Ⅰ. INTRODUCTION

Oxidative stress is usually induced by excess accumulation of reactive oxygen and nitrogen species in cells. Reactive oxygen species (ROS), especially hydrogen peroxide (H₂O₂), are well-known mediators of oxidative damage and cell demise, nevertheless the mechanisms by which they induce specific cell death modalities and the dose-dependent temporal/ spatial molecular signaling are not clearly elucidated. High concentration of H₂O₂ is known to induce cell death (apoptosis) in a variety of cell types by affecting intracellular signal transduction pathways. In particular, c-Jun N-terminal kinase (JNK) and p38 MAP (mitogen-activated protein) kinase are triggered by extracellular stimuli, e.g., ultraviolet radiation and free radicals including H₂O₂ and NO.^1-5)

H₂O₂, the non-radical 2-electron reduction product of oxygen, is normally produced by aerobic metabolic pathways to the level of about 10 nm intracellular concentration in different cells. In liver, it is produced at 50 μmol/min/kg of tissue, which is about 2% of total oxygen absorption at steady condition.⁶⁾ Metabolically generated H₂O₂ emerged from recent researches as a central hub in redox signaling and/or oxidative stress. Upon generation by major products, the NADPH oxidases or Complex III of mitochondrial respiratory system, H₂O₂ is under fine control of peroxiredoxins and glutathione peroxidases with their feedback systems as well as by different catalases.^7-10) Of note, H₂O₂ is recognized as a second messenger in insulin signaling and in several growth factor-induced signaling cascades.^11-13) H₂O₂ transport across membranes is facilitated by aquaporins, also known as peroxiporins.¹⁴⁾ As specialized protein cysteines operate as redox switches using H₂O₂ as thiol oxidant, this reactive oxygen species important for poising the set point of redox proteome, it was suggested that major cellular processes including proliferation, differentiation, apoptosis, tissue repair, inflammation, survival, circadian rhythm, and aging use this low molecular weight oxygen metabolite, H₂O₂, as signaling compound.⁶⁾

Hydrogen peroxide is known as major redox metabolite for signaling and redox regulation under physiological conditions of H₂O₂ (1-10nM), whereas higher concentrations induce adaptive stress responses via master signaling switches such as Nrf2/Keap1 or NF-kB.¹⁴⁾ Supra-physiological concentrations of H₂O₂ (>100nM) result in damage of biomolecules, represented as oxidative distress.¹⁴⁾

It has been known that the p38 and JNK activations by H₂O₂ are required for apoptosis.^1-3,15,16) Furthermore, it has been reported that H₂O₂ increases the expression of apoptosis-related surface molecule FAS/APO-1 (CD95) and that it induces apoptosis via CD95-mediated signaling.^5,17,18)

In the present in vitro study, the first of all, the optimal biohazard concentration of H₂O₂ was determined by cell viability analysis, which showed favorable cell survival more than 80% in 3 hours and about 50% cell survival (LD50) in 16 hours. The selected H₂O₂ concentration, 40 μM, was applied in RAW 264.7 macrophages, and assessed their protein expressions by IP-HPLC to know the dynamic molecular mechanism of H₂O₂ in cells, which was hardly recognized by ordinary detection methods, e.g., Western blot, RT-PCR (reverse transcription polymerase chain reaction), ELISA (enzyme-linked immunosorbent assay), proteomics, DNA microarray, etc.

Immunoprecipitation high-performance liquid chromatography (IP-HPLC) has been developed to resolute the previous complication in technique and limited applicability.^19,20) A new IP-HPLC protocol was used to assess protein expression levels in various biological fluids, such as urine, blood serum, saliva,²¹⁾ inflammatory exudates,^22-24) and different protein extracts from cells,^25,26) liver,²⁷⁾ and cancer tissues.²⁸⁾ IP-HPLC is similar to ELISA, but the former measures protein expression level using protein A/G agarose beads in buffer solution and UV spectroscopy, whereas the latter uses fluorescence-c onjugated antibodies pre-attached on plastic wells and fluoroscopy. Furthermore, multiple trials have shown that IP-HPLC is available to get multiple results of protein levels accurately (±5% standard deviation) and reproducibly.

Ⅱ. MATERIALS & METHODS

1. RAW264.7 cell culture in the presence of H₂O₂

RAW 264.7 cells, an immortalized murine macrophage cell line (ATCC, USA), were cultured in Dulbecco’s modified Eagle’s medium (DMEM, WelGene Inc. Korea) supplemented with 10% heat-inactivated fetal bovine serum (WelGene Inc. Korea), 100μg/mL streptomycin, 100 unit/mL penicillin, and 250ng/mL amphotericin B (WelGene Inc. Korea) in 5% CO₂ at 37.5°C.²⁶⁾ To detect protein expressional changes, RAW 264.7 cells were cultured in antigen-free media.

About 70% confluent RAW 264.7 cells spreaded on Petri dish surfaces were treated with 40 μM H₂O₂ (Sigma, USA) for 30 min, 1, and 2 hours; the control cells were treated with 1 mL of normal saline. Cultured cells were gathered with protein lysis buffer (PRO-PREPTM, iNtRON Biotechnology INC, Korea) on ice and immediately preserved in deep freezer (-70°C) until required.

2. Determination of optimal concentration of H₂O₂ in RAW 264.7 cell culture by cell viability analysis

A series of MTT assay were carried out using different concentration of H₂O₂ in RAW 264.7 cell culture in order to determine the optimal responsible concentration of H₂O₂ for the cells. Viable cells were detected using MTT dye, which precipitates blue formazan crystals that are reduced by mitochondrial dehydrogenase present in living cells.²⁹⁾

Briefly, 5 x 10³ cells were seeded in a 96-well plate and cultured overnight. The cells were treated with various concentrations of H₂O₂ (0-200μM) for 3 and 16 h until treatment with MTT. MTT assays were performed in accordance with general procedures.²⁹⁾ The reduced MTT was measured at 570 nm spectrophotometry in a micro plate reader (BioTek, USA).

3. Immunoprecipitation high-performance liquid chromatography (IP-HPLC)

About 100μg of protein extracts were utilized for immunoprecipitation using a protein A/G agarose column (Amicogen, Korea). Each protein A/G agarose columns was separately pre-incubated with 0.25 μg of 51 different antisera for; proliferation-related proteins (n=7), cMyc/MAX/MAD network proteins (n=4(1)), epigenetic modification-related proteins (n=6), inflammatory proteins (n=8), apoptosis and survival-related proteins (n=13), and protection-related proteins (n=14(3)) (( ); overlapped number) (Table 1).

Briefly, each protein sample was mixed with 5 mL of immunoprecipitation binding buffer (150mM NaCl, 10mM Tris pH 7.4, 1mM EGTA, 1mM EDTA, 0.2mM PMSF, 0.2mM sodium vanadate, and 0.5% NP-40) and incubated in protein A/G agarose (Amicogen, Korea) columns pre-incubated with different antibodies on a rotating stirrer for 1 hour at 4°C.²⁶⁾ After washing columns with Tris-buffer (50 mM Tris, pH 7.5, 0.15M NaCl), target proteins were eluted using 250μL of IgG elution buffer (Pierce, USA). Immunoprecipitated proteins were detected using a HPLC apparatus (1100 series, Agilent, USA) equipped with a reverse phase silicate column and a micro-analytical detector system (SG Highteco, Korea). HPLC analysis was performed using mobile buffer solution (0.15M NaCl/20% acetonitrile) at 0.4 mL/min for 15 min, and proteins were detected using a UV spectrometer at 280 nm. Control and experimental samples were analyzed sequentially to allow comparisons. For IP-HPLC, immunoprecipitated protein peak areas (mAU*s) were mathematically calculated by subtracting negative control antibody peak areas from the whole protein peak areas, and square roots of immunoprecipitated protein peak areas were calculated to normalize concentrations.^25,26) Protein percentages in total proteins in experimental and control groups were plotted into a line graph and a star plot. Results were analyzed using the Chi-squared test.^25-27)

The housekeeping proteins, α-tubulin, β-actin, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were used as internal controls. Expressional changes of housekeeping proteins were mathematically adjusted to < ±5% using a proportional basal line algorithm. Protein expressional changes of ≤ ±5%, ±5-10%, ±10-20%, and ≥±20% change were defined as minimal, slight, meaningful, or marked, respectively. It was reported that when the IP-HPLC results were compared with the western blot data of cytoplasmic housekeeping protein (β-actin), the IP-HPLC results exhibiting minute error ranges less than ±5% could be analyzed statistically, while the western blot results showed a large error range of more than 20%, thereby it was almost impossible to analyze them statistically.²⁶⁾

4. Statistical analysis

Proportional data (%) of experimental and control groups were plotted into line graphs and star plots, and results were analyzed using the Chi-squared test. The expressions of control housekeeping proteins, that is, α-tubulin, β-actin, and GAPDH were almost non-responsive (≤5%) to 30 min, 1, and 2 hours of 40 μM H₂O₂ treatment.

Ⅲ. RESULTS

1. Determination of optimal concentration of H₂O₂ in RAW 264.7 cell culture

In order to know the reactive cellular responses against H₂O₂-induced oxidative stresses in RAW 264.7 cells, the first of all, the optimal concentration of H₂O₂ was determined by MTT assay for cell viability test. Among 0 to 200 μM H₂O₂ applied to RAW 264.7 cells for 3 and 16 hours, 40 μM H₂O₂ showed 85.7% cell survival in 3 hours and LD50 cell survival (48%) in 16 hours (Fig. 1), therefore, the optimal responsible concentration of H₂O₂ was determined as 40 μM in this study for the H₂O₂-induced oxidative stresses in RAW 264.7 cells.

2. Effects of 40 μM H₂O₂ on the expressions of proliferation-related proteins in RAW 264.7 cells

RAW 264.7 cells treated with 40 μM H₂O₂ for 30 min, 1, or 2 hours exhibited decreases in levels of proliferationactivating proteins, Ki-67 (by 9.7% at 2 hours), proliferation cell nuclear antigen (PCNA, 3.4% at 1 hour), cyclin dependent kinase 4 (CDK4, 9.8% at 2 hours), polo-like kinase 4 (PLK4, 18.9% at 1 hour), and cyclin D2 (9.7% at 30 min), and increases in proliferation-inhibiting proteins, p15/16 (42.9% at 2 hours) and p27 (9.3% at 30 min) levels versus non-treated controls. Particularly, GAPDH necessary for important energy- yielding step in carbohydrate metabolism was downregulated by 40 μM H₂O₂, 15.7% at 2 hours, contrast to other housekeeping proteins, β-actin and α-tubulin which were minimally changed (< ±7%) (Figs. 2 A1 and A2).

The expressions of proliferation-activating proteins, Ki-67, PCNA, PLK4, and cyclin D2 were decreased after 30 min of 40 μM H₂O₂ treatment, while the expressions of proliferationinhibiting proteins, p15/16 and p27 were markedly increased, so that these results suggest 40 μM H₂O₂ might have a strong anti-proliferative effect on RAW 264.7 cells.

3. Effects of 40 μM H₂O₂ on the expressions of cMyc/MAX/MAD network-related proteins in RAW 264.7 cells

RAW 264.7 cells treated with 40 μM H₂O₂ for 30 min, 1, or 2 hours exhibited decreases in the expressions of cMyc and MAX by 30.1% and 9.2%, respectively, after 2 hours treatment versus non-treated controls, while MAD-1 expression increased by a maximum of 18.9% at 30 min and maintained by 15.9% at 2 hours of treatment (Figs. 2 B1 and B2). The reduction of cMyc and MAX expressions while elevation of MAD-1 expression in cMyc/MAX/MAD network may contribute to the anti-proliferative effect of 40 μM H₂O₂ on RAW 264.7 cells. And subsequently, p27 expression was upregulated by 9.3% at 30 min but gradually decreased by 11.6% at 2 hours.

4. Effects of 40 μM H₂O₂ on the expressions of epigenetic modification-related proteins in RAW 264.7 cells

Histone H1 expression slightly decreased in 40 μM H₂O₂ treated cells to 91.3% at 1 hour and to 94.7% at 2 hours versus non-treated controls. Regarding histone methylation/ deacetylation, the expression of lysine-specific demethylase 4D (KDM4D) was increased by 13.8% at 30 min but gradually decreased by 12% at 2 hours, and that of histone deacetylase 10 (HDAC-10) and DNA methyltransferase 1-associated protein 1 (DMAP1) were similarly increased by 46.7% and 60.7%, respectively, at 1 hours and maintained by 21.1% and 32.1%, respectively, at 2 hours versus non-treated controls. With respect to DNA methylation, DNA (cytosine-5)- methyltransferase 1 (DNMT1) expression was greatly reduced by 43.5% and 34% at 30 min and 1 hour, respectively, but its decrease was recovered by 0.1% at 2 hours. And the expressions of methyl-CpG binding domain 4 (MBD4) was 10.2% higher at 2 hours (Figs. 2 C1 and C2). These results suggest 40 μM H₂O₂ increased histone methylation/deacetylation and DNA methylation simultaneously, and subsequently repressed DNA transcription in RAW 264.7 cells, and that this epigenetic effect of 40 μM H₂O₂ might be associated with genome-wide DNA damage.

5. Effects of 40 μM H₂O₂ on the expressions of inflammation-related proteins in RAW 264.7 cells

The most inflammatory proteins were upregulated by 40 μM H₂O₂ treatment; tumor necrosis factor α (TNFα, 11.9%), lysozyme (6.3%), leukotriene A4 hydrolase (LTA4H, 6.2%), C-X-C chemokine receptor type 4 (CXCR4, 11.9%), macrophage colony-stimulating factor (M-CSF, 6.9%), CD68 (13.4%), cyclooxygenase-2 (COX-2, 20.1%), and monocyte chemotactic proteins (MCP-1, 18.9%) versus non-treated controls. Particularly, the expressions of TNFα, CXCR4, and M-CSF were decreased by 9.2%, 10.3%, and 3.5% at 30 min, respectively, but reactively increased by 11.9%, 11.9%, and 6.9% at 2 hours (Figs. 3 A1 and A2). Resultantly, 40 μM H₂O₂ markedly upregulated the expressions of M1 macrophage differentiation cytokines, e.g., TNFα, LTA4H, CXCR4, CD68, COX-2, and MCP-1, indicating that RAW 264.7 cells were activated into M1 type macrophages by 40μM H₂O₂ treatment.

6. Effects of 40 μM H₂O₂ on the expressions of apoptosis and survival-related proteins in RAW 264.7 cells

40 μM H₂O₂ affected the expressions of p53-mediated apoptosis-related proteins for 2 hours treatment, particularly p53 protein, which was increased by 14% at 1 hour and maintained by 9.2% at 2 hours versus non-treated controls. The expressions of BAX and APAF-1 were increased by 16% at 30 min and 13% at 1 hour, respectively, while BCL2 expression was decreased by 18.4% at 1 hour. These expressions were resulted in the increases of caspase 9 (12.3%) and c-caspase 9 (8.6%) expressions (Figs. 3 B1 and B2). On the other hand, the expressions of PARP-1 and AIF were markedly increased by 27.2% at 1 hour and 48.2% at 2 hours, respectively, which suggest the increase of ssDNA damage and nuclear apoptosis of parthanatos.

RAW 264.7 cells treated with 40 μM H₂O₂ showed upregulation of cell survival-related proteins as compared with non-treated controls. The expressions of pAKT1/2/3, telomerase reverse transcriptase (TERT), nuclear factor (erythroid-derived 2)-like 2 (NRF2), and AMP-activated protein kinase (AMPK) were increased by 19.7% at 2 hours, 13.9% at 30 min, 13.7% at 2 hours, and 8.5% at 1 hour, respectively. These results suggest that 40 μM H₂O₂ induced survival signaling in RAW 264.7 cells.

7. Effects of 40 μM H₂O₂ on the expressions of cell protection-related proteins in RAW 264.7 cells

The expressions of cell protection-related proteins in RAW 264.7 cells were increased by 40 μM H₂O₂; heat shock protein-70 (HSP-70) by 10.7% at 2 hours, HSP-90 by 14.3% at 30 min, heme oxygenase 1 (HO-1) by 17.8% at 1 hour, glutathione S-transferase ω 1 (GSTO1) by 18.1% at 30 min, Cu-Zn superoxide dismutase-1 (SOD-1) by 43.1% at 30 min, NRF2 by 13.7% at 2 hours, LC by 29.5% at 1 hour, sodiumdependent vitamin C transporter 2 (SVCT2) by 15.1% at 2 hours, leptin by 13.7% at 2 hours, and AMPK by 8.5% at 1 hour as compared with non-treated controls, while the expressions of nitric oxide synthase 1 (NOS1), nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB), and mammalian target of rapamycin (mTOR) were decreased by 9.3% at 1 hour, 4.8% at 30 min, and by 9.3% at 30 min, respectively. Particularly, the expressions of antioxidantrelated proteins including heme oxygenase-1 (HO-1), glutathione S-transferase ω 1 (GSTO1), Cu-Zn superoxide dismutase-1 (SOD1), and sodium-dependent vitamin C transporter 2 (SVCT2) were consistently elevated, NRF2, a transcription factor the expression of antioxidant proteins was gradually increased for 2 hours, and the expressions of autophagosome protein (LC3) and anti-obesity protein (leptin) were markedly increased by 40 μM H₂O₂. (Figs. 3 C1 and C2).

Ⅳ. DISCUSSION

RAW 264.7 cells, murine macrophage cell line, are invaluable biological models for mechanistic studies of oxidative stress, because they usually play roles for anti-bacterial and scavenging effects using ROS. RAW 264.7 cells are relatively resistant to the hazard reaction of low H₂O₂ concentration but eventually succumbed to high H₂O₂ concentration similar to other cells. In order to know the primitive molecular signaling occurred by H₂O₂ treatment in RAW 264.7 cells, the optimal H₂O₂ concentration was determined by cell viability analysis and 40 μM H₂O₂ produced favorable cell survival over 80% for 3 hours and LD50 survival for 16 hours. Therefore, the present study was performed the H₂O₂-induced biohazard experiments using 40 μM H₂O₂ for 2 hours in RAW 264.7 cells. It was estimated that the cells would actively respond to the exogenous ROS derived from H₂O₂ treatment, survive well for 2 hours of experiment, represent early stages of molecular signaling for cellular proliferation, protection, apoptosis, survival, and adaptation in this study

It is known that high levels of exogenous H₂O₂ induce extensive DNA damage, ATP depletion, and severe cytotoxicity, while low levels of exogenous H₂O₂ trigger a modest drop in ATP activity, delayed toxicity, G2/M arrest, and cell senescence.³¹⁾ Oxygen free radicals derived from mitochondria induced a reversible ATP drop without affecting cell viability. Instead, cells became senescent after accumulating in the G1/S phase of the cell cycle.^32-34) Concomitant inhibition of glycolysis was shown to make cells highly sensitive to death in the presence of different nontoxic concentrations of H₂O₂, perhaps by inhibiting the ATP-restoring mechanisms.^31,35)

In the present study, 40 μM H₂O₂ decreased the expressions of proliferation-activating proteins, Ki-67, PCNA, CDK4, cyclin D2, and PLK4, while increased the expressions of proliferation-inhibiting proteins, p15/16 and p27. Concurrently, cMyc and MAX were downregulated by H₂O₂, while MAD-1 expressions was upregulated. These expressional changes may indicate RAW 264.7 cell proliferation was potently arrested by H₂O₂ together with diminishment of carbohydrate metabolism in cells. As p27 expression is intimately regulated by cMyc/MAX/MAD network signaling, the increase of p27 expression may closely related to the reduction of MAX homodimer formation but the elevation of MAX-MAD 1 heterodimer formation in cMyc/MAX/MAD network.^36,37)

40 μM H₂O₂ greatly influenced on the inflammatory protein expressions of RAW 264.7 cells, murine macrophages. Most of inflammatory cytokines were upregulated for 2 hours of H₂O₂ treatment, among the affected inflammatory proteins, the increases in the expressions of M1 differentiation cytokines, e.g., TNFα (11.9%), LTA4H (6.2%), CXCR4 (11.9%), CD68 (13.4%), COX-2 (20.1%), and MCP-1 (18.9%) more dominant than the minimal increases of M2 differentiation cytokines, e.g., lysozyme (6.3%), and macrophage colony-stimulating factor (M-CSF, 6.9%). These expressional changes may indicate that 40 μM H₂O₂ potently activate RAW 264.7 cells to differentiate into M1 type macrophages which are utilized for the first line defense of inflammatory reaction.^38,39)

It has been reported that apoptosis is closely associated with oxidative stress in biological systems, and that caspases have been considered to play a pivotal role in the execution phase of apoptosis.^3,40,41) In the present study, the expressions of p53-mediated apoptosis proteins, p53, BAX, APAF-1, caspase 9, and c-caspase 9, were increased after 30 min of 40 μM H₂O₂ treatment, but these expressions were diminished until 2 hours. On the other hands, PARP-1 and AIF were dominantly upregulated at 1 and 2 hours of 40 μM H₂O₂ treatment.

H₂O₂ upregulates Bfl-1, an anti-apoptotic member of the BCL-2 family, which is responsible for the anti-apoptotic activity against ROS. When Jurkat cells and human leukemic T cells, were pretreated with 100 μM H₂O₂ and then treated with anti-FAS antibody, apoptosis was impaired without change of cell surface Fas expression.¹⁸⁾ And 50 μM H₂O₂ increased telomerase activity, so that this expression was suspected for the BCL-2 increment and the oncogenic reactivation.⁴²⁾ In the present study, 40 μM H₂O₂ increased TERT expression by 13.9% at 30 min but decreased it by 5.3% at 1 hour and 9.1% at 2 hours, and the expression of BCL-2 was consistently downregulated for 2 hours. Therefore, it was considered that the transient elevation of TERT expression by 40 μM H₂O₂ would be only reactive to the apoptotic impact of ROS.

In the present study, 40 μM H₂O₂ induced the intense expressions of cell protection proteins, including heat shock proteins (HSP-70 and HSP-90), antioxidant-related proteins (HO-1, GSTO1, SOD-1, and SVCT2), autophagosome-related protein (LC3), cellular energy homeostasis-related protein (AMPK), and a transcription factor regulating for the expression of antioxidant proteins (NRF2). Increasing evidence has demonstrated that heme oxygenase-1 (HO-1) is a major enzyme induced by cellular stress, exhibits cellular protective, antioxidant, and anti-inflammatory abilities.⁴³⁾ HO-1 significantly restored the cell viability under H₂O₂ injury via reducing the generation of ROS and increasing the levels of SOD and glutathione (GSH) activity.⁴³⁾

H₂O₂-stimulated AMPK activation is pro-survival and antiapoptosis in osteoblasts. Autophagy induction and NADPH maintenance are involved in AMPK-mediated pro-survival effects.⁴⁴⁾ H₂O₂-induced increases in JNK phosphorylation and cell apoptosis were largely activated by ASK1 overexpression whereas the similar transfection did not affect ROS generation in the cells. Moreover, inhibition of H₂O₂-increased ROS generation, JNK phosphorylation, and cellular apoptosis by upregulation of heat shock transcription factor-1 HSF1 tended to be disappeared.⁴⁵⁾ HSP-70 protects cardiomyocytes from apoptosis under oxidative stress by reducing intracellular ROS generation and by inhibiting p38 phosphorylation.⁴⁶⁾ The present study showed that 40 μM H₂O₂ induced the expressions of cell protection-related proteins (HSP-70, HSP-90, AMPK, etc.) and phase II type antioxidant proteins (HO-1, GSTO1, and SOD-1), which conjugate drug metabolites or endobiotics via acetylation, glucuronidation, glutathionylation, methylation, and sulphation, making them more hydrophilic; or degrade heme or quinone.⁴⁷⁾

This preliminary study was aimed to determine the optimal concentration of H₂O₂, which can give minimal ROS impact on RAW 264.7 cells in order to observe the reactive molecular signaling of viable cells maintaining their cellular proliferation, protection, apoptosis, and survival. 40 μM H₂O₂ determined in this study showed significant ROS injury on RAW 264.7 cells, resulted in the repression of proliferation, inactivation of epigenetic modification, and caspase 3/PARP-1/AIF-associated apoptosis, parthanatos. On the other hand, 40 μM H₂O₂ activated RAW 264.7 cells to be differentiated into M1 type macrophages, induced cell survival signaling via marked increases in the expressions of heat shock proteins, antioxidantrelated proteins, autophagosome-related protein, cellular energy homeostasis-related protein, and a transcription factor regulating for the expression of antioxidant proteins. Therefore, it is suggested that 40 μM H₂O₂ is proper concentration to investigate the early stage of H₂O₂-induced ROS biohazard in RAW 264.7 cells. However, further precise molecular biological investigation should be followed in order to elucidate H₂O₂-induced molecular signaling mechanisms in cells.