Neuroprotection against NMDA excitotoxicity by group I metabotropicglutamate receptors is associated with reduction of

NMDA stimulated currents

Morten Blaabjerg,a Liwei Fang,b Jens Zimmer,a and Andrius Baskysb,c,*a Anatomy and Neurobiology, University of Southern Denmark-Odense, Denmark

b Long Beach VA Medical Center, 06/116A, 5901 E, Seventh Street, Long Beach, CA 90822, USAc Department of Psychiatry and Human Behavior, University of California, Irvine, CA 92668, USA

Received 12 December 2002; revised 13 March 2003; accepted 26 March 2003

Abstract

The neurotransmitter glutamate can have both excitotoxic and protective effects on neurons. The excitotoxic effects have beenintensively studied, whereas the protective effects, including the involvement of metabotropic glutamate receptors (mGluRs), remainunclear. In the present study, we tested the protective effects of the group-I-mGluR agonist (S)-3,5-dihydroxyphenylglycine (DHPG) onorganotypic hippocampal slice cultures exposed to excitotoxic concentrations of N-methyl-D-aspartate (NMDA). Effects of DHPG onelectrophysiological responses induced by NMDA receptor activation were also recorded. Experiments were performed on organotypichippocampal slice cultures derived from 7-day-old rats, with cellular uptake of propidium iodide as a marker for neuronal cell death. Slicecultures pretreated with DHPG (10 or 100 �M) for 2 h prior to exposure to 50 �M NMDA for 30 min displayed reduced propidium iodideuptake, compared to cultures exposed to NMDA only. The neuroprotective effect was confirmed by Hoechst 33342 staining, where theappearance of pycnotic nuclei after NMDA treatment was prevented by the DHPG pretreatment. Using caspase-3 activity to monitor thepresence of apoptosis, failed to demonstrate this type of cell death in CA1 after NMDA application. The protective effect of DHPG wasabolished by the mGluR1 selective antagonist (S)-(�)-�-amino-4-carboxy-2-methylbenzeneacetic acid (LY367385; 5 or 10 �M), whereasthe mGluR5-selective antagonist 2-methyl-6-phenylethynylpyridine (MPEP; 1 �M) had no effect. Voltage-clamping of CA1 pyramidal cellsin cultures treated with 10 �M DHPG for 2 h showed a significant depression of NMDA-induced inward currents compared to untreatedcontrols. We conclude that neuroprotection induced by activation of group-I-mGluRs involve mGluR1 and is associated with decreasedNMDA-stimulated currents.© 2003 Elsevier Science (USA). All rights reserved.

Keywords: mGluR; N-methyl-D-aspartate; Propidium iodide; Neurodegeneration; LY367385; MPEP

Introduction

Glutamate is the major excitatory neurotransmitter in thebrain but is also a potent excitotoxin involved in neurode-generative disorders such as ischemia and Alzheimer’s dis-ease. Pretreatment with moderate levels of glutamate ago-nists can, however, protect neurons from damage caused bysubsequent exposure to excitotoxic concentrations of gluta-

mate. Recent evidence suggests that activation of metabo-tropic glutamate receptors (mGluRs) is crucial for this neu-roprotection (Adamchik and Baskys, 2000; Chiamulera etal., 1992; Koh et al., 1991; Mount et al., 1993; Opitz andReymann, 1993; Siliprandi et al., 1992), particularly againstischemic nerve cell death (Kalda and Zharkovsky, 1999;Maiese et al., 1996; Pizzi et al., 1996b; Sagara and Schubert,1998; Schroder et al., 1999).

Since mGluR-mediated neuroprotection requires proteinkinase C (PKC) activation (Kalda and Zharkovsky, 1999;Kalda et al., 2000; Koh et al., 1991; Pizzi et al., 1996b;Sagara and Schubert, 1998), group-I-mGluRs (mGluR1 and

* Corresponding author. Fax: �1-562-826-5969.E-mail address: abaskys@uci.edu or andrius.baskys@med.va.gov (A.

Baskys).

R

Available online at www.sciencedirect.com

Experimental Neurology 183 (2003) 573–580 www.elsevier.com/locate/yexnr

0014-4886/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved.doi:10.1016/S0014-4886(03)00204-8

mGluR5) are likely to be involved, because these receptorsare coupled to the IP3/DAG signaling pathway leading torelease of Ca2� from intracellular stores and activation ofPKC respectively (for review see Conn and Pin, 1997). Therole of these receptors, however remain poorly understoodand controversial.

There is evidence for both neurotoxic and neuroprotec-tive actions of both group-I-mGluR agonists and antagonists(for review see Nicoletti et al., 1999). For example, there arereports that the mGluR1 and -5 agonist (S)-3,5-dihydroxy-phenylglycine (DHPG), facilitates NMDA receptor function(Awad et al., 2000; Bandrowski et al., 2001; Fitzjohn et al.,1996; Koga et al., 1996; Orlando et al., 2001; Pisani et al.,2001; Skeberdis et al., 2001) just as the nonselective mGluRagonist 1S,3R-1-aminocyclopentane-trans-1,3-dicarboxylicacid (ACPD) or the group-I-mGluR-selective agonistDHPG potentiated NMDA-induced toxicity in spinal cordneurons or hippocampal slice cultures (Blaabjerg et al.,2001; Faden, 1997; Young et al., 1998) and neuronal de-generation by oxygen and glucose deprivation (Pellegrini-Giampietro et al., 1999) There is, however, at least onereport that coapplication of NMDA and ACPD in acutehippocampal slice preparations induced neuroprotection(Pizzi et al., 1996a).

Other evidence that group-I-mGluRs are neuroprotectivecomes from studies of nitric oxide (NO) or ischemia-in-duced cell death (Lin and Maiese, 2001a, 2001b; Maiese etal., 1995, 2000; Vincent et al., 1997), decrease in apoptoticmarkers (Allen et al., 2000; Anneser et al., 1998; Maieseand Vincent, 1999; Vincent and Maiese, 2000), and reduc-tion of NMDA toxicity (Adamchik and Baskys, 2000; Col-well and Levine, 1999; Colwell et al., 1996). Bruno et al.(2001) reported that the group-I-mGluR agonist DHPG wasprotective when repeatedly coapplied with NMDA despitean increase in NMDA-receptor-mediated currents.

In the present study, we investigated whether selectiveactivation of group-I-mGluRs induced neuroprotection andto what extent mGluR1 and mGluR5 are involved. Theeffect of group-I-mGluR activation on NMDA-stimulatedcurrents was also investigated.

Materials and methods

Animals

Organotypic hippocampal slice cultures were preparedand grown by the interface method (Stoppini et al., 1991) aspreviously described (Baskys and Adamchik, 2001).Briefly, 7-day-old Wistar rat pups (Charles River, Raleigh,NC) were anesthetized with halothane and rapidly decapi-tated, and the brains placed into an ice-cold stabilizationmedium [50% minimal essential medium with no bicarbon-ate or glutamate, 50% calcium and magnesium-free Hanks’balanced salt solution, 7.5 mM D-glucose, and 20 mM N-2-hydroxyethyl piperazine-N�-2-ethanosulfonic acid (HEPES),

pH 7.15]. The middorsal segments of the two hippocampiwere dissected out and cut into 400-�m transverse slices.The slices were separated, and excess tissue was removedand placed on 30-mm Millicel-CM 0.4-�m-thick inserts(Millipore, Bedford, MA, USA). The inserts were trans-ferred to a 6-well culture plate (Falcon, Becton DickinsonLabware, NJ, USA) on top of 1 ml of culture medium (50%MEM, 25% horse serum, 25% Earl’s balanced salt solution,D-glucose, HEPES, 5000 units/ml penicillin G, and 5 �l/mlstreptomycin sulfate, pH 7.15). Slices were grown for aminimum of 2 weeks at 36.5°C in 100% humidity, with95% air and 5% CO2 and fed twice weekly via 50% mediumexchange. All experiments were performed according to thestandards described in Animal Welfare Regulations and theGuide for the Care and Use of Laboratory Animals.

Experimental protocol for propidium iodide (PI)experiments

To investigate the potential neuroprotective effects of thegroup-I-mGluR agonist DHPG it was applied to the culturemedium for 2 h prior to a 30-min exposure to 50 �MN-methyl-D-aspartate (NMDA), followed by change tonormal control medium. In the experiments in which thecompetitive mGluR1 antagonist (S)-(�)-�-amino-4-car-boxy-2-methylbenzeneacetic acid (LY367385) or the non-competitive mGluR5 antagonist 2-methyl-6-phenylethy-nylpyridine (MPEP) was used, these compounds wereapplied 30 min before and during the pretreatment period toensure maximal receptor saturation. The same time sched-ule was used in experiments where antagonists were appliedwithout DHPG exposure. Cell death, measured as cellularuptake of PI (3,8-diamino-5-[3-(diethylethylamino)propyl]-6-phenyl phenanthridinium diiodide; Molecular Probes, Eu-gene, OR) was recorded by laser scanning confocal micros-copy (PCM2000, Nikon). Images of the cultures wereobtained before the experiments to determine the baselinecellular uptake of PI and 24 and 48 h after exposure toNMDA. All data were normalized to data obtained after theslices had been kept for 48 h at an ambient temperature of4°C (except for cultures used for Hoechst 33342 andcaspase-3 staining). Measurements of the PI uptake in slicesexposed to 4°C was assumed to reflect the maximal celldeath level in that same slice and were used for two pur-poses. First, the high level of PI uptake allowed the visual-ization of all hippocampal subfields and exclusion of dam-aged slices from analysis. Second, the PI uptake normalizedto the maximal PI uptake in the same slice allowed toquantify and compare cell death levels in individual slicesregardless of their size or thickness. Since all comparisonsof cell death were made in relation to cell death induced byNMDA alone, data were expressed as a percentage of theNMDA-induced cell death (taken as 100%) in that particu-lar experiment. In order to gage the basal level of cell death,one well with five to six cultures in each experimental runwas exposed to culture medium only. The basal cell death

574 M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580

level measured in 8 wells (n � 5–6 slices each) after 24 hof culturing was 0.96 � 0.10 % in relation to the maximallevel of cell death in all areas of the hippocampus. This lowlevel of basal cell death was ignored and not considered incell death calculations. PI images were analyzed off-line bymeasuring the pixel intensity, using ScionImage analysissoftware (available at http://www.scioncorp.com). The ar-eas of interest (CA1, CA3 pyramidal cell layers) were out-lined and superimposed on the images obtained at baseline,24 h and 48 h.

DHPG, MPEP and LY367385 were obtained from TocrisCookson Inc. (Ellisville, MO) and NMDA from Sigma (St.Louis, MO). The drugs were dissolved in sterile distilledwater, aliquoted, and stored at �20°C until use.

Processing of tissue for immunohistochemistry

To investigate whether the lesion induced by NMDAtreatment resulted in apoptotic or necrotic cell death, weperformed double stainings of Hoechst 33342 and activatedcaspase-3. Cultures were treated with NMDA as describedabove and fixed after 24 and 48 h in 4% paraformaldehydefor 1 h followed by cryoprotection in 20% sucrose in phos-phate buffer for 24 h. Cultures were then frozen on CO2 iceand cryostat-sectioned in 20-�m thick sections and kept at�20°C until processed for immunohistochemistry.

Immunohistochemistry

After thawing to room temperature, sections were rinsedin 0.15 M Tris-buffered saline (TBS) with 1% Triton-X for3 � 15 min and incubated in 5% goat serum for 30 min.Sections were then incubated with anti-caspase-3 active(1:5000; R&D Systems, Abingdon, UK) at 4°C for 48 hfollowed by 3 � 15 min rinse in TBS with 1% Triton-X. Toavoid unspecific binding of the secondary antibody, sectionswere again incubated in 5% goat-serum followed by incu-bation with anti-rabbit/CY3 (1:100) and Hoechst 33342 (2�g/ml) for 2 h. After a final 3 � 15 min rinse in TBS,sections were washed with distilled water air-dried andcover-slipped in Flouromount (BDH). As a positive controlfor the caspase-3 staining we included sections from cul-tures treated with colchicine (1 �M for 48 h) known toinduce apoptosis in the dentate granule cells (Kim et al.,2002; Kristensen et al., 2003).

Digital images of the stained sections where obtained byusing a fluorescence microscope (Olympus, Vanox-T)equipped with a digital camera (Sensys KAF 1400 G2:Photometric, Tucson, AZ).

Electrophysiology

Cultured hippocampal slices treated with DHPG (10 �Mfor 2 h) or untreated age-matched controls were transferredto a submersion-type chamber with the attached membranefor recording and superfused with ASCF (in micromolar

concentrations): 129 NaCl, 3 KCl, 1.25 NaH2PO4 � H2O, 2MgCl2 � 6H2O, 10 D-glucose, 26 NaHCO3, and 2 CaCl2,pH 7.5, heated to 30°C, osmolarity � 300 mOsm) suppliedat a rate of 1.8–2 ml/min. Recording microelectrodes wereprepared from borosilicate glass (WPI Inc; Sarasota, FL) bya Kopf Instruments micropipette puller (Tujunga, CA) (re-sistance ranging from 4 to 6 M�) and filled with solution ofthe following composition (in micromolar concentrations):K� gluconate 142.5, potassium methylsulfate 20, NaCl 8,Hepes 10, EGTA 0.1, MgATP 2, and GTP 0.2. The pH ofthe internal solution was adjusted to 7.2 with KOH and theosmolarity was adjusted to 300 mOsm with H2O and su-crose. After making a high-resistance seal the whole-cellconfiguration was established by rupturing the membraneunder the patch pipette. Recordings were done using aAxopatch 1D preamplifier (Axon Instruments; Foster City,CA) and filtered at 5 kHz. Pipette and cell capacitancetransients were compensated with the appropriate capaci-tance controls on the amplifier and series resistance com-pensation was applied to a level of 80–90%. Cells includedin analysis were voltage-clamped between �53 and �77mV and had action potential amplitudes �70 mV and stableinput resistance for at least 10 min prior to the NMDAapplication.

Statistics

Statistical significance was assessed in GraphPad Instat(GraphPad Software, San Diego, CA), using one-wayANOVA with Bonferonni correction for comparison be-tween groups of interest. In voltage clamp studies Student’st-test was used for comparison. Differences were consideredsignificant at P 0.05.

Results

Neuroprotection induced by selective activation ofgroup-I-mGluRs

To investigate the potential neuroprotective effects ofgroup-I-mGluR activation, cultures were pretreated withDHPG (1, 10, or 100 �M) for 2 h followed by exposure toa toxic 50 �M concentration of NMDA for 30 min. It hasbeen shown earlier that at this concentration NMDA causesdamage to approximately 30% of all neurons in an organo-typic hippocampal culture (Adamchik and Baskys, 2000).After 24 h cultures exposed to NMDA showed an increasedPI uptake in hippocampal subfields (Fig. 1B and M). Whenpretreated with 1 �M DHPG for 2 h prior to the NMDAexposure, cultures displayed no reduction in PI uptake com-pared to cultures exposed to NMDA only (Fig. 1M). Pre-treatment with DHPG at 10 or 100 �M did, however, inducesignificant concentration-dependent reduction in PI uptakein CA1 and CA3 pyramidal cell areas (Fig. 1C–D and M).To verify that reduced PI uptake in DHPG � NMDA-

575M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580

treated cultures was indeed associated with reduced neuro-nal cell death, cryostat sectioned slices was stained withHoechst 33342. This revealed small pycnotic nuclei in CA1pyramidal cells at 24 and 48 h (Fig. 1F; images at 48 h arenot shown) after the NMDA treatment compared to un-treated controls (Fig. 1E). In DHPG � NMDA-treated cul-tures we observed less pycnotic CA1 pyramidal cells (Fig.1G and H) when compared to cultures exposed to NMDAonly (Fig. 1F). Since presence of pycnotic cells followingNMDA treatment could be interpreted as an indicator ofnecrotic cell death, we measured caspase-3 activity in con-trol and NMDA-treated cultures. (Increases in caspase-3activity has been associated with apoptotic cell death; Kimet al., 2002; Kristensen et al., 2003.) We found no increasein caspase-3 activity following 24 or 48 h after NMDA

treatment (Fig. 1J; image of 48 h not shown) compared tountreated controls (Fig. 1I). As a control for the caspase-3staining we included sections from cultures treated withcolchicine (1 �M for 48 h) known to induce apoptosis in thedentate granule cells. In Hoechst staining dentate granulecells from these cultures had fragmented nuclei (Fig. 1K)with increased caspase-3 activity (Fig. 1L), suggesting thatthe staining protocol was appropriate and that the NMDAtreatment did not induce caspase-3 activity.

To investigate whether the timing of agonist applicationwas crucial for neuroprotection we also performed experi-ments with 2 h DHPG treatment immediately after theNMDA exposure. Using this protocol, DHPG did, however,not induce neuroprotection (DHPG 10 �M � 91.1 �28.1%; DHPG 100 �M � 85.6 � 10.0%).

Fig. 1. DHPG-induced neuroprotection against NMDA excitotoxicity. (A–D) PI uptake after 24 h in organotypic hippocampal slice cultures exposed toNMDA (50 �M for 30 min) compared to slices pretreated with DHPG before the NMDA exposure. (E–H) High magnification (100�) of CA1 pyramidalcells after similar treatment. I and J correspond to E and F but in caspase-3 staining. As a positive control for the caspase-3 staining, sections from culturestreated with colchicine (1 �M for 48 h) were included (K and L). Untreated age-matched control cultures displayed a small baseline level of PI uptake (A),normal nuclear morphology (E), and no caspase-3 activity (I). Cultures treated with NMDA had increased PI uptake in hippocampal subfields (B) with nucleiappearing pycnotic and densely stained (F; arrow) but without caspase-3 activity (J). In contrast nuclei in cultures treated with colchicine showedfragmentation (K; arrowhead) and increased caspase-3 activation (L; arrowhead). Pretreatment with DHPG for 2 h before the NMDA exposure reduced theNMDA-induced PI uptake (C and D) as well as the number of pycnotic nuclei (G and H). Graph (M) shows quantification of PI uptake in CA1 (blackcolumns) and CA3 subfields (gray columns) after 24 h with NMDA set to 100%. Pretreatment for 2 h with DHPG at 1 �M had no effect on NMDA-inducedPI uptake, whereas pretreatment with DHPG at 10 and 100 �M significantly reduced PI uptake in both CA1 and CA3 in a concentration dependent manner.Data are shown as mean � SEM, ***P 0.001 (compared to NMDA in CA1 alone). ##P 0.01, and ###P 0.001 (compared to NMDA in CA3 alone)with n � 21–42 cultures per group.

576 M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580

Effects of group-I-mGluR antagonists on DHPG-inducedneuroprotection

We examined the contribution of mGluR1 and mGluR5receptors to the DHPG-induced neuroprotection by applica-tion of selective antagonists during the 2 h DHPG pretreat-ment. An mGluR1 selective antagonist LY367385 (5 and 10�M) abolished neuroprotection in a concentration-depen-dent manner (Fig. 2A), whereas the mGluR5-selective an-tagonist MPEP (1 �M) had no significant effect (Fig. 2B).

Effects of DHPG on NMDA receptor/channel function

To examine whether DHPG pretreatment had any effecton the NMDA receptor/channel function we performedwhole-cell recordings of the NMDA-stimulated current inCA1 pyramidal cells in organotypic slice cultures treatedwith DHPG (10 �M, 2 h) and untreated control cultures. Tomeasure the NMDA-stimulated current, neurons were volt-age-clamped (Vhold of control neurons � �73, �72, �56,�53 mV; DHPG-treated neurons � �77, �76, �66 mV)and the holding current (Ihold) values were recorded eachminute after application of NMDA (1 �M, 5 min) for 10min. The average Ihold prior to NMDA application was 115� 42 pA (210, 160, 60, 30 pA) and 113 � 49 pA (210, 50,80 pA) in the control and DHPG-treated cultures respec-tively. The average membrane input resistance (RN) prior toNMDA application was 872 � 322 and 904 � 335 M� inthe control and DHPG-treated groups, which was similar toRN values reported in cultured hippocampal cells (Mynlieff,1999). In the control but not the DHPG-treated cultures,NMDA produced a robust inward current recorded as anincrease in the negative Ihold (Fig. 3), which was associatedwith a decreased RN(490 � 107 M�, not significant, P �

0.05) in all four neurons. The mean values of the NMDA-stimulated inward current were significantly (P 0.05,two-sided t test) different between the two groups at 5, 6,and 7 min after the start of NMDA application (Fig. 3).Thus, the peak value of the NMDA-stimulated current inneurons from the control cultures was 42.5 � 2.5 pA (n �4 neurons from four cultures) at 6 min after the onset ofNMDA application (Fig. 3). In contrast, in the group ofcultures treated with DHPG (10 �M, 2 h) NMDA-stimu-lated pA (n � 3 neurons from three cultures).

Fig. 2. Effect of group I antagonist on DHPG-induced neuroprotection. Graphs showing quantification of PI uptake in cultures coexposed to the mGluR1antagonist LY367385 (A) or the mGluR5 antagonist MPEP (B) during the 2-h pretreatment with DHPG 10 �M. Antagonists were applied 30 min before andduring DHPG treatment to ensure receptor saturation. Blocking the mGluR1 receptor with 10 �M LY367385 during the DHPG pretreatment attenuatedneuroprotection (A), whereas blocking the mGluR5 receptor with 1 �M MPEP was ineffective (B). Data are shown as mean � SEM, *P 0.05, **P 0.01 with n � 6–12.

Fig. 3. NMDA stimulated inward current in CA1 pyramidal cells fromhippocampal slice cultures. In control cultures (solid circles), NMDAinduced a clear inward current measured as holding current (Ihold). Thiscurrent was strongly suppressed in cultures treated with DHPG (10 �M,2 h, open circles). The difference between the DHPG-treated and controlcultures reached significance after 5, 6, and 7 min. Data are shown as mean� SEM, *P 0.05, **P 0.01 using Student’s t-test with n � 3–4neurons from individual cultures per group.

577M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580

Discussion

The results of the present study show that selectiveactivation of group-I-mGluRs protect neurons against a sub-sequent excitotoxic stimulation of NMDA receptors in aconcentration-dependent manner in hippocampal CA1 andCA3 subfields. This finding is consistent with previousresults in organotypic hippocampal slice cultures, where asimilar ACPD pretreatment induced neuroprotection (Ad-amchik and Baskys, 2000), confirming the important role ofthese receptors. DHPG-induced neuroprotection could bereversed by inhibition of the mGluR1 receptor withLY367385, but not by the mGluR5 receptor antagonistMPEP. Earlier studies with LY367385 show a lack of adiscernible effect on NMDA receptors, which, taken to-gether, suggest that the neuroprotective effects of DHPGwere likely due to inhibition of group-I-mGluRs.

Interestingly, the neuroprotective pretreatment withDHPG also significantly decreased the NMDA-stimulatedinward currents in voltage-clamped CA1 pyramidal cells,suggesting that activation of group-I-mGluRs could lead toinhibition of NMDA-receptor function and although thenature of the current recorded in the voltage-clamp experi-ments was not investigated in this study, it most likelyrepresents the NMDA receptor/channel mediated inwardcurrent (Collingridge and Lester 1989). Several possiblemechanisms could underlie this inhibition. Moderate in-creases in [Ca2�] cause inactivation of NMDA receptors(Krupp et al., 1998; Tong et al., 1995) and activation ofgroup-I-mGluRs may bring about such an increase in[Ca2�] (for review see Baskys, 1994). Other possible mech-anisms may involve modifications of NMDA receptors thatsignificantly reduce agonist binding (Bi et al., 1998),changes in NMDA receptor pentameric structure followingmGluR stimulation (Nicoletti et al., 1999), or downregula-tion of NMDA receptors (Yu et al., 1997). Consistent withthis, it has recently been shown that DHPG treatment resultsin internalization of NMDA and AMPA receptors in pri-mary hippocampal neurons (Snyder et al., 2001), whichcould explain neuroprotection. That the mechanism bywhich DHPG induce neuroprotection is due to an increasein [Ca2�]i is supported by electrophysiological evidenceshowing that LY367385 blocks the DHPG-induced increasein [Ca2�]i and the direct depolarization in CA1 pyramidalcells, whereas MPEP blocks Ca2� activated K� currentsand potentiation of NMDA receptors (Mannaioni et al.,2001). Taken together these data are consistent with theresults of the present study and show that DHPG-mediatedneuroprotection against NMDA toxicity occurs via themGluR1 receptor and that neuroprotective DHPG treatmentcan greatly reduce NMDA-stimulated currents. A similarobservation has been made in cerebellar granule cells, inwhich treatment with trans-ACPD or (R,S)-DHPG signifi-cantly reduces Ca2� influx induced by application ofNMDA or glutamate (Pizzi et al., 1996b) and in mousecortical cultures where DHPG reduced NMDA whole-cell

currents (Yu et al., 1997). Moreover, it has been suggestedthat functional activity and the number of functional NMDAreceptors are regulated by the strength of the glutamatergicinput (Cebers et al., 2001). Group-I-mGluRs have beenshown to increase glutamate release (Rodriguez-Moreno etal., 1998; Ye and Sontheimer, 1999) from both neurons andastrocytes, and as a result, increased concentration of am-bient glutamate may decrease NMDA receptor-mediatedeffects.

It should be noted that neuroprotection could only befound if the activation of group-I-mGluRs was prior to theNMDA exposure and no neuroprotection was observed ifDHPG was applied for 2 h immediately after the insult. Theconflicting results on the role of glutamate in neuroprotec-tion and neurodegeneration, discussed in the introduction,may be due to the timing of agonist application. Coappli-cation of a group-I-mGluR agonist and NMDA may poten-tiate NMDA responses through phosphorylation of theNMDA receptors relieving the Mg2� block and increasingCa2� influx through the associated channel to a toxic level(Bruno et al., 1995). Pretreatment with the group-I-mGluRagonist, however, would increase [Ca2�]i causing inhibitionof NMDA receptors or resulting in their internalization, sothat a subsequent exposure to NMDA would yield a de-creased response. This idea is supported by our earlierfinding that coapplication of NMDA with ACPD potenti-ated NMDA toxicity and that this potentiation was reversedby a 30-min pretreatment with ACPD before the coapplica-tion (Blaabjerg et al., 2001). Another mechanism ofmGluR-induced neuroprotection could be mGluR receptordesensitization. Due to a high intracellular concentration ofglutamate, initial disintegration of neurons after the insultwould result in a massive glutamate release that wouldactivate extrasynaptic mGluRs, thus increasing Ca2� influxand cell death. mGluRs, however, undergo rapid PKC-mediated receptor desensitization upon agonist binding(Herrero et al., 1994; Gereau and Heinemann, 1998). It isnot unlikely that, if mGluRs are activated prior to exposureto NMDA, the receptors may be desensitized and, therefore,respond weakly to the glutamate released by a subsequenttoxic NMDA exposure, resulting in neuroprotection.

Several studies have shown that activation of group-I-mGluR appears to reduce apoptotic cell death (Allen et al.,2000; Kalda and Zharkovsky, 1999; Vincent and Maiese,2000; Vincent et al., 1997) while potentiating necrotic celldeath (Allen et al., 2000). Interestingly we observed pyc-notic densely stained nuclei, with no caspase-3 activation,after NMDA treatment, corresponding to necrotic celldeath. One study on mouse hippocampal slice cultures re-port NMDA-induced apoptosis independent of caspase-3activation by measuring the intrahistonic DNA fragmenta-tion (Djebaili et al., 2002). We cannot rule out that the samemechanism could be present in our model; however, wefound no nuclear fragmentation 24 or 48 h after the NMDAexposure.

578 M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580

In summary, this study strongly supports the role ofmGluR1 in neuroprotection against NMDA-induced celldeath. It also points to changes in NMDA receptor/channelfunction as a putative mechanism of such protection.

Acknowledgments

This study was supported by the Codan Foundation, theFuhrmann Foundation (M.B.), the FP5 EU Grant QLK3-CT-2001-00407 (J.Z.), and Mental Illness Research andEducation Clinical Center (MIRECC) (A.B.). The authorsacknowledge the editorial services of Jennifer Kahle fromBPS International as well as technical assistance from DorteBramsen, University of Southern Denmark, Odense.

References

Adamchik, Y., Baskys, A., 2000. Glutamate-mediated neuroprotectionagainst N-methyl-D-aspartate toxicity: a role for metabotropic gluta-mate receptors. Neuroscience 99, 731–736.

Allen, J.W., Knoblach, S.M., Faden, A.I., 2000. Activation of group Imetabotropic glutamate receptors reduces neuronal apoptosis but in-creases necrotic cell death in vitro. Cell Death Differ. 7, 470–476.

Anneser, J.M., Horstmann, S., Weydt, P., Borasio, G.D., 1998. Activationof metabotropic glutamate receptors delays apoptosis of chick embry-onic motor neurons in vitro. NeuroReport 9, 2039–2043.

Awad, H., Hubert, G.W., Smith, Y., Levey, A.I., Conn, P.J., 2000. Acti-vation of metabotropic glutamate receptor 5 has direct excitatory ef-fects and potentiates NMDA receptor currents in neurons of the sub-thalamic nucleus. J. Neurosci. 20, 7871–7879.

Bandrowski, A.E., Aramakis, V.B., Moore, S.L., Ashe, J.H., 2001.Metabotropic glutamate receptors modify ionotropic glutamate re-sponses in neocortical pyramidal cells and interneurons. Exp. BrainRes. 136, 25–40.

Baskys, A., 1994. Metabotropic Glutamate Receptors. R.G. Landes Press,Austin, TX.

Baskys, A., Adamchik, Y., 2001. Neuroprotective effects of extracellularglutamate are absent in hippocampal organotypic cultures treated withthe amyloid peptide Abeta(25–35). Brain Res. 907, 188–194.

Bi, X., Rong, Y., Chen, J., Dang, S., Wang, Z., Baudry, M., 1998. Calpain-mediated regulation of NMDA receptor structure and function. BrainRes. 790, 245–253.

Blaabjerg, M., Kristensen, B.W., Bonde, C., Zimmer, J., 2001. Themetabotropic glutamate receptor agonist 1S,3R-ACPD stimulates andmodulates: NMDA receptor mediated excitotoxicity in organotypichippocampal slice cultures. Brain Res. 898, 91–104.

Bruno, V., Battaglia, G., Copani, A., Cespedes, V.M., Galindo, M.F., Cena,V., Sanchez-Prieto, J., Gasparini, F., Kuhn, R., Flor, P.J., Nicoletti, F.,2001. An activity-dependent switch from facilitation to inhibition in thecontrol of excitotoxicity by group I metabotropic glutamate receptors.Eur. J. Neurosci. 13, 1469–1478.

Bruno, V., Copani, A., Knopfel, T., Kuhn, R., Casabona, G., Dell’Albani,P., Condorelli, D.F., Nicoletti, F., 1995. Activation of metabotropicglutamate receptors coupled to inositol phospholipid hydrolysis ampli-fies NMDA-induced neuronal degeneration in cultured cortical cells.Neuropharmacology 34, 1089–1098.

Cebers, G., Cebere, A., Kovacs, A.D., Hogberg, H., Moreira, T., Liljequist,S., 2001. Increased ambient glutamate concentration alters the expres-sion of NMDA receptor subunits in cerebellar granule neurons. Neu-rochem. Int. 39, 151–160.

Chiamulera, C., Albertini, P., Valerio, E., Reggiani, A., 1992. Activation ofmetabotropic receptors has a neuroprotective effect in a rodent modelof focal ischaemia. Eur. J. Pharmacol. 216, 335–336.

Collingridge, G.L., Lester, R.A., 1989. Excitatory amino acid receptors inthe vertebrate nervous system. Pharmacol. Rev. 41, 143–210.

Colwell, C.S., Altemus, K.L., Levine, M.S., 1996. Metabotropic glutamatereceptor activation selectively limits excitotoxic damage in the intactneostriatum. Brain Res. 726, 223–226.

Colwell, C.S., Levine, M.S., 1999. Metabotropic glutamate receptor mod-ulation of excitotoxicity in the neostriatum: role of calcium channels.Brain Res. 833, 234–241.

Conn, P.J., Pin, J.P., 1997. Pharmacology and functions of metabotropicglutamate receptors. Annu. Rev. Pharmacol. Toxicol. 37, 205–237.

Faden, A.I., 1997. Therapeutic approaches to spinal cord injury. Adv.Neurol. 72, 377–386.

Fitzjohn, S.M., Irving, A.J., Palmer, M.J., Harvey, J., Lodge, D., Col-lingridge, G.L., 1996. Activation of group-I-mGluRs potentiatesNMDA responses in rat hippocampal slices. Neurosci. Lett. 203, 211–213.

Gereau, R.W., Heinemann, S.F., 1998. Role of protein kinase C phosphor-ylation in rapid desensitization of metabotropic glutamate receptor 5.Neuron 20, 143–151.

Herrero, I., Miras-Portugal, M.T., Sanchez-Prieto, J., 1994. Rapid desen-sitization of the metabotropic glutamate receptor that facilitates gluta-mate release in rat cerebrocortical nerve terminals. Eur. J. Neurosci. 6,115–120.

Kalda, A., Kaasik, A., Vassiljev, V., Pokk, P., Zharkovsky, A., 2000.Neuroprotective action of group I metabotropic glutamate receptoragonists against oxygen-glucose deprivation-induced neuronal death.Brain Res. 853, 370–373.

Kalda, A., Zharkovsky, A., 1999. Metabotropic glutamate receptor agonistsprotect from oxygen-glucose deprivation- and colchicine-induced ap-optosis in primary cultures of cerebellar granule cells. Neuroscience 92,7–14.

Kim, J.A., Mitsukawa, K., Yamada, M.K., Nishiyama, N., Matsuki, N.,Ikegaya, Y., 2002. Cytoskeleton disruption causes apoptotic degener-ation of dentate granule cells in hippocampal slice cultures. Neurophar-macology 42, 1109–1118.

Koga, T., Sakai, N., Tanaka, C., Saito, N., 1996. Presynaptic and Ca(2�)-independent PKC subspecies modulates NMDAR1 current. NeuroRe-port 7, 477–480.

Koh, J.Y., Palmer, E., Cotman, C.W., 1991. Activation of the metabotropicglutamate receptor attenuates N-methyl-D-aspartate neurotoxicity incortical cultures. Proc. Natl. Acad. Sci. USA 88, 9431–9435.

Kristensen, B.W., Noer, H., Gramsbergen, J.B., Zimmer, J., Noraberg, J.,2003. Colchicine induces apoptosis in organotypic hippocampal slicecultures. Brain Res. 964, 269–278.

Krupp, J.J., Vissel, B., Heinemann, S.F., Westbrook, G.L., 1998. N-termi-nal domains in the NR2 subunit control desensitization of NMDAreceptors. Neuron 20, 317–327.

Maiese, K., Greenberg, R., Boccone, L., Swiriduk, M., 1995. Activation ofthe metabotropic glutamate receptor is neuroprotective during nitricoxide toxicity in primary hippocampal neurons of rats. Neurosci. Lett.194, 173–176.

Maiese, K., Swiriduk, M., TenBroeke, M., 1996. Cellular mechanisms ofprotection by metabotropic glutamate receptors during anoxia andnitric oxide toxicity. J. Neurochem. 66, 2419–2428.

Maiese, K., Vincent, A., Lin, S.H., Shaw, T., 2000. Group I and group IIImetabotropic glutamate receptor subtypes provide enhanced neuropro-tection. J. Neurosci. Res. 62, 257–272.

Maiese, K., Vincent, A.M., 1999. Group I metabotropic receptors down-regulate nitric oxide induced caspase-3 activity in rat hippocampalneurons. Neurosci. Lett. 264, 17–20.

Mannaioni, G., Marino, M.J., Valenti, O., Traynelis, S.F., Conn, P.J., 2001.Metabotropic glutamate receptor 1 and 5 differentially regulate CA1pyramidal cell function. J. Neurosci. 21, 5925–5934.

579M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580

Mount, H.T., Dreyfus, C.F., Black, I.B., 1993. Purkinje cell survival isdifferentially regulated by metabotropic and ionotropic excitatoryamino acid receptors. J. Neurosci. 13, 3173–3179.

Movsesyan, V.A., O’Leary, D.M., Fan, L., Bao, W., Mullins, P.G., Kno-blach, S.M., Faden, A.I., 2001. mGluR5 antagonists 2-methyl-6-(phe-nylethynyl)-pyridine and (E)-2-methyl-6-(2-phenylethenyl)-pyridinereduce traumatic neuronal injury in vitro and in vivo by antagonizingN-methyl-D-aspartate receptors. J. Pharmacol. Exp. Ther. 296, 41–47.

Mynlieff, M., 1999. Identification of different putative neuronal subtypesin cultures of the superior region of the hippocampus using electro-physiological parameters. Neuroscience 93, 479–486.

Nicoletti, F., Bruno, V., Catania, M.V., Battaglia, G., Copani, A., Bar-bagallo, G., Cena, V., Sanchez-Prieto, J., Spano, P.F., Pizzi, M., 1999.Group-I metabotropic glutamate receptors: hypotheses to explain theirdual role in neurotoxicity and neuroprotection. Neuropharmacology 38,1477–1484.

O’Leary, D.M., Movsesyan, V., Vicini, S., Faden, A.I., 2000. SelectivemGluR5 antagonists MPEP and SIB-1893 decrease NMDA or gluta-mate-mediated neuronal toxicity through actions that reflect NMDAreceptor antagonism. Br. J. Pharmacol. 131, 1429–1437.

Opitz, T., Reymann, K.G., 1993. (1S, 3R)-ACPD protects synaptic trans-mission from hypoxia in hippocampal slices. Neuropharmacology 32,103–104.

Orlando, L.R., Alsdorf, S.A., Penney Jr., J.B., Young, A.B., 2001. The roleof group I and group II metabotropic glutamate receptors in modulationof striatal NMDA and quinolinic acid toxicity. Exp. Neurol. 167,196–204.

Pellegrini-Giampietro, D.E., Cozzi, A., Peruginelli, F., Leonardi, P., Meli,E., Pellicciari, R., Moroni, F., 1999. 1-Aminoindan-1,5-dicarboxylicacid and (S)-(�)-2-(3�-carboxybicyclo[1.1.1] pentyl)-glycine, twomGlu1 receptor-preferring antagonists, reduce neuronal death in invitro and in vivo models of cerebral ischaemia. Eur. J. Neurosci. 11,3637–3647.

Pisani, A., Gubellini, P., Bonsi, P., Conquet, F., Picconi, B., Centonze, D.,Bernardi, G., Calabresi, P., 2001. Metabotropic glutamate receptor 5mediates the potentiation of N-methyl-D-aspartate responses in mediumspiny striatal neurons. Neuroscience 106, 579–587.

Pizzi, M., Consolandi, O., Memo, M., Spano, P.F., 1996a. Activation ofmultiple metabotropic glutamate receptor subtypes prevents NMDA-induced excitotoxicity in rat hippocampal slices. Eur. J. Neurosci. 8,1516–1521.

Pizzi, M., Galli, P., Consolandi, O., Arrighi, V., Memo, M., Spano, P.F.,1996b. Metabotropic and ionotropic transducers of glutamate signalinversely control cytoplasmic Ca2� concentration and excitotoxicity in

cultured cerebellar granule cells: pivotal role of protein kinase C. Mol.Pharmacol. 49, 586–594.

Rodriguez-Moreno, A., Sistiaga, A., Lerma, J., Sanchez-Prieto, J., 1998.Switch from facilitation to inhibition of excitatory synaptic transmis-sion by group-I-mGluR desensitization. Neuron 21, 1477–1486.

Sagara, Y., Schubert, D., 1998. The activation of metabotropic glutamatereceptors protects nerve cells from oxidative stress. J. Neurosci. 18,6662–6671.

Schroder, U.H., Opitz, T., Jager, T., Sabelhaus, C.F., Breder, J., Reymann,K.G., 1999. Protective effect of group I metabotropic glutamate recep-tor activation against hypoxic/hypoglycemic injury in rat hippocampalslices: timing and involvement of protein kinase C. Neuropharmacol-ogy 38, 209–216.

Siliprandi, R., Lipartiti, M., Fadda, E., Sautter, J., Manev, H., 1992.Activation of the glutamate metabotropic receptor protects retinaagainst N-methyl-D-aspartate toxicity. Eur. J. Pharmacol. 219, 173–174.

Skeberdis, V.A., Lan, J., Opitz, T., Zheng, X., Bennett, M.V., Zukin, R.S.,2001. mGluR1-mediated potentiation of NMDA receptors involves arise in intracellular calcium and activation of protein kinase C. Neuro-pharmacology 40, 856–865.

Snyder, E.M., Philpot, B.D., Huber, K.M., Dong, X., Fallon, J.R., Bear,M.F., 2001. Internalization of ionotropic glutamate receptors in re-sponse to mGluR activation. Nat. Neurosci. 4, 1079–1085.

Stoppini, L., Buchs, P.A., Muller, D., 1991. A simple method for organo-typic cultures of nervous tissue. J. Neurosci. Methods 37, 173–182.

Tong, G., Shepherd, D., Jahr, C.E., 1995. Synaptic desensitization ofNMDA receptors by calcineurin. Science 267, 1510–1512.

Vincent, A.M., Maiese, K., 2000. The metabotropic glutamate systempromotes neuronal survival through distinct pathways of programmedcell death. Exp. Neurol. 166, 65–82.

Vincent, A.M., Mohammad, Y., Ahmad, I., Greenberg, R., Maiese, K.,1997. Metabotropic glutamate receptors prevent nitric oxide-inducedprogrammed cell death. J. Neurosci. Res. 50, 549–564.

Ye, Z.C., Sontheimer, H., 1999. Metabotropic glutamate receptor agonistsreduce glutamate release from cultured astrocytes. Glia 25, 270–281.

Young, M.R., Blackburn-Munro, G., Dickinson, T., Johnson, M.J., Ander-son, H., Nakalembe, I., Fleetwood-Walker, S.M., 1998. Antisense ab-lation of type I metabotropic glutamate receptor mGluR1 inhibits spinalnociceptive transmission. J. Neurosci. 18, 10180–10188.

Yu, S.P., Sensi, S.L., Canzoniero, L.M., Buisson, A., Choi, D.W., 1997a.Membrane-delimited modulation of NMDA currents by metabotropicglutamate receptor subtypes 1/5 in cultured mouse cortical neurons.J. Physiol. 499 (3), 721–732.

580 M. Blaabjerg et al. / Experimental Neurology 183 (2003) 573–580