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

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ l oca te /b ra in res

Research Report

Degenerative alterations in the visual pathway afterNMDA-induced retinal damage in mice

Yasushi Itoa, Masamitsu Shimazawaa, Yuta Inokuchia, Hidefumi Fukumitsub,Syouei Furukawab, Makoto Araiec, Hideaki Haraa,⁎aDepartment of Biofunctional Evaluation,Molecular Pharmacology, Gifu Pharmaceutical University, 5-6-1Mitahora-higashi, Gifu 502-8585, JapanbDepartment of Biofunctional Evaluation, Molecular Biology, Gifu Pharmaceutical University, 5-6-1 Mitahora-higashi, Gifu 502-8585, JapancDepartment of Ophthalmology, University of Tokyo School of Medicine, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan

A R T I C L E I N F O

⁎ Corresponding author. Fax: +81 58 237 8596.E-mail address: hidehara@gifu-pu.ac.jp (HAbbreviations: BDNF, brain-derived neurot

geniculate nucleus; NeuN, neuronal nuclear

0006-8993/$ – see front matter © 2008 Elsevidoi:10.1016/j.brainres.2008.03.021

A B S T R A C T

Article history:Accepted 5 March 2008Available online 21 March 2008

In the present study, intravitreal injection of N-methyl-D-aspartate (NMDA) into the left eyeinduced retinal damage (decreases in the number of retinal ganglion cells) at 1 day after theinjection. At 7 days after the injection, atrophy of the optic tract was observed on thecontralateral side, but not on the ipsilateral side. Number of neuronal nuclear specificprotein (NeuN)-immunostained neurons were decreased in the contralateral dorsal LGN(dLGN) and contralateral ventral LGN-lateral (vLGN-l) at 90 and 180 days, respectively, afterthe injection. Furthermore, expressions of glial fibrillary acid protein (GFAP) were increasedin the contralateral dLGN and contralateral vLGN-l at 7 and 30 days, respectively, and thoseof brain-derived neurotrophic factor (BDNF) were increased in the contralateral dLGN at 30and 90 days and in the contralateral vLGN-l at 7 and 30 days. All NeuN-positive neuronalcells exhibited BDNF, whereas only some GFAP-positive astroglial cells exhibited BDNF.However, the contralateral ventral LGN-medial (vLGN-m) and ipsilateral LGN displayed nosignificant differences related to NeuN, GFAP, or BDNF immunohistochemistry. Takentogether, these results indicate that time-dependent alterations occurred after the NMDAinjection along the retinogeniculate pathway (from retina to LGN), and that the degree ofdamage in the LGN was region-dependent. In addition, the increased activated astroglialcells and expressions of BDNF in the damaged regions may play some roles in the cell-survival process of the LGN.

© 2008 Elsevier B.V. All rights reserved.

Keywords:Brain-derived neurotrophic factor(BDNF)Glial fibrillary acid protein (GFAP)Neuronal nuclear specific protein(NeuN)Optic tractRetinal ganglion cells (RGC)

1. Introduction

Glaucoma is an optic neuropathy resulting from the death ofretinal ganglion cells (RGC). In clinical studies, a loss of morethan 50% of RGC has been reported to induce visual field loss.However, the initial loss of RGC does not lead to visual field

. Hara).rophic factor; CNS, centrspecific protein; NMDA, N

er B.V. All rights reserved

loss in humans (Quigley et al., 1989). Possibly, a compensatoryaction of the visual cortex may protect the visual field againstsuch a decrease. However, there is recent evidence that theRGC death that occurs in glaucoma leads to neuronaldegeneration within the lateral geniculate nucleus (LGN), themajor relay center between eye and visual cortex (Yücel et al.,

al nervous system; GFAP, glial fibrillary acid protein; LGN, lateral-methyl-D-aspartate; RGC, retinal ganglion cells

.

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

2000). Further, it was reported that time- and region-depen-dent morphological changes had occurred in LGN at 120 daysafter intraocular pressure (IOP) was elevated in rats (Wanget al., 2000). These reports suggest that the visual field lossinduced by glaucoma may not result only from RGC loss, butalso from neuronal degeneration in LGN. However, noprevious investigation has been made of time-dependentalterations along the retinogeniculate pathway (i.e., retina,optic tract, and LGN) after retinal injury in mice. In addition,the possible pathophysiological mechanisms underlying neu-ronal cell death in LGN following RGC loss remain uncertain.

Excessive activation of glutamate receptors by glutamatereleased from injured RGC is implicated in the glaucomatousRGC death process (Osborne et al., 1999). Glutamate is theprincipal excitatory neurotransmitter within the central ner-vous system (CNS), and it has been found to be increased in thevitreous body in glaucoma (Dreyer et al., 1996). In contrast, thisfact was not confirmed by Honkanen et al. (2003). However, infact the toxic effects of elevated levels of glutamate arepredominantly mediated by the overstimulation of ionotropicreceptors. Overstimulation of the class of these receptors thatrespond specifically to the glutamate analog N-methyl-D-aspartate (NMDA) leads to an overload of intracellular Ca2+.Such elevations in Ca2+ elicit various cytotoxic biochemicalreactions including the activation of nitric oxide (NO) synthaseand the generation of reactive NO free radicals. Two otherclasses of ionotropic receptors, which respond to the agonists

Fig. 1 – Intravitreal injection of N-methyl-D-aspartate (NMDA) inHematoxylin and eosin staining of sections obtained from miceRepresentativemicrophotographs showing non-treated control reinjection (B–G, respectively). Vertical bars show thickness of inneArrows (in D–G) indicate retinal ganglion cells (RGC). Each animal's30, 90, or 180 days after the NMDA injection. Number of neuronsmorphological differences in retina were observed among controand among the control and sham-treated mice. Each value repres(Dunnett's test).

kainate and α-amino-3-hydroxy-5-methyl-4-isoxazolepropia-nate (AMPA), respectively, can also mediate Ca2+ overloadwhen overstimulated, but they are somewhat less permeableto this ion than the NMDA receptor (NMDAR). In fact, a singleintravitreal injection of NMDA has been reported to damagethe cells in GCL and the IPL without affecting the other retinallayers in rats 7 days after the injection (Akaike et al., 1998).Among the glutamate receptors, the NMDA receptor's role incell death has been extensively studied: excessive doses ofNMDA induce apoptotic cell deathof RGCandamacrine cells inrat retina (Lamet al., 1999; Inomata et al., 2003). Furthermore, ithas been reported that NMDAR positive cells are either RGC oramacrine cells (Jakobs et al., 2007). In addition, a previous studyhas shown that NMDAR-mediated neurotoxicity in the RGC isdependent on the influx of extracellular Ca2+ (Sucher et al.,1997). In fact, blockade of glutamate activity by modulation ofits receptors has been advocated as an important strategy forneuroprotection in glaucoma, and memantine, an NMDARantagonist, displays a neuroprotective effect in experimentalglaucoma (Li et al., 2002; Yücel et al., 2006). In either case, theanimal model employed in the present study (involvingintravitreal injection of NMDA) exhibits high sensitivity andstability, as well as good reproducibility, and is widely used forinvestigating the mechanisms underlying neuronal cell deathin the retina (Yoneda et al., 2001). We therefore used it toinvestigate time-dependent alterations in the murine LGNfollowing NMDA-induced retinal damage.

to left eye induced time-dependent retinal damage in mice.at 1, 3, 7, 30, 90, and 180 days after NMDA injection.tina (A), and retinas at 1, 3, 7, 30, 90, and 180 days after NMDAr plexiform layer (IPL). Horizontal bar represents 20 µm.left eyewas enucleated togetherwith its optic nerve at 1, 3, 7,in the GCL (H) and thickness of IPL (I) were measured. No

l mice at 1, 30, 90, and 180 days (n=4, 3, 3, and 3, respectively)ents the mean±S.E.M. for 6 to 8 eyes. **P<0.01 versus Control

91B R A I N R E S E A R C H 1 2 1 2 ( 2 0 0 8 ) 8 9 – 1 0 1

2. Results

2.1. NMDA-induced retinal damage

Intravitreal injection of NMDA at 40 nmol/eye decreased boththe cell-count in the GCL and the thickness of the IPL in theretina, as compared with those in the non-treated controlretina (Fig. 1). The cell-count in GCL was decreased to 85.3,47.6, 38.5, 32.5, 29.8, and 29.6% of control at 1, 3, 7, 30, 90, and180 days, respectively, after NMDA injection. The thickness ofIPL was increased to 138.8% of control at 1 day after the NMDAinjection (Fig. 1B and I), then time-dependently decreased to75.9, 53.7, 45.8, 43.1, and 34.0% of control at 3, 7, 30, 90, and180 days, respectively. Retinal damage occurred dramaticallyuntil 7 days after NMDA injection, and then the degree ofdamage was gradually increased by day 180. In fact, NMDAinjection significantly decreased cell number of GCL and IPLthickness between day 7 and day 180. No morphologicaldifferences in retina were observed among control mice at 1,30, 90, and 180 days (n=4, 3, 3, and 3, respectively) and amongthe control and sham-treated mice (data not shown).

2.2. Projection from the retina to the LGN

The distribution of labeling in the contralateral and ipsilateralretinogeniculate pathways is shown for normal mice in Fig. 2.Contralateral retinogeniculate labeling was almost all distrib-uted in dLGN and vLGN-l. Ipsilateral retinogeniculate labelingwas much more restricted within dLGN and vLGN-l. In dLGNand vLGN-l, the regions of weak labeling on the contralateralside overlapped with the labeled regions on the ipsilateralside. No contralateral and ipsilateral retinogeniculate labelingwas observed at all in vLGN-m.

2.3. Atrophy of optic tract

Both the thickness and the area of the optic tract were time-dependently decreased after intravitreal injection of NMDA

Fig. 2 – Representative photographs from control animals (non-tr(A) and ipsilateral (B) sides following intravitreal injection of whethe left eye. Scale bar=200 µm.

(Fig. 3). The thickness was decreased to 89.2, 59.7, 56.5, and41.5% of control at 7, 30, 90, and 180 days, respectively, afterNMDA injection, and the area was decreased to 84.2, 46.1, 44.7,and 43,7% of control at 7, 30, 90, and 180 days, respectively,after NMDA injection.

2.4. Neuronal cells immunostained for NeuN

NeuN is one of the best markers when it is important todistinguish neurons from glial cells (Gittins andHarrison, 2004).In control mice, NeuN labels were seen in almost all neurons indLGN (Fig. 4A), vLGN-l (Fig. 4D), and vLGN-m (Fig. 4G). NeuN-labeledneuronswere first decreased innumber in thedLGNandvLGN-l of the contralateral side at 180 days (Fig. 4B and E) afterNMDA injection. The number of NeuN-labeled neurons in thecontralateral dLGNwasdecreased to 82.1 and 81.5%of control at90 and 180 days, respectively, after NMDA injection (Fig. 4C). Inthe contralateral vLGN-l, it was decreased to 89.9% of control at180 days (Fig. 4F). However, no significant decrease in thenumber of NeuN-labeled neurons (versus control) was observedin the contralateral vLGN-m (Fig. 4G, H, and I) or in any region ofthe ipsilateral LGN (Fig. 4C, F, and I). No morphologicaldifferences in the LGN were observed among the control andsham-treated mice (data not shown).

Furthermore, the cell-size of NeuN-labeled neuronswas firstdecreased in vLGN-l and dLGN at 7 and 30 days, respectively(Fig. 5, Table 1). However, no significant decrease (versuscontrol) was observed in the contralateral vLGN-m (Table 1)and in any region of the ipsilateral LGN (data not shown).

2.5. Astrocytes immunostained for GFAP

GFAP immunoreactivity was elevated in the contralateraldLGN (Fig. 6A, B, and C) and vLGN-l (Fig. 6D, E, and F) at both 7and 30 days after NMDA injection. The density of GFAPexpression in the contralateral dLGN was increased to 706%and 677% of control at 7 and 30 days, respectively, after NMDAinjection (Fig. 6C), while in the contralateral vLGN-l it wasincreased to 438%and 450%of control at the same time-points(Fig. 6F). However, no significant increase (versus control) was

eated) showing sections through the LGN on the contralateralat germ agglutinin–horseradish peroxidase (WGA–HRP) into

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

observed in the contralateral vLGN-m (Fig. 6G, H, and I) and or inany region of the ipsilateral LGN (Fig. 6C, F, and I).

2.6. BDNF expression in LGN cells

In our previous study, the specificity of the antibody againstBDNF used in this experiment was characterized using immu-nohistochemistry and Western blotting (Furukawa et al., 1998).The antibodywas unique to BDNF, but not other neurotrophins.In control mice, BDNF-positive cells were observed in dLGN(Fig. 7A), vLGN-l (Fig. 7D), and vLGN-m (Fig. 7G). BDNF-positivecellswere increasedat 30and 90daysandat 7 and30days in thecontralateral dLGN and vLGN-l, respectively, after NMDAinjection (Fig. 7). The number of BDNF-positive cells in thecontralateral dLGN was increased to 121 and 124% of control at30 and 90 days, respectively, after the NMDA injection (Fig. 7C).In the contralateral vLGN-l, the number was increased to 121and 122% at 7 and 30 days, respectively, after the NMDAinjection (Fig. 7F), but it had decreased to 89% of control at180 days. In contrast, no significant decrease (versus control) inthe number of BDNF-positive cells was observed in thecontralateral vLGN-m (Fig. 7G, H, and I) or in any region of theipsilateral LGN (Fig. 7C, F, and I).

Fig. 3 – Time-dependent atrophy of optic tract induced by NMDfrom control (non-treated)mice (A) andmice at 180 days after NMDmagnification. Horizontal scale bar=50 µm. Vertical bars indicateshow area of optic tract. Thickness (C) and area (D) of optic tract wamong control mice at 1, 30, 90, and 180 days. Sham group (miceuthanized, and nomorphological differences were observed amthe mean±S.E.M. for 6 brains. *P<0.05, **P<0.01 versus Control (

2.7. Double immunofluorescence

To identify BDNF-positive cells, double immunofluorescencewas performed for BDNF and NeuN or for BDNF and GFAPusing LGN sections. NeuN-positive neuronal cells expressedBDNF (Fig. 8A, B, and C) during the study period after NMDAinjection, while GFAP-positive astroglial cells partly exhibitedBDNF in the contralateral dLGN and vLGN specifically at both 7and 30 days (Fig. 8D, E, and F) after NMDA injection.

3. Discussion

In the present study, unilateral intravitreal injection of NMDAinduced neuronal damage that was detected, in turn, in RGC,the optic tract, and in neurons contralateral dLGN, andcontralateral vLGN-l, the maximal extent of the neuronaldamage in these tissues (versus control, non-treated mice)being about 70, 60, 18, and 10%, respectively. This suggeststhat the retinal damage induced by intravitreal injection ofNMDA in mice lead to neuronal degeneration in the LGNconnected to that particular retina. Although a small numberof optic fibers (about 10% in the mouse) are uncrossed and

A injection. Representative microphotographs of optic tractA injection (B). Boxed areas in (A) and (B) are shown at higherthickness of optic tract. Regions surrounded by broken linesere measured. No morphological differences were observed

e that received an intravitreal injection of vehicle) wasong the control and sham-treatedmice. Each value representsDunnett's test).

93B R A I N R E S E A R C H 1 2 1 2 ( 2 0 0 8 ) 8 9 – 1 0 1

project to the ipsilateral LGN (Dräger and Olsen, 1980), therewas no significant change of neuronal loss in the ipsilateralLGN. Presumably since the damage was confined to smallregions, we could not detect neuronal loss in the ipsilateralLGN. During the period of neuronal degeneration, increasedactivated astrocytes and expressions of BDNF were observedin the contralateral dLGN and vLGN-l. Our results areconsistent with some previous reports on human glaucomaor on experimental primate and rat glaucoma models (Wanget al., 2000; Yücel et al., 2001; Gupta et al., 2006). However, thisis the first report that, by using mice, has shown time-dependent alterations in LGN as well as in the retina and optictract after NMDA injection. In addition, we demonstrated thepresence of GFAP-positive astroglial cells exhibiting BDNFimmunoreactivity in LGN following NMDA injection at anearly stage (at 7 days) of neuronal degeneration.

Recently, mice deficient in the glutamate transportersglutamate/aspartate transporter (GLAST) or excitatory aminoacid carrier 1 (EAAC1) have demonstrated spontaneous RGCand optic nerve degeneration without elevated (intraocularpressure) IOP (Harada et al., 2007). In addition, it has been

Fig. 4 – NeuN immunostaining of sections obtained frommousesham-treated mice. Representative microphotographs are shownvLGN-m (G)] and for mice at 180 days after NMDA injection [dLGNnumber of NeuN-labeled neurons was counted in dLGN (C; per 0.1and vLGN-m (I; per 0.0292 mm2 tissue area). No morphological dand180 days. Shamgroupwas euthanized, andnomorphological dmice. Each value represents the mean±S.E.M. for 6 to 8 brains. *P<

reported that GLAST is significantly reduced in an experi-mental rat glaucoma model. Reductions of GLAST mayincrease the potential for glutamate-induced injury to RGC inglaucoma (Martin et al., 2002). These permit us to speculatethat these glutamate transporters are involved in retinaldamage induced by NMDA injection. In addition, NMDARsubunits (NR1, NR2A-D) composition and the splicing of theNR1 are involved in neurotoxicity. When coexpressed withNR1, each of the NR2 subunits can form an ion channel(Meguro et al., 1992; Ishii et al., 1993) characterized by high Ca2+

conductance (MacDermott et al., 1986). Changing the splicevariants of the carboxyl-terminal domain of the NR1 subunitmodulates cell death, most likely by altering the calcium/calmodulin-dependent inactivation of the receptor (Lu et al.,2000; Rameau et al., 2000). Furthermore Ca2+ entry throughsynaptic NMDAR stimulates the transcription-regulating com-plex cAMP response element-binding (CREB)/CREB-bindingprotein and enhances neuronal survival (Papadia et al., 2005).In contrast, extrasynaptic NMDAR couples to CREB shut-offand cell death pathways (Hardingham et al., 2002). Thus, thedecision of whether a neuron survives or dies after glutamate

brains at 1, 3, 7, 30, 90, and 180 days after NMDA injection and: for control (non-treated) mice [dLGN (A), vLGN-l (D), and(B), vLGN-l (E), and vLGN-m (H)]. Scale bar=20 µm. Average551mm2 tissue area), vLGN-l (F; per 0.0972 mm2 tissue area),ifferences were observed among control mice at 1, 30, 90,ifferenceswere observed among the control and sham-treated0.05, **P<0.01 versus Control (Dunnett's test).

Fig. 5 –The cell size of NeuN-labeled neuronal cells was reduced by NMDA treatment in dLGN and vLGN-l (on the sidecontralateral to the injection). Percentage frequency distribution histograms comparing cross-sectional measurements (radius,µm) between each NMDA-treated group and the control (non-treated group). Data are shown for contralateral dLGN (A–D),vLGN-l (E–H), and vLGN-m (I–L). Each value represents the mean±S.E.M. for 6 brains.

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

exposure is dependent on the localization of the NMDAR(Soriano and Hardingham, 2007). Therefore focusing onglutamate is effective in elucidating the mechanisms of RGCloss in glaucoma.

In the present study, time-dependent decreases in RGC andin IPL thickness were observed (although IPL thickness wasactually increased at 1 day after NMDA injection, an affectpossibly due to retinal edema following the injection sinceedema has been reported at 1 day after NMDA exposure in thechick retina) (Pisani et al., 2006). As the result, retinal damageoccurred dramatically until 7 days after NMDA injection, andthen it was gradually increased by day 180. In fact, NMDAinjection showed marginally significant decreases of cellnumber of GCL and IPL thickness between day 7 and day

Table 1 – Radius of LGN cells in cross-sectional views in vLGN-

Time dLGN

Control 3.65±0.18 (100)7 days 3.27±0.06 (89.6)30 days 3.15±0.05 (86.3) ⁎90 days 2.71±0.05 (74.2) ⁎⁎180 days 3.17±0.15 (86.8) ⁎

Range of cell-counts total 1381–1682

Data were obtained from percentage frequency distribution histogramsNMDA-treated group and the control group for right vLGN-l, vLGN-m, anvalues±S.E.M. (µm) for 6 animals, with the value within parentheses show(Dunnett's test).

180. On the other hand, there were no significant differencesin cell number of GCL and IPL thickness of sham or controlmice during day 180 in this experiment. In addition, in aprevious report cells number of GCL was significantlydecreased 5 days after NMDA (200 nmol/eye) injection andthere was no further RGC loss from 7 to 14 days (Nakazawaet al., 2005). Therefore, we think that retinal damagemay havemainly occurred until 7 days after NMDA injection.

Following the RGC loss, a significant atrophic change in theoptic tract (optic fibers) was observed from 7 days after theNMDA injection, indicating that atrophy of the optic tract mayinfluence the reception by LGN of signals from the retina. Infact, neuronal loss in LGN has been reported to be present at14 months after optic fiber loss in the cynomolgus monkey

l, vLGN-m, and dLGN at various times after NMDA injection

vLGN-l vLGN-m

3.87±0.12 (100) 3.16±0.28 (100)3.41±0.10 (88.1) ⁎ 3.33±0.15 (105.4)3.35±0.09 (86.6) ⁎⁎ 3.01±0.10 (95.3)3.30±0.06 (85.3) ⁎⁎ 3.04±0.11 (96.2)2.86±0.08 (73.9) ⁎⁎ 3.25±0.04 (102.3)

839–983 374–505

comparing cross-sectional measurements (radius, µm) between thed dLGN (connected to the left damaged eye). Data shown are meaning % of control. ⁎P<0.05, ⁎⁎P<0.01 versus Control (non-treated mice)

Fig. 6 – GFAP immunostaining of sections obtained from mice at 1, 3, 7, 30, 90, and 180 days after NMDA injection.Representativemicrophotographs are shown: for control (non-treated) mice [dLGN (A), vLGN-l (D), and vLGN-m (G)] and formiceat 30 days after NMDA injection [dLGN (B), vLGN-l (E), and vLGN-m (H)]. Scale bar=20 µm. Average area of GFAP-IR astroglial cellsis shown for dLGN (C; per 0.1551 mm2 tissue area), vLGN-l (F; per 0.0972 mm2 tissue area), and vLGN-m (I; per 0.0292 mm2

tissue area). No morphological differences were observed among control mice at 1, 30, 90, and 180 days. Sham group waseuthanized, and nomorphological differences were observed among the control and sham-treatedmice. Each value representsthe mean±S.E.M. for 6 to 8 brains. **P<0.01 versus Control (Dunnett's test).

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

(Yücel et al., 2000). Therefore the optic nerve loss induced byNMDA injection may lead to a decrease in soma-size inneurons connected to the optic nerve. We also showedneuronal loss in LGN following atrophy of the optic tract, theloss occurring selectively in the dLGN and vLGN-l, but not inthe vLGN-m, on the side contralateral to the NMDA injection.These damaged regions correspond to the regions receivingprojection from RGC (see Fig. 2). Such neuronal lossmay resultfrom a reduction in the visual stimuli transmitted from RGCand from a depletion of neurotrophins accompanying by adysfunction of anterograde transport from RGC (Vrabec andLevin, 2007). These may lead to “Wallerian degeneration”, in aprocess where the part of the axon separated from theneuron's cell nucleus will degenerate (Finn et al., 2000). Inrats, neurons in the vLGN receive inputs from RGC as well asfrom such structures as the visual cortex, superior colliculus,optic tectum, cerebellum, and vestibular nucleus. Our resultsin mice suggest that even if neurons in the vLGN-m receive

inputs from the above-mentioned structures, they do notreceive from RGC (Fig. 3).

NeuN is an antibody which recognizes a neuron-specificantigen and which selectively and clearly stains neuronalperikarya and nuclei (Mullen et al., 1992). NeuN is also beingused in studies of neurodegenerative disorders and experi-mental lesions (Falke et al., 2003; Jongen-Rêlo and Feldon 2002;Kobayashi et al., 2003; Kordower et al., 2001). Recently, Gittinsand Harrison have reported that NeuN and Nissl stainingproduce highly correlated estimates of neuronal density, size,and shape and that values for neuronal density, size, andshape are consistently higher with NeuN than with Nissl(Gittins and Harrison, 2004). Therefore, we used NeuN stainingto evaluate the changes in neuronal density and size in theLGN after retinal damage. In the present study, neuronal cell-size decreased in the contralateral dLGN and contralateralvLGN-l (Fig. 5, Table 1) before any significant neuronal losswasdetected in LGN (Fig. 4).

Fig. 7 – BDNF immunostaining of sections obtained from mice at 1, 3, 7, 30, 90, and 180 days after NMDA injection.Representativemicrophotographs are shown: for control (non-treated) mice [dLGN (A), vLGN-l (D), and vLGN-m (G)] and formiceat 30 days after NMDA injection [dLGN (B), vLGN-l (E), and vLGN-m (H)]. Scale bar=20 µm. Average number of BDNFimmunopositive puncta counts is shown for dLGN (C; per 0.1551 mm2 tissue area), vLGN-l (F; per 0.0972 mm2 tissue area), andvLGN-m (I; per 0.0292 mm2 tissue area). No morphological differences were observed among control mice at 1, 30, 90, and180 days. Sham group was euthanized, and no morphological differences were observed among the control and sham-treatedmice. Each value represents the mean±S.E.M. for 4 brains. *P<0.05, **P<0.01 versus Control (Dunnett's test).

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

It is known that glial cells are activated at an early stage ofneuronal damage (during ischemia etc.) (Zoli et al., 1997). Thecytostructural protein of astrocytes in the brain, glial fibrillaryacidic protein (GFAP), is known to be a marker of astrocytesreacting to neuronal injury, and GFAP has proved to be a goodindex of the early changes induced by neuronal damage inbrain. Increased expressions of GFAP have been observed (a) inastroglial cells in the LGN following a lesion of the visualcortex in the rat (Agarwala and Kalil, 1998), and (b) in theprimary visual cortex after unilateral destruction of dLGN inadult albino rats (Hajos et al., 1990). Moreover, it has beenreported that number of GFAP-positive astroglial cellsincreased in the contralateral LGN from 7 days until at least168 days after monocular enucleation in rats (Gonzalez et al.,2006). In the present study, GFAP expression was alsoincreased from 7 days (until at least 30 days) in the

contralateral LGN after NMDA injection. Anti-GFAP antibodyused in this study was little response to resting astroglial cellsin the LGN of control mice (Fig. 7). Since the original number ofastroglial cells could not be measured in control mice, wecould not confirm whether the astroglial cells proliferated orbecame hypertrophied. However it has been reported toincrease number of astroglial cells in the contralateral LGNof monocular enucleated rat (Gonzalez et al., 2006). Therefore,these findings indicate that astroglial cells may be proliferatedin this study. Increased GFAP expression in astroglial cells isknown to be associatedwith synaptic regulation (Meshul et al.,1987; Canady et al., 1994), and it has been suggested thatastrocytic responses to afferent transmitter release are essen-tial for the survival of neurons (Brennemanet al., 1987). Indeed,in the present study neuronal death was not observed in LGNduring the period in which GFAP expression was increased.

Fig. 8 – Representative photographs showing BDNF (A), NeuN (B), and BDNF/NeuN (C) double-immunostaining and BDNF (D),GFAP (E), and BDNF/GFAP (F) double-immunostaining of vLGN at 30 days after NMDA treatment. Scale bar=20 µm. SomeBDNF-expressing cells (A, green) were co-localized with NeuN-labeled neurons (B, red), as indicated by yellow color in C (mergeof A and B). BDNF-expressing cells (D, green) were partly co-localizedwith GFAP-positive astroglial cells (E, red), as indicated byyellow color in F (merge of D and E).

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

Neurotrophins are growth factors that play an essentialrole in neuronal development and maintenance, includingmorphological differentiation. Brain-derived neurotrophicfactor (BDNF), a member of the family of neurotrophic factors,is selectively, but very widely, distributed within the develop-ing and adult nervous systems (Ernfors et al., 1990; Merlioet al., 1992; Ringstedt et al., 1993; Yan et al., 1997), and appearsto play a variety of previously unsuspected roles in synapticfunction and plasticity in the adult CNS. Many neurotrophicfactors (BDNF, GDNF, etc.) are produced by activated astroglialcells. Thus, functional changes in astroglial cells are importantfor the neuronal repair process in the damaged CNS (Koyama,2002). Reportedly, BDNFhas a neuroprotective effect in variousneuronal injury models (Weibel et al., 1995; Ikeda et al., 1999;Ota et al., 2002), promotes neuroregeneration (Sawai et al.,1996). In fact, tropomyosin receptor kinase B (TrkB) expressedin the LGN neurons (Avwenagha et al., 2006) and protectscortical neurons through the extracellular signal-regulatedkinase (ERK) and phosphatidylinositol 3-kinase (PI3K) path-ways (Hetman et al., 1999; Nakazawa et al., 2002). In thepresent study, we observed that the increased expressions ofBDNF in the contralateral dLGN and contralateral vLGN-l afterNMDA injection partly merged with the increased expressionsof GFAP, revealing that BDNF was expressed in GFAP-positiveastroglial cells. These data may be interpreted to suggest thatBDNF from GFAP-positive astroglial cells protected neuronalcells in LGN. In fact, expressions of BDNF in GFAP-positiveastroglial cells showed protective effect on transient forebrainischemia in gerbils (Lee et al., 2002). In contrast, BDNF bindingto p75 neurotrophin receptor has been shown to stimulate cell

death (Roux and Barker, 2002; Reichardt, 2006). In thisexperiment, neuronal death was observed after the increasedexpressions of BDNF. Therefore, we cannot rule out thepossibility that neuronal cell death in the contralateral dLGNand contralateral vLGN-l after NMDA injection is triggered bythe increased expressions of BDNF. Further studies will beneeded to clarify the precise roles performed by BDNF in GFAP-positive astroglial cells during the processes leading to LGNdamage following NMDA injection.

In summary, we demonstrated time- and region-dependentalterations in LGN following NMDA-induced retinal damage inmice. Furthermore, BDNF may be a beneficial factor protectingLGN neurons after NMDA-induced retinal damage. Indeed, inour previous report, BDNF has been reported to exhibit aprotective effect against shrinkage of RGC at in the early phaseafter optic nerve axotomy in rats (Ota et al., 2002). Although themechanisms underlying neuronal degeneration and protectionare not yet fully understood, further studies using transgenicmicemay be an effectiveway of elucidating thesemechanisms,to judge from the findings made in the present study.

4. Experimental procedures

4.1. Animals

Male adult C57BL/6J mice weighing 20–32 g (Clea Japan, Inc.Fujimiya, Japan) were kept under lighting conditions of 12 hlight and 12 h dark. All experiments were performed inaccordance with the ARVO Statement for the Use of Animals

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

in Ophthalmic and Vision Research and were approved andmonitored by the Institutional Animal Care and Use Commit-tee of Gifu Pharmaceutical University.

4.2. NMDA injection

Mice were anesthetized with 3.0% isoflurane (Merck, Osaka,Japan) and maintained with 1.5% isoflurane in 70% N2O and30% O2 via an animal general anesthesia apparatus (SoftLander; Sin-ei Industry Co. Ltd., Saitama, Japan). Retinaldamage was induced by the intravitreal injection (2 µl/eye) ofN-methyl-D-aspartate (NMDA; Sigma-Aldrich, St. Louis, MO)dissolvedat 20mMin0.01Mphosphate-buffered saline (PBS) atpH 7.4. This was injected into the vitreous body of the left eyeunder the above anesthesia. One drop of levofloxacin ophthal-mic solution (Santen Pharmaceuticals Co. Ltd., Osaka, Japan)was applied topically to the treated eye immediately after theintravitreal injection. NMDA-treated mice were euthanized at1, 3, 7, 30, 90, and 180 days. Control (non-treated) group, (totaln=13) was consisted ofmice euthanized at 1 (n=4), 30 (n=3), 90(n=3), and 180 days (n=3). Sham group (mice received anintravitreal injection of PBS instead of NMDA) was euthanizedat 1 (n=3), 30 (n=3), 90 (n=3), and 180 days (n=3).

4.3. Sample preparation

At the end of their assigned survival periods, mice wereanesthetized with sodium pentobarbital (80 mg/kg, i.p.) (Nem-butal, Dainippon, Osaka, Japan), and perfused with 4% (w/v)paraformaldehyde solution in 0.01 M PBS (except for immuno-histochemistry of brain-derived neurotrophic factor (BDNF),when 2% (w/v) paraformaldehyde solution in 0.01 M PBS wasused). The brains were removed after 15-min perfusion at 4 °C,

Fig. 9 – Illustrations showing (A) pathway from retina to LGN, (Bmice; boxed area is shown diagrammatically in (C). dLGN, dorsanucleus (lateral); vLGN-m, ventral lateral geniculate nucleus (me

immersed in the same fixative solution for 24 h, soaked in 25%(w/v) sucrose for 1 day, then frozen in embedding compound(Tissue-Tek, Sakura Finetechnical Co. Ltd., Tokyo, Japan).Coronal sections through the lateral geniculate nucleus (LGN)(Bregma −2.30 mm) were cut at 10-µm thickness.

4.4. Histochemistry

Each eye was enucleated at the time of brain removal and 4%paraformaldehyde solution was injected into the vitreousbody. The eye was then kept immersed for at least 24 h in thesame fixative solution at 4 °C. Six paraffin-embedded sec-tions (thickness, 4 µm) cut through the optic disc of each eyewere prepared in a standard manner, and stained withhematoxylin and eosin. Retinal damage was evaluated aspreviously described (Yoneda et al., 2001), three sectionsfrom each eye being used for the morphometric analysis.Light-microscope images were photographed, and (i) the cell-counts in the ganglion cell layer (GCL) at 375–625 µm from theoptic disc, and (ii) the thickness of the inner plexiform layer(IPL) were measured on the photographs in a masked fashionby a single observer (Y.I.). Data from three sections (selectedrandomly from the six sections) were averaged for each eye,and these were used to evaluate the GCL cell-count and IPLthickness.

4.5. Anatomical tracing

Intravitreal injection (2 µl/eye) of an anterograde axonal tracer,wheatgerm agglutinin–horseradish peroxidase (WGA–HRP;Sigma-Aldrich), dissolved in saline at 5% were made into the lefteye under isoflurane anesthesia. Animals were perfused with 2%paraformaldehydesolutionunderpentobarbital anesthesia at 46–

) coronal section through level of LGN (Bregma −2.30 mm) inl lateral geniculate nucleus; vLGN-l, ventral lateral geniculatedial).

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

50 h after tracer injection, then the brains were removed. Brainswere soaked in 25% (w/v) sucrose in 0.1 M phosphate buffer, pH7.4, at 4 °C for 1 day, then frozen in embedding compound(Sakura). Coronal sections through the LGN (Bregma −2.30 mm)were cut at 10-µm thickness. Coronal sections containing LGNwere placed on slides (MASCOAT; Matsunami, Osaka, Japan) andwashed with 0.01 M PBS. WGA–HRP-labeled sections werevisualized using True blue (Kimble Glass Inc, Texas, CO) retro-grade fluorescent labeling, it was determined that the vastmajority of LGN neurons projecting form retina.

4.6. Immunohistochemistry

For immunohistochemistry, coronal sections containing LGNwere placed on slides (MASCOAT; Matsunami, Osaka, Japan)and washed with 0.01 M PBS, then treated with 0.3% hydrogenperoxidase inmethanol for 30min at room temperature. Next,sections were blocked with Mouse onMouse (M.O.M.) blockingreagent (M.O.M. immunodetection kit; Vector, Burlingame,CA), then incubated either withmouse anti-NeuNmonoclonalantibody (1: 250 dilution) (MAB377, Chemicon, Temecula, CA)or with mouse anti-glial fibrillary acidic protein (GFAP)monoclonal antibody (CIT611, Ylem, Rome, Italy) for 1 day at4 °C. They were washed with 0.01 M PBS and then incubatedwith biotinylated anti-mouse IgG before being incubated withavidin–biotin–peroxidase complex for 30 min at room tem-perature, and finally visualized using DAB as a peroxidasesubstrate.

In the immunostaining procedures for brain-derived neuro-trophic factor (BDNF), coronal sections containing LGN werewashedwith0.01MPBScontaining0.05%Tween20 (PBS-T), thentreated with 0.3% hydrogen peroxidase inmethanol. Next, theywere preincubated with 10% normal goat serum (Vector) in0.01 M PBS for 30 min, then incubated for 1 day at 4 °C withspecific rabbit anti-BDNF polyclonal antibody (Furukawa et al.,1998) (1: 1000 dilution) in the following solution: 10% normalgoat serum in 0.01 M PBS containing 0.3% (v/v) Triton X-100.They were washed with 0.05% PBS-T and then incubated withbiotinylated anti-rabbit IgG before being incubated with theavidin–biotin–peroxidase complex for 30min at room tempera-ture, and finally visualized using DAB as a peroxidase substrate.

To visualize co-localization of BDNF with NeuN or GFAP,double immunofluorescence was performed on LGN sectionsfrom NMDA-treated mice. Coronal sections containing LGNwere washed with PBS-T, then treated with 0.3% hydrogenperoxidase inmethanol for 30min at room temperature. Next,they were preincubated with 10% normal goat serum in PBSfor 30 min, then incubated overnight at 4 °C with rabbit anti-BDNF monoclonal antibody (1: 1000 dilution) in the followingsolution: 10% normal goat serum in PBS with 0.3% (v/v) TritonX-100. Sections were blocked with M.O.M. blocking reagent(M.O.M. immunodetection kit; Vector), then incubated eitherwith mouse anti-NeuN monoclonal antibody (1: 250 dilution)(Chemicon) or with mouse anti-GFAP monoclonal antibody(Ylem) for 1 day at 4 °C. They were washed with PBS-T andthen incubated for 3 h at room temperature with amixture ofAlexa Fluor 488 F(ab′)2 fragment of goat anti-rabbit IgG (H+L)(1:1000 dilution) (Molecular Probes, Eugene, OR) and AlexaFluor 488 F(ab′)2 fragment of goat anti-mouse IgG (H+L)(1:1000 dilution) (Molecular Probes).

4.7. Data analysis

Some sections from each brain were used for morphometricanalysis of the optic tract. Light-microscope images werephotographed, (i) the thickness of the optic tract at a positionneighboring the intergeniculate leaf (IGL), and (ii) the area ofthe optic tract in the region from dLGN to vLGN weremeasured on the photographs by a single observer (Y.I.).

For quantification of immunostained sections, the area ofthe LGN was divided into three regions, dLGN; dorsal lateralgeniculate nucleus, vLGN-l; ventral lateral geniculate nucleus(lateral), vLGN-m; ventral lateral geniculate nucleus (medial)(Fig. 9). The quantification wasmade within a field outlined ineach region. The average area of each field (n=6) was0.1551 mm2 for the tissue area of dLGN, 0.0972 mm2 for thetissue area of vLGN-l, and 0.0292 mm2 for the tissue area ofvLGN-m. Immunostained specimens were used for countingNeuN- and BDNF-positive cells, and for measuring the densityof GFAP immunoreactivity.

For measurement of the cell-size of NeuN-immunostainedneurons, a microdensitometry system was set up to scan theselected area of the digitized image and recognize those pixelsin the image with an optical density exceeding a predeter-mined threshold. The threshold was selected so as to detectintensely NeuN-immunostained neurons. For each samplearea, the fraction containing suprathreshold pixels wascalculated, then expressed as the radius of the NeuN-immu-nostained neurons.

Data are presented asmeans±S.E.M. Statistical comparisons(one-way ANOVA followed by a Student's t-test, Fisher's PLSDtest, Dunnett's test, or Bonferroni-test) were made using STATVIEW version 5.0 (SAS Institute, Inc., Cary, NC, USA). A value ofP<0.05 was considered to indicate statistical significance.

Acknowledgments

Thisstudywassupported inpartbyGrants-in-Aid forexploratoryresearch from theMinistry of Education, Culture, Sports, Science,and Technology, Japan (Nos. 18209053 and 18210101).

R E F E R E N C E S

Agarwala, S., Kalil, R.E., 1998. Axotomy-induced neuronal deathand reactive astrogliosis in the lateral geniculate nucleusfollowing a lesion of the visual cortex in the rat. J. Comp.Neurol. 392, 252–263.

Akaike, A., Adachi, K., Kaneda, K., 1998. Techniques for evaluatingneuronal death of the retina in vitro and in vivo. NipponYakurigaku Zasshi 111, 97–104.

Avwenagha, O., Bird, M.M., Lieberman, A.R., Yan, Q., Campbell, G.,2006. Patterns of expression of brain-derived neurotrophicfactor and tyrosine kinase B mRNAs and distribution andultrastructural localization of their proteins in the visualpathway of the adult rat. Neurosci. 140, 913–928.

Brenneman, D.E., Neale, E.A., Foster, G.A., d'Autremont, S.W.,Westbrook, G.L., 1987. Nonneuronal cellsmediate neurotrophicaction of vasoactive intestinal peptide. J. Cell Biol. 104,1603–1610.

Canady, K.S., Hyson, R.L., Rubel, E.W., 1994. The astrocyticresponse to afferent activity blockade in chick nucleus

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

magnocellularis is independent of synaptic activation age andneuronal survival. J. Neurosci. 14, 5973–5985.

Dräger, U.C., Olsen, J.F., 1980. Origins of crossed and uncrossedretinal projections in pigmented and albino mice. J. Comp.Neurol. 191, 383–412.

Dreyer, E.B., Zurakowski, D., Schumer, R.A., Podos, S.M., Lipton, S.A.,1996. Elevated glutamate levels in the vitreous body of humansand monkeys with glaucoma. Arch. Ophthalmol. 114, 299–305.

Ernfors, P., Wetmore, C., Olson, L., Persson, H., 1990. Identificationof cells in rat brain and peripheral tissues expressing mRNAfor members of the nerve growth factor family. Neuron 5,511–526.

Falke, E., Nissanov, J.,Mitchell, T.W., Bennett, D.A., Trojanowski, J.Q.,Arnold, S.E., 2003. Subicular dendritic arborization inAlzheimer'sdisease correlates with neurofibrillary tangle density. Am. J.Pathol. 163, 1615–1621.

Finn, J.T., Weil, M., Archer, F., Siman, R., Srinivasan, A., Raff, M.C.,2000. Evidence that Wallerian degeneration and localized axondegeneration induced by local neurotrophin deprivation do notinvolve caspases. J. Neurosci. 20, 1333–1341.

Furukawa, S., Sugihara, Y., Iwasaki, F., Fukumitsu, H., Nitta, A.,Nomoto, H., Furukawa, Y., 1998. Brain-derived neurotrophicfactor-like immunoreactivity in the adult rat central nervoussystem predominantly distributed in neurons with substantialamounts of brain-derived neurotrophic factor messenger RNAor responsiveness to brain-derived neurotrophic factor.Neurosci. 82, 653–670.

Gittins, R., Harrison, P.J., 2004. Neuronal density, size and shape inthe human anterior cingulate cortex: a comparison of Nissl andNeuN staining. Brain Res. Bull. 63, 155–160.

Gonzalez, D., Satriotomo, I., Miki, T., Lee, K.Y., Yokoyama, T.,Touge, T., Matsumoto, Y., Li, H.P., Kuriyama, S., Takeuchi, Y.,2006. Changes of parvalbumin immunoreactive neurons andGFAP immunoreactive astrocytes in the rat lateral geniculatenucleus following monocular enucleation. Neurosci. Lett. 395,149–154.

Gupta, N., Ang, L.C., de Tilly, Noel, Bidaisee, L., Yucel, Y.H., 2006.Human glaucoma and neural degeneration in intracranialoptic nerve, lateral geniculate nucleus, and visual cortex. Br. J.Ophthalmol. 90, 674–678.

Hajos, F., Kalman, M., Zilles, K., Schleicher, A., Sotonyi, P., 1990.Remote astrocytic response as demonstrated by glial fibrillaryacidic protein immunohistochemistry in the visual cortex ofdorsal lateral geniculate nucleus lesioned rats. Glia. 3,301–310.

Harada, T., Harada, C., Nakamura, K., Quah, H.M., Okumura, A.,Namekata, K., Saeki, T., Aihara, M., Yoshida, H., Mitani, A.,Tanaka, K., 2007. The potential role of glutamate transportersin the pathogenesis of normal tension glaucoma. J. Clin. Invest.117, 1763–1770.

Hardingham, G.E., Fukunaga, Y., Bading, H., 2002. ExtrasynapticNMDARs oppose synaptic NMDARs by triggering CREB shut-offand cell death pathways. Nat. Neurosci. 5, 405–414.

Hetman, M., Kanning, K., Cavanaugh, J.E., Xia, Z., 1999.Neuroprotection by brain-derived neurotrophic factor ismediated by extracellular signal-regulated kinase andphosphatidylinositol 3-kinase. J. Biol. Chem. 274,22569–22580.

Honkanen, R.A., Baruah, S., Zimmerman, M.B., Khanna, C.L.,Weaver, Y.K., Narkiewicz, J., Waziri, R., Gehrs, K.M., Weingeist,T.A., Boldt, H.C., et al., 2003. Vitreous amino acid concentrationsin patients with glaucoma undergoing vitrectomy. Arch.Ophthalmol. 121, 183–188.

Ikeda, K., Tanihara, H., Honda, Y., Tatsuno, T., Noguchi, H.,Nakayama, C., 1999. BDNF attenuates retinal cell death causedby chemically induced hypoxia in rats. Invest. Ophthalmol. Vis.Sci. 40, 2130–2140.

Inomata, Y., Hirata, A., Yonemura, N., Koga, T., Kido, N., Tanihara,H., 2003. Neuroprotective effects of interleukin-6 on

NMDA-induced rat retinal damage. Biochem. Biophys. Res.Commun. 302, 226–232.

Ishii, T.,Moryoshi, K., Sugihara, H., Sakurada, K., Kadotani,H., Yokoi,M., Akazawa, C., Shigemoto, R., Mizuno, N., Masu, M., Nakanishi,N., 1993. Molecular characterization of a family ofN-methyl-D-aspartate receptor subunits. J. Biol. Chem. 268,2834–2843.

Jakobs, T.C., Ben, Y., Masland, R.H., 2007. Expression of mRNA forglutamate receptor subunits distinguishes the major classes ofretinal neurons, but is less specific for individual cell types.Mol. Vis. 13, 933–948.

Jongen-Rêlo, A.L., Feldon, J., 2002. Specific neuronal protein: a newtool for histological evaluation of excitotoxic lesions. Physiol.Behav. 76, 449–456.

Kobayashi, M., Wen, X., Buckmaster, P.S., 2003. Reduced inhibitionand increased output of layer II neurons in the medialentorhinal cortex in a model of temporal lobe epilepsy.J. Neurosci. 23, 8471–8479.

Kordower, J.H., Chu, Y., Stebbins, G.T., DeKosky, S.T., Cochran, E.J.,Bennett, D., Mufson, E.J., 2001. Loss and atrophy of layer IIentorhinal cortex neurons in elderly peoplewithmild cognitiveimpairment. Ann. Neurol. 49, 202–213.

Koyama, Y., 2002. Functional alterations of astroglia on brainpathologiesandtheir intracellularmechanisms.Folia Pharmacol.Jpn. 119, 135–143.

Lam, T.T., Abler, A.S., Kwong, J.M., Tso, M.O., 1999.N-methyl-D-aspartate (NMDA)-induced apoptosis in rat retina.Invest. Ophthalmol. Vis. Sci. 40, 2391–2397.

Lee, T.H., Kato, H., Chen, S.T., Kogure, K., Itoyama, Y., 2002.Expression disparity of brain-derived neurotrophic factorimmunoreactivity and mRNA in ischemic hippocampalneurons. Neuroreport 13, 2271–2275.

Li, Y., Schlamp, C.L., Poulsen, G.L., Jackson, M.W., Griep, A.E.,Nickells, R.W., 2002. p53 regulates apoptotic retinal ganglioncell death induced by N-methyl-D-aspartate. Mol. Vis. 8,341–350.

Lu, W.Y., Jackson, M.F., Bai, D., Orser, B.A., MacDonald, J.F., 2000. InCA1 pyramidal neurons of the hippocampus protein kinase Cregulates calcium-dependent inactivation of NMDA receptors.J. Neurosci. 20, 4452–4461.

MacDermott, A.B., Mayer, M.L., Westbrook, G.L., Smith, S.J., Barker,J.L., 1986. NMDA-receptor activation increases cytoplasmiccalcium concentrations in cultured spinal cord neurons.Nature 321, 519–522.

Martin, K.R., Levkovitch-Verbin, H., Valenta, D., Baumrind, L.,Pease, M.E., Quigley, H.A., 2002. Retinal glutamatetransporter changes in experimental glaucoma and after opticnerve transection in the rat. Invest. Ophthalmol. Vis. Sci. 43,2236–2243.

Meguro, H., Mori, H., Araki, K., Kushiya, E., Kutsuwada, T.,Yamazaki, M., Kumanishi, T., Arakawa, M., Sakimura, K.,Mishina, M., 1992. Functional characterization of heteromericNMDA receptor channel expressed from cloned cDNAs. Nature357, 70–74.

Merlio, J.P., Ernfors, P., Jaber, M., Persson, H., 1992. Molecularcloning of rat trkC and distribution of cells expressingmessenger RNAs for members of the trk family in the ratcentral nervous system. Neurosci. 5, 513–532.

Meshul, C.K., Seil, F.J., Herndon, R.M., 1987. Astrocytes play a rolein regulation of synaptic density. Brain Res. 402, 139–145.

Mullen, R.J., Buck, C.R., Smith, A.M., 1992. NeuN, a neuronalspecific nuclear protein in vertebrates. Development 116,201–211.

Nakazawa, T., Tamai, M., Mori, N., 2002. Brain-derivedneurotrophic factor prevents axotomized retinal ganglion celldeath through MAPK and PI3K signaling pathways. Invest.Ophthalmol. Vis. Sci. 43, 3319–3326.

Nakazawa, T., Shimura, M., Endo, S., Takahashi, H., Mori, N.,Tamai, M., 2005. N-Methyl-D-aspartic acid suppresses Akt

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

activity through protein phosphatase in retinal ganglion cells.Mol. Vis. 11, 1173–1182.

Osborne, N.N., Ugarte, M., Chao, M., 1999. Neuroprotection inrelation to retinal ischemia and relevance to glaucoma. Surv.Ophthalmol. 43, 102–128.

Ota, T., Hara, H., Miyawaki, N., 2002. Brain-derived neurotrophicfactor inhibits changes in soma-size of retinal ganglion cellsfollowing optic nerve axotomy in rats. J. Ocular Pharmacol.Ther. 18, 241–249.

Papadia, S., Stevenson, P.,Hardingham,N.R., Bading,H.,Hardingham,G.E., 2005. Nuclear Ca2+ and the cAMP response element-bindingprotein family mediate a late phase of activity-dependentneuroprotection. J. Neurosci. 25, 4279–4287.

Pisani, F., Costa, C., Caccamo, D., Mazzon, E., Gorgone, G., Oteri, G.,Calabresi, P., Ientile, R., 2006. Tiagabine and vigabatrin reducethe severity of NMDA-induced excitotoxicity in chick retina.Exp. Brain Res. 17, 511–515.

Quigley, H.A., Dunkelberger, G.R., Green, W.R., 1989. Retinalganglion cell atrophy correlated with automated perimetryin human eyes with glaucoma. Am. J. Ophthalmol. 107,453–464.

Rameau, G.A., Akaneya, Y., Chiu, L., Ziff, E.B., 2000. Role of NMDAreceptor functional domains in excitatory cell death.Neuropharmacology 39, 2255–2266.

Reichardt, L.F., 2006. Neurotrophin-regulated signalling pathways.Philos. Trans. R Soc. Lond., B Biol. Sci. 361, 1545–1564.

Ringstedt, T., Lagercrantz, H., Persson, H., 1993. Expression ofmembers of the trk family in the developing postnatal ratbrain. Brain Res. Dev. Brain Res. 72, 119–131.

Roux, P.P., Barker, P.A., 2002. Neurotrophin signaling through thep75 neurotrophin receptor. Prog. Neurobiol. 67, 203–233.

Sawai, H., Clarke, D.B., Kittlerova, P., Bray, G.M., Aguayo, A.J., 1996.Brain-derived neurotrophic factor and neurotrophin-4/5stimulate growth of axonal branches from regenerating retinalganglion cells. J. Neurosci. 16, 3887–3894.

Soriano, F.X., Hardingham, G.E., 2007. Compartmentalized NMDAreceptor signalling to survival and death. J. Physiol. 584, 381–387.

Sucher, N.J., Lipton, S.A., Dreyer, E.B., 1997. Molecular basis ofglutamate toxicity in retinal ganglion cells. Vision. Res. 37,3483–3493.

Vrabec, J.P., Levin, L.A., 2007. The neurobiology of cell death inglaucoma. Eye 21, 11–14.

Wang, X., Sam-Wah, T.S., Ng, Y.K., 2000. Nitric oxide, microglialactivities and neuronal cell death in the lateral geniculatenucleus of glaucomatous rats. Bain Res. 878, 136–147.

Weibel, D., Kreutzberg, G.W., Schwab, M.E., 1995. Brain-derivedneurotrophic factor (BDNF) prevents lesion-induced axonaldie-back in young rat optic nerve. Brain Res. 679, 249–254.

Yan, Q., Rosenfeld, R.D., Matheson, C.R., Hawkins, N., Lopez, O.T.,Bennett, L., Welcher, A.A., 1997. Expression of brain-derivedneurotrophic factor protein in the adult rat central nervoussystem. Neurosci. 78, 431–448.

Yoneda, S., Tanihara, H., Kido, N., Honda, Y., Goto, W., Hara, H.,Miyawaki, N., 2001. Interleukin-1β mediates ischemic injury inthe rat retina. Exp. Eye Res. 73, 661–667.

Yücel, Y.H., Zhang, Q., Gupta, N., Kaufman, P.L., Weinreb, R.N.,2000. Loss of neurons in magnocellular and parvocellularlayers of the lateral geniculate nucleus in glaucoma. Arch.Ophthalmol. 118, 378–384.

Yücel, Y.H., Zhang, Q., Weinreb, R.N., Kaufman, P.L., Gupta, N., 2001.Atrophy of relay neurons in magno- and parvocellular layers inthe lateral geniculate nucleus in experimental glaucoma. Invest.Ophthalmol. Vis. Sci. 42, 3216–3222.

Yücel, Y.H., Gupta, N., Zhang, Q., Mizisin, A.P., Kalichman, M.W.,Weinreb, R.N., 2006. Memantine protects neurons fromshrinkage in the lateral geniculate nucleus in experimentalglaucoma. Arch. Ophthalmol. 124, 217–225.

Zoli, M., Grimaldi, R., Ferrari, R., Zini, I., Agnati, L.F., 1997.Short- and long-term changes in striatal neurons and astrogliaafter transient forebrain ischemia in rats. Stroke 28, 1049–1058.