an activity-dependent switch from facilitation to inhibition in the control of excitotoxicity by...

An activity-dependent switch from facilitation to inhibitionin the control of excitotoxicity by group I metabotropicglutamate receptors

Valeria Bruno,1 Giuseppe Battaglia,1 Agata Copani,2 Virtudes M. CespeÂdes,3 MarõÂa F. Galindo,3 ValentõÂn CenÄa,3

Jose SaÂnchez-Prieto,4 Fabrizio Gasparini,5 Rainer Kuhn,5 Peter J. Flor5 and Ferdinando Nicoletti1,2

1I.N.M. Neuromed, LocalitaÁ Camerelle, 86077 Pozzilli, Italy2Department of Human Physiology and Pharmacology, University of Roma `La Sapienza', Italy3Instituto de Neurociencias and Departamento de FarmacologõÂa, Universidad Miguel HernaÂndez, Alicante, Spain4Departamento de BioquõÂmica, Facultad de Veterinaria, Universitad Complutense, Madrid, Spain5Novartis Pharma AG, Nervous System Research, CH-4002 Basel, Switzerland

Keywords: activity-dependent switch, cortical cultures, excitotoxicity, group I metabotropic glutamate receptors, hippocampalcultures, NMDA

Abstract

Activation of group I metabotropic glutamate receptors (mGlu1 or -5 receptors) is known to either enhance or attenuate excitotoxic

neuronal death depending on the experimental conditions. We have examined the possibility that these receptors may switchbetween two different functional modes in regulating excitotoxicity. In mixed cultures of cortical cells, the selective mGlu1/5

agonist, 3,5-dihydroxyphenylglycine (DHPG), ampli®ed neurodegeneration induced by a toxic pulse of NMDA. This effect was

observed when DHPG was either combined with NMDA or transiently applied to the cultures prior to the NMDA pulse. However,two consecutive applications of DHPG consistently produced neuroprotection. Similar effects were observed with DHPG or

quisqualate (a potent agonist of mGlu1/5 receptors) in pure cultures of cortical neurons virtually devoid of astrocytes. In cultures of

hippocampal pyramidal neurons, however, only protective effects of DHPG were seen suggesting that, in these particular

cultures, group I mGlu receptors were endogenously switched into a `neuroprotective mode'. The characteristics of the activity-dependent switch from facilitation to inhibition were examined in mixed cultures of cortical cells. The switch in the response to

DHPG was observed when the two applications of the drug were separated by an interval ranging from 1±45 min, but was lost

when the interval was extended to 90 min. In addition, this phenomenon required the initial activation of mGlu5 receptors (asindicated by the use of subtype-selective antagonists) and was mediated by the activation of protein kinase C. We conclude that

group I mGlu receptors are subjected to an activity-dependent switch in regulating excitotoxic neuronal death and, therefore, the

recent `history' of these receptors is critical for the response to agonists or antagonists.

Introduction

Metabotropic glutamate (mGlu) receptors are considered promising

`targets' for neuroprotective drugs (Schoepp & Conn, 1993; Nicoletti

et al., 1996). These receptors form a family of eight subtypes, which

have been subdivided into three groups. Group I includes mGlu1

and -5 receptors, which are coupled to polyphosphoinositide (PI)

hydrolysis; group II includes mGlu2 and -3, whereas, group III

includes mGlu4, and mGlu6±8 receptors. All group II and III subtypes

are negatively coupled to adenylyl cyclase activity when hetero-

logously expressed in non-neuronal cells (Nakanishi, 1994; Pin &

Duvoisin, 1995; Conn & Pin, 1997). While activation of group II and

-III mGlu receptors has been proven to be neuroprotective (Bruno

et al., 1995a, 1997, 1998; Buisson & Choi, 1995; Gasparini et al.,

1999a; but see also Behrens et al., 1999), the role of group I mGlu

receptors in neurodegeneration/neuroprotection is controversial.

Activation of mGlu1 and -5 receptors leads to mobilization of

intracellular Ca2+ and activation of protein kinase C (PKC), as a

result of inositol-1,4,5-trisphosphate and diacylglycerol formation,

respectively. As increases in free cytosolic Ca2+ and activation of

PKC may contribute to the development of neuronal death (reviewed

by Choi, 1992), one expects that pharmacological activation of

mGlu1 and -5 receptors exacerbates neuronal toxicity. However,

group I mGlu receptor agonists either amplify or attenuate neuronal

degeneration in different experimental models (reviewed by Nicoletti

et al., 1999). In mixed murine cortical cultures, for example, agonists

like 3-hydroxyphenylglycine, 3,5-dihydroxyphenylglycine (DHPG)

or quisqualate enhance excitotoxic neuronal damage (Bruno et al.,

1995b; Buisson & Choi, 1995), and at least the action of DHPG and

quisqualate is sensitive to PKC inhibitors (Bruno et al., 1995b). In

contrast, group I mGlu receptor agonists protect cultured cerebellar

granule cells against excitotoxic damage (Pizzi et al., 1993, 1996a),

and are neuroprotective in brain slices challenged with excitotoxins

or oxygen±glucose deprivation (Opitz & Reymann, 1993; Pizzi et al.,

1996b; Schroder et al., 1999). The origin of these contrasting results

is unclear and may be related to factors inherent to the cell

environment (selection of a particular neuronal type, absence or

presence of glial cells, heteromeric assembly of NMDA receptor

Correspondence: Dr Ferdinando Nicoletti, at 1I.N.M. Neuromed, as above,E-mail: [email protected]

Received 19 December 2000, accepted 1 February 2001

European Journal of Neuroscience, Vol. 13, pp. 1469±1478, 2001 ã Federation of European Neuroscience Societies

subunits) or may depend on factors that are intrinsic to group I mGlu

receptors (Nicoletti et al., 1999; Pizzi et al., 1999). We have

examined the latter possibility moving from the description of an

experience-dependent switch in the modulation of glutamate release

by group I mGlu receptors. In cortical or hippocampal synaptosomal

preparations, a ®rst application of group I mGlu receptor agonists

facilitates glutamate release via PKC activation (Herrero et al., 1992,

1994). However, when agonists are applied for a second time, shortly

after a previous application, they inhibit, rather than facilitate,

glutamate release (Herrero et al., 1998) through a membrane-

delimited mechanism insensitive to PKC inhibitors. A similar form of

experience-dependent switch has been described for the modulation

of excitatory synaptic transmission by group I mGlu receptors in

hippocampal slice preparations (Rodriguez-Moreno et al., 1998). We

now report that pharmacological activation of group I mGlu receptors

may either facilitate or attenuate excitotoxic neuronal death depend-

ing on the functional state of the receptors. This in turn depends on

how long before, and to what extent, group I mGlu receptors have

been previously activated.

Materials and methods

Materials

N-Methyl-D-aspartate (NMDA) (S)-3,5-dihydroxyphenylglycine

(DHPG), 7-(hydroxyimino)cyclopropa[b]chromen-1a-carboxylate

ethyl ester (CPCCOEt), GYKI 52466 and quisqualate were purchased

from Tocris Cookson (Bristol, U.K.). Calphostin C and (6)-1-(5-

isoquinolinesulphonyl)-2-methylpiperazine (H7) were obtained from

Calbiochem-Novabiochem (San Diego, CA, USA). 6-(2-Phenyl-1-

ethynyl)-pyridine (MPEP) was synthesized by Novartis Pharma AG

and kindly provided by Dr P. J. Flor (Novartis Pharma AG, Basel,

Switzerland). (E)-2-methyl-6-(2-phenylethenyl)piridine (SIB-1893)

and 6-methyl-2-(phenylazo)-3-pyridinol (SIB-1757) were also kindly

provided by Dr P. J. Flor (Novartis Pharma AG, Basel, Switzerland).

All other drugs were purchased from Sigma (Milano, Italy); pertussis

toxin (PTX) was purchased from Alexis Biochemicals (Lausen,

Switzerland).

Mixed cortical cultures

Mixed cortical cell cultures, containing both neurons and astrocytes,

were prepared from fetal mice at 14±16 day (E14±16) of gestation, as

described previously (Rose et al., 1992). Brie¯y, mice were killed by

cervical dislocation under chloroform anaesthesia, and dissociated

cortical cells were plated in 15 mm multiwell vessels (Falcon

Primaria, Lincoln Park, NJ, USA) on a layer of con¯uent astrocytes

[7±14 day in vitro (DIV)], using a plating medium of MEM-Eagle's

(supplied glutamine free) supplemented with 5% heat-inactivated

horse serum, 5% fetal bovine serum, glutamine (2 mM), and glucose

(®nal concentration 21 mM). Cultures were kept at 37 °C in a

humidi®ed 5% CO2 atmosphere. After 3±5 DIV, non-neuronal cell

division was halted by 1±3 day exposure to 10 mM cytosine

arabinoside, and cultures were shifted to a maintenance medium

identical to plating medium but lacking fetal serum. Subsequent

partial medium replacement was carried out twice a week. Only

mature cultures (13±14 DIV) were used for the experiments.

Glial cell cultures

Glial cell cultures were prepared as described previously (Rose et al.,

1992) from postnatal mice [1±3 day after birth, (P1±3)]. Dissociated

cortical cells were grown in 15 mm multiwell vessels (Falcon

Primaria) using a plating medium of MEM-Eagle's salts supple-

mented with 10% fetal bovine serum, 10% horse serum, glutamine

(2 mM), and glucose (®nal concentration 21 mM). Cultures were kept

at 37 °C in a humidi®ed CO2 atmosphere until they reached

con¯uence (13±14 DIV). Con¯uent cultures were then used as a

support for mixed cultures.

Pure cultures of cortical neurons

Cultures of pure cortical neurons were obtained from E15 rat

embryos. Rats were killed by decapitation and cortices were dissected

in a Ca2+/Mg2+-free buffer; pieces were collected by slow speed

centrifugation, and cells were mechanically dissociated in a plating

medium consisting of DMEM/Ham's F12 (1 : 1) supplemented

with the following components: 10 mg/mL bovine serum albumin,

10 mg/mL insulin, 100 mg/mL transferrin, 100 mM putrescine, 20 nM

progesterone, 30 nM selenium, 2 mM glutamine, 6 mg/mL glucose,

50 units/mL penicillin and 50 mg/mL streptomycin. Cortical cells

were plated at a density of 2 3 106 per dish on 35 mm Nunc dishes

precoated with 0.1 mg/mL poly D-lysine. Cytosine arabinoside

(10 mM) was added to the cultures 18 h after plating to avoid the

proliferation of non-neuronal elements, and was kept for three days

before medium replacements. Subsequent partial medium replace-

ment was carried out every two days. This method yields more than

99% pure neuronal cultures, as judged by immunocytochemistry for

glial ®brillary acidic protein (GFAP) and neuron speci®c micro-

tubule-associated protein 2 (MAP-2; Copani et al., 1999). Cultures

were used at 10±12 DIV, when > 90% of cells were viable, as

demonstrated by ¯uorescent staining with ¯uorescein diacetate and

propidium iodide (not shown).

Hippocampal cultures

Pyramidal neurons were prepared from the hippocampi of fetal rats at

E17 as described by Prehn & Miller (1996). Brie¯y, rats were killed

by decapitation and hippocampi were dissected in Ca2+/Mg2+-free

Hanks balanced salt solution (HBSS) and incubated in 0.1% trypsin

for 15 min. Hippocampi were triturated by aspirating through a

Pasteur pipette and cells were plated in Dulbecco's modi®ed Eagle's

medium (GIBCO BRL, Life Technologies Inc., Grand Island, NY,

USA) supplemented with 10% horse serum (GIBCO BRL) on poly

L-lysine (Sigma, 0.5 mg/mL in borate buffer, pH 8.0) -coated round

glass coverslips, and allowed to adhere for 2±4 h. Coverslips were

then transferred to 60 mm dishes containing supporting astrocytes

attached to the bottom of the culture dishes. Cytosine arabinoside

(5 mM) was added to each plate 2 days later to inhibit non-neuronal

cell proliferation. Cultures were grown for 7±9 DIV before perform-

ing all the experiments. Pyramidal cells were identi®ed by their

typical morphology. Their soma had a width of 12±18 mm and at least

one apical dendrite emerging from it (Shah & Haylett 2000). The

number of GABAergic neurons was also assessed by glutamate

decarboxylase (GAD) immunostaining, as described previously

(Catania et al., 1999).

Exposure to excitatory amino acids

Brief exposure to NMDA (60 mM, 10 min) was carried out in mixed

cortical cultures at room temperature in a HEPES-buffered salt

solution containing (in mM): NaCl, 120; KCl, 5.4; MgCl2, 0.8; CaCl2,

1.8; HEPES, 20 and glucose, 15. After 10 min, the drug was washed

out and cultures were incubated at 37 °C for the following 24 h in

medium stock (MS; MEM-Eagle's supplemented with 15.8 mM

NaHCO3 and glucose < 25 mM). In a series of experiments, mixed

cortical cultures or pure cultures of rat cortical neurons were

pretreated for 1 min with the group I mGluR agonists, DHPG

(100 mM) or quisqualate (50 mM), and then challenged after 5 min

1470 V. Bruno et al.

ã 2001 Federation of European Neuroscience Societies, European Journal of Neuroscience, 13, 1469±1478

with NMDA in the presence or absence of DHPG or quisqualate.

When present, the PKC inhibitors, H7 (100 mM) and calphostin C

(100 nM), were added 10 min before DHPG or quisqualate pretreat-

ment. The group I mGlu receptor antagonists, SIB-1757, SIB-1893,

CPCCOEt and MPEP (all at 30 mM), where present 2 min before and

during the 1 min pretreatment with DHPG, 5 min before the NMDA

pulse. In some experiments, MPEP was also present during the

NMDA pulse.

In experiments performed in hippocampal cultures, coverslips

containing hippocampal neurons were rinsed twice with HEPES-

buffered salt solution, preincubated for 5 min either with HEPES-

buffered salt solution alone or containing the indicated drugs, and

exposed for 10 min to NMDA (30 or 100 mM) at room temperature in

the presence or absence of DHPG (200 mM) and/or the PKC inhibitor,

calphostin C (1 mM). When used alone, DHPG was present 5 min

before and during the NMDA pulse. When used together with DHPG,

calphostin C was present 10 min before NMDA (the last 5 min in the

presence of DHPG). Drug exposure was terminated by washing out

the drugs with HEPES-buffered salt solution, and coverslips were

transferred back to dishes containing astrocytes and kept in the

incubator for the following 24 h. In another set of experiments, cells

were treated using the same protocol as in cortical cultures.

Assessment of neuronal injury

In all experiments performed in mixed and pure cortical cultures,

neuronal injury was assessed by phase-contrast microscopy at 100±

4003, 20±24 h after the insult, when the process of cell death was

largely complete. Neuronal damage was quanti®ed by estimation of

dead neurons by trypan-blue staining. Stained neurons were counted

from three random ®elds per well. The accuracy of cell counting was

validated by measuring the activity of lactate dehydrogenase (LDH)

released into the extracellular medium in pilot experiments.

Neuronal death in hippocampal cultures was determined using

¯uorescein diacetate/propidium iodide double staining procedure

(Prehn & Miller, 1996). Stained cells were examined with a standard

epi-illumination ¯uorescence microscope and counted from ®ve

microscopic ®elds.

Measurement of glutamate release

Analysis of glutamate was performed by precolumn derivatization

with o-phthalaldehyde and mercaptoethanol followed by HPLC with

¯uorescence detection. Culture medium was collected at different

times after the application of NMDA alone or in combination with

DHPG. One-hundred mL of sample was diluted with 0.1 N HCl and

mixed with equal volumes of ¯uorescent reagent. The mixture was

kept at room temperature for 1 min to derivatize the sample before

being injected into the column by a 200-mL loop. The system utilized

an autosampler 507 (Beckman Instruments, Inc., Fullerton, CA), a

programmable solvent module 126 (Beckman Instruments, Inc.,

Fullerton, CA, USA), an analytical reverse phase C-18 column at

30 °C (Ultrasphere ODS 3 mm Spherical, 80 AÊ pore,

2 mm 3 250 mm, Beckman Instruments, Inc., Fullerton, CA,

USA), a RF-551 spectro¯uorimetric detector (Shimadzu, Kyoto,

Japan) and a computer running a Gold Nouveau software (Beckman

Instruments, Inc., Fullerton, CA, USA). The excitation and emission

wavelengths were set at 360 and 450 nm, respectively. The mobile

phase consisted of (A) 50 mM sodium phosphate, pH 7.2, containing

10% methanol and (B) 50 mM sodium phosphate, pH 7.2, containing

70% methanol, at a ¯ow rate of 0.3 mL/min. Both buffers were

®ltered through a 0.45-mm ®lter and degassed under vacuum for

5 min. Gradient elution consisted of 98% A and 2% B initially

for 16 min, was then increased to 98% B over 1 min, maintained for

12 min to elute other substances, and then returned to the initial

conditions before running the next sample. From peak areas, culture

medium concentrations of glutamate were computed by the use of

external standard.

Electrophysiological studies

NMDA-evoked inward currents in hippocampal neurons were

determined, at room temperature, under voltage-clamp, using the

nystatin-perforated patch-clamp recording technique (Horn & Marty,

1988) and an EPC-7 patch-clamp ampli®er (List Medicals,

Greenvale, NY, USA). The following solution was used as bathing

solution (in mM): NaCl, 140; KCl, 5; CaCl2, 2.5; HEPES, 10; glucose,

10; glycine, 10 mM; pH, 7.5. The pipette contained (in mM): KCl, 40;

K2SO4, 50; NaCl, 10; MgCl2, 1; HEPES, 10; nystatin, 250 mg/mL;

pH, 7.5. Patch electrodes were ®re-polished and had open tip

resistances of 2±4 MW. Neuronal capacitance was compensated using

the EPC7 internal circuitry and holding potential set at ±60 mV.

Current signals were ®ltered at 1 KHz using a Bassel ®lter (Frequency

Devices, NY. USA) and drugs were applied using a fast application

system (DAD-12; Adams & List Associates, NY, USA). To avoid

interneuronal variability in NMDA-induced currents we used each

cell as its own control. Control recordings were performed by two

consecutive (3 s; 10 min apart) applications (S1, S2) of NMDA.

During the interval between S1 and S2, the neuron was continuously

perfused with bath solution. When DHPG or calphostin C were used

alone, a control (S1) exposure to NMDA was performed. DHPG or

calphostin C were then continously perfused 5 min before and during

the second NMDA application (S2). Alternatively, DHPG was

perfused only during the second NMDA application. In the experi-

ments where DHPG and calphostin C were used together, after a

control NMDA exposure (S1), calphostin C was added to the

perfusion system and was present during 10 min interval between S1

and S2 and during the second application (S2). DHPG was added

5 min before and maintained during S2.

Currents were analysed using the pClamp (Axon Instruments,

Foster City, CA, USA). Difference between data sets was examined

by Student's two-tailed t-test, with signi®cance level at P < 0.05.

Average values are expressed as means 6 SEM.

Results

In mixed cultures of mouse cortical cells, a 10-min pulse of NMDA

induced neuronal degeneration (see Rose et al., 1992). Submaximal

concentrations of NMDA (30 or 60 mM, depending on responses of

individual culture preparations) were used in our experiments. These

concentrations killed 45±60% of the neuronal population leaving the

underlying astrocytes intact. When coapplied with NMDA, the group

I mGlu receptor agonist, DHPG, ampli®ed excitotoxic neuronal

death, in agreement with previous results (Bruno et al., 1995b).

DHPG could also enhance NMDA toxicity when applied for 1 min

and then removed 5 min prior to the addition of NMDA. The extent

of this potentiation was similar to that observed when DHPG was

coapplied with NMDA (Table 1). However, when applied twice, both

5 min before and during the NMDA pulse, DHPG substantially

protected against NMDA toxicity (Table 1). We observed a similar

switch in the response to DHPG when we measured the release of

endogenous glutamate associated with the NMDA pulse, an event

that contributes to the development of excitotoxic death (Monyer

et al., 1992). When measured immediately after the NMDA pulse,

extracellular glutamate levels increased by about two-fold as

compared with control cultures. One single exposure to DHPG

Functional switch of group I mGluRs in neurodegeneration 1471


(either 5 min before or during the NMDA pulse) further enhanced the

NMDA-stimulated glutamate release. In contrast, glutamate levels

returned back to normal when DHPG was ®rst applied 5 min before

and then reapplied in combination with NMDA (Table 1).

To study the time-dependency of the change in the response to

DHPG, we performed experiments in which the interval between the

®rst and the second application of the drug varied from 1 to 90 min.

Neuroprotection was consistently observed with time intervals of 1, 5,

15, 30 and 45 min. However, when the interval was extended to

90 min, the double exposure to DHPG enhanced, rather than reduced,

NMDA toxicity (Fig. 1). To examine whether PKC was involved in

the functional switch induced by a pre-exposure to DHPG (see

Herrero et al., 1998), we have treated the cultures with either

calphostin C (a selective PKC inhibitor) or H7 (a nonselective protein

kinase inhibitor). When calphostin C or H7 were present during the

®rst exposure to DHPG, a second application of DHPG failed to

produce neuroprotection (Fig. 2) and to reduce glutamate release

(Fig. 3). Both drugs abolished the potentiation of NMDA toxicity

(Fig. 2) and NMDA-stimulated glutamate release (not shown)

induced by the ®rst exposure to DHPG. At least H7 did not attenuate

neuroprotection when present exclusively during the second exposure

to DHPG (not shown). To con®rm a role for PKC in the activity-

dependent switch, we have pretreated the cultures with the PKC

activator (E)-2-methyl-6-(2-phenylethenyl)piridine (TPA, 100 nM)

for 1 min, and then applied DHPG and NMDA 5 min later. Under

these conditions, DHPG was neuroprotective, although to a lesser

extent than in cultures pretreated with DHPG (Table 2).

FIG. 1. The switch in the regulation of NMDA toxicity by DHPG in mixedcortical cultures is time-dependent. Cultures were pretreated with 100 mM

DHPG for 1 min (DHPGpt) at different time intervals before application of60 mM NMDA combined or not with a second application of DHPG(DHPGco). Time intervals between DHPTpt and NMDA (orNMDA + DHPGco) are indicated on the x-axis. Note that theneuroprotective activity of the sequence DHPGpt + DHPGco is maintainedwhen the interval between the two treatment ranges from 1±45 min, but isno longer observed with a time interval of 90 min. In this experiment,NMDA alone killed 45% of the neuronal population. DHPG alone (i.e.without NMDA) practically had no effect on neuronal viability. Values aremeans + SEM of three to six determinations. All values are statisticallysigni®cant (P < 0.05, One-way ANOVA + Fisher's PLSD), as compared withNMDA alone.

FIG. 2. Pretreatment of mixed cortical cultures with calphostin C or H7prevents the switch in the regulation of NMDA toxicity by DHPG. Cultureswere pretreated with 100 mM DHPG for 1 min (DHPGpt) and then treatedwith 60 mM NMDA either combined or not with DHPG (DHPGco), 5 minlater. When present, calphostin C (100 nM) or H7 (100 mM) were applied10 min before and maintained during the 1-min pretreatment with DHPG.NMDA alone killed about 50% of the neuronal population in controlcultures. Values are means + SEM of six determinations. *P < 0.01 (One-way ANOVA + Fisher's PLSD), as compared with NMDA alone.

FIG. 3. Pretreatment with calphostin C (100 nM) prevents the reduction ofglutamate release induced by a double application of DHPG (100 mM) inmixed cultures of cortical cells. The experiment was performed as describedin the legend of Fig. 2. Values are means + SEM of six determinations.*P < 0.01 (Student's t-test), as compared with NMDA alone.

TABLE 1. Activity-dependent switch in the regulation of NMDA toxicity and

NMDA-stimulated glutamate release by group I mGlu receptors in mixed

cultures of cortical cells

TreatmentNMDAtoxicity (%)

Extracellularglutamate (nM)

Control 7 6 2 57 6 5NMDA, 60 mM 137 6 7 135 6 21DHPGpt + NMDA 185 6 11* 210 6 18*NMDA + DHPGco 189 6 9* 186 6 11*DHPGpt + NMDA + DHPGco 72 6 4* 71 6 9*

Values are means 6 SEM of 4±8 determinations. *P < 0.01 (One-wayANOVA + Fisher's PLSD), as compared to NMDA alone. DHPGpt, DHPGpretreatment (1 min exposure, 5 min prior to the NMDA pulse); DHPGco,DHPG coadded with NMDA, during the excitotoxic pulse.



To unravel the identity of the individual mGlu receptor subtype

subjected to the activity-dependent switch in the regulation of NMDA

toxicity, we have used MPEP, SIB-1757 and SIB-1893, which behave

as potent and selective noncompetitive mGlu5 receptor antagonists

(Gasparini et al., 1999b; Varney et al., 1999), or CPCCOEt, which

behaves as a selective noncompetitive mGlu1 receptor antagonist

(Annoura et al., 1996; Litschig et al., 1999). Antagonists were

applied to the cultures 2 min before, and maintained during the ®rst

application of DHPG (total exposure, 3 min). MPEP, SIB-1757 and

SIB-1893 (all at 30 mM) attenuated (MPEP) or abolished (SIB-1757

or SIB-1893) the potentiation of NMDA toxicity produced by the ®rst

application of DHPG and, more important, completely abolished the

experience-dependent switch in the response to DHPG. Thus, the

second application of DHPG enhanced (rather than reduced) NMDA

toxicity when the ®rst application was performed in the presence of

mGlu5 receptor antagonists (Fig. 4A±C). It is noteworthy that MPEP,

but not SIB-1893 or SIB-1757, was neuroprotective when applied

alone 5 min prior to the NMDA pulse, although all drugs attenuate

NMDA toxicity when combined with NMDA (Bruno et al. 2000a).

Pretreatment with CPCCOEt (30 mM) did not affect the potentiation

of NMDA toxicity produced by the ®rst application of DHPG but

could also prevent the switch in the response to DHPG. In this

particular case, however, the second exposure to DHPG enhanced

NMDA toxicity to a lesser extent than the ®rst exposure (Fig. 4D),

differently to that observed in cultures pretreated with mGlu5 receptor

antagonists. Using mixed cultures of cortical cells we also examined

whether the switch in the response to DHPG was affected by PTX.

PTX (1 mg/mL) was applied to cultures 16 h prior to the experiment.

PTX treatment enhanced NMDA toxicity and this effect occluded the

ampli®cation of excitotoxic death produced by DHPG applied either

before or during the NMDA pulse. Opposed to that observed in

control cultures, a second application of DHPG failed to protect

against NMDA toxicity in cultures treated with PTX (Table 3). PTX

treatment also reduced the ability of the group III mGlu receptor

agonist, L-AP4, to attenuate NMDA toxicity (Table 3).

As glial mGlu5 receptors have been implicated in the regulation of

excitotoxic neuronal death (Nicoletti et al., 1999), we have examined

the effect of group I mGlu receptor agonists in pure cultures of

cortical neurons. These cultures are virtually devoid of astrocytes, as

assessed by double ¯uorescent staining for the neuronal marker

MAP-2 and the glial marker GFAP, followed by cyto¯uorimetric

analysis (Copani et al., 1999). In these cultures, quisqualate (in the

presence of the AMPA receptor antagonist, GYKI 52466, 10 mM)

enhanced NMDA toxicity when applied either 5 min before or during

the NMDA pulse, although the extent of this potentiation was lower

than that observed with DHPG in mixed cultures (Fig. 5A). Also in

this model, however, two consecutive applications of quisqualate

TABLE 2. Pretreatment with TPA produces a switch in the modulation of

NMDA toxicity by DHPG

Treatment Number of dead neurons

Control 13 6 2NMDA, 60 mM 218 6 13TPApt + NMDA 219 6 3NMDA + DHPGco 265 6 15*TPApt + NMDA + DHPGco 173 6 5*

Values are means 6 SEM of 4 determinations. *P < 0.01 (One-wayANOVA + Fisher's PLSD), as compared to NMDA alone. TPApt, TPApretreatment (1 min exposure, 5 min prior to the NMDA pulse); DHPGco,DHPG coadded with NMDA, during the excitotoxic pulse.

FIG. 4. Effects of MPEP (A), SIB-1757 (B), SIB-1893 (C) or CPCCOEt (D) on the induction of the experience-dependent switch in the regulation of NMDAtoxicity by DHPG in mixed cultures of cortical cells. MPEP, SIB-1757, SIB-1893 or CPCCOEt (all at 30 mM) were applied 2 min before and maintainedduring the 1-min pretreatment with DHPG (DHPGpt). Toxicity was induced with 60 mM NMDA, which killed 42% of the neuronal population. Values aremeans + SEM of six determinations. *P < 0.01 (One-way ANOVA + Fisher's PLSD), as compared with NMDA alone.



produced neuroprotection (Fig. 5A). Similar results were obtained

with DHPG, except that a single application of this drug did not

signi®cantly enhance NMDA toxicity in pure cultures of cortical

neurons (Fig. 5B).

Finally, we have assessed NMDA toxicity in cultures of

hippocampal pyramidal neurons, to establish whether the effect we

have observed in cortical neurons could be extended to other neuronal

types. Hippocampal pyramidal neurons were grown on coverslips

placed over a monolayer of con¯uent astrocytes. Under our experi-

mental conditions, neurons bearing the typical morphology of

pyramidal cells accounted for 81 6 8% of the cell population

(n = 5), similarly to that reported by Scholz & Miller (1991). This

estimate was indirectly con®rmed by the percentage of GAD positive

neurons, which was consistently less than 10%. Coverslips were

removed from the supporting astrocytes during exposure to NMDA

and/or DHPG and then returned to astrocytes at the end of the NMDA

pulse. Using this model, NMDA was applied at concentrations of

either 30 or 100 mM, which produced the death of 25 and 45% of the

neuronal population, respectively. With 30 mM NMDA, a coincuba-

tion with DHPG did not affect neuronal toxicity, whereas a 1-min

pre-exposure to DHPG (carried out 5 min before the NMDA pulse)

was neuroprotective per se. Neuroprotection was also observed when

DHPG was applied for the second time during the NMDA pulse

(Fig. 6). Similar results were obtained using 100 mM NMDA with the

exception that coapplication of DHPG was also neuroprotective

(Fig. 6). Hippocampal pyramidal neurons were also treated with

NMDA and DHPG using a slightly different experimental protocol.

In particular, cultures were pretreated with DHPG for 5 min and then

exposed to NMDA and DHPG, 1 min later. This double application

of DHPG induced neuroprotection, which was abolished by 1 mM

calphostin C (not shown). Hippocampal neurons plated on coverslips

were particularly suitable to the examination of how DHPG affects

electrophysiological responses to NMDA. An initial application of

NMDA (100 mM for 3 s) to cultured hippocampal neurons evoked an

inward current of 245 6 38 pA (n = 9), which remained constant in

response to a second application performed 10 min later (compare S1

and S2 in Fig. 7A). When applied for 5 min and then reapplied in

combination with NMDA (i.e. under conditions in which we observed

neuroprotection), DHPG enhanced, rather than reduced, NMDA

currents and its effect was prevented by calphostin C, which did not

affect NMDA currents on its own (Fig. 7B±C). DHPG combined with

NMDA without any pretratment also enhanced NMDA currents

(Fig. 7C). Thus, at least in hippocampal neurons, neuroprotection by

DHPG could not be ascribed to a direct modulation of NMDA

receptors by group I mGlu receptors.

Discussion

Data obtained in cultured cortical neurons indicate that group I mGlu

receptors differentially modulate NMDA toxicity depending on their

state of activation. While activation of pharmacologically `naõÈve'

receptors facilitates excitotoxic death, activation of receptors that

have already been stimulated results in neuroprotection. This

experience-dependent `switch' in the modulation of excitotoxicity

was induced by a very short pre-exposure to group I mGlu receptor

agonists, and lasted for at least 45 min. Induction of a functional

switch suggests that in cultured cortical neurons the endogenous

glutamate is too low to maintain group I mGlu receptors in a

`switched', protective, mode. Accordingly, extracellular glutamate

levels in mixed cultures of cortical cells were about 50±60 nM, a

value which is substantially lower than the reported EC50 for the

TABLE 3. Effect of DHPG on NMDA toxicity in mixed cultures of cortical

cells treated with PTX

NMDA toxicity in control cultures (%)

Controls PTX (1 mg/mL)

NMDA, 60 mM 100 6 3.6 152 6 7.9DHPGpt + NMDA 143 6 10* 158 6 5.4NMDA + DHPGco 160 6 9.5* 155 6 11DHPGpt + NMDA + DHPGco 66 6 4.4* 148 6 6.6NMDA + L-AP4co 51 6 5.5* 139 6 12

Values were calculated from 8 individual determinations. In control cultures,NMDA killed 65% of the neuronal population. *P < 0.01 (One-wayANOVA + Fisher PLSD) vs the respective NMDA values. pt, Pretreatmentfor 1 min, 5 min prior to NMDA; co, coapplication with NMDA. DHPG andL-AP4 were applied at concentrations of 100 mM.

FIG. 5. A double application of quisqualate (QUIS; A) or DHPG (B)protects pure cultures of cortical neurons against NMDA toxicity. QUIS(50 mM) or DHPG (100 mM) were preapplied 5 min before (-pt) and thenapplied again (-co) in combination with 100 mM NMDA. This concentrationof NMDA killed 63% of the cell population. The experiment with QUISwas performed in the presence of the AMPA receptor antagonist, GYKI52466. Values are means + SEM of four to eight determinations. *P < 0.01(One-way ANOVA + Fisher's PLSD), as compared with NMDA alone.



activation of recombinant mGlu1 or -5 receptors by glutamate

(ranging from 1±20 mM, Schoepp et al., 1999). The much higher

concentrations of extracellular glutamate found in cultured cerebellar

granule cells (about 4±5 mM, Aronica et al., 1993) might help to

explain why a single application of group I mGlu receptor agonists is

neuroprotective rather than neurotoxic in this particular model (Pizzi

et al., 1993, 1996a, 1996b). Similar to cultured granule cells, group I

mGlu receptors may exist in a switched mode in cultured

hippocampal pyramidal neurons, where DHPG was already neuro-

protective when applied for the ®rst time. This might re¯ect the high

percentage of glutamatergic pyramidal neurons found in these

cultures, as well as the high ®ring rate of this particular neuronal

type. The importance of the endogenous glutamatergic tonus in the

functional state of group I mGlu receptors is emphasized by

Rodriguez-Moreno et al. (1998), who showed that group I agonists

inhibit transmission at the synapses between Schaeffer collaterals and

CA1 pyramidal cells, but produce an opposite effect when the tonic

action of endogenous glutamate is removed, i.e. when receptors

return back to their `naõÈve' state. This has potential implications for

the treatment of brain disorders, such as stroke or status epilepticus,

in which elevated extracellular levels of glutamate contribute to

neuronal death (see Choi, 1992). Under these conditions, group I

mGlu receptor antagonists are expected to be highly effective before

or during the induction phase of neuronal death, whereas, they should

lose ef®cacy afterwards, when receptors are switched into a

protective mode by the endogenous glutamate.

The molecular mechanisms underlying the activity-dependent

switch from facilitation to inhibition may involve phosphorylation

processes mediated by PKC. This applies to the modulation of both

glutamate release (Herrero et al., 1998; see also present data) and

excitotoxicity, as indicated by the effect of PKC inhibitors. However,

protein kinases other than PKC might also be involved, because

neuroprotection by DHPG was less substantial after a previous

exposure to phorbol esters than after exposure to DHPG itself. G-

protein coupled receptor kinases (GRKs) are potential candidates,

because they have been shown to phosphorylate group I mGlu

receptors (Dale et al. 2000; Sallese et al. 2000). Phosphorylation of

group I mGlu receptors has been implicated in the mechanism of

receptor desensitization, i.e. in the loss of receptor response induced

by a prolonged exposure to agonists (Catania et al., 1991; Aramori &

Nakanishi, 1992; Aronica et al., 1993; Gereau IV & Heinemann,

1998). In this particular case, however, preactivated receptors not

only lose their original function, but also acquire a novel `opposite'

function. We expect that receptor phosphorylation switches the

coupling mechanism from one G-protein to another, although it

cannot be excluded that two different receptors with similar

pharmacology exert opposite functions and that agonist-dependent

FIG. 6. Effect of DHPG on NMDA toxicity in cultured hippocampalpyramidal neurons. DHPG (200 mM) was preapplied and then combinedwith NMDA (DHPGco). Note that DHPG failed to amplify NMDA toxicityin this particular preparation. Values are means + SEM of four to sixdeterminations and are expressed as percentage of neuronal toxicity inducedby 100 mM NMDA. *P < 0.01 (One-way ANOVA + Fisher's PLSD), ascompared with NMDA alone.

FIG. 7. DHPG enhances NMDA-induced inward currents in culturedhippocampal pyramidal neurons. Representative records of NMDA currentsare in A and B. (A) Inward currents induced by two consecutive 3-secapplications of 100 mM NMDA (S1 and S2; 10 min interval between thetwo applications). (B) Same as in A, but DHPG (200 mM) was present5 min before, and then reapplied during S2. (C) Calphostin C (1 mM,applied 10 min before and during S2) prevented the increase in NMDAcurrent induced by two consecutive applications of DHPG (pt and co).Values are means + SEM of eight to twelve determinations. *P < 0.01(One-way ANOVA + Fisher's PLSD), as compared with all other groups).



phosphorylation desensitizes exclusively the receptor that facilitates

NMDA toxicity. `NaõÈve' group I mGlu receptors are coupled to Gq or

Go, which transduce the extracellular signal into the stimulation of PI

hydrolysis via the activation of phospholipase Cb (Nakanishi, 1994).

Switched receptors might instead be coupled to different G-proteins

able to promote speci®c events that are more compatible with a

neuroprotective effect, such as the inhibition of Ca2+ channels

(Herrero et al., 1998). As an attempt to characterize the G-protein

type that couples to group I mGlu receptors before and after the

activity-dependent switch, we have treated the cultures with PTX.

The ef®cacy of the treatment was shown by the lower effect of L-AP4

(which activates group III mGlu receptors coupling to Gi) to protect

neurons against NMDA toxicity. PTX treatment alone ampli®ed

NMDA toxicity, perhaps by limiting the endogenous activation of

protective group II and group III mGlu receptors. A similar

potentiation of NMDA toxicity as observed in knock-out mice

lacking mGlu4 receptors (Bruno et al. 2000b). As the intrinsic effect

of PTX might have occluded the potentiation of NMDA toxicity

produced by a single application of DHPG, we cannot conclude

whether `naõÈve' group I mGlu receptors are coupled or not with a

PTX-sensitive G-protein. The lack of neuroprotection induced by a

second application of DHPG suggests that either the `switched'

receptor is coupled to a PTX-sensitive G-protein or that the switch

process requires the activation of a naõÈve receptor coupled to a PTX-

sensitive G-protein. This question cannot be solved by applying PTX

after the ®rst application of DHPG because the interval between the

®rst and the second application of DHPG is too short (< 90 min) to

allow the penetration and activation of PTX in cultured neurons.

Perhaps the identity of the G-protein could be examined by using

peptide inhibitors of Ga subunits fused with viral proteins that allow

fast penetration and release in neurons.

The mechanisms by which group I mGlu receptors modulate

excitotoxic neuronal death are still uncertain. Electron microscopy

analysis shows that mGlu1 and -5 receptors are preferentially, if not

exclusively, localized in the peripheral region of postsynaptic

densities (Baude et al., 1993; Shigemoto et al., 1997). However,

the existence of presynaptic receptors has been inferred from the

ability of group I mGlu receptor agonists to modulate glutamate

release (Herrero et al., 1992, 1994). One of the functions of

postsynaptic group I mGlu receptors is the positive modulation of

NMDA-gated ion currents (Aniksztejn et al., 1992; Bleakman et al.,

1992; Kelso et al., 1992; Harvey & Collingridge, 1993; Yu et al.,

1997). Activation of these receptors may facilitate excitotoxic

neuronal death by amplifying NMDA currents. However, it is

unlikely that these are the receptors that change their function in an

experience-dependent manner, because, at least in cultured hippo-

campal pyramidal cells, a second application of DHPG still enhanced

NMDA currents, but produced neuroprotection. In contrast, in mixed

cultures of cortical cells, a single application of group I mGlu

receptor agonists ampli®es NMDA toxicity (Bruno et al., 1995b;

Buisson & Choi, 1995; present data) but reduces NMDA currents (Yu

et al., 1997). One possible explanation is that the `functional switch'

is an exclusive property of presynaptic group I mGlu receptors

controlling glutamate release. These receptors might have the basic

function of facilitating glutamate release, but they `switch' to avoid

an excessive, potentially harmful, release of glutamate. This may

explain our results in mixed cortical cultures, where NMDA toxicity

has two components; a component intrinsic to the activation of

NMDA receptors, and a second component mediated by the

endogenously released glutamate (Monyer et al., 1992) acting on

other `harmful' receptors, such as AMPA/kainate or neurotoxic mGlu

receptors. The possibility that only this second component is

regulated by group I mGlu receptors, in an experience-dependent

fashion, is suggested by the evidence that a second application of

DHPG completely abolished NMDA-stimulated glutamate release,

but only partially protected against NMDA toxicity.

We searched for the identity of the mGlu receptor undergoing this

experience-dependent switch by using a battery of selective

noncompetitive antagonists of mGlu1 or -5 receptors. There was an

interesting difference between these two classes of antagonists. In

cultures pretreated with mGlu5 antagonists (particularly with SIB-

1757 or SIB-1893), the ®rst application of DHPG virtually lost its

ability to potentiate NMDA toxicity, and the second application of

DHPG ampli®ed the action of NMDA to an extent similar to what

observed in cultures that had not been pretreated with DHPG. In

cultures pretreated with the mGlu1 antagonist CPCCOEt, the ®rst

application of DHPG fully retained its ability to potentiate NMDA

toxicity, and this potentiation was reduced by more than 70% when

DHPG was applied for the second time. However, the second

application of DHPG did not reduce neuronal toxicity below the level

observed with NMDA alone, and it was regularly observed in cultures

which had not been pretreated with any antagonist. This suggests that

a preapplication of DHPG potentiates NMDA toxicity through the

activation of mGlu5 receptors and that activation of mGlu5 receptors

is largely required for the induction of the experience-dependent

switch in the response to DHPG. This parallels results obtained by

examining the regulation of glutamate release, in which the

experience-dependent switch is regularly observed in cerebrocortical

nerve terminals prepared from mGlu1 knockout mice (Sistiaga et al.,

1998). However, the lower extent of neuroprotection induced by the

second application of DHPG, in cultures pretreated with CPCCOEt,

suggests that activation of mGlu1 receptors may also contribute to the

induction phase of the `switch process'. It remains to be established

whether it is the same mGlu5 receptor that produces neuroprotection

in response to the second application of DHPG. Unfortunately, we

could not address this issue because all mGlu5 receptor antagonists

are neuroprotective by themselves when applied in combination with

NMDA (Bruno et al. 2000a), and MPEP was protective even when

applied prior to the NMDA pulse (see Fig. 4A).

Besides providing a way to explain contrasting data on the role of

group I mGlu receptors in neurodegeneration, present results indicate

the effectiveness of group I related neuroprotective drugs. The

ef®cacy of these neuroprotective drugs could be related to the

functional state of group I mGlu receptors, which depends on how

frequently these receptors are activated by endogenous glutamate.

Thus, group I mGlu receptor antagonists may be ef®cient neuropro-

tectants when they block `naõÈve' receptors, such as perisynaptic

receptors which can be acutely recruited by supraphysiological levels

of glutamate in the initial phase of an excitotoxic event (for example

during stroke or brain trauma). In contrast, group I mGlu receptor

antagonists may be less effective when they target receptors which

spend most of their lifespan in a `switched con®guration'; as may

occur in chronic disorders, such as amyotrophic lateral sclerosis, in

which extracellular glutamate levels remain high due to a defect in

the activity of glutamate transporters or other clearing mechanisms.

Acknowledgements

This work has been supported in part by grant 1238 from Telethon-Italy(F.N.), grants SAF96-0169 and 1FD97-0500 from CICYT and a grant fromFundacioÂn Navarro TrõÂpodi to V.C.; M.F.G. and M.V.C. are recipients offellowships from the University Miguel HernaÂndez.



Abbreviations

CPCCOEt, 7-(hydroxyimino)cyclopropa[b]chromen-1a-carboxylate ethylester; DHPG, 3,5-dihydroxyphenylglycine; DIV, days in vitro; H7, (6)-1-(5-isoquinolinesulphonyl)-2-methylpiperazine; GAD, glutamate decarboxylase;GRKs, G-protein coupled receptor kinases; HBSS, Hanks balanced saltsolution; mGlu, metabotropic glutamate; MPEP, 6-(2-phenyl-1-ethynyl)-piridine; NMDA, N-methyl-D-aspartate; PI, polyphosphoinositide; PKC,protein kinase C; SIB-1757, 6-methyl-2-(phenylazo)-3-pyridinol; SIB-1893,(E)-2-methyl-6-(2-phenylethenyl)piridine; TPA, (E)-2-methyl-6-(2-pheny-lethenyl)piridine.

References

Aniksztejn, L., Otani, S. & Ben-Ari, Y. (1992) Quisqualate metabotropicreceptors modulate NMDA currents and facilitate induction of long-termpotentiation through protein kinase C. Eur. J. Neurosci., 4, 500±505.

Annoura, H., Fukunaga, A., Uesugi, M., Tatsuoka, T. & Horikawa, Y. (1996)A novel class of antagonists for metabotropic glutamate receptors, 7-(hydroxyimino) cyclopropa[b]chromen-1a-carboxylates. Bioorg. Med.Chem. Lett., 6, 7763±7766.

Aramori, I. & Nakanishi, S. (1992) Signal transduction and pharmacologicalcharacteristics of a metabotropic glutamate receptor, mGluR1, in transfectedCHO cells. Neuron, 8, 757±765.

Aronica, E., Dell'Albani, P., Condorelli, D.F., Nicoletti, F., Hack, N. &Balazs, R. (1993) Mechanisms underlying developmental changes in theexpression of metabotropic glutamate receptors in cultured cerebellargranule cells: homologous desensitization and interactive effects involvingN-methyl-D-aspartate receptors. Mol. Pharmacol., 44, 981±989.

Baude, A., Nusser, Z., Robert, J.D.B., Mulvihill, E., McIlhinney, R.A.J. &Somogyi, P. (1993) The 1a form of metabotropic glutamate receptor(mGluR1a) is concentrated at extra and perisynaptic membrane of discretesubpopulations of neurons as detected by immunogold reaction in the rat.Neuron, 11, 771±787.

Behrens, M.M., Strasser, U., Heidinger, V., Lobner, D., Yu, S.P., McDonald,J.W., Won, M. & Choi, D.W. (1999) Selective activation of group IImGluRs with LY354740 does not prevent neuronal excitotoxicity.Neuropharmacology, 38, 1621±1630.

Bleakman, D., Rusin, K.I., Chard, P.S., Glaum, S.R. & Miller, R.J. (1992)Metabotropic glutamate receptors potentiate ionotropic glutamate responsesin the rat dorsal horn. Mol. Pharmacol., 42, 192±196.

Bruno, V., Battaglia, G., Casabona, G., Copani, A., Caciagli, F. & Nicoletti, F.(1998) Neuroprotection by glial metabotropic glutamate receptors ismediated by transforming growth factor-beta. J. Neurosci., 18, 9594±9600.

Bruno, V., Battaglia, G., Copani, A., Giffard, R.G., Raciti, G., Raffaele, R.,Shinozaki, H. & Nicoletti, F. (1995a) Activation of class II or IIImetabotropic glutamate receptors protects cultured cortical neuronsagainst excitotoxic degeneration. Eur. J. Neurosci., 7, 1906±1913.

Bruno, V., Battaglia, G., Ksiazek, I., van der Putten, H., Catania, M.V.,Giuffrida, R., Lukic, S., Leonhardt, T., Inderbitzin, W., Gasparini, F., Kuhn,R., Hampson, D.R., Nicoletti, F. & Flor, P.J. (2000b) Selective activation ofmGlu4 metabotropic glutamate receptors is neuroprotective againstexcitotoxic neuronal death. J. Neurosci., 20, 6413±6420.

Bruno, V., Copani, A., KnoÈpfel, T., Kuhn, R., Casabona, G., Dell'Albani, P.,Condorelli, D.F. & Nicoletti, F. (1995b) Activation of metabotropicglutamate receptors coupled to inositol phospholipid hydrolysis ampli®esNMDA-induced neuronal degeneration in cultured cortical cells.Neuropharmacology, 34, 1089±1098.

Bruno, V., Ksiazek, I., Battaglia, G., Lukic, S., Leonhardt, T., Sauer, D.,Gasparini, F., Kuhn, R., Nicoletti, F. & Flor, P.J. (2000a) Selective blockadeof metabotropic glutamate receptor subtype 5 is neuroprotective.Neuropharmacology, 39, 2223±2230.

Bruno, V., Sureda, F.X., Storto, M., Casabona, G., Caruso, A., KnoÈpfel, T.,Kuhn, R. & Nicoletti, F. (1997) The neuroprotective activity of group-IImetabotropic glutamate receptors requires new protein synthesis andinvolves a glial-neuronal signaling. J. Neurosci., 17, 1891±1897.

Buisson, A. & Choi, D.W. (1995) The inibitory mGluR agonist, S-4-carboxy-3-hydroxy-phenylglycine selectively attenuates NMDA neurotoxicity andoxygen-glucose deprivation-induced neuronal death. Neuropharmacology,34, 1081±1087.

Catania, M.V., Aronica, E., Sortino, M.A., Canonico, P.L. & Nicoletti, F.(1991) Desensitization of metabotropic glutamate receptors in neuronalcultures. J. Neurochem., 56, 1329±1335.

Catania, M.V., Copani, A., Calogero, A., Ragonese, G.I., Condorelli, D.F. &

Nicoletti, F. (1999) An enhanced expression of the immediate early gene,Egr-1, is associated with neuronal apoptosis in culture. Neuroscience., 91,1529±1538.

Choi, D.W. (1992) Excitotoxic cell death. J. Neurobiol., 23, 1261±1276.Conn, P.J. & Pin, J.-P. (1997) Pharmacology and functions of metabotropic

glutamate receptors. Annu. Rev. Pharmacol. Toxicol., 37, 205±237.Copani, A., Condorelli, F., Caruso, A., Vancheri, C., Sala, A., Giuffrida Stella,

A.M., Canonico, P.L., Nicoletti, F. & Sortino, M.A. (1999) Mitoticsignaling by beta-amyloid causes neuronal death. FASEB J., 13, 2225±2234.

Dale, L.B., Bhattacharrya, M., Anborgh, P.H., Murdoch, B., Bhadia, M.,Nakanishi, S. & Ferguson, S.S. (2000) G protein coupled receptor kinasemediated desensitization of metabotropic glutamate receptor 1a protectsagainst cell death. J. Biol. Chem., 275, 38213±38220.

Gasparini, F., Bruno, V., Battaglia, G., Lukic, S., Leonhardt, T., Inderbitzin,W., Laurie, D., Sommer, B., Varney, M.A., Hess, S.D., Johnson, E.C.,Kuhn, R., Urwyler, S., Sauer, D., Portet, C., Schmutz, M., Nicoletti, F. &Flor, P.J. (1999a) (R,S) -4-phosphonophenylglycine, a potent and selectivegroup III metabotropic glutamate receptor agonist, is anticonvulsive andneuroprotective in vivo. J. Pharmacol. Exp. Ther., 289, 1678±1687.

Gasparini, F., LingenhoÈhl, K., Stoehr, N., Flor, P.J., Heinrich, M., Vranesic, I.,Biollaz, M., Allgeirer, H., Heckendorn, R., Urwyler, S., Varney, M.A.,Johnson, E.C., Hess, S.D., Rao, S.P., Sacaan, A.I., Santori, E.M., VelicËelebi,G. & Kuhn, R. (1999b) 2-Methyl-6-(phenylethynyl) -pyridine (MPEP), apotent, selective and systemically active mGlu5 receptor antagonist.Neuropharmacology, 38, 1493±1503.

Gereau IV, R.W. & Heinemann. S.F. (1998) Role of protein kinase Cphosphorylation in rapid desensitization of metabotropic glutamate receptor5. Neuron, 20, 143±151.

Harvey, J. & Collingridge, G.L. (1993) Signal transduction pathways in theacute potentiation of NMDA responses by 1S,3R-ACPD in rat hippocampalslices. Br. J. Pharmacol., 109, 1085±1090.

Herrero, I., Miras-Portugal, M.T. & Sanchez-Prieto, J. (1992) Positivefeedback of glutamate exocytosis by metabotropic presynaptic receptorstimulation. Nature, 360, 163±166.

Herrero, I., Miras-Portugal, M.T. & Sanchez-Prieto, J. (1994) Rapiddesensitization of the presynaptic metabotropic receptor that facilitatesglutamate exocytosis. Eur. J. Neurosci., 6, 115±120.

Herrero, I., Miras-Portugal, M.T. & Sanchez-Prieto, J. (1998) Functionalswitch from facilitation to inhibition in the presynaptic control of glutamaterelease by metabotropic glutamate receptors. J. Biol. Chem., 273,1951±1958.

Horn, R. & Marty, A. (1988) Muscarinic activation of ionic currents measuredby a new whole-cell recording method. J. Gen. Physiol., 92, 145±159.

Kelso, S.R., Nelson, T.E. & Leonard, J.P. (1992) Protein kinase C-mediatedenhancement of NMDA currents by metabotropic glutamate receptors inXenopus oocytes. J. Physiol. (Lond.), 449, 705±718.

Litschig, S., Gasparini, F., Ruegg, D., Munier, N., Flor, P.J., Vranesic, I.-T.,PreÁzau, L., Pin, J.-P., Thomsen, C. & Kuhn, R. (1999) CPCCOEt, a non-competitive mGluR1 antagonist, inhibits receptor signaling withoutaffecting glutamate binding. Mol. Pharmacol., 55, 453±461.

Monyer, H., Giffard, R.G., Hartley, D.M., Dugan, L.L., Goldberg, M.P. &Choi, D.W. (1992) Oxygen and glucose-deprivation-induced neuronalinjury in cortical cell cultures is reduced by tetanus toxin. Neuron, 8,967±973.

Nakanishi, S. (1994) Metabotropic glutamate receptors: synaptic transmission,modulation and plasticity. Neuron, 13, 1031±1037.

Nicoletti, F., Bruno, V., Catania, M.V., Battaglia, G., Copani, A., Barbagallo,G., CenÄa, V., Sanchez-Prieto, J., Spano, P.F. & Pizzi, M. (1999) Group Imetabotropic glutamate receptors: hypotheses to explain their dual role inneurotoxicity and neuroprotection. Neuropharmacology, 38, 1477±1484.

Nicoletti, F., Bruno, V., Copani, A., Casabona, G. & KnoÈpfel, T. (1996)Metabotropic glutamate receptors: a new target for the teraphy ofneurodegenerative disorders? Trends Neurosci., 19, 267±271.

Opitz, T. & Reymann, K.G. (1993) (1S,3R)-ACPD protects synaptictransmission from hypoxia in hippocampal slices. Neuropharmacology,32, 103±104.

Pin, J.P. & Duvoisin, R. (1995) The metabotropic glutamate receptors:structure and functions. Neuropharmacology, 34, 1±26.

Pizzi, M., Boroni, F., Moraitis, K., Bianchetti, A., Memorandum, M. & Spano,P.F. (1999) Reversal of glutamate excitotoxicity by activation of PKC-associated metabotropic glutamate receptors in cerebellar granule cellsrelies on NR2C subunit expression. Eur. J. Neurosci., 11, 2489±2496.

Pizzi, M., Consolandi, O., Memorandum, M. & Spano, P.F. (1996b)Activation of multiple metabotropic glutamate receptor subtypes prevents



NMDA-induced excitotoxicity in rat hippocampal slices. Eur. J. Neurosci.,8, 1516±1521.

Pizzi, M., Fallacara, C., Arrighi, V., Memorandum, M. & Spano, P.F. (1993)Attenuation of excitatory amino acid toxicity by metabotropic glutamatereceptor agonists and aniracetam in primary cultures of cerebellar granulecells. J. Neurochem., 61, 683±689.

Pizzi, M., Gallic, P., Consolandi, O., Arrighi, V., Memorandum, M. & Spano,P.F. (1996a) Metabotropic and ionotropic transducers of glutamate signalsinversely control cytoplasmic Ca2+ concentration and excitotoxicity incultured cerebellar granule cells: pivotal role of protein kinase C. Mol.Pharmacol., 49, 586±594.

Prehn, J.H. & Miller, R.J. (1996) Opposite effects of TGF-beta1 on rapidly-and slowly-triggered excitotoxic injury. Neuropharmacology, 35, 249±256.

Rodriguez-Moreno, A., Sistiaga, A., Lerma, J. & Sanchez-Prieto, J. (1998)Switch from facilitation to inhibition of excitatory synaptic transmission bythe desensitization of group I mGluRs in the rat hippocampus. Neuron, 21,1477±1486.

Rose, K., Goldberg, M.P. & Choi, D.W. (1992) Cytotoxicity in murineneocortical cell cultures. In Tyson, C.A. & Frazier, J.M. (eds), Methods inToxicology, Vol. 1. San Diego Academics, San Diego, pp. 46±60.

Sallese, M., Salvatore, L., D'Urbano, E., Sala, G., Storto, M., Launey, T.,Nicoletti, F., KnoÈpfel, T. & De Blasi, A. (2000) The G-protein-coupledreceptor kinase GRK4 mediates homologous desensitization ofmetabotropic glutamate receptor 1. FASEB J., 14, 2569±2580.

Schoepp, D.D. & Conn, P.J. (1993) Metabotropic glutamate receptors in brainfunction and pathology. Trends Pharmacol. Sci., 14, 13±20.

Schoepp, D.D., Jane, D.E. & Monn, J.A. (1999) Pharmacological agents actingat subtypes of metabotropic glutamate receptors. Neuropharmacology, 38,1431±1476.

Scholz, K.P. & Miller, R.J. (1991) Analysis of adenosine actions on Ca2+

currents and synaptic transmission in cultured rat hippocampal pyramidalneurones. J. Physiol. (Lond.), 435, 373±393.

Schroder, U.H., Opitz, T., Jager, T., Sabelhaus, C.F., Breder, J. & Reymann,K.G. (1999) Protective effect of group I metabotropic glutamate receptoractivation against hypoxic/hypoglycemic injury in rat hippocampal slices:timing and involvement of protein kinase C. Neuropharmacology, 38,209±216.

Shah, M. & Haylett, D.G. (2000) Ca2+ channels involved in the generation ofthe slow afterhyperpolarization in cultured rat hippocampal pyramidalneurons. J. Neurophysiol., 83, 2554±2561.

Shigemoto, R., Kinoshita, A., Wada, E., Nomura, S., Ohishi, H., Takada, M.,Flor, P.J., Neki, A., Abe, T., Nakanishi, S. & Mizuno, N. (1997) Differentialpresynaptic localization of metabotropic glutamate receptor subtypes in rathippocampus. J. Neurosci., 17, 7503±7522.

Sistiaga, A., Herrero, I., Conquet, F. & Sanchez-Prieto, J. (1998) Themetabotropic glutamate receptor 1 is not involved in the facilitation ofglutamate release in cerebrocortical nerve terminals. Neuropharmacology,3, 1485±1492.

Varney, M.A., Cosford, N., Jachec, C., Rao, S.P., Sacaan, A., Lin, F.F.,Bleicher, L., Santori, E.M., Flor, P.J., Allgeier, H., Gasparini, F., Kuhn, R.,Hess, S.D., Velicelebi, G. & Johnson, C. (1999) SIB-1757 and SIB-1893:selective non-competitive antagonists at metabotropic glutamate receptortype 5. J. Pharmacol. Exp. Ther., 290, 170±181.

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



an activity-dependent switch from facilitation to inhibition in the control of excitotoxicity by...

Documents