Sage Journals: Discover world-class research

Abstract

Astrocytes are intimately involved in both glutamate and γ-aminobutric acid (GABA) synthesis, and ischemia-induced disruption of normal neuroastrocytic interactions may have important implications for neuronal survival. The effects of middle cerebral artery occlusion (MCAO) on neuronal and astrocytic intermediary metabolism were studied in rats 30, 60, 120, and 240 minutes after MCAO using in vivo injection of [1-¹³C]glucose and [1,2-¹³C]acetate combined with ex vivo¹³C magnetic resonance spectroscopy and high-performance liquid chromatography analysis of the ischemic core (lateral caudoputamen and lower parietal cortex) and penumbra (upper frontoparietal cortex). In the ischemic core, both neuronal and astrocytic metabolism were impaired from 30 minutes MCAO. There was a continuous loss of glutamate from glutamatergic neurons that was not replaced as neuronal glucose metabolism and use of astrocytic precursors gradually declined. In GABAergic neurons astrocytic precursors were not used in GABA synthesis at any time after MCAO, and neuronal glucose metabolism and GABA-shunt activity declined with time. No flux through the tricarboxylic acid cycle was found in GABAergic neurons at 240 minutes MCAO, indicating neuronal death. In the penumbra, the neurotransmitter pool of glutamate coming from astrocytic glutamine was preserved while neuronal metabolism progressively declined, implying that glutamine contributed significantly to glutamate excitotoxicity. In GABAergic neurons, astrocytic precursors were used to a limited extent during the initial 120 minutes, and tricarboxylic acid cycle activity was continued for 240 minutes. The present study showed the paradoxical role that astrocytes play in neuronal survival in ischemia, and changes in the use of astrocytic precursors appeared to contribute significantly to neuronal death, albeit through different mechanisms in glutamatergic and GABAergic neurons.

Keywords

Astrocytes GABA Glutamate Glutamine MCAO

Astrocytes are intimately involved in both glutamate and γ-aminobutyric acid (GABA) synthesis, and changes in trafficking of metabolites between astrocytes and glutamatergic or GABAergic neurons during ischemia in vivo may have important implications for neuronal survival. In a previous paper on middle cerebral artery occlusion (MCAO), we demonstrated that astrocytic and neuronal intermediary metabolism, and the trafficking of metabolites between these two cellular compartments, can be analyzed in great detail by combining in vivo injection of¹³C-labeled acetate and glucose with ex vivo¹³C magnetic resonance (MR) spectroscopy of deproteinized brain extracts (Håberg et al., 1998a). In the present study, this method was used to identify changes in glutamate and GABA synthesis, and in astrocytic metabolism during the first 4 hours MCAO. In cerebral ischemia, glutamate excitotoxicity is considered an important mediator of neuronal death (Choi, 1992; Meldrum, 2000); however, GABA is recognized as an endogenous neuroprotectant in the mature brain (Schwartz-Bloom and Sah, 2001).

The lateral caudoputamen and lower parietal cortex, and the upper frontoparietal cortex ipsilateral to the MCAO, were analyzed separately to gain insight into regionally specific alterations in the metabolic activity of glutamatergic and GABAergic neurons and astrocytes. In MCAO induced with the intraluminal filament technique, the ipsilateral lateral caudoputamen and lower parietal cortex are rendered severely ischemic, whereas the upper frontoparietal cortex represents moderately ischemic tissue. The former is often referred to as the ischemic core, and the latter as the penumbra (Memezawa et al., 1992a; Back, 1998). Inherent to the definition of the penumbra is the possibility of reversing the neuronal failure, but with time the penumbra is incorporated into the ischemic core (Back, 1998). In the present study, neuronal and astrocytic tricarboxylic acid (TCA) cycle activity and the use of astrocytic intermediates for neurotransmitter synthesis in these two regions were followed closely during the initial 240 minutes MCAO to identify changes influencing neuronal survival.

MATERIALS AND METHOD

Animal experiments

All animal procedures adhered to institutionally approved protocol in accordance with the guidelines set by the Norwegian Committee for Animal Research.

Male Wistar rats (Møllegaard Breeding Center, Copenhagen, Denmark) weighing 320 to 340 g were fasted overnight. Anesthesia was induced with 3.5% isoflurane in 70% nitrogen/30% oxygen. During surgery, the animals spontaneously breathed 2% isoflurane in 70% nitrogen/30% oxygen delivered through a face mask. Body temperature was maintained using a feedback-controlled heating blanket connected to a rectal temperature probe. A catheter was introduced into the right femoral vein, externalized at the tail radix, and taped in place. Middle cerebral artery occlusion was induced with the intraluminal filament technique (Longa et al., 1989; Memezawa et al., 1992a). All incisions were sprayed with lidocaine (10 mg/doses) before closure. The rats subjected to MCAO were assigned to four different groups with MCAOs of 30 (n = 7), 60 (n = 7), 120 (n = 7), or 240 minutes (n = 7). Ten rats were sham operated and allowed to recover for either 30 (n = 3) or 240 minutes (n = 7). Isoflurane administration was discontinued at time of MCAO except in rats assigned to MCAO for 120 and 240 minutes, who underwent diffusion-weighted MR imaging immediately after surgery. The imaging was performed 18 minutes after MCAO on a 2.35-T Bruker Biospec (Bruker AG, Fällanden, Switzerland) with a single b value of 1468 mm²/second, echo time 32/repetition time 1500, and number of excitations two-in-eighth transaxial slices with 1.5-mm thick slices that covered the entire hemisphere. Isoflurane was maintained at 2% during MR imaging. To avoid unwanted effects of anesthesia on the cerebral metabolism, rats assigned to either 30 or 60 minutes MCAO did not undergo diffusion-weighted imaging. The rats were returned to individual cages and allowed to recover immediately after surgery or MR imaging. An intravenous injection (1 mL/100 g) 0.3 mmol/L [1-¹³C]glucose and [1,2-¹³C]acetate (Cambridge Isotopes Laboratories, Woburn, MA, U.S.A.) in sterile water was given for 2 minutes starting 15 minutes before decapitation. The rats were decapitated into liquid nitrogen 30, 60, 120, or 240 minutes after MCAO or sham operation. Blood was collected from the severed neck vessels of rats subjected to 240 minutes MCAO. A 5-mm coronal slice extending caudally from the chiasma opticum was cut from the frozen brains using a brain matrix (RBM-40000; Activational Systems, Warren, MI, U.S.A.), and the right lateral caudoputamen, lower parietal cortex, and upper frontoparietal cortex were sampled. The brain and plasma samples were extracted as described previously (Håberg et al., 1998a).

¹³C magnetic resonance spectroscopy

Proton-decoupled 125.5-MHz¹³C MR spectra were obtained with a Bruker DRX-500 spectrometer (Bruker AG) using a 35 pulse angel, 25-kHz spectral width, and 64-K data points. The acquisition time was 1.3 seconds/scan plus 2.5 seconds of relaxation delay. The number of scans was typically 15,000 for brain samples. To avoid nuclear Overhauser effects, some spectra were broadband decoupled during acquisition only. The respective correction factors were then applied to the spectra obtained.

¹H magnetic resonance spectroscopy

The 500-MHz¹H MR spectra of the plasma extracts were obtained with a Bruker DRX-500 spectrometer using a 90 pulse angle, 7.5-kHz spectral width, and 16-K data points. The acquisition time was 1.1 second/scan plus 15.0 seconds of relaxation delay. The number of scans was typically 200.

High-performance liquid chromatography analysis

After¹³C MR spectroscopy, total amino acid concentrations were determined with high-performance liquid chromatography (Specta System Gradient Pump, Freemont, CA, U.S.A.), and fluorescence detection (Shimadzu RF 530, Tokyo, Japan) after derivatization with (o-phthaldialdehyde (Sigma Chemical Co., St. Louis, MO, U.S.A.) using L-aminobutyric acid as an internal standard.

Analysis of¹³C and¹H magnetic resonance spectra and the rationale behind data interpretation

Relevant peaks in the¹³C and¹H MR spectra were identified, and the total amount of¹³C or¹H in the resonance of a particular metabolite was quantified from the integral of the peak area using ethyleneglycol as an internal standard. The total amount of¹³C includes the naturally abundant¹³C (1.1% of all carbon atoms). Data analysis is described in detail in Håberg et al. (1998a).

The neuronal and astrocytic compartments can be analyzed simultaneously by concomitant administration of [1-¹³C]glucose and [1,2-¹³C]acetate (Badar-Goffer et al., 1990; McLean et al, 1993; Sonnewald et al., 1996; Hålberg et al., 1998a) because of the uneven distribution of enzymes and transporter proteins between neurons and astrocytes.

The astrocytic compartment is recognized by its ability to use acetate as a substrate for the TCA cycle (Van den Berg, 1973) due to the presence of a specific acetate-uptake mechanism (Waniewski and Martin, 1998). In astrocytes, [1,2-¹³C]acetate is metabolized by acetyl coenzyme A synthetase (EC 6.2.1.1) to acetyl coenzyme A, which enters the TCA cycle and finally gives rise to [4,5-¹³C]glutamate. [4,5-¹³C]Glutamate is rapidly converted to [4,5-¹³C]glutamine by the glia-specific enzyme glutamine synthetase (EC 6.3.1.2) (Norenberg and Martinez-Hernandez, 1979). Glutamate is present only in low concentrations in astrocytes (Ottersen, 1989). [4,5-¹³C]Glutamine is released from astrocytes and taken up by a high-affinity glutamine transporter that is present on neurons (Varoqui et al., 2000). Glutamate is regenerated from glutamine in the neurons by phosphate-activated glutaminase (PAG) (EC 3.4.1.2) (Kvamme et al., 2000), which converts [4,5-¹³C]glutamine to [4,5-¹³C]glutamate. In GABAergic neurons, GABA is synthesized from glutamate by glutamate decarboxylase (GAD) (EC 4.1.1.15), and the direct conversion of [4,5-¹³C]glutamate results in [1,2-¹³C]GABA. GABA synthesis only takes place in neurons (Sze, 1979; Sloviter et al., 1996). Label in [1,2-¹³C]GABA is distributed equally into [3-¹³C] and [4-¹³C]GABA in the second turn of the TCA cycle due to the symmetrical succinate step. Likewise, label in [4,5-¹³C]glutamate/glutamine is distributed into [1,2-¹³C]- and [3-¹³C]glutamate/glutamine.

Astrocytes can also metabolize glucose, but to a lesser extent than neurons. However, glucose can enter the TCA cycle of astrocyte cells via the astrocyte-specific anaplerotic enzyme pyruvate carboxylase (PC) (EC 6.4.1.1) (Yu et al., 1983), the brain's principal anaplerotic enzyme (Patel, 1974). In astrocytes, [1-¹³C]glucose metabolized by PC activity gives rise to [2-¹³C]glutamine. [2-¹³C]Glutamine is converted to [2-¹³C]glutamate in neurons (or to [4-¹³C]GABA in GABAergic neurons).

Neurons primarily metabolize glucose (Van den Berg, 1973; Künnecke et al., 1993). [1-¹³C]Glucose enters the neuronal TCA cycle as acetyl coenzyme A solely via pyruvate dehydrogenase (PDH) (EC 1.2.4.1), and finally results in [4-¹³C]glutamate formation. However, glutamate de novo synthesis occurs in astrocytes and not in neurons (Hertz et al., 1999). The majority of glutamate is found in glutamatergic neurons (Ottersen, 1989). Synaptically released glutamate is predominantly taken up by high-affinity glutamate transporters located on astrocytes (Erecinska, 1987; Tanaka, 2000). In the astrocyte, [4-¹³C]glutamate either is converted directly to [4-¹³C]glutamine, or is reintroduced to the TCA cycle (Farinelli and Nicklas, 1992). The exchange of glutamate and glutamine between astrocytes and neurons is termed the glutamate-glutamine cycle (Westergaard et al., 1995; Daikhin and Yudkoff, 2000). After one turn in the astrocytic TCA cycle, the [4-¹³C]glutamate label will be recovered in [3-¹³C]- and [2-¹³C]glutamine. In GABAergic neurons, [4-¹³C]glutamate is rapidly converted to [2-¹³C]GABA, and little glutamate is present in these neurons (Ottersen, 1989). After one turn of the TCA cycle, the label that originated in [2-¹³C]GABA will be distributed equally to [3-¹³C]- and [4-¹³C]GABA.

Neurons depend on the transfer of metabolites from astrocytes because of their inability to synthesize glutamate de novo and their lack of anaplerotic activity, combined with the continuous drain of metabolic intermediates from the neuronal TCA cycle due to neurotransmitter production and release (Hertz, 1979; Kaufmann and Driscoll, 1993; Hertz et al., 1999). The contribution from astrocytic precursors to neuronal glutamate and GABA formation can be derived from the PC versus PDH activity ratio, which is an estimate of the anaplerotic (astrocytic) pathway compared with the oxidative (neuronal) pathway in glutamate (Taylor et al., 1996). Furthermore, the acetate versus glucose utilization ratio approximates the relative contribution of the astrocytic TCA cycle precursors compared with the neuronal TCA cycle precursors glutamate, glutamine, and GABA formation (Taylor et al., 1996).

All values are mean ± SD. Statistical comparisons between the groups subjected to MCAO of varying duration and the sham-operated rats who recovered for 240 minutes before decapitation were performed with analysis of variance followed by least-significant difference post-hoc test for multiple comparisons. An unpaired two-way Student's t-test was used to compare the sham-operated rats recovering for 30 and 240 minutes after termination of anesthesia. P < 0.05 was considered significant.

RESULTS

Diffusion-weighted magnetic resonance imaging

The total volume of ischemic tissue 18 minutes after MCAO, as estimated by diffusion-weighted imaging, was 125.5 ± 52.15 mm³(n = 14). The volume of ischemia for each separate slice, from caudal to rostral, was 2.2 ± 4.8, 18.9 ± 12.1, 31.1 ± 16.2, 38.5 ± 10.5, 22.2 ± 10.9, 8.1 ± 10.0, 8.3 ± 2.1, and 1.3 ± 4.2 mm³. The 5-mm slice cut after decapitation to sample brain tissue for further analysis with¹³C MR spectroscopy was located between slice three and five, the region in which the largest areas of ischemic tissue were found.

Analysis of plasma extracts

In rats subjected to 240 minutes MCAO, the total glucose concentration in plasma as estimated from the¹H MR spectra was 16.1 ± 1.1 mmol/L, and only glucose enriched in the C-1 position was detected in significant amounts. The concentrations of [1,2-¹³C]acetate, lactate, and hydroxybutyrate were 1.7 ± 0.8, 2.1 ± 0.5, and 1.2 ± 0.4 mmol/L, respectively. There was no significant¹³C enrichment in lactate or hydroxybutyrate; therefore, the observed amino acid labeling originated from cerebral [1-¹³C]glucose and [1,2-¹³C]acetate metabolism. Furthermore, neither the duration of fasting nor the stress of cerebral ischemia influenced the plasma composition, because similar amounts of¹³C-labeled substrate were present in plasma at 90 minutes MCAO (Håberg et al., 1998a).

Effects of middle cerebral artery occlusion on total amino acid content

Lateral caudoputamen and lower parietal cortex.

In this severely ischemic region, the total glutamate content was significantly decreased 30 minutes after MCAO, and continued to decline for the next 240 minutes (Fig. 1A). The glutamine content remained at control levels until 240 minutes when it was significantly decreased compared with 60 minutes MCAO (Fig. 1A). The total GABA content was more than doubled at 240 minutes MCAO (Fig. 1C). The alanine levels started to increase markedly from 60 minutes MCAO, and had doubled by 240 minutes (Fig. 1E). The total amount of aspartate decreased from 60 minutes MCAO, and was approximately half of that found in sham-operated rats after 240 minutes MCAO (Fig. 1E).

FIG. 1.

Effect of MCAO on total amino acid content. Time zero represents sham-operated rats allowed to recover for 240 minutes before time of death. The amino acid content was analyzed with high-performance liquid chromatography (see Materials and Methods). The lateral caudoputamen and lower parietal cortex (A, C, E) represent severely ischemic tissue. The upper frontoparietal cortex (B, D, F) represents moderately ischemic tissue. All values are mean ± SD; n = 7 in each group. Statistical comparisons were performed with analysis of variance followed by post-hoc test for multiple comparisons.^aSignificantly different from sham operated, P < 0.05,^aaP < 0.01,^aaaP < 0.001.^bSignificantly different from group subjected to 30 minutes MCAO, P < 0.05,^bbP < 0.01,^bbbP < 0.001.^cSignificantly different from group subjected to 60 minutes MCAO, P < 0.05,^cccP < 0.001.^dSignificantly different from group subjected to 120 minutes MCAO, P < 0.05,^dddP < 0.001.

Upper frontoparietal cortex.

In this moderately ischemic region, glutamate remained at control levels during the 240 minutes MCAO (Fig. 1B). The glutamine content was significantly increased at 60 minutes MCAO and nearly doubled at 240 minutes (Fig. 1B). The GABA content was also significantly increased at and beyond 60 minutes MCAO (Fig. 1D). Alanine levels were not markedly increased before 120 minutes (Fig. 1F), and aspartate concentration was at control levels during 240 minutes MCAO (Fig. 1F).

¹³C magnetic resonance spectroscopy of brain extracts: effects of middle cerebral artery occlusion on cerebral metabolism

[1-¹³C]Glucose and [1,2-¹³C]acetate.

In the lateral caudoputamen and lower parietal cortex, [1-¹³C]glucose content was significantly increased to a similar degree at all time points after MCAO (Fig. 2A). Unmetabolized [1,2-¹³C]acetate was present at all times after MCAO (Fig. 2A).

FIG. 2.

The total amount of [1-¹³C]glucose and [1,2-¹³C]acetate in ischemic tissue after MCAO. Time zero represents sham-operated rats allowed to recover for 240 minutes before time of death. The total amount of¹³C (10⁻⁸mol/g tissue) was analyzed with¹³C MR spectroscopy (see Materials and Methods), and was not corrected for naturally abundant¹³C (1.1%). The lateral caudoputamen and lower parietal cortex (A) represent severely ischemic tissue. The upper frontoparietal cortex (B) represents moderately ischemic tissue. All values are mean ± SD; n = 7 in each group. Statistical comparisons were performed with analysis of variance followed by the post-hoc test for multiple comparisons.^aSignificantly different from sham operated rats, P < 0.05,^aaP < 0.01,^aaaP < 0.001.^bSignificantly different from group subjected to 30 minutes MCAO, P < 0.05,^bbP < 0.01,^bbbP < 0.001.^cSignificantly different from group subjected to 60 minutes MCAO, P < 0.05,^ccP < 0.01.^dSignificantly different from group subjected to 120 minutes MCAO, P < 0.05,^ddP < 0.01.

In the frontoparietal cortex, [1-¹³C]glucose content increased steadily and significantly from 60 to 240 minutes MCAO (Fig. 2B). [1,2-¹³C]Acetate was present at 30 and 120 minutes MCAO, but not at 60 minutes. At 240 minutes, the [1,2-¹³C]acetate content was significantly increased compared with all previous time points (Fig. 2B).

Glutamate and glutamine.

In the lateral caudoputamen and lower parietal cortex, the total amount of [4-¹³C]glutamate and [4-¹³C]glutamine decreased rapidly during the first 60 minutes MCAO, then leveled out before decreasing 240 minutes after MCAO (Fig. 3A). This decrease was most notable in [4-¹³C]glutamate. The total amount of label in [4,5-¹³C]glutamate was also significantly reduced from 30 minutes MCAO (Fig. 3C), but the use of acetate compared with glucose as precursor for glutamate was markedly increased throughout the 240 minutes MCAO (Table 1). In glutamine, the total amount of label in [4,5-¹³C]glutamine was slightly reduced during the initial 120 minutes, then declined steeply at 240 minutes MCAO (Fig. 3C). The acetate-glucose utilization ratio was significantly higher at 120 minutes MCAO, but returned to control levels at 240 minutes (Table 1). In [3-¹³C]glutamate, the total amount of¹³C decreased markedly to the same range as the total amount of label in [3-¹³C]glutamine from 60 minutes MCAO (Fig. 3E). Pyruvate carboxylase activity in glutamate was present at all times after MCAO, but was significantly reduced from 120 minutes MCAO (Table 1). In glutamine, the PC/PDH activity ratio was markedly increased at 60 minutes MCAO, but at 240 minutes MCAO no PC activity was detectable (Table 1).

FIG. 3.

Effect of MCAO on incorporation of label from [1-¹³C]glucose and [1,2-¹³C]acetate in amino acids. Time zero represents sham-operated rats allowed to recover for 240 minutes before time of death. The total amount of¹³C (10⁻⁸mol/g tissue) was analyzed with¹³C MR spectroscopy (see Materials and Methods), and was not corrected for naturally abundant¹³C (1.1%). All values are mean ± SD; n = 7 in each group. The lateral caudoputamen and lower parietal cortex (A,C, E,G, I) represent severely ischemic tissue. The upper frontoparietal cortex (B,D,F,H, J) represents moderately ischemic tissue. Statistical comparisons were performed with analysis of variance followed by post-hoc test for multiple comparisons.^aSignificantly different from sham operated, P < 0.05,^aaP < 0.01,^aaaP < 0.001.^bSignificantly different from group subjected to 30 minutes MCAO, P < 0.05,^bbP < 0.01,^bbbP < 0.001.^cSignificantly different from group subjected to 60 minutes MCAO, P < 0.05,^cccP < 0.001.^dSignificantly different from group subjected to 120 minutes MCAO, P < 0.05,^ddP < 0.01,^dddP < 0.001.

TABLE 1.

Effects of MCAO on the use of precursors derived from astrocytic metabolism compared with neuronal metabolism in glutamate, glutamine, and GABA formation

	Lateral caudoputamen and lower parietal cortex			Upper frontoparietal cortex
	Glutamate	Glutamine	GABA	Glutamate	Glutamine	GABA
Acetate/glucose utilization ratio
Sham operated	0.20 ± 0.01	1.30 ± 0.15	0.20 ± 0.04	0.24 ± 0.02	1.34 ± 0.13	0.21 ± 0.02
30 min of MCAO	0.26 ± 0.04^*	1.79 ± 0.28^*	NQ	0.26 ± 0.03	2.22 ± 0.09^†	NQ
60 min of MCAO	0.27 ± 0.08^*	2.30 ± 0.52^‡^§	NQ	0.32 ± 0.04^†^§	3.71 ± 0.53^‡^¶	NQ
120 min of MCAO	0.34 ± 0.04^‡^§^#	2.17 ± 0.42^‡	NQ	0.30 ± 0.06^*	2.77 ± 0.60^‡^#	NQ
240 min of MCAO	0.31 ± 0.11^†	1.75 ± 0.51	NQ	0.43 ± 0.07^‡^¶^**^§§	3.27 ± 0.78^‡^‖	NQ
PC/PDH activity ratio
Sham operated	0.20 ± 0.02	0.45 ± 0.09	0.06 ± 0.03	0.17 ± 0.05	0.76 ± 0.07	0.06 ± 0.02
30 min of MCAO	0.16 ± 0.06	0.36 ± 0.15	ND	0.10 ± 0.02	0.73 ± 0.18	0.07 ± 0.03
60 min of MCAO	0.18 ± 0.05	0.76 ± 0.08^‡^¶	ND	0.11 ± 0.08	0.68 ± 0.46^‡^¶	0.06 ± 0.05
120 min of MCAO	0.11 ± 0.05^†^#	0.37 ± 0.18^††	ND	0.03 ± 0.01^†	0.77 ± 0.13^†^†	0.04 ± 0.02
240 min of MCAO	0.07 ± 0.04^‡^§^**	ND	ND	0.05 ± 0.02^*	0.20 ± 0.13^†^‖^#^‡‡	ND

All values are mean ± SD. Ratios for PC/PDH activity and acetate/glucose utilization were calculated as described in Materials and Methods. Statistical comparisons between groups were performed with ANOVA followed by post hoc test for multiple comparisons.

Significantly different from sham operated, P < 0.05

†

P < 0.01,

‡

P < 0.001.

Significantly different from group subjected to 30 minutes of MCAO, P < 0.05

‖

P < 0.01

P < 0.001

Significantly different from group subjected to 60 min of MCAO, P < 0.05

P < 0.01,

††

P < 0.001.

‡‡

Significantly different from group subjected to 120 min of MCAO,P < 0.01

§§

P < 0.001.

ND, not detectable; NQ, not quantifiable because [1,2-¹³C]GABA was not detected.

The total amount of label in [4-¹³C]glutamate in the upper frontoparietal cortex was significantly reduced to about the same level the initial 120 minutes MCAO, and then declined further at 240 minutes MCAO (Fig. 3B). In [4-¹³C]glutamine the total amount of¹³C was markedly reduced to the same extent at all times after MCAO (Fig. 3B). In [4,5-¹³C]glutamate the total amount of label was reduced to the same level from 30 to 240 minutes MCAO (Fig. 3D). However, from 60 minutes MCAO acetate became increasingly important as precursor in glutamate formation indicated by the significantly elevated acetate/glucose utilization ratio in glutamate which was almost doubled at 240 minutes (Tbl. 1). There was no decline in the total amount of¹³C in [4,5-¹³C]glutamine for 240 minutes MCAO (Fig. 3D), and the acetate/glucose utilization ratio was elevated from 30 minutes MCAO and continued to increase with time (Table 1). The total amount of¹³C in [3-¹³C]glutamate declined rapidly and was in the same range as the total amount of¹³C in [3-¹³C]glutamine from 60 minutes MCAO (Fig. 3F). In glutamate the PC/PDH activity ratio was significantly lower from 120 minutes MCAO, whereas in glutamine the PC/PDH activity was markedly lowered first at 60 minutes and then at 240 minutes MCAO (Table 1).

γ-Aminobutyric acid.

In the lateral caudoputamen and lower parietal cortex, the total amount of label in [2-¹³C]GABA was significantly reduced at all times after MCAO (Fig. 3G). In [3-¹³C]GABA, the total amount of¹³C decreased significantly after MCAO, but between 120 and 240 minutes MCAO the amount of label in [3-¹³C]- and [4-¹³C]GABA (data not shown) increased markedly (Fig. 3G). At 240 minutes MCAO, the total amount of¹³C was similar in [2-¹³C]-, [3-¹³C]- and [4-¹³C]GABA (Fig. 3G). This finding implied a cessation of GABAergic TCA cycle activity, because the amount of label in [2-¹³C]GABA must exceed the amount of label found in the subsequent turns using the present infusion scheme. The stop in GABAergic TCA-cycle activity was verified by the lack of incorporation of¹³C above the natural abundance level of 1.1% in the GABA isotopomers at 240 minutes MCAO. [1,2-¹³C]GABA was not detected at any time after MCAO, indicating that there was no direct use of [4,5-¹³C]glutamine in GABA synthesis after MCAO (Fig. 3G). Furthermore, there was no detectable PC activity in GABA (Table 1).

In the upper frontoparietal cortex, the total amount of label in [2-¹³C]GABA decreased rapidly during the initial 60 minutes MCAO and then leveled out during the remainder of the experimental period (Fig. 3H). The total amount of label in [3-¹³C]- and [4-¹³C]GABA (data nor shown) was significantly decreased in this region at all times after MCAO, but not as markedly as for [2-¹³C]GABA (Fig. 3H). [1,2-¹³C]GABA was not detected at any time after MCAO, but PC activity was present from 30 to 120 minutes MCAO (Table 1).

Aspartate and alanine.

In the lateral caudoputamen and lower parietal cortex, the total amount of¹³C in [3-¹³C]aspartate was approximately halved at 30 minutes MCAO and continued to decline for 240 minutes (Fig. 3I). However, the total amount of¹³C in [3-¹³C]alanine was significantly increased at 60 and 120 minutes MCAO, but at 240 minutes had decreased to a level similar to that in sham-operated rats (Fig. 3I).

The total amount of amount of¹³C in [3-¹³C]aspartate in the upper frontoparietal cortex was significantly reduced to the same level at 30 to 120 minutes MCAO, followed by a further marked reduction at 240 minutes MCAO (Fig. 3J). In [3-¹³C]alanine, the total amount of label increased steadily and significantly throughout the 240 minutes MCAO (Fig. 3J).

Effects of isoflurane on cerebral metabolism

Lateral caudoputamen and lower parietal cortex.

In sham-operated rats recovering for 30 minutes after termination of anesthesia, the total amount of label in [3-¹³C]aspartate, [2-¹³C]GABA, [4-¹³C]glutamate, and [4-¹³C]glutamine was significantly lower compared with sham-operated rats recovering for 240 minutes (Table 2). Moreover, [1,2-¹³C]acetate (4.1 ± 1.7 10⁻⁸mol/g) was detected in all samples from the 30 minutes sham group, but was not found in the 240 minutes group.

TABLE 2.

Delayed effects of isoflurane on cerebral metabolism in sham-operated rats

	Lateral caudoputamen and lower parietal cortex		Upper frontoparietal cortex
Amino acid isotopomer	Sham 30 min (n = 3)	Sham 240 min (n = 7)	Sham 30 min (n = 3)	Sham 240 min (n = 7)
[3-¹³C]Alanine	7.1 ± 1.4	7.3 ± 1.6	7.0 ± 0.5	6.0 ± 0.8
[3-¹³C]Aspartate	7.3 ± 0.9	12.1 ± 1.3^‡	12.2 ± 4.5	12.3 ± 1.5
[2-¹³C]GABA	14.1 ± 2.0	18.6 ± 2.0^*	11.0 ± 0.4	14.8 ± 1.9^†
[1,2-¹³C]GABA	3.7 ± 0.3	3.5 ± 0.5	3.2 ± 0.5	3.5 ± 0.8
[4-¹³C]Glutamine	15.9 ± 2.3	24.6 ± 1.9^‡	20.7 ± 4.8	25.3 ± 3.9
[4,5-¹³C]Glutamine	30.4 ± 0.2	31.7 ± 3.8	33.9 ± 7.8	32.2 ± 3.5
[4-¹³C]Glutamate	67.6 ± 8.4	100.9 ± 5.9^‡	84.7 ± 3.0	108.9 ± 8.1^†
[4,5-¹³C]Glutamate	18.0 ± 1.6	19.1 ± 0.6	23.8 ± 0.9	28.4 ± 5.4

The sham-operated rats were allowed to recover for 30 or 240 minutes after termination of anesthesia before sacrifice. All values are total amount of ¹³C (10⁻⁸ mol/g tissue) and was calculated with ¹³C MRS as described in Materials and Methods. The total amount of ¹³C was not corrected for naturally abundant ¹³C (1.1%). The values are given as mean ± SD. Two-way unpaired Student's t-test was used to determined statistical differences.

Significantly different from sham-operated rats allowed to recover for 30 minutes, P < 0.05

†

P < 0.01

‡

P < 0.001

Upper frontoparietal cortex.

In the sham-operated group recovering for 30 minutes after anesthesia, the total amount of label was significantly decreased in [2-¹³C]GABA and [4-¹³C]glutamate compared with 240 minutes sham group (Table 2). [1,2-¹³C]Acetate (4.5 ± 1.7 10⁻⁸mol/g) was present in the former group but not in the latter sham group.

These results clearly show that isoflurane exerted prolonged effects on neuronal and, to some extent, astrocytic metabolism. Thirty minutes after termination of isoflurane administration, the anesthetic agent still affected incorporation of the label from [1-¹³C]glucose into some amino acids via PDH. Previously reduced incorporation of the label into amino acids from glucose but not from acetate has been shown during pentobarbital anesthesia (Cremer and Lucas, 1971; Shank et al., 1993). The reduction in glucose metabolism was more prominent in the lateral caudoputamen and lower parietal cortex than in the upper frontoparietal cortex; however, isoflurane affects subcortical structures more than cortical structures (Hansen et al., 1988; Hoffman et al., 1991). Furthermore, the marked reduction of label in [4-¹³C]glutamine in this region, which is mostly derived from [4-¹³C]glutamate, implied that isoflurane specifically reduced neuronal glutamate release there. These results show that isoflurane predominantly affects neuronal metabolism and activity. Agreeing with that, astrocytic metabolism was also reduced, as shown by the presence of unmetabolized [1,2-¹³C]acetate in the 30-minute sham group but not in the 240-minute group. Therefore, the presence of unmetabolized [1,2-¹³C]acetate in the upper frontoparietal cortex 30 minutes after MCAO was most likely an isoflurane artifact because [1,2-¹³C]acetate was not detected at 60 minutes MCAO in the same region. Although isoflurane reduced cerebral metabolism 30 minutes after it had been turned off, it was evident only 30 minutes after induction that MCAO exerted its own distinct effects on the cerebral metabolism, as illustrated by the significant reduction in label incorporation from [1,2-¹³C]acetate into amino acids.

DISCUSSION

The present study clearly shows that MCAO rapidly induces profound changes in neuronal and astrocytic metabolism, and in the trafficking of metabolites between these two cellular compartments. The changes depended both on the duration of ischemia and on the degree of ischemia. There were both similarities and distinct differences in the metabolic response to MCAO in severely ischemic tissue in the lateral caudoputamen and lower parietal cortex, and in moderately ischemic tissue in the upper frontoparietal cortex.

Label incorporation into amino acids from [1-¹³C]glucose via PDH and the TCA cycle was depressed almost to the same extent in the two ischemic regions, showing the high sensitivity of neuronal metabolism to reduction of CBF. However, decreased neuronal glucose metabolism in MCAO does not indicate irreversible neuronal damage, because reperfusion within 30 minutes salvages substantial tissue volumes from infarction in the lateral caudoputamen, and the upper frontoparietal cortex can be saved with reperfusion within 90 to 120 minutes (Memezawa et al., 1992b; Müller et al., 1995; Garcia et al., 1995). Thus, the development of metabolic insufficiency in neurons appeared to arise from a combination of metabolic disturbances, including alterations in the use of astrocytic precursor.

The most obvious difference between the severely and moderately ischemic tissue was found in the acetate metabolism. Reduced astrocytic metabolism, as indicated by reduced formation of [4,5-¹³C]glutamine and/or the presence of unmetabolized [1,2-¹³C]acetate, was detected at 30 minutes in severely ischemic tissue and at 120 minutes MCAO in moderately ischemic tissue. These time points coincide with the time-windows for successful reperfusion of the lateral caudoputamen and lower parietal cortex and of the upper frontoparietal cortex, respectively, suggesting that reduced metabolism in astrocytes may be an indicator of irreversible tissue injury.

MCAO induced changes in the glutamate-glutamine cycle

Lateral caudoputamen and lower parietal cortex.

Profound changes in the glutamate metabolism were revealed in this severely ischemic region. The normal equilibrium between glutamate and glutamine content and, thus, the glutamate-glutamine cycle were disturbed from 30 minutes MCAO. There was a rapid decline in glutamate synthesis from glucose-derived acetyl coenzyme A, representing neuronal-derived glutamate synthesis. At the same time, a reduction was seen in use of glutamine in glutamate formation. This reduction may result from reduced activity in astrocytic or neuronal glutamine transporters in ischemia and/or from reduced PAG activity, because both ammonia accumulation and acidosis, which are present in ischemia, inhibit PAG (Benjamin, 1981; Kvamme et al., 1982; Hogstad et al., 1988). However, astrocytic glutamine was a more important precursor for glutamate than precursors derived from neuronal TCA cycle activity at all times after MCAO, suggesting that astrocytic glutamine contributed significantly to excitotoxicity in vivo. This observation concurs with previous studies using cell cultures (Goldberg et al., 1988; Rosenberg, 1991; Huang and Hertz, 1994). The reduced glutamate synthesis present from 30 minutes MCAO cannot fully explain the reduced glutamate content detected in the same region. The decreased glutamate levels were most likely connected to increased glutamate use, including oxidation and/or loss of glutamate into systemic circulation and cerebrospinal fluid during MCAO. Astrocytic glutamate uptake and subsequent conversion to glutamine was shown by the better preservation of label in [4-¹³C]glutamine than in [4-¹³C]glutamate throughout the 240 minutes MCAO, which is in agreement with earlier studies (Aas et al., 1993; Takagi et al., 1993; Swanson, 1992; Torp et al., 1993; Ottersen et al., 1996). Increased entry of glutamate into the TCA cycle in astrocytes was suggested by the better preservation of the total amount of label in [3-¹³C]glutamine than in [3-¹³C]glutamate at 60 minutes MCAO and beyond. Previous in vitro and in vivo data have shown increased glutamate metabolism via the astrocytic TCA cycle when extracellular glutamate concentrations were increased and during ischemia (McKenna et al., 1996; Håberg et al., 1998a; Pascual et al., 1998). Glutamate may also be consumed in GABA and glutamine synthesis, and increased alanine synthesis may reduce glutamate content by transamination. In addition, glutamate may be lost into the systemic circulation and cerebrospinal fluid. Elevated glutamate levels have been demonstrated in the cerebrospinal fluid of rats subjected to focal ischemia (Cataltepe et al., 1996), and in both plasma and cerebrospinal fluid of patients with acute ischemic stroke (Castillo et al., 1996). Loss of glutamate from the brain parenchyma gives an indication of the mismatch between glutamate synthesis and consumption and of glutamate release and uptake during MCAO, illustrating the serious metabolic insufficiency in this region. Despite this continuous loss of metabolites, glutamatergic neurons maintained a low level of glutamate synthesis throughout the 240 minutes MCAO.

Upper frontoparietal cortex.

Glutamate metabolism was also affected in moderately ischemic tissue, but the changes in the glutamate-glutamine cycle were somewhat distinctive in this region. As in the severely ischemic tissue, glutamate synthesis was reduced from both neuronal and astrocytic precursors at 30 minutes MCAO and beyond, but in moderately ischemic tissue the contribution from astrocytic precursors compared with neuronal precursors did not increase significantly before 60 minutes MCAO. That neuronal glutamine use in glutamate formation was reduced despite the presence of normal [4,5-¹³C]glutamine levels, and that the total amount of [4,5-¹³C]glutamate was at the same level at all times after MCAO, indicates some functional inhibition of either glutamine transporters and/or PAG activity independent of the duration of ischemia. For example, PAG activity may be reduced by the presence of normal glutamate concentrations (Hogstad et al., 1988). From 60 minutes MCAO, glutamine became increasingly important in glutamate synthesis while the neuronal TCA cycle activity gradually declined. Glutamine principally replenishes the neurotransmitter pool of glutamate (Laake et al., 1995), the pool from which glutamate release is considered to originate in moderately ischemic tissue (Takagi et al., 1993; Obrenovitch, 1996). The constellation of gradually declining neuronal metabolic activity and continued replenishment of the neurotransmitter pool of glutamate renders neurons extremely susceptible to glutamate excitotoxicity. There is increasing evidence for a delayed component in glutamate-mediated neuronal damage in cerebral ischemia, and based on the present data it appears to be connected to the continued astrocytic glutamine production and the subsequent neuronal use of glutamine in glutamate neurotransmitter synthesis. The current finding of persistent glutamate synthesis from glutamine and, to some extent, from the neuronal TCA cycle precursors for 240 minutes MCAO implies that administration of glutamate receptor antagonists has to be continued for a prolonged period. In line with this observation is the finding that AMPA (α-amino-3-hydroxy-5-methyl-4-isoxasole propionate) receptor antagonist administration lasting 4 hours after MCAO does not reduce final infarct volume, whereas administration lasting 24 hours salvages the upper frontoparietal cortex (Håberg et al., 1998b). Furthermore, inhibiting glutamine synthetase activity and thereby reducing glutamine synthesis has been shown to reduce infarct size in the cortex of rats subjected to MCAO (Swanson et al., 1990). Finally, because increased extracellular glutamine concentrations stimulate neuronal glutamate release (Szerb and O'Regan, 1985), the increased glutamine concentrations detected from 60 minutes MCAO may have deleterious effects.

Results from the sham-operated rats showed that astrocytic glutamine was a more important precursor for glutamate in the upper frontoparietal cortex than in the lateral caudoputamen and lower parietal cortex. This finding is in agreement with the theory that PAG density and activity is higher in cortex than in caudate-putamen/striatum (Aoki et al., 1991; Wallace and Dawson, 1993).

MCAO-induced changes in GABA metabolism

Middle cerebral artery occlusion completely changed GABA metabolism in severely ischemic tissue in the lateral caudoputamen and lower parietal cortex, and also in moderately ischemic tissue in the upper frontoparietal cortex. [1,2-¹³C]GABA was not detectable in either region showing a dramatic reduction in the conversion of astrocytic glutamine to GABA after MCAO. Furthermore, the PC activity was undetectable at all times after MCAO in severely ischemic tissue, suggesting the complete cessation of astrocytic precursor use for GABA formation. In moderately ischemic tissue, PC activity was present in GABA from 30 to 120 minutes MCAO, indicating that astrocytic precursors participated to some extent in GABA synthesis. This reduction in glutamine use in GABA formation was at great variance from the increased use of glutamine in glutamate synthesis during MCAO. However, as in glutamatergic neurons, GABA synthesis from neuronal precursors was markedly lowered after MCAO.

The build up of GABA, which can proceed in the absence of functioning mitochondria, was present from 60 minutes MCAO in moderately ischemic tissue and at 240 minutes in severely ischemic tissue. Cerebral ischemia stimulates GABA synthesis via GAD (Sze, 1979; Erecinska et al., 1996), and at the same time inhibits GABA breakdown by GABA transaminase (E.C.2.6.1.19) (Schousboe et al., 1973) consequently increasing the GABA content (Erecinska et al., 1984). Decreased GABA transaminase activity will impede the re-entry of carbon-atoms from GABA into the TCA cycle, thereby decreasing the activity of the GABA-shunt (Balazs, 1970), resulting in loss of TCA cycle intermediates and subsequently also of aspartate (see below). However, in severely ischemic tissue additional mechanisms causing GABA build up may be operating. The substantial increase in unlabeled GABA from 120 to 240 minutes MCAO, which was accompanied by a corresponding decline in glutamate levels, may arise from direct conversion of unlabeled glutamate to GABA taking place either intracellularly or extracellularly. A substantial increase in GABA levels have been demonstrated both in the cells (Torp et al., 1993) and in the extracellular compartment (Matsumoto et al., 1993) during cerebral ischemia. The increased extracellular GABA level is considered to be due to reversal of GABA transporters; however, based on the present results, the possibility that GAD leaks into the extracellular space and leads to extracellular production of GABA from extracellular glutamate cannot be excluded.

In the GABAergic neurons, MCAO compromised the flux through the TCA cycle by a (1) significant decrease or complete stop in the use of astrocytic glutamine, (2) rapid and marked reduction in use of neuronal precursors, and (3) progressive decrease in GABA-shunt activity. At 240 minutes MCAO, no TCA cycle flux was detected in the region rendered severely ischemic in the lateral caudoputamen and lower parietal cortex, suggesting the death of GABAergic neurons. Selective death of GABAergic neurons in the striatum has previously been shown in rats after transient forebrain ischemia (Francis and Pulsinelli, 1982). However, GABAergic neurons in the hippocampus and cortex are considered to be particularly resistant to ischemia and N-methyl- d -aspartate excitotoxicity (Tecoma and Choi, 1989; Gonzales et al., 1992; Johansen and Diemer, 1991). These differences in ischemia susceptibility between different populations of GABAergic neurons may be because GABAergic neurons are diverse in size and function. Based on the present data, GABAergic neurons in the lateral caudoputamen and lower parietal cortex were more sensitive to ischemia than glutamatergic neurons in the same region. However, it is not certain what properties lead to the increased sensitivity of GABAergic neurons in the lateral caudoputamen and lower parietal cortex to focal ischemia. The current findings imply that the complete interruption in the use of glutamine in GABA synthesis in this region may contribute substantially to the death of GABAergic neurons, because the reduction in glucose use and GABA-shunt activity was at least as noticeable in the region with moderate ischemia during the initial 120 minutes MCAO.

Because aspartate is found mainly in GABAergic neurons (Ottersen and Storm-Mathisen, 1985), the MCAO-induced changes in aspartate labeling and content were most likely connected to the changes in GABA metabolism. The significant reduction in aspartate synthesis in both ischemic regions from 30 minutes MCAO and beyond could be connected to the reduced glutamine use in GABAergic neurons, because in vitro data suggest that aspartate is predominantly synthesized from glutamine in GABAergic neurons (McKenna et al., 2000). The rapid decline in aspartate labeling found between 120 and 240 minutes MCAO in the upper frontoparietal cortex coincided with the complete stop in the use of astrocytic precursors, lending further support to the notion that glutamine is important in aspartate synthesis. In addition, aspartate consumption is increased by reduced GABA transaminase activity (Hassel et al., 1998), which was also found in the current study. In summary, aspartate consumption was probably increased because of decreased GABA shunt activity, whereas aspartate synthesis decreased because of reduced glutamine use.

MCAO-induced changes in astrocyte metabolism

Lateral caudoputamen and lower parietal cortex.

In this region of severe ischemia, astrocytic metabolism was reduced at 30 minutes, but was not severely impaired before 240 minutes MCAO. The first sign of declining astrocytic metabolism was reduced acetate metabolism, both oxidative and for use in amino acid synthesis. Still, acetyl coenzyme A synthetase, PC, and glutamine synthetase, all of which are adenosine triphosphate-dependent enzymes, operated throughout the experimental period. All enzymes appeared to maintain a slightly lower level of activity during the initial 120 minutes MCAO, but we detected dramatically reduced activity in the label incorporation via these enzymes at 240 minutes MCAO. No PC activity was detected in glutamine, and only a low level of PC activity was present in glutamate; furthermore, the acetate-glucose use ratio was at control levels in glutamine but increased in glutamate at 240 minutes MCAO. These findings suggest that glutamate metabolism in astrocytes is compartmentalized and that the compartment synthesizing glutamate for glutamine production is more susceptible to ischemia. Alternatively, ischemia reduces the transport of glutamate produced in the astrocytic mitochondria over the mitochondrial membrane to the cytosol, where glutamine synthetase is located. Previously, mitochondrial glutamate accumulation has been demonstrated in astrocytes in global ischemia (Torp et al., 1993). Alanine synthesis may also be considered a marker for astrocytic metabolism because it is preferentially synthesized in astrocytes in vitro (Westergaard et al., 1993). Furthermore, alanine synthesis is considered to increase specifically in astrocytes in cerebral ischemia (Bachelard, 1998). In severely ischemic tissue, alanine synthesis was significantly increased between 60 and 120 minutes MCAO but declined from 120 to 240 minutes MCAO, giving further evidence of rapidly declining astrocyte metabolism at the end of the experimental period.

Ischemic frontoparietal cortex.

In moderately ischemic tissue, astrocytic metabolism was relatively unaffected by MCAO until 120 minutes MCAO, at which time unmetabolized [1,2-¹³C]acetate appeared. [1,2-¹³C]Acetate has previously been detected in this region after 90 minutes MCAO (Håberg et al., 1998a). There was reduced PC activity at 60 minutes MCAO, but activity was notably reduced at 240 minutes MCAO. Both acetyl coenzyme A synthetase and glutamine synthetase activity were unaffected by MCAO. In addition, the increasing alanine synthesis present throughout the 240 minutes MCAO gave yet another indication of better preservation of metabolism in astrocytes of moderately ischemic tissue.

CONCLUSION

The present study showed the paradoxical role that astrocytes play in neuronal survival in ischemia. Although the presence of normal astrocytic metabolism appeared to be critical for neuronal survival, it was evident that alterations in the use of astrocytic precursors contributed significantly to neuronal death, albeit through different mechanisms in glutamatergic and GABAergic neurons. In glutamatergic neurons astrocytic precursors sustained the neurotransmitter pool of glutamate despite clear indications of failing neuronal metabolism. This finding was especially notable in the upper frontoparietal cortex (often referred to as the penumbra), thus showing the importance of glutamine in glutamate excitotoxicity. In GABAergic neurons, the complete stop in the use of glutamine in GABA formation in the ischemic core in the lateral caudoputamen and lower parietal cortex may have contributed significantly to the death of these neurons at 240 minutes MCAO. Therefore, neuronal survival during focal cerebral ischemia depends on astrocytic metabolism and neuronal use of astrocytic precursors.

Footnotes

Acknowledgment:

The authors thank Kristin Antonsen and Torild Krogstad for performing the high-performance liquid chromatography analysis at the Institute of Biotechnology, Norwegian University of Science and Technology.

Abbreviations Used:

References

Aas

J-E

Berg-Johnsen

Hegstad

Laake

Langmoen

Ottersen

(1993) Redistribution of glutamate and glutamine in slices of human neocortex exposed to combined hypoxia and glucose deprivation in vitro. J Cereb Blood Flow Metab 13:503–515

Aoki

Kaneko

Starr

Pickel

(1991) Identification of mitochondrial and non-mitochondrial glutaminase within select neurons and glia of rat forebrain by electron microscopic immunocytochemistry. J Neurosci Res 28:531–548

Bachelard

(1998) Landmarks in the application of ¹³C-magnetic resonance spectroscopy to studies of neuronal/glial relationships. Dev Neurosci 20:277–288

Back

(1998) Pathophysiology of the ischemic penumbra – revision of a concept. Cell Mol Neurobiol 18:621–638

Badar-Goffer

Bachelard

Morris

(1990) Cerebral metabolism of acetate and glucose studied by ¹³C NMR spectroscopy. A technique for investigating metabolic compartmentation in the brain. Biochem J 266:133–139

Balazs

Machiyama

Hammond

Julian

Richter

(1970) The operation of the γ-aminobutyrate bypath of the tricarboxylic acid cycle in brain tissue in vitro. Biochem J 116:445–467

Benjamin

(1981) Control of glutaminase activity in rat brain cortex in vitro: Influence of glutamate, phosphate, ammonia, calcium and hydrogen ions. Brain Res 2:363–377

Castillo

Dávalos

Naveiro

Noya

(1996) Neuroexcitatory amino acids and their relation to infarct size and neurological deficit in ischemic stroke. Stroke 27:1060–1065

Cataltepe

Towfighi

Vannucci

(1996) Cerebrospinal fluid concentrations of glutamate and GABA during perinatal cerebral hypoxia-ischemia and seizures. Brain Res 709:326–330

10.

Choi

(1992) Excitotoxic cell death. J Neurobiol 23:1261–1276

11.

Cremer

Lucas

(1971) Sodium pentobarbitone and metabolic compartments in rat brain. Brain Res 35:619–621

12.

Daikhin

Yudkoff

(2000) Compartmentation of brain glutamate metabolism in neurons and glia. J Nutr 130 (Suppl 4):1026S–1031S

13.

Erecinska

(1987) The neurotransmitter amino acid transport systems. A fresh outlook on an old problem. Biochem Pharmacol 36: 3547–3555

14.

Erecinska

Nelson

Wilson

Silver

(1984) Neurotransmitter amino acids in the CNS: I. Regional changes in amino acid levels in rat brain during ischemia and reperfusion. Brain Res 304:9–22

15.

Erecinska

Nelson

Daikhin

Yudkoff

(1996) Regulation of GABA level in rat brain synaptosomes: Fluxes through enzymes of the GABA shunt and effects of glutamate, calcium, and ketone bodies. J Neurochem 67:2325–2334

16.

Farinelli

Nicklas

(1992) Glutamate metabolism in rat cortical astrocyte cultures. J Neurochem 58:1905–1915

17.

Francis

Pulsinelli

(1982) The response of GABAergic and cholinergic neurons to transient cerebral ischemia. Brain Res 243:271–278

18.

Garcia

Liu

K-F

K-L

(1995) Neuronal necrosis after middle cerebral artery occlusion in Wistar rats progress at different time intervals in the caudoputamen and the cortex. Stroke 26:636–643

19.

Goldberg

Monyer

Choi

(1988) Hypoxic neuronal in vitro injury depends on extracellular glutamine. Neurosci Lett 94:52–57

20.

Gonzales

Lin

Chesselet

(1992) Relative sparing of GABAergic interneurons in the striatum of gerbils with ischemia induced lesions. Neurosci Lett 135:53–58

21.

Håberg

Haraldseth

Unsgård

Sonnewald

(1998a) In vivo injection of [1-¹³C]glucose and [1,2-¹³C]acetate combined with ex vivo ¹³C NMR spectroscopy: A novel approach to the study of MCA occlusion in rat. J Cereb Blood Flow Metab 18: 1223–1232

22.

Håberg

Takahashi

Yamaguchi

Hjelstuen

Haraldseth

(1998b) Neuroprotective effect of the novel glutamate AMPA receptor antagonist YM872 assessed with in vivo MR imaging of rat MCA occlusion. Brain Res 811:63–70

23.

Hansen

Warner

Todd

Vust

Trawick

(1988) Distribution of cerebral blood flow during halothane versus isoflurane anesthesia in rats. Anesthesiology 69:332–337

24.

Hassel

Johannessen

Sonnewald

Fonnum

(1998) Quantification of the GABA shunt and the importance of the GABA shunt versus the 2-oxoglutarate dehydrogenase pathway in GABAergic neurons. J Neurochem 71:1511–1518

25.

Hertz

(1979) Functional interactions between neurons and astrocytes. I. Turnover and metabolism of putative amino acid transmitters. Prog Neurobiol 13:277–323

26.

Hertz

Dringen

Schoousboe

Robinson

(1999) Astrocytes: Glutamate producers for neurons. J Neurosci Res 57:417–428

27.

Hoffman

Edelman

Kochs

Werner

Segil

Albrecht

(1991) Cerebral autoregulation in awake versus isoflurane-anesthetized rats. Anesth Analg 73:753–757

28.

Hogstad

Svenneby

Torgner

IAa

Kvamme

Hertz

Schousboe

(1988) Glutaminase in neurons and astrocyte cultured from mouse brain: Kinetic properties and effects of phosphate, glutamate and ammonia. Neurochem Res 13:383–388

29.

Huang

Hertz

(1994) Effect of anoxia on glutamate formation from glutamine in cultured neurons: Dependence on neuronal subtypes. Brain Res 660:129–137

30.

Johansen

Diemer

(1991) Enhancement of GABA neurotransmission after cerebral ischemia in the rat reduces loss of hippocampal CA1 pyramidal cells. Acta Neurol Scand 84:1–6

31.

Kaufman

Driscoll

(1993) Evidence for cooperativity between neurons and astroglia in the regulation of CO₂ fixation in vitro. Dev Neurosci 15:299–305

32.

Künneck

Ceredan

Seelig

(1993) Cerebral metabolism of [1,2-¹³C₂]glucose and [U-¹³C₄]3-hydroxybutyrate in rat brain as detected by ¹³C NMR spectroscopy. NMR Biomed 6:264–277

33.

Kvamme

Svenneby

Hertz

Schousboe

(1982) Properties of phosphate activated glutaminase in astrocytes cultured from mouse brain. Nerochem Res 6:761–770

34.

Kvamme

Roberg

Torgner

(2000) Phosphate-activated glutaminase and mitochondrial glutamine transport in the brain. Neurochem Res 25:1407–1419

35.

Laake

Slyngstad

Haug

F-MS

Ottersen

(1995) Glutamine from glial cells is essential for the maintenance of the nerve terminal pool of glutamate: Immunogold evidence from hippocampal slice cultures. J Neurochem 65:871–881

36.

Longa

Weinstein

Cummins

(1989) Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 20:84–91

37.

Matsumoto

Graf

Rosner

Taguchi

Heiss

W-D

(1993) Elevation of neuroactive substances in the cortex of cats during prolonged focal ischemia. J Cereb Blood Flow Metab 13:586–594

38.

McKenna

Sonnewald

Huang

Stevenson

Zielke

(1996) Exogenous glutamate concentration regulates the metabolic fate of glutamate in astrocytes. J Neurochem 66:386–393

39.

McKenna

Stevenson

Huang

Hopkins

(2000) Distribution of the enzymes glutamate dehydrogenase and aspartate aminotransferase in cortical synaptic mitochondria contributes to metabolic compartmentation in cortical synaptic terminals. Neurochem Int 37:229–241

40.

McLean

Badar-Goffer

Bachelard

Morris

(1993) Simultaneous measurement of neuronal and glial metabolism via ¹³C-MRS isotopomer analysis. Abstract presented at the 12th annual meeting of Society of Magnetic Resonance in Medicine; August 14–20; New York, NY

41.

Meldrum

(2000) Glutamate as transmitter in the brain: Review of physiology and pathology. J Nutr 130(Suppl 4):1007S–1015S

42.

Memezawa

Minamisawa

Smith

M-L

Siesjö

(1992a) Ischemic penumbra in a model of reversible middle cerebral artery occlusion in the rat. Exp Brain Res 89:67–78

43.

Memezawa

Smith

Siesjö

(1992b) Penumbral tissues salvaged by reperfusion following middle cerebral artery occlusion in rats. Stroke 23:552–560

44.

Müller

Haraldseth

Jones

Sebastiani

Lindboe

Øksendal

Unsgård

(1995) Combined perfusion and diffusion weighted MR imaging in a rat model of reversible middle cerebral artery occlusion. Stroke 26:451–458

45.

Norenberg

Martinez-Hernandez

(1979) Fine structural localization of glutamine synthetase in astrocytes of rat brain. Brain Res 161:303–310

46.

Obrenovitch

(1996) Origins of glutamate release in ischemia. Acta Neurochir Suppl (Wein) 66:50–55

47.

Ottersen

(1989) Quantitative electron microscopic immunocytochemistry of neuroactive amino acids. Anat Embryol 180:1–15

48.

Ottersen

Storm-Matisen

(1985) Different neuronal localization of aspertate-like and glutamate-like immunoreactivities in the hippocampus of rat, guinea-pig and Senegalese baboon (Papio papio), with a note on the distribution γ-aminobutyrate. Neurosci 6:589–606

49.

Ottersen

Laake

Reichelt

Haug

Torp

(1996) Ischemic disruption of glutamate homeostasis in brain: Quantitative immunocytochemical analyses. J Chem Neuroanat 12:1–14

50.

Pascual

Carceller

Roda

Cerdán

(1998) Glutamate, glutamine, and GABA as substrates for the neuronal and glial compartments after focal cerebral ischemia in rats. Stroke 29:1048–1057

51.

Patel

(1974) The relative significance of CO₂-fixing enzymes in the metabolism of rat brain. J Neurochem 22:717–724

52.

Rosenberg

(1991) Accumulation of extracellular glutamate and neuronal death in astrocyte-poor cortical cultures exposed to glutamine. Glia 4:91–100

53.

Schousboe

J-Y

Roberts

(1973) Purification and characterization of the 4-aminobyturtate-2-ketoglutatate transaminase from mouse brain. Biochem 12:2868–2873

54.

Schwartz-Bloom

Sah

(2001) γ-Aminobutyric acid_A neurotransmission and cerebral ischemia. J Neurochem 77:353–371

55.

Shank

Leo

Zielke

(1993) Cerebral metabolic compartmentation as revealed by nuclear magnetic resonance analysis of d-[1-¹³C]glucose metabolism. J Neurochem 61:315–323

56.

Sloviter

Dichter

Rachinsky

Dean

Goodman

Sollas

Martin

(1996) Basal expression and induction of glutamate decarboxylase and GABA in excitatory granule cells of the rat and monkey hippocampal dentate gyrus. J Comp Neurol 373:593–618

57.

Sonnewald

Håberg

Haraldseth

Müller

Unsgård

(1996) Glial/neuronal interactions studied by MRS of rat brain after MCAO. In: Pharmacology of cerebral ischemia. ( Kriegelstein

, ed), Stuttgart: Medpharm, pp. 285–290

58.

Swanson

Shiraishi

Morton

Sharp

(1990) Methionine sulfoximine reduces cortical infarct size in rats after middle cerebral artery occlusion. Stroke 21:322–327

59.

Swanson

(1992) Astrocyte glutamate uptake during chemical hypoxia in vitro. Neurosci Lett 147:143–146

60.

Sze

(1979) L-Glutamate decarboxylase. Adv Exp Med Biol 123: 59–78

61.

Szerb

O'Regan

(1985) Effect of glutamine on glutamate release from hippocampal slices induced by high K⁺ or by electrical stimulation: Interactions with different Ca²⁺ concentrations. J Neurochem 44:1724–1731

62.

Takagi

Ginsberg

Globus

MY-T

Dietrich

Martinez

Kraydieh

Busto

(1993) Changes in amino acid neurotransmitters and cerebral blood flow in the ischemic penumbral region following middle cerebral artery occlusion in the rat: Correlation with histopathology. J Cereb Blood Flow Metab 13:575–585

63.

Tanaka

(2000) Functions of glutamate transporters in the brain. Neurosci Res 37:15–19

64.

Taylor

McLean

Morris

Bachelard

(1996) Approaches to studies on neuronal/glial relationships by ¹³C-MRS analysis. Dev Neurosci 18:434–442

65.

Tecoma

Choi

(1989) GABAergic neocortical neurons are resistant to NMDA receptor-mediated injury. Neurology 39:676–682

66.

Torp

Arvin

Le Peillet

Chapman

Ottersen

Meldrum

(1993) Effect of ischemia and reperfusion on the extra- and intracellular distribution of glutamate, glutamine, aspartate and GABA in the rat hippocampus, with a note on the effect of the sodium channel blocker BW1003C87. Exp Brain Res 96:365–376

67.

Van den Berg

(1973) A model of compartmentation in the mouse brain based on glucose and acetate metabolism. In: Metabolic compartmentation in the brain ( Balazs

Cremer

, eds), London: Maxmillian Press, pp. 137–166

68.

Varoqui

Zhu

Yao

Ming

Erickson

(2000) Cloning and functional identification of a neuronal glutamine transporter. J Biol Chem 275:4049–4054

69.

Wallace

Dawson

Jr (1993) Regional differences in glutaminase activation by phosphate and calcium in rat brain: Impairment in aged rats and implications for regional glutaminase isoenzymes. Neurochem Res 12:1271–1279

70.

Waniewski

and Martin

(1998) Preferential utilization of acetate by astrocytes is attributable to transport. J Neurosci 18:5225–5233

71.

Westergaard

Varming

Peng

Sonnewald

Hertz

Schousboe

(1993) Uptake, release and metabolism of alanine in neurons and astrocytes in primary cultures. J Neurosci Res 35:540–545

72.

Westergaard

Sonnewald

Schousboe

(1995) Metabolic trafficking between neurons and astrocytes: The glutamine/glutamate cycle revisited. Dev Neurosci 17:203–211

73.

ACH

Drejer

Hertz

Schousboe

(1983) Pyruvate carboxylase activity in primary cultures of astrocytes and neurons. J Neurochem 41:1484–1487

Differences in Neurotransmitter Synthesis and Intermediary Metabolism between Glutamatergic and GABAergic Neurons during 4 Hours of Middle Cerebral Artery Occlusion in the Rat: The Role of Astrocytes in Neuronal Survival

Abstract

Keywords

MATERIALS AND METHOD

Animal experiments

13C magnetic resonance spectroscopy

1H magnetic resonance spectroscopy

High-performance liquid chromatography analysis

Analysis of13C and1H magnetic resonance spectra and the rationale behind data interpretation

RESULTS

Diffusion-weighted magnetic resonance imaging

Analysis of plasma extracts

Effects of middle cerebral artery occlusion on total amino acid content

Lateral caudoputamen and lower parietal cortex.

Upper frontoparietal cortex.

13C magnetic resonance spectroscopy of brain extracts: effects of middle cerebral artery occlusion on cerebral metabolism

[1-13C]Glucose and [1,2-13C]acetate.

Glutamate and glutamine.

γ-Aminobutyric acid.

Aspartate and alanine.

Effects of isoflurane on cerebral metabolism

Lateral caudoputamen and lower parietal cortex.

Upper frontoparietal cortex.

DISCUSSION

MCAO induced changes in the glutamate-glutamine cycle

Lateral caudoputamen and lower parietal cortex.

Upper frontoparietal cortex.

MCAO-induced changes in GABA metabolism

MCAO-induced changes in astrocyte metabolism

Lateral caudoputamen and lower parietal cortex.

Ischemic frontoparietal cortex.

CONCLUSION

Footnotes

Acknowledgment:

Abbreviations Used:

References

¹³C magnetic resonance spectroscopy

¹H magnetic resonance spectroscopy

Analysis of¹³C and¹H magnetic resonance spectra and the rationale behind data interpretation

¹³C magnetic resonance spectroscopy of brain extracts: effects of middle cerebral artery occlusion on cerebral metabolism

[1-¹³C]Glucose and [1,2-¹³C]acetate.