Research Article - (2018) Volume 8, Issue 2

Increased Levels of Reactive Oxygen Species in Brain Slices after Transient Hypoxia Induced By a Reduced Oxygen Supply

*Corresponding Author:
Toru Sasaki
Department of Medical Engineering and Technology
Kitasato University School of Allied of Health Sciences
1-15-1 Kitasato, Sagamihara, Kanagawa 252-0373, Japan
Tel: +81 42 778 8157
Fax: +81 42 778 9628
email: tsasaki@ kitasato-u.ac.jp

Abstract

Abstract

Background: Reactive oxygen species (ROS) have been suggested to be involved in cellular damage caused by ischemia-reperfusion, anoxia-reperfusion, and hypoxia-reperfusion. We previously demonstrated that the generation of ROS was enhanced following hypoxia caused by an increased oxygen demand, and this was related to a shift in the tissue redox balance toward reduction. The aim of the present study was to elucidate the relationship among changes in ROS generation, tissue pO2 levels, and redox balance changes in brain slices following hypoxia caused by a decreased oxygen supply.

Methods We measured ROS-dependent chemiluminescence in cerebral cortex slices using a photonic imaging method as well as tissue pO2 levels and the redox balance using micro sensors during reoxygenation after hypoxia caused by the deprivation of an adequate oxygen supply.

Results ROS-dependent chemiluminescent intensity was transiently enhanced during reoxygenation after the hypoxic treatment. Tissue pO2 levels decreased and the tissue redox balance shifted towards reduction with the hypoxic treatment, followed by restoration to the steady-state condition. Increased ROS generation following hypoxia was related to a transient decrease in tissue pO2 levels and a shift in the tissue redox balance towards reduction.

Conclusions The present results demonstrated that ROS generation increased following hypoxia caused by a decreased oxygen supply. In addition, a transient redox shift to “hyper-reduction” with pO2 changes may be involved in ROS generation in tissue.

Keywords

Reactive oxygen species, Hypoxia-reoxygenation, Brain slice, Redox, Hyper-reduction

Introduction

Reactive oxygen species (ROS) are considered to play a significant role in injury responses to ischemia and reperfusion as well as hypoxia and reoxygenation [1-3]. ROS generation increases during reoxygenation after hypoxia caused by a decreased oxygen supply. The status, in which oxygen consumption is enhanced beyond its supply, is regarded as another type of hypoxia. A decrease in tissue oxygen partial pressure (pO2) has been reported in a seizure-like animal model [4,5]. However, it currently remains unclear whether ROS generation in the brain also increases during the restoration of hypoxia caused by a greater oxygen demand.

Ex-vivo brain slices have been used as a wellknown model for electro-physiological studies in animal brain for half century [6]. Such measurements are performed in brain slices being a kind of steady state, which is keeping a living ability. This steady state is usually kept for several hours. In these experiments, slices were submersed with oxygenated 95%O2/5%CO2 media, because tissue pO2 steeply decreased from the surface to tissue core [7-9]. Therefore, the oxygen concentration 95%O2/5%CO2 may be hyperoxic condition at slice surface cells whereas it may be hypoxic at the slice core. The thin-slicing has a merit to derive the sufficient oxygen into the mid of the slice, however it cannot be adopted for slice experiments because it causes neuronal damage by disrupting the long-range connections. The electrophysiological studies with direct measurements of slice tissue oxygenation indicated that a strong dependence of neuronal functions on oxygen in the range of pO2 (10–400 Torr) [10-12] exceeded the physiological pO2 range (10–60 Torr) in brain tissue in vivo [13,14]. These facts suggest that the oxygen concentration requirement for neuronal function is much higher in slices than in vivo.

We previously demonstrated that ROSdependent chemiluminescence in the ex-vivo brain slices increased during the restoration of hypoxia caused by an enhanced oxygen demand using a novel photonic imaging method. Our findings suggested that ROS generation does not directly depend on the energy metabolic rate; it may be related to a transient decrease in tissue pO2 levels and a shift in the tissue redox balance towards reduction. We also reported that ROSdependent chemiluminescence in the brain increased during reoxygenation after hypoxia caused by the deprivation of an adequate oxygen supply [15-19]. However, the relationships among ROS generation, tissue pO2 levels, and the redox balance during the restoration of hypoxia caused by a decreased oxygen supply have not yet been investigated. Therefore, we herein examined superoxide-dependent chemiluminescence in cerebral cortex slices as well as tissue pO2 levels and the redox balance during reoxygenation after hypoxia caused by the deprivation of an oxygen supply.

Materials and Methods

▪ Preparation of brain slices

Three-month-old male Wistar rats were purchased from Japan SLC Inc. (Shizuoka, Japan), and bred in the animal facility of Kitasato University School of Allied of Health Sciences. Animals were sacrificed by decapitation under anesthesia with an intraperitoneal injection of sodium pentobarbital (45 mg/kg body weight), and their brains were rapidly removed and placed on a tissue cutter (Microslicer DTK-3000W; Dosaka EM, Kyoto, Japan). Coronal slices cut at a thickness of 300 μm were transferred into ice-cold Krebs-Ringer solution (124 mM NaCl, 3 mM KCl, 2 mM MgCl2, 1.3 mM NaH2PO4, 26 mM NaHCO3, 10 mM glucose, and 200 mM sucrose) [20] equilibrated with 95% O2/5% CO2. Eight brain slices from each rat were preincubated at 34°C for 45 min in a chamber filled with 50 mL of 95% O2/5% CO2 gas-saturated Krebs-Ringer solution. The animal experimental protocol entitled “Elucidation of the molecular mechanism of reactive oxygen species generation using a molecular imaging technique” was approved by the Kitasato University School of Allied of Health Sciences Animal Care and Use Committee on 28 April, 2014. The approval number is 14-39. All procedures on animals were performed in accordance with the Kitasato University School of Allied of Health Sciences Guide for the Care and Use of Laboratory Animals.

▪ System components

The apparatus comprised a culturing chamber, photon-counting camera (Intensified CCD camera C-2400-35; Hamamatsu Photonics K.K., Hamamatsu, Japan), imaging chamber (temperature-controlled dark box; Aloka Co. Ltd., Tokyo, Japan), and image controller (ARGUS-20; Hamamatsu Photonics K.K.), as previously reported [15]. Brain slices were arranged on a nylon net in the inner chamber and were lightly fixed in place by covering them with a fine net that was stretched and glued to the upper side of a 300-µm-thick stainless-steel ring. The field-of-view of image acquisition was 4 cm × 3 cm and contained 752 × 582 pixels.

Chemiluminescent image acquisition in cerebral cortex slices during normoxia, hypoxia, and reoxygenation.

The treatment of brain slices for the investigation of hypoxia-reoxygenation (30 min normoxia, 15 min hypoxia, and 120 min reoxygenation) was performed as described previously [21]. Briefly, after a 45-min preincubation period, slices were transferred into Krebs-Ringer solution (124 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 1.2 mM KH2PO4, 26 mM NaHCO3, and 10 mM glucose) containing 100 µM of N, N’-dimethyl-9, 9’-biacridinium dinitrate (lucigenin) equilibrated with 95% O2/5% CO2 for normoxic conditions and were incubated for additional 120 min in an imaging chamber at 34oC. They were then incubated under hypoxic (95% N2/5% CO2) conditions for 15 min before being returned to normoxic conditions for 120 min. Images of brain slices were acquired every 15 minutes during normoxia and hypoxiareoxygenation for up to 255 min (11 frames). Images were acquired under a 9 cm × 12 cm field-of-view. A region of interest (ROI) was placed on the cerebral cortex. Chemiluminescent emission was expressed as “counts/pixel/min”, which represented chemiluminescent emission per unit area in 15 min. The steady-state level of chemiluminescent emission under normoxia was expressed as “counts/pixel/min”, which was calculated by averaging the chemiluminescent intensity of eight brain slices from each rat in the 30 minutes prior to the hypoxic treatment, while that of hypoxia-reoxygenation was calculated during the 45-minute post-hypoxic treatment. The experiment was performed using eight different animals, and the values of chemiluminescence emission (counts/pixel/ min) under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (150-195 min) conditions were averaged and expressed as the mean ± SD, as shown in Figure 1.

neuropsychiatry-normoxic-hypoxic

Figure 1: Time course of superoxide-dependent chemiluminescent intensity in the cerebral cortex during oxygenation and hypoxia-reoxygenation (A). Chemiluminescent images in cerebral cortex slices were acquired every 15 minutes under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (135-255 min) conditions. Chemiluminescent intensity was expressed as “counts/pixel/min”, which represents chemiluminescent emission per unit area in 15 min. The experiment was performed using eight different animals, and the value of chemiluminescent intensity was averaged and expressed as the mean ± SD. Comparison of averaged superoxide-dependent chemiluminescent intensities in cerebral cortex slices among normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (150-195 min) conditions (B). The experiment was performed using eight different animals, and the value of chemiluminescent emission (counts/pixel/min) was averaged and expressed as the mean ± SD. The significance of differences was analyzed using a one-way ANOVA with Scheffé’s test (p<0.01** significantly different from normoxic and hypoxic levels).

▪ Measurement of tissue pO2 levels and redox potential in cerebral cortex slices during normoxia, hypoxia, and reoxygenation

Brain slices (400-µm-thick) were prepared from male Wistar rats, as described above. After a 45- min preincubation period, slices were transferred into Krebs-Ringer solution equilibrated with 95% O2/5% CO2 in an imaging chamber at 34oC, as described above. A Clark type oxygen microelectrode (Ox-10; diameter 10 μ; Unisense, Aarhus, Denmark) was used to measure tissue pO2 levels. The electrode was connected to a high-impedance millivoltmeter (Microsensor Multimeter; Unisense, Aarhus, Denmark). The electrode was calibrated by a two-point calibration using 100% oxygen-saturated (100%-point) and 100% nitrogen-saturated (0%-point) distilled water. A redox microelectrode (Ox-10; diameter 10 µm; Unisense, Aarhus, Denmark) and reference Ag/AgCl electrode (REF-321; diameter 8 mm; Unisense, Aarhus, Denmark) connected to a high-impedance millivoltmeter were used to measure tissue redox potential. The electrodes were calibrated by a two-point calibration using quinhydrone redox solutions (1% w/v of quinhydrone in pH 4 or pH 7 buffer solution (HORIBA, Ltd., Kyoto, Japan)). The electrode was positioned on the surface of the cerebral cortex slice, and then lowered at 20-µm increments to a depth of 200 µm using a micromanipulator (Unisense, Aarhus, Denmark). pO2 levels and the redox potential in the cerebral cortex slice were recorded every minute under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (135-255 min) conditions.

The experiment was performed using four different animals, and the values of pO2 (hPa) and redox potential (mV) in cerebral cortex slices at each time point were averaged and plotted in Figures 2 and 3. Mean pO2 and redox potential levels under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (150-195 min) conditions were averaged from four rats and expressed as the mean ± SD (Figures 2 and 3). The significance of differences was analyzed using a one-way ANOVA with Scheffé’s test.

neuropsychiatry-tissue-oxygen

Figure 2: Time course of tissue oxygen partial pressure (pO2) in the cerebral cortex during oxygenation and hypoxia-reoxygenation (A). pO2 was recorded in cerebral cortex slices every minute under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (135-255 min) conditions. Mean pO2 levels under normoxic (90-120 min), hypoxic (150-195 min), and reoxygenic (135-180 min) conditions were averaged from four rats and expressed as the mean ± SD (B). The significance of differences was analyzed using a one-way ANOVA with Scheffé’s test (p<0.01** significantly different from the normoxic level).

Results

▪ Time course of superoxide-dependent chemiluminescent intensity in cerebral cortex slices during normoxia, hypoxia, and reoxygenation

The time course of superoxide-dependent chemiluminescent intensity in cerebral cortex slices collected from rats was shown in Figure 1. The steady-state level of chemiluminescent intensity under normoxic conditions was decreased by the hypoxic treatment, enhanced during reoxygenation, and then reached a maximum (10.5-fold the normoxic level) 30 min after the hypoxic treatment. Average chemiluminescent emission intensity in reoxygenation was 7.6-fold and 42.3-fold those in normoxia and hypoxia, respectively (Figure 1).

Time course of pO2 and redox potential levels in cerebral cortex slices during normoxia, hypoxia, and reoxygenation

The steady-state level of tissue pO2 in the cerebral cortex was significantly decreased by the hypoxic treatment. Decreased pO2 levels were then restored to normoxic levels (Figure 2). The steady-state level of the tissue redox potential under normoxic conditions shifted to a more negative potential (from 252 mV to -87 mV). This redox balance shift towards reduction by the hypoxic treatment was restored to its normoxic level (Figure 3).

neuropsychiatry-redox-potential

Figure 3: Time course of tissue redox potential (mV) in the cerebral cortex during oxygenation and hypoxia-reoxygenation (A). Redox potential was recorded in cerebral cortex slices every minute under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (135-255 min) conditions. Mean redox potential levels under normoxic (90-120 min), hypoxic (120-135 min), and reoxygenic (150-195 min) conditions were averaged from four rats and expressed as the mean ± SD (B). The significance of differences was analyzed using a one-way ANOVA with Scheffé’s test (p<0.01** significantly different from the normoxic level).

Discussion

We previously demonstrated that ROSdependent chemiluminescence in the brain increased during reoxygenation after hypoxia caused by the deprivation of an adequate oxygen supply using a photonic imaging method with lucigen, a chemilumigenic probe [15-19]. Due to its high sensitivity, lucigenin has been frequently used to detect superoxide anion radicals produced by enzymatic and cellular systems in tissue and isolated heart slices [20-26]. Lucigenin reacts not only with superoxide, yielding a product that emits chemiluminescence, but also with hydrogen peroxide under alkaline conditions. However, at physiological pH, lucigenin-derived chemiluminescence represents superoxide levels in cells and tissues [27,28]. We indicated that lucigenin-derived chemiluminescence is derived from intracellular superoxide because lucigeninderived chemiluminescence intensity was significantly decreased by EUK-8, a membranepermeable superoxide dismutase (SOD)/catalase mimic, but not by Cu,Zn-SOD [17]. However, a major limitation of lucigenin is artifactual luminescence generation due to the production of superoxide by redox cycling with molecular oxygen, which may cause a high background [23,29].

Hypoxia is defined as a status in which tissue oxygen concentrations are insufficient to meet demands. Another type of hypoxia is conceivable as enhanced tissue oxygen consumption beyond its supply. Decreases in tissue pO2 levels have been reported in a seizure-like animal model [4,5]. In ex vivo experiments, decreases in pO2 levels, reflecting enhanced neural activity-dependent oxygen consumption, was demonstrated in electrical-, high potassium-, and odor stimulanttreated brain slices [10,30-34], while biphasic changes in tissue pO2 responses, an initial decrease and subsequent increase due to enhanced cerebral blood flow, were found in activated brain regions in in vivo experiments [30,31]. We previously showed that ROS generation may be increased during the restoration of hypoxia caused by a greater oxygen demand [35]. In that study, ROS generation in tissue was enhanced after high potassium-induced hypoxia, and the high potassium treatment induced a transient decrease in tissue pO2 levels as well as a shift in the tissue redox balance towards reduction. We herein investigated whether the tissue redox balance transiently shifted towards reduction due to a decreased oxygen supply and also if its recovery correlated with ROS generation. We also confirmed that superoxide-dependent chemiluminescence in cerebral cortex slices was transiently enhanced in response to tissue pO2 levels and the redox balance (Figure 1-3).

Mitochondria have been suggested as major sites for the intracellular generation of ROS, and the electrons that leak from the mitochondrial respiratory chain cause the partial reduction of molecular oxygen to the superoxide anion [36]. The intra-mitochondrion sites of ROS generation under the restoration process of transiently enhanced energy expenditure have been suggested to be complexes I and III. The blockage of the mitochondrial electron transfer chain enhances ROS generation from upstream of redox centers, whereas it minimizes ROS generation from downstream of these centers [37-41]. Previously, we analyzed lucigenineenhanced chemiluminescence emission in slices of brain tissue prepared from Mn-SOD (located in mitochondria) and Cu, Zn-SOD (located in cytoplasm)-deficient mice to estimate the superoxide levels in mitochondria and cytoplasm. The enhanced chemiluminescence with Mn-SOD and Cu,Zn-SOD deficiency indicated that superoxide level in mitochondria was lower than that in cytoplasm, however the superoxide concentration was assumed to be higher in mitochondrial than the cytoplasm [18]. We also demonstrated that complexes I, III, and IV of the mitochondrial electron transfer chain are major sites for the generation of superoxide during the reoxygenation of hypoxia induced by a decreased oxygen supply, because the chemiluminescence in brain tissues was transiently enhanced by complex I inhibition by rotenone, complex III inhibition by antimycin A, and complex IV inhibition by cyanide. The upstream electron-rich status could be responsible for transiently increased superoxide generation by mitochondrial electron transfer chain inhibition [17].

Electrical stimuli are known to decrease pO2 levels due to the early oxidation of mitochondrial NADH, followed by a prolonged redox balance of the NADH/NAD+ shift towards reduction [33,42] because increased energy generation requires more oxygen and NADH for mitochondrial oxidative phosphorylation. In hippocampal slices, the redox shift of mitochondrial NADH/NAD+ towards reduction also occurs during anoxia, while hyperoxidation has been reported during postanoxia [43,44]. The reduced oxygen supply in brain slices inhibited NAD(P)H oxidation parallel to the decrease in neuronal activity [45]. In in vivo models of seizure and anoxia, the mitochondrial redox balance of NADH/NAD+ was also found to shift towards reduction [46]. High succinate concentration accumulates under ischemia/ hypoxia in various tissues. Upon reperfusion, the accumulated succinate is rapidly re-oxidized by succinate dehydrogenase, driving the ROS generation by reverse electron transport at mitochondria [47]. Recent report indicates that the transition of mitochondrial complex I in the dormant form (D-form) during ischemia to the active form (A-form) after ischemia at early stage of reperfusion plays a significant role in ROS generation. In ischemic tissue, most of the complex I is present in the D-form, it is the state that an over-reduction of the upstream redox centers. If oxygen introduced to the reduced form of complex I ROS generation site, that results in an apparent increase in the rate of superoxide generation rates at flavin site [48]. Increased superoxide and hydrogen peroxide generation from the complex I expected for time needed for slow D-form to A-form transformation and restoration of the normal ubiquinone reductase activity. The ROS-generation activity of complex I depends on the matrix redox potential (NADH/ NAD+). Indeed, the superoxide and hydrogen peroxide generation from the complex I increased with NADH concentration dependence [49].

The results of present and our previous [35] studies suggested that the redox balance shift towards reduction, namely, “hyper-reduction” in tissue may participate in increased superoxide generation under the restoration of transient hypoxia induced by a decreased oxygen supply and increased oxygen demand.

Conclusion

ROS-dependent chemiluminescent intensity was transiently enhanced during reoxygenation after hypoxia induced by a decreased oxygen supply. Increased ROS generation following hypoxia was related to a transient decrease in tissue pO2 levels and a shift in the tissue redox balance towards reduction. We suggest that the redox balance shift towards reduction, namely, “hyper-reduction” in tissue and mitochondria may participate in increased superoxide generation under the restoration of transient hypoxia induced by a decreased oxygen supply and increased oxygen demand.

References