Directory UMM :Data Elmu:jurnal:B:Brain Research:Vol882.Issue1-2.2000:

(1)

www.elsevier.com / locate / bres

Research report

Crucial role of kainate receptors in mediating striatal kainate

injection-induced decrease in acetylcholine M receptor binding

₁

in rat forebrain

a ,

_*

a b a

Shaoyu Jin

, Jian Yang , Wei Ling Lee , Peter T.H. Wong

Department of Pharmacology, Faculty of Medicine, The National University of Singapore, MD2, 18 Medical Drive, Singapore 117597, Singapore b

The National Neuroscience Institute of Singapore, Irrawaddy Block, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore Accepted 15 August 2000

Abstract

We investigated the roles of kainate-,a-amino-3-hydroxy-5-methylisoxazol-4-propionate (AMPA)- and N-methyl-D-aspartate (NMDA)-receptors in mediating striatal kainate injection-induced decrease in the binding of acetylcholine M (NMDA)-receptors in rat forebrain. After₁

3 3 3

unilateral intrastriatal injection of kainate (4 nmol), the bindings of [ H]kainate (10 nM), [ H]MK-801 (4 nM) and [ H]pirenzepine (4 nM) to the rat ipsilateral forebrain membranes declined, reaching the lowest on day 2 to 4 and recovering on day 8. Saturation binding studies, performed on day 2 post-injection, showed that kainate (1, 2, 4 nmol) dose-dependently decreased Bmax and K of the threed

ligands. (1)-5-Methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine (MK-801), a selective NMDA receptor channel blocker,

3 3

antagonised (from a dose of 4 nmol) kainate-induced decreases in the bindings of [ H]kainate (up to|20%), [ H]MK-801 (up to|90%)

and [ H]pirenzepine (up to |70%). In contrast, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), a selective non-NMDA receptor

antagonist, almost completely abolished (from a dose of 12 nmol) kainate-induced decreases in the bindings of all the three ligands (up to

|95–98%). Cyclothiazide, a selective potentiator that enhances AMPA receptor-mediated responses, significantly enhanced (from a dose

3 3 3

of 4 nmol) kainate-induced decrease in the binding of [ H]kainate but not that of [ H]pirenzepine or [ H]MK-801. In summary, these results indicate that striatal kainate injection-induced decrease in the binding of acetylcholine M receptors in rat forebrain is dependent₁ on activation of kainate receptors and, to a certain extent, a consequent involvement of NMDA receptors. These and previous studies provide some evidence showing that kainate receptors might play a crucial role in regulating excitatory amino acids (EAA)-modulated cholinergic neurotransmission in the central nervous system (CNS).  2000 Elsevier Science B.V. All rights reserved.

Theme: Neurotransmitters, modulators, transporters, and receptors

Topic: Excitatory amino acid receptors: physiology, pharmacology and modulation 3

Keywords: Cholinergic neurotransmission; CNQX; Cyclothiazide; Kainate injection; [ H]Pirenzepine binding

1. Introduction role in mediating excitatory amino acids (EAA)-induced

acetylcholine release in rat striatum in vitro [16], whereas a L-Glutamate, the principal excitatory neurotransmitter in demonstration of kainate receptor-mediated control of the mammalian central nervous system (CNS), mediates cholinergic neurotransmission in the CNS under in vivo most of the excitatory neurotransmission through activa- conditions is still lacking.

tion of three classes of ionotropic glutamate receptors Local injection of a glutamate analogue, such as kainate, [31,34], named after the agonists N-methyl-D-aspartate is widely used as a method for destroying neurones in vivo (NMDA), a-amino-3-hydroxy-5-methylisoxazol-4-prop- [34]. Previous studies have documented that local injection ionate (AMPA) and kainate [39]. Among the three sub- of kainate in rat brain induced massive neuronal loss and types, kainate receptors have been shown to play a distinct over-expression of amyloid precursor protein (APP) [7,12,21] as well as corresponding decreases in the expres-sions of choline acetyltransferase (ChAT) and glutamate

*Corresponding author. Tel.:165-874-6061; fax:165-773-0579.

E-mail address: [email protected] (S. Jin). decarboxylase [12,14] and in the bindings of ionotropic

(2)

glutamate- and acetylcholine muscarinic-receptors [15,25]. hypotonic buffer (1 mM EGTA buffered by Tris–HCl to However, the roles of NMDA-, AMPA- and kainate-re- pH 8.0) and left on ice for 30 min. The lysed membranes ceptors in these mechanisms are poorly understood. Gener- were collected by a 30-min centrifugation at 40 0003g,

ally, the effect of kainate is thought to be indirect through and then the lysis / centrifugation step was repeated one innervations of glutamatergic neurons [10]. It is the release more time. The membranes were then suspended and of glutamate [11,26] that leads to neuronal death through washed twice in a Tris–acetate buffer (100 mM, pH 7.4). activation of postsynaptic NMDA receptors [27,34]. These After each centrifugation, the membrane suspension was studies focusing on rat hippocampus suggested an excitat- sonicated with a tip sonicator at high intensity. The final ory role for presynaptic kainate receptors on glutamatergic pellet was suspended in the assay buffer (100 mM HEPES terminals [27]. On the other hand, conflicting observations containing 50 mM EGTA buffered to pH 7.4 by Tris– regarding the roles of presynaptic kainate- and postsynap- HCl). The membrane suspension was stored in aliquots at tic NMDA-receptors have also been reported [5,6,12]. 2808C.

We have recently observed that striatal kainate injection

induced a loss in cholinergic neuron and an increase in 2.3. Protein analysis APP expression in rat forebrain. These effects are

depen-dent on the activation of kainate receptors with a con- On the day of the binding assay, the frozen membrane sequent involvement of NMDA receptors [42]. In the preparations were thawed on ice and their protein contents present studies, we, therefore, investigated the roles of were determined by Bradford’s method [4]. Bovine serum kainate-, AMPA- and NMDA-receptors in mediating striat- albumin was used for the standard curve. The membrane al kainate injection-induced decreases in the bindings of preparations (protein 0.4 mg / ml) were kept on ice until

3 3 3

ionotropic glutamate- and acetylcholine M -receptors by1 used in [ H]kainate, [ H]MK-801 and [ H]pirenzepine 3

examining the characteristics of the specific [ H]kainate, binding assays.

3 3

[ H]MK-801 and [ H]pirenzepine binding to the

mem-branes prepared from the rat ipsilateral forebrain. 2.4. Radioligand binding assays

3 3 3

[ H]Kainate, [ H]MK-801 and [ H]pirenzepine bindings

2. Materials and methods were studied, using the filtration method [8,23,38]. All the

binding incubations were performed in disposable sterile

2.1. Animal treatment 96-well plates in a total volume of 300ml and terminated

by rapidly filtering the incubation medium through What-Male Sprague–Dawley rats (200–250 g) were housed man GF / B glass fiber filters under suction by means of a under controlled conditions with a 12-h day-night cycle Skatron cell harvester.

and with food and water available ad libitum. Rats were For [ H]kainate binding, aliquots of the membrane anesthetized with chloral hydrate (400 mg / kg, i.p.) and suspension (in a final concentration of 200 mg / ml) were

mounted in a stereotaxic apparatus (Stoelting, Wood Dale, incubated with [ H]kainate (10 nM) on ice for 1 h. IL, USA). One ml of phosphate buffer (4 mM, pH 7.4) Non-specific binding, determined by inclusion of 1 mM with kainate was slowly injected unilaterally into the L-glutamate, amounted to 10–12% of the total binding. For striatum (AP510.7 mm, ML563.0 mm, DV526.5 mm the saturation experiments, the membranes were incubated

from bregma), with 5 min allowed for local diffusion with 0.8–400 nM [ H]kainate. Non-specific binding before removing the microsyringe. The antagonists (1)-5- amounted to 10–40% of the total binding. The Skatron cell Methyl-10,11-dihydro-5H-dibenzo[a, d]cyclohepten-5,10- harvester was set at 111 which meant the filters were imine (MK-801) or 6-cyano-7-nitroquinoxaline-2,3-dione washed with three 1-ml aliquots of ice-cold 50 mM Tris– (CNQX) was injected 15 min before the kainate injection. HCl buffer, pH 7.4 (total filtering and washing time was

Cyclothiazide was coinjected with kainate. 3 s).

For [ H]MK-801 binding, aliquots of the membrane

2.2. Membrane preparation suspension (in a final concentration of 200mg / ml with 10

mM glutamate and 10 mM glycine) were incubated with 3

Rats were decapitated without prior stunning or anaes- [ H]MK-801 (4 nM) at 258C for 2.5 h. Non-specific thesia, and the forebrains were rapidly removed on ice. binding, determined by inclusion of 100 mM MK-801, Membrane preparations were prepared as previously de- amounted to 5–8% of the total binding. For the saturation scribed [22]. In brief, the ipsilateral forebrains were experiments, the membranes were incubated with 0.5–150

homogenized in ten volumes of ice-cold sucrose (300 mM) nM [ H]MK-801. Non-specific binding amounted to 5– containing 1 mM EGTA (pH 7.4) in a glass / Teflon 30% of the total binding. The Skatron cell harvester was homogenizer. The homogenate was centrifuged at 10003g set at 333 which meant the filters were washed with three for 10 min, and then the supernatant was recentrifuged at 3-ml aliquots of ice-cold 50 mM Tris–HCl buffer, pH 7.4 30 0003g for 20 min. The pellet was suspended in (total filtering and washing time was 9 s).

(3)

For [ H]pirenzepine binding, aliquots of the membrane from Research Biochemicals (RBI, Natick, MA, USA). All suspension (in a final concentration of 200mg / ml with 1 other chemicals were of analytical purity and of commer-mM KCl and 1 commer-mM MgCl ) were incubated with2 cial origin.

[ H]pirenzepine (4 nM) at 258C for 3 h. Non-specific binding, determined by inclusion of 1 mM atropine,

amounted to 6–8% of the total binding. For the saturation _{3. Results} experiments, the membranes were incubated with 0.5–100

3 3

nM [ H]pirenzepine. Non-specific binding amounted to _{3.1. Time-courses of [ H]kainate, [ H]MK-801 and} 3

6–45% of the total binding. The Skatron cell harvester was _{[ H]pirenzepine binding} set at 333 which meant the filters were washed with three

3-ml aliquots of ice-cold 50 mM Tris–HCl buffer, pH 7.4 _{After unilateral intrastriatal injection of kainate (4} 3

(total filtering and washing time was 9 s). _{nmol), the specific bindings of [ H]kainate (10 nM),}

3 3

[ H]MK-801 (4 nM) and [ H]pirenzepine (4 nM) to the

2.5. Scintillation counting membranes prepared from the ipsilateral striatum (results

not shown), cortex (results not shown) and forebrain (Fig. The filters were transferred to scintillation vials and 1) of rat clearly declined. Among them, the specific

3 3

soaked in 3 ml of BCS (Amersham, Arlington Heights, IL, bindings of [ H]kainate and [ H]MK-801 declined to the USA) scintillation cocktail. The vials were shaken and left lowest (about 33 and 19%, respectively) on day 2 and overnight at room temperature, and shaken again before recovered on day 8 post-injection (Fig. 1A and 1B),

being counted in a Beckman LS 3801 liquid scintillation whereas the specific binding of [ H]pirenzepine declined counter with an efficiency of 56% for tritium. to the lowest (about 38%) on day 4 and did not recover

until day 8 post-injection (Fig. 1C). 2.6. Calculation and statistical analysis

3 3

3.2. Saturation-curves of [ H]kainate, [ H]MK-801 and 3

Specific binding was calculated as the difference be- [ H]pirenzepine binding tween total and non-specific binding and expressed on a

protein basis (pmol / mg protein). Receptor density (Bmax) Saturation binding experiments were performed on day and dissociation constant (K ) values were determinedd 2 after unilateral intrastriatal injections of kainate (1, 2, 4 from saturation binding data by a computerized non-linear nmol). Kainate, in a dose-dependent manner, declined

3 3

regression analysis (GRAPHPAT PRISM 2.0, GraphPat Soft- saturation-curves of the specific [ H]kainate, [ H]MK-801

ware, San Diego, CA, USA). and [ H]pirenzepine binding to the rat ipsilateral forebrain

Scatchard analysis of the data was not used for the membranes (Fig. 2(1)). Their Kd and Bmax values are determination of binding parameters (Kd and Bmax), but shown in Table 1.

plotted primarily to visualize the different binding models, The Scatchard plot of the saturation binding of 3

curvilinearity for multiple binding sites and linearity for a [ H]kainate was curvilinear, indicating more than one single binding site. The best-fit model was determined population of receptor binding sites [Fig. 2(2A)]. F-test

using the F-test. indicated that a two-site fit of the data was significantly

Total binding and non-specific binding for each sample better than a one-site fit (P,0.05). In contrast, the 3 were determined in triplicate. All data were the results of Scatchard plot of the saturation binding data of [

H]MK-3

at least three independent experiments using at least three 801 [Fig. 2(2B)] or [ H]pirenzepine [Fig. 2(2C)] was rats per experiment and given as means6S.E.M. One-way linear, indicating only one apparent population of receptor analysis of variance (ANOVA) was employed followed by binding site. F-test indicated that a two-site fit of the data Bonferroni post-hoc analysis for multiple comparisons. A was not significantly better than a one-site fit (P.0.05). probability level of,0.05 was considered significant.

3.3. Differential antagonism by MK-801 and CNQX of

3 3

2.7. Chemicals kainate-induced decreases in[ H]kainate, [ H]MK-801

and[ H]pirenzepine binding 3

[ H-Vinylidene]kainate (specific activity: 58 Ci / mmol), 3

[ H](1)MK-801 (specific activity: 22.5 Ci / mmol), and MK-801 and CNQX dose-dependently antagonised kain-3

ate (4 nmol)-induced decreases in the specific bindings of [ H]pirenzepine (specific activity: 85.6 Ci / mmol) were

3 3

[ H]kainate (10 nM), [ H]MK-801 (4 nM) and purchased from NEN Life Science Products (Boston, MA,

USA). Glycine, kainic acid, L-glutamic acid and HEPES [ H]pirenzepine (4 nM) to the rat ipsilateral forebrain (N-[2-hydroxyethyl] piperazine-N9-[2-ethanesulfonic acid]) membranes (Fig. 3). The antagonism by MK-801, from a were purchased from Sigma Chemical (St. Louis, MO, dose of 4 nmol, reached a maximum of only about 20% for

(4)

3 3 3

Fig. 1. Time-courses of the bindings of [ H]kainate (10 nM, A), [ H]MK-801 (4 nM, B) and [ H]pirenzepine (4 nM, C) to the rat ipsilateral forebrain membranes after kainate (4 nmol) was unilaterally injected into the striatum. Each point is the mean6S.E.M of nine to 14 observations from at least three independent experiments, using at least three rats per experiment. A significant difference (by one-way ANOVA with Bonferroni) vs. corresponding uninjected controls is represented by: * P,0.05, ** P,0.01, *** P,0.001.

(5)

3 3 3

Fig. 2. (1) Saturation-curves of the bindings of [ H] kainate (0.8–400 nM, A), [ H]MK-801 (0.5–150 nM, B) and [ H]pirenzepine (0.5–100 nM, C) to the rat ipsilateral forebrain membranes. Saturation binding assays were performed on day 2 after kainate (1, 2, 4 nmol) was unilaterally injected into the striatum. Control curves come from uninjected rats. Their K and Bd maxare shown in Table 1. Each point is the mean6S.E.M of nine to 12 observations from at least three independent experiments, using at least three rats per experiment. In some cases, the low variability resulted in error bars that were too small to plot.

(6)

specific [ H]MK-801 binding and about 70% for the 3

specific [ H]pirenzepine binding. On the other hand, CNQX at a dose of 12 nmol almost completely abolished kainate (4 nmol)-induced decreases in the specific bindings of all the three ligands (about 95–98%).

3.4. Selective potentiation by cyclothiazide of

kainate-3 3

induced decreases in[ H]kainate, [ H]MK-801 and 3

[ H]pirenzepine binding

Cyclothiazide, in a dose-dependent manner, significantly enhanced kainate (4 nmol)-induced decrease in the specific

3 3

binding of [ H]kainate (10 nM) but not that of [ H]MK-3

801 (4 nM) or [ H]pirenzepine (4 nM) to the rat ipsilateral forebrain membranes (Fig. 4). The potentiation by cyclothiazide, from a dose of 4 nmol, reached a maximum

of about 15% for the specific [ H]kainate binding.

4. Discussion

4.1. Kainate-induced decrease in acetylcholine M1 receptor binding

The present studies demonstrated that kainate injected unilaterally into the striatum dose-dependently decreased the bindings of ionotropic glutamate- and acetylcholine M -receptors in the ipsilateral forebrain of rat, particularly1 on day 2 to 4 post-injection. These results are in agreement with previous observations, using immunohistochemistry, histochemistry, histofluorescence, and electron microscopic techniques [7,42], that on day 2 after intrastriatal injection of kainate, there were a massive neuronal loss and a significant reduction of ChAT expression in the ipsilateral forebrain of rat. Considered together, these findings sug-gest that the decrease in the binding of acetylcholine M1 receptors in the present experimental model, at least to a certain extent, reflects a loss of cholinergic neurons [28,40].

4.2. Roles of NMDA-, AMPA- and kainate-receptors Saturation binding experiments of radiolabeled kainate, MK-801 and pirenzepine were performed to mainly assess kainate-, NMDA- and acetylcholine M -receptor binding1 sites. The selective NMDA receptor channel blocker, MK-801, and the selective non-NMDA receptor antagonist, CNQX, were used to distinguish between NMDA- and non NMDA-receptor-mediated responses. The present results showed that CNQX and MK-801 differentially antagonised

Fig. 2. (2) Scatchard plots of the saturation bindings of [ H] kainate the kainate-induced decrease in acetylcholine M receptor1

3 3

(0.8–400 nM, A), [ H]MK-801 (0.5–150 nM, B) and [ H]pirenzepine _{binding. We have further extended this observation to} (0.5–100 nM, C) to the rat ipsilateral forebrain membranes. Data come

include the kainate-induced neuronal loss and decrease in

from controls in Fig. 2(1). The Scatchard analysis of data was not used

ChAT expression [42]. Cyclothiazide is a selective

poten-for any determinations of binding parameters (K and Bd max), but plotted

tiator that enhances AMPA receptor-mediated responses

(7)

Table 1

3 3 3

K (nM) and Bd max(pmol / mg protein) of the bindings of [ H]kainate (0.8–400 nM), [ H]MK-801 (0.5–150 nM) and [ H]pirenzepine (0.5–100 nM) to the a

rat ipsilateral forebrain membranes

3 3 3

Injection [ H]Kainate binding [ H]MK-801 binding [ H]Pirenzepine binding

Kd Bmax Kd Bmax Kd Bmax

Site 1 Site 2 Site 1 Site 2

None 5.5260.23 49.0461.37 0.4760.02 1.7160.02 23.3161.07 3.8060.05 8.8860.83 3.1060.09 Buffer 5.4960.25 47.2462.22 0.5060.02 1.8060.07 19.1261.28 3.7360.07 9.5361.00 3.1960.10

b c b

KA 1 nmol 5.6560.31 45.6561.52 0.6060.05 1.7860.05 18.1361.02 3.4560.06 7.6061.03 2.6260.10

c b c d d c

KA 2 nmol 5.3560.28 40.2361.15 0.3260.06 1.4760.03 17.6260.88 3.2960.05 6.6960.87 2.4660.09

b d d d d d d

KA 4 nmol 3.5360.21 34.3160.86 0.2260.02 1.2660.01 16.4060.48 2.9460.02 6.3560.83 2.0360.08 a

Saturation assays were performed on day 2 after kainate (1, 2, 4 nmol) was unilaterally injected into rat striatum. Data are expressed as the mean6S.E.M of nine to 12 observations from at least three independent experiments, using at least three rats per experiment.

A significant difference P,0.05 (by one-way ANOVA with Bonferroni) vs. corresponding uninjected control. c

A significant difference P,0.01 (by one-way ANOVA with Bonferroni) vs. corresponding uninjected control. d

A significant difference P,0.001 (by one-way ANOVA with Bonferroni) vs. corresponding uninjected control.

[2,16]. In the present studies, cyclothiazide was used to bral cortex and hippocampus have established the localiza-differentiate the involvement of the AMPA- and kainate- tion of AMPA receptors at postsynaptic densities [33] as receptors since CNQX was able to almost completely well as at the developed dendrites [9,31]. However, an block the kainate-induced decreases in all three bindings. observation using autoradiography indicated the localiza-The decrease in the binding of radiolabeled kainate in- tion of AMPA receptors on presynaptic endings of dopa-duced by cyclothiazide shows the potentiation of AMPA mine fibres [41], which is supported by our previous receptor-mediated responses, which is in agreement with studies regarding dopamine release from the presynaptic previous findings [22]. However, cyclothiazide did not endings of dopamine fibres [16,17] and by a previous enhance kainate-induced decrease in the binding of acetyl- finding that dopamine positively mediated only GluR1

choline M receptors.1 gene expression [1]. To date, there is no evidence showing

These results indicate that the activation of NMDA the presence of excitatory presynaptic AMPA receptors on receptors accounts for a substantial fraction of kainate- cholinergic and glutamatergic neurons.

induced decrease in the binding of acetylcholine M1 More importantly, these results demonstrate a key role receptors in rat forebrain, which support most of the of kainate receptors in mediating kainate-induced decrease previous studies regarding the involvement of NMDA in the binding of acetylcholine M receptors in rat fore-1 receptors in kainate-induced brain injury [11,26,27,42]. brain, consistent with our other findings regarding the Radioligand binding and immunohistochemical studies distinct role of kainate receptors in mediating kainate-have shown that NMDA receptors are distributed through- induced acetylcholine release in rat striatum [16] and out the rat brain but predominantly within the forebrain, cholinergic neuron loss in rat forebrain [42]. Immuno-and mainly localized at postsynaptic densities Immuno-and the cytochemical studies, using antibodies against kainate associated dendrites of most neurons [31], including receptor subunits, have clearly shown the presence of cholinergic neurons [17]. The activation of postsynaptic kainate receptors at both the presynaptic and postsynaptic NMDA receptors on target cells may trigger cytotoxic densities in the rat brain [32,35]. In the present experimen-actions by opening calcium channels [3] and evoke exces- tal model, the involvement of kainate and NMDA re-sive release of neurotransmitters including acetylcholine by ceptors suggest that both the pre- and post-synaptic kainate depolarising cell membranes [17,18]. Both the cytotoxic receptors might participate in mediating the effects of actions and acetylcholine release induced by excessive kainate. The mechanism underlying the link between the activation of postsynaptic NMDA receptors might directly activation of kainate receptors and the involvement of contribute to the effects of NMDA receptors in the present NMDA receptors is a controversial issue. It might be

experimental model. attributed to a presynaptic excitation on glutamatergic

In addition, these results show that the potentiation of neurons [27], a presynaptic disinhibitory effect on AMPA receptors by cyclothiazide has no significant effects GABAergic neurons [6], or a postsynaptic excitatory on kainate-induced decrease in the binding of acetylcholine action on target cells [5] through kainate receptors. In the M1 receptors in rat forebrain, in agreement with our latter case, it might be necessary to assume the presence of previous observation that AMPA receptors apparently did a direct functional interaction between the postsynaptic not participate in mediating kainate-induced acetylcholine kainate- and NMDA-receptors on the same target cells, release in the rat striatum [16]. Immunolabelings with like that between adenosine A2A- and dopamine D -re-2 anti-GluR1, anti-GluR2 / 3 and anti-GluR4 in the rat cere- ceptors [19,30].

(8)

Fig. 3. Effects of increasing doses of MK-801 (filled symbols) or CNQX (unfilled symbols) on kainate (4 nmol)-induced decreases in the bindings of

3 3 3

[ H]kainate (10 nM, A), [ H]MK-801 (4 nM, B) and [ H]pirenzepine (4 nM, C) to the rat ipsilateral forebrain membranes. Binding assays were performed on day 2 after kainate (4 nmol) was unilaterally injected into rat striatum. MK-801 or CNQX was injected into the ipsilateral striatum 15 min before the kainate injection. Responses are expressed as a percentage of kainate-induced decreases in the bindings of the three ligands. Each point is the mean6S.E.M of nine to 18 observations from at least three independent experiments, using at least three rats per experiment. A significant difference (by one-way ANOVA with Bonferroni) vs. corresponding control is represented by: * P,0.05, ** P,0.01, *** P,0.001. In some cases, the low variability resulted in error bars that were too small to plot.

(9)

3 3

Fig. 4. Effects of increasing doses of cyclothiazide on kainate (4 nmol)-induced decreases in the bindings of [ H]kainate (10 nM), [ H]MK-801 (4 nM) 3

and [ H]pirenzepine (4 nM) to the rat ipsilateral forebrain membranes. Binding assays were performed on day 2 after kainate (4 nmol) with or without cyclothiazide was unilaterally injected into rat striatum. Each bar is the mean6S.E.M of nine to 16 observations from at least three independent experiments, using at least three rats per experiment. A significant difference (by one-way ANOVA with Bonferroni) vs. corresponding control (kainate 4 nmol) is represented by: * P,0.05, ** P,0.01, *** P,0.001.

4.3. Kainate receptors and the cholinergic kainate receptors could not be ruled out because of the lack

neurotransmission in the CNS of high selectivity of NBQX and CNQX on AMPA- and

kainate-receptors.

The present receptor binding results are consistent with Loss of cholinergic neurons in the basal forebrain is those obtained in neurotransmitter release [16] and im- believed to account for much of the learning and memory munohistochemistry [42] studies. These results indicate deficits in Alzheimer’s disease. It was reported that that kainate receptors are predominantly involved in the domoate, a glutamate analogue produced by mussels, EAA-induced increase in acetylcholine release, decrease in caused a syndrome with features that resemble those of M receptor binding and ChAT immunoreactivity, increase1 Alzheimer’s disease in a group of Newfoundlanders after in APP immunoreactivity, and neuronal loss. They might consuming domoate-contaminated mussels in 1987 suggest a crucial role for kainate receptors in regulating [29,36,37]. Domoate has been established to be a relatively EAA-modulated cholinergic neurotransmission in the CNS selective agonist at kainate receptors although it can still as well as a similar role for the receptors in the CNS act on AMPA receptors [20]. The present studies may help diseases associated with cholinergic disorders e.g. Al- to understand the mechanism of neurodegeneration under-zheimer’s disease on the basis of the cholinergic hypoth- lying the domoate-induced syndrome.

esis of this neurodegenerative disease [13,24].

The main functions ascribed to cholinergic pathways are

related to arousal and learning, and motor control. Block- 5. Conclusion ade of glutamatergic transmission by CNQX and NBQX

(another antagonist at non-NMDA receptors) has been In summary, the present studies demonstrated that shown to produce cognitive deficits, which are attributed to striatal kainate injection decreased the binding of acetyl-blockade of AMPA receptors [3,43]. An involvement of choline M1 receptors in rat forebrain. This decrease is

(10)

[13] P.T. Francis, A.M. Palmer, M. Snape, G.K. Wilcock, The cholinergic

dependent on the activation of kainate receptors and, to a

hypothesis of Alzheimer’s disease: a review of progress, J. Neurol.

certain extent, a consequent involvement of NMDA

re-Neurosur. Psychiat. 66 (1999) 137–147.

ceptors. These results support the possibility that both the _{[14] B.H. Guevara, I.S. Hoffmann, L.X. Cubeddu, Ventral and dorsal} pre- and post-synaptic kainate receptors participate in striatal cholinergic neurons have different sensitivities to kainic acid,

mediating the effects of kainate in the present experimental Neurochem. Int. 31 (1997) 723–730.

[15] R.E. Hruska, R. Schwarcz, J.T. Coyle, H.I. Yamamura, Alterations

model. These and previous findings provide some evidence

of muscarinic cholinergic receptors in the rat striatum after kainic

that kainate receptors may play a crucial role in regulating

acid injections, Brain Res. 152 (1978) 620–625.

EAA-modulated cholinergic neurotransmission in the _{[16] S. Jin, AMPA- and kainate-receptors differentially mediate}

excitat-CNS. ory amino acid-induced dopamine and acetylcholine release from rat

striatal slices, Neuropharmacology 36 (1997) 1503–1510. [17] S. Jin, B.B. Fredholm, Role of NMDA, AMPA and kainate receptors

in mediating glutamate- and 4-AP-induced dopamine and

acetyl-Acknowledgements choline release from rat striatal slices, Neuropharmacology 331

(1994) 1039–1048.

[18] S. Jin, B.B. Fredholm, Glucose deprivation increases basal and

This work was supported by grants (3981319 / 3981320)

electrically evoked transmitter release from rat striatal slices. Role of

to PTHW from The National University of Singapore.

NMDA- and adenosine A receptors, Eur. J. Pharmacol. 340 (1997)1 169–175.

[19] S. Jin, B. Johansson, B.B. Fredholm, Effects of adenosine A and1 A receptor activation on electrically evoked dopamine and acetyl-2

References _{choline release from rat striatal slices, J. Pharmacol. Exp. Ther. 267}

(1993) 801–808.

[20] T.H. Johansen, J. Drejer, F. Watjen, E. Nielsen, A novel non-NMDA [1] V. Bernard, A. Gardiol, B. Faucheux, B. Bloch, Y. Agid, E.C.

receptor antagonist shows selective displacement of low-affinity Hirsch, Expression of glutamate receptors in the human and rat basal

[ H] kainate binding, Eur. J. Pharmacol. 246 (1993) 195–204. ganglia: effect of the dopaminergic denervation on AMPA receptor

[21] T. Kawarabayashi, M. Shoji, Y. Harigaya, H. Yamaguchi, S. Hirai, gene expression in the striatopallidal complex in Parkinson’s disease

Expression of APP in the early stage of brain damage, Brain Res. and rat with 6-OHDA lesion, J. Compar. Neurol. 368 (1996)

563 (1991) 334–338. 553–568.

[22] M. Kessler, A. Arai, A. Quan, G. Lynch, Effect of cyclothiazide on [2] B. Bettler, C. Mulle, Review: Neurotransmitter receptors II AMPA

binding properties of AMPA-type glutamate receptors: lack of and kainate receptors, Neuropharmacology 34 (1995) 123–139.

competition between cyclothiazide and GYKI 52466, Mol. Phar-[3] D. Bleakman, D. Lodge, Neuropharmacology of AMPA and kainate

macol. 49 (1996) 123–131. receptors, Neuropharmacology 37 (1998) 1187–1204.

[23] V.C. Kurschner, R.L. Petruzzi, G.T. Golden, W.H. Berrettini, T.N. [4] M.M. Bradford, A rapid and sensitive method for quantitation of

Ferraro, Kainate and AMPA receptor binding in seizure-prone and microgram quantities of protein utilizing the principle of protein-dye

seizure-resistant inbred mouse strains, Brain Res. 780 (1998) 1–8. binding, Anal. Biochem. 72 (1976) 248–254.

[24] J. Lemiere, D. Van Gool, R. Dom, Treatment of Alzheimer’s disease: [5] R. Chittajallu, S.P. Braithwaite, V.R. Clarke, J.M. Henley, Kainate

an evaluation of the cholinergic approach, Acta Neurologica Belgica receptors: subunits, synaptic localization and function, Trends

99 (1999) 96–106. Pharmacol. Sci. 20 (1999) 26–35.

[25] E.D. London, J.T. Coyle, Specific binding of [3H]kainic acid to [6] V.R. Clarke, B.A. Ballyk, K.H. Hoo, A. Mandelzys, A. Pellizzari,

receptor sites in rat brain, Molecular Pharmacol. 15 (1978) 492– C.P. Bath et al., A hippocampal GluR5 kainate receptor regulating

505. inhibitory synaptic transmission, Nature 389 (1997) 599–603.

[26] J.O. Malva, A.P. Carvalho, C.M. Carvalho, Domoic acid induces the [7] J.T. Coyle, M.E. Molliver, M.J. Kuhar, In situ injection of kainic

release of glutamate in the rat hippocampal CA3 subregion, Neuro-acid: a new method for selectively lesioning neural cell bodies while

report 7 (1996) 1330–1334. sparing axons of passage, J. Compar. Neurol. 180 (1978) 301–323.

[27] J.O. Malva, A.P. Carvalho, C.M. Carvalho, Kainate receptors in [8] R.F. Cowburn, B. Wiehager, E. Trief, M. Li-Li, E. Sundstrom,

hippocampal CA3 subregion: evidence for a role in regulating Effects of beta-amyloid-(25–35) peptides on radioligand binding to

neurotransmitter release, Neurochem. Int. 32 (1998) 1–6. excitatory amino acid receptors and voltage-dependent calcium

[28] M.M. Nicolle, A. Shivers, T.M. Gill, M. Gallagher, Hippocampal channels: evidence for a selective affinity for the glutamate and

N-methyl-D-aspartate and kainate binding in response to entorhinal glycine recognition sites of the NMDA receptor, Neurochem. Res.

cortex aspiration or 192 IgG-saporin lesions of the basal forebrain, 22 (1997) 1437–1442.

Neuroscience 77 (1997) 649–659. [9] A.M. Craig, C.D. Blackstone, R.L. Huganir, G. Banker, The

[29] J.W. Olney, Excitotoxicity: an overview, Can. Dise. Week. Rep. 16 distribution of glutamate receptors in cultured rat hippocampal

(1990) 47–57. neurons: postsynaptic clustering of AMPA-selective subunits,

Neu-[30] E. Ongini, B.B. Fredholm, Pharmacology of adenosine A2A re-ron 10 (1993) 1055–1068.

ceptors, Trends Pharmacol. Sci. 17 (1996) 364–372. [10] J.L. Gaiarsa, L. Zagrean, Y. Ben-Ari, Neonatal irradiation prevents

[31] S. Ozawa, H. Kamiya, K. Tsuzuki, R.J. Wenthold, Glutamate the formation of hippocampal mossy fibers and the epileptic action

receptors in the mammalian central nervous system, Prog. Neuro-of kainate on rat CA3 pyramidal neurons, J. Neurophysiol. 71

biol. 54 (1998) 581–618. (1994) 204–215.

[32] R.S. Petralia, Y.X. Wang, Histological and ultrastructural localization [11] R.L. Gannon, D.M. Terrian, Presynaptic modulation of glutamate

of the kainate receptor subunits, KA2 and GluR6 / 7, in the rat and dynorphin release by excitatory amino acids in the guinea-pig

nervous system using selective antipeptide antibodies, J. Compar. hippocampus, Neuroscience 41 (1991) 401–410.

Neurol. 349 (1994) 85–110. [12] A.C. Foster, R. Gill, G.N. Woodruff, Neuroprotective effects of

[33] R.S. Petralia, R.J. Wenthold, Light and electron immunocytochemi-MK-801 in vivo: selectivity and evidence for delayed degeneration

cal localization of AMPA-selective glutamate receptors in the rat mediated by NMDA receptor activation, J. Neurosci. 8 (1988)

brain, J. Compar. Neurol. 318 (1992) 329–354. 4745–4754.

(11)

[34] H.P. Rang, M.M. Dale, J.M. Ritter, Pharmacology, in: The Central mRNA in cerebral cortex in response to kainate lesion of nucleus Nervous System, 3rd Edition, Churchill Livingstone, Edinburgh, basalis: involvement of NMDA receptors, Neurosci. Lett. 111

1995, pp. 519–521, Section 4. (1990) 176–182.

[35] A. Represa, E. Tremblay, Y. Ben-Ari, Kainate binding sites in the [41] K. Zavitsanou, A. Mitsacos, P. Giompres, E.D. Kouvelas, Changes hippocampal mossy fibers: localization and plasticity, Neuroscience of [3H]AMPA and [3H]kainate binding in rat caudate-putamen and 20 (1987) 739–748. nucleus accumbens after 6-hydroxydopamine lesions of the medial [36] G.R. Stewart, C.F. Zorumski, M.T. Price, J.W. Olney, Domoic acid: forebrain bundle: an autoradiographic study, Brain Res. 731 (1996)

a dementia inducing excitotoxic food poison with kainic acid 132–140.

receptor specificity, Exp. Neurol. 110 (1990) 127–138. [42] X. Zhu, S. Jin, Y.K. Ng, P.T.H. Wong, Positive and negative [37] R.J. Sutherland, J.M. Hoesing, I.Q. Whishaw, Domoic acid, an modulation by AMPA- and kainate-receptors of striatal kainate environmental toxin, produces hippocampal damage and severe injection-induced lesions in rat forebrain, Brain Res. (2000) sub-memory impairment, Neurosci. Lett. 120 (1990) 221–223. mitted.

[38] J.X. Wang, L. Mei, H.I. Yamamura, W.R. Roeske, Solubilization with [43] I. Zivkovic, D.M. Thompson, M. Bertolino, D. Uzunov, M. DiBella, digitonin alters the kinetics of pirenzepine binding to muscarinic E. Costa et al., 7-Chloro-3-methyl-3-4-dihydro-2H-1, 2,4 ben-receptors from rat forebrain and heart, J. Pharmacol. Exp. Ther. 242 zothiadiazine S,S-dioxide (IDRA 21): a benzothiadiazine derivative (1987) 981–990. that enhances cognition by attenuatingDL -alpha-amino-2,3-dihydro-[39] J.C. Watkins, R.H. Evans, Excitatory amino acid transmitters, Ann. 5-methyl-3-oxo-4-isoxazolepropanoic acid (AMPA) receptor

desen-Rev. Pharmacol. Toxicol. 21 (1981) 165–204. sitisation, J. Pharmacol. Exp. Ther. 272 (1995) 300–309. [40] H. Wood, J. de Belleroche, Induction of ornithine decarboxylase

(1)

specific [ H]MK-801 binding and about 70% for the

specific [ H]pirenzepine binding. On the other hand,

CNQX at a dose of 12 nmol almost completely abolished

kainate (4 nmol)-induced decreases in the specific bindings

of all the three ligands (about 95–98%).

3.4. Selective potentiation by cyclothiazide of

kainate-3 3

induced decreases in

[ H]kainate, [ H]MK-801 and

[ H]pirenzepine binding

Cyclothiazide, in a dose-dependent manner, significantly

enhanced kainate (4 nmol)-induced decrease in the specific

3 3

binding of [ H]kainate (10 nM) but not that of [

H]MK-3

801 (4 nM) or [ H]pirenzepine (4 nM) to the rat ipsilateral

forebrain membranes (Fig. 4). The potentiation by

cyclothiazide, from a dose of 4 nmol, reached a maximum

of about 15% for the specific [ H]kainate binding.

4. Discussion

4.1. Kainate-induced decrease in acetylcholine M

receptor binding

The present studies demonstrated that kainate injected

unilaterally into the striatum dose-dependently decreased

the bindings of ionotropic glutamate- and acetylcholine

M -receptors in the ipsilateral forebrain of rat, particularly

on day 2 to 4 post-injection. These results are in agreement

with previous observations, using immunohistochemistry,

histochemistry, histofluorescence, and electron microscopic

techniques [7,42], that on day 2 after intrastriatal injection

of kainate, there were a massive neuronal loss and a

significant reduction of ChAT expression in the ipsilateral

forebrain of rat. Considered together, these findings

sug-gest that the decrease in the binding of acetylcholine M

receptors in the present experimental model, at least to a

certain extent, reflects a loss of cholinergic neurons

[28,40].

4.2. Roles of NMDA-, AMPA- and kainate-receptors

Saturation binding experiments of radiolabeled kainate,

MK-801 and pirenzepine were performed to mainly assess

kainate-, NMDA- and acetylcholine M -receptor binding

sites. The selective NMDA receptor channel blocker,

MK-801, and the selective non-NMDA receptor antagonist,

CNQX, were used to distinguish between NMDA- and non

NMDA-receptor-mediated responses. The present results

showed that CNQX and MK-801 differentially antagonised

Fig. 2. (2) Scatchard plots of the saturation bindings of [ H] kainate

the kainate-induced decrease in acetylcholine M receptor

3 3

(0.8–400 nM, A), [ H]MK-801 (0.5–150 nM, B) and [ H]pirenzepine

_{binding. We have further extended this observation to}

(0.5–100 nM, C) to the rat ipsilateral forebrain membranes. Data come

include the kainate-induced neuronal loss and decrease in

from controls in Fig. 2(1). The Scatchard analysis of data was not used

ChAT expression [42]. Cyclothiazide is a selective

poten-for any determinations of binding parameters (K and Bd max), but plotted

tiator that enhances AMPA receptor-mediated responses

(2)

Kd Bmax Kd Bmax Kd Bmax

Site 1 Site 2 Site 1 Site 2

None 5.5260.23 49.0461.37 0.4760.02 1.7160.02 23.3161.07 3.8060.05 8.8860.83 3.1060.09 Buffer 5.4960.25 47.2462.22 0.5060.02 1.8060.07 19.1261.28 3.7360.07 9.5361.00 3.1960.10

b c b

KA 1 nmol 5.6560.31 45.6561.52 0.6060.05 1.7860.05 18.1361.02 3.4560.06 7.6061.03 2.6260.10

c b c d d c

KA 2 nmol 5.3560.28 40.2361.15 0.3260.06 1.4760.03 17.6260.88 3.2960.05 6.6960.87 2.4660.09

b d d d d d d

KA 4 nmol 3.5360.21 34.3160.86 0.2260.02 1.2660.01 16.4060.48 2.9460.02 6.3560.83 2.0360.08

A significant difference P,0.05 (by one-way ANOVA with Bonferroni) vs. corresponding uninjected control.

A significant difference P,0.01 (by one-way ANOVA with Bonferroni) vs. corresponding uninjected control.

A significant difference P,0.001 (by one-way ANOVA with Bonferroni) vs. corresponding uninjected control.

[2,16]. In the present studies, cyclothiazide was used to

bral cortex and hippocampus have established the

localiza-differentiate the involvement of the AMPA- and kainate-

tion of AMPA receptors at postsynaptic densities [33] as

receptors since CNQX was able to almost completely

well as at the developed dendrites [9,31]. However, an

block the kainate-induced decreases in all three bindings.

observation using autoradiography indicated the

localiza-The decrease in the binding of radiolabeled kainate in-

tion of AMPA receptors on presynaptic endings of

dopa-duced by cyclothiazide shows the potentiation of AMPA

mine fibres [41], which is supported by our previous

receptor-mediated responses, which is in agreement with

studies regarding dopamine release from the presynaptic

previous findings [22]. However, cyclothiazide did not

endings of dopamine fibres [16,17] and by a previous

enhance kainate-induced decrease in the binding of acetyl-

finding that dopamine positively mediated only GluR1

choline M receptors.

gene expression [1]. To date, there is no evidence showing

These results indicate that the activation of NMDA

the presence of excitatory presynaptic AMPA receptors on

receptors accounts for a substantial fraction of kainate-

cholinergic and glutamatergic neurons.

induced decrease in the binding of acetylcholine M

More importantly, these results demonstrate a key role

receptors in rat forebrain, which support most of the

of kainate receptors in mediating kainate-induced decrease

previous studies regarding the involvement of NMDA

in the binding of acetylcholine M receptors in rat fore-

receptors in kainate-induced brain injury [11,26,27,42].

brain, consistent with our other findings regarding the

Radioligand binding and immunohistochemical studies

distinct role of kainate receptors in mediating

kainate-have shown that NMDA receptors are distributed through-

induced acetylcholine release in rat striatum [16] and

out the rat brain but predominantly within the forebrain,

cholinergic neuron loss in rat forebrain [42].

Immuno-and mainly localized at postsynaptic densities Immuno-and the

cytochemical studies, using antibodies against kainate

associated dendrites of most neurons [31], including

receptor subunits, have clearly shown the presence of

cholinergic neurons [17]. The activation of postsynaptic

kainate receptors at both the presynaptic and postsynaptic

NMDA receptors on target cells may trigger cytotoxic

densities in the rat brain [32,35]. In the present

experimen-actions by opening calcium channels [3] and evoke exces-

tal model, the involvement of kainate and NMDA

re-sive release of neurotransmitters including acetylcholine by

ceptors suggest that both the pre- and post-synaptic kainate

depolarising cell membranes [17,18]. Both the cytotoxic

receptors might participate in mediating the effects of

actions and acetylcholine release induced by excessive

kainate. The mechanism underlying the link between the

activation of postsynaptic NMDA receptors might directly

activation of kainate receptors and the involvement of

contribute to the effects of NMDA receptors in the present

NMDA receptors is a controversial issue. It might be

experimental model.

attributed to a presynaptic excitation on glutamatergic

In addition, these results show that the potentiation of

neurons

[27],

a

presynaptic

disinhibitory

effect

on

AMPA receptors by cyclothiazide has no significant effects

GABAergic neurons [6], or a postsynaptic excitatory

on kainate-induced decrease in the binding of acetylcholine

action on target cells [5] through kainate receptors. In the

M

receptors in rat forebrain, in agreement with our

latter case, it might be necessary to assume the presence of

previous observation that AMPA receptors apparently did

a direct functional interaction between the postsynaptic

not participate in mediating kainate-induced acetylcholine

kainate- and NMDA-receptors on the same target cells,

release in the rat striatum [16]. Immunolabelings with

like that between adenosine A

- and dopamine D -re-

(3)

Fig. 3. Effects of increasing doses of MK-801 (filled symbols) or CNQX (unfilled symbols) on kainate (4 nmol)-induced decreases in the bindings of

3 3 3

(4)

3 3

Fig. 4. Effects of increasing doses of cyclothiazide on kainate (4 nmol)-induced decreases in the bindings of [ H]kainate (10 nM), [ H]MK-801 (4 nM)

4.3. Kainate receptors and the cholinergic

kainate receptors could not be ruled out because of the lack

neurotransmission in the CNS

of high selectivity of NBQX and CNQX on AMPA- and

kainate-receptors.

The present receptor binding results are consistent with

Loss of cholinergic neurons in the basal forebrain is

those obtained in neurotransmitter release [16] and im-

believed to account for much of the learning and memory

munohistochemistry [42] studies. These results indicate

deficits in Alzheimer’s disease. It was reported that

that kainate receptors are predominantly involved in the

domoate, a glutamate analogue produced by mussels,

EAA-induced increase in acetylcholine release, decrease in

caused a syndrome with features that resemble those of

M receptor binding and ChAT immunoreactivity, increase

Alzheimer’s disease in a group of Newfoundlanders after

in APP immunoreactivity, and neuronal loss. They might

consuming

domoate-contaminated

mussels

in

1987

suggest a crucial role for kainate receptors in regulating

[29,36,37]. Domoate has been established to be a relatively

EAA-modulated cholinergic neurotransmission in the CNS

selective agonist at kainate receptors although it can still

as well as a similar role for the receptors in the CNS

act on AMPA receptors [20]. The present studies may help

diseases associated with cholinergic disorders e.g. Al-

to understand the mechanism of neurodegeneration

under-zheimer’s disease on the basis of the cholinergic hypoth-

lying the domoate-induced syndrome.

esis of this neurodegenerative disease [13,24].

The main functions ascribed to cholinergic pathways are

related to arousal and learning, and motor control. Block-

5. Conclusion

ade of glutamatergic transmission by CNQX and NBQX

(another antagonist at non-NMDA receptors) has been

In summary, the present studies demonstrated that

shown to produce cognitive deficits, which are attributed to

striatal kainate injection decreased the binding of

(5)

[13] P.T. Francis, A.M. Palmer, M. Snape, G.K. Wilcock, The cholinergic

dependent on the activation of kainate receptors and, to a

hypothesis of Alzheimer’s disease: a review of progress, J. Neurol.

certain extent, a consequent involvement of NMDA

re-Neurosur. Psychiat. 66 (1999) 137–147.

ceptors. These results support the possibility that both the

_{[14] B.H. Guevara, I.S. Hoffmann, L.X. Cubeddu, Ventral and dorsal}

pre- and post-synaptic kainate receptors participate in

striatal cholinergic neurons have different sensitivities to kainic acid,

mediating the effects of kainate in the present experimental

Neurochem. Int. 31 (1997) 723–730.

[15] R.E. Hruska, R. Schwarcz, J.T. Coyle, H.I. Yamamura, Alterations

model. These and previous findings provide some evidence

of muscarinic cholinergic receptors in the rat striatum after kainic

that kainate receptors may play a crucial role in regulating

acid injections, Brain Res. 152 (1978) 620–625.

EAA-modulated cholinergic neurotransmission in the

_{[16] S. Jin, AMPA- and kainate-receptors differentially mediate}

excitat-CNS.

ory amino acid-induced dopamine and acetylcholine release from rat

striatal slices, Neuropharmacology 36 (1997) 1503–1510. [17] S. Jin, B.B. Fredholm, Role of NMDA, AMPA and kainate receptors

in mediating glutamate- and 4-AP-induced dopamine and

acetyl-Acknowledgements

choline release from rat striatal slices, Neuropharmacology 331 (1994) 1039–1048.

[18] S. Jin, B.B. Fredholm, Glucose deprivation increases basal and

This work was supported by grants (3981319 / 3981320)

electrically evoked transmitter release from rat striatal slices. Role of

to PTHW from The National University of Singapore.

NMDA- and adenosine A receptors, Eur. J. Pharmacol. 340 (1997)1

169–175.

[19] S. Jin, B. Johansson, B.B. Fredholm, Effects of adenosine A and1

A receptor activation on electrically evoked dopamine and acetyl-2

References

_{choline release from rat striatal slices, J. Pharmacol. Exp. Ther. 267}