B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te /b ra i n res

Research Report

Changes in glutamate transporter expression in mouseforebrain areas following focal ischemia

Pirusha Ketheeswaranathana, Neil A. Turnera, Emma J. Sparya, Trevor F.C. Battena,Barry W. McCollb, Sikha Sahaa,⁎aDivision of Cardiovascular and Neuronal Remodelling, Leeds Institute for Genetics, Health and Therapeutics, University of Leeds, Leeds,LS2 9JT, UKbThe Roslin Institute, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK

A R T I C L E I N F O

⁎ Corresponding author at: Division of CardioWorsley Building, University of Leeds, Leeds

E-mail address: s.saha@leeds.ac.uk (S. SaAbbreviations: AMPA, α-amino-3-hydroxy

excitatory amino acid carrier-1; GLAST, glutaocclusion; NMDA, N-methyl-D-aspartate; PKCreaction

0006-8993/$ – see front matter © 2011 Elseviedoi:10.1016/j.brainres.2011.08.029

A B S T R A C T

Article history:Accepted 12 August 2011Available online 19 August 2011

Dysfunction of glutamate transporters has been proposed to promote neuronal death inmodelled cerebral ischemia. However, these studies have produced conflicting results and thechanges in glutamate transporter expression have not yet been examined in a mouse focalischemic strokemodel. This study used quantitative real-time reverse-transcription polymerasechain reaction to examine glutamate transporter mRNA expression in the hippocampus,cortex and striatum in amousemodel of focal ischemic stroke induced bymiddle cerebral arteryocclusion (MCAO). Effects on mRNA expression of glial (GLT-1, GLAST) and neuronal (EAAC1)glutamate transporters in these brain areas were assessed by comparing MCAO brains withsham-operated control brains. Changes in transporter proteins were also assessed by immuno-histochemistry using specific antibodies to GLT-1 and GLAST. Following focal ischemia, GLT-1mRNA expression was decreased significantly in the ipsilateral hippocampus and cortex com-pared to the sham-operated brains (p<0.05). There were no significant differences in GLAST orEAAC1 mRNA expression between MCAO and sham-operated brains. Immunohistochemistryalso confirmed a marked reduction in GLT-1 immunoreactivity in the cortex and hippocampus.Down regulation of GLT-1 in these brain areas may impair normal clearance of synaptically-released glutamate and contribute to neural damage following focal ischemic insult.

© 2011 Elsevier B.V. All rights reserved.

Keywords:Middle cerebral artery occlusionMouseGlutamate transporterStrokeQuantitative real time PCRGLT-1

1. Introduction

Ischemic stroke is a leading cause of death and amajor risk fac-tor for the later development of various neurological disordersincluding vascular dementia and Alzheimer's disease. Stroke

vascular and Neuronal R, LS2 9JT, UK. Fax: +44 113ha).-5-methylisoxazole-4-promate-aspartate transpor, protein kinase C; qRT-P

r B.V. All rights reserved.

produces cell death by increasing extracellular concentrationsof glutamate that excessively activate ionotropic glutamate re-ceptors and initiate calcium overload and neuronal death (forreview, Camacho and Massieu, 2006; Johnston, 2005). In thenormal brain, high-affinity plasma membrane glutamate

emodelling, Leeds Institute of Genetics, Health and Therapeutics,343 4803.

pionic acid; EAAT, excitatory amino acid transporter; EAAC-1,ter; GLT-1, glutamate transporter-1; MCAO, middle cerebral arteryCR, quantitative real-time reverse-transcription polymerase chain

94 B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

transporters actively clear the glutamate released in the synap-tic cleft tomaintain glutamate homeostasis and prevent neuro-toxicity (Anderson and Swanson, 2000; Camacho and Massieu,2006).

In rodents three high affinity glutamate transportersubtypes have been isolated: GLAST (glutamate-aspartatetransporter), GLT-1 (glutamate transporter-1) and EAAC-1 (ex-citatory amino acid carrier-1). Human homologues of thesethree transporters are EAAT1 (excitatory amino acid transporter1 = GLAST), EAAT2 (=GLT-1) and EAAT3 (=EAAC-1). Two addi-tional transporters (EAAT4 and EAAT5) have also been cloned(for review see, Beart and O'Shea, 2007; Danbolt, 2001). Theseplasma membrane glutamate transporters are driven by bothsodium and potassium gradients and recycle glutamate backinto neurones (Anderson and Swanson, 2000; Danbolt, 2001).GLAST (EAAT1) and GLT-1 (EAAT2) are primarily expressed byastrocytes (Sims and Robinson, 1999), which also express theenzyme glutamine synthetase (GS) to convert glutamate to glu-tamine; the latter amino acid is then ‘recycled’ into neurones(Danbolt, 2001). EAAC-1 (EAAT3), on the other hand, is predom-inantly expressed in neurones especially in γ-aminobutyric acid(GABA)-containing neurones, but possibly also in glutamatergic,cholinergic and aminergic neurones (Sims and Robinson, 1999).In addition to clearing the released glutamate, subtypes of glu-tamate transporters have also been proposed to release gluta-mate into the extracellular space by reverse action duringischemic conditions when the ion gradient or membrane po-tential drops (Phillis et al., 2000; Rossi et al., 2000).

Although dysfunction of glutamate transporters has beenproposed to initiate neuronal death following cerebral ischemiain rats and gerbils (Fujita et al., 1999; Rao et al., 2001a), thechanges in glutamate transporter subtype expression have notyet been examined in a mouse ischemic stroke model. Further-more, changes in mRNA coding for glutamate transporter sub-types have not been investigated by quantitative real-timereverse-transcription polymerase chain reaction (qRT-PCR),which is accepted to be the most sensitive method available todate to quantify mRNA levels with reliable specificity (Valasekand Repa, 2005). Considering the importance of the physiologicalfunction of glutamate transporters in the clearance of extracel-lular glutamate and the inconsistent results obtained in earlierstudies, the present study aimed to use qRT-PCR to investigatealterations in mRNA expression of the glial transporters GLASTandGLT-1, and the neuronal transporter EAAC-1 in the striatum,cerebral cortex and the hippocampus (brain areasmost vulnera-ble to stroke) in a mouse model of focal ischemia. The changesin transporter proteins in the cortex and the hippocampuswere also confirmed by immunohistochemistry using specificantisera. This information is important to take advantage ofthe prospective use of transgenic mice with mutation of specificglutamate transporter genes in future to understand the molec-ularmechanisms underlying stroke and stroke related disorders.

2. Results

2.1. Neurological deficit

After 1 h ischemia and 24 h reperfusion, mice exhibitedmoderate to severe neurological deficits such as circling,

decreased resistance to lateral push and reduced locomotoractivity, flexion of contralateral torso and forelimb upon lift-ing the animal by its tail, and loss of righting reflex. Anymice not showing behavioural deficit consistent with a 1 hMCAO were excluded from further study.

2.2. PCR efficiency

In order to compare relative mRNA levels of different tran-scripts it was necessary to establish that reaction efficiencieswere similar for the transcripts being compared. The efficien-cy of the qRT-PCR reactions for the different primer/probe setswas determined from CT values obtained across a 10-fold seri-al dilution of cDNA samples (data not shown). Efficiency (E)was calculated as E=(10−1/k−1), in which 1/k represents theslope of the serial dilution graph (Kubista et al., 2006). Primerefficiencies were calculated to be 91% (EAAC-1), 94% (GLAST)and 92% (GLT-1), thus allowing accurate comparison betweendifferent primer/probe sets.

2.3. Glutamate transporter mRNA expression innormal brains

The relative mRNA expression levels of GLT-1, GLAST andEAAC-1 in tissue samples (Fig. 1) from different areas of theforebrain (cortex, hippocampus, striatum) were compared incontrol mouse brains (Figs. 2A, B and C). GLT-1 appeared tobe the most highly expressed transporter in all three areas ofthe forebrain, followed by GLAST. EAAC-1 was expressed inrelatively small amounts. A lower expression of all three glu-tamate transporters was observed in striatum (Fig. 2C) com-pared to other forebrain areas. There were no significantdifferences in transporter expression between right and leftsides of the brain.

2.4. Glutamate transporter mRNA expression in strokeand sham-operated brains

To assess changes in the expression of GLT-1, GLAST andEAAC-1 following focal ischemia, the stroke affected brains(n=7) were compared with sham-operated (n=7) brains. Allstroke brains were separated into right and left sides, the lat-ter representing the ischemic side. Tissue samples from theleft sides of the stroke brains were compared to the sameareas of the brains of sham-operated controls, and to theright sides of the same stroke brains to determine whetherthe stroke had affected both sides of the brain (Figs. 3–5). Insham-operated control brains, no significant changes inGLT-1, GLAST or EAAC-1 mRNA expression were observed be-tween the left (ipsilateral) side of the cortex, hippocampusand the striatum compared to the right (contralateral) side(data not shown).

2.4.1. GLT-1 mRNA expressionFollowing 1 h MCAO and 24 h reperfusion, a significant de-crease in GLT-1 mRNAwas observed on the ipsilateral (stroke)side of the cortex (Fig. 3A) and the hippocampus (Fig. 3B) com-pared to the same side in sham-operated control brains. Therewas no significant change in GLT-1mRNA levels in the ipsilat-eral side of the striatum compared to the sham-operated

Fig. 1 – Thick (~0.5 mm) section of mouse forebrain showing areas of tissue sampling within the hippocampus (H), striatum (S)and parietal cortex (C). Scale bar=1 mm.

95B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

brains (Fig. 3C). Moreover, no significant changes in GLT-1mRNA levels were found for any of the areas investigatedwhen comparing tissue from the left and right (contralateral)sides of the brains of MCAO animals (Figs. 3A, B, C).

2.4.2. GLAST mRNA expressionA slight decrease in GLASTmRNA expression in the ipsilateralside of cortex (Fig. 4A) and the hippocampus (Fig. 4B) in strokebrains compared to sham-operated controls was not statisti-cally significant. However, in the cortex a significant higherGLAST mRNA expression was observed on the contralateral(right) side compared to the ipsilateral (left), stroke side(Fig. 4A). Similarly, in the striatum, GLAST mRNA expressionlevel was significantly higher on the contralateral side ofstroke brains compared to the infarcted side (Fig. 4C).

2.4.3. EAAC-1 mRNA expressionIn contrast, EAAC-1mRNA expression was unaltered in the ip-silateral stroke sides of the cortex (Fig. 5A), hippocampus(Fig. 5B) or striatum (Fig. 5C) when compared to sham-operated control brains. Similarly, there were no differences

A. Cortex

GLT1 R

GLT1 L

GLAST R

GLAST L

EAAC1 R

EAAC1 L0

5

10

15

mR

NA

leve

l (%

bet

a-ac

tin

)

B. Hippo

GLT1 R

GLT1 L

GLAST R

G

0

5

10

15

Fig. 2 – Comparison of relative mRNA expression levels of GLT-1,expressed as a percentage of β-actin mRNA levels (n=3). R and L

in the EAAC-1 mRNA level on the ipsilateral sides in the cor-tex, hippocampus or the striatum when compared to the con-tralateral sides following MCAO (Figs. 5A, B, C).

2.5. Immunohistochemical analysis

In control animals (n=4), GLAST and GLT-1 immunoreactiv-ities were localised principally to puncta and processes inthe neuropil throughout the forebrain areas, consistent withtheir expression in astroglial processes. GLAST and GLT-1 im-munoreactive processes outlined the neuronal cell bodies andprimary dendrites, with the most prominent labelling ob-served around the unlabelled pyramidal cells in the cortexand the hippocampus (e.g. Fig. 6G). In contrast, the distribu-tions of GLAST and GLT-1 immunoreactivities in the striatumwere patchier, with weaker labelling appearing to be restrictedto glial processes immediately surrounding the neuronal cellbodies (data not shown).

FollowingMCAO (n=4), GLT-1 labelling appeared to bemark-edly decreased in the cortex (Figs. 6C, F, I) on the infarcted sideof the brain when compared to the same area in sham-operated

campus

LAST L

EAAC1 R

EAAC1 L

C. Striatum

GLT1 R

GLT1 L

GLAST R

GLAST L

EAAC1 R

EAAC1 L0

5

10

15

GLAST and EAAC1 in normal mouse brain. qRT-PCR data arerefer to right and left sides of brain respectively.

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0

2

4

6

8

10

12 * P=0.016

GLT

-1 m

RN

A (%

bet

a-ac

tin)

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0

2

4

6

8

10

12* P=0.013

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0

2

4

6

8

A. Cortex B. Hippocampus C. Striatum

Fig. 3 – GLT-1 mRNA expression level in specific areas of stroke brains and sham-operated brains. qRT-PCR data are expressedrelative to β-actin mRNA levels (n=7). Stroke induced in left side of brain.

96 B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

brains (Figs. 6A, D, G) or the contralateral side of the same brain(Figs. 6B, E, H). This impression was confirmed by the quantita-tive analysis (Fig. 7), which showed significant reductions inboth number of immunolabelled puncta per unit area and per-centage of total image area occupied by immunolabelled struc-tures on the infarcted sides of MCAO brains compared to thecontralateral sides and to sham-operated brains. Thereappeared to be a redistribution of GLT-1 immunoreactivity,with a reduction in immunoreactive processes and punctaaround unlabelled pyramidal cells, but labelling in granularstructures resembling astrocytes or small neuronal cell bodiesremained similar (Figs. 6C, F, I). The labelling patterns forGFAP (Figs. 6A–C) and GS (Figs. 6D–F) appeared less affected.The changes in GLT-1 immunostaining observed in the hippo-campus were similar; labelling around CA1 pyramidal neuronswas drastically reduced and this was similarly replaced by la-belling of granular structures (data not shown). There was alsoreduction of labelling for GLAST in the cortex (Figs. 6J–L) onthe infarcted side of the brains of MCAO operated animals.

3. Discussion

This study is novel in that it is the first to use qRT-PCR to eval-uate expression of the glial (GLT-1 and GLAST) and neuronal

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0

1

2

3

4

5* P=0.030

GL

AS

T m

RN

A (

% b

eta-

acti

n)

Sham L

eft

Stroke

L

0

1

2

3

4

5

A. Cortex B. Hipp

Fig. 4 – GLASTmRNA expression level in specific areas of stroke brelative to β-actin mRNA levels (n=7). Stroke induced in left side

(EAAC-1) glutamate transporter mRNAs in the striatum, cor-tex and hippocampus after transient focal ischemia in mice.Our experimental protocol employed a 1 h MCAO and 24 h re-perfusion, as this produces neuronal damage in the striatumand cortex both in rats and mice (McColl et al., 2004; Rao etal., 2001a). In agreement with other studies (Butler et al.,2002; McColl et al., 2004), we also consistently observed neuro-nal degeneration in the hippocampus following 1 hMCAO and24 h reperfusion. Our results suggest that alterations in gluta-mate transporter mRNA expression in the cortex and hippo-campus induced by focal ischemia involve mainly GLT-1,since we observed no significant changes in GLAST andEAAC-1 mRNA levels compared to sham-operated brains.Quantitative immunohistochemical analysis confirmed amarked decrease in area occupied by GLT-1 immunoreactivestructures, suggesting that GLT1 protein levels in ischemicareas of the cortex are reduced compared to sham-operatedbrains. A clearly different pattern of transporter immunoreac-tivity was observed, consistent with a redistribution of immu-noreactivity from numerous fine glial processes into granularcell bodies. This observation may be a consequence of astro-cytes undergoing pathophysiological changes as a protectiveresponse to ischemic insult and energy failure (Giffard andSwanson, 2005; Sullivan et al., 2010). Confirmation of ourdata at the cellular level using immunohistochemistry is

eft

Stroke

Rig

ht

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0

1

2

3

4 * P=0.035

ocampus C. Striatum

rains and sham-operated brains. qRT-PCR data are expressedof brain.

A. Cortex B. Hippocampus C. Striatum

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0.0

0.5

1.0

1.5

2.0

EA

AC

1 m

RN

A (

% b

eta-

acti

n)

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0.0

0.5

1.0

1.5

2.0

Sham L

eft

Stroke

Lef

t

Stroke

Rig

ht0.0

0.5

1.0

1.5

2.0

Fig. 5 – EAAC1mRNA expression level in specific areas of stroke brains and sham-operated brains. qRT-PCR data are expressedrelative to β-actin mRNA levels (n=7). Stroke induced in left side of brain.

97B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

important, since from PCR analysis of tissue samples alone itcannot be known whether changes in mRNA expression aretranslated into changes in functional transporter proteins,nor precisely where such changes occur in relation to the in-farcted area. While immunohistochemistry showed a signifi-cant reduction in GLT1 immunoreactivity in the cortex onthe stroke affected side of the brain compared to the contra-lateral side, such a difference inmRNA expression was not ob-served from the PCR data; however levels of both mRNA andimmunoreactivity appeared to be reduced on the contralateralside compared to sham operated controls. This might reflectan effect of MCAO on blood flow on the contralateral side ofthe brain that influences levels of GLT1 expression.

A down-regulation of GLT-1 protein was also observed inthe cerebral cortex at 24–72 h following MCAO in spontane-ously hypertensive rats, with GLAST levels unchanged (Raoet al., 2001a). Significant decreases in GLT-1 mRNA and pro-tein levels were also demonstrated in rat hippocampusfollowing transient global ischemia induced by 4 vessels'occlusion (Bruhn et al., 2000). This decrease in GLT-1 expres-sion was not due to a general dysfunction in astrocytic proteinsynthesis, as the astrocyte-specific marker glial fibrillaryacidic protein (GFAP) was increased following ischemicinsults. Transient global ischemia induced in gerbils by bilat-eral carotid artery occlusion also led to a reduction of GLT-1protein in hippocampus one day after reperfusion (Kim etal., 2006; Rao et al., 2000). Several studies showed decreasesin GLT-1 mRNA and protein levels in the rat hippocampusfollowing ischemic insult (Chen et al., 2005; Torp et al., 1995).A similar decrease in mRNA of GLT-1v (EAAT2v), a splice-variant form of GLT-1, in the rat cortex was observed follow-ing fluid percussion insult that models traumatic brain injury(Yi et al., 2005).

Although the half-lives of GLT-1 and the other glutamatetransporter subtypes have not been determined in vivo, thereis evidence that the half-life of GLT-1 is longer than 24 h in as-trocyte cultures (Zelenaia and Robinson, 2000). This suggeststhat decreases in GLT-1 observed in the present study in miceand earlier studies in rats and gerbils are dependent upondecreased transcription/translation due to ischemic insults.However, very little is known about the mechanisms thatmight contribute to decreases in GLT-1 protein levels followingischemia. Previous studies have demonstrated that acute

activation of various signalling molecules including proteinkinase C (PKC) and scaffolding protein causes redistributionof GLT-1 from the plasmamembrane to an intracellular com-partment (Beart and O'Shea, 2007; Danbolt, 2001). A recentstudy reported that the PKC-induced redistribution of GLT-1is dependent upon clathrin-mediated endocytosis (Susarlaand Robinson, 2003), a novel mechanism by which the levelsof GLT-1 could be rapidly down regulated via lysosomal deg-radation. McColl et al. (2003) earlier reported an increase inrabaptin-5 and rabaptin-4 proteins, markers of endocyticpathway activity, in neurons and glial cells in the hippocam-pus and neocortex of human post-mortem brains subjectedto an episode of global ischemia suggesting that global ische-mia may increase endocytosis in neurones and glia. Similarcellular machinery may be involved in the redistribution ofGLT-1 and contribute to decreases in total GLT-1 levelsunder ischemic conditions.

GLT-1 is abundant in the forebrain, particularly in the hip-pocampus, cerebral cortex and the striatum (Lehre et al., 1995)andmouse knockout studies have demonstrated that GLT-1 isresponsible for more than 90% of glutamate uptake in theforebrain with GLT-1 knockout mice showing more neuraldamage following cerebral injury than their wild-type coun-terparts (Tanaka et al., 1997). Similarly, Rao et al. (2001c)found that antisense knockdown of GLT-1, but not EAAC-1,exacerbated the ischemic infarct volume and neuronal dam-age in the cerebral cortex and the striatum of the spontane-ously hypertensive rat again using a 1 h transient MCAO and24 h reperfusion protocol. Antisense knockdown of GLT-1also exacerbated hippocampal neuronal damage followingtraumatic injury to rat brain (Rao et al., 2001b). This evidencestrongly suggests that GLT-1 is the key glutamate transportersubtype protecting neurones from exposure to excitotoxiclevels of glutamate in ischemic conditions.

Though we found no significant changes in GLAST mRNAlevel in the ipsilateral stroke brains compared to the sham-operated control brains, a significantly greater expression ofGLAST mRNA in the contralateral side of the stroke brainscompared to the ipsilateral side was observed. An increaseof GLAST expression contralaterally might represent a com-pensatory neuroprotective mechanism following stroke. Pre-vious studies also reported that GLAST expression was notaltered following ischemia in the rat forebrain (Chen et al.,

Fig. 6 – GLT-1 (A–I) and GLAST (J–L) immunoreactivities in similar areas of the parietal cortex, layers 5–6b. Astrocyte processeslabelled for GLT-1 (red fluorescence) are seen around unlabelled neuronal somata in the cortex of sham-operated mice (A, D, G)and contralaterally to the infarct in ischemic mice (B, E, H). On the ipsilateral (stroke) side of mice subjected to MCAO,immunoreactive processes are depleted in the infarcted areas (arrowheads), but immunoreactive structures resemblingastrocytes or small neurones are observed (arrows in C, F, I). The distributions of immunoreactivities for GFAP (green in A–C)and GS (green in D–F) appear largely unaffected. There is a similar, but less marked reduction in GLAST immunoreactivity(green; GFAP—red) in the infarcted area (arrowheads) (J–L). Scale bars=50 μm, except G–I, 20 μm.

98 B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

2005; Rao et al., 2001a) and GLAST mRNA levels in the gerbilcortex were largely unchanged during the first week of postischemic insult (Fujita et al., 1999). However, Kim et al. (2006)reported a decrease of GLAST protein level in CA1 followingcommon carotid artery occlusion in gerbils 12 h after ische-mia. These variable results could be explained by the expres-sion of splice variants of GLAST that are differentiallyregulated by hypoxia/ischemia. In the pig brain, hypoxia wasfound to cause down-regulation of full length GLAST in

astroglia, but induce expression of an exon 9-skipping form,GLAST1b, missing the N-terminal sequence in neurones: thismay be a more “energy-efficient” form for glial cells recover-ing from an energy depleted state (Sullivan et al., 2007).Under such conditions, a loss, increase or no change inGLAST expression might be observed depending on the spec-ificity of the antibody or mRNA probes used.

EAAC-1 mRNA levels remained unchanged in all the fore-brain areas examined suggesting that this neuronal glutamate

Sham Stroke Left Stroke Right0

5

10

15

20

Var

ico

siti

es p

er 1

00µm

2

Sham Stroke Left Stroke Right0

5

10

15

20

Are

a F

ract

ion

(%

)

Fig. 7 – Quantitative analysis of GLT-1 immunolabelling inthe parietal cortex. The number of GLT-1 immunoreactivepuncta on the left (infarct) side of MCAO brains issignificantly reduced compared to the contralateral side(paired t-test) and to sham-operated brains (mean value ofleft and right sides; unpaired t-test). The percentage ofsampled tissue areas occupied by GLT-1 immunoreactivestructures is similarly reduced compared to the contralateraland sham-operated. *p<0.05; **p<0.01; ***p<0.001.

99B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

transporter might play a less significant role in the ischemiccascade leading to neurodegeneration. A similar conclusionwas drawn previously, since no neurodegeneration was ob-served in EAAC-1 knockout mice (Peghini et al., 1997). In rats,however, up-regulation of immunoreactive EAAC-1, but notGLT-1 or GLAST was reported in both hippocampal CA1 corti-cal pyramidal neurones after 4 vessel occlusion-induced is-chemia (Gottlieb et al., 2000). In contrast, decreased EAAC-1expression was reported in the gerbil hippocampus followingbilateral carotid artery occlusion (Fujita et al., 1999) and therat cortex following MCAO (Rao et al., 2001a). These inconsis-tent results are possibly attributable to the differences in ani-mal species used.

Glutamate transporters are able to transport glutamateagainst a concentration gradient and modulate neurotrans-mission by maintaining low concentrations of extracellularglutamate, thus preventing toxic levels being reached(Anderson and Swanson, 2000). In addition to clearing thereleased glutamate, glutamate transporters are also pro-posed to release intracellular glutamate during ischemicconditions (Phillis et al., 2000; Rossi et al., 2000). However,Roettger and Lipton (1996) failed to observe any effect ofdihydrokainate, a non-transportable blocker of the reverseduptake by GLT-1, on ischemia-induced glutamate releasefrom hippocampal slices suggesting that GLT-1 reversalmay not be the primary mechanism for increased extracellu-lar glutamate concentration.

It should be pointed out that glutamate transporters arenot the only components of the brain's glutamate homeostat-ic system that undergo remodelling following an ischemic in-sult. Importantly, N-methyl-D-aspartate (NMDA) receptorsactivated by excess glutamate released into the intracellularspace and contributing to excitotoxicity, show changes in ex-pression of specific subunits occurring after a few hoursand persisting for several days following transient ischemia(Dos-Anjos et al., 2009a; Hsu et al., 1998; Zhang et al., 1997).Down regulation of α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor subtypes has also beenreported following ischemia/reperfusion (Dos-Anjos et al.,2009b; Gorter et al., 1997; Opitz et al., 2000) and changes inthe expression of key enzymes and proteins associated withthe glutamine–glutamate cycle, such as GS may occur (Leeet al., 2003; 2010). Judging by our immunohistochemical data,a reduction in GS expression did not accompany down regula-tion of GLT-1 in our mouse experimental stroke model. Indrawing conclusions about the role of glutamate transportersin ischemia, stroke and neurodegeneration it is therefore cru-cial to consider observations in the context of the temporalchanges occurring in the mechanisms of glutamate synthesis,release and homeostasis as a whole.

In conclusion, this is the first demonstration of alterationsin the expression of the glutamate transporters in the mouseforebrain following transient focal ischemia induced byMCAO. Furthermore, the results show a good correlationbetween changes in GLT1 mRNA and protein expressions inthe ipsilateral hippocampus and the cortex compared to thesham operated animals. This suggests that during the courseof MCAO, down-regulation of GLT-1 occurs both at the tran-scriptional and protein levels that persists even after 24 h ofreperfusion. This decrease in GLT-1 level may impair normalclearance of synaptically-released glutamate and this, com-bined with changes in glutamate receptor expression andactivation may contribute to ischemic neuronal death. Fur-ther work to establish the functional significance of inducedalterations in GLT-1 expression inmodels of cerebral ischemiain transgenic mice and at different time points of occlusion/reperfusion may help to develop better therapeutic strategiesaimed at reducing glutamate release bymodulating transport-er activity following ischemic events.

4. Experimental procedures

4.1. Animals

All experiments were performed on 10- to 12-week-old (25–30 g) C57BL/6J mice (Harlan-Olac, Bicester, UK) under appro-priate United Kingdom Home Office personal and project li-cences and adhered to regulations as specified in theAnimals (Scientific Procedures) Act (1986) and according to in-stitutional ethical guidelines.

4.2. Cerebral ischemia

Cerebral ischemia was induced by transient (1 h) MCAO as de-scribed previously (McColl et al., 2007). In brief, under isofluor-ane anaesthesia (1.5% in 30% O2/70% N2O), the carotid arteries

100 B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

were exposed and a 6-0 nylon monofilament (Dermalon) with2 mm tip (180 μm diameter) coated in thermo-melting glue(Jet Melt) was introduced into the external carotid artery andadvanced along the internal carotid artery until occluding theorigin of the middle cerebral artery. Core body temperaturewas regulated at 37±0.5 °C throughout the procedure by afeedback-controlled heating blanket. After 1 h, the filamentwas withdrawn to establish reperfusion. Sham-operated miceunderwent the same procedure except the filament was ad-vanced along the internal carotid artery, and then immediatelywithdrawn. Reperfusion was allowed for 24 h, as we have rou-tinely found that this occlusion protocol generates reproduciblestriatal and cortical damage 24 h after MCAO (67±17mm3). Themice were then killed by decapitation under anaesthesia,brains removed, rapidly frozen on dry ice and stored at −80 °Cfor qRT-PCR. For immunohistochemistry, following 24 h surviv-al stroke induced and sham-operated control mice wereperfused transcardially with 0.9% saline followed by 4% para-formaldehyde. Brains were removed, post-fixed, cryoprotectedin 15% sucrose and rapidly frozen on dry ice. Sections (20 μm)were cut on a cryostat (Leica Microsystems, Germany),mounted on gelatinised slides and stored at −20 °C. Tissuesfrom the sham operated animals and the stroke animals wereharvested and processed in parallel and in similar condition.

4.3. Assessment of neurological deficit

Stroke in the MCAO operated animals was confirmed after24 h by neurological deficit tests that were performed usinga neurological grading score of increasing severity of deficit(Bederson et al., 1986); i.e. 0, no observable deficit; 1, torso flex-ion to right; 2, spontaneous circling to right; 3, leaning/fallingto right; 4, no spontaneous movement.

4.4. Tissue preparation for qRT-PCR

Coronal slices of approximately 0.5 mm thickness were cutfrom the forebrain (in dry ice) with a scalpel blade and then tis-sue samples (25–35 mg) were punched from the striatum, hip-pocampus and cerebral cortex (somatosensory/parietal area 1)at the level of the forebrain corresponding to bregma −0.22 to−2.06 mm (Paxinos and Franklin, 2000) with 1 mm (for striatumand hippocampus) and 2 mm (for the cortex) corers under a ×5dissecting microscope. Separate tissue samples were takenfrom the three forebrain regions (Fig. 1) on both sides of thebrain in the stroke and sham-operated brains, the left side ofthe stroke brain being the side on which MCAOwas performed.

4.5. RNA extraction

Tissue samples were homogenised in RNA lysis buffer(Promega) by passing through a 20 gauge blunt needle beforeextracting total RNA using the Promega SV Total RNA Isola-tion System, according to the manufacturer's protocol. Puri-fied RNA (100 μl final volume) was stored at −80 °C.

4.6. Reverse transcription (cDNA synthesis)

Total RNA was reverse-transcribed to first-strand cDNA as de-scribed previously (Saha et al., 2004; Spary et al., 2008). In brief,

oligo (dT) primers (1 μl) and 10 mM dNTPmix (1 μl) were addedto 10 μl of the purified RNA and the mixture incubated at 65 °Cfor 5 min. After cooling on ice, 5× first strand buffer (4 μl),0.1 M DTT (2 μl) and RNaseOUT ribonuclease inhibitor (1 μl)were added. Samples were incubated at 37 °C for 3 min beforeaddition of M-MLV Reverse Transcriptase (1 μl) and further in-cubation at 37 °C for 1 h. After a final incubation at 70 °C for15 min, the cDNA (final volume 20 μl) was stored at −80 °C.All reverse transcription reagents were purchased from Invi-trogen (Paisley, Scotland, UK).

4.7. qRT-PCR

Analysis was performed in triplicate using the Applied Biosys-tems 7500 Real-Time PCR System with intron-spanning mouse-specific GLAST (Mm00600697_m1), GLT-1 (Mm00441457_m1),EAAC-1 (Mm00436590_m1) and β-actin (Mm00607939_s1)primers and Taqman probes (Applied Biosystems), as describedpreviously (Spary et al., 2008; Turner et al., 2007). Glutamatetransporter mRNA expression was calculated as a percentage ofβ-actin mRNA expression using the formula 2−ΔCT×100, inwhich ΔCT was the difference in CT values between each gluta-mate transporter and β-actin mRNA.

4.8. Negative control and internal standard

To avoid amplification of any contaminating genomic DNA,primers were designed to span one or more exon/exon bound-aries (Table 1). We further confirmed lack of genomic DNAcontamination by running negative controls (cDNA synthesisin the absence of reverse transcriptase enzyme), which gaveno amplification.

4.9. Immunohistochemistry

Endogenous peroxidase activity in the slide mounted cryostatsections was quenched with 0.3% H2O2 in ethanol and non-specific protein binding sites were blocked with 10% normalgoat serum (Sigma-Aldrich, Poole, Dorset, UK) diluted in0.1 M phosphate buffered saline pH 7.6 containing 0.1% TritonX-100 (PBS-T buffer). Sections were then incubated overnightat 4 °C in primary antibodies against GLAST (NC Danbolt,code number A522) or GLT-1 (NC Danbolt, code number B12),diluted in PBS-T buffer to protein concentrations of 0.28 μg/ml(GLT-1) and 0.6 μg/ml (GLAST).

Following three washes in PBS, the sections were incubat-ed for 2 h in species specific anti-rabbit immunoglobulin G(IgG) conjugated to Cy3 (Jackson Immunoresearch/StratechScientific, Soham, Cambs, UK), diluted 1/1000 in PBS-T buffer.Sections were then washed thoroughly in PBS, dried at 4 °C,and mounted under coverslips in PBS-glycerol. The primaryantibody incubation step was omitted for negative controlsections.

4.10. Dual fluorescence immunolabelling

For dual labelling, forebrain sections were incubated over-night at 4 °C in the primary antibody to GLT-1 or GLAST,raised in rabbit or guinea pig (Chemicon; dilution 1/1000)mixed with mouse antiserum to GFAP (Sigma, 1/1000) or

Table 1 – The details of real-time PCR primers.

Primers Assay name NCBI ref seq Position Exon boundary Size (bp) Supplier

GLAST Mm00600697_m1 NM_148938 810 3–4 71 Applied BiosystemsGLT-1 Mm00441457_m1 NM_011393 585 4–5 64 Applied BiosystemsEAAC-1 Mm00436590_m1 NM_009199 1401 11–12 66 Applied BiosystemsΒ-actin Mm00607939_s1 NM_007393.3 1226 6 115 Applied Biosystems

101B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

rabbit antiserum to GS (Sigma, 1/1000). To visualise thebound primary antibodies in the dual labelling prepara-tions, sections were incubated in a mixture of species-specific secondary antibodies raised against rabbit,guinea-pig or mouse immunoglobulins (as appropriate)conjugated to Cy3 (Jackson Immunoresearch) and Alexa488

(Molecular Probes BV, Leiden, Netherlands), each diluted1/500–1/1000 in PBS-T as described previously (Saha et al.,2005). Sections were then washed thoroughly in PBS, ar-ranged on gelatinised slides, dried at 4 °C, and mountedunder coverslips in PBS-glycerol.

4.10.1. Antibody specificityAntibodies against GLAST (NC Danbolt, code number A522) orGLT-1 (NC Danbolt, code number B12) were prepared byimmunising rabbits with C terminal peptide A522–541 (PYQ-LIAQDNEPEKPVADSET) or N-terminal peptide 12–26(KVEVRMHDSHLSSE), respectively. These antibodies wereraised, purified and extensively characterised as previouslydescribed (Lehre et al., 1995; Lehre and Danbolt, 1998).

4.11. Imaging and quantitative analysis

Sections were viewed on an AxioImager Z.1 epifluorescencemicroscope (Carl Zeiss, Welwyn Garden City, UK). A #10 filterset with excitation wavelength of 450–490 nm was used to vi-sualise Alexa488 and a #20 set (534–558 nm excitation) wasused to visualise Cy3. Fluorescence images were capturedusing a Zeiss AxioVision Imaging System, which included‘Apotome’ confocal imaging and ‘Colocalisation’ modules.All images were imported into Adobe Photoshop 7.0 forminor adjustment of brightness and contrast, resizing or crop-ping and assembling into figures. After lettering, the layerswere merged, resolution was adjusted to 500 dpi and the im-ages were saved as TIFF files.

Quantitative analysis was performed on sections fromMCAO and sham-operated brains (n=4) labelled for GLT-1using Cy3 as described above. Three×40 images were cap-tured from randomly selected areas of the somatosensory(parietal) cortex, layers III–V at levels between bregma −1.0and −1.3 on both sides of each brain, using fixed camerasettings. The resulting 8-bit grey scale JPEG images wereopened in ImageJ (http://rsbweb.nih.gov/ij/), thresholded withconstant pixel intensity level settings (85) to a white back-ground and then analysedwith the ‘Analyze Particles’ function.

4.12. Statistical analysis

Differences between sham and stroke (left sides) were ana-lysed by unpaired t-tests as they were derived from differentanimals. Differences between left and right sides of stroke

brains were analysed by paired t-tests as they were from thesame animals (Graph Pad Prism software, www.graphpad.com). Similar results were obtained when paired or unpairedANOVA was used, but the mixed nature of the data precludeda single ANOVA calculation and hence t-test values are pre-sented. p≤0.05 was considered statistically significant.

Disclosure/conflict of interest

There is no conflict of interest.

Acknowledgments

We acknowledge the financial support of the Yorkshire Alz-heimer's Research Trust Network, and thank Jean Kaye forher expert technical assistance. Antibodies to GLT-1 andGLAST were kindly provided by Dr. N.C. Danbolt (Universityof Oslo, Norway).

R E F E R E N C E S

Anderson, C.M., Swanson, R.A., 2000. Astrocyte glutamate transport:review of properties, regulation, and physiological functions.Glia 32, 1–14.

Beart, P.M., O'Shea, R.D., 2007. Transporters for L-glutamate: anupdate on their molecular pharmacology and pathologicalinvolvement. Br. J. Pharmacol. 150, 5–17.

Bederson, J.B., Pitts, L.H., Tsuji, M., Nishimura, M.C., Davis, R.L.,Bartkowski, H., 1986. Rat middle cerebral artery occlusion:evaluation of the model and development of a neurologicalexamination. Stroke 17, 472–476.

Butler, T.L., Kassed, C.A., Sanberg, P.R., Willing, A.E., Pennypacker,K.R., 2002. Neurodegeneration in the rat hippocampus andstriatum after middle cerebral artery occlusion. Brain Res. 929,252–260.

Bruhn, T., Levy, L.M., Nielson, M., Christensen, T., Johansen, F.F.,Diemer, N.H., 2000. Ischaemia induced changes in expressionof the astrocyte glutamate transporter GLT-1 in hippocampusof the rat. Neurochem. Int. 37, 277–285.

Camacho, A., Massieu, L., 2006. Role of glutamate transporters inthe clearance and release of glutamate during ischemia and itsrelation to neuronal death. Arch. Med. Res. 37, 11–18.

Chen, J.C., Hsu-Chou, H., Lu, J.L., Chiang, Y.C., Huang, H.M., Wang,H.L., Wu, T., Liao, J.J., Yeh, T.S., 2005. Down-regulation of theglial glutamate transporter GLT1 in rat hippocampus andstriatum and its modulation by a group III metabotropicglutamate receptor antagonist following transient globalforebrain ischemia. Neuropharmacology 49, 703–714.

Danbolt, N.C., 2001. Glutamate uptake. Prog. Neurobiol. 65, 1–105.Dos-Anjos, S., Martínez-Villayandre, B., Montori, S., Regueiro-

Purriños, M.M., Gonzalo-Orden, J.M., Fernández-López, A.,

102 B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

2009a. Transient global ischemia in rat brain promotesdifferent NMDA receptor regulation depending on the brainstructure studied. Neurochem. Int. 54, 180–185.

Dos-Anjos, S., Martínez-Villayandre, B., Montori, S., Regueiro-Purriños, M.M., Gonzalo-Orden, J.M., Fernández-López, A.,2009b. Global ischemia-induced modifications in theexpression of AMPA receptors and inflammation in ratbrain. Brain Res. 1287, 20–27.

Fujita, H., Sato, K., Wen, T., Peng, Y., Sakanaka, M., 1999.Differential expression of glycine transporter 1 and threeglutamate transporter mRNA in the hippocampus of gerbilsafter transient forebrain ischaemia. J. Cereb. Blood FlowMetab. 19, 604–615.

Giffard, R.G., Swanson, R.A., 2005. Ischemia-induced programmedcell death in astrocytes. Glia 50, 299–306.

Gorter, J.A., Petrozzino, J.J., Aronica, E.M., Rosenbaum, D.M., Opitz,T., Bennett, M.V., Connor, J.A., Zukin, R.S., 1997. Global ischemiainduces downregulation of Glur2 mRNA and increases AMPAreceptor-mediated Ca2+ influx in hippocampal CA1 neuronsof gerbil. J. Neurosci. 17, 6179–6188.

Gottlieb, M., Domercq, M., Matute, C., 2000. Altered expression ofthe glutamate transporter EAAC-1 in neurons and immatureoligodendrocytes after transient forebrain ischemia. J. Cereb.Blood Flow Metab. 20, 678–687.

Hsu, J.C., Zhang, Y., Takagi, N., Gurd, J.W., Wallace, M.C., Zhang, L.,Eubanks, J.H., 1998. Decreased expression and functionality ofNMDA receptor complexes persist in the CA1, but not in thedentate gyrus after transient cerebral ischemia. J. Cereb. BloodFlow Metab. 18, 768–775.

Johnston, M.V., 2005. Excitotoxicity in perinatal brain injury. BrainPathol. 15, 234–240.

Kim, D.S., Kwak, S.E., Kim, J.E., Jung, J.Y., Won, M.H., Choi, S.Y.,Kwon, O.S., Kang, T.C., 2006. Transient ischaemia affectsplasma membrane glutamate transporter, not vesicularglutamate transporter, expressions in the gerbil hippocampus.Anat. Histol. Embryol. 35, 265–270.

Kubista, M., Andrade, J.M., Bengtsson, M., Forootan, A., Jonák, J.,Lind, K., Sindelka, R., Sjöback, R., Sjögreen, B., Strömbom, L.,Ståhlberg, A., Zoric, N., 2006. The real-time polymerase chainreaction. Mol. Aspects Med. 27, 96–125.

Lee, A., Lingwood, B.E., Bjorkman, S.T., Miller, S.M., Poronnik, P.,Barnett, N.L., Colditz, P., Pow, D.V., 2010. Rapid loss of glutaminesynthetase from astrocytes in response to hypoxia: implicationsfor excitotoxicity. J. Chem. Neuroanat. 39, 211–220.

Lee, D.R., Helps, S.C., Gibbins, T.L., Nilsson, M., Sims, N.R., 2003.Losses of NG2 and NeuN immunoreactivity but not astrocyticmarkers during early reperfusion following severe focalcerebral ischemia. Brain Res. 989, 221–230.

Lehre, K.P., Levy, L.M., Ottersen, O.P., Storm-Mathisen, J., Danbolt,N.C., 1995. Differential expression of two glial glutamatetransporters in the rat brain: quantitative andimmunocytochemical observations. J. Neurosci. 15, 1835–1853.

Lehre, K.P., Danbolt, N.C., 1998. The number of glutamatetransporter subtype molecules at glutamatergic synapses:chemical and stereological quantification in young adult ratbrain. J. Neurosci. 18, 8751–8757.

McColl, B.W., Graham, D.I., Weir, C.J., White, F., Horsburgh, K.,2003. Endocytic pathway alterations in human hippocampusafter global ischemia and the influence of APOE genotype. Am.J. Pathol. 162, 273–281.

McColl, B.W., Carswell, H., McCulloch, J., Horsburgh, K., 2004.Extension of cerebral hypoperfusion and ischaemic pathologybeyond MCA territory after intraluminal filament occlusion inC57B1/6J mice. Brain Res. 997, 15–23.

McColl, B.W., Rothwell, N.J., Allan, S.M., 2007. Systemic inflammatorystimulus potentiates the acute phase and CXC chemokineresponses to experimental stroke and exacerbates brain damagevia interleukin-1- and neutrophil-dependent mechanisms.J. Neurosci. 27, 4403–4412.

Opitz, T., Grooms, S.Y., Bennett, M.V., Zukin, R.S., 2000. Remodelingof alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acidreceptor subunit composition in hippocampal neurons afterglobal ischemia. Proc. Natl. Acad. Sci. U.S.A. 97, 13360–13365.

Paxinos, G., Franklin, K.B.J., 2000. The Mouse Brain in StereotaxicCoordinates (2nd edition). Academic Press, San Diego.

Peghini, P., Janzen, J., Stoffel, W., 1997. Glutamate transporterEAAC-1-deficient mice develop dicarboxylic aminoaciduriaand behavioral abnormalities but no neurodegeneration.EMBO J. 16, 3822–3832.

Phillis, J.W., Ren, J., O'Regan, M.H., 2000. Transporter reversalas a mechanism of glutamate release from the ischemic ratcerebral cortex: studies with DL-threo-B benzyloxyasparte.Brain Res. 868, 105–112.

Rao, V.L.R., Rao, A.M., Dogna, A., Bowen, K.K., Hatcher, J.,Rothstein, J.D., Dempsey, R.J., 2000. Glial glutamate transporterGLT-1 downregulation precedes delayed neuronal deathin gerbil hippocampus following transient global cerebralischaemia. Neurochem. Int. 36, 531–537.

Rao, V.L.R., Bowen, K.K., Dempsey, R.J., 2001a. Transient focalcerebral ischemia down-regulates glutamate transportersGLT-1 and EAAC-1 expression in rat brain. Neurochem. Res.26, 497–502.

Rao, V.L.R., Dogan, A., Bowen, K.K., Todd, K.G., Dempsey, R.J.,2001b. Antisense knockdown of the glial glutamate transporterGLT-1 exacerbates hippocampal neuronal damage followingtraumatic injury to rat brain. Eur. J. Neurosci. 13, 119–128.

Rao, V.L.R., Dogan, A., Todd, K.G., Bowen, K.K., Kim, B.T.,Rothstein, J.D., Dempsey, R.J., 2001c. Antisense knockdownof the glial glutamate transporter GLT-1, but not the neuronalglutamate transporter EAAC-1, exacerbates transient focalcerebral ischemia-induced neuronal damage in rat brain.J. Neurosci. 21, 876–883.

Roettger, V., Lipton, P., 1996. Mechanism of glutamate releasefrom rat hippocampal slices during in vitro ischemia.Neuroscience 75, 677–685.

Rossi, D.J., Oshima, T., Attwell, D., 2000. Glutamate release insevere brain ischaemia is mainly by reversed uptake. Nature403, 316–321.

Saha, S., Drinkhill, M.J., Moore, J.P., Batten, T.F.C., 2005. Centralnucleus of amygdala projections to rostral ventrolateralmedulla neurones activated by decreased blood pressure Euro.J. Neurosci. 21, 1921–1930.

Saha, S., Spary, E.J., Maqbool, A., Corbett, E.K.A., Asipu, A.,Corbett, E.K., Batten, T.F.C., 2004. Increased expression ofglutamate AMPA receptor subunits in the nucleus of thesolitary tract in the spontaneously hypertensive rat. Mol.Brain Res. 121, 37–49.

Sims, K.D., Robinson, M.B., 1999. Expression patterns andregulation of glutamate transporters in the developing andadult nervous system. Crit. Rev. Neurobiol. 13, 169–197.

Spary, E.J., Maqbool, A., Saha, S., Batten, T.F.C., 2008. IncreasedGABAB receptor subtype expression in the nucleus of thesolitary tract of the spontaneously hypertensive rat. J. Mol.Neurosci. 35, 211–224.

Sullivan, S.M., Björkman, S.T., Miller, S.M., Colditz, P.B., Pow,D.V., 2010. Structural remodeling of gray matter astrocytesin the neonatal pig brain after hypoxia/ischemia. Glia 58,181–194.

Sullivan, S.M., Macnab, L.T., Björkman, S.T., Colditz, P.B., Pow,D.V., 2007. GLAST1b, the exon-9 skipping form of theglutamate-aspartate transporter EAAT1 is a sensitive markerof neuronal dysfunction in the hypoxic brain. Neuroscience149, 434–445.

Susarla, B.T., Robinson, M.B., 2003. Rottlerin, an inhibitor ofprotein kinase Cdelta (PKCdelta), inhibits astrocytic glutamatetransport activity and reduces GLAST immunoreactivity bya mechanism that appears to be PKCdelta-independent.J. Neurochem. 86, 635–645.

103B R A I N R E S E A R C H 1 4 1 8 ( 2 0 1 1 ) 9 3 – 1 0 3

Tanaka, K., Watase, K., Manabe, T., Yamada, K., Watanabe, M.,Takahashi, K., Iwama, H., Nishikawa, T., Ichihara, N., Kikuchi,T., Okuyama, S., Kawashima, N., Hori, S., Takimoto, M., Wada,K., 1997. Epilepsy and exacerbation of brain injury in micelacking the glutamate transporter GLT-1. Science 276,1699–1702.

Torp, R., Lekieffre, D., Levy, L.M., Haug, F.M., Danbolt, N.C.,Meldrum, B.S., Ottersen, O.P., 1995. Reduced postischemicexpression of a glial glutamate transporter, GLT1, in the rathippocampus. Exp. Brain Res. 103, 51–58.

Turner, N.A., Mughal, R.S., Warburton, P., O'Regan, D.J., Ball, S.G.,Porter, K.E., 2007. Mechanism of TNFα-induced IL-1α, IL-1β andIL-6 expression in human cardiac fibroblasts: effects of statinsand thiazolidinediones. Cardiovasc. Res. 76, 81–90.

Valasek, M.A., Repa, J.J., 2005. The power of real-time PCR. Adv.Physiol. Educ. 29, 151–159.

Yi, J., Pow, D.V., Hazell, A.S., 2005. Early loss of the glutamatetransporter splice variant GLT-1v in rat cerebral cortexfollowing lateral fluid-percussion injury. Glia 49, 121–133.

Zelenaia, O.A., Robinson,M.B., 2000. Degradation of glial glutamatetransporter mRNAs is selectively blocked by inhibition ofcellular transcription. J. Neurochem. 75, 2252–2258.

Zhang, L., Hsu, J.C., Takagi, N., Gurd, J.W., Wallace, M.C., Eubanks,J.H., 1997. Transient global ischemia alters NMDA receptorexpression in rat hippocampus: correlation with decreasedimmunoreactive protein levels of the NR2A/2B subunits, andan altered NMDA receptor functionality. J. Neurochem. 69,1983–1994.