Bilateral Changes After Neonatal Ischemia in the P7 Rat Brain

ORIGINAL ARTICLE

Bilateral Changes After Neonatal Ischemia in the P7 Rat Brain

Maria Spiegler, Sonia Villapol, MD, Valerie Biran, MD, Catherine Goyenvalle, MD,

Jean Mariani, MD, PhD, Sylvain Renolleau, MD, PhD, and Christiane Charriaut-Marlangue, PhD

AbstractNeurogenesis persists throughout life in the rodent subventricular

zone (SVZ) and subgranular zone (SGZ) and increases in the adult

after brain injury. In this study, postnatal day 7 rats underwent

middle cerebral artery electrocoagulation and transient homolateral

common carotid artery occlusion, a lesioning protocol that resulted

in ipsilateral (IL) forebrain ischemic injury, leading to a cortical

cavity 3 weeks later. The effects of neonatal ischemia on

hemispheric damage, cell death, cell proliferation, and neurogenesis

were examined 4 hours to 6 weeks later by the terminal

deoxynucleotidyl transferase dUTP nick-end labeling assay and

immunohistochemistry of Ki-67 in proliferating cells and of

doublecortin, a microtubule-associated protein expressed only by

immature neurons. Neonatal ischemic injury resulted in persistent

reduced IL and transient reduced contralateral (CL) hemispheric

areas, a consequence of sustained and transient cell death in the IL

and CL areas, respectively. Ki-67 immunostaining revealed 3 peaks

of newly generated cells in the dorsal SVZ and SGZ in the IL side

and also in the CL side at 48 hours and 7 and 28 days after

ischemia. Double immunofluorescence revealed that most of the

Ki-67-positive cells were astrocytes at 48 hours. Ischemic injury

also stimulated SVZ neurogenesis, based on increased doublecortin

immunostaining in both SVZs at 7 to 14 days after injury.

Doublecortin-positive neurons remained visible around the lesion

at 21 days but displayed an immature shape in discrete chains or

clusters. Although unilateral ischemic damage was produced,

results indicate successful regenerative changes in the CL hemi-

sphere, allowing anatomical recovery.

Key Words: Cavity, Contralateral, Doublecortin, Neonatal

ischemia, Subventricular zone.

INTRODUCTIONPerinatal hypoxia-ischemia (HI) followed by reperfu-

sion is an important cause of neonatal brain injury (1).

Despite recent advances in neonatal intensive care manage-ment, the incidence of long-term disability after neonatalbrain injury is not decreasing and has even been shown toincrease in some areas in the last 30 years (2). Many infantswith this problem have permanent neurologic deficits (suchas cerebral palsy, seizures, sensory deficits, and mentalretardation). Recent progress in regenerative medicine and inour understanding of neurogenesis with stem cell therapymay lead to the recovery of neuronal function that was lostby perinatal HI (3, 4).

In adult rodents, acute ischemic brain injury stimulatesneurogenesis in 2 forebrain regions: the subventricular zone(SVZ) and subgranular zone (SGZ) of the dentate gyrus (5).In contrast with these findings, in the neonatal brain, severeHI (elicited by unilateral carotid ligation and timed exposureto 8% oxygen) acutely depleted neural precursors in the rat(6) and mouse (7) SVZ. However, moderate HI (by reducedhypoxic time exposure) in postnatal day (P) 7 rats (8, 9) orP10 mice (10) resulted in a marked increase in cellproliferation and stimulation of neurogenesis in the ipsi-lateral (IL) side 1 to 3 weeks later. We also recently reportedthat neonatal stroke (elicited by middle cerebral artery[MCA] electrocoagulation and transient homolateral com-mon carotid artery [CCA] occlusion) (11), leading tocerebral damage restricted in the hemisphere IL to theoccluded artery increased astrocyte and oligodendrocyteneogenesis in the posterior corpus callosum (cingulum) overthe dorsal hippocampus (12). All previous data wereprovided from models that have been described as inducingcell injury in the IL hemisphere without damage contral-aterally (11, 13). Despite the absence of detectable differ-ences between IL and contralateral (CL) hemispheres whenexamined macroscopically, several studies reported CL cellinjury (14Y16).

The purpose of this study was to evaluate the effectof neonatal stroke on morphologic changes, cell death, andneurogenesis in the IL hemisphere but also in remoteregions and particularly in the CL hemisphere. P7 ratsunderwent ischemia, and brains were analyzed from 4 hoursup to 6 weeks for: 1) cell death using the terminaldeoxynucleotidyl transferase dUTP nick-end labeling(TUNEL) assay; 2) proliferation of precursors with Ki-67and nestin labeling; and 3) neurogenesis using doublecortin(DCx) immunofluorescence. Our data demonstrate thatischemia in the neonatal rat brain enhances cell prolifer-ation of progenitors in the IL and CL dorsolateral (DL)SVZ during the recovery phase after injury. Interestingly,

481J Neuropathol Exp Neurol � Volume 66, Number 6, June 2007

J Neuropathol Exp NeurolCopyright � 2007 by the American Association of Neuropathologists, Inc.

Vol. 66, No. 6June 2007

pp. 481Y490

From the Universite Pierre et Marie Curie-Paris6 (MS, VB, CG, JM, SR,CC-M), Unite Mixte de Recherche-Centre National de la RechercheScientifique 7102, Paris, France; Institute of Neuroscience (SV),Autonomous University of Barcelona, Barcelona, Spain; Service deNeonatologie (VB) and Service de Reanimation Neonatale et Pediatri-que (SR), APHP, Hopital Armand Trousseau, Paris, France; and APHP,Hopital Charles Foix, Unite d’Explorations Fonctionelles, Ivry-sur-Seine, France (JM).

Send correspondence and reprint requests to: Christiane Charriaut-Marlangue, PhD, UMR-CNRS 7102, HICD, case 14, 9 quai St-Bernard,75005 Paris, France; E-mail: [email protected]

Color figures are available online at http://www.jneuropath.com.

7Copyright @ 200 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

our data suggest that cell proliferation in the CL side com-pensates for early deficits.

MATERIALS AND METHODS

Perinatal IschemiaAll animal experimentation was conducted in accord-

ance with the French and European Community guidelinesfor the care and use of experimental animals. Ischemia wasperformed in 7-day-old rats (17Y21 g) of both sexes, aspreviously described (11), minimizing the number ofanimals used and their suffering. Rat pups were anesthetizedwith an intraperitoneal injection of chloral hydrate (350mg/kg). The anesthetized rat was positioned on its back anda median incision was made in the neck to expose the leftCCA. The rat was then placed on the right side, and anoblique skin incision was made between the ear and the eye.After excision of the temporal muscle, the cranial bone wasremoved from the frontal suture to a level below thezygomatic arch. Then the left middle cerebral artery,exposed just after its appearance over the rhinal fissure,was coagulated at the inferior level of the cerebral vein.After this procedure, a clip was placed to occlude the leftCCA. Rats were then placed in an incubator to avoidhypothermia, and after 50 minutes the clip was removed.Carotid blood flow restoration was verified with the aid of a

microscope. Both neck and cranial skin incisions were thenclosed. During the surgical procedure, body temperature wasmaintained at 37-C. After recovery, pups were transferred totheir mothers. Controls received neither MCA electrocoagu-lation nor transient CCA occlusion. Of 45 ischemic animals,4 died during the first 24 hours of recovery. Eighty-fivepercent (36 of 41) of the animals sustained a lesion (Fig. 1),as previously observed, and were used in this study.

Tissue Preparation and HistologyIschemic and control rats at 4, 12, 24, and 48 hours

(n = 3Y4 at each time point) and 7, 10, 14, 21, and 28 daysand 6 weeks (n = 2 at each time point) were anesthetizedwith chloral hydrate (400 mg/kg i.p.) and killed, and brainswere rapidly removed on an ice-cold plate. Brains were keptfor 2 to 24 hours in phosphate buffer (0.12 M, pH 7.4)solution containing 0.5 or 4% paraformaldehyde and placedin 0.12 M phosphate buffer containing 20% sucrose for 2 to3 days. The cryoprotected brains were frozen in isopentane(Y40-C), and serial cryostat sections (20-Km thick) werecollected on gelatin-coated slides. Coronal sections throughthe dorsal hippocampus were cresyl violet-stained to assesstissue damage. A computerized video camera-based imageanalysis (with Image-Pro version 4.1 software) was used tomeasure bilateral regional cross-sectional areas. From P8 toP28, bilateral cross-sectional areas of the entire cerebral

FIGURE 1. Neonatal stroke in postnatal day (P) 7 rat brain. (A) Examples of brain damage to the left ipsilateral hemisphere at48 hours and 10 and 21 days of recovery. Representative cresyl violet-stained sections at the level of dorsal hippocampus(bregma Y3.3 mm) showing a pale ischemic zone (48 hours) and a cortical cyst (10 and 21 days). Note that in some animals, theCA2YCA3 hippocampal field can be damaged (arrow in inset at 10 days). (B) Bilateral cerebral hemisphere areas were measuredin cresyl violet-stained sections of ischemic animals (from 48 hours [P9] to 21 days [P28]) compared with naive age-matchedanimals (n = 5 at each time point for both groups). Values are expressed as means T SD (mm2); at each age values werecompared with analysis of variance (*, p G 0.05 versus control). **, p G 0.01. (C) Body weight (in grams) of control (naive age-matched) and ischemic rats was measured at 48 hours after ischemia and once per week for 3 weeks. Ischemic animals had only asmaller body weight at 48 hours of recovery (*, p G 0.05) compared with naive animals.

Spiegler et al J Neuropathol Exp Neurol � Volume 66, Number 6, June 2007

� 2007 American Association of Neuropathologists, Inc.482


hemispheres (i.e. IL and CL) were measured in 4 coronalsections per brain at levels encompassing the dorsal hippo-campus (Fig. 1) corresponding to plate 31 at a bregma ofj3.3 mm in the brain atlas of Paxinos and Watson (17).Data are expressed in mm2 T SD.

Fluorescence In Situ Labelingof Fragmented DNA

Sections (n = 3 at each time point) were processed forDNA strand breaks (TUNEL assay) using the In Situ CellDeath Detection Kit, Fluorescein (Roche, Meylan, France)according to the manufacturer_s instructions. TUNEL-positive nuclei were counted under fluorescence microscopy,and results are expressed as means T SD per 0.1 mm2.

ImmunohistochemistryThe primary antibodies were directed against Ki-67, a

proliferative marker (rabbit polyclonal NCL-Ki-67p, 1:200;Novocastra, Tebu, Le Perray en Yvelines, France). Thesections, at bregma Y3.3 mm, were first incubated for 30minutes with 5% normal horse serum in PBS with 0.5%Triton X-100 (PBS-TX-NHS) 48 hours at 4-C withappropriately diluted primary antibody in PBS-TX-NHS.The primary antibodies were visualized after incubation withthe appropriate species-specific biotinylated secondary anti-body (Vectastain, 1:200; AbCys, Paris, France) followed bythe streptavidin-biotin-peroxidase complex (Elite ABC kit,Vectastain). Nonspecific peroxidase activity was abrogatedby incubating the sections in 1% hydrogen peroxide in 0.1 MPBS at the appropriate stage. Selected sections were photo-graphed using a Nikon Eclipse E800M microscope (Nikon,Paris, France) and DFC 300 FX camera (Leica Micro-systems, Rueil-Malmaison, France) interfaced with IM50imaging software.

ImmunofluorescenceThe primary antibodies were directed against DCx

(DCx goat polyclonal, 1:1000; Santa-Cruz, Tebu-Bio, LePerray-en-Yvelines, France) to label neo-formed neurons,NG2 (Chemicon, AB5320, 1:200; Euromedex, Souffelweyer-sheim France) to label pre-oligodendrocyte, Cy3-conjugatedglial fibrillary acidic protein (GFAP) (clone G-A-5, 1:1000;Sigma-Aldrich, St-Quentin Fallavier, France) to stain astro-cytes, and Ki-67 (see above) and nestin (mouse monoclonalclone Rat 401, 1:200; Chemicon) to detect proliferation andprogenitors, respectively. The sections (n = 6Y8 per brain),were first incubated for 30 minutes with PBS-TX-NHS andovernight at room temperature with appropriately dilutedprimary antibody in PBS-TX-NHS, except for Ki-67 (48hours incubation at 4-C). The primary antibodies were visu-alized after incubation with the appropriate species-specificfluorescein isothiocyanate- or Cy3-conjugated secondaryantibody (1:100; Jackson ImmunoResearch Laboratories,Interchim, Asnieres, France) for 2 hours at room temperature.The specificity of the primary antibodies that we used wastested by their omission. Combined double fluorescenceand/or DCx and TUNEL assays were also performed.Immunofluorescence staining was analyzed using a LeicaDM-IRBE confocal laser microscope, and z-series stack

(2 Km optical slice thickness) images were obtained andimported into Adobe Photoshop (version 7.0) for imageprocessing and analysis.

QuantificationKi-67-labeled nuclei were counted in the SVZDL and

SGZ in 3 coronal sections (20-Km thick, spaced 50 Kmapart) for each animal and at each time point of recovery inboth IL and CL sides (n = 3) and in age-matched control(n = 2) rats. All counts were performed using a 100�objective, and data are given as means T SD per 0.1 mm2.

Statistical AnalysisA microcomputer-based statistics program (Statview,

version 5.1) was used for data analysis. Analysis of variancewas used to compare area values and cell counts; theBonferroni post hoc test was applied to evaluate intergroupdifferences.

RESULTS

Histopathologic Effects of Neonatal Ischemiain P7 Rat Pups

The association of the left MCA electrocauterizationand transient (50 minutes) homolateral CCA occlusion led toan infarct in the parietal cortex at 48 hours, which evolved ina cavity and a complete loss of the parietal cortex at 7 to 10and 21 days after recovery, respectively (Fig. 1A). In 15% to20% of animals, the CA2YCA3 layers of the hippocampuswere also damaged, as shown in Figure 1A (arrow inenlarged panel at 10 days). To analyze the extent ofischemic damage, we measured the area of both hemispheresat various times of recovery and compared them with thenormal development of cerebral hemispheres from age-matched naive control animals. As reported in Figure 1B,ischemic animals demonstrated a significantly (between 18%and 27%) reduced IL hemispheric area for all time pointsstudied, compared with control data. A reduced CL hemi-spheric area was also detected at 24 hours (29.7 T 1.8 vscontrol 35.6 T 1.9 mm2, p G 0.05); in contrast, data at 48(31 T 3 mm2) and 72 (34.5 T 1.9 mm2) hours were notsignificantly different from those for naive pups (35.0 T 4.4and 41.1 T 6.2 mm2, respectively) although a small trend to areduced CL hemisphere area was observed (Fig. 1B).Thereafter, no difference was observed in hemisphere areasbetween CL sides of ischemic and control animals at thesame age. The mean lesion size represented 21% to 25% ofthe IL hemisphere at 48 hours of recovery, as classicallyfound in our previous studies (18). The ischemic animalsdemonstrated reduced body weight (~20%Y25%) comparedwith age-matched control animals at 48 hours after ischemia.A recovery in body weight was observed as soon as 1 weekafter lesioning, and no difference was shown betweencontrol and ischemic animals thereafter (Fig. 1C).

Cell Death in Both Ipsi- and ContralateralHemispheres After Neonatal Ischemia

Ischemic animals showed a significant increase inTUNEL-positive nuclei in the IL parietal cortex from 4 to 48

J Neuropathol Exp Neurol � Volume 66, Number 6, June 2007 Neonatal Cerebral Ischemia

� 2007 American Association of Neuropathologists, Inc. 483


hours after recovery (Fig. 2A, C), and most were observed inthe mediocortical layers III to V (Fig. 2A). Some columnarpreserved tissues without TUNEL staining could be detectedat 24 hours after lesioning (Fig. 2A, white arrow). TUNEL-positive nuclei were also detected in layer VI and underlyingwhite matter, and numerous dying cells were present in theIL hemisphere between 4 and 48 hours and also in the CLhemisphere at 12 hours of recovery (Fig. 2C). TUNEL-positive nuclei were still detected 7 and 10 days afterischemia around the cavity (Fig. 2B) in the IL hemisphere.TUNEL-positive nuclei have also been shown in the SVZDL

at 12 (Fig. 2D) and 24 hours after injury in both IL and CLhemispheres.

Neonatal Ischemia Induces Cell Proliferation inBoth Ipsilateral and Contralateral Hemispheres

Cell proliferation in both the SVZDL and SGZ wasdetermined using Ki-67 immunostaining, which visualizesall cells during late G1, S, M, and G2 phases of the cell

cycle, as previously reported (12). As shown in Figure 3, 3peaks of Ki-67-positive nuclei could be detected in boththe IL SVZDL and SGZ at 48 hours (P9), 7 days (P14), and28 days (P35) after lesioning. In the SVZDL, the numbers ofKi-67-positive nuclei were similar in the first 2 peaks,whereas the number in the third peak was lower. In the SGZ,the first peak was the most important of the 3. The 2 firstpeaks of proliferation were also detected in the CL hemi-sphere, but the number of Ki-67-positive nuclei was reducedby 50% compared with the IL side. The third peak (thesmallest) of Ki-67-positive nuclei was seen in both ILSVZDL and SGZ and only in the CL SVZDL. Some Ki-67-positive nuclei were still detected in both SVZDL at 6 weeksafter ischemia (Fig. 3).

Using a mouse monoclonal against nestin, an inter-mediate filament protein that is abundant and transientlyexpressed by multipotential neural stem cells and someprogenitors (19), we found that all rats showed increasednestin staining within the IL SVZDL at 48 hours and 7 days

FIGURE 2. Cell death shown by the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay after neonatalstroke in postnatal day (P) 7 rats. (A) Presence of TUNEL-positive nuclei in the parietal cortex at 24 hours (upper) and 10 days(lower) after reperfusion. Note the large number of TUNEL-positive nuclei in layers IVYV and also some preserved tissue at 24hours (arrow indicates unaffected columnar tissue). (B) Several TUNEL-positive nuclei can be still observed at 10 days around thecortical cavity (asterisk). (C) Quantification of TUNEL-positive nuclei in the lesion and the white matter (WM and layer VI) in boththe ipsi- and contralateral hemispheres of ischemic animals (n = 5). Values are expressed as means T SD (per mm2); at each agevalues were compared with analysis of variance (*, p G 0.05 vs contralateral values). (D, E) TUNEL-positive nuclei in the ipsilateraldorsolateral subventricular zone (SVZDL) and control SVZDL at 12 hours after ischemia. Scale bar = 25 Km.




after lesioning (Fig. 4), compared with the control SVZDL

but not thereafter. In the CL SVZDL, nestin was onlydetected at 48 hours (Fig. 4A). Re-expression of nestin hasalso been observed in reactive astrocytes after injury (20)and indeed was strongly expressed in the lesion (mainly inthe penumbra (Fig. 4C) at 48 hours and around the cavity at7 (Fig. 4F) and 14 (not shown) days.

In the SVZDL, Ki-67-positive cells were double-labeled with nestin at 48 hours after ischemia (Fig. 4G),suggesting proliferation of stem cells or progenitors. Aspreviously reported, double Ki-67 and GFAP immunostain-ing demonstrated that most of Ki-67-positive nuclei wereastrocytes at 48 hours after ischemia in the SVZDL, in theSGZ, and around the lesion (12). Some of them can be stilldetected at 7 days after ischemia in both SVZDL (notshown), but not thereafter.

Neonatal Ischemia Enhances Neurogenesis inBoth Ipsilateral and Contralateral Hemispheres

Immunofluorescence for DCx, a marker uniquelyexpressed by migrating neuronal precursors and immatureneurons (21, 22), demonstrated a faint band of labelingconcentrated in the SVZDL of P14 (Fig. 5A), P21, and P28(not shown) control rats. In animals exposed to ischemia,DCx was substantially increased in the IL SVZDL as earlyas 7 days (Fig. 5C) and markedly observed 10 to 14 days(Fig. 5D) after lesioning. DCx was also increased, but lessextensively, in the CL SVZDL (Fig. 5B, E). Migrating DCx-labeled neurons were conspicuously shown in both the CLand IL sides (arrows in Fig. 5D, E) at 10 to 14 days after

ischemia. At 21 days, DCx labeling was only detected in aband, which includes the IL side underlying the SVZDL andthe edge of the cavity (Fig. 5F), as discrete chains or clustersof DCx-labeled neurons exhibiting long processes (doublearrows in Fig. 5FYH). A great number of DCx-positive cellswere also detected in both SGZs, but without conspicuousdifferences between IL and CL hemispheres (not shown).Four weeks after ischemia, a few DCx-positive neurons werestill detected in the IL side, but none were detected afterthat.

As in our model, the lesion evolved in a cavity,demonstrating an ongoing damage process, and we per-formed double DCx and TUNEL assays at different timepoints of recovery. No double DCx-TUNEL-positive cellswere detected. In contrast, DCx-labeled cells were shownnear the lesion where numerous TUNEL-positive nucleiwere observed at 7 days after ischemia (Fig. 5IYK),suggesting that newly damaged neurons may survive at least1 week before disappearance.

Neonatal Ischemia Enhances Oligogenesis inBoth Ipsilateral and Contralateral Hemispheres

In age-matched control animals, only faint stainingwith NG2, which is a sulfate proteoglycan present inoligodendrocyte progenitors, was detected. After ischemia,NG2 labeling increased in both IL and CL SVZDL from 48hours (not shown) up to 14 days (Fig. 6AYC). NG2-positivecells represented approximatively 30% to 50% of nestin-positive cells in the SVZDL (Fig. 6AYC) and in the border ofthe lesion (Fig. 6DYF) at 14 days after ischemia. NG2

FIGURE 3. Proliferative cells in the dorsolateral subventricular zone (SVZDL) and subgranular zone (SGZ). (A) Representative Ki-67-labeled sections demonstrating the distribution of Ki-67-positive nuclei in an ischemic animal in both ipsi- and contralateralhemispheres. Scale bar = 50 Km. (B) Quantification of Ki-67-positive nuclei in the SVZDL (upper) and SGZ (lower) in bothhemispheres in ischemic (n = 5) and control rats (naive age-matched, n = 5). Note that 3 peaks of Ki-67-positive nuclei could bedetected at 48 hours (postnatal day [P] 9), 7 days (P14), and 28 (P35) days after ischemia in both ipsi- and contralateralhemispheres compared with control. Values are expressed as means T SD (per 0.1 mm2); at each age values were compared withanalysis of variance (*, p G 0.05 vs naive).




labeling was still detectable at 4 weeks (Fig. 6G) and to alesser extent at 6 weeks after ischemia.

DISCUSSIONOur results demonstrate that in neonatal rats,

ischemia with a reperfusion phase in the CCA stimulatescell proliferation and neurogenesis in the SVZDL; our mostimportant observation is that most of these regenerativechanges are detected both in the IL and CL hemispheres.These findings are surprising because this model of neonatalstroke was known to induce long-term brain damageexclusively in the IL hemisphere (11, 23).

In this neonatal stroke model, blood recanalization inthe left CCA was induced by the release of the clip. Atthis time, a reperfusion phase occurs via anastomosesbetween anterior, middle, and posterior arteries in theMCA territory as demonstrated by black ink perfusion(S. Renolleau and L.-M. Joly, unpublished observation,2004). This reperfusion has been detected before braininfarction was visible (pale zone at 48 hours). In addition,

we recently reported that reperfusion blocks edema volumeusing diffusion-weighted magnetic resonance imagingcompared with animals with a permanent CCA occlusion(24). Neonatal stroke induces a transiently reduced bodyweight associated with very significantly reduced hemi-spheric areas and a lesion in the IL side from 24 hours upto 4 weeks of recovery. Reduced CL hemispheric areaswere also found during the first 2 days after injury butdisappeared 1 week after ischemia. This transient differ-ence suggests that either body weight loss induced atransient growth restriction and/or that loss of cellconnections occurred, particularly during the first 24 to48 hours of recovery. Reduced hemispheric areas arerelated to the presence of TUNEL-positive nuclei leadingto cell death in both hemispheres. Several studies reportedCL cell injury/death after HI (14, 25), as well as thepresence of an apoptotic marker, cleaved poly(ADP-ribose)polymerase (16). In the IL cortex, TUNEL-positive nucleiwere observed in columnar patterns oriented at a right angleto the pial surface (Fig. 2), with alternative regions of

FIGURE 4. Nestin immunostaining in the dorsolateral subventricular zone (SVZDL) and border of the lesion of ischemic animalsat 48 hours (AYC) and 7 days (D, E) of recovery. Note a more pronounced nestin staining in the ipsi- compared with thecontralateral SVZDL at 48 hours and 7 days after ischemia. Note also a high nesting staining in the border of the lesion (doublewhite arrow indicates the process of a probably neural stem cell and asterisks indicate reactive astrocytes). Such staining is stillobserved around the cavity (Ca) at 7 days after ischemia. (G) The double immunofluorescence assay demonstrates thatnumerous nestin-positive cells are newly progenitors labeled by both Ki-67 and nestin (white arrows). Scale bars = (AYF) 40 Km;(G) 20 Km.




preserved tissue as observed at 24 hours after HI in P3 (26)and P7 (27Y31) rats. This initial columnar neurodegenerationseen in the neocortex correlates with areas of incompleteblood supply and altered NADH metabolism as described

after HI in the developing brain (27) and has also beobserved in our model using diffusion in magnetic resonanceimaging (S. Renolleau and L.-M. Joly, unpublished obser-vation, 2004).

FIGURE 5. Doublecortin (DCx) immunostaining in control and ischemic lesioned animals. (AYC) In control (postnatal day [P]14) brain, DCx expression is concentrated in the border of the dorsolateral subventricular zone (SVZDL). At 7 days after ischemia(pI), subtle increased DCx immunoreactivity is apparent in the SVZ contralateral to the stroke. The injured ipsilateral hemisphereshows extensive DCx labeling. (D, E) DCx expression in the SVZ of both contra- and ipsilateral hemispheres of P17 to P21 rats(10Y14 days after ischemia). Note the presence of DCx-labeled migrating neuroblasts (white arrowheads) from both SVZDL.(FYH) In P28 ischemic animals (21 days after ischemia), DCx-positive neuroblasts with long processes (white arrows) and clustersor chains of DCx-positive neuroblasts (white double arrows) could still be observed close to the lesion. Scale bars = (F) 100 Km;(AYE, G, H) 50 Km. (IYK) The presence of DCx-positive neurons among terminal deoxynucleotidyl transferase dUTP nick-endlabeling (TUNEL)-positive nuclei in the border of the lesion at 7 days after ischemia. Scale bar = 25 Km.




Cell death was also detected in both IL and CLSVZDL at 12 and 24 hours after ischemia, suggesting thatsome progenitors are vulnerable to ischemia according tofindings reported in the Rice-Vannucci model (32, 33). Celldeath was followed by proliferation evidenced by Ki-67protein upregulation and nestin immunostaining and wasdetected in both the SVZDL and SGZ. In the CL side, cellproliferation has not been described in studies using theRice-Vannucci model of HI but agrees with recoveryobserved in the CA1 subfield of the hippocampus after abrief period of anoxia in P1 rats (34). Two peaks of cellproliferation were observed in both the SVZDL and SGZ at48 hours and 7 days after ischemia. At 48 hours afterischemia, Ki-67-positive cells coexpressed GFAP or NG2, asulfate proteoglycan present in oligodendrocyte progenitors/polydendrocytes (35), giving rise to astrocytes and pre-oligodendrocytes, as we previously reported in the IL andCL corpus callosum (12). At 7 days after ischemia only afew Ki-67-positive/GFAP-positive cells were found in theSVZDL, suggesting that the second peak corresponds mainlyto neurogenesis and oligogenesis in agreement with studiesshowing that the activated neural stem/progenitor cells in theSVZ have an enhanced capacity to produce both neurons andoligodendrocytes (9). The third Ki-67 peak found at 28 daysafter ischemia mainly contained newly generated oligoden-droglia as previously reported (8, 36, 37).

To detect possible neurogenesis after neonatal stroke,we examined the expression of DCx, a marker of migratingneuroblasts, in the SVZ and peri-infarct territory. ExtendedDCx labeling suggests an extent of the SVZDL as previouslyreported at the level of the midstriatum (8, 10). IncreasedDCx was detected from 7 up to 14 days after ischemia inboth the IL and CL hemispheric SVZDL in contrast to its

restriction to IL SVZ in HI-treated P7 rats (8, 9) and P10mice (10). At 21 days of recovery, DCx-expressing cellswere identified only in the IL hemisphere in a ring includingthe cavity edge and underlying SVZDL. A distinctive featureof DCx-positive cells was their frequent arrangement inchains, and most of them displayed long processes, indicat-ing their possible migration towards the lesion in the IL side.Interestingly, these DCx-positive immature neurons maysurvive during several days as they migrate in the lesionamong numerous TUNEL-positive cells. However, theseneurons do not mature and die because a detrimentalongoing process occurs after neonatal stroke, leading to acavity in all layers of the IL parietal cortex (12, 23).Similarly, only 1% of mature neurons (NeuN-positive) couldbe detected between 5-bromo-2[prime]-deoxyuridine-immunopositive cells in the infarcted cortex at 14 days af-ter HI (31, 38). We also previously observed that newGABAergic interneurons are present 2 weeks after ischemiabut only as chains of immature neurons (39), suggesting thatthey do not mature.

An extensive inflammatory reaction in response tofocal ischemia evolves in the neonatal brain (40) and may beinvolved in accelerating the destructive process. We recentlydemonstrated that persistent microglial infiltration occursover several days after ischemic injury in P7 rat pups andthat macrophages engulf immature oligodendrocytes (12).Therefore, this chronic component of immunoinflammationprobably also prevents differentiation from neural precursorsinto mature neurons.

As in our previous work (12), animals were notseparated by gender but studies during the past few yearsindicate that neonatal HI or ischemia induced gender-specific mechanisms of cell death in mice (41, 42) and rats

FIGURE 6. Confocal images of double immunofluorescence with antibodies against nestin and NG2 at 14 days after neonatalstroke. (AYC) Nestin/NG2-positive cells are present in the dorsolateral subventricular zone (SVZDL) (white arrow). (DYF) In thelesion, both nestin/NG2-positive cells (double white arrow) and nestin-positive cells (white arrow) are present. Nestin-positiveand NG2-negative cells might represent immature neurons. Scale bar = 25 Km. (G) NG2-positive cells are still detected at 4 to 6weeks after ischemia in the ipsilateral SVZDL.




(18, 43). Although further experiments are necessary toevaluate the time course and density of proliferative cellsin each gender, infarct volumes of untreated animals arenot significantly different between males and females (18).

In conclusion, our results provide evidence that theinjured immature forebrain has regenerative potential.These repair processes occur not only in the IL but alsoin the CL sides, although unilateral ischemic injury wasperformed. In the IL hemisphere, repair processes arelimited in part because of a sustained important inflamma-tory component. In contrast, in the CL hemisphere, wesuppose that the intrinsic ability for neurologic self-repairis sufficient to replace dying cells, allowing anatomicalrecovery. What happens in the CL hemisphere mightrepresent part of the different mechanisms occurring inpreconditioning after a mild preinjury and may be relatedto changes in the IL hemisphere. Nevertheless, injury in theleft cortex may induce bradycardia or tachycardia inanimals, thus possibly related to hypotension, which leadsto a more or less ischemic damage in the CL hemisphere,giving rise to more or less neogenesis and subsequentlymore or less recovery. A greater understanding of themolecular mechanisms underlying recovery in the CLhemisphere after mild, moderate, or severe ischemic injurymay offer great prospects for the development of brainreparative therapies.

ACKNOWLEDGMENTSThe authors thank the Imaging Center (IFR 83, Uni-

versity Pierre and Marie Curie, Paris, France) for confocalimages.

REFERENCES1. Nelson KB, Lynch JK. Stroke in newborn infants. Lancet Neurol 2004;

3:150Y582. Edwards AD, Azzopardi DV. Perinatal hypoxia-ischemia and brain

injury. Pediatr Res 2000;47:431Y323. Zheng T, Rossignol C, Leibovici A, et al. Transplantation of multipotent

astrocytic stem cells into a rat model of neonatal hypoxic-ischemicencephalopathy. Brain Res 2006;1112:99Y105

4. Park KI, Himes BT, Stieg PE, et al. Neural stem cells may be uniquelysuited for combined gene therapy and cell replacement: Evidencefrom engraftment of neurotrophin-3-expressing stem cells in hypoxic-ischemic brain injury. Exp Neurol 2006;199:179Y90

5. Lichtenwalner RJ, Parent JM. Adult neurogenesis and the ischemicforebrain. J Cereb Blood Flow Metab 2006;26:1Y20

6. Levison SW, Rothstein RP, Romanko MJ, et al. Hypoxia/ischemiadepletes the rat perinatal subventricular zone of oligodendrocyteprogenitors and neural stem cells. Dev Neurosci 2001;23:234Y47

7. Brazel CY, Romanko MJ, Rothstein RP, et al. Roles of the mammaliansubventricular zone in brain development. Prog Neurobiol 2003;69:49Y69

8. Ong J, Plane JM, Parent JM, et al. Hypoxic-ischemic injury stimulatessubventricular zone proliferation and neurogenesis in the neonatal rat.Pediatr Res 2005;58:600Y606

9. Yang Z, Levison SW. Hypoxia/ischemia expands the regenerativecapacity of progenitors in the perinatal subventricular zone. Neuro-science 2006;139:555Y64

10. Plane JM, Liu R, Wang TW, et al. Neonatal hypoxic-ischemic injuryincreases forebrain subventricular zone neurogenesis in the mouse.Neurobiol Dis 2004;16:585Y95

11. Renolleau S, Aggoun-Zouaoui D, Ben-Ari Y, et al. A model of transientunilateral focal ischemia with reperfusion in the P7 neonatal rat:

Morphological changes indicative of apoptosis. Stroke 1998;29:1454Y60; discussion 1461

12. Biran V, Joly LM, Heron A, et al. Glial activation in white matterfollowing ischemia in the neonatal P7 rat brain. Exp Neurol 2006;199:103Y12

13. Rice JE III, Vannucci RC, Brierley JB. The influence of immaturityon hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9:131Y41

14. Hill IE, MacManus JP, Rasquinha I, et al. DNA fragmentationindicative of apoptosis following unilateral cerebral hypoxia-ischemiain the neonatal rat. Brain Res 1995;676:398Y403

15. Nakajima W, Ishida A, Lange MS, et al. Apoptosis has a prolonged rolein the neurodegeneration after hypoxic ischemia in the newborn rat.J Neurosci 2000;20:7994Y8004

16. Martin SS, Perez-Polo JR, Noppens KM, et al. Biphasic changes inthe levels of poly(ADP-ribose) polymerase-1 and caspase 3 in theimmature brain following hypoxia-ischemia. Int J Dev Neurosci 2005;23:673Y86

17. Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates. NewYork, NY: Academic Press, 1982

18. Renolleau S, Fau S, Goyenvalle C, et al. Specific caspase inhibitorQ-VD-OPh prevents neonatal stroke in P7 rat: A role for gender.J Neurochem 2007;100:1062Y71

19. Messam CA, Hou J, Major EO. Coexpression of nestin in neural andglial cells in the developing human CNS defined by a human-specificanti-nestin antibody. Exp Neurol 2000;161:585Y96

20. Schmidt-Kastner R, Humpel C. Nestin expression persists in astrocytesof organotypic slice cultures from rat cortex. Int J Dev Neurosci 2002;20:29Y38

21. Francis F, Koulakoff A, Boucher D, et al. Doublecortin is adevelopmentally regulated, microtubule-associated protein expressedin migrating and differentiating neurons. Neuron 1999;23:247Y56

22. Gleeson JG, Lin PT, Flanagan LA, et al. Doublecortin is a micro-tubule-associated protein and is expressed widely by migrating neurons.Neuron 1999;23:257Y71

23. Joly LM, Benjelloun N, Plotkine M, et al. Distribution of poly(ADP-ribosyl)ation and cell death after cerebral ischemia in the neonatal rat.Pediatr Res 2003;53:776Y82

24. Fau S, Po C, Mariani J, et al. Effect of reperfusion after cerebralischemia in P7 rat using diffusion-weighted MRI monitoring (Abstract).Proceedings of Neuroscience 2006, Society for Neuroscience, held inAtlanta, GA, October 14Y18, 2006. Washington, DC: Society forNeuroscience, 2006:282.9

25. Joashi UC, Greenwood K, Taylor DL, et al. Poly(ADP ribose)polymerase cleavage precedes neuronal death in the hippocampus andcerebellum following injury to the developing rat forebrain. Eur JNeurosci 1999;11:91Y100

26. Sizonenko SV, Sirimanne E, Mayall Y, et al. Selective corticalalteration after hypoxic-ischemic injury in the very immature rat brain.Pediatr Res 2003;54:263Y69

27. Welsh FA, Vannucci RC, Brierley JB. Columnar alterations of NADHfluorescence during hypoxia-ischemia in immature rat brain. J CerebBlood Flow Metab 1982;2:221Y28

28. Vannucci RC. Experimental models of perinatal hypoxic-ischemic braindamage. APMIS Suppl 1993;40:89Y95

29. Towfighi J. Mauger D, Vannucci RC, et al. Influence of age on thecerebral lesions in an immature rat model of cerebral hypoxia-ischemia:A light microscopic study. Brain Res Dev Brain Res 1997;100:149Y60

30. Northington FJ, Ferriero DM, Graham EM, et al. Early neurodegenera-tion after hypoxia-ischemia in neonatal rat is necrosis while delayedneuronal death is apoptosis. Neurobiol Dis 2001;8:207Y19

31. Ikeda T, Iwai M, Hayashi T, et al. Limited differentiation to neuronsand astroglia from neural stem cells in the cortex and striatum afterischemia/hypoxia in the neonatal rat brain. Am J Obstet Gynecol 2005;193:849Y56

32. Romanko MJ, Rothstein RP, Levison SW. Neural stem cells inthe subventricular zone are resilient to hypoxia/ischemia whereasprogenitors are vulnerable. J Cereb Blood Flow Metab 2004;24:814Y25

33. Rothstein RP, Levison SW. Gray matter oligodendrocyte progenitorsand neurons die caspase-3 mediated deaths subsequent to mild perinatalhypoxic/ischemic insults. Dev Neurosci 2005;27:149Y59




34. Daval JL, Pourie G, Grojean S, et al. Neonatal hypoxia triggerstransient apoptosis followed by neurogenesis in the rat CA1 hippo-campus. Pediatr Res 2004;55:561Y67

35. Komitova M, Perfilieva E, Mattsson B, et al. Enriched environment afterfocal cortical ischemia enhances the generation of astroglia and NG2 positivepolydendrocytes in adult rat neocortex. Exp Neurol 2006;199:113Y21

36. Levison SW, Druckman SK, Young GM, et al. Neural stem cells in thesubventricular zone are a source of astrocytes and oligodendrocytes, butnot microglia. Dev Neurosci 2003;25:184Y96

37. Zaidi AU, Bessert DA, Ong JE, et al. New oligodendrocytes aregenerated after neonatal hypoxic-ischemic brain injury in rodents. Glia2004;46:380Y90

38. Fagel DM, Ganat Y, Silbereis J, et al. Cortical neurogenesis enhancedby chronic perinatal hypoxia. Exp Neurol 2006;199:77Y91

39. Benjelloun N, Represa A, Joly L-M, et al. GABAergic neurogenesis

following hypoxia-ischemia in the neonatal rat (P7). J Cereb BloodFlow Metab 2003;23:426

40. Benjelloun N, Renolleau S, Represa A, et al. Inflammatory responses inthe cerebral cortex after ischemia in the P7 neonatal rat. Stroke 1999;30:1916Y23; discussion 1923Y24

41. Hagberg H, Wilson MA, Matsushita H, et al. PARP-1 gene disruptionin mice preferentially protects males from perinatal brain injury.J Neurochem 2004;90:1068Y75

42. Zhu C, Xu F, Wang X, et al. Different apoptotic mechanisms areactivated in male and female brains after neonatal hypoxia-ischaemia.J Neurochem 2006;96:1016Y27

43. Nijboer CH, Groenendaal F, Kavelaars A, et al. Gender-specificneuroprotection by 2-iminobiotin after hypoxia-ischemia in the neonatalrat via a nitric oxide independent pathway. J Cereb Blood Flow Metab2007;27:282Y92




Bilateral Changes After Neonatal Ischemia in the P7 Rat Brain

Documents

Transcript of Bilateral Changes After Neonatal Ischemia in the P7 Rat Brain