N

Ao

SIa

b

c

d

e

f

g

h

a

ARR1AA

KAEGGKLNNP

1

ntaocten

aT

t

0h

ARTICLE IN PRESSG ModelSR-3518; No. of Pages 9

Neuroscience Research xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Neuroscience Research

jo u r n al hom ep age: www.elsev ier .com/ locate /neures

mygdala kindling induces nestin expression in the leptomeningesf the neocortex

hogo Ninomiyaa,h, Shigeyuki Esumia, Kunimasa Ohtab, Takaichi Fukudac, Tetsufumi Itod,taru Imayoshie, Ryoichiro Kageyamae, Toshio Ikedaf, Shigeyoshi Itoharag, Nobuaki Tamamakia,∗

Department of Morphological Neural Science, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, JapanDepartment of Developmental Neurobiology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, JapanDepartment of Anatomy and Neurobiology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, JapanDepartment of Anatomy and Neuroscience, Faculty of Medical Sciences, University of Fukui, Fukui, JapanInstitute for Virus Research, Kyoto University, Kyoto, JapanNational Institute for Longevity Sciences, Aichi, JapanLaboratory for Behavioral Genetics, RIKEN, BSI, Wako, JapanCoruse of Physical Therapy, Department of Rehabilitation, Faculty of Nursing and Welfare, Kyushu University of Nursing and Social Welfare, Kumamoto, Japan

r t i c l e i n f o

rticle history:eceived 1 October 2012eceived in revised form3 December 2012ccepted 17 December 2012vailable online xxx

eywords:rachnoidpilepsy

a b s t r a c t

Nestin is an intermediate filament found in neurogenic progenitors and non-neuronal cells. Nestin-immunoreactivity (IR) in the brain often increases after brain damage. Here we show that amygdalakindling, which mimics the epileptic seizures, also induces nestin expression in the brain. Nestin-IR wasgreatly enhanced in the leptomeninges (pia and arachnoid maters) and neocortical parenchyma, but notmuch in the SVZ around the lateral ventricle, SGZ in the dentate gyrus, or the endothelial progenitor cellsof blood vessels, fimbria, or choroid plexus after kindling. Electron microscopy revealed that nestin-IRin the leptomeninges was localized to granule cells, where it co-localized with GAD67-IR after electricalstimulation. The nestin-positive granule cells in the leptomeninges, especially around the emissary vein,were proliferative. However, nestin-IR in the neocortical parenchyma was expressed in NG2 glia and did

ABAAD67indlingeptomeningesestineurogenesis

not co-localize with GAD67-IR. Deletion of nestin-positive cells resulted in a high susceptibility to electri-cal stimulation. Consequently, almost all of the mice died or dropped out during kindling progression in20 days, from naturally generated epileptic seizure or exhaustion. We speculate that the nestin-positivecells activated by amygdala kindling may involve in the protection of the brain from epilepsy.

© 2013 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved.

ia mater

. Introduction

Nestin is an intermediate filament that is often expressed ineurogenic sites (Lendahl et al., 1990). Nestin-IR is useful to iden-ify the neurogenic progenitors in the embryonic brain and in thedult brain, i.e., the SVZ around the lateral ventricle and the SGZf the dentate gyrus. However, non-neuronal cells, like NG2 glialells (Kronenberg et al., 2005) the endothelial progenitor cells and

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

he pericytes of blood vessels (Dahlstrand et al., 1995; Kobayashit al., 1998; Suzuki et al., 2010; Nakagomi et al., 2011), also expressestin in the adult brain. Notably, brain damage caused by ischemia,

∗ Corresponding author at: Department of Morphological Neural Science, Gradu-te School of Medical Sciences, Kumamoto University, Kumamoto 860-8556, Japan.el.: +81 96 373 5298; fax: +81 96 373 5300.

E-mail addresses: tamamaki@kumamoto-u.ac.jp,amamakidesu@gmail.com (N. Tamamaki).

168-0102/$ – see front matter © 2013 Elsevier Ireland Ltd and the Japan Neuroscience Sttp://dx.doi.org/10.1016/j.neures.2012.12.006

hemorrhage, epilepsy, or mechanical injury is known to inducenestin expression (Nakagomi et al., 2011; Sgubin et al., 2007;Crespel et al., 2005; Decimo et al., 2011). However, distributionof nestin-IR depends on the type of brain damage, respectively.Therefore, it is necessary to investigate the distribution of nestin-IRin each experimental paradigm of adding brain damage to experi-mental animals.

Epilepsy is a serious neurological disorder that affects 1% of thepopulation. Seizures are triggered at a focus formed after braindamage or by intrinsic unknown mechanisms and can recur persis-tently for much of a patient’s life. Moreover, epilepsy can becomeintractable, i.e., resistant to antiepileptic drugs, leading to brainatrophy. To study the progress of epilepsy in the brain, animalmodels that mimic epilepsy are produced by artificially kindling

induces nestin expression in the leptomeninges of the neocortex.

seizures through electrical stimulation of the amygdala. In thispaper, we investigated whether kindling in mice is accompaniedby nestin expression, and whether the distribution of nestin-IRis similar to that after other kind of brain damage. Moreover, we

ociety. All rights reserved.

ING ModelN

2 ence R

ee

2

2

Clmwaeb

tmait[

Fostited

ARTICLESR-3518; No. of Pages 9

S. Ninomiya et al. / Neurosci

xamined whether the nestin expression protects the brain frompileptic seizure.

. Materials and methods

.1. Animals and kindling procedure

Three strains of 10-week-old mice were used: Wild-type57BL/6J mice were purchased from SLC (Shizuoka Japan); NSE-

stopl-DTA (diphtheria toxin A-fragment)/NesCreERT2 (Nes-DTA)ice were prepared by mating homozygous NSE-lstopl-DTA miceith homozygous NesCreERT2 mice (Imayoshi et al., 2006, 2008);

nd the Nestin-GFP transgenic mice were kind gift (Yamaguchit al., 2000). All the animal experiments in this study were approvedy the Animal Experiment Committee of Kumamoto University.

Fig. 1A shows the experimental paradigm. As untreated con-rols, 3 C57/B6 wild-type mice, 3 Nes-DTA mice, and 3 Nestin-GFP

ice were quickly anesthetized with ketamine (100 mg/kg). An

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

dequate depth of anesthesia was verified by the absence of reflexesn response to toe pinches. Mice were then perfused with a fixa-ive (4% paraformaldehyde in 0.1 M phosphate buffered 0.9% salinePBS]) by cardiac puncture.

ig. 1. (A) Schematic diagram of the experimental paradigm. [1] On the first day of the ene-week recovery period, the mouse brain was stimulated once per day (2 mA maximumtimulation) were administered for [3] three days or [4] seven days, or longer upto 28 dayo Nes-DTA and wild-type mice in another elimination experiments (Figs. 5–7). (B) Schemnto the brain at 1.6 mm caudal, 3 mm lateral, and 4 mm deep from the bregma. The refehe bregma. (C) The seizure class was rated from class 0 to 5 during kindling progression.pileptic seizure during kindling progression. The duration of epileptic seizure reached a muring the epileptic seizure induced by electrical stimulation. Arrows indicate the time p

PRESSesearch xxx (2013) xxx–xxx

For kindling, a pair of stimulation electrodes was implanted intothe brain of 37 wild-type, 32 Nes-DTA, and 3 Nestin-GFP mice. Themice were anesthetized as described above and held in a stereo-taxic apparatus. The points for electrode insertion were mappedon the skull. The electrode was advanced into the brain at 1.6 mmcaudal, 3 mm lateral, and 4 mm deep from the bregma (Fig. 1B). Thereference electrode was placed in the neocortex at 1.6 mm caudaland 1 mm lateral to the bregma.

Three wild-type mice with implanted electrodes were subjectedto electrical stimulation (2 mA maximum, 60 Hz square wave for2 s) and given BrdU twice a day (50 mg/kg i.p., 30 min and 2 hafter the electrical stimulation) for 28 days. At each applicationof electrical stimulation, the progression of kindling in the wild-type mice was scored according to seizure class from 0 to 5 for 20days (Fig. 1C) (Racine, 1972). The duration of epilepsy was mea-sured based on the behavior (Fig. 1D, Supplemental Movie) andelectroencephalography (EEG) (Fig. 1E). The EEG was recorded byreconnecting the electrical stimulation electrodes to a preamplifier(BSA-ECG, Unique-Medical Co. Tokyo, Japan) and a pen-recorder.

induces nestin expression in the leptomeninges of the neocortex.

As controls, 3 wild-type mice without electrodes were given BrdUtwice a day (50 mg/kg i.p. with a 1.5 h interval) for 28 days. Afterthe BrdU injection on the 28th day, the 6 wild-type mice with andwithout stimulation electrodes, were anesthetized and perfused

xperiment, electrodes were implanted in the amygdala and neocortex. [2] After a, 60 Hz square wave for 2 s) and BrdU (50 mg/kg, 30 min and 2 h after the electrical

s. Additionally tamoxifen (1 mg dissolved in sunny oil/day orally) was administeredatic diagram showing the location of the electrodes. The electrode was advanced

rence electrode was placed in the neocortex at 1.6 mm caudal and 1 mm lateral to The seizure reached class 5 in 5.04 ± 0.59 days in 3 wild-type mice. (D) Duration of

aximum by 10 days in 3 wild-type mice. (E) Electroencephalogram (EEG) recordedoints of movie recording, the initiation of walking, and touches with the finger.

ING ModelN

ence R

wh

f2

ww6(ae1k

2

a((pg1b

Fmaitpscbr(s

ARTICLESR-3518; No. of Pages 9

S. Ninomiya et al. / Neurosci

ith fixative. The brain was removed and prepared for immuno-istochemistry.

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.neures.012.12.006.

To examine the effects of DTA expression by histology, 3ild-type mice and 3 Nes-DTA mice with electrodes were perfusedith fixative after a one-week recovery period (Fig. 1A). For the

wild-type mice and 6 Nes-DTA mice with electrodes, tamoxifen1 mg dissolved in sunny oil/day orally) and BrdU (50 mg/kg, 30 minnd 2 h after electrical stimulation) were administered, as well aslectrical stimulation, every day for 3 days (3 mice of each type) or

week (3 mice of each type). All the mice were anesthetized withetamine and perfused with fixative as described above.

.2. Immunohistochemistry

The fixed brains were used for immunohistochemistry. Thentibodies used in this experiment were anti-cleaved caspase-3rabbit 1/200 Cell Signaling), anti-BrdU (rat 1/200 Abcam), anti-DTAmouse 1/200-1/1000 Abcam), anti-GAD67 (mouse 1/1000 Milli-

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

ore), anti-GAD67 (rabbit 1/200 Abcam), anti-GFP (rabbit 1 �g/ml,uinea pig 1 �g/ml; Tamamaki et al., 2000), anti-GFAP (rabbit/2000 Dako), anti-nestin (mouse 1/300-1000 BD), anti-NG2 (rab-it 1/1000 Millipore). Secondary labeled antibodies were raised

ig. 2. Nestin immunohistochemistry in the mouse brain. (A) Immunohistochemistry ononoclonal antibody. DAB was used as the chromogen in Fig. 2. (B) Nestin immunohist

nti-nestin monoclonal antibody. (C–E) Nestin immunohistochemistry on a coronal sectionsertion (C), plus 3 (D), or 7 applications of electrical stimulation (E). (F) Nestin-positive

ricular zone (SVZ) around the lateral ventricles. (H) Nestin-positive cells in the subgranurogenitor cells of blood vessels in the parenchyma of the neocortex (I) and of blood vessections. Glial processes were observed in contact with the soma of pyramidal neurons. (L)ells (arrowheads) were visible in this semi-thin section. (M) Basal lamina (indicated by ary electron microscopy. (N) The leptomeninges observed by electron microscopy. The leectangle indicates the area shown in Fig. 2M. Arrowheads indicate the nestin-positive grA) applies to (B–E), and bar in (F) applies to (G–L). BV: blood vessel; Cap: capillary; Ch: cubventricular zone.

PRESSesearch xxx (2013) xxx–xxx 3

in donkey or goat and recognized a constant region of mouseIgG, or rabbit IgG, or rat IgG. Immunoreactive sites were visu-alized by the fluorescence of Alex-488 or Alex-594, or by theABC/diaminobenzidine (DAB) reaction as described previously(Tamamaki et al., 2003).

2.3. Electron microscopy

Six wild-type mice were used for electron microscopic (EM)observation of the pia and the arachnoid membrane. Stimula-tion electrodes were implanted in the brain of 3 mice, and thesemice were stimulated once a day for 7 days, as described above.The 3 mice with stimulation electrodes and 3 control mice wereanesthetized as described above, then perfused with fixative forEM (4% paraformaldehyde, 0.1% glutaraldehyde in PBS). After 2 hof fixation, the brain tissue was trimmed into small blocks andpost-fixed with 1% OsO4 in PBS. After dehydration in graded alco-hol/water solutions, the brain blocks were embedded in plasticresin (Epon-812, Oken, Tokyo, Japan). The blocks were trimmed,cut into semi-thin sections (0.5–1-�m thick), and observed undera light microscope. Selected plastic blocks were then trimmed fur-

induces nestin expression in the leptomeninges of the neocortex.

ther, and the area of interest was sectioned into ultrathin sections(60-nm thick). After staining with uranyl acetate and lead citrate,the ultra-thin sections were observed under an electron micro-scope (H-7500, Hitachi, Japan).

a coronal section of a wild-type mouse brain without primary mouse anti-nestinochemistry on a coronal section of wild-type mouse brain with a primary mousen of wild-type mouse brain after the one-week recovery period following electrodemyeloid progenitors in the choroid plexus. (G) Nestin-positive cells in the subven-lar zone (SGZ) in the dentate gyrus. (I–J) Nestin-positive pericytes and endothelialels of the leptomeninges (J). (K) Nestin-positive satellite glia observed in semi-thin

Nestin-positive granule cells in the leptomeninges. Several nestin-positive granulerowheads), the boundary between the neocortex and the leptomeninges, observedptomeninges consist of plexuses of vessels, fibroblast cells, and granule cells. Theanule cells. Scale bars are 1 mm in A, 10 �m in (F) and (N), and 1 �m in (M). Bar in

horoid plexus; Pia: pia matter; Py: pyramidal neuron; SGZ: subgranular zone; SVZ:

ARTICLE ING ModelNSR-3518; No. of Pages 9

4 S. Ninomiya et al. / Neuroscience R

Fig. 3. (A) Western blot analysis of Nestin and GAPDH expression in the neocortexof 3 control mice, 3 mice after 3 days of kindling, and 3 mice after 7 days of kin-dling, and in the E15.5 embryonic cortex. Nestin (350–170 kDa) and GAPDH (36 kDa)bands were detected. The predicted nestin band was at 170 kDa (arrowhead), andthe larger bands (asterisk) indicate nestin with post-translational modifications.The same amount of lysate was loaded in each column. (B) Quantitative analysis ofthe amount of unmodified nestin protein (around 170 kDa, see the ab11306: rat401antibody manual) normalized to the GAPDH band shows that it increased about four-fr

2

esst6bSftAaTamLwBsbGc

old after 3 days of kindling (P < 0.05, P < 0.01, t-test), but after 7 days of kindling wasestored to almost the control level.

.4. Immunoblotting

Immunoblotting was performed as described previously (Ohtat al., 2004). Neocortical tissue was collected from the left hemi-pheres of 3 wild control mice, 3 mice stimulated for 3 days, 3 micetimulated for 7 days, and from E15.5 mouse embryos. The collectedissue was lysed in SDS lysis buffer (4% SDS, 0.1 mM Tris–HCl [pH.8], 8.3% �-mercaptoethanol, 20% Glycerol, and 0.02% BPB) andoiled for 5 min. Each lysate was subjected to electrophoresis in 5%DS-polyacrylamide gel (PAGE). The proteins in the gel were trans-erred to a PVDF membrane (Immobilon, Millipore), which washen incubated with mouse anti-nestin (Rat401, 1:500; ab11306,bcam) or rabbit anti-GAPDH (1:2500, G9545, Sigma) as primaryntibodies. The GAPDH signal was used as an internal standard.he membrane was then reacted with the appropriate secondaryntibodies conjugated with horseradish peroxidase (goat anti-ouse IgG1 or goat anti-rabbit IgG, Jackson ImmunoResearch

aboratories), and developed using Immunostar (Wako). The bandsere quantified with a lumino-image analyzer (ChemiDoc XRS+;io Rad) and Image Lab (Bio Rad) software. The nestin band inten-

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

ities were normalized to GAPDH to account for loading variationsetween lanes. The mean relative band intensity (normalized toAPDH) of the control brain lysate was assigned a value of 1.0 foromparisons between kindling and control groups.

PRESSesearch xxx (2013) xxx–xxx

3. Results

To investigate the expression of nestin and other genes in themouse neocortex before and during kindling progression (Fig. 1A),we first determined the characteristics of kindling progressionin wild-type mice. Stimulation electrodes were implanted in theamygdala and neocortex of 3 wild-type mice (Fig. 1B). Kindlingprogression was scored according to the seizure class (Fig. 1C andSupplemental Movie) (Racine, 1972), duration of seizure behavior(Fig. 1D and Supplemental Movie), and duration of epileptic EEG(Fig. 1E), for 20 days. The seizure class reached a plateau by 6 days(Fig. 1C), and the duration of seizure reached a plateau by 10 days(Fig. 1D).

Next, we investigated the expression of nestin and other severalgenes in the mouse brain, using control mice (without electrodesFigs. 1A-[1] and 2B), or mice with electrodes that were allowed torecover for 1 week from implantation that received no stimulation(Figs. 1A-[2] and 2C), 3 days of stimulation (Figs. 1A-[3] and 2D), or7 days of stimulation (Figs. 1A-[4] and 2E). In addition, 22 C57/B6wild-type mice and 17 Nes-DTA mice were subjected to electricalstimulation as well as tamoxifen and BrdU administration (Fig. 1A-[1–4]). All the mice were perfused with fixative, and the fixed brainswere recovered for immunohistochemistry.

We first examined the suitability of using a mouse monoclonalantibody for immunohistochemistry on adult mouse brain tissue.Mouse normal serum contains immunoglobulin-G (IgG), whichcannot be completely expelled from mouse brain by perfusion withPBS. To estimate the background staining by mouse normal serum,we incubated 50-�m-thick mouse brain sections with or with-out the primary anti-nestin mouse monoclonal antibody (1/1000purified IgG) in incubation solution (1% donkey normal serum,0.25% �-Carrageenan, 0.3% Triton X-100, in PBS) overnight. Thedistribution of mouse IgG was then detected by biotinylated anti-mouse-IgG donkey IgG, ABC complex (Vector, USA), and the DABreaction. Fig. 2A and B shows the results of staining without andwith anti-nestin mouse IgG (Fig. 2A and B). The low backgroundin Fig. 2A indicated that nonspecific staining by the mouse-on-mouse effect was negligible. Thus, the staining in Fig. 2B specificallyrevealed the nestin-immunoreactive sites, which are describedbelow.

Adult mouse brain has a low level of nestin-IR as a whole (Fig. 2Aand B). However, nestin-IR most strongly labeled the floor of thethird ventricle, the hippocampal fimbria, and the choroid plexuses(Fig. 2B). The nestin-positive cells in the third ventricle are tany-cytes (Barrett et al., 2006); those in the fimbria and choroid plexusesare myeloid progenitors (Nataf et al., 2006) (Fig. 2F). The nextstrongest level of nestin-IR was seen in the subventricular zone(SVZ) around the lateral ventricles and the subgranular zone (SGZ)of the dentate gyrus (Fig. 2G and H). Proliferating endothelial pro-genitor cells and pericytes of blood vessels are often positive fornestin-IR (Dahlstrand et al., 1995; Kobayashi et al., 1998; Suzukiet al., 2010). We observed sparse labeling in capillaries of the neo-cortex (Fig. 2I) and in capillaries on the surface of the pia mater(Fig. 2J). The immunoreactivity in the parenchyma of the neocortexand that of the hippocampus was sparse (Fig. 2B) and did not revealthe shape of the nestin-positive cells. However, the immunoreac-tivity might have been due to NG2 cells in the parenchyma of theneocortex and hippocampus as reported previously (Kronenberget al., 2005; Tanaka et al., 2009).

The nestin-IR in the neocortex and hippocampus, especiallyaround the electrodes and the surface of the neocortex, was greatlyincreased after electrode insertion (Fig. 2C) and after 3 days of

induces nestin expression in the leptomeninges of the neocortex.

electrical stimulation (Fig. 2D), but it had been slightly reducedby 7 days of electrical stimulation (Fig. 2E). Since the immunore-activity in the parenchyma and the surface of the neocortex wasdiffuse, it was difficult to investigate the morphology of individual

ARTICLE IN PRESSG ModelNSR-3518; No. of Pages 9

S. Ninomiya et al. / Neuroscience Research xxx (2013) xxx–xxx 5

Fig. 4. Glial and neuronal marker distribution in the parenchyma of the neocortex. (A) Distribution of nestin-IR and NG2-IR in the in the parenchyma of the neocortex.N ressedm the ao (E) is

nwia(

n

estin-IR revealed a type of satellite glia. Some were NG2-positive while others exparker GAD67-IR in the neocortical parenchyma. (D) Nestin-IR did not overlap with

ften overlapped around the stimulation electrode (asterisk). The calibration bar in

estin-positive cells in thick sections. Therefore, we perfused miceith a fixative for EM observation, processed the brain for nestin

mmunohistochemistry, embedded the fixed tissue in epon resin,

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

nd cut it into semi-thin (0.5-�m thick) (Fig. 2K and L) or thin60-nm thick) sections (Fig. 2M and N).

The nestin-positive cells in the neocortical parenchyma hadumerous ramified thin processes radiating from a small cell body.

little or no NG2. (B–C) Neither nestin-IR nor NG2-IR co-localized with the neuronalstrocyte marker GFAP-IR in the neocortical parenchyma. (E) Nestin-IR and GFAP-IR

50 �m and applies to all panels.

Their processes were often in contact with the cell bodies of pyra-midal neurons, indicating they were a type of satellite glia (Fig. 2K)(Rakic, 1985; Kornack and Rakic, 2001; Koketsu et al., 2003). The

induces nestin expression in the leptomeninges of the neocortex.

nestin-IR on the surface of the pia mater was also enhanced afterelectrical stimulation (Fig. 2C–E). The nestin-IR was localized togranule cells (arrowheads in Fig. 2L and N) and their processes onthe surface of blood vessels. Arrowheads indicate the basal lamina,

ARTICLE IN PRESSG ModelNSR-3518; No. of Pages 9

6 S. Ninomiya et al. / Neuroscience Research xxx (2013) xxx–xxx

Fig. 5. (A) Nestin-IR in the leptomeninges after 7 days of kindling. Clusters of nestin-positive granule cells were found around the emissary vein (asterisk) from the neocorticalparenchyma. (B) During kindling progression, nestin-positive granule cells incorporated BrdU into their nuclei, indicating that they were proliferating. (C) Nestin-positivegranule cells in the leptomeninges were negative for NG2. (D–F) The leptomeninges of wild-type mice did not show any GAD67-IR. However, in parallel with kindlingprogression (3–7 days), the granule cells in the leptomeninges showed labeling with GAD67-IR. (G–H) The vast majority of the GAD67-positive cells also showed nestin-IRafter 3 days of kindling. (I) GFP-positive granule cells in the leptomeninges of Nestin-GFP mouse after 4 days kindling progression. (J) The continuous proliferation of cellsin the leptomeninges was revealed by BrdU incorporation. BrdU was administered to 3 wild-type mice for 28 days. (K) Kindling enhanced the proliferation of cells in theleptomeninges, as revealed by BrdU incorporation after 28 days of kindling. More BrdU-positive cells were observed in the leptomeninges of mice subjected to kindling thanin controls. (L) Most of the BrdU-positive cells in the neocortical parenchyma were positive for NG2. (M) Elimination of the BrdU-positive cells produced during kindling for7 days by tamoxifen administration. (N) The deletion of nestin-positive cells by DTA expression caused seizure and active-caspase-3-IR in many pyramidal neurons (arrows)and granule cells (double arrow) in the neocortex. (O) The number of BrdU-positive cells in the leptomeninges of a coronal section at 50 �m thick obtained from the wild-typemice and Nes-DTA mice (n = 3). That of wild-type mice was 190 ± 20.6, and that of Nes-DTA mice (n = 3) was 115 ± 6.6 (Welch’s t-test, *P < 0.01). The difference between thetwo numbers might be regarded as the number of nestin positive granule cells in one coronal section. Calibration bars are 100 �m in (A), (I), and (J), and 50 �m in (G) and(

wtr(c

L). The bar in (A), (G), and (J) applies to (B–F), (H), and (K), (M), respectively.

hich consists of extracellular matrix and is a boundary between

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

he brain and leptomeninges (Fig. 2M, which corresponds to theectangle in Fig. 2N). Therefore, the nestin-positive granule cellsarrowheads in Fig. 2N) were outside the brain, but were not bloodells or capillary cells.

The mouse brains shown in Fig. 2C–E were injured by the elec-

induces nestin expression in the leptomeninges of the neocortex.

trodes and damaged by the electrical stimulation. To exclude thepossibility of increased labeling caused by antibody mouse-on-mouse binding in brains with inflammation, we performed westernblotting for nestin with brain tissue taken from 3 control mice, 3

ARTICLE IN PRESSG ModelNSR-3518; No. of Pages 9

S. Ninomiya et al. / Neuroscience Research xxx (2013) xxx–xxx 7

Fig. 6. (A–C) Nestin immunohistochemistry on coronal sections of Nes-DTA mouse brains. (A) Nestin immunohistochemistry of an untreated Nes-DTA mouse brain, (B)N 7 daym Nes-Do

mtcmtdn

iacfpnmi(

lntcpsfNl(felHs(Gt

tpu

es-DTA mouse brain after 3 days of kindling, and (C) Nes-DTA mouse brain afterouse brain. (D) DTA immunohistochemistry of normal Nes-DTA mouse brain, (E)

f kindling.

ice stimulated for 3 days, and 3 mice stimulated for 7 days. Inhe immunohistochemistry results, the nestin-IR was lowest in theontrol mice, medium in mice stimulated for 7 days, and highest inice stimulated for 3 days (Fig. 2C–E). This pattern was repeated in

he western blotting results (Fig. 3). That is, brains stimulated for 3ays contained the highest amount of nestin and more unmodifiedestin (170 kDa) than those of the other mice (Fig. 3B).

We also investigated the co-localization of nestin-IR andmmunoreactivity for other proteins in the neocortical parenchymafter 7 days of electrical stimulation. A number of nestin-positiveells were positive for the glial marker NG2; others were negativeor NG2 or only moderately positive for it (Fig. 4A). Both the nestin-ositive and NG2-positive cells in the neocortical parenchyma wereegative for the neuronal marker GAD67 (Fig. 4B and C). Further-ore, Nestin-IR and GFAP-IR, an astrocyte marker, did not overlap

n the neocortex (Fig. 4D) except around the stimulation electrodeasterisk in Fig. 4E).

Under the condition of no electrode and no electrical stimu-ation, the granule cells in the leptomeninges were negative forestin (Fig. 2B and J). After three times of stimulation, however,he granule cells in the leptomeninges turned into nestin-positiveells. They incorporated BrdU into their nuclei, indicating that theyroliferated (Fig. 5A and B) during the 7 days of kindling progres-ion, especially around the emissary vein (asterisk in Fig. 5A). Weound that the nestin-positive cells in the leptomeninges wereG2-negative (Fig. 5C). We also examined the GAD67-IR in the

eptomeninges of control animals and did not find any labelingFig. 5D). However, in brains stimulated for 3 days, GAD67-IR wasound in the leptomeninges (Fig. 5E), and additional stimulationnhanced this labeling (Fig. 5F). The nestin-IR and GAD67-IR co-ocalized in the granule cells of the leptomeninges (Fig. 5G and). However, the level of nestin-IR peaked on the third day of

timulation (Figs. 2D and 3), and the GAD67-IR level peaked laterFig. 5D–F). In addition, granule cells in the leptomeninges showedFP fluorescence after 3 days of electrical stimulation in Nestin-GFP

ransgenic mice (Fig. 5I).

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

We also estimated the number of granule cells in the lep-omeninges over a wider area and a longer time. Although cellroliferation in the leptomeninges continued at low level evennder normal conditions as indicated by incorporation of BrdU

s of kindling. (D–F) DTA immunohistochemistry on a coronal section of Nes-DTATA mouse brain after 3 days of kindling, (F) and Nes-DTA mouse brain after 7 days

(Fig. 5J), kindling enhanced the proliferation of the cells in theleptomeninges and in the parenchyma of the neocortex (Fig. 5K).The cells newly produced in the leptomeninges seemed to benestin-positive granule cells (Fig. 5B), while those in theparenchyma of the neocortex were nestin-positive NG2 cells(Fig. 5L). To delete the BrdU- and nestin-double-positive cells, weadministered tamoxifen to Nes-DTA mice during the kindling pro-gression (Fig. 6). The Nes-DTA mice simultaneously expressed DTAand nestin in time and space, and nestin-IR in the Nes-DTA mousebrain was reduced earlier (Fig. 6C) than the case of wild-typemice (Fig. 2E). Therefore, we counted BrdU-positive cells buriedin laminin-IR (i.e., outside the basal lamina; see Fig. 2M and N) andcompared between wild-type mice and Nes-DTA mice. We stimu-lated brains of 3 wild-type mice and 3 Nes-DTA mice for kindlingprogression and administered tamoxifen and BrdU after stimula-tion for 10 days. Then a coronal section at 50 �m thick was obtainedfrom the 6 mice. The number of BrdU-positive cells in the lep-tomeninges of was 190 ± 20.6, and that of Nes-DTA mice (n = 3)was 115 ± 6.6 (Welch’s t-test, *P < 0.01). The difference between thetwo numbers might be regarded as the number of nestin positivegranule cells in one coronal section.

Twenty eight days after (28 applications of electrical stimula-tion), there were 188.1 ± 20.6 (mean ± s.e.m.) BrdU-positive cellsin the leptomeninges of one section obtained from wild-type mice(n = 6), whereas 115.8 ± 6.6 (mean ± s.e.m.) were found in the lep-tomeninges of Nes-DTA mice (n = 6). The difference was significant(P < 0.01) (Fig. 5O) and indicated that more than 38% (73 cells) ofthe BrdU-positive cells were positive for nestin.

We also tried to determine the function of the nestin-expressingcells which were induced by the kindling. The administration of thetamoxifen to the adult Nes-DTA mice deleted the nestin-positivecells in the SGZ of the dentate gyrus, those in the SVZ of the tele-ncephalon, and possibly also the endothelial progenitor cells. Theloss of these nestin-positive cells did not immediately induce asignificant change in the animals’ behavior or in the morpholog-ical features of the neocortex (Imayoshi et al., 2008). However,

induces nestin expression in the leptomeninges of the neocortex.

the administration of the tamoxifen to the Nes-DTA mice in kin-dling induced a high susceptibility to electrical stimulation andfaster kindling progression, while the administration of the tamox-ifen to the wild-type mice in kindling did not change kindling

ARTICLE ING ModelNSR-3518; No. of Pages 9

8 S. Ninomiya et al. / Neuroscience R

Fig. 7. (A) The seizure class was rated from class 0 to 5 during kindling progres-sion in 22 wild-type mice and 17 Nes-DTA mice. According to the experimentalparadigm shown in Fig. 1A, after each electrical stimulation, BrdU and tamoxifenwas administered to all the mice. The seizure reached class 5 in 3.6 ± 0.39 days inNes-DTA mice and 4.0 ± 0.33 days in wild-type mice (not statistically significant). (B)Duration of epileptic seizure during kindling progression. The duration of epilepticseizure reached a maximum by 10 days in wild-type mice. However, the maximumlevel in the seizure duration of Nes-DTA mice was not found. Moreover, after 13days, Nes-DTA mice were lost almost, possibly due to a naturally generated epilep-tsy

pafwoeso2(

4

intnedrd(Nt(ttv

ic seizure or due to exhaustion. The difference of duration of epileptic seizure wasignificant (*P < 0.05, **P < 0.01 t-test), but data from one Nes-DTA mouse could notield significant value after 13 days.

rogression (n = 22) (Fig. 7). Moreover, a significant number ofctive-caspase3-positive granule cells and pyramidal neurons wereound in the neocortex of the Nes-DTA mice in kindling pregressionith tamoxifen (Fig. 5N). In addition, as the result of the deletion

f nestin-positive cells, 7 of 17 Nes-DTA mice (41%) died before thend of experiment, possibly due to a naturally generated epilepticeizure and 9 Nes-DTA mice dropped out, due to exhaustion. Finallynly one Nes-DTA mouse remained in the Nes-DTA mice group for0 days, but could not yield significant data any more after 13 daysduration in Fig. 7).

. Discussion

This study showed that the artificial modification of brain activ-ty induces gene expression changes in the brain. In particular,estin expression was markedly induced by both electrode inser-ion and a few applications of electrical stimulation in kindling. Theestin expression in some cells may be related to CNS neurogen-sis, but in others it may be related to non-neural functions. Toetermine whether or not the nestin expression is related to neu-ogenesis, we used a transgenic mouse in-which GFP cDNA wasriven by the nestin promoter and the nestin second enhancerZimmerman et al., 1994; Yamaguchi et al., 2000). Here, using thisestin-GFP transgenic mouse, we visualized the granule cells in

he leptomeninges by their GFP expression. Although Tanaka et al.

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

2009) found that the NG2-positive satellite glia express GFP inhese mice, it is well known that the progenitors in the SVZ ofhe lateral ventricle and in the SGZ of the dentate gyrus are alsoisualized by their GFP expression.

PRESSesearch xxx (2013) xxx–xxx

There are several reports in which neural progenitors weredetected in the leptomeninges after a traumatic event, such asischemia, hemorrhage, or injury (Nakagomi et al., 2011; Sgubinet al., 2007; Decimo et al., 2011). However, the markers used tocharacterize the neural progenitors in these studies were ones thatappear only during cell proliferation or at a very immature stage.Nestin was one of them. Furthermore, all the neuronal markers inthese papers were seen only in culture, not in vivo. Since it is impos-sible to add the stress of ischemia, hemorrhage, or injury to a mousebrain repetitively, we chose amygdala kindling to apply repetitivestress to the brain in a well-defined manner. With this method, wewere able to detect leptomeningeal granule cells expressing bothnestin and GAD67 in vivo. GAD67 is a marker of inhibitory neuronsthat synthesize GABA as a transmitter, and it is also present in theproliferating intermediate progenitors of GABAergic neurons (Wuet al., 2011).

Our results using Nes-DTA mice suggested that the nestin-expressing cells are necessary for the brain to recover fromelectrical stimulation. When we performed kindling in Nes-DTAmice for longer than 13 days and delete nestin positive cells, mostof the mice died or dropped out before the end of the experi-ment, probably because of naturally generated repetitive epilepticseizures and exhaustion. Since nestin-positive NG2 satellite cellsin the parenchyma of the neocortex are reported to express GABAAreceptor and release BDNF (Tanaka et al., 2009), NG2-positive satel-lite glia may suppress epileptic seizure with BDNF.

Since the GAD67-positive granule cells in the leptomeningeslie outside the brain, they may originate from the neural crest(Nakagomi et al., 2011). Neural crest cells construct the entericneural plexus, which includes GAD67-positive GABA neurons(Williamson et al., 1995). Therefore, GAD67-positive granule cellsin the leptomeninges may also release GABA and regulate neuronalactivity in the neocortex. Moreover, as observed in the spinal cord(Decimo et al., 2011), nestin-positive leptomeningeal cells maymigrate into the parenchyma of the neocortex and acquire furtherimportant function in the neocortical circuits.

Here, we obtained novel information about changes in the lep-tomeninges occurring as a result of seizure. Further studies shouldaddress the function of the nestin- and GAD67-double-positivegranule cells in vivo.

Acknowledgements

This work was supported by a Grant in Aid from The JapanEpilepsy Research Foundation and a Grant in Aid from the JapanMinistry of Education, Culture, Sports, Science and Technology(2470032700).

References

Barrett, P., Ivanova, E., Graham, E.S., Ross, A.W., Wilson, D., Plé, H., Mercer, J.G., Ebling,F.J., Schuhler, S., Dupré, S.M., Loudon, A., Morgan, P.J., 2006. Photoperiodic regula-tion of cellular retinol binding protein, CRBP1 and nestin in tanycytes of the thirdventricle ependymal layer of the Siberian hamster. J. Endocrinol. 191, 687–698.

Crespel, A., Rigau, V., Coubes, P., Rousset, M.C., de Bock, F., Okano, H., Baldy-Moulinier,M., Bockaert, J., Lerner-Natoli., M., 2005. Increased number of neural progenitorsin human temporal lobe epilepsy. Neurobiol. Dis. 19, 436–450.

Dahlstrand, J., Lardelli, M., Lendahl, U., 1995. Nestin mRNA expression correlateswith the central nervous system progenitor cell state in many, but not all, regionsof developing central nervous system. Brain Res. Dev. Brain Res. 84, 109–129.

Decimo, I., Bifari, F., Rodriguez, F.J., Malpeli, G., Dolci, S., Lavarini, V., Pretto, S.,Vasquez, S., Sciancalepore, M., Montalbano, A., Berton, V., Krampera, M., Fuma-galli, G., 2011. Nestin- and doublecortin-positive cells reside in adult spinal cordmeninges and participate in injury-induced parenchymal reaction. Stem Cells29, 2062–2076.

Imayoshi, I., Ohtsuka, T., Metzger, D., Chambon, P., Kageyama, R., 2006. Temporal

induces nestin expression in the leptomeninges of the neocortex.

regulation of Cre recombinase activity in neural stem cells. Genesis 44, 233–238.Imayoshi, I., Sakamoto, M., Ohtsuka, T., Takao, K., Miyakawa, T., Yamaguchi, M., Mori,

K., Ikeda, T., Itohara, S., Kageyama, R., 2008. Roles of continuous neurogenesis inthe structural and functional integrity of the adult forebrain. Nat. Neurosci. 11,1153–1161.

ING ModelN

ence R

K

K

K

K

L

N

N

O

R

RS

ARTICLESR-3518; No. of Pages 9

S. Ninomiya et al. / Neurosci

obayashi, M., Sjöberg, G., Söderhäll, S., Lendahl, U., Sandstedt, B., Sejersen, T., 1998.Pediatric rhabdomyosarcomas express the intermediate filament nestin. Pedi-atr. Res. 43, 386–392.

oketsu, D., Mikami, A., Miyamoto, Y., Hisatsune, T., 2003. Nonrenewal of neuronsin the cerebral neocortex of adult macaque monkeys. J. Neurosci. 23, 937–942.

ornack, D.R., Rakic, P., 2001. Cell proliferation without neurogenesis in adult pri-mate neocortex. Science 294, 2127–2130.

ronenberg, G., Wang, L.P., Synowitz, M., Gertz, K., Katchanov, J., Glass, R., Harms,C., Kempermann, G., Kettenmann, H., Endres, M., 2005. Nestin-expressing cellsdivide and adopt a complex electrophysiologic phenotype after transient brainischemia. J. Cereb. Blood Flow Metab. 25, 1613–1624.

endahl, U., Zimmerman, L.B., McKay, R.D., 1990. CNS stem cells express a new classof intermediate filament protein. Cell 60, 585–595.

akagomi, T., Molnár, Z., Nakano-Doi, A., Taguchi, A., Saino, O., Kubo, S., Clausen,M., Yoshikawa, H., Nakagomi, N., Matsuyama, T., 2011. Ischemia-induced neuralstem/progenitor cells in the pia mater following cortical infarction. Stem CellsDev. 20, 2037–2051.

ataf, S., Strazielle, N., Hatterer, E., Mouchiroud, G., Belin, M.F., Ghersi-Egea, J.F.,2006. Rat choroid plexuses contain myeloid progenitors capable of differentia-tion toward macrophage or dendritic cell phenotypes. Glia 54, 160–171.

hta, K., Lupo, G., Kuriyama, S., Keynes, R., Holt, C.E., Harris, W.A., Tanaka, H.,Ohnuma, S., 2004. Tsukushi functions as an organizer inducer by inhibition ofBMP activity in cooperation with chordin. Dev. Cell 7, 347–358.

acine, R.J., 1972. Modification of seizure activity by electrical stimulation: II. Motor

Please cite this article in press as: Ninomiya, S., et al., Amygdala kindlingNeurosci. Res. (2013), http://dx.doi.org/10.1016/j.neures.2012.12.006

seizure. Electroencephalogr. Clin. Neurophysiol. 32, 281–294.akic, P., 1985. Limits of neurogenesis in primates. Science 227, 1054–1056.gubin, D., Aztiria, E., Perin, A., Longatti, P., Leanza, G., 2007. Activation of endogenous

neural stem cells in the adult human brain following subarachnoid hemorrhage.J. Neurosci. Res. 85, 1647–1655.

PRESSesearch xxx (2013) xxx–xxx 9

Suzuki, S., Namiki, J., Shibata, S., Mastuzaki, Y., Okano, H., 2010. The neuralstem/progenitor cell marker nestin is expressed in proliferative endothe-lial cells, but not in mature vasculature. J. Histochem. Cytochem. 58,721–730.

Tamamaki, N., Nakamura, K., Furuta, T., Asamoto, K., Kaneko, T., 2000. Neurons inGolgi-stain-like images revealed by GFP-adenovirus infection in vivo. Neurosci.Res. 38, 231–236.

Tamamaki, N., Yanagawa, Y., Tomioka, R., Miyazaki, J., Obata, K., Kaneko, T., 2003.Green fluorescent protein expression and colocalization with calretinin, parval-bumin, and somatostatin in the GAD67-GFP knock-in mouse. J. Comp. Neurol.467, 60–79.

Tanaka, Y., Tozuka, Y., Takata, T., Shimazu, N., Matsumura, N., Ohta, A., Hisatsune,T., 2009. Excitatory GABAergic activation of cortical dividing glial cells. Cereb.Cortex 19, 2181–2195.

Williamson, S., Faulkner-Jones, B.E., Cram, D.S., Furness, J.B., Harrison, L.C., 1995.Transcription and translation of two glutamate decarboxylase genes in the ileumof rat, mouse and guinea pig. J. Auton. Nerv. Syst. 55, 18–28.

Wu, S., Esumi, S., Watanabe, K., Chen, J., Nakamura, K.C., Nakamura, K., Kometani,K., Minato, N., Yanagawa, Y., Akashi, K., Sakimura, K., Kaneko, T., Tamamaki,N., 2011. Tangential migration and proliferation of intermediate progeni-tors of GABAergic neurons in the mouse telencephalon. Development 138,2499–2509.

Yamaguchi, M., Saito, H., Suzuki, M., Mori, K., 2000. Visualization of neurogene-sis in the central nervous system using nestin promoter-GFP transgenic mice.

induces nestin expression in the leptomeninges of the neocortex.

Neuroreport 11, 1991–1996.Zimmerman, L., Parr, B., Lendahl, U., Cunningham, M., McKay, R., Gavin, B., Mann, J.,

Vassileva, G., McMahon, A., 1994. Independent regulatory elements in the nestingene direct transgene expression to neural stem cells or muscle precursors.Neuron 12, 11–24.