Naunyn-Schmiedeberg's Arch Pharmacol (1994) 349:113 117 Naunyn-Schmiedeberg's

Archivesof Pharmacology © Springer-Verlag 1994

N-Methyl-D-aspartate (NMDA)-stimulated noradrenaline (NA) release in rat brain cortex is modulated by presynaptic H3-receptors K. Fink, E. Schlicker, M. G6thert

lnstitut for Pharmakologie und Toxikologie der Rheinischen Friedrich-Wilhelms-Universit~t Bonn, Reuterstrasse 2b, D-53113 Bonn, Germany

Received: 12 July 1993/Accepted: 18 October 1993

Abstract. In superfused rat brain cortex slices and synap- tosomes preincubated with [3H]noradrenaline the effect of agonists or antagonists at presynaptic H 3 receptors on NMDA-evoked [3H]noradrenaline release was investigat- ed. In experiments on slices, histamine and the preferen- tial H 3 receptor agonist R-(-)-a-methylhistamine inhib- ited NMDA-evoked tritium overflow (IC20 values 0.27 ~tmol/1 or 0.032 gmol/l , respectively); S-(+)-a- methylhistamine (up to 10 ~tmol/1) as well as the selective H~ receptor agonist (2-(2-thiazolyl)ethylamine and the selective H2 receptor agonist dimaprit (each up to 10 gmol/1) were ineffective. The H 3 receptor antagonist thioperamide abolished the inhibitory effect of histamine whereas the preferential H 1 receptor antagonist di- metindene and the preferential H 2 receptor antagonist ranitidine were ineffective. In experiments on synap- tosomes, histamine and R-(-)-a-methylhistamine in- hibited NMDA-evoked tritium overflow, whereas 2-(2- thiazolyl)ethylamine or dimaprit had no effect. The in- hibitory effect of histamine was abolished by thioper- amide. When tritium overflow was stimulated by NMDA in the presence of co-conotoxin GVIA (which by itself de- creased the response to NMDA by about 55%), R-( - ) - a-methylhistamine did not inhibit NMDA-evoked over- flow. It is concluded that NMDA-evoked noradrenaline release in the cerebral cortex can be modulated by inhibi- tory H3 receptors. NMDA receptors and H 3 receptors are both located presynaptically and may interact at the same noradrenergic varicosity. An unimpaired function of the N-type voltage-sensitive calcium channel probably is a prerequisite for the inhibition of NMDA-evoked nor- adrenaline release by H 3 receptor stimulation.

Key words: NMDA receptor - H3-receptor - Nor- adrenaline release - Presynaptic receptors - Rat brain cortex

Correspondence to." M. GOthert at the above address

Introduction

Evidence has been presented that NMDA induces nor- adrenaline release in rat or human brain cortex slices (Fink et al. 1989, 1992). Experiments on rat cortical and hippocampal synaptosomes provided similar results (Fink et al. 1990; Fink and G6thert 1992; Pittaluga and Raiteri 1990, 1992), suggesting that at least part of the NMDA receptors involved are located presynaptically on the noradrenergic varicosities. According to data ob- tained on slices, but not yet on synaptosomes, the noradrenergic axon terminals are also endowed with in- hibitory presynaptic H 3 receptors (Schlicker et al. 1989). Therefore, the same varicosities may be equipped with both types of receptors. The present study was carried out in order to examine whether the two receptor systems can functionally interact at the level of the same single varicosities. Such an interaction between an inhibitory presynaptic receptor and the presynaptic stimulatory NMDA receptor has recently been demonstrated between NMDA receptors and a2-autoreceptors (Pittaluga and Raiteri 1992; Fink and GOthert 1993), but analogous data for the interaction with a presynaptic inhibitory heterore- ceptor are not yet available.

A preliminary account of some of the present data has been given at the International Symposium on Presynaptic Receptors and Neuronal Transporters at Rouen 1990 (Fink et al. 1991).

Methods

Rat brain cortices were prepared from male Wistar rats weighing 200-300g. Slices (0.3 mm thick, diameter 3 mm) were incubated in Krebs' solution (37°C, 30min) composed as follows (mmol/1): NaCI 118, KCI 4.8, NaHCO 3 25, KH2PO 4 1.2, CaCI 2 1.3, MgSO 4 1.2, glucose 11.1, ascorbic acid 0.06, disodium EDTA 0.03 (equilibrated with 95% 02 and 5% CO2). During incubation this solution contained 50 nmol/1 [3Hlnoradrenaline ([3HINA).

Synaptosomes were prepared essentially as described by Gray and Whittaker (1962) with slight modifications (Fink and G6thert 1992). A 10% (w/v) homogenate of cortical tissue (without the frontal poles) was prepared in 0.32 mol/1 sucrose by means of a Potter-Elvehjem glass ho-

114

mogenizer with a rotating teflon pestle (1000 rpm, 6 strokes/2 rain). The homogenate was centrifuged at 1000×g for 10 min (4 °C), and 9 ml of the supernatant plus 6 ml of Krebs' solution were incubated for 7 rain at 37 °C. After addition of [3H]NA incubation was continued for an- other 7 rain. The labelled suspension was centrifuged at 600xg for 10 min (4°C) and the resulting pellet was resuspended in 2.25 ml ice- cold Krebs' solution (protein content 1023 _+ 83 gg protein/0.1 ml, deter- mined by the method of Lowry et al. 1951).

Single slices or aliquots of 100 gl of the synaptosomal suspension were distributed on Whatman GF/C or GF/B filters, respectively, and placed into 24 parallel superfusion chambers. Subsequently, the slices or synaptosomes were superfused with Mg2+-free Krebs solution for 62 min at a flow rate of 0.6 ml/min. In experiments on slices the super- fusion solution contained 1 ~tmol/1 desipramine and 1 ~tmol/1 idazoxan (unless stated otherwise). In the experiments on synaptosomes it con- tained 100 ~tmol/1 glycine (for discussion of reasons, see Fink et al. 1990; Pittaluga and Raiteri 1990). Substances under investigation were present in the buffer from 20 rain of superfusion onward (exception: ~o- conotoxin GVIA which was present during the 2-min stimulation period only). The superfusion technique in synaptosomes was based on that described by Raiteri et al. (1974).

Tritium overflow was stimulated by adding 100 gmol/1 (slices; unless stated otherwise) or 1 mmol/1 (synaptosomes) NMDA, for 2 min after 40 min of superfusion. The superfusate was continuously collected in 5-min (slices) or 4-min (synaptosomes) samples. At the end of the exper- iments the radioactivity in the superfusate samples, the synaptosomes (extracted with 0.25 tool/1 HCI) and slices (solubilized with 0.5 ml Soluene ®) was determined by liquid scintillation counting.

Tritium efflux was calculated as the fraction of the tritium content in the slice or synaptosomes at the beginning of the respective collection period (fractional rate of tritium efflux). To quantify effects of drugs on basal efflux, the fractional rate was determined in the sample collect- ed immediately before stimulation. Stimulation-evoked tritium over- flow was calculated by subtraction of the estimated basal from total ef- flux during the stimulation period and the following 10 (synaptosomes) or 13 min (slices) and was expressed as percentage of the tritium content in the slices or synaptosomes at the onset of stimulation (basal efflux was assumed to decline linearly from the sample before to that 12-16 (synaptosomes) or 15- 20 rain (slices) after onset of stimulation). Con- trol experiments were always run in parallel, and evoked tritium over- flow was calculated as percentage of these controls.

Results are given as means _+ SEM of n experiments in slices or n ex- periments in quadruplicate in synaptosomes. For comparison of mean values, Student's t-test was used. In case of multiple comparisons, Dun- nett's test was applied.

with time. Under the conditions applied in this study, the drugs at the concentrations investigated did not affect basal efflux (results not shown). The only exception was an increase by 87+ 12°70 which was caused by 10 ~tmol/1 histamine in the absence of desipramine and idazoxan (thus excluding the evaluation of the corresponding NMDA-evoked tritium overflow).

NMDA-evoked tritium overflow in slices

When no desipramine or idazoxan was present in the superfusion fluid, NMDA 30 to 100 gmol/1 induced a concentration-dependent overflow of tritium from cere- bral cortical slices (Table 1). In the presence of t ~tmol/1 desipramine, the tritium overflow evoked by 100 ~tmol/1 NMDA was not significantly affected (Table i); the fail- ure of desipramine to increase overflow in spite of its blocking effect on the neuronal noradrenaline transporter was probably due to the inhibitory effect of this drug on NMDA-induced responses (White et al. 1990). Presence of I ~tmol/1 idazoxan in addition to 1 gmol/1 desipramine in the superfusion fluid produced a 2.5-2.8 fold increase in NMDA (30 or 100 gmol/1)-evoked tritium overflow compared to the drug-free control experiments (Table 1).

Effects of histamine receptor agonists. In the presence of desipramine plus idazoxan, the NMDA (100 gmol/1)- evoked 3H overflow was concentration-dependently in- hibited by histamine (Fig. 1) and R-(-)-a-methylhista- mine (Fig. 2). The inhibitory effect of 10 gmol/1 hista- mine also occurred when under otherwise identical con- ditions NMDA was applied at a concentration of 30 instead of 100gmol/1 (result not shown). However, 0.1 - 10 gmol/1 histamine did not inhibit the tritium over- flow evoked by 100 gmol/I NMDA when idazoxan was omitted from the superfusion fluid, i.e. when the lat- ter contained only desipramine (Fig. 1). In the absence of both, idazoxan and desipramine, histamine

Drugs used. (-)-[ring 2,5,6-3H]-Noradrenaline (NEN, Dreieich, Ger- many); glycine, N-methyl-D-aspartate (Sigma Chemical Co., St. Louis, MO, USA); idazoxan (Reckitt & Colman, Hull, UK); desipramine hy- drochloride (CIBA-Geigy, Wehr, Germany; dimaprit dihydrochloride, 2-(2-thiazolyl)-ethylamine-dihydrochloride (SKF, Welwyn Garden City, UK); dimetindene maleate (Zyma, Mt~nchen, Germany); histamine dihydrochloride (Merck, Darmstadt, Germany); R-(-)-a-methylhista- mine, S-(+)-a-methylhistamine (Prof. Dr. Dr. W. Schunack, Institut ftir Pharmazie, Freie Universitgt, Berlin, Germany); ranitidine hydrochlo- ride (Glaxo, Ware, UK); thioperamide (Prof. Dr. J.-C. Schwartz, Centre Paul Broca de I'INSERM, Paris, France); ~o-conotoxin GVIA (RBI, Na- tick, MA, USA).

Stock solutions were prepared with dimethyl sulfoxide (thioper- amide), ascorbic acid 6 mmol/l ([3Hlnoradrenaline) or demineralized water (other drugs).

Results

Basal tritium efflux

In control experiments, basal tritium efflux from rat brain cortex slices and synaptosomes preincubated with 50nmol/1 [3H]-noradrenaline declined continuously

Table 1. NMDA-evoked tritium overflow above basal efflux from rat brain cortex slices in control experiments carried out in the absence or presence of auxiliary drugs throughout superfusion. The slices were preincubated with [3H]noradrenaline and superfused with Mg2+-free salt solution. NMDA was added for 2 min after 40 min of superfusion. Means +_ SEM of 6 - 8 experiments, n.d., not determined

NMDA-evoked overflow (% of tissue tritium a and nCi b)

NMDA No auxiliary Desipramine (gmol/1) drug (1 ~mol/l)

Desipramine (i gmol/1) plus Idazoxan (1 gmol/1)

30 1.97+_0.38% n.d. O. 70++_ 0.12 nCi

100 8.61 +_ i .06% 6.91 _+ 0.61% 3.35+_0.45nCi 2.13+_0.21nCi

300 17.46 _+ 2.36% n.d. 5.29 + 0.53 nCi

5.00 +_ 0.84% 1.88 +- 0.40 nCi

24.25 _+ 2.80% 8.64 +_ 1.39 nCi

n.d.

a At the onset of the stimulation period b In italics

,~- I00-

o

50-

$

0.1 1 10 Histamine (IJm01/I)

Fig. 1. Effects of histamine on the NMDA-evoked tritium overflow from rat brain cortex slices preincubated with [3H]noradrenaline and super- fused with Mg2+-free salt solution. NMDA was added for 2 min after 40 min of superfusion. Each concentration of histamine was present from 20 min of superfusion onward until the end of the experiments. • • The NMDA concentration was 100 gmot/1; the superfusion solution contained 1 ;imol/1 desipramine and 1 ~lmol/l idazoxan throughout the experiments. A A The NMDA concentration was 100 gmol/1; the superfusion solution contained 1 ~tmol/l desipramine but no idazoxan throughout the experiments. [] [] The NMDA concentration was 300 ~tmol/1; no auxiliary drugs were added to the superfusion solution. *P<0.05, compared to the corresponding con- trol; means _+ SEM of 6 - 8 experiments. For absolute values of NMDA- evoked tritium overflow under control conditions, see Table i

(0.1-3 ~tmol/1) did also not affect the NMDA (300 ~tmol/1)-evoked tritium overflow (Fig. 1).

In the experiments in which the superfusion fluid con- tained idazoxan and desipramine, the highest concentra- tions of histamine and R-(-)-a-methylhistamine investi- gated (10 and 1 ;lmol/1, respectively) caused an inhibition of the NMDA (100 ~lmol/1)-evoked 3H overflow by 41% and 35%, respectively (Figs. I and 2; for reasons of limited supply, R-(-)-a-methylhistamine concentrations exceeding i ;imol/1 could not be applied). IC20 values, as a measure of potency, were determined graphically for histamine (0.27 Blmol/1) and R-(-)-a-methylhistamine (0.032 ~tmol/1). S-(+)-a-methylhistamine was ineffective at all concentrations tested (0.01- 10 ~tmol/1). Dimaprit (1; 10 gmol/1) and 2-(2-thiazolyl)ethylamine (1; 10 ~imol/1) did also not affect the NMDA-evoked 3H overflow from slices (Fig. 2).

Effects of histamine receptor antagonists. The experi- ments described in this section were carried out in the presence of idazoxan plus desipramine and with NMDA 100;lmol/1 as a stimulus. The inhibitory effect of 10 ;lmol/1 histamine was abolished in the presence of 0.32 ~tmol/1, but not 0.032 timol/1, thioperamide and it remained unaffected in the presence of 1 limol/1 dimetindene or 320 ;imol/1 ranitidine (Fig. 3). Given alone, dimetindene, ranitidine or thioperamide at the concentrations indicated did not modify the NMDA- evoked tritium overflow (see legend to Fig. 3).

%- -~ 10o

oo "6

o

g 50

o

< O

z O

J i

!

0.01 0.1 1 1 10 R a M H SaMH

J 1 10 ThEA

115

1 10 p.mol/I Dimaprit

Fig. 2. Effects of histamine receptor agonists on the NMDA-evoked tri- tium overflow from rat brain cortex slices preincubated with [3H]nor- adrenaline and superfused with Mg2+-free salt solution. The superfu- sion buffer contained 1 ~mol/1 desipramine and 1 gmol/1 idazoxan. NMDA 100 ~mol/1 was added to the superfusion buffer for 2 rain after 40 min of superfusion. Each concentration of R-(-)-a-methylhistamine (RaMH," speckled columns), S-(+)-a-methylhistamine (SaMH; empty columns), 2-(2-thiazolyl)ethylamine (ThEA; hatched columns) or dimaprit (criss-crossed columns) was present from 20 min of superfu- sion onward until the end of the experiments. *P<0.01, compared to the corresponding control; means+SEM of 9-12 experiments. For ab- solute values of NMDA-evoked tritium overflow under control condi- tions, see Table 1

o ~

C

o

i

(23 ~K z

i00

o o ~6 50

Dimegndene Ranigd ine Thioperamide 1 /m~lA 320/~nol/1 0.032/mlOl/1 0.32 ,unlol/1

I

I

0 10

L j_

?

0 i0 0 i0

Histamine (!amol/I)

I i

0 i0 0 I0

Fig. 3. Effects of histamine receptor antagonists on the NMDA-evoked tritium overflow from rat brain cortex slices preincubated with [3H]noradrenaline and superfused with MgZ+-free salt solution. The superfusion buffer contained 1 pmol/1 desipramine and 1 ~lmol/1 idazoxan. NMDA I00 ~tmol/1 was added to the superfusion buffer for 2 rain after 40 rain of superfusion. Histamine, dimetindene, ranitidine or thioperamide were present from 20 rain of superfusion onward until the end of the experiments. *P< 0.01, compared to the corresponding controls; means+SEM of 6-12 experiments. In the control experi- ments in the absence of an antagonist or in the presence of dimetindene, ranitidine or thioperamide the NMDA-evoked tritium overflow above basal efflux was 8.64_+1.39nCi, 7.07_+1.31nCi, 9.06_+1,53nCi, 7.62+1.77nCi and 7.52_+2.09nCi, corresponding to 24.25_+2.80%; 18.60_+2.65%, 27.12+_ 1.61%, 23.02_+3.10% and 21.10-+2.20% of tissue tritium, respectively (no significant differences)

NMDA-evoked tritium overflow in synaptosomes

In synaptosomes, NMDA I retool/1 produced a much weaker stimulation of tritium overflow than did NMDA 30-300 ~tmol/1 in cortical slices (compare Table 1 to the

116

legend to Fig. 4). For reasons discussed below, desipra- mine and idazoxan could be omitted from the superfu- sion fluid.

Effects of histamine receptor ligands. Histamine ( 0 . l - 10 gmol/1; Fig. 4) inhibited the NMDA-evoked 3H overflow from rat brain cortex synaptosomes prein- cubated with [3H]noradrenaline concentration-depen- dently. R-(-)-a-Methylhistamine also produced an inhi- bition, but the degree was identical at 1 and 10 gmol/1 (a finding which may be related to limitations of evaluation due to the weak stimulatory effect of NMDA; Fig. 4). 2- (2-Thiazolyl)ethylamine (10 gmol/1; Fig. 4) and dimaprit (10 ~tmol/1; Fig. 4) did not influence the NMDA-evoked 3H overflow. The inhibitory effect of histamine was abolished in the presence of 0.32 gmol/1 thioperamide which given alone was ineffective (Fig. 4).

Experiments in synaptosomes were also carried out in the presence of co-conotoxin GVIA (co-CgTx). It was shown in a previous paper (Fink and GOthert 1993) that this drug applied at concentrations of 0.1 or 1 ~tmol/1 during stimulation with NMDA 1 mmol/1 caused an inhi- bition of overflow by about 55% indicating that co-CgTx 0.1 ~tmol/1 already caused the maximal effect. The NMDA (1 mmol)/1)-evoked tritium overflow in the pres- ence of e)-CgTx 0.1 gmol/1 was not inhibited by R-( - ) - a-methylhistamine 1 ~tmol/l: in the o)-CgTx controls (n = 15) the overflow amounted to 0.32+0.05% of tissue tritium and in the experiments with R-( - ) - a-methylhistamine (n = 15) it was 0.33+0.05% of tissue tritium.

g CI

C~

7

I00-

50-

2

! 0 0.1 I 10

Histamine

0 1

Thioperamide 0.32 ,u mol/l

2 L

0 0.I 1 I0 I0 I0 pmol/l

RaMH ThEA Dimaprit

Fig. 4. Effects of histamine receptor agonists and antagonists on the NMDA-evoked trit ium overflow from rat brain cortex synaptosomes preincubated with [3H]noradrenaline and superfused with Mg2+-free salt solution containing 100 gmol/1 glycine. NMDA 1 mmol/1 was add- ed to the superfusion buffer for 2 min after 40 min of superfusion. His- tamine, R-( - ) -a -methylh is tamine (RaMH), 2-(2-thiazolyl)ethylamine (ThEA), dimaprit and/or thioperamide were present from 20min of superfusion onward until the end of the experiments. *P<0.05 , com- pared to the corresponding controls; means + SEM of 5 - 7 experiments in quadruplicate. In the three series of control experiments, the NMDA- evoked tritium overflow above basal efflux ranged from 0.68 + 0.08 nCi to 0.89+_0.11 nCi, corresponding to 0.40_+0.06% to 0.55+_0.07°70 of synaptosomal trit ium

Discussion

The present study shows that both in slices and in synap- tosomes histamine inhibits NMDA-evoked release of nor- adrenaline. The involvement of H 3 receptors in this ef- fect is supported by the findings that the effect of hista- mine was mimicked by the selective H 3 receptor agonist R-(-)-a-methylhistamine and abolished by the selective H 3 receptor antagonist thioperamide, whereas agonists and/or antagonists at H 1- or H2-receptors were ineffec- tive. In slices, in which a more exact quantitative evalua- tion of NMDA-evoked noradrenaline release was possi- ble, R-(-)-a-methylhistamine was more potent than his- tamine, and the less active S-(+)-enantiomer of a- methylhistamine failed to affect NMDA-evoked nor- adrenaline release up to 10 gmol/1 in the present study. These results are in harmony with previous studies on H 3 receptors (Arrang et al. 1987; Schlicker et al. 1989, 1992b).

In most of the experiments on slices, desipramine was present in the superfusion fluid. Thus, the release of nor- adrenaline could be studied undisturbed by its neuronal reuptake. Furthermore, uptake of histamine into the noradrenergic axon terminals and displacement of [3H]noradrenaline from the preloaded storage sites could be avoided. In addition, the a2-adrenoceptor antagonist idazoxan was, as a rule, added to the superfusion fluid of slices, since the presence of this drug was a prerequisite for the disclosure of a histamine-induced inhibition of NMDA-evoked noradrenaline release. Similarly, previous experiments in which electrical stimuli were applied had revealed that the extent of inhibition obtainable with his- tamine was markedly increased by blockade of the a2- adrenoceptors (Schlicker et al. 1989, 1992a). These find- ings can be explained by the interactions between inhibi- tory presynaptic a2-adrenoceptors and H 3 receptors (see Schticker et al. 1992a). Obviously, this interaction plays a particularly important role when noradrenaline release is stimulated by NMDA. Desipramine plus idazoxan caused a 2 .5-2 .8 fold increase in noradrenaline release. However, the increase in release by itself was not the rea- son for the disclosure of the inhibitory effect of histamine since in the presence of the auxiliary drugs an inhibition was also obtained when NMDA-evoked noradrenaline re- lease was decreased to a lower level than in the absence of desipramine plus idazoxan by reducing the NMDA concentration from 100 to 30 gmol/1.

In the experiments on synaptosomes neither desipramine nor idazoxan was added to the superfusion fluid. Desipramine could be omitted since histamine did not affect basal tritium efflux. Idazoxan was not added since the amplifying action of a2-adrenoceptor blockade on the H 3 receptor-mediated effect only occurs if the a2-adrenoceptors are simultaneously activated by endo- genous noradrenaline (Schlicker et al. 1992 a). This prere- quisite is not fulfilled in superfused synaptosomes, in which the released neurotransmitters are flushed away so rapidly that they cannot act at the corresponding recep- tors on neighbouring synaptosomes (for discussion see Raiteri and Levi 1978; Starke et al. 1989). For reasons dis- cussed elsewhere, the maximum release for noradrenaline

obtainable with NMDA is much lower in synaptosomes than in slices (see Fink et al. 1990, 1992; Fink and GOthert 1992, 1993; Pittaluga and Raiteri 1990, 1992). The release values, expressed as °70 of tissue tritium and nCi, determined in the present study in both preparations are compatible with those previous findings.

The experiments on synaptosomes were mainly car- ried out in order to provide evidence that the NMDA and H 3 receptors mediating stimulation and inhibition of noradrenaline release, respectively, are at least in part lo- cated on the same single varicosities. The thioperamide- sensitive inhibition of NMDA-evoked release by H 3 re- ceptor agonists in synaptosomes (see above) indicates that a functional interaction between the respective recep- tors occurs at the level of single varicosities. Furthermore, the present results in synaptosomes conclusively prove the hypothesis (previously based on experiments on slices on- ly; Schlicker et al. 1989) that the H 3 heteroreceptors are located on the noradrenergic nerve terminals themselves.

Previous experiments on synaptosomes (Fink and G6thert 1993), revealed that o~-CgTx, a blocker of the N- type voltage-sensitive calcium channel (VSCC; Dooley et al. 1987; Sher and Clementi 1991), partially inhibits (by maximally about 55°70) NMDA-evoked noradrenaline re- lease, suggesting that the N-type VSCC is involved in cou- pling between NMDA receptor stimulation and noradren- aline release. Furthermore, when stimulation by NMDA was carried out in the presence of co-CgTx, the remaining noradrenaline release was not decreased by the a2-adre- noceptor agonist talipexol (Fink and G6thert 1993). Basi- cally the same finding was obtained here in experiments with R-(-)-a-methylhistamine. Accordingly, an unim- paired function of the N-type VSCC seems to be a prere- quisite for the inhibition of the NMDA-evoked noradren- aline release by stimulation not only of a2-adrenoceptors (Fink and GOthert 1993) but also of H 3 receptors. Con- versely, the inhibitory effect probably is ultimately due to the inhibition of a step in stimulus-release coupling in which Ca 2+ influx via the N-type VSCC plays a role.

The functional interaction between H 3 and NMDA receptors may be involved in an inhibitory histaminergic influence on NMDA receptor-mediated responses. Thus, the histaminergic neuronal system in the cerebral cortex may contribute to an inhibitory control of excitotoxic ef- fects of glutamatergic neurotransmission and may modu- late physiological functions such as neuronal plasticity in which the NMDA receptor system is involved.

Acknowledgements. We thank Mrs. H. Burisch and Mrs. I. Konrad for their skilled technical assistance and Professor Schunack (Berlin), Pro- fessor Schwartz (Paris) as well as the companies CIBA-GEIGY, Glaxo, Reckitt & Colman, SK & Beecham and Zyma for gifts of drugs. This study was supported by the Deutsche Forschungsgemeinschaft.

References

117

Arrang JM, Garbarg M, Lancelot JC, Lecomte JM, Pollard M, Robba M, Schunack W, Schwartz JC (1987) Highly potent and selective li- gands for histamine H 3 receptors. Nature 327:117-123

Dooley D J, Lupp A, Hertting G (1987) Inhibition of central neurotrans- mitter release by co-conotoxin GVIA, a peptide modulator of the N- type voltage-sensitive calcium channel. Naunyn-Schmiedebergs Arch Pharmacol 336:467-470

Fink K, GOthert M (1992) Presynaptic site of action underlying the eth- anol-induced inhibition of norepinephrine release evoked by stimu- lation of N-methyl-D-aspartate (NMDA) receptors in rat cerebral cortex. Brain Research 572:27-32

Fink K, G6thert M (1993) Modulation of N-methyl-D-aspartate (NMDA)-stimulated noradrenaline release in rat brain cortex by presynaptic a2-adrenoceptors. Naunyn-Schmiedebergs Arch Phar- macol 348:372-378

Fink K, G6thert M, Molderings G, Schlicker E (1989) N-Methyl-D- aspartate (NMDA) receptor-mediated stimulation of noradrenaline release, but not release of other neurotransmitters, in the rat brain cortex: receptor location, characterization and desensitization. Naunyn-Schmiedebergs Arch Pharmacol 339:514-521

Fink K, BOnisch H, G0thert M (1990) Presynaptic NMDA receptors stimulate noradrenaline release in the cerebral cortex. Eur J Phar- macol 185:115- 117

Fink K, Schlicker E, G6thert M (1991) Modulation of N-methyl-D- aspartate (NMDA)-stimulated noradrenaline release in rat brain cortex by presynaptic a2-adrenoceptors and histamine H 3 receptors. In Langer SZ, Galzin AM, Costentin J (eds) Presynaptic receptors and neuronal transporters. Advances in the biosciences, vol 82 Pergamon Press, Oxford New York Seoul Tokyo, pp 125-126

Fink K, Schultheif3 R, G6thert M (1992) Stimulation of noradrenaline release in human cerbral cortex mediated by N-methyl-D-aspartate (NMDA) and non-NMDA receptors. Br J Pharmacol 106:67-72

Gray EG, Whittaker VP (1962) The isolation of nerve endings from brain: an electron microscope study of cell fragments derived by ho- mogenization and centrifugation. J Anat (London) 96:79-88

Lowry OH, Rosebrough N J, Farr AK, Randall RJ (1951) Protein mea- surement with the folin phenol reagent. J Biol Chem 193:265-275

Pittaluga A, Raiteri M (1990) Release-enhancing glycine-dependent presynaptic NMDA receptors exist on noradrenergic terminals of hippocampus. Eur J Pharmacol 191:231-234

Pittaluga A, Raiteri M (1992) N-Methyl-D-aspartic acid (NMDA) and non-NMDA receptors regulating hippocampal norepinephrine re- lease. I. Location on axon terminals and pharmacological character- ization. J Pharmacol Exp Ther 260:232-237

Raiteri M, Levi G (1978) Release mechanisms for catecholamines and serotonin in synaptosomes. Rev Neurosci 3:77-130

Raiteri M, Angelini F, Levi G (1974) A simple apparatus for studying the release of neurotransmitters from synaptosomes. Eur J Phar- macol 25:411-414

Schlicker E, Fink K, Hinterthaner M, GOthert M (1989) Inhibition of noradrenaline release in the rat brain cortex via presynaptic H 3 re- ceptors. Naunyn-Schmiedebergs Arch Pharmacol 340:633-638

Schlicker E, Behling A, Lfimmen G, Malinowska B, GOthert M (i992 a) Mutual interaction of histamine H3-receptors and a2-adrenoceptors on noradrenergic terminals in mouse and rat brain cortex. Naunyn- Schmiedebergs Arch Pharmacol 345:639-646

Schlicker E, Behling A, Lammen G, G6thert M (1992b) Histamine H3A receptor-mediated inhibition of noradrenaline release in the mouse brain cortex. Naunyn-Schmiedebergs Arch Pharmacol 345:489 - 493

Sher E, Clementi F (199i) co-Conotoxin-sensitive voltage operated cal- cium channel in vertebrate cells. Neuroscience 42:301-307

Starke K, GOthert M, Kilbinger H (1989) Modulation of neurotransmit- ter release by presynaptic autoreceptors. Physiol Rev 69:864-989

White G, Lovinger DM, Peoples RW, Weight FF (1990) Inhibition of N- methyl-D-aspartate activated ion current by desmethylimipramine. Brain Res 537:337-339