Oxaloacetate Decreases the Infarct Size and Attenuates the Reduction in Evoked Responses after Photothrombotic Focal Ischemia in the Rat Cortex

Abstract

A traumatic brain injury or a focal brain lesion is followed by acute excitotoxicity caused by the presence of abnormally high glutamate (Glu) levels in the cerebrospinal and interstitial fluids. It has recently been demonstrated that this excess Glu in the brain can be eliminated into the blood following the intravenous administration of oxaloacetate (OxAc), which, by scavenging the blood Glu, induces an enhanced and neuroprotective brain-to-blood Glu efflux. In this study, we subjected rats to a photothrombotic lesion and treated them after the illumination with a single 30-min-long administration of OxAc (1.2 mg/100 g, i.v.). Following induction of the lesion, we measured the infarct size and the amplitudes of the somatosensory evoked potentials (SEPs) as recorded from the skull surface. The photothrombotic lesion resulted in appreciably decreased amplitudes of the evoked potentials, but OxAc administration significantly attenuated this reduction, and also the infarct size assessed histologically. We suggest that the neuroprotective effects of OxAc are due to its blood Glu-scavenging activity, which, by increasing the brain-to-blood Glu efflux, reduces the excess Glu responsible for the anatomical and functional correlates of the ischemia, as evaluated by electrophysiological evoked potential (EP) measurements.

Introduction

Glutamate (Glu), the major excitatory amino acid neurotransmitter in the central nervous system, mediates a number of physiological processes, and it is involved in the pathological process of excitotoxicity (Choi 1988; Choi 1994). The nerve cells are damaged and killed by the excess Glu present in the brain fluids after various acute and chronic neurodegenerative disorders. The latter brain insults include, among others, stroke (Castillo et al. 1996; Castillo et al. 1997; Choi and Rothman 1990), traumatic brain injury (Bullock et al. 1998), amyotrophic lateral sclerosis (Shaw et al. 1995; Spreux-Varoquaux et al. 2002), and HIV dementia.

In acute ischemic insults, Glu causes irreversible neuronal damage, which is generally observed at the histological level by the presence of a cortical infarction, and at the electrophysiological level by the correlated loss of somatosensory evoked potentials (SEPs) (Lye et al. 1987), cortical disinhibition (Farkas et al. 2003), and peri-infarct depolarizations (Mies et al. 1993). Since the most of the functional consequences of ischemia can be blocked by the administration of Glu receptor antagonists (Bordi et al. 1997; Molchanova et al. 2004), one can expect that the elimination of the excess Glu present in the brain interstitial fluids upon ischemia will be of beneficial value.

Recently, blood Glu scavenging has been shown to cause an increased brain-to-blood Glu efflux (Gottlieb et al. 2003), to eliminate the excess Glu in the brain interstitial fluids (Teichberg et al. 2009), and to provide neuroprotection after a traumatic brain injury (Zlotnik et al. 2007; Zlotnik et al. 2008). The intravenous administration of oxaloacetate (OxAc) or pyruvate, which activates the blood-resident glutamate–oxaloacetate transaminase and glutamate–pyruvate transaminase, respectively, leads to blood Glu scavenging by the transamination of Glu into 2-ketoglutarate, as a result of an accelerated efflux of the excess brain Glu into the blood, and neuroprotection is achieved.

On these premises, we tested here the prediction that the intravenous administration of the blood Glu scavenger OxAc should cause neuroprotection after an acute ischemic insult, and improve both histological and functional manifestations of ischemia, as evaluated by Fluoro-Jade B (FJB)-staining histology and the measurement of SEPs. In this study, the histology was carried out as early as 4 h after the lesion. The reasons for such early monitoring were twofold: one is that magnetic resonance-diffusion-weighted imaging clearly delineates cerebral ischemic lesions as early as 90 min post-occlusion (Back et al. 2004), and the diffusion-weighted imaging can be used as a marker of successful neuroprotective drug action (Ebisu et al. 2001; Muller et al. 1995). The second reason is that previous publications on blood Glu scavengers (Zlotnik et al. 2008; Zlotnik et al. 2007) emphasized the fact that the major neuroprotective effects of these compounds (OxAc and pyruvate) are readily observed within the first few hours.

As acute ischemic insult, we elected to use the photothrombosis model introduced by Watson et al. (1985), since it has the advantages of being noninvasive and of producing lesions with good reproducibility. During photothrombosis, a photosensitive dye circulating through the cerebral vasculature is exposed to an externally applied light beam. Light exposure generates highly reactive oxygen radicals in the local blood stream. The reactive oxygen species disrupt the capillary endothelium, causing microvascular platelet aggregation and disruption of the blood–brain barrier. Moreover, photothrombosis results in the occurrence of repetitive episodes of cortical spreading depression and is accompanied by a massive rise in extracellular Glu level lasting several hours (Scheller et al. 2000). The photothrombosis was induced and electrophysiological recordings were made without opening the skull.

Materials and Methods

Materials

OxAc and Rose Bengal sodium were purchased from Sigma-Aldrich (Munich, Germany). Urethane was from Reanal (Hungary).

Animals

The experimental procedure used in this study adhered to the protocol for animal care approved by the Hungarian Health Committee (1998) and the European Communities Council Directives (86/609/EEC). Wistar male rats (200–250 g) were used in these experiments. The animals for electrophysiology were divided into three groups: the sham-operated animals (n = 4), the photothrombotic-lesioned group (n = 5), and the OxAc-treated group (n = 5). For the histological studies, 12 rats received a photothrombotic lesion, while eight animals received a photothrombotic lesion + OxAc administration. All the efforts were made to minimize the number of animals used and their suffering. The animals were kept under 12-h light and 12-h dark conditions, with lights on at 7 a.m., and were raised with free access to water and food pellets. The room temperature was 22 ± 1°C.

Surgical and Experimental Procedures

All of the surgical procedures were carried out under deep anesthesia. During the experiments, the rats were anesthetized with an intraperitoneal urethane injection (1.3 g/kg body weight), and their heads were then fixed in a stereotaxic head-holder (David Kopf Instr.). On the left side, the skull was exposed above the primary somatosensory cortex (SSI) by removal of the skin and connective tissue. The cortical photothrombotic lesion was induced by the tail vein injection of Rose Bengal (3 mg/100 g) and cold light exposure. The illumination (through the intact skull, with a cold light source) was started just after the Rose Bengal injection, and lasted for 20 min. During the experiment, the rectal temperature was held constant at 37.0 ± 0.2°C. In the OxAc-treated group, the animals received OxAc solution (1.2 mg/100 g, i.v., pH 7.2–7.4) at a constant rate through the tail vein, during a 30-min period. This dose was used because it was shown in previous studies to cause both blood Glu scavenging (Gottlieb et al. 2003) and neuroprotection (Zlotnik et al. 2007). The OxAc injection started just after the cold light exposure.

Four hours later, the animals were transcardially perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (Fig. 1a). The brains were removed and post-fixed overnight in the same fixative. Coronal sections (36 μm) were cut with a freezing microtome (Frigomobil Model 1206, Reichert-Jung, Nußloch, Germany) and the sections were stained with FJB (Chemicon, Millipore Ltd, Hungary).

Fig. 1.

The effect of oxaloacetate (OxAc) administration (1.2 mg/100 g; i.v.) on the cortical lesion produced by Rose Bengal-induced photothrombosis. a The protocol used for the photothrombotic lesion. b OxAc treatment resulted in a decrease in the extent of Fluoro-Jade B (FJB) staining in each coronal slice of the cortex including the lesion. The border of the staining is denoted by white lines in a representative slice from a lesioned animal (lesioned) and from a lesioned + OxAc-treated (OxAc) rat. c OxAc treatment resulted in a decrease in the volume of the lesion by about 30% (calculation based on serial sections). The reduction was significant. Means ± SEM, * P < 0.05. Lesioned group: n = 12 (2448 slices), lesioned + OxAc treated group: n = 8 (1536 slices)

Fluoro-Jade B Staining

Rehydrated sections were immediately transferred to a solution of 0.06% potassium permanganate for 15 min. The sections were then rinsed in distilled water for 2 min and placed in a 0.001% FJB staining solution made by adding 10 ml of a 0.01% stock solution of FJB to 90 ml of 0.1% acetic acid. After 30 min in the FJB staining solution, the sections were rinsed through three changes of distilled water for 1 min per change. Next, the sections were air–dried overnight and coverslipped with Fluoromount (SERVA, Germany).

The sections were subsequently analyzed with a fluorescent microscope (BX51, Olympus, Tokyo, Japan) at an excitation wavelength of 470–490 nm and an emission wavelength of 520 nm. The volume of the hemispheric lesion and the number of FJB-positive cells were calculated for each animal.

Electrophysiology

SEPs induced in the contralateral primary SSI by electrical stimulation of the right whisker pad were recorded from the surface of the skull. The details relating to the stimulation and cortical recordings were published earlier (Farkas et al. 2000; Toldi et al. 1999). In brief, the right whisker pad was stimulated electrically with a bipolar needle electrode (0.1 Hz, 0.3 ms duration and 4–6 V) to evoke visible whisker movements. Stimulation for 1 h was applied prior to the start of the experiment in order to attain equilibrium. The recordings were made on the surface of the skull with the aid of a silver electrode. Electrode paste was used to provide appropriate contact. The seven recording points were situated along a virtual line parallel to and 5 mm laterally from the midline, including the SSI, i.e., at the points 0, −1, −2, −3, −4, −5, and −6 mm frontal to the bregma (Fig. 3b) in all the animals. These points are localized above the upper lip region, in the barrel field of the SSI and partly above the parietal association cortex (Paxinos and Watson 1998). The amplified responses were fed into a computer via an interface (Digidata 1200, pClamp 6.0.4. software, Axon Instruments, USA) and stored for further processing. The EPs (6/min) were observed during 3-min periods at all the recording points. The data were averaged for each 3-min period (18 EPs altogether at each point). Control series were recorded in each group prior to induction of the photothrombotic lesion. The following recording period started 1 h later, after the induction of the photothrombotic lesion, while the successive series of recordings started at 2, 3, and 4 h post-insult (Fig. 3a).

Fig. 3.

Schematic drawing showing the protocol of the electrophysiological experiments, the recording points and the location of the illuminated area on the skull. a Protocol of the experiments. Somatosensory evoked potentials (SEPs) were recorded before induction of the cortical lesion (−60 min). Rose Bengal injection (3 mg/100 g, i.v.) was followed by cold light illumination for 20 min. OxAc administration (1.2 mg/100 g, i.v.) followed for 30 min. In the “lesioned” group, which did not receive OxAc, a 30 min pause followed, and during the subsequent 240 min, SEPs were recorded hourly in both groups from the coordinates shown in panel b. b Potentials were recorded along a virtual line 5 mm laterally from the midline. The punctum maximum of the responses (D) is localized at the coordinates: frontal-3 (3 mm behind the bregma) and lateral five (5 mm laterally to the midline). Rostrally and caudally from the punctum maximum, (C, B, A) and (E, F, G), respectively, the amplitudes of the EPs gradually decreased

Statistical Analysis

To test the statistical significance of differences between the responses at the punctum maximum in the different groups, as a function of time (Fig. 4a), one-way ANOVA followed by the post hoc Bonferroni test was applied. When the EPs were compared inside the groups at the various coordinates, the paired-sample t-test was applied (Fig. 4b, c). A P value of < 0.05 was considered significant.

Fig. 4.

Amplitudes of the EPs of the sham-operated, photothrombotic-lesioned, and lesioned + OxAc-treated animals. a Amplitudes of the EPs at the punctum maximum of the sham-operated (control), the lesioned, and the lesioned + OxAc-treated animals as a function of time. The percentage changes in the amplitudes were compared with the controls, i.e., the respective EP amplitudes at t = 0. The normalized means of the EPs are presented. b Amplitudes of the EPs recorded at seven coordinates in the control period (values at the punctum maximum were taken as 100%), and at 1 h and 4 h after the intervention in the lesioned animals. c Amplitudes of the EPs recorded at seven coordinates in the control period (values at the punctum maximum were taken as 100%), and at 1 h and 4 h after the intervention in the lesioned + OxAc-treated animals. Means ± SEM. a * Significant difference relative to the sham-operated group; ^# significant difference between the lesioned and lesioned + OxAc-treated groups. b * Significant difference from the control period; ^# significant difference between 1 and 4 h. c * Significant difference from the control period; ^# significant difference between 1 and 4 h. *^{, #} P < 0.05; **^, ## P < 0.01

Results

Figure 1 shows the protocol used for the photothrombotic lesion (panel a) and the extent of FJB staining with or without treatment with OxAc (panels b and c). FJB is a polyanionic fluorescein derivative which sensitively and specifically binds to damaged neurons, with increased contrast and resolution during acute neuronal stress. It may label activated microglia too if used at later times (Damjanac et al. 2007). One can observe very distinct FJB staining (denoted by two white lines) indicative of the presence of neuronal cell damage. Serial analysis of the FJB staining in the contiguous brain sections carried out to determine the lesion volume revealed that the treatment with OxAc reduced the volume of the thrombotic lesion by about 30% (Fig. 1c). At higher resolution, the counting of the FJB- positive cells confirmed these results and showed that the OxAc treatment reduced the number of stained cells by about 30% (Fig. 2a, b). Inspection of Fig. 2a reveals a conspicuous labeling of capillaries, which is unexpected, since FJB staining is well known to be highly selective to neurodegenerating neurons. The most likely explanation is that the capillary labeling is not due to FJB staining, but due to the fluorescence of the erythrocytes trapped within the capillaries as a result of photoactivated platelet aggregation. The photoactivation of Rose Bengal is well known to cause peroxidation of various cellular targets in a singlet oxygen-driven process (Girotti et al. 1985; Rozanowska et al. 1999; Wright et al. 2003). In the case of peroxidation of the erythrocytes, the reaction is accompanied by an increase in erythrocyte fluorescence at 525 nm following excitation at 460 nm (el-Rahman et al. 1995; Nagababu and Rifkind 1998; Nagababu and Rifkind 2004). These spectral properties are almost equivalent to those of FJB. A comparison of the capillary staining in A1, A2 (Fig. 2) reveals a more fragmented labeling in A2 than in A1, which may be suggestive of better blood perfusion in the brain of the OxAc-treated rats. However, a direct study of blood perfusion is needed to establish this point with certainty.

Fig. 2.

OxAc treatment reduced the number of FJB-positive cells in the lesioned area of the cortex. a Labeled neurons (and capillaries) in a photothrombotically lesioned cortex (A1, lesioned) and in a lesioned cortical area of an animal which received OxAc (A2, OxAc). Both images are representative examples. b The statistical analysis revealed a significant decrease in the number of FJB-positive cells/mm². Means ± SEM, * P < 0.05. Lesioned group: n = 12 (2448 slices), lesioned + OxAc treated group: n = 8 (1536 slices)

We then proceeded to measure the functional consequences of the photothrombotic lesion. For this purpose, we measured the responses evoked in the contralateral hemisphere by stimulation of the C2 and C5 whisker follicles in control rats. SEPs were recorded at seven points (Fig. 3b) with maximal amplitudes (punctum maximum) obtained at the coordinates frontal: -3 mm, lateral: 5 mm with respect to the bregma.

The amplitudes of these SEPs facilitated slightly in the course of the 4-h experiments (Fig. 4a, control), during which repeated stimulations were carried out at 1-h intervals.

The above experiments were repeated on rats subjected to a photothrombotic lesion, using a protocol described in Fig. 3a. The diameter of the illuminated area on the skull was 3 mm, centered at the punctum maximum (Fig. 3b). The photothrombotic lesion resulted in an immediate reduction in the amplitude of the EPs measured at the punctum maximum (Fig. 4a) and at all the other points (Fig. 4b). The amplitudes of the EPs at the punctum maximum were decreased to 26% of the control level after the illumination, but subsequently increased slightly and after 2 h had reached 40–43% of the control level, where they remained until the end of the experiment (Fig. 4a).

Rats treated with OxAc after induction of the photothrombotic lesion displayed less profound decreases in the EP amplitudes at the punctum maximum than those in the lesioned rats (Fig. 4a), and the EPs exhibited a very distinct recovery, reaching 80–82% of the control levels at 3–4 h post-illumination (Fig. 4c, lesioned + OxAc). Normalization of the EP amplitudes at the punctum maximum to those of the control EPs measured during pre-illumination revealed that the extent of recovery of the EP amplitudes measured at 3–4 h post-illumination in the lesioned and OxAc-treated rats was much higher than that in the lesioned and untreated rats (Fig. 5).

Fig. 5.

Changes in EP amplitudes at all the recording points as a function of time. a Rostrally and caudally from the punctum maximum, the amplitudes gradually became smaller throughout the experiments. b The photothrombotic lesion resulted in severe decreases in the EP amplitudes. A slight time-dependent recovery was nevertheless observed. c In the OxAc-treated animals, smaller decreases in the EP amplitudes were observed after induction of the photothrombotic lesion

Comparison of the EP amplitudes of the sham-operated (control) animals with those in the lesioned and lesioned + OxAc-treated groups indicated that the patterns of changes at all the recorded points were similar to those observed at the punctum maximum (Fig. 4a–c). To summarize, normalization of the amplitudes measured at the various coordinates to those measured at the punctum maximum demonstrated that the response amplitudes gradually became smaller in both the rostral and caudal directions throughout the experiment, and in the lesioned group, but not in the lesioned + OxAc-treated group (Fig. 5a–c), were significantly different from those in the sham-operated control group.

Discussion

Focal cerebral ischemia induces a complex series of events that damage the brain cells. The very early events include massive depolarization due to Glu release and the loss of SEPs (Minamide et al. 1994). In photothrombotic ischemia, repetitive episodes of cortical spreading depression lasting up to 3 h are observed and are accompanied by a massive rise in extracellular Glu level lasting up to 5 h post-illumination (Scheller et al. 2000). The role of Glu is clearly critical since the anatomical correlates of the photothrombotic ischemia can be significantly attenuated following the administration of Glu receptor antagonists (Kharlamov et al. 1996; Stieg et al. 1999; Umemura et al. 1997). Similarly, the results of Bordi et al. (1997) also underline the importance of NMDA receptor blocking in neuroprotection. They used the glycine antagonist GV150526 to block glycine, as one of the two agonists (in addition to Glu itself) necessary for the activation of NMDA receptors. They found that GV150526 administration protected the SEPs and reduced the infarct area in the middle cerebral artery occlusion model of focal ischemia in the rat. On the above premises, this study was motivated by the expectation that the elimination of excess Glu after ischemia would have a positive impact on both the anatomical and functional correlates of ischemia. First of all, we predicted that the volume of the ischemic lesion would be decreased after OxAc administration, because of the increased brain-to-blood Glu efflux (Gottlieb et al. 2003). Indeed, the administration of the Glu scavenger OxAc at a dose shown in previous studies to cause both blood Glu scavenging (Gottlieb et al. 2003) and neuroprotection (Zlotnik et al. 2007) resulted in a reduction in the volume of the ischemia-induced cortical damage. This effect could be observed as early as 4 h after the intervention. This result is in accordance with the finding of Zlotnik et al. who found that the major neuroprotective effects of OxAc and pyruvate were readily observed within the first few hours (Zlotnik et al. 2007; Zlotnik et al. 2008). Moreover, we expected that the photothrombotic lesion would result in decreased amplitudes of the SEPs as these are early consequences of ischemia (Baik et al. 1990; Ladds et al. 1988; Liu et al. 1992). We also anticipated that OxAc would attenuate the decreases in the EP amplitudes.

The results we obtained clearly fulfil these expectations since the administration of the blood Glu scavenger OxAc after the photothrombotic lesion attenuated both the volume of the infarct and the decreases in the amplitudes of the EPs. Thus, as predicted from its blood Glu-scavenging activity and its ability to increase the efflux of excess Glu from the brain into the blood (Gottlieb et al. 2003), OxAc exerts a neuroprotective effect here, as it did after a traumatic brain injury (Zlotnik et al. 2007).

It could be argued that the neuroprotective effects of OxAc are not due to its blood Glu-scavenging activity, but rather due to a therapeutic activity exerted within the brain. OxAc, for instance, could contribute to an improvement in the NAD-linked mitochondrial energetics via an enhancement of the malate–aspartate shuttle. It could also, as the other ketoacids do, scavenge hydrogen peroxide, which is known to play a crucial role in the pathogenesis of ischemia–reperfusion injury (Traystman et al. 1991), or contribute to the transamination of kynurenine into kynurenic acid, a competitive broad-spectrum antagonist of all the ionotropic excitatory amino acid receptors (Hodgkins et al. 1999; Stone 1993). However, several very compelling arguments have been raised against a central action of OxAc, including the observation that the neuroprotective action of OxAc depends on the presence of GOT in the blood (Zlotnik et al. 2007), and they apply to this study as well. Moreover, hydrogen peroxide was recently reported to produce a neuroprotective effect in an in vitro model of brain ischemia (Nistico et al. 2008).

The neuroprotective effects of OxAc can readily be explained in light of the very specific events taking place upon ischemia. The decrease in the extent of tissue oxygenation causes the excess of Glu to be released into the interstitial fluid, which in turn produces a deep depolarization and an exaggerated influx of sodium ions, with resultant edema and marked neuronal swelling. The latter further contributes to the tissue hypoperfusion, while the diffusion of excess Glu in the perineuronal fluid puts the peri-infarct region at risk.

It can be suggested, therefore, that the beneficial effects of OxAc are due to its blood Glu-scavenging activity, which causes an enhanced brain-to-blood Glu efflux and a decrease in the level of excess Glu. This limits the size of the penumbra, reduces the edema, improves the tissue perfusion, increases the oxygenation level and reduces the ischemia-related damage.

In conclusion, the attenuation of the anatomical and electrophysiological correlates of the brain ischemia produced by OxAc are in line with the neurological improvements it causes in closed head injuries and provide additional evidence for the neuroprotective activity of OxAc.

Abbreviations

Glu

Glutamate

OxAc

Oxaloacetate

EP

Evoked potential

SEPs

Somatosensory evoked potentials

SSI

Somatosensory cortex

FJB

Fluoro-Jade B