. Collectively, our findings establish a cardiac-enriched EV proteomic framework for identifying molecular readouts of COP-associated myocardial injury and treatment response. Importantly, this work provides proof-of-concept evidence supporting cardiac EV proteins as candidate biomarkers for future blood-based monitoring of COP-associated myocardial injury. While the observed associations among mitochondrial dynamics, calcium-handling protein alterations, and cardiac dysfunction remain hypothesis-generating, they highlight biologically relevant pathways for future mechanistic investigation. Further studies using circulating EV proteomics, biomarker classifier analyses, and dedicated redox proteomics will be required to establish clinical utility and redox-regulated mechanistic relevance.
Supplementary Information
Acknowledgements
The authors acknowledge the Consortia of Key Technologies and Instrumentation Center, National Taiwan University for the provision of mass spectrometry technical research services. The authors gratefully acknowledge the JEM-1400 of the machine equipment belonging to the Instrument Development Center of the National Cheng Kung University.
Abbreviations
- CO
Carbon monoxide
- COP
Carbon monoxide poisoning
- COHb
Carboxyhemoglobin
- EVs
Extracellular vesicles
- HBO
Hyperbaric oxygen
- NBA
Normobaric air
- H&E
Hematoxylin and eosin
- SEC
Size exclusion
- NTA
Nanoparticle tracking analysis
- nLC-MS/MS
Nano-liquid chromatography-tandem mass spectrometry
- LDH
Lactate dehydrogenase
- LVIDs
Left ventricular internal dimension in systole
- LVIDd
Left ventricular internal dimension in diastole
- EDV
Left ventricular end-diastolic volume
- ESV
End-systolic volume
- EF
Left ventricular ejection fraction
- FS
Fractional shortening
- KHB
Krebs–Henseleit buffer
- PBS
Phosphate buffered saline
- TEM
Transmission electron microscopy
- MCI
Mitochondrial complexity index
- RIPA buffer
Radioimmunoprecipitations Assay buffer
- ELISA
Enzyme-Linked Immunosorbent Assay
- ATP
Adenosine Triphosphate
- HCD
Energy collisional dissociation
- NCE
Normalized collision energy
- DAVID
Database for Annotation, Visualization, and Integrated Discovery
- GO
Gene Ontology
- KEGG
Kyoto Encyclopedia of Genes and Genomes
- ANOVA
Analysis of Variance
Authors’ contributions
C.W.H.: investigation, formal, analysis and interpretation of the data, and writing the original draft of the manuscript. Y.C.L.: analysis and interpretation of data, and reviewing and editing the manuscript. T.H.C.: analysis and interpretation of data. W.T.C.: conceptualization, and reviewing and editing the manuscript. C.P.C: study concept and design, and reviewing and editing the manuscript. C.C.H.: study concept and design, and reviewing and editing the manuscript. P.C.L: conceptualization, resources, reviewing and editing the manuscript, and supervision.
Funding
This work was supported by the National Science and Technology Council, Taiwan [grant number NSTC 111–2314-B-006–045].
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
All animal experiments adhered to protocols approved by the Institutional Animal Care and Use Committee of Chi Mei Medical Center, Tainan, Taiwan (approval no. 110122316), and were performed in accordance with the U.S. National Institutes of Health guidelines for the Guide for the Care and Use of Laboratory Animals. Considerations were given to minimize pain and suffering in this study.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Ching-Ping Chang, Email: a50831@mail.chimei.org.tw.
Chien-Cheng Huang, Email: cmher11001@mail.chimei.org.tw.
Pao-Chi Liao, Email: liaopc@mail.ncku.edu.tw.
References
- Alonso JR, Cardellach F, Lopez S, Casademont J, Miro O. Carbon monoxide specifically inhibits cytochrome c oxidase of human mitochondrial respiratory chain. Pharmacol Toxicol. 2003;93:142–6. [DOI] [PubMed] [Google Scholar]
- Bround MJ, et al. Cardiac ryanodine receptors control heart rate and rhythmicity in adult mice. Cardiovasc Res. 2012;96:372–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Casillas S, Galindo A, Camarillo-Reyes LA, Varon J, Surani SR. Effectiveness of hyperbaric oxygenation versus normobaric oxygenation therapy in carbon monoxide poisoning: a systematic review. Cureus. 2019;11:e5916. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cha YS, et al. Incidence and patterns of cardiomyopathy in carbon monoxide-poisoned patients with myocardial injury. Clin Toxicol (Phila). 2016;54:481–7. [DOI] [PubMed] [Google Scholar]
- Chang H, et al. Hydrodynamic-based delivery of an interleukin-22-Ig fusion gene ameliorates experimental autoimmune myocarditis in rats. J Immunol. 2006;177:3635–43. [DOI] [PubMed] [Google Scholar]
- Chi YJ, Pan HY, Cheng FJ, Chang YI, Chuang PC. Experience of carbon monoxide poisoning and the outcome predicting score: a multicenter retrospective study. Am J Emerg Med. 2022;58:73–8. [DOI] [PubMed] [Google Scholar]
- Cho DH, Ko SM, Son JW, Park EJ, Cha YS. Myocardial injury and fibrosis from acute carbon monoxide poisoning: a prospective observational study. JACC Cardiovasc Imaging. 2021;14:1758–70. [DOI] [PubMed] [Google Scholar]
- Choi IS. Carbon monoxide poisoning: systemic manifestations and complications. J Korean Med Sci. 2001;16:253–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Claridge B, et al. Proteome characterisation of extracellular vesicles isolated from heart. Proteomics. 2021;21:e2100026. [DOI] [PubMed] [Google Scholar]
- D’Acunzo P, et al. Mitovesicles are a novel population of extracellular vesicles of mitochondrial origin altered in Down syndrome. Sci Adv. 2021;7(7):eabe5085. [DOI] [PMC free article] [PubMed]
- Degechisa ST, Dabi YT, Gizaw ST. The mitochondrial associated endoplasmic reticulum membranes: a platform for the pathogenesis of inflammation-mediated metabolic diseases. Immun Inflamm Dis. 2022;10:e647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Doenst T, Nguyen TD, Abel ED. Cardiac metabolism in heart failure: implications beyond ATP production. Circ Res. 2013;113:709–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dong X, et al. Consensus statement on extracellular vesicles in liquid biopsy for advancing laboratory medicine. Clin Chem Lab Med. 2025;63:465–82. [DOI] [PubMed] [Google Scholar]
- Edwards ML. Hyperbaric oxygen therapy. Part 1: history and principles. J Vet Emerg Crit Care (San Antonio). 2010;20:284–8. [DOI] [PubMed] [Google Scholar]
- Gomes LC, Di Benedetto G, Scorrano L. During autophagy mitochondria elongate, are spared from degradation and sustain cell viability. Nat Cell Biol. 2011;13:589–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hagemann D, Xiao RP. Dual site phospholamban phosphorylation and its physiological relevance in the heart. Trends Cardiovasc Med. 2002;12:51–6. [DOI] [PubMed] [Google Scholar]
- Haghighi K, et al. A mutation in the human phospholamban gene, deleting arginine 14, results in lethal, hereditary cardiomyopathy. Proc Natl Acad Sci U S A. 2006;103:1388–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hampson NB. U.S. Mortality Due to Carbon Monoxide Poisoning, 1999–2014. Accidental and Intentional Deaths. Ann Am Thorac Soc. 2016;13:1768–74. [DOI] [PubMed] [Google Scholar]
- Henry CR, et al. Myocardial injury and long-term mortality following moderate to severe carbon monoxide poisoning. JAMA. 2006;295:398–402. [DOI] [PubMed] [Google Scholar]
- Heyn J, Heuschkel MA, Goettsch C. Mitochondrial-derived vesicles-link to extracellular vesicles and implications in cardiovascular disease. Int J Mol Sci. 2023;24(3):2637. [DOI] [PMC free article] [PubMed]
- Hill AF. Extracellular vesicles and neurodegenerative diseases. J Neurosci. 2019;39:9269–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hinton A Jr, et al. ATF4-dependent increase in mitochondrial-endoplasmic reticulum tethering following OPA1 deletion in skeletal muscle. J Cell Physiol. 2024;239:e31204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hu Q, et al. Hyperbaric oxygen reduces infarction volume and hemorrhagic transformation through ATP/NAD(+)/Sirt1 pathway in hyperglycemic middle cerebral artery occlusion rats. Stroke. 2017;48:1655–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang CC, et al. Increased risk of congestive heart failure following carbon monoxide poisoning. Circ Heart Fail. 2021;14:e007267. [DOI] [PubMed] [Google Scholar]
- Ikeda G, et al. Mitochondria-rich extracellular vesicles from autologous stem cell-derived cardiomyocytes restore energetics of ischemic myocardium. J Am Coll Cardiol. 2021;77:1073–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Iorio R, Petricca S, Di Emidio G, Falone S, Tatone C. Mitochondrial Extracellular Vesicles (mitoEVs): Emerging mediators of cell-to-cell communication in health, aging and age-related diseases. Ageing Res Rev. 2024;101:102522. [DOI] [PubMed] [Google Scholar]
- Jeon SB, et al. Acute brain lesions on magnetic resonance imaging and delayed neurological sequelae in carbon monoxide poisoning. JAMA Neurol. 2018;75:436–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jheng HF, et al. Mitochondrial fission contributes to mitochondrial dysfunction and insulin resistance in skeletal muscle. Mol Cell Biol. 2012;32:309–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kalay N, et al. Cardiovascular effects of carbon monoxide poisoning. Am J Cardiol. 2007;99:322–4. [DOI] [PubMed] [Google Scholar]
- Kilkenny C, et al. Animal research: reporting in vivo experiments: the ARRIVE guidelines. Br J Pharmacol. 2010;160:1577–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Konig T, McBride HM. Mitochondrial-derived vesicles in metabolism, disease, and aging. Cell Metab. 2024;36:21–35. [DOI] [PubMed] [Google Scholar]
- Konig T, et al. MIROs and DRP1 drive mitochondrial-derived vesicle biogenesis and promote quality control. Nat Cell Biol. 2021;23:1271–86. [DOI] [PubMed] [Google Scholar]
- Kosaka N, et al. Exploiting the message from cancer: the diagnostic value of extracellular vesicles for clinical applications. Exp Mol Med. 2019;51:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee FY, Chen WK, Lin CL, Kao CH. Carbon monoxide poisoning and subsequent cardiovascular disease risk: a nationwide population-based cohort study. Medicine (Baltimore). 2015;94:e624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee H, et al. Association between early phase serum lactate levels and occurrence of delayed neuropsychiatric sequelae in adult patients with acute carbon monoxide poisoning: a systematic review and meta-analysis. J Pers Med. 2022;12(4):651. [DOI] [PMC free article] [PubMed]
- Lin MS, et al. Myocardial injury was associated with neurological sequelae of acute carbon monoxide poisoning in Taiwan. J Chin Med Assoc. 2018;81:682–90. [DOI] [PubMed] [Google Scholar]
- Munoz Moreno JF, Rubio Prieto E, Ortiz Diaz-Miguel R. Cardiogenic shock secondary to severe carbon monoxide poisoning. Med Intensiva (Engl Ed). 2023 Jun;47(6):365-367. [DOI] [PubMed]
- Okura Y, et al. Recombinant murine interleukin-12 facilitates induction of cardiac myosin-specific type 1 helper T cells in rats. Circ Res. 1998;82:1035–42. [DOI] [PubMed] [Google Scholar]
- Peng T, et al. Mitochondrial-derived vesicles: Gatekeepers of mitochondrial response to oxidative stress. Free Radic Biol Med. 2022;188:185–93. [DOI] [PubMed] [Google Scholar]
- Picca A, et al. Mitochondrial-Derived Vesicles: The Good, the Bad, and the Ugly. Int J Mol Sci. 2023;24(18):13835. [DOI] [PMC free article] [PubMed]
- Rambold AS, Kostelecky B, Elia N, Lippincott-Schwartz J. Tubular network formation protects mitochondria from autophagosomal degradation during nutrient starvation. Proc Natl Acad Sci U S A. 2011;108:10190–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rodriguez P, Kranias EG. Phospholamban: a key determinant of cardiac function and dysfunction. Arch Mal Coeur Vaiss. 2005;98:1239–43. [PubMed] [Google Scholar]
- Rose JJ, et al. Carbon monoxide poisoning: pathogenesis, management, and future directions of therapy. Am J Respir Crit Care Med. 2017;195:596–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shah R, Patel T, Freedman JE. Circulating extracellular vesicles in human disease. N Engl J Med. 2018;379:958–66. [DOI] [PubMed] [Google Scholar]
- Spanos M, et al. Cardiomyocyte-derived circulating extracellular vesicles allow a non-invasive liquid biopsy of myocardium in health and disease. medRxiv : the preprint server for health sciences. 2024 Sep 22:2024.09.19.24314009.
- Suner S, Jay G. Carbon monoxide has direct toxicity on the myocardium distinct from effects of hypoxia in an ex vivo rat heart model. Acad Emerg Med. 2008;15:59–65. [DOI] [PubMed] [Google Scholar]
- Thom SR. Oxidative stress is fundamental to hyperbaric oxygen therapy. J Appl Physiol (Bethesda, Md: 1985). 2009;106:988–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Todkar K, et al. Selective packaging of mitochondrial proteins into extracellular vesicles prevents the release of mitochondrial DAMPs. Nat Commun. 2021;12:1971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Towers CG, et al. Mitochondrial-derived vesicles compensate for loss of LC3-mediated mitophagy. Dev Cell. 2021;56(2029–2042):e2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Turner M, Esaw M, Clark RJ. Carbon monoxide poisoning treated with hyperbaric oxygen: metabolic acidosis as a predictor of treatment requirements. J Accid Emerg Med. 1999;16:96–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Urabe F, et al. Extracellular vesicles as biomarkers and therapeutic targets for cancer. Am J Physiol Cell Physiol. 2020;318:C29–39. [DOI] [PubMed] [Google Scholar]
- van Niel G, et al. Challenges and directions in studying cell-cell communication by extracellular vesicles. Nat Rev Mol Cell Biol. 2022;23:369–82. [DOI] [PubMed] [Google Scholar]
- Vanderboom PM, et al. A size-exclusion-based approach for purifying extracellular vesicles from human plasma. Cell Rep Methods. 2021;1(3):100055. [DOI] [PMC free article] [PubMed]
- Varanita T, et al. The OPA1-dependent mitochondrial cristae remodeling pathway controls atrophic, apoptotic, and ischemic tissue damage. Cell Metab. 2015;21:834–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams PA, et al. Retinal ganglion cell dendritic degeneration in a mouse model of Alzheimer’s disease. Neurobiol Aging. 2013;34:1799–806. [DOI] [PubMed] [Google Scholar]
- Wu S, et al. Extracellular vesicles meet mitochondria: potential roles in regenerative medicine. Pharmacol Res. 2024;206:107307. [DOI] [PubMed] [Google Scholar]
- Yang M, et al. Mitochondria-Associated ER Membranes - The Origin Site of Autophagy. Front Cell Dev Biol. 2020;8:595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Youle RJ, van der Bliek AM. Mitochondrial fission, fusion, and stress. Science. 2012;337:1062–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.