www.elsevier.com/locate/expgero

Experimental Gerontology 42 (2007) 343–354

Delayed neurodegeneration and early astrogliosis after excitotoxicityto the aged brain

M.M. Castillo-Ruiz *, O. Campuzano, L. Acarin *, B. Castellano, B. Gonzalez

Department of Cell Biology, Physiology and Immunology, Unit of Medical Histology, School of Medicine, and Institute of Neurosciences,

Autonomous University Barcelona, Bellaterra 08193, Spain

Received 12 May 2006; received in revised form 14 September 2006; accepted 10 October 2006Available online 28 November 2006

Abstract

Excitotoxicity is well recognised as a mechanism underlying neuronal cell death in several brain injuries. To investigate age-dependentdifferences in neurodegeneration, edema formation and astrogliosis, intrastriatal N-methyl-D-aspartate injections were performed inyoung (3 months) and aged (22–24 months) male Wistar rats. Animals were sacrificed at different times between 12 h and 14 dayspost-lesion (DPL) and cryostat sections were processed for Toluidine blue, Fluoro-Jade B staining, NeuN and GFAP immunohistochem-istry. Our results show that both size of tissue injury and edema were reduced in the old subjects only up to 1DPL, correlating with aslower progression of neurodegeneration with peak numbers of degenerating neurons at 3DPL in the aged, contrasting with maximumneurodegeneration at 1DPL in the young. However, old animals showed an earlier onset of astroglial response, seen at 1DPL, and alarger area of astrogliosis at all time-points studied, including a greater glial scar. In conclusion, after excitotoxic striatal damage, pro-gression of neurodegeneration is delayed in the aged but the astroglial response is earlier and exacerbated. Our results emphasize theimportance of using aged animals and several survival times for the study of acute age-related brain insults.� 2006 Elsevier Inc. All rights reserved.

Keywords: Aging; Brain; Neurodegeneration; Fluoro-Jade B; Edema; Astrogliosis; Experimental stroke; Excitotoxicity; Striatum

1. Introduction

Normal brain aging is characterised by macroscopicchanges including decreased brain weight and volume,and microscopic alterations such as changes in vasculature,blood–brain barrier permeability, modifications in theextracellular compartment, and both cellular and biochem-ical alterations in glia and neuronal cells (Scahill et al.,2003; Shah and Mooradian, 1997). Glial cells show activat-ed/reactive phenotypes with changes in cell morphologyand metabolism and in some cases also an increase in cellnumbers (Amenta et al., 1998; Peinado, 1998; Unger,1998). Neuronal cells typically accumulate intracellularmacromolecules like lipofuscin, which, together with thewell known aging-induced increase in oxidative stress and

0531-5565/$ - see front matter � 2006 Elsevier Inc. All rights reserved.

doi:10.1016/j.exger.2006.10.008

* Corresponding authors. Tel.: +34 935811826; fax: +34 935812392.E-mail addresses: mariadelmar.castillo@uab.es (M.M. Castillo-Ruiz),

Laia.acarin@uab.es (L. Acarin).

an important decline in neurotransmission, induces a pro-gressive loss of neuronal function (Finch, 2003; Finkeland Holbrook, 2000; Floyd and Hensley, 2002; Segoviaet al., 2001). Undoubtedly, these changes influence the sus-ceptibility and response of the aging brain to damage andmay explain why aging is the major risk factor for higherincidence of brain injuries and worse outcome after aninsult (Brown et al., 2003; Bruns and Hauser, 2003; Nakay-ama et al., 1994; Wang et al., 2003; Woo et al., 1992). How-ever, it should be noted that most of the availableexperimental work use young animals to study acuteCNS insults which are epidemiologically associated withaging, like brain ischemia or stroke (Bonita et al., 1997;Truelsen and Bonita, 2003).

A common mechanism underlying neuronal cell death insuch acute injuries is excitotoxicity (Dirnagl et al., 1999;Doble, 1999; Mattson, 2003), a phenomenon that takesplace when excessive activation of glutamate receptorsleads to a series of intracellular processes inducing cell

344 M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354

death (Mattson, 2003; Olney, 1969). Of all glutamate recep-tors, N-methyl-D-aspartate (NMDA) receptors are consid-ered the main leaders of these processes due to their highcalcium permeability, as well as its induction of calciummobilization from intracellular compartments (Mody andMacDonald, 1995). As such, overactivation of NMDAreceptors by exogenous application of agonists has largelybeen used as a model of experimental brain injury in thepostnatal (Acarin et al., 2001), adult (Dietrich et al.,1992; Kollegger et al., 1993; Stewart et al., 1986) and alsothe aged brain (Suzuki et al., 2003).

Moreover, it is now well established that besides pri-mary neuronal cell death, inflammatory, toxic and protec-tive molecules produced by surrounding activated glial cellsdetermine the final outcome of brain damage. Reactiveastrocytes play many important roles, both detrimentaland beneficial for neuron survival and are the main compo-nents of the glial scar (Anderson et al., 2003; Nedergaardand Dirnagl, 2005; Pekny and Nilsson, 2005), which pro-vides structural support to injured tissue, but at the sametime it is extremely inhibitory for axonal regeneration(Bovolenta et al., 1993). Nevertheless, besides these generalfeatures of astrogliosis, the intensity and temporal patternof the astroglial response observed after brain injury canvary depending on the type of stimulus, the CNS regionand age of the animals (Acarin et al., 1999; Raivichet al., 1999; Stoll et al., 1998). In the normal aging brain,up to date, the most outstanding and characteristic changedescribed in astrocytes is cell hypertrophy associated with asignificant increase in the expression of the intermediatefilament protein glial fibrillary acidic protein (GFAP)(Bronson et al., 1993; Finch, 2003; Landfield et al., 1977),changes that are commonly found in reactive gliosis inthe adult brain (Eng et al., 2000) and that will obviouslyinfluence the astroglial response in the lesioned aging brain.Despite this fact, again most of the studies on the astroglialresponse are based on the use of young adult animals andlittle and contradictory data is available on the extent andtemporal pattern of reactive gliosis in the lesioned agingbrain.

In this context, based on the hypothesis that neurode-generation and the glial response may differ after damageto the aged CNS, in comparison to the young adult brain,the aim of the present study was to investigate, after anexcitotoxic lesion, the age-dependent alterations in theextent and the temporal profile of neurodegeneration,brain edema and astroglial reactivity, three importantparameters that determine final lesion outcome. For thispurpose, brain damage response was followed from 12 huntil 14 days post-lesion in young and aged rats.

2. Materials and methods

2.1. Study population

Experimental animal work was conducted using maleWistar rats of two different age groups: 46 young animals

(3- to 4-month-old) and 51 old animals (22- to 24-month-old) were used. The subjects were distributed in threegroups which will be referred as: intact controls, saline-in-jected controls and NMDA-injected lesioned animals. Ani-mals were housed individually in cages with free food andwater supply, in an environment with controlled tempera-ture (22 �C), humidity (55%) and light/dark cycle (12 h).All experimental work was conducted according to estab-lished European Union bioethic directives and wasapproved by the ethical commission of the UniversitatAutonoma de Barcelona. During all the process unneces-sary animal suffering was avoided.

2.2. Excitotoxic lesions

Under isofluorane gas anesthesia, rats were placed in aKopf stereotaxic frame. A hole in the skull was openedwith a drill, and 120 nmol of NMDA (Sigma, M-3262,St. Louis, USA), diluted in 1 ll of saline solution(0.9% NaCl) were injected into the striatum (caudate–pu-tamen, coordinates A = +0.12, L = �0.3; V = �0.45 cm)of the right hemisphere, with a Hamilton microsyringecoupled to an automatic microinjector (Stoelting, Illinois,USA) at a speed of 0.2 ll/min. After the injection, theneedle was kept 10 min in order to facilitate diffusioninto the striatum and minimize reflux. Finally, operatedanimals were maintained at normothermia by means ofa thermal pad inside an incubator until they recuperatedfrom anesthesia. Saline-injected controls followed thesame procedure but were injected with 1 ll of saline solu-tion. The groups of intact control animals did not receiveany injection.

2.3. Sacrifice and histologic processing

After the survival times of 12 h post-lesion (HPL) 1day, 3 days, 5 days, 7 days, and 14 days post-lesion(DPL), animals were anaesthetised and intracardially per-fused for 10 min with 4% paraformaldehyde in 0.1 Mphosphate buffer (pH 7.4). At least 5 NMDA-injectedlesioned animals and 2 saline controls per age and sur-vival time were sacrificed, in addition to 4 intact controladult animals and 8 intact control aged animals. Afterperfusion, brains were quickly removed, postfixed for4 h at 4 �C in the same fixative and immersed for cryo-protection in a 30% sacarose solution in 0.1 M phosphatebuffer (pH 7.4) until freezing. For each animal, 15 seriesof 30 lm thick parallel frozen coronal sections (thus,each section 450 lm apart from adjacent ones) wereobtained with a Leitz cryostat, and either stored free-floating in antifreeze solution at �20 �C or mounted ongelatin-coated slides and stained with toluidine blue.After washing, slides were dehydrated in alcohol, clearedin xylene, and coverslipped with DPX. These slides wereused for the histologic control of the intracerebral injec-tion, the study of the size of injury and ipsilateral hemi-spheric swelling.

M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354 345

2.4. Fluoro-Jade B staining

For the study of degenerating neurons, sections werestained with Fluoro-Jade B as described (Schmued andHopkins, 2000). Briefly, free-floating sections were mount-ed and air dried overnight. After dehydration and rehydra-tion in graded ethanol, sections were rinsed in water andoxidized with 0.06% potassium permanganate (MnO4K inwater) for 15 min. Next, sections were rinsed with distilledwater and immersed in 0.0004% Fluoro-Jade B (1FJ-B,Histo-Chem Inc., Jefferson USA) + 1% glacial acetic acidfor 20 min. After rinsing, slides were air dried, cleared inxylene and coverslipped in DPX.

2.5. Immunohistochemistry

For the visualization of astrocytes, parallel free floatingsections were processed for the immunohistochemical dem-onstration of GFAP and for neuronal labelling sectionswere immunohistochemically processed for the Neuronalnuclear antigen (NeuN). In both cases, sections werewashed in tris buffered saline (TBS) 0.05 M, pH 7.4, andafter endogenous peroxidase blocking (10 min with 2%H2O2 in 70% methanol), sections were washed in TBS, inTBS + 1% Triton X-100 and treated with blocking buffer(BB) (10% foetal calf serum in TBS + 1% Triton X-100)for 30 min. Afterwards, sections were incubated overnightat 4 �C and 1 h at room temperature, with either a poly-clonal rabbit anti-cow GFAP antibody (DakopattsZ-0334, Dako A/S, Glostrup, Denmark) (1:1800 in BB);or a monoclonal mouse anti- NeuN (Chemicon Interna-tional, MAB 377, Temecula, CA, USA) (1:1000 in BB).After the incubation, sections were rinsed in TBS + 1% Tri-ton X-100 and incubated for one hour at room temperaturewith either biotinylated secondary antibodies: anti-rabbitIg (Amersham Biosciences RPN1004V1, UK) (diluted1:200 in BB); or anti-mouse Ig (Amersham BiosciencesRPN1001V1, UK) (diluted 1:200 in BB). Afterwards,sections were rinsed again and incubated for 1 h at roomtemperature with avidin-peroxidase (Dakopatts P0364,Dako A/S, Denmark) diluted 1:400 in BB and washed inTBS and in tris buffer (TB). Peroxidase reaction productwas visualized with 100 ml of TB containing 50 mg of3,3 0-diaminobenzidine (DAB) and 33 ll of hydrogen per-oxide for 4–5 min. Finally sections were rinsed in TB,mounted on gelatine-coated slides, dehydrated, cleared inxylene, and coverslipped in DPX. As negative controlsthe primary antibodies were omitted.

2.6. Quantification: general considerations

Sections from intact controls, NMDA injected and sa-line-injected control animals were studied through a NikonEclipse E 600 microscope interfaced to a DMX 1200 cam-era and a PC. Pictures of representative areas were taken atdifferent magnifications using the software ACT-1 2.20(Nikon Corporation). A minimum of 4 NMDA-injected

animals for each survival time and age group were used.Researchers were blinded to sample identity. For each ani-mal and staining procedure, 4 sections in the injury corecomprised between 1 mm/�0.3 mm from bregma were ana-lysed. Anatomical landmarks (aspect, size and position ofthe anterior commissures, corpus callosum, septum, lateralventricles, striatum, nucleus accumbens) were used toensure that parameters were analysed at similar levels with-in and between groups.

2.6.1. Quantification of neurodegenerative area, brain edema

and astroglial reactivity

Selected brain sections were scanned with a Nikon LS-1000 assembled to a Macintosh computer. National Insti-tute of Health Image software (N.I.H. 1.62) was used toobtain the following area values in mm2: (i) contralateralhemisphere (CLH); (ii) contralateral ventricle (CLV); (iii)corrected contralateral hemisphere (cCLH) obtained asCLH-CLV; (iv) ipsilateral hemisphere (ILH); (v) ipsilateralventricle (ILV); (vi) corrected ipsilateral hemisphere (cILH)obtained as ILH-ILV; (vii) area of tissue injury in the ipsi-lateral hemisphere, identified in toluidine blue processedsections in basis of its decreased staining; and (viii) areaof astroglial reactivity in the ipsilateral hemisphere, mea-sured in sections processed for GFAP, as the area showinghigher staining than the contralateral hemisphere of thesame section.

In order to avoid misinterpretation of data due to thesignificantly smaller hemisphere size in the old control sub-jects (see Section 3), results are shown as percentage oftotal area in the ipsilateral hemisphere: (i) % of tissuedegeneration = (area of tissue degeneration/cILH) · 100;and (ii) % of astroglial reactivity = (area of astroglial reac-tivity/cILH) · 100. As previously described (Fujimuraet al., 2001), brain edema was calculated indirectly fromvalues of the ipsilateral hemisphere as follows: % of hemi-spheric swelling (brain edema) = (cILH/cCLH) · 100.Accordingly, positive results were considered indicatorsof brain edema, whereas negative results would point tohemispheric collapse or atrophy.

2.6.2. Quantitative study of neurodegeneration

In order to avoid misinterpretation of data due to pos-sible reduction in striatal neuronal cell numbers with aging(see Section 3), NeuN+ profiles were assessed in controlaged and young striatum. Additionally, to address the per-centage of remaining neurons, NeuN+ profiles wereassessed in the 14DPL group. Fluoro-Jade B+ profileswere assessed in all control and injured animals. Regionof interest (striatum) was delimited by the lateral ventriclemedially, the corpus callosum dorsally and laterally, andby a line drawn between the two anterior commissures ven-trally. For each section, 4 pictures were taken at 10·, com-prising the whole striatal area, and all profiles clearlyidentified as specifically labeled neurons were manuallycounted. The number of Fluoro-Jade B+ neurodegenerat-ing profiles and NeuN+ remaining profiles were expressed

346 M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354

in relation to age-matched control striatal NeuN+ neuro-nal profiles as follows: (i) % of degenerating neu-rons = (FJ-B+ profiles/NeuN+ profiles in age-matchedcontrol animals) · 100; (ii) % of remainingneurons = (NeuN+ profiles 14DPL/NeuN+ profiles inage-matched control animals) · 100.

2.7. Statistical analysis

Statistical analysis was performed with SPSS 11.01. Stri-atal neuronal cell numbers, hemispheric area and ventriclesize in control animals were studied by means of one-wayANOVA followed by Scheffe post hoc comparison. Tissueinjury, neurodegeneration, edema and astroglial reactivitypercentage results were studied by means of two-wayANOVA. All results are presented as mean values ± SEM,p 6 0.05 was considered significant and p 6 0.10 was con-sidered marginally significant.

3. Results

3.1. Normal brain aging

In control animals, the quantitative analysis showed a17% reduction in cerebral hemispheric area in aged animalswhen compared to young ones (Fig. 1A), but no changes inthe percentage of this hemispheric area occupied by thelateral ventricles (Fig. 1B). At the microscopic level,although aged animals showed no apparent differences inneuronal morphology, as observed with toluidine bluestaining and NeuN immunolabeling (Fig. 1D–G), a 13%reduction in NeuN+ profiles was seen in the striatum ofaged animals (Fig. 1C). In agreement, sparse Fluoro-JadeB+ staining found in the aged striatum, whereas no degen-erating neurons were found in the young (data not shown).In addition, qualitatively thicker white matter bundles wereseen in the striatum of aged brains, especially in the dorso-lateral striatum and surrounding the lateral ventricles(Fig. 1F and G). In regards to astroglial cells, no remark-able differences in the pattern of GFAP immunoreactivitybetween young and aged control animals were observed.As expected, both groups presented stronger immunoreac-tive astrocytes in white matter tracts and in cortical layer Iforming the glia limitans (Fig. 1H and I). However, ahigher variability in GFAP staining was observed in thegroup of aged animals, with brains presenting more immu-noreactive astrocytes in cortex and striatum.

3.2. Tissue injury, edema and neurodegeneration

Control saline-injected aged and young animals showeda mild cortical lesion in the frontal cortex. In the striatum,injured tissue was only observed around the needle track,comprising a very small region, where a few scatteredFluoro-Jade B+ degenerating neurons were seen, similarin both age groups (data not shown). In contrast, in bothyoung and aged rats, the intrastriatal injection of NMDA

induced tissue degeneration mainly in the striatum (cau-date–putamen) but also in the cortex and the corpus callo-sum (Fig. 2A). The cortical lesion affected mainly thefrontal cortex (areas 1 and 2), and in the most injured ani-mals, it expanded caudally affecting part of the forelimbsensorimotor area, the cingulate and the parietal cortex.Quantitative analysis of degenerating tissue area(Fig. 2B), showed strong differences between ages andpost-lesion times. Injury size was significantly larger inyoung rats at 12HPL and 1DPL but there were no signifi-cant differences between age groups at later time points.Overall, in aged animals, evolution of the degeneratingarea was more regular, varying from 15.5 ± 3.6% to7.50 ± 1.2% of ipsilateral hemisphere, whereas in the youngit changed from 27.7 ± 4.5% at 1DPL to 2.6 ± 0.6% of ipsi-lateral hemisphere at the last time studied.

Brain edema, shown as the percentage of ipsilateralhemispheric swelling, was marginally significantly biggerin young subjects only at 1DPL (Fig. 2C). As observedwith injury size, no differences between age groups werefound at later time points, although differences betweensurvival times were observed at both ages. Whereas inyoung subjects brain edema increased between 12HPLand 1DPL and diminished between 3 and 5DPL, this pat-tern was delayed in the aged animals, where brain edemaslightly increased between 1 and 3DPL and hemisphericswelling was reduced between 5 and 7DPL (Fig. 2C). After7DPL brain edema was not observed at any age group.Instead, in the aged group a tendency to hemisphere col-lapse was observed, giving negative values both at 7 and14DPL.

The temporal analysis of neurodegeneration within thelesioned striatum was carried out by Fluoro-Jade B label-ling, a marker for degenerating neurons. Fluoro-Jade Bwas observed in the soma of cells with the morphology ofstriatal neurons, in both young and aged rats and at alltime points studied (Fig. 3). Although Fluoro-Jade B label-ling was observed in the whole striatal injured area at bothages, the higher number of stained cells as well as the stron-gest staining intensity were always found at the lesion core,especially in aged animals. Towards the periphery of thelesion, cellular labelling decreased in amount and intensity.The quantitative analysis of Fluoro-Jade B+ stainingshowed that the temporal dynamics of neurodegenerationwere different between young and aged animals (Fig. 3I).Although young rats presented a slightly higher percentageof degenerating neurons than the aged at 12HPL, greatestdifferences were seen in the time of maximum neurodegen-eration. Whereas young rats reached a peak percentage ofFluoro-Jade B+ labeling by 1DPL, in the aged, neurode-generation developed to a maximum percentage of degen-erating neurons at 3DPL. At 5DPL both ages showedsimilar percentages of neurodegeneration that were not dif-ferent from 12HPL. It should be noted that aged animalsshowed Fluoro-Jade B labelling in white matter bundlesat 7 and 14DPL, hindering a reliable counting at the latesttime points. Finally, no changes in the percentage of

Fig. 1. Quantitative and qualitative analysis of different parameters in control young (n = 4) and aged (n = 5) brains. Aged rats show a decrease inhemispheric tissue area (p < 0.02) (A) but no differences in the percentage of area occupied by lateral ventricles (B). Quantification of NeuN+ profilesshowed a significative reduction in the number of neurons in the striatum of normal aged animals (p = 0.03) (C–E). Toluidine Blue staining (F and G) donot display apparent differences in the striatal neuronal morphology but qualitatively thicker white matter bundles in the aged are seen. Both age groupspresent a similar pattern of GFAP immunoreactivity, with strong immunoreactive astrocytes forming the glia limitans (H and I). Scale bars (D–L) = 50 lm.

M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354 347

remaining striatal NeuN+ profiles were seen between theyoung and the aged groups at 14DPL (Fig. 3J), the lasttime studied.

3.3. Astroglial reactivity

Aged and young animals injected with saline solutiondid not show differences in the pattern of astroglialresponse, both groups presented hypertrophied astrocytesonly adjacent to the needle track (data not shown). In con-trast, in animals injected with NMDA, the quantitativeanalysis of the area occupied by reactive astrocytes showedsignificant differences between ages and post-lesion times.The astroglial response developed earlier and was signifi-cantly greater in the aged subjects at all time points studied(Fig. 4). At both ages there was a significant increase in thearea of reactive astrocytes between 1 and 3DPL and the

maximum area was achieved at 3DPL, when 32.9 ± 6.1%of hemispheric area in the young and 58 ± 6.1% in the agedwere occupied by reactive astrocytes. Longer survival timeswere characterised by a progressive reduction until 7DPLbut no further reduction was detected at 14DPL(Fig. 4B), when the area occupied by reactive astrocyteswas still significantly bigger in the aged subjects, implyingthe formation of a larger glial scar.

At 12HPL, the lesion core of both young and aged ani-mals presented a punctuated background of GFAP stain-ing but only in the aged animals scattered GFAP+astrocytes were clearly distinguished (Fig. 5A–D). At1DPL, whereas in the young animals the lesion core wasstill devoid of astrocytes, in the aged, signs of astroglialhypertrophy were perceived (Fig. 5E–H): In the cortex,GFAP overexpressing cells were observed, and in the stri-atum thick immunoreactive cellular extensions were mainly

Fig. 2. Tissue injury and edema formation after excitotoxic damage to the young adult and aged brain. Representative toluidine blue stained sections (A)show the area of tissue injury (marked in every section with a broken line). The quantitative analysis of the area of tissue injury (B) show a significantlylarger injury size (*) in young subjects at 12HPL (p = 0.04) and 1DPL (p = 0.01) but no significant differences between age groups later on. In both agegroups, the injury size did not change significantly within the first 3DPL but showed a marginal reduction between 3 and 5DPL (p = 0.08 in young andp = 0.09 in the aged). In addition, young animals showed a further significant decrease in the area of tissue degeneration between 5 and 7DPL (p = 0.036).Quantification of the percentage of brain edema (C) shows marginally significantly bigger edema in young subjects only at 1DPL (*) (p = 0.06). Edemadevelops more slowly in the aged, with a marginal increase (&) between 1DPL and 3DPL (p = 0.10) and a significant diminution between 5 and 7DPL(p = 0.02) (*), in contrast to the young rats that show a marginal increase (&) between 12HPL and 1DPL (p = 0.9) and a significant decrease between 3 and5DPL (p = 0.03) (*). Animals analysed: aged; 12HPL and 14DPL n = 5, 1DPL and 3DPL n = 8, and 5DPL and 7DPL n = 4. Young adults; 12HPL,1DPL and 14DPL n = 5, 3DPL and 7DPL n = 4, and 5DPL n = 6. Scale bar (A) = 5 mm.

348 M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354

detected. Interestingly, at 3DPL, the central core of thecortical or striatal lesion and in both young and aged rats,was devoid of GFAP+ cells (and filled mainly with microg-lia/macrophages, data not shown) and surrounded by reac-tive astrocytes (Fig. 5I and J). In aged animals, theseastrocytes were more strongly hypertrophied, especiallystriatal astrocytes that showed the thickest and shortestextensions (Fig. 5K and L). At 5DPL, cellular hypertrophywas increased and aged reactive astrocytes were still morehypertrophied than in the young brains (Fig. 5M–P). At7DPL and in both age groups, the diminution in the areaoccupied by reactive astroglial cells (Fig. 4A) coincidedwith a concentration of reactive astrocytes in the lesioncore. At this survival time, the grade of astroglial reactivitywas similar between ages, and higher than the observed atearlier time-points (Fig. 5Q–T). At both ages, corticalastrocytes were more numerous, had thinner cellular pro-

cesses and showed a different pattern of hypertrophy thanstriatal astrocytes where GFAP immunoreactivity concen-trated in the soma, and cells displayed thicker and shorterextensions, especially in aged animals (Fig. 5T). At 14DPLin both ages, the presence of the glial scar was clearlyobserved, with more strongly packed and more hypertro-phied astrocytes than at 7DPL (Fig. 5U–X). At this lasttime-point, the presence of vacuoles in the astrocytic somawas clearly evident (Fig. 5Y) although they were alreadyobserved for the first time at 3DPL in both age groups.

4. Discussion

4.1. Neurodegeneration and edema in the aged brain

By analysing the evolution of lesion size and neurode-generation after an NMDA-induced excitotoxic lesion in

Fig. 3. Neurodegeneration after excitotoxic damage to the young adult and aged brain. Representative images from sections stained with Fluoro-Jade B(A–H) and quantitative analysis of positive cells (I) shown as the percentage of Fluoro-Jade B in relation to the density of striatal NeuN+ profiles incontrol animals (see Fig. 1). Young subjects present a higher percentage of degenerating neurons at 12HPL (*) (p < 0.01) than the aged (A,B,I), and reachmaximum percentage of stained cells by 1DPL (C,G). In contrast, in the aged (B,D,F,H), neurodegeneration was slower and the maximum percentage ofdegenerating neurons is not seen until 3DPL (p < 0.01) (F,H), when the percentage of degenerating neurons in the young striatum (E) is already decreased(p < 0.01). Global percentage of remaining NeuN+ profiles in the striatum at 14DPL was not different between age groups (J). Scale bars (A–H) = 50lm.Animals analysed: n = 4 for each age and survival time.

M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354 349

the striatum of young and aged rats we here describe, incomparison to the young rats, a decreased lesion size inaged rats until 1DPL and a concomitant slower progres-sion of neuronal degeneration in the injury core. Interest-ingly, available studies using several experimental modelsof brain ischemia in aged rats show a remarkable variabil-ity in the results reported. Studies have either shown biggerinjury size in the young brain (Shapira et al., 2002), no dif-ferences in the injury size between age groups (Badan et al.,2003; Brown et al., 2003; Duverger and MacKenzie, 1988;Fotheringham et al., 2000; Wang et al., 2003) or a biggerinjury size in the aged brain (Davis et al., 1995; Futrellet al., 1991; Kharlamov et al., 2000; Sutherland et al.,1996). The only previous study which has used NMDAto induce excitotoxic damage in old (24 months old) andyoung adult (8 weeks) (Suzuki et al., 2003) showed no sig-nificant differences in injury size at day 1, but they found atrend towards more damage in the young (p = 0.16), theonly time-point analysed. The several times lower dosageof NMDA used, the strain, and the site of NMDA injec-tion could account for the lack of significance in theirresults.

It is not surprising that the variability of neuropatho-logical findings becomes especially highlighted in studiesbased on aged subjects, where it sums to the intrinsic vari-ability of aging. We here describe a significant reductionin the aged brains’ hemispheric area and in the numberof striatal neuronal profiles, as it has been previouslydescribed (McNeill and Koek, 1990; Scahill et al., 2003;Stemmelin et al., 2000) and paralleling the presence ofscattered Fluoro-Jade B+ degenerating neurons in thenormal aged striatum. These differences stand out theneed of presenting results as percentages of age-matchedcontrols, when comparing aged versus young brains. Inthis regard, many existing work obviate the decrease inbrain size (Fotheringham et al., 2000; Kharlamov et al.,2000; Suzuki et al., 2003; Wang et al., 2003), which couldlead to misleading interpretations, as an equal injury sizemeans a bigger percentage of lesion in a smaller brain. Inthis sense, variations in the lesion model, experimentalprocedure, specie and strain, age, post-lesion time ana-lysed and the way of presenting results are particularlyimportant when comparing studies on the young and agedbrain’s response to an insult.

Fig. 4. Area of astroglial response after excitotoxic damage to the youngadult and aged brain. Representative GFAP immunolabeled sections (A)show the area of astroglial response at different post-lesion times (markedin every section with a broken line). The quantitative analysis of the areaof astroglial response (B) is shown as the percentage of hemispheric areaoccupied by reactive astrocytes. GFAP overexpression was initially foundat 1DPL in the aged and at 3DPL in the young and the area of astrogliosisis significatively larger in the aged subjects at all time points studied (*).However, both ages show similar temporal dynamics with a significantincrease in the area of reactive astrocytes from 1 to 3DPL (p < 0.01),maximum values at 3DPL and a reduction until 7DPL (p < 0.03). Animalsanalysed: Aged; n = 4 for each survival time. Young adults; 12HPL, 1DPLand 7DPL n = 4, 3DPL, 5DPL and 14DPL n = 5. Scale bar (A) = 5 mm.

350 M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354

It is now well established that the final lesion outcomedepends both on the injury size and the extent of neuronalcell death within the injured area. By using Fluoro-Jade B,a marker for degenerating neurons (Schmued and Hop-

kins, 2000), we have shown that whereas neurodegenera-tion is more rapid in the adult, showing maximumnumbers of dying neurons at 1DPL, in the aged striatumthe neurodegenerative process is slower and peaks at3DPL. The delay in the neurodegenerative process maybe probably explained by intrinsic factors in neuronal cellsoccurring during aging. In agreement, the most significantand consistent finding in glutamatergic transmission alter-ations occurring with aging is a decrease in NMDA recep-tor density which induces a previously reported diminutionin the response mediated by this receptor (Segovia et al.,2001; Yamada and Nabeshima, 1998). Moreover, the lowermetabolic rate, increased oxidative stress, inflammationand other biochemical CNS changes associated with aging(Finkel and Holbrook, 2000; Floyd and Hensley, 2002; Leeet al., 2000; Moeller et al., 1996), are stimulus reported toinduce moderate tolerance against ischemia-induced neuro-nal death in the adult brain, a phenomenon known asischemic tolerance or preconditioning (Dirnagl et al.,2003; Trendelenburg and Dirnagl, 2005). As examples,increased basal levels of nitric oxide, heat shock proteins(HSPs) and Nuclear Factor kappa B (NFjB) found in aged(Calabrese et al., 2004; Korhonen et al., 1997; Toliver-Kin-sky et al., 1997), have all been related to an increased resis-tance to ischemic damage (Dirnagl et al., 2003) and couldexplain why the aging brain may respond better to theinsult at first. However, it is well established that aged sub-jects present poor final lesion outcome attributed to adecreased capacity to adapt to new situations, which someauthors have related to a slower and reduced over-induc-tion of protective genes like HSPs and NFjB (Nakayamaet al., 1994; Shamovsky and Gershon, 2004; Verbekeet al., 2001), and that could explain the delayed neurode-generation we observe. However, it should be mentionedthat in our model, the global percentage of remaining stri-atal neuronal profiles was not different between ages, sug-gesting that both the decreased amount of striatalneurons in control aged striatum as well as the impairmentof neurochemical and neurophysiological parameters ofremaining neurons may contribute to a worse final out-come characteristic of aged subjects. In this sense, severalstudies have demonstrated that aging is associated with areduction in striatal dopamine as well as trophic activity(Ling et al., 2000) and that following neurotoxic treatmentto the aged striatum, the regenerative-associated increase inthe expression of neurotrophic factors like brain-derivedneurotrophic factor (BDNF) and glial-derived neurotro-phic factor (GDNF) is highly diminished (Collier et al.,2005; Yurek and Fletcher-Turner, 2000, 2001), pointingto the idea that the aged striatum loses the ability to bio-chemically and trophically compensate for neural damage.

Finally, slower neurodegeneration observed in agedbrains, may also be related to the delayed marginal increasein edema formation, as brain edema causing swelling of theaffected hemisphere is considered an important componentof the neuropathological process (Kimelberg, 2004; Linet al., 1993; Wang et al., 2003). Formation of edema

Fig. 5. Astroglial reactivity after excitotoxic damage to the young adult and aged brain from 12HPL to 14DPL. Only in the aged animals GFAP+astrocytes are observed within the first 1DPL (A–H): in the cortex, GFAP+ cells are clearly observed (B,F), while in the striatum immunoreactive thickcellular extensions are mainly detected (D,H). At 3DPL, in the cortex, a core free of GFAP+ cells (asterisks in I and J) surrounded by reactive astrocytes isseen at both ages (I,J), although astrocytes are more reactive in the aged (J, L in comparison to I, K, respectively). At 5DPL, cellular hypertrophy isincreased (M–P) and glial scar is observed at 7DPL and 14DPL at both ages (Q–X), although greater cell hypertrophy is clearly seen in the aged. Thepresence of vaquoles in the astrocytic soma is clearly evident in astrocytic soma at 14DPL (arrow in Y). The morphological differences between corticaland striatal astrocytes are maintained. Scale bars (A–Y) = 50 lm.

M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354 351

depends on blood–brain barrier breakdown and the move-ment of liquid from the extracellular to the intracellularcompartment, producing secondary injury by alteration

of brain homeostasis and mechanic pressure (Kimelberg,2004). Although brain edema has been sparsely studied inaged animals, it is clear that many factors implicated in

352 M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354

its formation differ in the aged brain. In this sense, a clin-ical study (Heinsius et al., 1998), shows that patients thatdied owing to brain edema were younger and died soonerthan those who died for other causes. In agreement, wehave observed increased and earlier edema formation inyoung animals. As it has been demonstrated that NMDAinhibition reduces edema formation (Ohtani et al., 2000),the previously mentioned decreased activity of these recep-tors in aging may account for the reduction in edema.However, it should be noted that previous studies using dif-ferent experimental injury models have described increasededema formation in aged animals at 1DPL (Fotheringhamet al., 2000; Gong et al., 2004).

4.2. Differences in the astroglial response

Many of the existing studies analysing the astroglialresponse after an insult in the aged, have been performedusing axotomy models (Goss and Morgan, 1995; Hurleyand Coleman, 2003; Kane et al., 1997), and most ofthem have reported a decreased astroglial response. Thisdiminished responsiveness has been either explained by arise in the threshold for astroglial reactivity due to intrin-sic age-induced astroglial activation (Scheff et al., 1980),or because aged astrocytes are less sensitive to signallingfrom injured neurons, especially in situations like axoto-my in rats, where neuronal death is limited. In contrast,available results on experimental models where massiveneuronal death takes place, like in our model, are con-tradictory: some studies carried out in a model of ische-mia show less astroglial activation in the aged brain(Futrell et al., 1991; Kharlamov et al., 2000), whereasother studies using permanent focal ischemia or traumahave shown, as we do, increased astroglial activation inthe aged (Badan et al., 2003; Fotheringham et al.,2000; Kyrkanides et al., 2001). However, to ourknowledge, this is the first study reporting astroglialhypertrophy in aged animals as soon as 1DPL.

In our study, the larger area of astroglial response in theaged rats at all times studied and in the absence of anincreased area of neurodegeneration, implies the activationof astrocytes in the tissue border surrounding damage. Inthis sense, it has been reported that both pro and anti-in-flammatory cytokines such as interleukin-1beta, tumournecrosis factor alpha and interleukin-6, well known modu-lators of astroglial reactivity (Raivich et al., 1999; Stollet al., 1998), are elevated in the aged rodent (Bodles andBarger, 2004), undoubtedly influencing the astroglialresponse. In addition, it is thought that astroglial activa-tion in distal sites depends on the propagation of intracel-lular calcium (Finkbeiner, 1992; Raivich et al., 1999). Inrelation to this, it has been postulated that in the first stagesof neurodegeneration, affected neurons can trigger calciumwave propagation to adjacent astrocytes by gap junctionsor by glutamate release (Budd and Lipton, 1998), initiatingneuronal death (Anderson et al., 2003), but also triggeringastrogliosis. The findings that astroglial gap junctions

remain opened for hours in ischemic conditions (Cotrinaet al., 1998), and that astrocytes of the aged brain expresshigh amounts of gap junction proteins and that these junc-tions are functional (Cotrina et al., 2001), suggest that cel-lular intermediates can diffuse freely to distal astrocytes inthe penumbra. In addition, it has been previously describedthat an amplification of calcium responses is promoted bythe oxidation of thiol groups in the 3-inositol triphosphate(IP3) receptor (Peuchen et al., 1996). As oxidative stress is awell known event in the aging process (Finkel and Hol-brook, 2000; Floyd and Hensley, 2002; Lee et al., 2000) thisamplified response of aged astrocytes could further explainthe larger extension of reactive astrocytes reported hereand hypothesized in a previous study (Cotrina and Nederg-aard, 2002). Finally, we have also reported a significantlylarger glial scar in aged animals. As it is commonly knownthat the glial scar is inhibitory for axonal regeneration(Bovolenta et al., 1993), a larger glial scar could also con-tribute to the worsen lesion outcome found in the agedbrain (Badan et al., 2003). However, further studies areneeded to elucidate the role of the astroglial response inthe delay of the neurodegenerative processes occurring inthe aged and the putative differences that may occur inthe microglial reactivity and the inflammatory response.It is now well established that besides primary neuronal celldeath, inflammatory, toxic and protective molecules pro-duced by surrounding activated glial cells and infiltratingleukocytes determine the final outcome of brain damage.Among other mechanisms, oxidative stress and inflamma-tion mediated by proinflammatory cytokines and enzymes(Iadecola, 1997; Minghetti and Levi, 1998) are consideredto be the leading factors contributing to delayed cell death(Dirnagl et al., 1999).

In conclusion, these results emphasize the importance ofusing aged animals for experimental models of insults relat-ed to aging, and stands out the need of taking into accountthe changes related to normal aging in addition of multipleparameters along different survival times to evaluate dam-age progression in the aged. Noteworthy, the delay in theneurodegenerative process may imply a longer therapeuticwindow in the aged.

Acknowledgements

Supported by BFI2002-02079, ‘‘La Caixa’’ (00/074-00)and BFU2005-02783. The authors thank Miguel A. Martiland Lola Mulero for their outstanding technical help, Dr.Perez-Clausell’s group for introducing us to the FJ-B stain-ing, Dr. Armario’s group and Dr. Giraldo for statisticalaid, and Hugo Peluffo for critical reading of themanuscript.

References

Acarin, L., Gonzalez, B., Castellano, B., 2001. Glial activation in theimmature rat brain: implication of inflammatory transcription factorsand cytokine expression. Prog. Brain Res. 132, 375–389.

M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354 353

Acarin, L., Gonzalez, B., Hidalgo, J., Castro, A.J., Castellano, B., 1999.Primary cortical glial reaction versus secondary thalamic glial responsein the excitotoxically injured young brain: astroglial response andmetallothionein expression. Neuroscience 92, 827–839.

Amenta, F., Bronzetti, E., Sabbatini, M., Vega, J.A., 1998. Astrocytechanges in aging cerebral cortex and hippocampus: a quantitativeimmunohistochemical study. Microsc. Res. Tech. 43, 29–33.

Anderson, M.F., Blomstrand, F., Blomstrand, C., Eriksson, P.S., Nilsson,M., 2003. Astrocytes and stroke: networking for survival? Neurochem.Res. 28, 293–305.

Badan, I., Platt, D., Kessler, C., Popa-Wagner, A., 2003. Temporaldynamics of degenerative and regenerative events associated withcerebral ischemia in aged rats. Gerontology 49, 356–365.

Bodles, A.M., Barger, S.W., 2004. Cytokines and the aging brain – whatwe don’t know might help us. Trends Neurosci. 27, 621–626.

Bonita, R., Solomon, N., Broad, J.B., 1997. Prevalence of stroke andstroke-related disability. Estimates from the Auckland stroke studies.Stroke 28, 1898–1902.

Bovolenta, P., Wandosell, F., Nieto-Sampedro, M., 1993. Neuriteoutgrowth inhibitors associated with glial cells and glial cell lines.Neuroreport 5, 345–348.

Bronson, R.T., Lipman, R.D., Harrison, D.E., 1993. Age-related gliosis inthe white matter of mice. Brain Res. 609, 124–128.

Brown, A.W., Marlowe, K.J., Bjelke, B., 2003. Age effect on motorrecovery in a post-acute animal stroke model. Neurobiol. Aging 24,607–614.

Bruns Jr., J., Hauser, W.A., 2003. The epidemiology of traumatic braininjury: a review. Epilepsia 44 (Suppl. 10), 2–10.

Budd, S.L., Lipton, S.A., 1998. Calcium tsunamis: do astrocytes transmitcell death messages via gap junctions during ischemia? Nat. Neurosci.1, 431–432.

Calabrese, V., Scapagnini, G., Ravagna, A., Colombrita, C., Spadaro, F.,Butterfield, D.A., Giuffrida Stella, A.M., 2004. Increased expression ofheat shock proteins in rat brain during aging: relationship withmitochondrial function and glutathione redox state. Mech. AgeingDev. 125, 325–335.

Collier, T.J., Dung Ling, Z., Carvey, P.M., Fletcher-Turner, A., Yurek,D.M., Sladek Jr., J.R., Kordower, J.H., 2005. Striatal trophic factoractivity in aging monkeys with unilateral MPTP-induced parkinson-ism. Exp. Neurol. 191 (Suppl. 1), S60–S67.

Cotrina, M.L., Gao, Q., Lin, J.H., Nedergaard, M., 2001. Expression andfunction of astrocytic gap junctions in aging. Brain Res. 901, 55–61.

Cotrina, M.L., Kang, J., Lin, J.H., Bueno, E., Hansen, T.W., He, L., Liu,Y., Nedergaard, M., 1998. Astrocytic gap junctions remain openduring ischemic conditions. J. Neurosci. 18, 2520–2537.

Cotrina, M.L., Nedergaard, M., 2002. Astrocytes in the aging brain. J.Neurosci. Res. 67, 1–10.

Davis, M., Mendelow, A.D., Perry, R.H., Chambers, I.R., James, O.F.,1995. Experimental stroke and neuroprotection in the aging rat brain.Stroke 26, 1072–1078.

Dietrich, W.D., Halley, M., Alonso, O., Globus, M.Y., Busto, R., 1992.Intraventricular infusion of N-methyl-D-aspartate. 2. Acute neuronalconsequences. Acta Neuropathol. (Berl.) 84, 630–637.

Dirnagl, U., Iadecola, C., Moskowitz, M.A., 1999. Pathobiologyof ischaemic stroke: an integrated view. Trends Neurosci. 22,391–397.

Dirnagl, U., Simon, R.P., Hallenbeck, J.M., 2003. Ischemic tolerance andendogenous neuroprotection. Trends Neurosci. 26, 248–254.

Doble, A., 1999. The role of excitotoxicity in neurodegenerative disease:implications for therapy. Pharmacol. Ther. 81, 163–221.

Duverger, D., MacKenzie, E.T., 1988. The quantification of cerebralinfarction following focal ischemia in the rat: influence of strain,arterial pressure, blood glucose concentration, and age. J. Cereb.Blood Flow Metab. 8, 449–461.

Eng, L.F., Ghirnikar, R.S., Lee, Y.L., 2000. Glial fibrillary acidic protein:GFAP-thirty-one years (1969–2000). Neurochem. Res. 25, 1439–1451.

Finch, C.E., 2003. Neurons, glia, and plasticity in normal brain aging.Neurobiol. Aging 24 (Suppl. 1), S123–S127, discussion S131.

Finkbeiner, S., 1992. Calcium waves in astrocytes-filling in the gaps.Neuron 8, 1101–1108.

Finkel, T., Holbrook, N.J., 2000. Oxidants, oxidative stress and thebiology of ageing. Nature 408, 239–247.

Floyd, R.A., Hensley, K., 2002. Oxidative stress in brain aging. Implica-tions for therapeutics of neurodegenerative diseases. Neurobiol. Aging23, 795–807.

Fotheringham, A.P., Davies, C.A., Davies, I., 2000. Oedema and glial cellinvolvement in the aged mouse brain after permanent focal ischaemia.Neuropathol. Appl. Neurobiol. 26, 412–423.

Fujimura, M., Morita-Fujimura, Y., Copin, J., Yoshimoto, T., Chan,P.H., 2001. Reduction of copper, zinc-superoxide dismutase inknockout mice does not affect edema or infarction volumes and theearly release of mitochondrial cytochrome c after permanent focalcerebral ischemia. Brain Res. 889, 208–213.

Futrell, N., Garcia, J.H., Peterson, E., Millikan, C., 1991. Embolic strokein aged rats. Stroke 22, 1582–1591.

Gong, Y., Hua, Y., Keep, R.F., Hoff, J.T., Xi, G., 2004. Intracerebralhemorrhage: effects of aging on brain edema and neurological deficits.Stroke 35, 2571–2575.

Goss, J.R., Morgan, D.G., 1995. Enhanced glial fibrillary acidic proteinRNA response to fornix transection in aged mice. J. Neurochem. 64,1351–1360.

Heinsius, T., Bogousslavsky, J., Van Melle, G., 1998. Large infarcts in themiddle cerebral artery territory. Etiology and outcome patterns.Neurology 50, 341–350.

Hurley, S.D., Coleman, P.D., 2003. Facial nerve axotomy in aged andyoung adult rats: analysis of the glial response. Neurobiol. Aging 24,511–518.

Iadecola, C., 1997. Bright and dark sides of nitric oxide in ischemic braininjury. Trends Neurosci. 20, 132–139.

Kane, C.J., Sims, T.J., Gilmore, S.A., 1997. Astrocytes in the aged ratspinal cord fail to increase GFAP mRNA following sciatic nerveaxotomy. Brain Res. 759, 163–165.

Kharlamov, A., Kharlamov, E., Armstrong, D.M., 2000. Age-dependentincrease in infarct volume following photochemically induced cerebralinfarction: putative role of astroglia. J. Gerontol. A Biol. Sci. Med. Sci.55, B135–B141, discussion B142–3.

Kimelberg, H.K., 2004. Water homeostasis in the brain: basic concepts.Neuroscience 129, 851–860.

Kollegger, H., McBean, G.J., Tipton, K.F., 1993. Reduction of striatal N-methyl-D-aspartate toxicity by inhibition of nitric oxide synthase.Biochem. Pharmacol. 45, 260–264.

Korhonen, P., Helenius, M., Salminen, A., 1997. Age-related changes inthe regulation of transcription factor NF-kappa B in rat brain.Neurosci. Lett. 225, 61–64.

Kyrkanides, S., O’Banion, M.K., Whiteley, P.E., Daeschner, J.C.,Olschowka, J.A., 2001. Enhanced glial activation and expressionof specific CNS inflammation-related molecules in aged versusyoung rats following cortical stab injury. J. Neuroimmunol. 119,269–277.

Landfield, P.W., Rose, G., Sandles, L., Wohlstadter, T.C., Lynch, G.,1977. Patterns of astroglial hypertrophy and neuronal degeneration inthe hippocampus of ages, memory-deficient rats. J. Gerontol. 32, 3–12.

Lee, C.K., Weindruch, R., Prolla, T.A., 2000. Gene-expression profile ofthe ageing brain in mice. Nat. Genet. 25, 294–297.

Lin, T.N., He, Y.Y., Wu, G., Khan, M., Hsu, C.Y., 1993. Effect of brainedema on infarct volume in a focal cerebral ischemia model in rats.Stroke 24, 117–121.

Ling, Z.D., Collier, T.J., Sortwell, C.E., Lipton, J.W., Vu, T.Q., Robie,H.C., Carvey, P.M., 2000. Striatal trophic activity is reduced in theaged rat brain. Brain Res. 856, 301–309.

Mattson, M.P., 2003. Excitotoxic and excitoprotective mechanisms:abundant targets for the prevention and treatment of neurodegener-ative disorders. Neuromolecular Med. 3, 65–94.

McNeill, T.H., Koek, L.L., 1990. Differential effects of advancing age onneurotransmitter cell loss in the substantia nigra and striatum ofC57BL/6N mice. Brain Res. 521, 107–117.

354 M.M. Castillo-Ruiz et al. / Experimental Gerontology 42 (2007) 343–354

Minghetti, L., Levi, G., 1998. Microglia as effector cells in brain damageand repair: focus on prostanoids and nitric oxide. Prog. Neurobiol. 54,99–125.

Mody, I., MacDonald, J.F., 1995. NMDA receptor-dependent excitotox-icity: the role of intracellular Ca2+ release. Trends Pharmacol. Sci. 16,356–359.

Moeller, J.R., Ishikawa, T., Dhawan, V., Spetsieris, P., Mandel, F.,Alexander, G.E., Grady, C., Pietrini, P., Eidelberg, D., 1996. Themetabolic topography of normal aging. J. Cereb. Blood Flow Metab.16, 385–398.

Nakayama, H., Jorgensen, H.S., Raaschou, H.O., Olsen, T.S., 1994. Theinfluence of age on stroke outcome. The Copenhagen Stroke Study.Stroke 25, 808–813.

Nedergaard, M., Dirnagl, U., 2005. Role of glial cells in cerebral ischemia.Glia 50, 281–286.

Ohtani, K., Tanaka, H., Yasuda, H., Maruoka, Y., Kawabe, A.,Nakamura, M., 2000. Blocking the glycine-binding site of NMDAreceptors prevents the progression of ischemic pathology induced bybilateral carotid artery occlusion in spontaneously hypertensive rats.Brain Res. 871, 311–318.

Olney, J.W., 1969. Brain lesions, obesity, and other disturbances in micetreated with monosodium glutamate. Science 164, 719–721.

Peinado, M.A., 1998. Histology and histochemistry of the aging cerebralcortex: an overview. Microsc. Res. Tech. 43, 1–7.

Pekny, M., Nilsson, M., 2005. Astrocyte activation and reactive gliosis.Glia 50, 427–434.

Peuchen, S., Clark, J.B., Duchen, M.R., 1996. Mechanisms of intracellularcalcium regulation in adult astrocytes. Neuroscience 71, 871–883.

Raivich, G., Jones, L.L., Werner, A., Bluthmann, H., Doetschmann, T.,Kreutzberg, G.W., 1999. Molecular signals for glial activation: pro-and anti-inflammatory cytokines in the injured brain. Acta Neurochir.Suppl. 73, 21–30.

Scahill, R.I., Frost, C., Jenkins, R., Whitwell, J.L., Rossor, M.N., Fox,N.C., 2003. A longitudinal study of brain volume changes in normalaging using serial registered magnetic resonance imaging. Arch.Neurol. 60, 989–994.

Scheff, S.W., Benardo, L.S., Cotman, C.W., 1980. Decline in reactive fibergrowth in the dentate gyrus of aged rats compared to young adult ratsfollowing entorhinal cortex removal. Brain Res. 199, 21–38.

Schmued, L.C., Hopkins, K.J., 2000. Fluoro-Jade B: a high affinityfluorescent marker for the localization of neuronal degeneration. BrainRes. 874, 123–130.

Segovia, G., Porras, A., Del Arco, A., Mora, F., 2001. Glutamatergicneurotransmission in aging: a critical perspective. Mech. Ageing Dev.122, 1–29.

Shah, G.N., Mooradian, A.D., 1997. Age-related changes in the blood–brain barrier. Exp. Gerontol. 32, 501–519.

Shamovsky, I., Gershon, D., 2004. Novel regulatory factors of HSF-1activation: facts and perspectives regarding their involvement in theage-associated attenuation of the heat shock response. Mech. AgeingDev. 125, 767–775.

Shapira, S., Sapir, M., Wengier, A., Grauer, E., Kadar, T., 2002. Aginghas a complex effect on a rat model of ischemic stroke. Brain Res. 925,148–158.

Stemmelin, J., Lazarus, C., Cassel, S., Kelche, C., Cassel, J.C., 2000.Immunohistochemical and neurochemical correlates of learning defi-cits in aged rats. Neuroscience 96, 275–289.

Stewart, G.R., Price, M., Olney, J.W., Hartman, B.K., Cozzari, C., 1986.N-Methylaspartate: an effective tool for lesioning basal forebraincholinergic neurons of the rat. Brain Res. 369, 377–382.

Stoll, G., Jander, S., Schroeter, M., 1998. Inflammation and glialresponses in ischemic brain lesions. Prog. Neurobiol. 56, 149–171.

Sutherland, G.R., Dix, G.A., Auer, R.N., 1996. Effect of age in rodentmodels of focal and forebrain ischemia. Stroke 27, 1663–1667,discussion 1668.

Suzuki, Y., Takagi, Y., Nakamura, R., Hashimoto, K., Umemura, K.,2003. Ability of NMDA and non-NMDA receptor antagonists toinhibit cerebral ischemic damage in aged rats. Brain Res. 964, 116–120.

Toliver-Kinsky, T., Papaconstantinou, J., Perez-Polo, J.R., 1997. Age-associated alterations in hippocampal and basal forebrain nuclearfactor kappa B activity. J. Neurosci. Res. 48, 580–587.

Trendelenburg, G., Dirnagl, U., 2005. Neuroprotective role of astrocytes incerebral ischemia: focus on ischemic preconditioning. Glia 50, 307–320.

Truelsen, T., Bonita, R., 2003. Advances in ischemic stroke epidemiology.Adv. Neurol. 92, 1–12.

Unger, J.W., 1998. Glial reaction in aging and Alzheimer’s disease.Microsc. Res. Tech. 43, 24–28.

Verbeke, P., Fonager, J., Clark, B.F., Rattan, S.I., 2001. Heat shockresponse and ageing: mechanisms and applications. Cell Biol. Int. 25,845–857.

Wang, R.Y., Wang, P.S., Yang, Y.R., 2003. Effect of age in rats followingmiddle cerebral artery occlusion. Gerontology 49, 27–32.

Woo, J., Kay, R., Yuen, Y.K., Nicholls, M.G., 1992. Factors influencinglong-term survival and disability among three-month stroke survivors.Neuroepidemiology 11, 143–150.

Yamada, K., Nabeshima, T., 1998. Changes in NMDA receptor/nitricoxide signaling pathway in the brain with aging. Microsc. Res. Tech.43, 68–74.

Yurek, D.M., Fletcher-Turner, A., 2000. Lesion-induced increase ofBDNF is greater in the striatum of young versus old rat brain. Exp.Neurol. 161, 392–396.

Yurek, D.M., Fletcher-Turner, A., 2001. Differential expression ofGDNF, BDNF, and NT-3 in the aging nigrostriatal system followinga neurotoxic lesion. Brain Res. 891, 228–235.