d-Serine regulates CREB phosphorylation induced by NMDA receptor activation in Müller glia from the retina

Page 1

A

rtaCr©

K

TasggaifiwcrthRbbiot

0d

Available online at www.sciencedirect.com

Neuroscience Letters 427 (2007) 55–60

d-Serine regulates CREB phosphorylation induced by NMDA receptoractivation in Muller glia from the retina

Monica Lamas a,∗,1, Irene Lee-Rivera b, Monica Ramırez a, Ana Marıa Lopez-Colome b,1

a Departamento de Farmacobiologıa, CINVESTAV Sede Sur, Calzada de los Tenorios 235, Mexico, D.F., Mexicob Instituto de Fisiologıa Celular, Departamento de Neurociencias, Universidad Nacional Autonoma de Mexico (UNAM), Mexico

Received 19 June 2007; received in revised form 28 August 2007; accepted 9 September 2007

bstract

d-Serine is an N-methyl-d-aspartate (NMDA) receptor coagonist predominantly produced by glial cells in the brain and the retina. Whereas aole for d-serine as a modulator of NMDA receptors in neurons has been suggested, its function in glial cells has not been analyzed. We here showhat d-serine modulates gene expression in Muller glial cells from the retina through the induction of transcription factor CREB phosphorylation

nd the expression of the immediate-early gene c-fos. Pharmacological analysis indicates that d-serine effect involves NMDA receptor activation.omparison of the effect of d-serine in Muller cells, hippocampal astrocytes and hippocampal neurons suggests that d-serine could function as a

etinal NMDA receptor coagonist activating functionally relevant transcription factor pathways in glial cells. 2007 Elsevier Ireland Ltd. All rights reserved.

ctors;

Hc

trtpsippb

ndfc

eywords: Glutamate receptors; Muller glia; Retina; d-Serine; Transcription fa

he N-methyl-d-aspartate (NMDA) receptors require the oblig-tory binding of a coagonist to the glycine-site of the NR1ubunit in order to exert their physiological actions [8]. Arowing body of evidence suggests that d-serine rather thanlycine, the “classic” coagonist, might be the endogenous lig-nd activating this site [24,17,9]. Specifically in the retina,t has been shown that endogenous d-serine is a necessaryactor for neuronal NMDA receptor full activation [20]. Orig-nally, d-serine and its synthesizing enzyme serine racemaseere described as glial-derived factors [24], promoting the

onsideration of d-serine as a novel unique putative glial neu-otransmitter [3]. In fact, in the retina, glial cells appear to behe primary source of d-serine since its synthesizing enzymeas been exclusively found in Muller cells and astrocytes [20].egarding neuronal-glial communication, recent research haseen mainly directed to the clarification of the mechanismsy which non-neuronal cells can modulate the processing of

nformation within the nervous system, focusing particularlyn the participation of glia-derived d-serine in the regula-ion of NMDA receptor responses in neurons [24,17,9,20].

∗ Corresponding author. Tel.: +52 55 50612876; fax: +52 55 50612863.E-mail address: [email protected] (M. Lamas).

1 These authors contributed equally to this work.

omobctec

304-3940/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.neulet.2007.09.009

c-fos

owever, little is known about the effect of d-serine on glialells.

NMDA receptors exert their physiological effects throughhe generation of intracellular Ca2+ transients which ultimatelyesult in the activation of several transcription factors [6]. Amonghese, the transcription factor cAMP-response element-bindingrotein (CREB) plays a key role in NMDA-receptor-mediatedynaptic plasticity and neuronal survival [19]. Consistent withts role as an activity-dependent transcription factor, CREB ishosphorylated on a serine residue critical for its function inhysiologically active brain areas, controlling a wide range ofehaviors [4].

In the vertebrate retina, NMDA receptors are present ineurons as well as in Muller cells [14,22]. We have recentlyemonstrated that NMDA receptors in neurons and Muller gliarom the retina show cell-specific structural and pharmacologi-al properties, which could underlie the differential participationf neurons and glia in the physiologic control of glutamate trans-ission in the retina [10,11,25]. Here we undertook the analysis

f the modulation of NMDA receptor-mediated gene expressiony d-serine in Muller cells from the rat retina. Our findings indi-

ate that co-activation of NMDA receptors by d-serine induceshe rapid phosphorylation of CREB and early response genexpression in rat Muller glia and thus d-serine co-activatesrucial transcription factor pathways opening a new perspec-

Page 2

5 ence L

tr

daf

pG(t3ldccrpBK2bowgdc(ffb

tca

anwode5Catircas

f

tcm

di

FcCN

6 M. Lamas et al. / Neurosci

ive in the physiologic effect of glutamate transmission in theetina.

All experiments and laboratory animal treatments were con-ucted as established by the Association for Research in Visionnd Ophthalmology (ARVO) Statement for the Use of Animalsor Ophthalmic and Vision Research.

Enucleated eyes of Long-Evans rats 8–10 days of age werelaced in Dulbecco’s Minimal Essential Medium (DMEM;ibco BRL, Gaithersburg, MD) with 10% fetal bovine serum

FBS) and 1:1000 penicillin–streptomycin overnight at roomemperature, in the dark. The eyes were incubated at 37 ◦C for0 min in DMEM containing 0.1% trypsin and 70 IU/ml col-agenase and transferred to 10% FBS-DMEM. Cells from twoissociated retinas were seeded onto a six-well Petri dish andultured in OPTIMEM (Gibco BRL) containing 4% FBS, untilonfluent. Hippocampal neurons were isolated from surgicallyemoved rat embryos collected “in utero”. The dissected hip-ocampi were incubated for 10 min at 37 ◦C in Krebs–Ringericarbonate buffer (KRB) pH 7.4 containing (mM): NaCl, 118;Cl, 4.7; CaCl2, 2.5; MgSO4, 1.17; KH2PO4, 2.0; NaHCO3,5; glucose, 5.6, plus 0.25% trypsin. The reaction was stoppedy 2 min incubation in KRB containing 3.8% MgSO4, 1 unitf DNAse I (Worthington) and trypsin inhibitor. The tissueas mechanically dissociated, the cells collected by centrifu-ation and plated on poly-l-lysine-coated 24-well tissue cultureishes at a density of 1.5 × 106 cells/well in Neurobasal mediumontaining 0.5 mM glutamine, 25 �M glutamic acid and 1%v/v) B27 supplement. Hippocampal astrocytes were obtained

rom 6-day-old rats by enzymatic dissociation, as describedor hippocampal neurons. The trypsin reaction was stoppedy the addition of serum. Cells were separated by centrifuga-

be1

ig. 1. (A) Photomicrograph of a confluent culture of rat Muller cells. (B) Immunofluounterstained with DAPI. Calibration bar: 50 �m. (C) Immunostaining control: ratalibration bar: 200 �m. (D) Immunofluorescence image of cultured Muller cells imo immunoreactive cells are detected. Calibration bar: 200 �m.

etters 427 (2007) 55–60

ion, resuspended and seeded in OPTIMEM plus 4% FBS. Allultures were maintained in a humidified 5% CO2 incubatort 37 ◦C.

Prior to every experiment, cell cultures were washed in PBSnd incubated in KRB for 24 h in order to minimize the exoge-ous d-serine present in commercial media. Control cultureells for each experimental condition were maintained with-ut treatment; the sister cultures were treated with the followingrugs, alone or in combination, as indicated in the figure leg-nds: 100 �M NMDA, 100 �M d-serine, 100 �M glycine and0 �M MK801 (Tocris). In calcium-dependence experiments,aCl2, was omitted from the KRB solution and EGTA (500 �M)nd BAPTA-AM (10 �M) were included for 15 min prior to thereatment. When glycine was tested, 500 �M strychnine wasncluded, in order to prevent the activation of inhibitory glycineeceptors. d-Amino acid oxidase (DAAOX) was added to theells for 2 h. After this time the cells were washed extensivelynd incubated in KRB (control cells) or KRB + NMDA + d-erine for an additional hour.

Except when stated, all chemicals and reagents used wererom Sigma Chemical Co. (St. Louis, MO).

Cell suspension was mixed with equal volume of 0.4% iso-onic trypan blue solution. Total number of viable cells wasounted after 2 min in a Neubauer haemocytometer under lighticroscope.The cells were cultured on poly-d-lysine coated slides for five

ays and fixed with 4% paraformaldehyde for 30 min, washedn PBS and blocked in 1% gelatin, in PBS. CRALBP anti-

ody (Affinity BioReagents) and rabbit anti-neuron specificnolase (NSE) polyclonal antibody (Chemicon) were diluted:30 and 1:100, respectively, in the blocking solution and added

orescence image of cultured Muller cells positive for CRALB immunostainingMuller cells incubated in secondary antibody and counterstained with DAPI.

munostained for neuron specific enolase (NSE) and counterstained with DAPI.

Page 3

scien

ti((3(m

b0ta

pitP1dda

F3tCampoaEoisN

M. Lamas et al. / Neuro

o the slides for 10 h at 4 ◦C. After extensive washing, the stain-ng was developed using FITC-coupled anti-mouse antibody1:200 dilution; Jacksons Laboratory). Counterstaing with DAPI4′,6-diamidino-2-phenylindole) (0.1 mg/ml) was performed for0 min in the dark. The slides were mounted with VectashieldVector Laboratories) and observed using a fluorescence invertedicroscope (NIKON Eclipse E600).Cultured cells were lysed in boiling Laemmli Buffer (Trizma-

ase pH 6.8 0.5 M; glycerol 5%; SDS 2%; �-mercaptoethanol.5%) to avoid dephosphorylation and degradation of the pro-eins. The extracts were quantified with the Biorad quantificationssay (Biorad Laboratories). Approximately 10 �g/well of total

tweP

ig. 2. (A) CREB phosphorylation levels are increased by NMDA + d-serine treatme0 min with 100 �M NMDA and increasing concentrations of d-serine: 1 �M, 10 �Mo PVDF membranes and probed with rabbit anti-P-CREB antibody. Following analyorresponding densitometric analysis is shown on the right. The graph represents ts the relative level of P-CREB phosphorylation in treated samples with respect toean ± S.E.M. of three independent experiments. *P < 0.5, significant difference betw

hosphorylation involves NMDA receptor and intracellular Ca2+ concentration increar 100 �M d-serine in the absence of NMDA; medium panel: cells were left untreatedbsence of 50 �M MK801; right panel: cells were either untreated (control) or treateGTA and 10 �M BAPTA-AM in the absence of extracellular calcium. Total proteinf CREB phosphorylation in cultured Muller cells. Cells were either untreated (contrndicated time period. Total protein extracts were analyzed as in (A) and probed in adignal relative to actin control signal. Results are plotted and expressed as in (A).MDA + d-serine.

ce Letters 427 (2007) 55–60 57

rotein extracts from cultured cells were resolved by denatur-ng (SDS 10%) polyacrylamide gel electrophoresis (PAGE),ransferred to poly(vinylidene difluoride) membranes (Hybond-, Amersham Pharmacia) and probed with mouse anti-CRALBP/500 dilution, rabbit anti p-CREB (Cell Signaling) 1/2000ilution, rabbit anti-CREB (Chemicon International) 1/1000ilution, rabbit anti c-fos (Sigma Chemicals) 1/1000 dilutionnd mouse anti-actin (Chemicon International) 1/1000 dilu-

ion. Horseradish peroxidase-conjugated secondary antibodiesere applied for detecting the primary antibody signal using the

nhanced chemiluminiscence method detection system (ECLlus, Amersham). The films were subjected to densitometric

nt in cultured Muller cells. Cells were either untreated (control) or treated forand 100 �M. Total protein extracts were separated by SDS/PAGE, transferred

sis, the blots were re-probed with mouse anti-actin antibody as loading control.he level of P-CREB signal relative to actin control signal. Results are plottedcontrol cells (control cell level = 1). Results on the graph are expressed as theeen treated and control Muller cells. (B) d-Serine mediated induction of CREB

se. Left panel: cells were either untreated (control) or treated with 1 �M, 10 �M(control) or treated with 100 �M NMDA + 100 �M d-serine in the presence or

d with 100 �M NMDA + 100 �M d-serine alone or in the presence of 500 �Mextracts were analyzed for P-CREB levels as in (A). (C) Time-course analysisol) or treated with 100 �M NMDA in the presence of 100 �M d-serine for thedition with an anti-CREB antibody. The graph represents the level of P-CREB*P < 0.5, significant difference between control and Muller cells treated with

Page 4

5 ence L

att

itcopsetiOccA9n

tf

RiTrato

iwnpbrppac

F1wster(v

8 M. Lamas et al. / Neurosci

nalysis, and results are expressed as the mean ± S.E.M. Statis-ical significance of the differences was determined by Student’s-test using a Sigma-stat computer program.

Muller cells isolated from the rat retina achieved confluencen 10 days (Fig. 1A). The vascular mammalian retina containswo types of glia: astrocytes and Muller cells, which share severalharacteristics in vitro, including morphology and expressionf several glial markers, whereas cellular retinaldehyde-bindingrotein (CRALBP) and glutamine synthetase (GS1) have beenhown to label exclusively Muller cells [1]. The techniquemployed here for Muller cell culture has been shown to reducehe survival of astrocytes, due to the decreased oxygen supplynduced by overnight incubation of the enucleated eyeballs inPTIMEM [7]. In order to assess the purity of our Muller cell

ulture, we analyzed the expression of CRALBP and neuron spe-ific enolase (NSE) using immunocytochemistry (Fig. 1B–D).nalysis of three independent cultures showed that more than0% of the cells expressed the Muller cell markers (Fig. 1B) and

o expression of NSE was detected (Fig. 1D).

The induction of CREB phosphorylation in Muller cellsreated with NMDA and increasing concentrations of d-serineor 30 min was analyzed by Western blot analysis of cell lysates.

a

ri

ig. 3. Induction of c-fos expression by d-serine, acting as a coagonist of NMDA rec00 �M NMDA and 100 �M d-serine for the indicated period. Total protein extractith rabbit anti-c-fos antibody. The blots were stripped and re-probed with mouse an

hown on the right. The graph represents the densitometric units of c-fos signal relathree independent experiments. (B) Cells were treated with increasing concentrationsither left untreated (control) or treated with 100 �M NMDA and 100 �M d-serine fepresents the densitometric units of c-fos signal relative to actin signal. Results on thC) Trypan blue exclusion assay of control cells and cells treated with increasing conalues of trypan blue-negative (living cells) of three independent experiments. No sig

etters 427 (2007) 55–60

esults in Fig. 2A show that d-serine potentiates NMDA-nduced CREB phosphorylation in a dose-dependent manner.he effect of d-serine is due to an interaction with the NMDA

eceptor coagonist site, since no stimulation was observed in thebsence of NMDA (Fig. 2B, left panel). Conversely, the induc-ion of CREB phosphorylation was not observed in the absencef the coagonist (not shown).

In order to further confirm that CREB phosphorylation wasnduced by the specific activation of NMDA receptors, the cellsere stimulated in the presence of the NMDA receptor chan-el blocker MK-801 (dizocilpine maleate) [5]. Fig. 2B (middleanel) shows that the inhibition of NMDA receptor signalingy this compound abolishes the induction of CREB phospho-ylation by NMDA plus d-serine in Muller cells. Experimentserformed in the absence of calcium show no detectable CREBhosphorylation (Fig. 2B, left panel), suggesting that CREBctivation is triggered by the increase in intracellular Ca2+

oncentration resulting from NMDA plus d-serine receptor-

ctivation.

Analysis of the kinetics of d-serine action on the NMDAeceptor-induced activation of CREB in Muller cells shows thatncrease in CREB phosphorylation can be detected as early

eptors in Muller cells. (A) Cells were either untreated (control) or treated withs were separated by SDS/PAGE, transferred to PVDF membranes and probedti-actin antibody for loading control. The densitometric analysis of the gels is

ive to actin signal. Results on the graph are expressed as the mean ± S.E.M. ofof d-amino acid oxidase (Daaox) for 2 h. After extensive washing, cells were

or 1 h. The densitometric analysis of the gels is shown on the right. The graphe graph are expressed as the mean ± S.E.M. of three independent experiments.centrations of d-amino acid oxidase (Daaox) for 2 h, showing mean ± S.E.M.nificant differences between control and treated cells were observed.

Page 5

science Letters 427 (2007) 55–60 59

a3wutis

Clsbea(aasc(csoe

wipapit(sh

aipotcgitsedgidLestao

Fig. 4. Comparative effect of NMDA receptor coagonists on CREB phospho-rylation in Retinal Muller glia, hippocampal neurons and astrocytes in culture.Cells were either untreated or treated for 30 min with 100 �M NMDA + 100 �Md-serine; 100 �M NMDA + 100 �M glycine; 100 �M NMDA or 100 �M d-serine. Total protein extracts were separated by SDS/PAGE, transferred to PVDFmembranes and probed with rabbit anti-P-CREB antibody. The blots were re-probed with mouse anti-actin antibody as loading control. The correspondingdensitometric analyses are shown. The graph represents the level of P-CREBsignal relative to actin control signal. Results are plotted as the relative level of P-CREB phosphorylation in treated samples with respect to control cells (controlcell level = 1). Results on the graph are expressed as the mean ± S.E.M. of threei *

a#

Ne

dbasorcdgprNstudtad

A

o

M. Lamas et al. / Neuro

s 15 min following stimulation, reaches maximal value after0 min, and slowly decays to basal levels (Fig. 2C) in agreementith previous reports on CREB phosphorylation after cell stim-lation [4]. Analysis of total CREB protein, using an antibodyhat recognizes phosphorylated and unphosphorylated CREB,ndicates that the treatment does not affect the protein levels pere (Fig. 2C).

In addition to the activation of transcription factors such asREB, calcium influx through NMDA receptor channel also

eads to the rapid induction of c-fos expression [21]. We herehow that a twofold induction of c-fos expression is activatedy d-serine acting as an NMDA-coagonist (Fig. 3A). Specificnzymatic degradation of the Muller cell derived d-serine by d-minoacid oxidase (Daaox) reduces the basal expression of c-fosFig. 3B). However, when Daaox is eliminated from the mediumnd the cells are treated with NMDA and d-serine, we observen induction of CREB phosphorylation. As degradation of d-erine by Daaox produces hydrogen peroxide creating potentialell harm, we analyzed cell viability after Daaox treatmentFig. 3C) and we demonstrate that no significant differencesan be detected between control and treated cells. These resultsupport a role for Muller-cell derived d-serine in the activationf NMDA receptor-linked signaling pathways in Muller cells,quivalent to those described in neurons.

NMDA receptor-induced phosphorylation of CREB has beenidely studied in hippocampal neurons, and shown to be

nvolved in a number of physiological processes [2]. Com-arison of the induction of CREB phosphorylation by NMDAnd d-serine in cultured rat hippocampal neurons, hippocam-al astrocytes and Muller glia at the time point of maximumnduction (30 min) shows that the effect of d-serine was consis-ently higher in Muller cells than in brain neurons or astrocytesFig. 4). These findings suggest a differential response to d-erine coagonist-action in Muller glia from the retina andippocampal cells (neurons and astrocytes).

The results from this study demonstrate that d-serine, actings NMDA receptor coagonist, can modulate gene expressionn Muller glia, through the specific activation of CREB phos-horylation and immediate early gene expression. Activationf CREB bears considerable physiological relevance, sincehis transcription factor has a number of potential targets thatould drive different effects on the cells including neuronal-lial communication and processing of information [18,23]. Themmediate-early gene c-fos is a member of the activator pro-ein (AP-1) transcription factor family that has been extensivelytudied in neurons where it couples neuronal activity to genexpression [16]. Glial c-fos expression in the brain has beenocumented in pathological conditions where it contributes tolial activation [15]. The mechanisms by which neuronal activ-ty induces expression of these transcription factors are wellocumented and involve calcium entry via NMDA receptors or-type calcium channels [2]. However, little is known about theffect of NMDA receptor activation on glial cells. Our analysis

how that the effect of d-serine in Muller glia is comparableo its effect on neurons and astrocytes from the hippocampuslthough with distinct efficiency (Fig. 4), suggesting a degreef cell specificity in the function of d-serine as a coagonist of

itIt

ndependent experiments. P < 0.5, significant difference between Muller cellsnd hippocampal cells (neurons and astrocytes) receiving the same treatment.P < 0.5, significant difference between control and treated cells.

MDA receptors in the nervous system, in addition to the wellstablished regional differences [18].

Further understanding of the physiological consequences of-serine modulation of transcription factor expression inducedy NMDA receptor activation in retinal glial cells may haven impact on the treatment of retinal degeneration in diseasesuch as diabetic retinopathy where glutamate metabolism dis-rder and over-stimulation of the NMDA subtype of glutamateeceptors has been reported [13]. In spite of the fact that Mullerells might not be the primary sites affected during retinalisease, studies of diabetic retina and animal models of hyper-lycemia have revealed changes in Muller cell morphology,rotein expression and physiology well in advance of detectableetinopathy [12,13]. We now show that d-serine interaction withMDA receptors in rat Muller glia drive changes in downstream

ignaling and gene transcription, activating CREB and c-fosranscription factors. It is tempting to speculate that, in a sit-ation of increased glutamate concentration, as is the case ofiabetic retinopathy, expression and activity patterns of theseranscription factors might be de-regulated and therefore couldccount for the involvement of Muller cells in the onset of theisease.

cknowledgements

The authors acknowledge the expert technical assistancef Edith Lopez. We thank Dr. Lourdes Massieu for training

n hippocampal cell culture and Dr. Mark Sefton for revisinghe manuscript. This work was partially supported by GrantsN228203 from PAPIIT/UNAM and 42640-Q from CONACYTo A.M.L.-C.

Page 6

6 ence L

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

0 M. Lamas et al. / Neurosci

eferences

[1] J.G. Arroyo, S. Ghazvini, D.H. Char, An immunocytochemical study ofisolated human retinal Muller cells in culture, Graefes Arch. Clin. Exp.Ophthalmol. 235 (1997) 411–414.

[2] H. Bading, D.D. Ginty, M.E. Greenberg, Regulation of gene expression inhippocampal neurons by distinct calcium signaling pathways, Science 260(1993) 181–186.

[3] D.E. Baranano, C.D. Ferris, S.H. Snyder, Atypical neural messengers,Trends Neurosci. 24 (2001) 99–106.

[4] W.A. Carlezon Jr, R.S. Duman, E.J. Nestler, The many faces of CREB,Trends Neurosci. 28 (2005) 436–445.

[5] R. Dingledine, K. Borges, D. Bowie, S.F. Traynelis, The glutamate receptorion channels, Pharmacol. Rev. 51 (1999) 7–61.

[6] A. Gosh, D.D. Ginty, H. Bading, M.E. Greenberg, Calcium regulation ofgene expression in neuronal cells, J. Neurobiol. 25 (1994) 294–303.

[7] D. Hicks, Y. Courtois, The growth and behaviour of rat retinal Muller cellsin vitro. 1. An improved method for isolation and culture, Exp. Eye Res.51 (2) (1990) 119–129.

[8] J.W. Johnson, P. Ascher, Glycine potentiates the NMDA response in cul-tured mouse brain neuron, Nature 325 (1987) 529–531.

[9] E. Kartvelishvily, M. Shelper, L. Balan, E. Dumin, H. Wolosker, Neuron-derived d-serine: novel means to activate N-methyl-d-aspartate receptors,J. Biol. Chem. 281 (2006) 14151–14162.

10] M. Lamas, I. Lee-Rivera, A.M. Lopez-Colome, Cell-specific expressionof N-methyl-d-aspartate receptor subunits in Muller glia and neurons fromthe chik retina, Invest. Ophtal. Vis. Sci. 46 (2005) 3570–3577.

11] I. Lee, A. Zarain-Herzberg, A.M. Lopez-Colome, Developmental expres-sion of NMDAR1 splice variants in the chick retina, J. Neurosci. Res. 73(2003) 369–383.

12] Q. Li, E. Zemel, B. Miller, I. Perlman, Early retinal damage in experimentaldiabetes: electroretinographical and morphological observations, Exp. Eye

Res. 74 (5) (2002) 615–625.

13] E. Lieth, A.J. Barber, B. Xu, C. Dice, M.J. Ratz, D. Tanase, J.M. Strother,Glial reactivity and impaired glutamate metabolism in short-term diabeticretinopathy, Penn State Retina Research Group, Diabetes 47 (5) (1998)815–820.

[

etters 427 (2007) 55–60

14] A.M. Lopez-Colome, M. Romo de Vivar, Excitatory amino acid receptorsin primary cultures of glial cells from the retina, Glia 4/5 (1991) 431–439.

15] W. Matsunaga, S. Osawa, S. Miyata, T. Kiyohara, Astrocytic Fos expressionin the rat posterior pituitary following LPS administration, Brain Res. 898(2001) 215–223.

16] J.L. Morgan, T. Curran, Stimulus-transcription coupling in the nervoussystem: involvement of the inducible proto-oncogenes fos and jun, Ann.Rev. Neurosci. 14 (1991) 421–451.

17] J.P. Mothet, A. Parent, H. Wolosker Jr., D.J. OıBrady, C. Linden, D. Ferris,M.A. Rogawski, S.H. Snyder, d-Serine is an endogenous ligand for theglycine site of the N-methyl-d-aspartate receptor, Proc. Natl. Acad. Sci.U.S.A. 97 (2000) 4926–5491.

18] M.J. Schell, R.O. Bardy, M.E. Molliver, D.H. Snyder, d-Serine as a neu-romodulator: regional and developmental localizations in rat brain gliaresemble NMDA receptors, J. Neurosci. 17 (1997) 1604–1615.

19] A.J. Shaywitz, M.E. Greenberg, CREB: a stimulus-induced transcriptionfactor activated by a diverse array of extracellular signals, Annu. Rev.Biochem. 68 (1999) 821–861.

20] E.R. Stevens, M. Esquerra, P.M. Kim, E.A. Newman, S.H. Snyder, K.R.Zahs, R.F. Miller, d-Serine and serine racemase are present in the vertebrateretina and contribute to the physiological activation of NMDA receptors,Proc. Natl. Acad. Sci. U.S.A. 100 (2003) 6789–6794.

21] A.M. Szekely, M.L. Barbaccia, H. Alho, E. Costa, In primary culturesof cerebellar granule cells the activation of N-methyl-d-aspartate-sensitiveglutamate receptors induces c-fos mRNA expression, Mol. Pharmacol. 35(4) (1989) 401–408.

22] Y. Uchihori, D.G. Puro, Glutamate as a neuron-to glial signal for mitoge-nesis: role of glial N-methyl-d-aspartate receptors, Brain Res. 613 (1993)212–220.

23] M.R. Walton, N. Dragunow, Is CREB a key to neuronal survival? TrendsNeurosci. 23 (2000) 48–53.

24] H. Wolosker, S. Blackshaw, S.H. Snyder, Serine racemase: a glial enzyme

synthesizing d-serine to regulate glutamate-N-methyl-d-aspartate neuro-transmission, Proc. Natl. Acad. Sci. U.S.A. 96 (1999) 13409–13414.

25] A. Zarain-Herzberg, I. Lee-Rivera, G. Rodrıguez, A.M. Lopez-Colome,Cloning and characterization of the chick NMDA receptor subunit-1 gene,Mol. Brain Res. 137 (2005) 235–251.