Regulation of glutamatergic neurotransmission in the striatum by presynaptic adenylyl cyclase-dependent processes

Regulation of glutamatergic neurotransmission in the striatum by presynaptic adenylyl cyclase-dependent processes

Neurochemistry International 42 (2003) 1–7 Regulation of glutamatergic neurotransmission in the striatum by presynaptic adenylyl cyclase-dependent pr...

110KB Sizes 0 Downloads 68 Views

Neurochemistry International 42 (2003) 1–7

Regulation of glutamatergic neurotransmission in the striatum by presynaptic adenylyl cyclase-dependent processes Róbert Dohovics a,b,∗ , Réka Janáky a , Vince Varga a,b , András Hermann a,b , Pirjo Saransaari a , Simo S. Oja a,c b

a Brain Research Center, University of Tampere Medical School, Tampere, Finland Department of Animal Anatomy and Physiology, University of Debrecen, Debrecen, Hungary c Department of Clinical Physiology, Tampere University Hospital, Tampere, Finland

Received 3 December 2001; received in revised form 1 April 2002; accepted 9 April 2002

Abstract The aim here was to examine the possible roles of adenylyl cyclase- and protein kinase A (PKA)-dependent processes in ionotropic glutamate receptor (iGluR)-mediated neurotransmission using superfused mouse striatal slices and a non-metabolized l-glutamate analogue, d-[3 H]aspartate. The direct and indirect presynaptic modulation of glutamate release and its susceptibility to changes in the intracellular levels of cyclic AMP (cAMP), Ca2+ and calmodulin (CaM) and in protein phosphorylation was characterized by pharmacological manipulations. The agonists of iGluRs, 2-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) and kainate, stimulated the basal release of d-[3 H]aspartate, while N-methyl-d-aspartate (NMDA) was without effect. Both the AMPA- and kainate-mediated responses were accentuated by the ␤-adrenoceptor agonist isoproterenol. These facilitatory effects were mimicked by the permeable cAMP analogue dibutyryl-cAMP. The ␤-adrenoceptor antagonist propranolol, the adenylyl cyclase inhibitor MDL12,330A, the inhibitor of PKA and PKC, H-7, and the PKA inhibitor H-89 abolished the isoproterenol effect on the kainate-evoked release. The dibutyryl-cAMP-induced potentiation was also attenuated by H-7. Isoproterenol, propranolol and MDL12,330A failed to affect the basal release of d-[3 H]aspartate, but dibutyryl-cAMP was inhibitory and MDL12,330A activatory. In Ca2+ -free medium, the kainate-evoked release was enhanced, being further accentuated by the CaM antagonists calmidazolium and trifluoperazine, though these inhibited the basal release. The potentiating effect of calmidazolium on the kainate-stimulated release was counteracted by both MDL12,330A and H-7. We conclude that AMPA- and kainate-evoked glutamate release from striatal glutamatergic terminals is potentiated by ␤-adrenergic receptor-mediated adenylyl cyclase activation and cAMP accumulation. Glutamate release is enhanced if the Ca2+ - and CaM-dependent, kainate-evoked processes do not prevent the excessive accumulation of intracellular cAMP. © 2003 Elsevier Science Ltd. All rights reserved. Keywords: Ionotropic glutamate receptors; ␤-Adrenoceptors; Adenylyl cyclases; Calmodulin antagonists; Striatum

1. Introduction The neostriatum receives glutamatergic projections from the cerebral cortex and thalamus (Smith and Bolam, 1990). Synaptic transmission in these corticostriatal and thalamostriatal pathways seems to play a crucial role in certain cognitive functions, motor coordination and plasticity (Calabresi et al., 2000) and in neurodegenerative disorders of the basal ganglia, i.e. Parkinson’s and Huntington’s diseases (Alexi et al., 2000). Glutamate released from the nerve endings may act at presynaptic ∗

Corresponding author. Tel.: +358-3-215-6879; fax: +358-3-215-6170. E-mail address: [email protected] (R. Dohovics).

or postsynaptic ionotropic glutamate receptors (iGluRs), which gate cation channels and mediate fast and prolonged excitatory effects. Of the iGluRs, N-methyl-d-aspartate (NMDA) receptors consist of NR1 and NR2A–D subunits, 2-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptors of GluR1–4 subunits and kainate receptors of GluR5–7 and KA1–2 subunits (Gasic and Hollmann, 1992; Dingledine et al., 1999). The AMPA and kainate receptors are rapidly activated and desensitized in response to glutamate and pass Na+ and Ca2+ ions. The presence of GluR2 subunit determines the Ca2+ permeability of recombinant AMPA receptors, whereas the homomeric GluR6 kainate receptors are fairly permeable to Ca2+ (Ozawa et al., 1998). The NMDA receptors are activated and desen-

0197-0186/03/$ – see front matter © 2003 Elsevier Science Ltd. All rights reserved. PII: S 0 1 9 7 - 0 1 8 6 ( 0 2 ) 0 0 0 6 6 - 9

2

R. Dohovics et al. / Neurochemistry International 42 (2003) 1–7

sitized slowly and their Ca2+ permeability is regulated by the voltage-dependent Mg2+ block (Ozawa et al., 1998). The kainate and AMPA receptors can be upregulated by protein kinase A (PKA) via cyclic AMP (cAMP)-dependent phosphorylation (Keller et al., 1992; Raymond et al., 1994). The modulation of presynaptic iGluRs by the cAMP-dependent phosphorylation cascade may thus be important for the regulation of glutamate release. Indeed, direct activation of the adenylyl cyclase/cAMP/PKA cascade enhances glutamatergic transmission in the striatum (Colwell and Levine, 1995). On the other hand, ␤-adrenoceptor-mediated activation of this pathway has also been shown to evoke glutamate release from cerebral cortical synaptosomes (Herrero and Sánchez-Prieto, 1996). Two different intracellular pathways, Ca2+ - and cAMP-dependent, may thus interact in the regulation of glutamate release from nerve endings. G-protein coupled receptors localized at presynaptic terminals may play a fundamental role in this regulation via activation of adenylyl cyclases, which are also influenced by the intracellular Ca2+ and calmodulin (CaM) levels (Cooper et al., 1995). The nine isoforms of mammalian adenylyl cyclases so far cloned possess distinct characteristics, including sensitivity to Ca2+ . They are classified as Ca2+ -stimulated (AC1, AC8, AC3), Ca2+ -inhibited (AC5 and AC6) and Ca2+ -insensitive (AC2, AC4 and AC7) enzymes (Mons et al., 1998). A further type IX (AC9) is inhibited by Ca2+ and calcineurin, being a Ca2+ and CaM-dependent protein phosphatase (Antoni et al., 1995; Paterson et al., 1995). In situ hybridization and immunohistochemical studies have shown that corticostriatal afferents, originating mainly from layers V–VI, are rich in AC5 and AC9 (Mons et al., 1995; Antoni et al., 1998a,b). Although the role of postsynaptic iGluRs in the interplay of Ca2+ , CaM, adenylyl cyclase and cAMP is fairly well-characterized (Rajadhyaksha et al., 1998) and the importance of Ca2+ -dependent inhibitory actions of kainate receptors at hippocampal excitatory terminals (Kamiya and Ozawa, 1988, 2000) recognized, knowledge of the involvement of presynaptic iGluRs in inducing long-lasting changes is still scant. Stimulation of D1 dopamine receptors, which are positively coupled to adenylyl cyclase, has been reported to promote phosphorylation of AMPA receptors by PKA and to increase the amplitude of currents through AMPA-gated channels in cultured striatal neurons (Priece et al., 1999). Ca2+ - and CaM-dependent changes in synaptic glutamate release may be responsible for this iGluR phosphorylation. In addition, recent electrophysiological studies suggest that ␤-adrenergic activation can induce longlasting enhancement of glutamatergic transmission in the corticostriatal pathway and ␤-adrenoceptors are probably localized at the corticostriatal terminals (Niittykoski et al., 1999). Our aim was now to investigate interactions of presynaptic ␤-adrenoceptors, iGluRs and adenylyl cyclase-, cAMP- and Ca2+ - and CaM-dependent processes in the regulation of striatal glutamatergic neurotransmission.

2. Experimental procedures 2.1. Materials d-[3 H]aspartate (specific activity 1.78 PBq/mol) was from Amersham International (Bristol, UK), agonists of iGluRs NMDA, AMPA, kainate and (2S,4R)-4-methylglutamate (SYM 2081), agonist and antagonist of ␤-adrenoceptors (±)-isoproterenol hydrochloride and (±)-propranolol hydrochloride, respectively, cAMP analogue N6 -2 -O-dibutyryladenosine 3 :5 -cyclic monophosphate (dibutyryl-cAMP), Ca2+ chelators ethylenediaminetetraacetic acid (EDTA) and ethyleneglycol-bis(␤-aminoethyl)-N,N,N ,N -tetraacetic acid (EGTA) from Sigma or Tocris Cookson (Bristol, UK), CaM antagonists calmidazolium chloride and trifluoperazine dihydrochloride, adenylyl cyclase inhibitor cis-N-(2-phenylcyclopentyl)azacyclotridec-1-en-2-amine monohydrochloride (MDL12,330A), inhibitor of PKA and protein kinase C (PKC) 1-(5-isoquinolinesulfonyl)-2-methylpiperazine dihydrochloride (H-7) from RBI (Natick, MA, USA) and PKA inhibitor N-[2-((p-bromoeinnamyl)amino)ethyl]-5-isoquinolinesulfonamide dihydrochloride (H-89) from Calbiochem (La Jolla, CA, USA). The drug concentrations were chosen to equal those commonly applied in other neuropharmacological experiments. At these concentrations, the effects of the compounds on the intracellular messengers and enzymes have been reported to be selective for their targets. All drugs were soluble in the superfusion medium, except for calmidazolium chloride and H-89, which were solubilized in 2% dimethylsulfoxide. At this concentration this solvent did not influence d-aspartate release. 2.2. Release experiments Adult NMRI mice (Orion, Helsinki, Finland) of both sexes were used throughout. The striata were excised and 0.3 mm thick slices prepared with a McIlwain tissue chopper on ice. The slices were first preloaded for 30 min with 0.16 ␮M (285 MBq/l) d-[3 H]aspartate at 37 ◦ C in preoxygenated standard Krebs–Ringer–HEPES (KRH) solution containing (in mM) 126 NaCl, 1.3 MgSO4 , 1.3 NaH2 PO4 , 5.0 KCl, 0.8 CaCl2 , 15.0 HEPES and 10.0 d-glucose (pH 7.4), then transferred to closed chambers (i.d. 6 mm) and superfused (0.25 ml/min) with oxygenated KRH solution using a peristaltic pump. After a 30 min washing period, the superfusion medium was pooled during the first 20 min, whereafter 2 min fractions were collected. The superfusion experiments were of two types. In the first set of experiments, the effectors were present from 60 to 70 min of superfusion. In the second set, the medium was supplemented by the effectors 6 min earlier, and kainate or AMPA was then added from 60 to 70 min, as detailed in the table and figure legends. From 70 min onwards, the superfusions were continued with standard KRH medium. After the experiments, the slices were homogenized in ice-cold 5% (w/v) trichloracetic acid solution and centrifuged, and

R. Dohovics et al. / Neurochemistry International 42 (2003) 1–7

3

Fig. 1. Time-course of d-[3 H]aspartate release from mouse striatal slices. (A) Effects of different iGluR agonists (all 1 mM): kainate (䊊), NMDA () and AMPA (䉱) on the basal (䊉) release. (B) Concentration-dependent stimulation of the basal release by 0.1 (䊐), 0.5 (), 2.0 (䉱), and 5.0 mM (䊊) kainate. The presence of agonists is indicated by the bars. The results are given as mean values ± S.E.M. (indicated when greater than the size of symbols) of three to 12 independent experiments.

the radioactivity remaining in the slices determined from the clear supernatants. The effluent samples were likewise subjected to liquid scintillation counting. In low-Mg2+ media, the medium contained only 0.1 mM MgSO4 and the remaining MgSO4 was substituted by equimolar choline chloride. In Ca2+ -free media, Ca2+ (0.8 mM) was omitted, substituted by choline chloride, and EDTA or EGTA (2 mM) added at 54 min of superfusion. 2.3. Estimation of efflux rate constants The first-order fractional efflux rate constants were calculated as follows (Kontro, 1979; Korpi and Oja, 1979). First, the release of d-[3 H]aspartate from the slices was plotted as the percentage of radioactivity remaining in the slices at each time point of superfusion, recovered in the collected superfusate fractions and in the slices at the end (cf. Fig. 1), then, the logarithm of the radioactivity remaining at the slices at each time point was plotted against the superfusion time. The straight regression lines were fitted to these data for the both experimental phases. The slopes of these straight lines are the estimated efflux rate constants k1 and k2 ; basal release from 50 to 60 min (k1 ) and stimulated release from 62 to 70 min (k2 ) (cf. Table 1). The results are given either as a time-course of percentage release (Fig. 1) or as percentages of k2 of the corresponding rate constant k1 for the same slice (Table 1). Statistical significances were estimated by two-way Student’s t-test.

GluR5–7 and KA1–2 iGluRs, enhanced the basal release of d-[3 H]aspartate at the 1 mM concentration, while the NR1–2 selective agonist NMDA had no effect (Fig. 1A). NMDA was tested in low-Mg2+ (0.1 mM) medium to avoid the voltage-dependent Mg2+ block. The stimulatory effect of Table 1 Modulation of kainate-evoked d-[3 H]aspartate release from mouse striatal slices by different receptor ligands, signal transduction effectors and enzyme modulatorsa Compound (mM)

Efflux rate constants k2 (62–70 min), percentage of control

Kainate, 1.0 (control A) MDL12,330A, 0.025 H-7, 0.1 H-89, 0.01 Propranolol, 0.01

226 175 227 249 224

± ± ± ± ±

12 (12) 11 (A)∗ (5) 10 (9) 8 (4) 9 (6)

Isoproterenol, 0.01 (control B) + propranolol, 0.01 + MDL12,330A, 0.025 + H-7, 0.1 + H-89, 0.01

306 225 197 205 241

± ± ± ± ±

9 (A)∗∗ (12) 10 (B)∗∗ (6) 8 (B)∗∗ (6) 12 (B)∗∗ (9) 14 (B)∗∗ (5)

Dibutyryl-cAMP, 1.0 (control C) + H-7, 0.1 Trifluoperazine, 1.0

326 ± 13 (A)∗∗ (5) 188 ± 7 (C)∗∗ (9) 327 ± 28 (A)∗∗ (5)

Calmidazolium, 0.05 (control D) + MDL12,330A, 0.025 + H-7, 0.1 Ca2+ -free medium + EGTA, 2.0

368 229 194 318

a

3. Results 3.1. Stimulation of the basal efflux by iGluR agonists Both AMPA, a specific agonist with high-affinity for GluR1–4 iGluRs, and kainate, an agonist preferably at

± ± ± ±

41 16 25 18

(A)∗∗ (5) (D)∗ (5) (D)∗ (4) (A)∗∗ (6)

Kainate, CaM antagonists and propranolol were present from 60 to 70 min of superfusion and the other effectors from 54 to 70 min. The fractional efflux rate constants k2 for the stimulation period (62–70 min) are given in percent of the corresponding control k2 . Mean values with S.E.M. are shown. Number of independent experiments in parentheses. Significance of differences from the corresponding control (indicated in parentheses). ∗ P < 0.05. ∗∗ P < 0.01.

4

R. Dohovics et al. / Neurochemistry International 42 (2003) 1–7

kainate (0.1–5.0 mM) was concentration-dependent, varying from 24% to 310% (Fig. 1B). SYM 2081, a high-affinity selective agonist of kainate receptors, increased the release of d-[3 H]aspartate by 353% ± 25% (mean ± S.E.M., n = 6) and 893% ± 92% (mean ± S.E.M., n = 5) at the 25 and 100 ␮M concentrations, respectively.

versus control k2 98.6% ± 2.1%, n = 8; of k1 ± S.E.M.), whereas the kainate-evoked release was strongly enhanced (Table 1). These effects were also similar in Ca2+ -free media containing EDTA (data not shown).

3.2. Effects of isoproterenol and dibutyryl-cAMP and their modification by adenylyl cyclase inhibitor

Ten micromoles H-89, a selective and potent inhibitor of PKA, did not significantly affect the basal efflux (k2 90.3%± 4.3% of k1 , mean ± S.E.M.; n = 5) or the release induced by kainate alone (Table 1), whereas at the same concentration it completely abolished the potentiating effect of 10 ␮M isoproterenol on the kainate-evoked release (Table 1). Similarly, 100 ␮M H-7, a potent inhibitor of PKA and PKC, did not alter the basal (k2 90.2% ± 2.7% of k1 , mean ± S.E.M.; n = 3) or kainate-evoked release (Table 1), but attenuated the effect of isoproterenol (Table 1). The potentiations of kainate-evoked release by dibutyryl-cAMP and calmidazolium were also completely blocked by H-7 (Table 1).

Isoproterenol, a selective ␤-adrenoceptor agonist, had no effect at the 10 ␮M concentration on the basal release (k2 93.9% ± 3.5%, n = 4 versus control k2 98.6% ± 2.1%, n = 8; of k1 ± S.E.M.) but increased the release evoked by 1 mM AMPA from 165%±5% to 188%±6% (mean±S.E.M.; n = 5; P < 0.05). This effect was mimicked by 1 mM dibutyrylcAMP, a non-metabolized cAMP analogue, which increased the release up to 209% ± 11% (mean ± S.E.M.; n = 5; P < 0.01). It also significantly reduced the basal release to 77.1% ± 0.9% (mean ± S.E.M., n = 3, P < 0.01). The release evoked by 1 mM kainate was also enhanced by isoproterenol with k2 values of 289% ± 11% and 306% ± 9% of control (mean ± S.E.M.; n = 5 and n = 12; P < 0.01) at the 1 and 10 ␮M concentrations, respectively. Similarly to isoproterenol, dibutyryl-cAMP also enhanced the kainate-evoked release. The effect of isoproterenol was antagonized by the selective ␤-adrenoceptor antagonist propranolol at the 10 ␮M concentration (Table 1). The permeable and irreversible inhibitor of adenylyl cyclase MDL12,330A did not alter the basal release at the 25 ␮M concentration but at 0.1 mM it was activatory (k2 133% ± 5% of k1 , mean ± S.E.M.; n = 3; P < 0.01). At the 25 ␮M concentration, MDL12,330A inhibited the effect of 1 mM kainate and completely abolished the potentiation of kainate stimulation by 10 ␮M isoproterenol (Table 1). 3.3. Effects of Ca2+ -free media and calmodulin antagonists Calmidazolium (50 ␮M), a potent antagonist of CaM and an inhibitor of brain CaM-dependent phosphodiesterase, markedly enhanced the kainate-evoked release (Table 1), but significantly reduced the basal release (k2 77.9% ± 6.7% of k1 , mean ± S.E.M.; n = 5; P < 0.05). Another CaM antagonist trifluoperazine, which is also a phosphodiesterase inhibitor, now concentration-dependently enhanced the kainate-evoked release. Trifluoperazine was a weaker activator than calmidazolium, but both at 0.5 mM (k2 296%±32% of control, mean ± S.E.M.; n = 5; P < 0.01) and 1 mM (Table 1) its effects were clearly significant. Trifluoperazine also diminished the basal release (k2 79.8% ± 6.6% of k1 , mean ± S.E.M.; n = 6; P < 0.05) similarly to calmidazolium. The effects of calmidazolium and trifluoperazine on the kainate-evoked release were completely blocked by 25 ␮M MDL12,330A (Table 1). The basal release was significantly (P < 0.05) reduced in Ca2+ -free media containing EGTA (k2 87.0% ± 5.0% of k1 , mean ± S.E.M.; n = 6,

3.4. Effects of protein kinase inhibitors

4. Discussion In order to investigate the presynaptic regulation of iGluR-mediated glutamate release in the striatum we used d-[3 H]aspartate, a non-metabolized compound, widely employed as a marker for l-glutamate release. Within nerve terminals, l-glutamate is distributed into two pools, cytosolic and vesicular, and most likely released from them in Ca2+ -independent and Ca2+ -dependent manners, respectively (Bernath, 1992; Vizi, 2000). Although d-aspartate is known to be a poor substrate of vesicular uptake and mainly resides in the cytosolic pool, it can be taken up by synaptic vesicles and is released from synaptic terminals by electric stimulation in a Ca2+ -dependent manner (Muzzolini et al., 1997; Savage et al., 2001). These findings, together with the observation that nerve endings take up d-[3 H]aspartate more readily than do glial cells (Gundersen et al., 1993), validate the use of this marker as an indicator of synaptic glutamate release. Our present finding that AMPA and kainate, but not NMDA, enhance the release of d-aspartate would indicate that iGluRs of the AMPA and kainate classes are localized at glutamatergic terminals in the striatum. This inference is supported by in vivo microdialysis studies (Smolders et al., 1996; Patel et al., 2001). In agreement, immunolabeling studies have provided evidence for the presence of GluR6/7 and KA2 receptor subunits at corticostriatal nerve endings in the monkey striatum (Charara et al., 1999) and GluR1 and GluR2/3 subunits in the rat striatum (Bernard et al., 1997). The effects of receptor agonists on the activation and desensitization of the AMPA and kainate receptor-coupled ion channels are different depending on the subunit composition (Bleakman and Lodge, 1998; Lerma et al., 2001). The existence of presynaptic NMDA receptors at glutamatergic terminals in the striatum remains a matter of debate,

R. Dohovics et al. / Neurochemistry International 42 (2003) 1–7

but according to immunocytochemical evidence they are co-localized postsynaptically rather than presynaptically with the GluR1/3 subunits of AMPA receptors (Bernard and Bolam, 1998). It thus seems unlikely that NMDA receptors are involved in presynaptic regulation of glutamate release in the striatum. SYM 2081, the high-affinity, fastacting and selective agonist of kainate receptors (Zhou et al., 1997), enhanced now very effectively the release of d-[3 H]aspartate from striatal slices. The assumption that there exists autoregulation of presynaptic kainate receptors in the striatum is also otherwise in line with our present data. This autoregulation may normally operate at relatively low glutamate concentrations, though the physiological conditions under which the receptors in the striatum are activated remain to be determined (Chergui et al., 2000). On the other hand, very high extracellular concentrations are needed to produce excessive ion fluxes in pathological states (Obrenovitch et al., 1997). Involvement of ␤-adrenoceptors in the presynaptic regulation of glutamate release from striatal nerve endings has been proposed by Niittykoski et al. (1999). These authors reported that isoproterenol enhances AMPA-mediated transmission in the corticostriatal pathway. In accordance with this observation, isoproterenol now markedly enhanced d-[3 H]aspartate release evoked by kainate and AMPA and propranolol antagonized its effect on the kainate-evoked release. ␤-Adrenoceptors are thus likely to be present at striatal glutamatergic terminals and to cooperate with presynaptic kainate and AMPA receptors in glutamatergic transmission. All agents which inhibit the consecutive steps in the ␤-adrenoceptor-mediated cAMP/PKA-dependent signaling pathway, i.e. the ␤-adrenoceptor antagonist propranolol, the adenylyl cyclase inhibitor MDL12,330A and the nonselective (H-7) and selective (H-89) inhibitors of PKA now abolished the enhancement of kainate-evoked release by isoproterenol. It is thus probable that the effect of isoproterenol is receptor-mediated and involves PKA-dependent protein phosphorylation. The involvement of PKA-dependent protein phosphorylation in the interactions of striatal kainate and AMPA iGluRs and ␤-adrenoceptors is corroborated by the finding that dibutyryl-cAMP mimicked the effects of isoproterenol on both AMPA- and kainate-evoked releases, and this potentiation was also sensitive to the PKC/PKA inhibitor H-7. A possible mechanism of these enforcements is cAMP-PKA-dependent phosphorylation of kainate and AMPA receptors, known to be upregulated by PKA-dependent phosphorylation. Phosphorylation of the GluR6 kainate receptor by PKA enhances the current amplitude (Raymond et al., 1993; Wang et al., 1993) and opening probability (Traynelis and Wahl, 1997) of the associated ion channel. Phosphorylation of the GluR1 AMPA receptor by PKA also increases the peak opening probability of the corresponding channel (Banke et al., 2000). The release evoked by kainate alone was now inhibited by the adenylyl cyclase inhibitor MDL12,330A, but not by propranolol, and somewhat unexpectedly, not by the protein

5

kinase inhibitors H-7 and H-89. This shows that the kainate effect does not result from the kainate-evoked release of endogenous ␤-adrenoceptor agonists. It is partially mediated by cAMP but does not involve PKA activation. Kainate can elevate the intracellular cAMP level by at least two mechanisms, though the exact coupling to cAMP generation is still obscure. It may directly activate adenylyl cyclase via a G-protein (Varga et al., 1996) or cause depolarizationdependent adenylyl cyclase activation by Na+ influx through kainate/AMPA receptor-gated ionophores or by Na+ /Ca2+ influx via the l-type Ca2+ -channels (Cooper et al., 1998). This PKA-independent action of kainate may have been abolished in our study by MDL12,330A. In order to establish whether indirect or direct alterations in the intracellular levels of Ca2+ and CaM are involved in the kainate-receptor-mediated release of d-[3 H]aspartate, we used Ca2+ chelators and the CaM antagonists calmidazolium and trifluoperazine. These are known to inhibit brain CaM-dependent phosphodiesterases (Gietzen et al., 1982) and to block directly the l-type Ca2+ -channels in smooth muscle cells (Nakazawa et al., 1993). In the present study, blocking of Ca2+ - and CaM-dependent processes, either by omission of extracellular Ca2+ or by inhibition of CaM, also enhanced the kainate-evoked release. This enhancement, similarly to that caused by isoproterenol, was prevented by MDL12,330A and H-7, indicating that a Ca2+ - and CaMdependent process indeed inhibits the cAMP-PKA-dependent facilitation of kainate/AMPA-mediated release of glutamate from striatal terminals. This, in turn, suggests that there are at least two alternative mechanisms involved in the regulation of cAMP levels by presynaptic kainate/AMPA receptors. First, kainate evokes depolarization and consequent influx of Ca2+ through activated voltage-sensitive Ca2+ -channels. An elevation of intracellular Ca2+ may then inhibit the adenylyl cyclases AC5 and AC9, which abound in the striatum (Hellevuo et al., 1996; Antoni et al., 1998b) and which are known to be inhibited in a Ca2+ - and CaM-dependent manner (Antoni et al., 1995, 1998a). This block is relieved indirectly by omission of extracellular Ca2+ or directly by the CaM antagonists calmidazolium and trifluoperazine. The antagonism of such inhibition would enhance the accumulation of cAMP evoked by kainate. Second, kainate may activate Ca2+ - and CaM-dependent phosphodiesterases by elevating intracellular Ca2+ , with a consequent perturbation of cAMP breakdown (Antoni et al., 1998a). In addition, the resulting high levels of cAMP may reprofile the phosphodiesterase activities by stimulating PKA, which activates Ca2+ -independent and inhibits Ca2+ - and CaM-dependent phosphodiesterases (Ang and Antoni, 1996). This autoregulation, which is most probably a physiological phenomenon and would aid in bridling any overexcitation, is thwarted by omission of Ca2+ or in the presence of CaM antagonists. Isoproterenol alone failed to evoke d-[3 H]aspartate release, not had propranolol any effect on the basal release. This bespeaks the absence of ␤-adrenoceptor-mediated glutamate release under non-depolarized conditions. On the

6

R. Dohovics et al. / Neurochemistry International 42 (2003) 1–7

other hand, dibutyryl-cAMP and the effectors which inhibit Ca2+ - and CaM-dependent processes, i.e. Ca2+ -free medium and CaM antagonists, inhibited the basal release, while the adenylyl cyclase inhibitor was activatory. These results clearly indicate that the basal and kainate/AMPAevoked releases of glutamate are regulated in a different way. The basal release, in contrast to that evoked by activation of kainate/AMPA iGluRs, is not modulated by ␤-adrenoceptors but tonically inhibited by adenylyl-cyclase-mediated cAMP generation and activated by Ca2+ - and CaM-dependent intracellular processes. The crucial point in this delicate balance could be the regulation of functions of glutamate transporters, which are known to be activated by both PKCand PKA-mediated phosphorylation (Pisano et al., 1996). In addition to its well-characterized protein-kinase-mediated actions, cAMP can also have an autocoid-like (local hormone-like) effect in the hippocampal formation, independent of PKA, leading to a marked elevation of the metabolites adenosine and 5 -AMP in the extracellular space. They act at A1 adenosine receptors and hence inhibit synaptic transmission at adjacent excitatory synapses (Gereau and Conn, 1994). We conclude that activation of presynaptic ␤-adrenoceptors located at corticostriatal glutamatergic terminals enhances the positive feedback regulation mediated by presynaptic kainate/AMPA iGluRs. This enhancement probably occurs via PKA-dependent phosphorylation of presynaptic iGluRs. Kainate may elevate presynaptic cAMP but also attenuate its excessive accumulation via a Ca2+ - and CaM-dependent process, either by inhibiting the adenylyl cyclases AC5 and AC9 or perturbing cAMP breakdown. On the other hand, long-lasting co-activation of presynaptic ␤-adrenoceptors and AMPA/KA iGluRs initiates neurotoxic processes and the Ca2+ - and CaM-dependent inhibition of adenylyl cyclases may then be protective. This inhibitory autoregulation is blocked by CaM antagonists.

Acknowledgements The skillful technical assistance of Ms. Irma Rantamaa and Ms. Sari Luokkala and the financial support of the Medical Research Fund of Tampere University Hospital, Finland, are gratefully acknowledged. References Alexi, T., Borlongan, C.V., Faull, R.L.M., Williams, C.E., Clark, R.G., Gluckman, P.D., Hughes, P.E., 2000. Neuroprotective strategies for basal ganglia degeneration: Parkinson’s and Huntington’s disease. Prog. Neurobiol. 60, 409–470. Ang, K.L., Antoni, F.A., 1996. Rolipram-inhibitable cyclic nuleotide phosphodiesterase (PDE) in rat adenohypophysis: potential functional role in corticotrophs and somatotrophs. Br. J. Pharmacol. 119, 361. Antoni, F.A., Barnard, R.J.O., Shipston, M.J., Smith, S.M., Simpson, J., Paterson, J.M., 1995. Calcineurin feedback inhibition of agonist-evoked cAMP formation. J. Biol. Chem. 270, 28055–28061.

Antoni, F.A., Smith, S.M., Simpson, J., Rosie, R., Fink, G., Paterson, J.M., 1998a. Calcium control of adenylyl cyclase: the calcineurin connection. Adv. Second Messenger Phosphoprotein Res. 32, 153–172. Antoni, F.A., Palkovits, M., Simpson, J., Smith, S.M., Leitch, A.L., Rosie, R., Fink, G., Paterson, J.M., 1998b. Ca2+ /calcineurin-inhibited adenylyl cyclase, highly abundant in forebrain regions, is important for learning and memory. J. Neurosci. 18, 9650–9661. Banke, T.G., Bowie, D., Lee, H.-K., Huganir, R.L., Schousboe, A., Traynelis, S.F., 2000. Control of GluR1 AMPA receptor function by cAMP-dependent protein kinase. J. Neurosci. 20, 89–102. Bernard, V., Bolam, J.P., 1998. Subcellular and subsynaptic distribution of the NR1 subunit of the NMDA receptor in the neostriatum and globus pallidus of the rat: co-localization at synapses with the GluR2/3 subunit of the AMPA receptor. Eur. J. Neurosci. 10, 3721–3736. Bernard, V., Somogyi, P., Bolam, J.P., 1997. Cellular, subcellular, and subsynaptic distribution of AMPA-type glutamate receptor subunits in the neostriatum. J. Neurosci. 12, 819–833. Bernath, S., 1992. Calcium-independent release of amino acid neurotransmitters: fact or artifact? Prog. Neurobiol. 38, 57–91. Bleakman, D., Lodge, D., 1998. Neuropharmacology of AMPA and kainate receptors. Neuropharmacology 37, 1187–1204. Calabresi, P., Centonze, D., Gubellini, P., Marfia, G.A., Pisani, A., Sancesario, G., Bernardi, G., 2000. Synaptic transmission in the striatum: from plasticity to neurodegeneration. Prog. Neurobiol. 61, 231–265. Charara, A., Blankstein, E., Smith, Y., 1999. Presynaptic kainate receptors in the monkey striatum. Neuroscience 91, 1195–1200. Chergui, K., Bouron, A., Normand, E., Mulle, C., 2000. Functional GluR6 kainate receptors in the striatum: indirect downregulation of synaptic transmission. J. Neurosci. 20, 2175–2182. Colwell, C.S., Levine, M.S., 1995. Excitatory synaptic transmission in neostriatal neurons: regulation by cyclic AMP-dependent mechanisms. J. Neurochem. 15, 1704–1713. Cooper, D.M.F., Mons, N., Karpen, J.W., 1995. Adenylyl cyclases and the interaction between calcium and cAMP signalling. Nature 374, 421–424. Cooper, D.M.F., Schell, M.J., Thorn, P., Irvine, R.F., 1998. Regulation of adenylyl cyclase by membrane potential. J. Biol. Chem. 273, 27703– 27707. Dingledine, R., Borges, K., Bowie, D., Traynelis, S.F., 1999. The glutamate receptor ion channels. Pharmacol. Rev. 51, 7–61. Gasic, G.P., Hollmann, M., 1992. Molecular neurobiology of glutamate receptors. Annu. Rev. Physiol. 54, 507–536. Gereau, R.W., Conn, P.J., 1994. Potentiation of cAMP responses by metabotropic glutamate receptors depresses excitatory synaptic transmission by a kinase-independent mechanism. Neuron 12, 1121– 1129. Gietzen, K., Sadorf, I., Bader, H., 1982. A model for the regulation of the calmodulin-dependent enzymes erythrocyte Ca2+ -transport ATPase and brain phosphodiesterase by activators and inhibitors. Biochem. J. 207, 541–548. Gundersen, V., Danbolt, N.C., Ottersen, O.P., Storm-Mathisen, J., 1993. Demonstration of glutamate/aspartate uptake activity in nerve endings by use of antibodies recognizing exogenous d-aspartate. Neuroscience 57, 97–111. Hellevuo, K., Hoffman, P.L., Tabakoff, B., 1996. Adenylyl cyclases: mRNA and characteristics of enzyme activity in three areas of brain. J. Neurochem. 67, 177–185. Herrero, I., Sánchez-Prieto, J., 1996. The cAMP-dependent facilitation of glutamate release by ␤-adrenergic receptors in cerebrocortical nerve terminals. J. Biol. Chem. 271, 30554–30560. Kamiya, H., Ozawa, S., 1988. Kainate receptor-mediated inhibition of presynaptic Ca2+ influx and EPSP in area CA1 of the rat hippocampus. J. Physiol., London 509, 833–845. Kamiya, H., Ozawa, S., 2000. Kainate receptor-mediated presynaptic inhibition at the mouse hippocampal mossy fibre synapse. J. Physiol., London 523, 653–665.

R. Dohovics et al. / Neurochemistry International 42 (2003) 1–7 Keller, B.U., Hollmann, M., Heinemann, S., Konnerth, A., 1992. Calcium influx through subunits GluR1/GluR3 of kainate/AMPA receptor channels is regulated by cAMP dependent protein kinase. EMBO J. 11, 891–896. Kontro, P., 1979. Components of taurine efflux in rat brain synaptosomes. Neuroscience 11, 1745–1749. Korpi, E.R., Oja, S.S., 1979. Efflux of phenylalanine from rat cerebral cortex slices as influenced by extra- and intracellular amino acids. J. Neurochem. 32, 789–796. Lerma, J., Paternain, A.V., Rodr´ıguez-Moreno, A., López-Garc´ıa, J.C., 2001. Molecular physiology of kainate receptors. Physiol. Rev. 81, 971–998. Mons, N., Harry, A., Dubourg, P., Premont, R.T., Iyengar, R., Cooper, D.M.F., 1995. Immunohistochemical localization of adenylyl cyclase in rat brain indicates a highly selective concentration at synapses. Proc. Natl. Acad. Sci. U.S.A. 92, 8473–8477. Mons, N., Decorte, L., Jaffard, R., Cooper, D.M.F., 1998. Ca2+ -sensitive adenylyl cyclases, key integrators of cellular signalling. Life Sci. 62, 1647–1652. Muzzolini, A., Bregola, G., Bianchi, C., Beani, L., Simonato, M., 1997. Characterization of glutamate and [3 H]d-aspartate outflow from various in vitro preparations of the rat hippocampus. Neurochem. Int. 31, 113– 124. Nakazawa, K., Higo, K., Abe, K., Tanaka, Y., Saito, H., Matsuki, N., 1993. Blockade by calmodulin inhibitors of Ca2+ channels in smooth muscle from rat vas deferens. Br. J. Pharmacol. 109, 137–141. Niittykoski, M., Ruotsalainen, S., Haapalinna, A., Larson, J., Sirviö, J., 1999. Activation of muscarinic M3-like receptors and ␤-adrenoceptors, but not M2-like muscarinic receptors or ␣-adrenoceptors, directly modulates corticostriatal neurotransmission in vitro. Neuroscience 90, 95–105. Obrenovitch, T.P., Urenjak, J., Zilkha, E., 1997. Effects of increased extracellular glutamate levels on the local field potential in the brain of anaesthetized rats. Br. J. Pharmacol. 122, 372–378. Ozawa, S., Kamiya, H., Tsuzuki, K., 1998. Glutamate receptors in the mammalian central nervous system. Prog. Neurobiol. 54, 581–618. Patel, D.R., Young, A.M.J., Croucher, M.J., 2001. Presynaptic ␣-amino-3-hydroxy-5-methyl-4-isoxazoleproprionate receptor-mediated stimulation of glutamate and GABA release in the rat striatum in vivo: a dual-label microdialysis study. Neuroscience 102, 101–111. Paterson, J.M., Smith, S.M., Harmar, A.J., Antoni, F.A., 1995. Control of a novel adenylyl cyclase by calcineurin. Biochem. Biophys. Res. Commun. 214, 1000–1008.

7

Pisano, P., Samuel, D., Nieoullon, A., Kerkerian-Le Goff, L., 1996. Activation of the adenylate cyclase-dependent protein kinase pathway increases high-affinity glutamate uptake into rat striatal synaptosomes. Neuropharmacology 35, 541–547. Priece, C.J., Kim, P., Raymond, L.A., 1999. D1 dopamine receptor-induced cyclic AMP-dependent protein kinase phosphorylation and potentiation of striatal glutamate receptors. J. Neurochem. 73, 2441– 2446. Rajadhyaksha, A., Leveque, J.-C., Macias, W., Barczak, A., Konradi, C., 1998. Molecular components of striatal plasticity: the various routes of cyclic AMP pathways. Dev. Neurosci. 20, 204–215. Raymond, L.A., Blackstone, C.D., Huganir, R.L., 1993. Phosphorylation and modulation of recombinant GluR6 glutamate receptors by cAMP-dependent protein kinase. Nature 361, 637–641. Raymond, L.A., Tingley, W.G., Blackstone, C.D., Roche, K.W., Huganir, R.L., 1994. Glutamate receptor modulation by protein phosphorylation. J. Physiol., Paris 88, 181–192. Savage, D.D., Galindo, R., Queen, S.A., Paxton, L.L., Allan, A.M., 2001. Characterization of electrically evoked [3 H]-d-aspartate release from hippocampal slices. Neurochem. Int. 38, 255–267. Smith, A.D., Bolam, J.P., 1990. The neural network of the basal ganglia as revealed by the study of synaptic connections of identified neurones. Trends Neurosci. 13, 259–265. Smolders, I., Sarre, S., Vanhaesendonck, C., Ebinger, G., Michotte, Y., 1996. Extracellular striatal dopamine and glutamate after decortication and kainate receptor stimulation, measured by microdialysis. J. Neurochem. 66, 2373–2380. Traynelis, S.F., Wahl, P., 1997. Control of rat GluR6 glutamate receptor open probability by protein kinase A and calcineurin. J. Physiol., London 503, 513–531. Varga, V., Janáky, R., Takáts, A., Oja, S.S., Dohovics, R., Rechardt, L., 1996. Interaction of ␥-l-glutamyltaurine with kainate-induced cyclic AMP formation in the rat hippocampus. Adv. Exp. Med. Biol. 403, 473–479. Vizi, E.S., 2000. Role of high-affinity receptors and membrane transporters in non-synaptic communication and drug action in the central nervous system. Pharmacol. Rev. 52, 63–89. Wang, L.-Y., Taverna, F.A., Huang, X.-P., MacDonald, J.F., Hampson, D.R., 1993. Phosphorylation and modulation of a kainate receptor (GluR6) by cAMP-dependent protein kinase. Science 259, 1173–1175. Zhou, L.M., Gu, Z.Q., Costa, A.M., Yamada, K.A., Mansson, P.E., Giordano, T., Skolnick, P., Jones, K.A., 1997. The (2S, 4R)-4-methylglutamic acid (SYM 2081): a selective, high-affinity ligand for kainate receptors. J. Pharmacol. Exp. Therap. 280, 422–427.