Secondary Logo

Journal Logo

Clinical Aspects

Hyperoxia Induces Inflammation and Cytotoxicity in Human Adult Cardiac Myocytes

Hafner, Christina; Wu, Jing; Tiboldi, Akos; Hess, Moritz; Mitulovic, Goran; Kaun, Christoph; Krychtiuk, Konstantin Alexander; Wojta, Johann; Ullrich, Roman; Tretter, Eva Verena; Markstaller, Klaus; Klein, Klaus Ulrich

Author Information
doi: 10.1097/SHK.0000000000000740



Oxygen (O2) therapy is regarded as a critical adjunct treatment in anesthesia, emergency, and intensive care medicine to prevent hypoxia. However, O2 therapy may result in excessive O2 levels (hyperoxia) that can injure the cardiovascular system and other organs (1). Indeed, hyperoxia in general surgical patients has been linked to higher patient mortality and morbidity (2, 3). Likewise, hyperoxic ventilation of intensive care patients after cardiac arrest and stroke was associated with poorer outcomes (4, 5).

The mechanisms involved in hyperoxic acute lung injury have been characterized in great detail (6). In contrast, the effects of hyperoxia on the cardiovascular system and especially the heart are far less studied. In patients at risk for myocardial infarction and critically-ill patients, high O2 levels are frequently administered to prevent myocardial hypoxia. Indeed, hyperoxia may have beneficial effects on cardiovascular function, including preconditioning-induced protection against reperfusion injury after short-term intermittent hyperoxia (7) and reduction of gas emboli (8). The evidence on prevention of surgical site infections by hyperoxia remains inconclusive. On the other hand, detrimental effects have been reported that include oxidative stress, coronary vasoconstriction, impaired microcirculation, and myocardial dysfunction (9). To date, few studies have examined the cellular effects of hyperoxia on cardiomyocytes, and no study has investigated the effects of hyperoxic exposure on human adult myocardial cells. Neither a comprehensive analysis of protein expression after hyperoxic exposure on human adult myocardial cells has been performed. Proteomic analysis may provide information regarding altered protein expression levels, which may play a crucial role in cell signaling and regulation. Proteomic analysis is also an important tool to achieve data for further biomarker research (10).

The present in vitro study compared the effects of constant hyperoxia (95% O2), intermittent hyperoxia (5%–95% O2 every 10 min, average value 50% O2), constant normoxia (21% O2), and constant mild hypoxia (5% O2) on human adult cardiac myocytes (HACMs). We initially speculated that constant but not intermittent hyperoxic exposure would have detrimental effects on HACMs. In fact, constant hyperoxia was in general more deleterious on cell viability than other conditions and induced the strongest inflammatory response. However, relatively brief intermittent hyperoxic episodes also disrupted cell morphology and induced the release of inflammatory mediators from HACMs. Further, quantitative proteomic analysis revealed broad effects of hyperoxia on cell-cycle regulation, metabolism, and cell signaling.


Isolation and culture of HACMs

Ethical approval for this study was provided by the Medical University of Vienna Ethics Committee, Vienna, Austria on August 23, 2008 (approval number: 171/2008; chairperson: Professor E. Singer). Experiments were performed after obtaining written informed consent from donor patients between December 2013 and July 2014. Primary cultures of HACMs were prepared from ventricular tissue obtained from donor hearts from three patients undergoing heart transplantation for severe ischemic or dilative cardiomyopathy as described (11). Isolation of cells was performed 2 to 6 h after explantation.

Ventricular tissue was cut into small pieces in phosphate buffered saline (PBS) pH 7.4 without enzymatic digestion (PBS (Gibco, Invitrogen, Carlsbad, Calif). The cell mass was filtered though a 40 μm cell strainer and centrifuged at 1,200 rpm for 10 min. The pellet was washed twice with PBS. Cells were resuspended in Dulbecco modified eagle medium (DMEM) containing 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin and plated into a Petri-dish and incubated for 60 min at 37 C with 5% carbon dioxide (CO2) in the atmosphere to separate myocytes from fibroblasts by preplating. The nonattached cells were centrifuged and rinsed twice with PBS. The cell pellet was resuspended in DMEM containing 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, 10 μg/mL transferrin, 10 μg/mL insulin. HACMs were seeded into fibronectin-coated flasks at a density of 104 cells/cm2. Cultured HACMs were characterized by staining for troponin I, tropomyosin, cardiotin, and myocardial muscle-actin. Contamination of the cultures with smooth muscle cells, endothelial cells, and fibroblasts was ruled out by staining for desmin, vWF, and fibroblast specific antigens.

Only cultures in which >95% of the cells stained positive for cardiac myocyte markers (troponin I, tropomyosin, cardiotin, and myocardial muscle actin) were used in this study. In these cultures contamination with smooth muscle cells, endothelial cells, and fibroblasts as judged by staining for smooth muscle actin, von Willebrand factor, and fibroblast specific antigens was <2% (11). All in vitro experiments were performed with HACMs at passage 2 to 4. The harvesting and maintenance of HACMs was described prior to experimentation (12). Briefly, HACMs were recovered in a 1% gelatin-coated (Sigma-Aldrich, St. Louis, Mo) 75 cm2 culture flask (Falcon, BD Biosciences, Schwechat, Austria) and maintained in Medium 199 (M199, Gibco, Invitrogen, Carlsbad, Calif) supplemented with 20% heat-inactivated fetal bovine serum (FBS superior, Biochrom, Berlin, Germany), 10 mM L-glutamine, and 1% penicillin/streptomycin (Gibco, Invitrogen, Carlsbad, Calif) under a humidified atmosphere containing 5% CO2 at 37°C.

For passaging and seeding, the M199 was removed and cells were incubated with 5 mL of 0.25% trypsin–EDTA solution (Sigma-Aldrich) for 2 min at 37°C. Cells were detached by agitation and trypsin was activity stopped by the addition of M199 containing 20% FBS. At 80% to 90% confluence, HACMs were seeded within OptiCell (OC) plates (Nunc, Thermo Scientific, Waltham, Mass) precoated with 1% gelatin at 0.6 × 105 cells/plate. After 2 h, the plates were turned over to achieve 50% density on both OC plate membranes. The HACMs were cultivated for 6 days before treatments. Experiments were conducted in fresh medium after 24 h incubation in serum-free M199 containing 0.1% bovine serum albumin (BSA, Sigma-Aldrich) and 1% penicillin/streptomycin.

Controlled oxygen exposure

A custom-made bioreactor consisting of four units, each able to hold 10 OC plates, was used for controlled O2 exposure as recently described (13, 14). One OC plate holds two 75 μm respiratory active membranes (RAMs) with cells growing on the inner surfaces of the RAMs. The space between the RAMs forms a chamber-holding medium. Oxygen, CO2, and nitrogen (N2) exchange is maintained via rapid diffusion through the RAMs. The four units and their gas supply tubes were stored in a standard incubator (HeraTherm, Thermo Fisher Scientific, Waltham, Mass) at 37°C. To maintain humidity >95%, sterile water bubblers (Kendall, Covidien, Germany) were connected in between the gas supplies and the units using non-permeable gas supply tubes. Humidity was monitored by an electronic sensor (SHT71, Sensirion, Staefa, Switzerland). Different gas concentrations were supplied from premixed bottles (AirLiquide, Schwechat, Austria) and pO2 levels monitored in real time using an O2 sensor system (NeoFox AL300, Ocean Optics, Dunedin, Fla) and RedEye oxygen sensing patches (Ocean Optics). The intermittent hyperoxia unit was connected to an electronic gas flow meter and a computer-controlled valve (EL-Flow Select Series, Bronkhorst, The Netherlands) to switch between 5% and 95% O2. Oxygen oscillations were set to a 10 min switching frequency, resulting in three full oscillation cycles per hour. Gas flow for each box was set to 0.8 L/min. Cultures were exposed to the following premixed O2 conditions: 95% O2: 5% CO2 (constant relative severe hyperoxia); 21% O2: 5% CO2: 74% N2 (constant normoxia); 5% O2: 5% CO2: 90% N2 (constant relative mild hypoxia); or 5% O2: 5% CO2: 90% N2 switching to 95% O2: 5% CO2 (intermittent relative severe hyperoxia). All experiments were performed four times with three internal replicates per trial.

Recovery of cells and media

Cultures were analyzed at 0 h (baseline) and after 8, 24, and 72 h, as described previously (11, 12). The supernatant was stored at 4°C for cytotoxicity testing (LDH activity) or at −20°C for assay of released inflammatory cytokines. Cells were harvested with 5 mL of 0.25% trypsin–EDTA solution and centrifuged at 1,000 rpm for 5 min at 4°C. The supernatant was discarded and the pellet suspended in 500 μL PBS for further analysis (below).

Cell viability and cytotoxicity

Cell viability and cytotoxicity were measured by TB staining and LDH release. A hemocytometer (Hausser Scientific, Horsham, Pa) was used to count cells after TB (0.4%, Sigma-Aldrich) staining. A colorimetric enzymatic assay (Cytotoxicity Detection Kit, Roche, Mannheim, Germany) was used to measure LDH activity in the media. Briefly, 100 μL of supernatant was mixed with 100 μL of freshly prepared LDH reaction mixture and incubated for 30 min at room temperature. The absorbance of the samples was measured at 490 nm on a microplate reader (Victor3, Perkin Elmer, Waltham, Mass). For determination of per cent cytotoxicity, high and low LDH controls were performed. For the high control, 100 μL cell suspension was treated with 100 μL Triton X-100 solution. For low control, 100 μL cell suspension was treated with fresh medium.

Cytokine secretion assays

Media concentrations of released IL-1β, IL-6, IL-8, VEGF, and MIF were measured by enzyme-linked immunosorbent assays (ELISA, DuoSet, R&D Systems, Minneapolis, Minn) at 0, 8, 24, and 72 h. Media concentrations of CXC-1, CXC-10, CCL-2, CCL-3, CCL-5, G-CSF, and ICAM-1 were measured by immunoassays (Bio-Plex Multiplex System, Bio-Rad Laboratories, Hercules, Calif).

Proteomic analysis

Changes in protein expression were analyzed by quantitative proteomic analysis. Proteins were extracted from supernatants and cell pellets using the Total Protein Extraction Kit (Merck-Millipore, Vienna, Austria). Extracted proteins were precipitated using a modified Wessel–Fluegge method and dissolved in 50 mM triethylammonium bicarbonate, pH 8.5 (Sigma-Aldrich, Vienna, Austria). Total protein concentration was determined using the Direct Detect infrared-based measurement system (Merck-Millipore, Vienna, Austria). Proteins were labeled using the ICPL Quadruplex reagent kit (Serva, Heidelberg, Germany). Labeled proteins were mixed in equimolar ratios and digested overnight at 37°C using sequencing grade trypsin (Promega, Vienna, Austria) at a trypsin/protein ratio of 1:50. Digestion was stopped by acidification with 10 μL of 1% trifluoroacetic acid (TFA) (Sigma-Aldrich, Vienna, Austria), and 20 μL portions of digested peptides were further diluted with 30 μL 0.1% TFA. Peptides were separated and analyzed by nano-high performance liquid chromatography (nano-HPLC)-tandem mass spectrometry (UltiMate3000, ThermoFisher Scientific, Vienna, Austria). Mass spectrometric analysis was performed using maXis qToF (Bruker, Bremen, Germany) and data acquisition rate was 2 Hz for full scan MS and masses of lower intensity (40,000 counts), and 20 Hz for signals of high intensity (960,000 counts).

Statistical analysis

Cell viability and ELISA cytokine assays were repeated using three independent cultures from different explanted hearts and in triplicate for each cell culture. Multiplex system assays were performed using one additional culture in triplicate. Proteomic analysis was performed using one additional culture in triplicate. Cells were counted before plating and at timepoints of 0, 8, 24, and 72 h. Pg/ng per 100,000 cells relates to the total cell count at individual time points. Data were analyzed (assumed in case of small sample size) for Gaussian distribution using the Kolmogorov–Smirnov normality test. Dare expressed as mean ± standard deviation. Means were compared by two-way analysis of variance for independent samples with pair-wise Bonferroni post hoc tests. If not indicated otherwise, adjusted P values are considered significant at a P value <0.05. However, as the study was conducted in an explorative approach, all P values are considered explorative. Statistical analysis and data plotting were performed using Prism 6.0 software (GraphPad, San Diego, Calif).

Proteomic data analysis

Proteomic data were used as queries for Mascot 2.5.1 (Matrix Science, London, UK) searches of the International Protein Index human database, with a mass tolerance of 50 ppm for the parent mass and 0.4 Da for MS. Raw files were converted into Mascot generic format files by a script, and Mascot database searches were performed using ProteinScape, Version 3.1.5 (Bruker, Bremen, Germany). Quantitation of identified proteins was performed using WARP-LC Version 1.3 (Bruker). Protein contents in samples exposed to hypoxia, constant hyperoxia, or intermittent hyperoxia in two independent experiments were compared with samples exposed to normoxia. Only proteins successfully quantified in both experiments were retained. If the average log2[fold-change] over both experiments was >1, proteins were considered upregulated relative to normoxic conditions. Similarly, proteins were considered downregulated relative to normoxic conditions if the average log2[fold-change] was <−1. Since the proteome analysis was explorative and conducted with only two replicates, no statistical tests were performed.


Exposure conditions

Stable target O2 concentrations in the different units were confirmed for constant normoxia (21% ± 0.5% O2), constant hypoxia (5% ± 0.5% O2), and constant hyperoxia (95% ± 0.5% O2). Consistent hyperoxic oscillations were produced in the intermittent hyperoxia unit (5%–95% O2 around a mean O2 value of 50% resulting in three full O2-oscillation cycles per hour). Temperature (37.0°C ± 0.5°C), relative humidity (95% ± 5%) and ambient pressure in all units remained unchanged (<0.1 kPa) during experiments.

Reduced cell viability and cytotoxicity under hyperoxia

Under examination with phase contrast microscopy, HACMs exposed to constant or intermittent hyperoxia for 72 h appeared more spindle shaped than cells incubated under constant normoxia or constant hypoxia for the same duration (Fig. 1, A–D). Cells exposed to constant hyperoxia showed a significantly higher TB-positive after 72 h (P <0.05, Fig. 2A). Cells exposed to constant hyperoxia showed the highest LDH release of all conditions after 24 and 72 h (P <0.01, Fig. 2B).

Fig. 1:
Morphological signs of cell degeneration during constant and intermittent hyperoxia in human adult cardiac myocytes (HACMs).Representative example from one cell culture. A–D, Changes in cell morphology during altered O2 conditions under phase contrast microscopy. After 72 h, HACMs exposed to constant (C) and intermittent hyperoxia (D) appeared spindle shaped than cells exposed to constant normoxia (A) or constant hypoxia (B).
Fig. 2:
Trypan blue (TB)-positive cells and lactate dehydrogenase (LDH) release under different oxygen (O2) conditions.Comparison of TB-positive cells (A) and LDH release (B) after exposure to normoxia (21% O2), hypoxia (5% O2), hyperoxia (95% O2), and intermittent hyperoxia (shifts between 5% and 95% O2 every 10 min) at different time points (0, 8, 24, and 72 h). Cytotoxicity assays were repeated using three independent cultures from different explanted hearts and in triplicate for each cell culture. * P <0.05, values differ from normoxia at indicated time points; § P <0.05, values differ from all other conditions at indicated time points.

Induction of cytokine and chemokine release by hyperoxia

Release of CXCL-1 was significantly higher in the prolonged hyperoxia group compared with normoxia after only 8 h (P <0.05, Fig. 3A). After 24 h, constant hyperoxia significantly increased release of CXCL-1 (P <0.01, Fig. 3A), CXCL-10 (P <0.01, Fig. 3B), and ICAM-1 (P <0.05, Fig. 3D) compared with normoxia. At this time, intermittent hyperoxia exposure resulted in the greatest release of VEGF (P <0.01, Fig. 4D) of any treatment and significant higher CXCL-10 (P <0.01). After 72 h, VEGF release was enhanced significantly from HACMs exposed to constant hyperoxia (P <0.01), intermittent hyperoxia (P <0.05), and constant hypoxia (P <0.01), compared with constant normoxia. Only constant hyperoxia significantly enhanced release of MIF (P <0.01, Fig. 4E) and G-CSF (Fig. 3C). Release levels of IL-1β (Fig. 4A), IL-6 (Fig. 4B), CXCL-10 (Fig. 3B), G-CSF (Fig. 3C), ICAM-1 (Fig. 3D), CCL-3 (Fig. 5A), and CCL-5 (Fig. 5B) were highest at 72 h in HACMs exposed to constant hyperoxia (all P <0.05). Exposure of HACMs to intermittent hyperoxia also significantly enhanced the release of IL-1β, IL-6, IL-8 (highest of all groups) compared with normoxia (Fig. 4), and the release of CXCL-1 and ICAM-1 (Fig. 3) compared with constant hyperoxia. Analysis of CCL-2 did not show any significant results between groups at any time point.

Fig. 3:
Chemokine release (chemokine C-X-C motif ligand, CXCL; granulocyte-colony stimulating factor, G-CSF; and intracellular adhesion molecule, ICAM) under different oxygen (O2) conditions.Comparisons of CXCL-1 (A), CXCL-10 (B), G-CSF (C), and ICAM-1 (D) release after exposure to normoxia (21% O2), hypoxia (5% O2), hyperoxia (95% O2), and intermittent hyperoxia (shifts between 5% and 95% O2 every 10 min) at different time points (0, 8, 24, and 72 h). Multiplex assays were performed using one culture from one explanted heart in triplicate. * P <0.05, values differ from normoxia at indicated time points. * P <0.05, values differ from normoxia at indicated time points; # P <0.05, values differ from hyperoxia at indicated time points; § P <0.05, values differ from all other conditions at indicated time points.
Fig. 4:
Cytokine release (interleukin (IL); vascular endothelial growth factor (VEGF); macrophage migration inhibitory factor (MIF)) under different oxygen (O2) conditions.Comparisons of IL-1β (A), IL-6 (B), IL-8 (C), VEGF (D), and MIF (E) release in the supernatant after exposure to normoxia (21% O2), hypoxia (5% O2), hyperoxia (95% O2), and intermittent hyperoxia (shifts between 5% and 95% O2 every 10 min) at different time points (0, 8, 24, and 72 h). ELISA assays were repeated using three independent cultures from different explanted hearts and in triplicate for each cell culture. * P <0.05, values differ from normoxia at indicated time points; + P <0.05, values differ from hypoxia at indicated time points; § P <0.05, values differ from all other conditions at indicated time points. ELISA indicates enzyme-linked immunosorbent assays.
Fig. 5:
Chemokine release (chemokine (C-C motif ligand, CCL)) under different oxygen (O2) conditions.Comparisons of CCL-3 (A) and CCL-5 (B) release after exposure to normoxia (21% O2), hypoxia (5% O2), hyperoxia (95% O2), and intermittent hyperoxia (shifts between 5% and 95% O2 every 10 min) at different time points (0, 8, 24, and 72 h). Multiplex assays were performed using one culture from one explanted heart in triplicate. * P <0.05, values differ from normoxia at indicated time points.

Proteomic analysis

Quantitative proteomic analysis was performed from both cell pellets and supernatant samples (Appendix 1, After HACM exposure to constant hyperoxia, nine proteins were upregulated (log2[fold-change] >1) and 10 proteins were downregulated (log2[fold-change] <−1), compared with normoxia. Intermittent hyperoxia exposure led to higher levels of 14 proteins and lower levels of seven proteins. Constant hypoxia resulted in 14 proteins upregulated and six proteins downregulated. In pellet, Q13813-3, I3L0K7, and Q9A4Q5-2 were upregulated after constant hyperoxia exposure and downregulated after constant hypoxia exposure. In supernatant, O60762 and H0YAK2 were upregulated after constant hyperoxia exposure and downregulated after constant hypoxia exposure. No significant differences could be detected between intermittent hyperoxia and constant hypoxia or constant hyperoxia exposure (Table 1).

Table 1:
Significant changes (up- and down regulation, P <0.05) in protein expression (accession number and function) in the supernatant and pellet analyzed by quantitative proteomic analysis after 48 h after exposure to normoxia (21% O2), hypoxia (5% O2), hyperoxia (95% O2), and intermittent hyperoxia (5–95% O2 every 10 min)


Both constant (95% O2) and intermittent hyperoxia (5%–95% O2, average value 50% O2) induced cytotoxic and inflammatory effects on HACMs as evidenced by higher TB-positive cell numbers, LDH release, and secretion of VEGF, MIF, IL-1β, IL-6, IL-8, CXCL-1, CXCL-10, G-CSF, ICAM-1, CCL-3, and CCL-5. Cell injury gradually increased over time up to 72 h and was highest in the constant hyperoxia group, but was also present in the intermittent hyperoxia group. Quantitative proteomic analysis after 48 h revealed that proteins involved in cell-cycle regulation, energy metabolism, and cell signaling were up- or down-regulated by hyperoxia. The present in vitro results suggest that prolonged intermittent hyperoxia and especially constant hyperoxia can induce substantial damage to HACMs. HACM injury may account for some of the aberrant effects of therapeutic hyperoxia on the cardiovascular system.

All treatment groups exhibited sustained VEGF secretion over time, suggesting that prolonged lack of or excess oxygen can stimulate vascular remodeling in the heart. In contrast to this relatively non-specific VEGF response, MIF was induced only by constant hyperoxia, suggesting that hyperoxia has unique pro-inflammatory effects. Indeed, MIF represents a key indicator of myocardial inflammation and injury (15). Cells exposed to constant or intermittent hyperoxia also appeared more spindle shaped, a macroscopic finding that suggests cell injury. Several other cytokines and chemokines were released specifically by constant and intermittent hyperoxia but not by mild hypoxia (IL-1β, IL-6, and IL-8). IL-1β and IL-6 are primarily produced at sites of acute and chronic inflammation and can activate several pro-inflammatory pathways (16), while IL-8 acts as a chemotactic factor in the inflammatory response and is also a potent angiogenic factor (17).

Several chemokines and cytokines were induced by both hyperoxia and mild hypoxia, for instance CCL3. This cytokine activates granulocytes leading to acute neutrophilic inflammation and further induces the synthesis and release of other pro-inflammatory cytokines from fibroblasts and macrophages. CXCL1, an inflammatory chemokine that has neutrophil chemo-attractant activity, was also induced by both hypoxia and hyperoxia. CCL5 (RANTES) is an inflammatory mediator chemotactic for inflammatory immune cells like granulocytes and monocytes. CCL5 also promotes matrix degradation by activating matrix metalloproteinase 9. Thus, a shift away from optimal O2 in either direction may induce cardiac inflammation and vascular remodeling. On the other hand, CCL5 also recruits inhibitory monocyte subsets and regulatory T-cells. Similarly, G-CSF release solely by constant hyperoxia may represent a compensatory protective mechanism, as treatment with G-CSF can prevent non-ischemic cardiomyopathy (18).

Quantitative proteomic analysis revealed that altered O2 exposure differently influenced the expression levels of proteins involved in cell-cycle regulation, energy metabolism, and cell signaling. In pellet, SPTN1_HUMAN, I3L0K7_HUMAN, and TRRAP_HUMAN were upregulated after constant hyperoxia exposure, while downregulated after constant hypoxia exposure. Spectrin (SPTN1_HUMAN) is essential for maintaining cell membrane stability and structure. This specific protein plays a pivotal role in signaling cascades and interacts with structural proteins, proteins in DNA repair, chromatin remodeling proteins, and transcription and RNA-processing factors (19). Mitochondrial heat shock protein 75 (I3L0K7_HUMAN) is upregulated after stress and provides cardioprotection by ameliorating mitochondrial dysfunction (20). Transformation/transcription domain-associated protein (TRRAP_HUMAN), a histone modifier, causes multiple mitotic defects if overexpressed and thereby targeted for destruction during cell cycle (21). In supernatant, DPM1_HUMAN and DPM1_HUMAN were upregulated after constant hyperoxia exposure and downregulated after constant hypoxia exposure. Dolichol-phosphate mannosyltransferase (DPM1_HUMAN) is involved in cellular protein modification and the first committed step enzyme in the N-glycosylation pathway. Notably, increased levels of DPM1 have been detected after ischemia-reperfusion injury of the mouse myocardium (22). Upregulation of inorganic pyrophosphatase 2 (H0YAK2_HUMAN) after constant hyperoxia exposure suggests increase in fatty acid lipid degradation (23). Collectively, our findings support the theory that hypoxia exposure might induce vascular remodeling and repair in HACMs. Instead, prolonged hyperoxia exposure might damage HACMs by triggering inflammation—a theory that needs further investigation.

Two studies investigating the effects of short-term hyperoxia on cardiac fibroblasts in vitro demonstrated reversible growth inhibition (24, 25). G2/M arrest and associated differentiation to myofibroblasts was mediated by induction of p21, cyclins D1, D2 and G1, and activation of TGFβ1 and p38. These results are in accord with an earlier observation that low O2 concentrations extend the lifespan of human diploid cells (26). Further, higher mitochondrial production of ROS was observed. A recent study investigating the effects of long-term hyperoxia on human umbilical vein endothelial cells (HUVECs) found that exposure induced inflammation, apoptosis, and cell death (14). The authors of this former study suggested that hyperoxia may cause injury in HUVECs potentially severe enough to influence perioperative outcome. Thus, this former in vitro study in HUVECs and the present in vitro study in HACMs indicate that hyperoxia may exert direct injurious effects on the cardiovascular system.

In both dogs and primates (baboons), hyperoxia causes hemodynamic compromise, including decreased total lung capacity, cardiac output, and stroke volume, increased systemic vascular resistance, and extravascular lung edema (27–29). Histopathological analysis revealed endothelial injury, neutrophil accumulation, and interstitial edema in the cardiovascular system (30). On the other hand, short-term intermittent hyperoxia has been shown to elicit myocardial protection, similar to that induced by ischemic or anesthetic preconditioning (7). However, studies that have investigated the efficacy of preconditioning used a relatively brief exposure. It appears therefore that these preconditioning-like changes are mitigated by repeated hyperoxia.

As in animal experiments, heart rate, stroke volume, and cardiac index were reduced while systemic vascular resistance was increased after hyperoxia administration to healthy adults (27–29). Similar effects on cardiac output, as well as increased pulmonary artery wedge pressure, were observed in studies of patients exposed to hyperoxia (31, 32). Despite these effects, a retrospective study in cardiac patients found no association between hyperoxia and hospital mortality (33), although a small but statistically significant increase in length of stay was detected in the hyperoxia group. Another small study in cardiac surgery patients found that intensive care unit stay and mechanical ventilation time tended to be slightly (but not significantly) longer in the hyperoxia group (34). Two clinical studies found normoxia during bypass lowered inflammation (as measured by polymorphonuclear leucocyte elastase) and oxidative stress-related parameters (such as malonaldehyde, an index of lipid peroxidation) (35, 36). While the number of studies and patients investigated are small and current evidence is insufficient to specify optimal O2 targets, clinical evidence suggests that hyperoxia exposure may cause more harm than good (8, 9). With this in mind, a recent study in general surgery populations found that perioperative hyperoxia increased risk for acute coronary syndrome and myocardial infarction (37).

In hospitalized patients, O2 delivery to tissues may be threatened by impaired macrocirculation and organ microcirculation, anemia, hypothermia, edema, and/or inflammation (9). Hyperoxia is frequently administered to augment O2 delivery to vital organs in at-risk patients. However, hyperoxia can cause injury to the pulmonary, cardiovascular, immune, and nervous systems possibly by oxidative stress and effects secondary to reduced organ blood flow. The toxic effects of prolonged O2 exposure on the lungs have been investigated in great detail (6). While hyperoxic pulmonary injury in healthy volunteers takes 6 to 12 h to evolve, which is long relative to some clinical applications, adverse effects of hyperoxia on pre-injured lungs can occur much faster (38, 39). However, hyperoxia also produces reactive O2 species, thereby improving oxidative killing of bacterial pathogens by neutrophils (40). The beneficial and harmful effects of high inspiratory O2 fraction around the time of surgery have been investigated in a recent Cochrane Systematic Review (41). The authors conclude that the risk of adverse events, including mortality, may be increased by a fraction of inspired O2 of 60% and higher. A recent post hoc analysis suggests that perioperative hyperoxia may be associated with an increased long-term risk of myocardial infarction and other heart disease (37). These findings in general surgical patients are in agreement with the fact that careful titration of O2 supply is recommended for intensive care patients as some studies suggest detrimental outcome after hyperoxia exposure (42).

In our study, HACMs were used as representatives of the cardiovascular system. Cells were isolated from damaged hearts of patients who received heart transplantation; therefore, findings are likely different to findings in cells isolated from healthier donors and results must be interpreted with caution. We proposed that the present findings should be confirmed in cells isolated from healthier donors. Also, clinical studies should be performed to investigate effects on endothelial and myocardial injury (e.g., biomarkers, electrocardiography, imaging) to confirm that hyperoxia causes harm to the cardiovascular system (endothelium, myocardium). In the clinical setting, hyperoxia causes vasoconstriction that might decrease myocardial O2 levels. Myocardial O2 levels depend on the diffusional gradient for O2 in tissue. Therefore, O2 levels cells near blood vessels might be exposed to higher O2 concentrations than cells located more distant to blood vessels. Hyperoxic ranges used in this in vitro study were outside the clinical range and results have to be interpreted with caution. We propose to investigate if short-term intermittent hyperoxia exerts cardioprotective effects, while long-term constant or intermittent hyperoxia might be deleterious. We also used an extreme level of hyperoxia, which was chosen because cell culture studies may require stronger or longer challenges compared with in vivo conditions. As another limitation of study, we did not replicate any common clinical situation. For instance, we did not expose the cells to any additional injurious event such as ischemia-reperfusion, which may also alter the sensitivity to hyperoxia. Another major limitation of study is the fact that we analyzed proteomic data after 48 h. We have chosen this time point for the fact that markers of cell injury and inflammation started to increase after 24 h and were highest after 72 h; however, results ideally should be performed at more time points. Although it remains unknown, it appears possible that injury may be aggravated in the clinical setting. Thus, further in vivo preclinical studies are warranted to assess the range, duration, and pathological contexts in which hyperoxia might be deleterious to the heart.

Hyperoxia is used in the clinical setting to prevent hypoxia. Moreover, hyperoxia exposure frequently occurs in the clinical setting even when aiming at normoxic O2 levels (9). Thus, deleterious effects of hyperoxia may be under-recognized. Here, we demonstrated that both constant (95% O2) and intermittent (5%–95% O2, average value 50%) hyperoxic exposure injured HACMs and induced a substantial inflammatory response. Cell injury may gradually increase after the start of hyperoxic exposure. Proteomic analysis showed that constant and intermittent hyperoxia altered the expression patterns of multiple proteins involved in cell-cycle regulation, energy metabolism, and cellular signaling. Our in vitro findings suggest that prolonged constant and intermittent hyperoxia exposure, as potentially occur during administration of O2 therapy, may produce rapid and substantial damage to the human heart.


1. Martin DS, Grocott MPW. Oxygen therapy and anaesthesia: too much of a good thing? Anaesthesia 2015; 70 5:522–527.
2. Meyhoff CS, Jorgensen LN, Wetterslev J, Christensen KB, Rasmussen LS. PROXI Trial Group. Increased long-term mortality after a high perioperative inspiratory oxygen fraction during abdominal surgery: follow-up of a randomized clinical trial. Anesth Analg 2012; 115 4:849–854.
3. Meyhoff CS, Jorgensen LN, Wetterslev J, Siersma VD, Rasmussen LS. PROXI Trial Group. Risk of new or recurrent cancer after a high perioperative inspiratory oxygen fraction during abdominal surgery. Br J Anaesth 2014; 113 (suppl 1):i74–i81.
4. de Jonge E, Peelen L, Keijzers PJ, Joore H, de Lange D, van der Voort PH, Bosman RJ, de Waal RA, Wesselink R, de Keizer NF. Association between administered oxygen, arterial partial oxygen pressure and mortality in mechanically ventilated intensive care unit patients. Crit Care 2008; 12 6:R156.
5. Janz DR, Hollenbeck RD, Pollock JS, McPherson JA, Rice TW. Hyperoxia is associated with increased mortality in patients treated with mild therapeutic hypothermia after sudden cardiac arrest. Crit Care Med 2012; 40 12:3135–3139.
6. Kallet RH, Matthay MA. Hyperoxic acute lung injury. Respir Care 2013; 58 1:123–141.
7. Kunst G, Klein AA. Perioperative anaesthetic myocardial preconditioning and protection—cellular mechanisms and clinical relevance in cardiac anaesthesia. Anaesthesia 2015; 70 4:467–481.
8. Young RW. Hyperoxia: a review of the risks and benefits in adult cardiac surgery. J Extra Corpor Technol 2012; 44 4:241–249.
9. Spoelstra-de Man AM, Smit B, Oudemans-van Straaten HM, Smulders YM. Cardiovascular effects of hyperoxia during and after cardiac surgery. Anaesthesia 2015; 70 11:1307–1319.
10. Stastna M, Van Eyk JE. Secreted proteins as a fundamental source for biomarker discovery. Proteomics 2012; 12 (4–5):722–735.
11. Weiss TW, Speidl WS, Kaun C, Rega G, Springer C, Macfelda K, Losert UM, Grant SL, Marro ML, Rhodes AD, et al. Glycoprotein 130 ligand oncostatin-M induces expression of vascular endothelial growth factor in human adult cardiac myocytes. Cardiovasc Res 2003; 59 3:628–638.
12. Macfelda K, Weiss TW, Kaun C, Breuss JM, Zorn G, Oberndorfer U, Voegele-Kadletz M, Huber-Beckmann R, Ullrich R, Binder BR, et al. Plasminogen activator inhibitor 1 expression is regulated by the inflammatory mediators interleukin-1alpha, tumor necrosis factor-alpha, transforming growth factor-beta and oncostatin M in human cardiac myocytes. J Mol Cell Cardiol 2002; 34 12:1681–1691.
13. Wu J, Stefaniak J, Hafner C, Schramel JP, Kaun C, Wojta J, Ullrich R, Tretter VE, Markstaller K, Klein KU. Intermittent hypoxia causes inflammation and injury to human adult cardiac myocytes. Anesth Analg 2016; 122 2:373–380.
14. Wu J, Hafner C, Schramel JP, Kaun C, Krychtiuk KA, Wojta J, Boehme S, Ullrich R, Tretter EV, Markstaller K, et al. Cyclic and constant hyperoxia cause inflammation, apoptosis and cell death in human umbilical vein endothelial cells. Acta Anaesthesiol Scand 2016; 60 4:492–501.
15. Dayawansa NH, Gao XM, White DA, Dart AM, Du XJ. Role of MIF in myocardial ischaemia and infarction: insight from recent clinical and experimental findings. Clin Sci (Lond) 2014; 127 3:149–161.
16. Bhandari V, Elias JA. Cytokines in tolerance to hyperoxia-induced injury in the developing and adult lung. Free Radic Biol Med 2006; 41 1:4–18.
17. Simonini A, Moscucci M, Muller DW, Bates ER, Pagani FD, Burdick MD, Strieter RM. IL-8 is an angiogenetic factor in human coronary atherectomy tissue. Circulation 2000; 101 13:1519–1526.
18. Li L, Takemura G, Li Y, Miyata S, Esaki M, Okada H, Kanamori H, Ogina A, Maruyama R, Nakagawa M, et al. Granulocyte colony-stimulating factor improves left ventricular function of doxorubicin-induced cardiomyopathy. Lab Invest 2007; 87 5:440–455.
19. Sridharan DM, McMahon LW, Lambert MW. AlphaII-Spectrin interacts with five groups of functionally important proteins in the nucleus. Cell Biol Int 2006; 30 11:866–878.
20. Zhang P, Lu Y, Yu D, Zhang D, Hu W. TRAP1 provides protection against myocardial ischemia-reperfusion injury by ameliorating mitochondrial dysfunction. Cell Physiol Biochem 2015; 36 5:2072–2082.
21. Ichim G, Mola M, Finkbeiner MG, Cros MP, Herceg Z, Hernandez-Vargas H. The histone acetyltransferase component TRRAP is targeted for destruction during the cell cycle. Oncogene 2014; 33 2:181–192.
22. Gao J, Xu D, Sabat G, Valdivia H, Xu W, Shi NQ. Disrupting KATP channels diminishes the estrogen-mediated protection in female mutant mice during ischemia-reperfusion. Clin Proteomics 2014; 11 1:19.
23. Carman GM, Han GS. Roles of phosphatidate phosphatase enzymes in lipid metabolism. Trends Biochem Sci 2006; 31 12:694–699.
24. Roy S, Khanna S, Bickerstaff AA, Subramanian SV, Atalay M, Bierl M, Pendyala S, Levy D, Sharma N, Venojarvi M, et al. Oxygen sensing by primary cardiac fibroblasts: a key role of p21(Waf1/Cip1/Sdi1). Circ Res 2003; 92 3:264–271.
25. Roy S, Khanna S, Wallace WA, Lappalainen J, Rink C, Cardounel AJ, Zweier JL, Sen CK. Characterization of perceived hyperoxia in isolated primary cardiac fibroblasts and in the reoxygenated heart. J Biol Chem 2003; 278 47:47129–47135.
26. Packer L, Fuehr K. Low oxygen concentration extends the lifespan of cultured human diploid cells. Nature 1977; 267 5610:423–425.
27. Lodato RF, Michael JR, Murray PA. Multipoint pulmonary vascular pressure-cardiac output plots in conscious dogs. Am J Physiol 1985; 249 (2 Pt 2):H351–357.
28. Lodato RF. Effects of normobaric hyperoxia on hemodynamics and O2 utilization in conscious dogs. Adv Exp Med Biol 1990; 277:807–815.
29. Fracica PJ, Knapp MJ, Piantadosi CA, Takeda K, Fulkerson WJ, Coleman RE, Wolfe WG, Crapo JD. Responses of baboons to prolonged hyperoxia: physiology and qualitative pathology. J Appl Physiol 1985; 71 6:2352–2362.
30. Tähepõld P, Ruusalepp A, Li G, Vaage J, Starkopf J, Valen G. Cardioprotection by breathing hyperoxic gas-relation to oxygen concentration and exposure time in rats and mice. Eur J Cardiothorac Surg 2002; 21 6:987–994.
31. Waring WS, Thomson AJ, Adwani SH, Rosseel AJ, Potter JF, Webb DJ, Maxwell SR. Cardiovascular effects of acute oxygen administration in healthy adults. J Cardiovasc Pharmacol 2003; 41 2:245–250.
32. Thomson AJ, Drummond GB, Waring WS, Webb DJ, Maxwell SR. Effects of short-term isocapnic hyperoxia and hypoxia on cardiovascular function. J Appl Physiol 2006; 101 3:809–816.
33. Sutton AD, Bailey M, Bellomo R, Eastwood GM, Pilcher DV. The association between early arterial oxygenation in the ICU and mortality following cardiac surgery. Anaesth Intensive Care 2014; 41 6:730–735.
34. Brown DM, Holt DW, Edwards JT, Burnett RJ. Normoxia vs. hyperoxia: impact of oxygen tension strategies on outcomes for patients receiving cardiopulmonary bypass for routine cardiac surgical repair. J Extra Corpor Technol 2006; 38 3:241–248.
35. Inoue T, Ku K, Kaneda T, Zang Z, Otaki M, Oku H. Cardioprotective effects of lowering oxygen tension after aortic unclamping on cardiopulmonary bypass during coronary artery bypass grafting. Circ J 2002; 66 8:718–722.
36. Ihnken K, Winkler A, Schlensak C, Sarai K, Neidhart G, Unkelbach U, Mülsch A, Sewell A. Normoxic cardiopulmonary bypass reduces oxidative myocardial damage and nitric oxide during cardiac operations in the adult. J Thorac Cardiovasc Surg 1998; 116 2:327–334.
37. Fonnes S, Gögenur I, Søndergaard ES, Siersma VD, Jorgensen LN, Wetterslev J, Meyhoff CS. Perioperative hyperoxia—long-term impact on cardiovascular complications after abdominal surgery, a post hoc analysis of the PROXI trial. Int J Cardiol 2016; 215:238–243.
38. Van De Water JM, Kagey KS, Miller IT, Parker DA, O’Connor NE, Sheh JM, MacArthur JD, Zollinger RM, Moore FD. Response of the lung to six to 12 hours of 100 per cent oxygen inhalation in normal man. N Engl J Med 1970; 283 12:621–626.
39. Nader-Djalal N, Knight PR, Davidson BA, Johnson K. Hyperoxia exacerbates microvascular lung injury following acid aspiration. Chest 1997; 112 6:1607–1614.
40. Allen DB, Maguire JJ, Mahdavian M, Wicke C, Marcocci L, Scheuenstuhl H, Chang M, Le AX, Hopf HW, Hunt TK. Wound hypoxia and acidosis limit neutrophil bacterial killing mechanisms. Arch Surg 1997; 132 9:991–996.
41. Wetterslev J, Meyhoff CS, Jorgensen LN, Gluud C, Lundschou J, Rasmussen LS. The effects of high perioperative inspiratory oxygen fraction for adult surgical patients. Cochrane Database Syst Rev 2015; 25 6:CD008884.
42. Helmerhorst HJ, Schultz MJ, van der Voort PHJ, de Jonge E, van Westerloo DJ. Bench-to-bedside review: the effects of hyperoxia during critical illness. Crit Care 2015; 19:284.

Cytotoxicity; human adult cardiac myocytes; hyperoxia; inflammation; injury; oxygen

Supplemental Digital Content

© 2017 by the Shock Society