Role of metabotropic glutamate receptor subclasses in modulation of adenylyl cyclase activity by a nootropic NS-105
Abstract
The involvement of metabotropic glutamate ŽmGlu. receptors in the modulatory actions of a novel cognition enhancer, Žq.-5-oxo-D- prolinepiperidinamide monohydrate ŽNS-105., on adenylyl cyclase activity in rat cerebrocortical membranes and primary neuronal cultures was investigated using selective antagonists and antisense oligodeoxynucleotides for mGlu receptor subclasses. In rat cerebrocortical membranes, the inhibitory action of NS-105 Ž0.1 mM. on forskolin-stimulated cAMP formation was blocked by a group II mGlu receptor antagonist, Ž”.-a-ethylglutamic acid, and by a group III antagonist, Žq.-2-amino-2-methyl-4-phosphonobutanoic acid ŽMAP-4., but not by a group I antagonist, Ž”.-1-aminoindan-1,5-dicarboxylic acid ŽAIDA., whereas the facilitation of cAMP formation by NS-105 Ž1 mM. in pertussis toxin-pretreated membranes was abolished by AIDA but not by Ž”.-a-ethylglutamic acid or MAP-4. In primary cultured neurons of mouse cerebral cortex, the inhibitory action of NS-105 on adenylyl cyclase activity disappeared after treatment with antisense oligodeoxynucleotides for group II ŽmGlu2 and mGlu3 receptors. and group III ŽmGlu4 and mGlu7 receptors. but not group I ŽmGlu5 receptor. mGlu receptor subclasses. These findings suggest that the inhibitory action of NS-105 on adenylyl cyclase activity is mediated through group II and group III mGlu receptor subclasses while the facilitatory action is dependent on the group I mGlu receptor subclass.
Keywords: NS-105; Glutamate receptor, metabotropic; Adenylyl cyclase; Antisense oligodeoxynucleotide, knockdown; Primary neuronal culture
1. Introduction
Žq.-5-Oxo-D-prolinepiperidinamide monohydrate ŽNS-105. is a novel cognition enhancer with potent anti-de-been shown that spatial learning in the Morris water maze task is impaired in type I adenylyl cyclase mutant mice ŽWu et al., 1995.. Moreover, long-term potentiation, which is considered to be an electrophysiological model that pressant activity ŽShimidzu et al., 1997.. This compound shows anti-amnesic actions in a variety of animal models of dementia, including those induced by the dysfunction of central cholinergic neurons ŽNakagawa et al., 1988; Oga- sawara et al., 1999., although the cellular mechanisms underlying the cognition-enhancing action are not fully understood.
Several lines of evidence have shown that the intra- cellular cAMP-signal transduction pathway is involved in the processes of learning and memory ŽFrey et al., 1993; Weisskopf et al., 1994.. Stimulation of Gs-coupled recep- tors such as b-adrenoceptors in the rat brain is reported to reflects certain types of learning and memory ŽBliss and Collingridge, 1993., is perturbed in the hippocampal CA1 subfield of these mutant mice ŽWu et al., 1995.
We have recently found that NS-105 modulates adeny- lyl cyclase activity in the rat brain as well as in primary cultured neurons of the mouse cerebral cortex ŽOka et al., 1997a,b.. Namely, it inhibited forskolin-stimulated cAMP formation in a pertussis toxin-sensitive G protein-depen- dent manner, while it enhanced cAMP formation after pretreatment of membranes with pertussis toxin in a cholera toxin-sensitive G protein-dependent fashion. In addition, both of these actions were blocked by a non-selective aminocyclopentane-1,3-dicarboxylic acid Ž1S,3 R-ACPD.
2.2. Membrane preparation from rat cerebral cortex
inhibited forskolin-stimulated cAMP formation in cultured neurons ŽPre´zeau et al., 1994. and cells expressing recom- binant mGlu receptors ŽTanabe et al., 1992. in a manner dependent on pertussis toxin-sensitive G proteins. More- over, the stimulation of mGlu receptors increased cAMP 1994. and cells expressing certain types of mGlu receptors ŽAramori and Nakanishi, 1992.. Both pharmacological and expression cloning studies have demonstrated the existence of different subclasses of mGlu receptors. To date, eight genes encoding mGlu receptors have been cloned ŽNakanishi, 1992; Tanabe et al., 1992; Simoncini et al., 1993; Pin and Duvoisin, 1995., and multiplicity in this receptor family is augmented by the existence of splice variants ŽPin et al., 1992; Minakami et al., 1993; Pickering et al., 1993.. Based on their similarity of amino acid sequences and pharmacological profiles, mGlu receptors supernatant was further centrifuged at 30,000 = g for 20 min. The pellet was washed twice with buffer A and used for cAMP assay. In a set of experiments, membranes were treated with pertussis toxin, as described previously ŽOka et al., 1997a.. Briefly, the pellet of cerebrocortical mem- branes was suspended in ADP-ribosylating buffer contain- ing 50 mM Tris–HCl ŽpH 7.4., 1 mM disodium dihydro- gen ethylenediamine tetraacetate ŽEDTA., 1 mM dithio- threitol, 1 mM MgCl2 , 1 mM adenosine triphosphate ŽATP., 10 mM thymidine and 10 mM nicotine adenine dinucleotide ŽNAD.. Seventy-five micrograms of membrane protein was incubated at 378C for 1 h with 20 mgrml pertussis toxin pre-activated by incubation at 308C for 10 min with 50 mM dithiothreitol.
2.3. Assay of cAMP formation in membrane fraction of group II ŽmGlu2 and mGlu3 receptors. or group III
ŽmGlu4 , mGlu6 , mGlu7 and mGlu8 receptors. mGlu receptors results in the inhibition of adenylyl cyclase activity measured as follow. The reaction mixture, which contained via pertussis toxin-sensitive G proteins ŽTanabe et al., 1992, 1993; Wu et al., 1998.Therefore, the present study was designed to clarify which subclasses or subgroups of mGlu receptors partici- pate in the modulatory action of NS-105 on adenylyl cyclase activity. For this purpose, the effects of various subclass-selective mGlu receptor antagonists on NS-105- induced modulation of adenylyl cyclase activity were ex- amined in rat brain. Furthermore, the effect of NS-105 on adenylate cyclase activity was investigated in the antisense knockdown of mGlu receptor subclasses in primary cul- tured neurons from the mouse cerebral cortex.
2. Materials and methods
2.1. Animals
Seven-week-old male Wistar rats ŽCharles River Japan, Kanagawa. and pregnant ddY strain mice were used. Rats were housed in groups of five to six in a room controlled at 21–258C and 45–65% humidity and maintained under a 12-h lightrdark cycle Žlights automatically on at 0800 h.10 mM forskolin, 2 mM MgCl2 , 1 mM dithiothreitol, 0.3 mM EDTA, 10 mM guanosine triphosphate ŽGTP., 10 mM ATP, 1 mM 3-isobutyl-1-methylxanthine ŽIBMX., 50 mM phosphocreatine and 50 Urml creatine phosphokinase in 500 ml of 50 mM HEPES–NaOH ŽpH 7.4., was incubated for 15 min at 378C with membrane fraction Ž100–200 mg of protein. in the presence or absence of NS-105 and various mGlu receptor antagonists. The reaction was termi- nated by addition of an equal volume of ice-cold 0.1 N HCl. After centrifugation at 30,000 = g for 20 min, the cAMP content in the supernatant was assayed using the cAMP enzyme immunoassay system ŽAmersham; Buck- inghamshire, UK.. Simultaneously, the resulting pellet was suspended and its protein content was measured by the method of Bradford Ž1976. using bovine serum albumin as a standard.
2.4. Primary neuronal cultures
Primary neuronal cultures were prepared from the cere- bral cortex of fetal mice, as described previously ŽOhkuma et al., 1986; Hirouchi et al., 1992.. Briefly, the neopal- lidum was dissected from the 15-day-old fetal mouse brain, and the meninges were carefully removed. Tissues were minced and trypsinized in Ca2q-free Puck’s solution at 378C for 5 min and then triturated with a Pasteur pipette. The cell suspension was centrifuged at 900 = g for 2 min. The resultant pellet was resuspended in Dulbecco’s modi- fied Eagle medium ŽDMEM., followed by filtration through a nylon sieve Žmesh size: 60 mm.. The dissociated cells 1282–1263.; mGlu receptor FW: 5X-TGA GCT ACG TGC TGC TGG CG-3X Ž1943–1962., RV: 5X-TGT CGG CTG ACT GTG AGG TG-3X Žantisense: 2509–2490.; mGlu receptor FW: 5X-GTC TCC TGA TGT CAA GTG GT-3X Ž1254–1273., RV: 5X-GGA CCA CAC TTC GTC ATC AT-3X Žantisense: 1767–1748.; mGlu receptor FW: 5X-CGC TAT GAC TTC TTC TCT CG-3X Ž968–987., RV:
dishes precoated with poly-L-lysine, and maintained for 3 days in DMEM supplemented with 15% fetal bovine serum. On the fourth day of culture, the cortical cells were exposed to DMEM containing 20 mM cytosine arabinoside and 15% fetal bovine serum for 24 h to prevent the growth of non-neuronal cells. Thereafter, the cells were continu- ously cultured in DMEM containing 15% fetal bovine serum.
2.5. Assay of cAMP accumulation in primary neurons
The cAMP accumulation in cultured neurons was mea- sured as described previously ŽWeiss et al., 1985; Oka et al., 1997b.. On the 10th day of culture, neuronal cells were washed and preincubated for 10 min at 378C Ž5% CO2r95% air mixture. with HEPES-buffered saline Ž140 mM NaCl, 4.7 mM KCl, 2.2 mM CaCl2 , 1.2 mM KH 2 PO4 , 11 mM glucose and 15 mM HEPES: pH 7.4.. The neuronal cells were incubated for 10 min with 1 mM NS-105 in the presence of 1 mM forskolin and 1 mM IBMX. The reaction was terminated by aspiration of the incubation medium followed by addition of 1 ml of ice-cold 0.1 N HCl. Cells were scraped off with a rubber policeman. After centrifugation at 10,000 = g for 10 min, the cAMP content in the supernatant was determined as described above.
2.6. ReÕerse transcription–polymerase chain reaction (RT–PCR) of mGlu receptor subtypes in primary neurons
1590–1571.; b-actin FW: 5X-TGG TGG GTA TGG GTC AGA AGG ACT C-3X Ž131–155., RV: 5X-CAT GGC TGG GGT GTT GAA GGT CTC A-3X Žantisense: 396–372.x.The reaction of RT–PCR was performed using Takara EX Taq polymerase and a Takara RNA PCR kit ŽAMV ver. 2.1, Kyoto, Japan.. The PCR conditions were 25 cycles of 30 s denaturation at 948C, 30 s annealing at 548C, and 60 s extension at 728C, after preheating at 948C for 2 min. PCR products were electrophoresed on 1.5% agarose gel and detected by ethidium bromide staining. The bands of PCR product were analyzed by NIH Image Žver. 1.6.. In this study, the reaction was performed under the experimental condition that both the content of tem- plate mRNA and the number of amplification cycles were sufficient for a linear increase in amplified product.
2.7. Treatment with antisense oligodeoxynucleotides for mGlu receptor subclasses of primary neurons
In order to clarify the involvement of each mGlu recep- tor subclass, several antisense oligodeoxynucleotides for mGlu receptor were used in the primary neuronal cells. Namely, the knockdown of group I mGlu receptor ŽmGlu5 receptor., group II mGlu receptor ŽmGlu2 and mGlu3
receptors. and group III mGlu receptor ŽmGlu4 and mGlu7 receptors. were examined using antisense oligodeoxynu- cleotides synthesized by phosphorothioate residues. The antisense sequences were indicated as follows: 5X-CCC AAG CAG TGA TTC CAT-3X ŽmGlu receptor: 209–192.;
receptor subclasses.
The antisense oligodeoxynucleotide for each mGlu re- ceptor subtype was added at 2 mgr35-mm dish. The antisense oligodeoxynucleotides were premixed with lipo- fectin reagent ŽLife Technologies, Rockville, MD, USA. in DMEM, which is devoid of antibiotics and serum, accord- ing to the instruction manual. On the 7th day of cultured, neuronal cells were washed with phosphate-buffered saline and treated with this mixture in antibiotic-free DMEM containing 15% fetal bovine serum ŽChiang et al., 1991; Chen et al., 1996.. After incubation for 3 h at 378C, the neuronal cells were washed with phosphate-buffered saline and further cultured in normal growth medium. On the 3rd day after application of the antisense oligodeoxynu- cleotides, neurons were used for the cAMP assay and RT–PCR analysis.
2.8. Statistical analysis
Data on cAMP formation were analyzed by using the SAS program ŽSASrSTAT, ver. 6, 4th edition, 1990, SAS
3. Results
3.1. Effects of Õarious antagonists of mGlu receptor sub- classes on NS-105-induced modulation of adenylyl cyclase actiÕity in membrane fractions from rat cerebral cortex
The effects of various antagonists of mGlu receptor subclasses on NS-105-induced inhibitory and facilitatory actions of adenylyl cyclase activity were examined in membrane fractions from the rat cerebral cortex. As shown in Fig. 1, NS-105 Ž0.1 mM. inhibited forskolin-stimulated cAMP formation in membrane fractions of the rat cerebral cortex. This inhibition was counteracted by 100 mM MCPG, an antagonist of group I and group II mGlu receptor ŽEaton et al., 1993., and by 100 mM Ž”.-a-ethyl- glutamic acid, a selective antagonist of group II mGlu receptor ŽJane et al., 1996.. In addition, 100 mM MAP-4, a selective antagonist of group III mGlu receptor ŽJane et al., 1994., also reversed the NS-105-induced inhibition. How- ever, 10 mM AIDA, a selective antagonist of group I mGlu receptor ŽPellicciari et al., 1995., failed to block NS-105- induced inhibition of cAMP formation. The concentrations of NS-105 for producing a consistent inhibition of adeny- lyl cyclase activity Ž0.01–0.1 mM. were 10 times lower than those for inducing stimulation Ž0.1–1 mM. according to our previous report ŽOka et al., 1997a.. Therefore, in our present study, the concentrations of various mGlu receptor antagonists used for antagonizing the inhibitory action of NS-105 Ž0.1 mM. were also 10 times lower than those for blocking the facilitatory action of NS-105 Ž1 mM..
4. Discussion
We have previously reported, for membranes of the rat cerebral cortex, that NS-105 inhibited forskolin-stimulated cAMP formation, but it markedly increased cAMP forma- tion in membranes pretreated with pertussis toxin ŽOka et al., 1997a.. Similar bidirectional actions of NS-105 on adenylyl cyclase activity were observed in primary cul- tured neurons of the mouse cerebral cortex ŽOka et al., 1997b.. It is probable that both the inhibitory and facilita- tory actions of NS-105 are mediated by mGlu receptors, since both actions of NS-105 were blocked by AP-3 and mimicked by 1S,3 R-ACPD, a mGlu receptor agonist ŽOka et al., 1997a.. However, a lack of selectivity of AP-3 for mGlu receptors has been claimed by several investigators ŽMistry et al., 1996..
Multiple mGlu receptor subtypes have been cloned and a variety of agonists and antagonists selective for each mGlu receptor subclass or subgroup are currently avail- able. Thus, to confirm our previous findings and to further examined in membranes of rat cerebral cortex. The in- hibitory action of NS-105 on forskolin-stimulated cAMP formation was blocked by the group II selective antagonist, Ž”.-a-ethylglutamic acid, and by MAP-4, a group III mGlu receptor antagonist, but not by AIDA, an antagonist of group I mGlu receptor. In our previous study, this inhibition by NS-105 in brain membranes and cultured neurons was dependent on GirGo proteins, since pretreat- ment with pertussis toxin completely blocked the in- hibitory action ŽOka et al., 1997a,b.. It has been demon- strated that both group II and group III mGlu receptors are coupled to pertussis toxin-sensitive G proteins and nega- tively regulate adenylyl cyclase activity ŽTanabe et al., 1992, 1993.. Taken together, NS-105 may inhibit adenylyl cyclase activity by acting on group II and group III mGlu receptorsrGi protein system.
The involvement of mGlu receptor in the modulatory action of NS-105 on adenylyl cyclase activity was further determined. In the primary neuronal cells of the mouse cerebral cortex, the expression of mRNAs for mGlu2 , receptor subclasses was attempted using primary neurons from the mouse cerebral cortex.
Knockdown by using specific antisense oligodeoxynu- cleotides for certain neurotransmitter receptors has been widely used for analyzing the function of receptors ŽGhelardini et al., 1999.. A study of the knockdown of specific mGlu receptor subtypes has also been performed in vivo ŽDorri et al., 1997.. In this study, treatment with antisense oligodeoxynucleotides specific for each mGlu receptor subclass selectively decreased the expression of the mRNA for the corresponding mGlu receptor subclass. However, we did not determine the changes in protein level for mGlu receptor subtypes, since specific antibodies or radiolabeled ligands for mGlu receptor subclasses are not currently available. In the case of the a1B-adrenocep- tor, it has been shown that treatment with antisense oligodeoxynucleotide weakly reduces the expression of a1B-adrenoceptor mRNA, while it showed the clear knock- down of this protein level ŽGonzalez-Cabrera et al., 1998.. In the present study, the lack of functional mGlu recep- tors in the antisense oligodeoxynucleotide-transfected neu- rons was confirmed as follows. DCG IV, a selective agonist for group II mGlu receptor, no longer inhibited forskolin-stimulated cAMP accumulation in neurons trans- fected with antisense oligodeoxynucleotides for group II mGlu receptor, while this mGlu receptor agonist signifi- cantly inhibited the cAMP accumulation in normal cells and those transfected with antisense oligodeoxynucleotides for group I or group III mGlu receptors. In addition, the knockdown of group II mGlu receptor was specific for the antisense sequence, since the treatment of sense oligodeoxynucleotide had no influence on the DCG IV-in- duced inhibition of forskolin-stimulated cAMP accumula- tion. In these neurons, the inhibitory action of NS-105 disappeared in neurons deficient in mRNAs for group II or group III mGlu receptors but not in neurons lacking group I mGlu receptor mRNA. Therefore, it is suggested that group II andror group III mGlu receptors are major targets of NS-105 for inhibiting adenylyl cyclase activity. Unfor- tunately, in the present study, we could not determine the effects of antisense oligodeoxynucleotides on NS-105-in- duced facilitation of cAMP accumulation in pertussis toxin-pretreated neurons because of the neuronal damage due to the pertussis toxin treatment after application of antisense oligodeoxynucleotides.
In both studies using specific antagonists and antisense oligodeoxynucleotides for mGlu receptor subclasses, the inhibitory action of NS-105 was almost completely blocked by the group II mGlu receptor antagonist, Ž”.-a-ethyl- glutamic acid ŽFig. 1. or prior treatment with antisense oligodeoxynucleotides for group II mGlu receptor ŽFig. 6., although the antagonist ŽMAP-4. or the antisense oligodeoxynucleotides for group III mGlu receptor par- tially attenuated the action of NS-105. Therefore, it is likely that the inhibitory action of NS-105 on adenylyl mGlu receptor. However, we do not know at present why NS-105 did not cause any suppressive action on adenylyl cyclase activity by acting on the remaining group such as group III mGlu receptor in tissues or neurons whose group II mGlu receptor was blocked or disappeared. It seems likely that group III mGlu receptor may also participate in the inhibitory action of NS-105, provided that group II mGlu receptor is activated by this compound.
In conclusion, NS-105 inhibited forskolin-stimulated cAMP formation in rat cerebrocortical membranes as well as primary neuronal cultures of the mouse cerebral cortex, and concurrently enhanced cAMP formation in pertussis toxin-pretreated preparations. A pharmacological study us- ing selective antagonists of mGlu receptor subgroups demonstrated that the inhibitory action of NS-105 was mediated by stimulation of group II and group III mGlu receptors. Consistent with these findings, NS-105 no longer decreased forskolin-stimulated cAMP accumulation in primary cultured neurons pretreated with antisense oligodeoxynucleotides for group II or group III mGlu receptor mRNAs. In contrast, the enhancement of forskolin-stimulated cAMP formation by NS-105 observed in pertussis toxin-pretreated membranes was due to the stimulation of group I mGlu receptor NS 105 but not group II or group III mGlu receptor.