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

A critical review of the neurovascular nature of migraine and the main mechanisms of action of prophylactic antimigraine medications

INTRODUCTION

Migraine involves neurovascular, functional, and anatomical alterations. Migraineurs experience an intense unilateral and pulsatile headache frequently accompanied with vomiting, nausea, photophobia, etc. Although there is no ideal preventive medication, frequency in migraine days may be partially decreased by some prophylactics, including antihypertensives, antidepressants, antiepileptics, and CGRPergic inhibitors. However, the mechanisms of action involved in antimigraine prophylaxis remain elusive.

AREAS COVERED

This review recaps some of the main neurovascular phenomena related to migraine and currently available preventive medications. Moreover, it discusses the major mechanisms of action of the recommended prophylactic medications.

EXPERT OPINION

In the last three years, migraine prophylaxis has evolved from nonspecific to specific antimigraine treatments. Overall, nonspecific treatments  mainly involve neural actions, whereas specific pharmacotherapy (represented by CGRP receptor antagonists and CGRPergic monoclonal antibodies) is predominantly mediated by neurovascular mechanisms that may include, among others: (i) reduction in the cortical spreading depression (CSD)-associated events; (ii) inhibition of pain sensitization; (iii) blockade of neurogenic inflammation; and/or (iv) increase in cranial vascular tone. Accordingly, the novel antimigraine prophylaxis promises to be more effective, devoid of significant adverse effects (unlike nonspecific treatments), and more beneficial for the quality of life of migraineurs.

Presynaptic AMPA Receptors in Health and Disease

AMPA receptors (AMPARs) are ionotropic glutamate receptors that play a major role in excitatory neurotransmission. AMPARs are located at both presynaptic and postsynaptic plasma membranes. A huge number of studies investigated the role of postsynaptic AMPARs in the normal and abnormal functioning of the mammalian central nervous system (CNS). These studies highlighted that changes in the functional properties or abundance of postsynaptic AMPARs are major mechanisms underlying synaptic plasticity phenomena, providing molecular explanations for the processes of learning and memory. Conversely, the role of AMPARs at presynaptic terminals is as yet poorly clarified. Accruing evidence demonstrates that presynaptic AMPARs can modulate the release of various neurotransmitters. Recent studies also suggest that presynaptic AMPARs may possess double ionotropic-metabotropic features and that they are involved in the local regulation of actin dynamics in both dendritic and axonal compartments. In addition, evidence suggests a key role of presynaptic AMPARs in axonal pathology, in regulation of pain transmission and in the physiology of the auditory system. Thus, it appears that presynaptic AMPARs play an important modulatory role in nerve terminal activity, making them attractive as novel pharmacological targets for a variety of pathological conditions.

Long-Term Shaping of Corticostriatal Synaptic Activity by Acute Fasting

Food restriction is a robust nongenic, nonsurgical and nonpharmacologic intervention known to improve health and extend lifespan in various species. Food is considered the most essential and frequently consumed natural reward, and current observations have demonstrated homeostatic responses and neuroadaptations to sustained intermittent or chronic deprivation. Results obtained to date indicate that food deprivation affects glutamatergic synapses, favoring the insertion of GluA2-lacking α-Ammino-3-idrossi-5-Metil-4-idrossazol-Propionic Acid receptors (AMPARs) in postsynaptic membranes. Despite an increasing number of studies pointing towards specific changes in response to dietary restrictions in brain regions, such as the nucleus accumbens and hippocampus, none have investigated the long-term effects of such practice in the dorsal striatum. This basal ganglia nucleus is involved in habit formation and in eating behavior, especially that based on dopaminergic control of motivation for food in both humans and animals. Here, we explored whether we could retrieve long-term signs of changes in AMPARs subunit composition in dorsal striatal neurons of mice acutely deprived for 12 hours/day for two consecutive days by analyzing glutamatergic neurotransmission and the principal forms of dopamine and glutamate-dependent synaptic plasticity. Overall, our data show that a moderate food deprivation in experimental animals is a salient event mirrored by a series of neuroadaptations and suggest that dietary restriction may be determinant in shaping striatal synaptic plasticity in the physiological state.

Dual contradictory roles of cAMP signaling pathways in hydroxyl radical production in the rat striatum

Studies have suggested that cAMP signaling pathways may be associated with the production of reactive oxygen species. In this study, we examined how modifications in cAMP signaling affected the production of hydroxyl radicals in rat striatum using microdialysis to measure extracellular 2,3-dihydroxybenzoic acid (2,3-DHBA), which is a hydroxyl radical adduct of salicylate. Up to 50 nmol of the cell-permeative cAMP mimetic 8-bromo-cAMP (8-Br-cAMP) increased 2,3-DHBA in a dose-dependent manner (there was no additional increase in 2,3-DHBA at 100 nmol). Another cAMP mimetic, dibutyryl cAMP (db-cAMP), caused a nonsignificant increase in 2,3-DHBA at 50 nmol and a significant decrease at 100 nmol. Up to 20 nmol of forskolin, which is a direct activator of adenylyl cyclase, increased 2,3-DHBA, similar to the effect of 8-Br-cAMP; however, forskolin resulted in a much greater increase in 2,3-DHBA. A potent inhibitor of protein kinase A (PKA), H89 (500 μM), potentiated the 8-Br-cAMP- and forskolin-induced increases in 2,3-DHBA and antagonized the inhibitory effect of 100 nmol of db-cAMP. Interestingly, the administration of 100 nmol of 8-bromo-cGMP alone or in combination with H89 had no significant effect on 2,3-DHBA levels. Doses of 100 nmol of a preferential PKA activator (6-phenyl-cAMP) or a preferential PKA inhibitor (8-bromoadenosine-3',5'-cyclic monophosphorothionate, Rp-isomer; Rp-8-Br-cAMPS), which also inhibits the cAMP-mediated activation of Epac (the exchange protein directly activated by cAMP), suppressed or enhanced, respectively, the formation of 2,3-DHBA. Up to 100 nmol of 8-(4-chlorophenylthio)-2'-O-methyladenosine-cAMP, which is a selective activator of Epac, dose-dependently stimulated the formation of 2,3-DHBA. These findings suggest that cAMP signaling plays contradictory roles (stimulation and inhibition) in the production of hydroxyl radicals in rat striatum by differential actions of Epac and PKA. These roles might contribute to the production of hydroxyl radicals concomitant with cAMP in carbon monoxide poisoning, because the formation of 2,3-DHBA was potentiated by the PKA inhibitor H89 and suppressed by Rp-8-Br-cAMPS, which inhibits PKA and Epac.

Taurine attenuates D-[3H]aspartate release evoked by depolarization in ischemic corticostriatal slices

Taurine is thought to be protective in ischemia due to its neuroinhibitory effects. The present aim was to assess the ability of taurine to attenuate glutamate release evoked by ischemia and to determine which component of this release is affected. The release of preloaded D-[(3)H]aspartate (a non-metabolized analog of glutamate) from superfused murine corticostriatal slices was used as index of glutamate release. Preincubation of corticostriatal slices with 10 mM taurine reduced the D-[(3)H]aspartate release evoked by either chemical ischemia (0.5 mM NaCN in glucose-free medium) or oxygen-glucose deprivation. The taurine uptake inhibitor guanidinoethanesulfonate (5 mM), the glycine receptor antagonist strychnine (0.1 mM) and the GABA(A) receptor antagonist bicuculline (0.1 mM) did not block the taurine effect. To determine which component of ischemia-induced glutamate release is affected by taurine, three pathways of this release were pharmacologically modeled. Unlabeled D-aspartate (0.5 mM) and hypo-osmotic medium (NaCl reduced by 50 mM) evoked D-[(3)H]aspartate release via homoexchange and hypo-osmotic release pathways, respectively. Taurine did not influence these pathways. However, it suppressed the synaptic release of D-[(3)H]aspartate evoked by the voltage-gated sodium channel opener veratridine (0.1 mM). Taurine thus reduces glutamate release under ischemic conditions by affecting the depolarization-evoked component.

Modulation of adenylyl cyclase activity in young and adult rat brain cortex. Identification of suramin as a direct inhibitor of adenylyl cyclase

Adenylyl cyclase (AC) in brain cortex from young (12-day-old) rats exhibits markedly higher activity than in adult (90-day-old) animals. In order to find some possibly different regulatory features of AC in these two age groups, here we modulated AC activity by dithiothreitol (DTT), Fe(2+), ascorbic acid and suramin. We did not detect any substantial difference between the effects of all these tested agents on AC activity in cerebrocortical membranes from young and adult rats, and the enzyme activity was always about two-fold higher in the former preparations. Nevertheless, several interesting findings have come out of these investigations. Whereas forskolin- and Mn(2+)-stimulated AC activity was significantly enhanced by the addition of DTT, increased concentrations of Fe(2+) ions or ascorbic acid substantially suppressed the enzyme activity. Lipid peroxidation induced by suitable combinations of DTT/Fe(2+) or by ascorbic acid did not influence AC activity. We have also observed that PKC- or protein tyrosine kinase-mediated phosphorylation apparently does not play any significant role in different activity of AC determined in cerebrocortical preparations from young and adult rats. Our experiments analysing the presumed modulatory role of suramin revealed that this pharmacologically important drug may act as a direct inhibitor of AC. The enzyme activity was diminished to the same extent by suramin in membranes from both tested age groups. Our present data show that AC is regulated similarly in brain cortex from both young and adult rats, but its overall activity is much lower in adulthood.

Down-regulation of microglial activation may represent a practical strategy for combating neurodegenerative disorders

Chronic neurodegenerative disorders are characterized by activation of microglia in the affected neural pathways. Peroxynitrite, prostanoids, and cytokines generated by these microglia can potentiate the excitotoxicity that contributes to neuronal death and dysfunction in these disorders--both by direct effects on neurons, and by impairing the capacity of astrocytes to sequester and metabolize glutamate. This suggests a vicious cycle in which the death of neurons leads to microglial activation, which in turn potentiates neuronal damage. If this model is correct, measures which down-regulate microglial activation may have a favorable effect on the induction and progression of neurodegenerative disease, independent of the particular trigger or target involved in a given disorder. Consistent with this possibility, the antibiotic minocycline, which inhibits microglial activation, shows broad utility in rodent models of neurodegeneration. Other agents which may have potential in this regard include PPARgamma agonists, genistein, vitamin D, COX-2 inhibitors, statins (and possibly policosanol), caffeine, cannabinoids, and sesamin; some of these agents could also be expected to be directly protective to neurons threatened with excitotoxicity. To achieve optimal clinical outcomes, regimens which down-regulate microglial activation could be used in conjunction with complementary measures which address other aspects of excitotoxicity.

Dopamine-glutamate reciprocal modulation of release and motor responses in the rat caudate-putamen and nucleus accumbens of "intact" animals

Functional interactions between dopaminergic neurotransmission and glutamatergic neurotransmission are well known to play a crucial integrative role in the striatum, the major input structure of the basal ganglia now widely recognized to contribute to the control of motor activity and movements but also to the processing of cognitive and limbic functions. However, the nature of these interactions is still a matter of debate and controversy. This review (1) summarizes anatomical data on the distribution of dopaminergic and glutamatergic receptors in the striatum-accumbens complex, (2) focuses on the dopamine-glutamate interactions in the modulation of each other's release in the striatum-accumbens complex, and (3) examines the dopamine-glutamate interactions in the entire striatum involved in the control of locomotor activity. The effects of dopaminergic and glutamatergic receptor selective agonists and antagonists on dopamine and glutamate release as well on motor responses are analyzed in the entire striatum, by reviewing both in vitro and in vivo data. Regarding in vivo data, only findings from focal injections studies in the nucleus accumbens or the caudate-putamen of "intact" animals are reviewed. Altogether, the available data demonstrate that dopamine and glutamate do not uniformly interact to modulate each others' release and postsynaptic modulation of striatal output neurons. Depending on the receptor subtypes involved, interactions between dopaminergic and glutamatergic transmission vary as a multiple and complex combination of tonic, phasic, facilitatory, and inhibitory properties.

Modification of the synaptic glutamate turnover in the hippocampal tissue exposed to low-frequency, pulsed magnetic fields

The influence of pulsed magnetic fields (PMF) on the release and uptake of glutamate was investigated. While the release was examined using hippocampal slices, synaptosomes were chosen to characterize the uptake process. (3)H-D-aspartate was used as a marker of glutamergic transmission. The pulsed magnetic fields (9-15 mT) applied according to the pattern which induced epileptic discharges in hippocampus amplified and attenuated the release and uptake of glutamate, respectively. However, the magnetic fields which induced an increase in neuronal excitability without concomitant seizures amplified both processes. These results confirm our previous reports and indicate that the glutamergic synapses are the target of magnetic fields action.

Clozapine, ziprasidone and aripiprazole but not haloperidol protect against kainic acid-induced lesion of the striatum in mice, in vivo: Role of 5-HT1A receptor activation

Excessive activation of non-NMDA receptors, AMPA and kainate, contributes to neuronal degeneration in acute and progressive pathologies, possibly including schizophrenia. Because 5-HT(1A) receptor agonists have neuroprotective properties (e.g., against NMDA-induced neurotoxicity), we compared the effects of the antipsychotics, clozapine, ziprasidone and aripiprazole, that are partial agonists at 5-HT(1A) receptor, with those of haloperidol, which is devoid of 5-HT(1A) agonist properties, on kainic acid (KA)-induced striatal lesion volumes, in C57Bl/6N mice. The involvement of 5-HT(1A) receptors was determined by antagonist studies with WAY100635, and data were compared with those obtained using the potent and high efficacy 5-HT(1A) receptor agonist, F13714. Intra-striatal KA lesioning and measurement of lesion volumes using cresyl violet staining were carried out at 48 h after surgery. F13714, antipsychotics or vehicle were administered ip twice, 30 min before and 3 1/2 h after KA injection. WAY100635 (0.63 mg/kg) or vehicle were given sc 30 min before each drug injection. Clozapine (2 x 10 mg/kg), ziprasidone (2 x 20 mg/kg) and aripiprazole (2 x 10 mg/kg) decreased lesion volume by 61%, 59% and 73%, respectively. WAY100635 antagonized the effect of ziprasidone and of aripiprazole but only slightly attenuated that of clozapine. In contrast, haloperidol (2 x 0.16 mg/kg) did not affect KA-induced lesion volume. F13714 dose-dependently decreased lesion volume. The 61% decrease of lesion volume obtained with F13714 (2 x 0.63 mg/kg) was antagonized by WAY100635. WAY100635 alone did not affect lesion volume. These results show that 5-HT(1A) receptor activation protects against KA-induced striatal lesions and indicate that some atypical antipsychotic agents with 5-HT(1A) agonist properties may protect against excitotoxic injury, in vivo.

Presynaptic localization of an AMPA-type glutamate receptor in corticostriatal and thalamostriatal axon terminals

The neostriatum is known to receive glutamatergic projections from the cerebral cortex and thalamic nuclei. Vesicular glutamate transporters 1 and 2 (VGluT1 and VGluT2) are located on axon terminals of corticostriatal and thalamostriatal afferents, respectively, whereas VGluT3 is found in axon terminals of cholinergic interneurons in the neostriatum. In the present study, the postsynaptic localization of ionotropic glutamate receptors was examined in rat neostriatum by the postembedding immunogold method for double labelling of VGluT and glutamate receptors. Immunoreactive gold particles for AMPA receptor subunits GluR1 and GluR2/3 were frequently found not only on postsynaptic but also on presynaptic profiles immunopositive for VGluT1 and VGluT2 in the neostriatum, and GluR4-immunoreactive particles were observed on postsynaptic and presynaptic profiles positive for VGluT1. Quantitative analysis revealed that 27-45% of GluR1-, GluR2-, GluR2/3- and GluR4-immunopositive particles found in VGluT1- or VGluT2-positive synaptic structures in the neostriatum were associated with the presynaptic profiles of VGluT-positive axons. In contrast, VGluT-positive presynaptic profiles in the neostriatum showed almost no immunoreactivity for NMDA receptor subunits NR1 or NR2A/B. Furthermore, almost no GluR2/3-immunopositive particles were observed in presynaptic profiles of VGluT3-positive (cholinergic) terminals that made asymmetric synapses in the neostriatum, or in those of VGluT1- or VGluT2-positive terminals in the neocortex. The present results indicate that AMPA receptor subunits but not NMDA receptor subunits are located on axon terminals of corticostriatal and thalamostriatal afferents, and suggest that glutamate released from these axon terminals controls the activity of the terminals through the presynaptic AMPA autoreceptors.

An increase in cAMP concentration in mouse hippocampal slices exposed to low-frequency and pulsed magnetic fields

Our previous studies revealed that magnetic fields amplified evoked potentials recorded from mouse hippocampal slices. In search for the mechanism of this effect, we evaluated the concentration of cAMP in slices exposed to low-frequency and pulsed magnetic fields. Low-frequency magnetic fields of 15 mT applied at 0.16 Hz for 30 min enhanced the concentration of cAMP almost three-fold. The concentration of cAMP continued to rise through the first hour after turning magnetic fields off, reaching almost a four-fold increase, and then returned to control levels at the end of the second hour. Neither static magnetic fields nor magnetic fields applied with the frequency of 0.5 Hz had any effect on cAMP concentration. The increase in cAMP levels was dependent on the strength of the magnetic field and required the presence of extracellular calcium. A pulsed magnetic field applied with variable intensity (9-15 mT) and in cycles lasting from 5 to 20 min doubled the cAMP concentration. These results support our previous electrophysiological observations and provide biochemical correlates for their interpretation.

The effects of nitric oxide on striatal serotoninergic transmission involve multiple targets: an in vivo microdialysis study in the awake rat

The role of endogenous nitric oxide (NO) in N-methyl-D-aspartate (NMDA)-induced modulation of serotonin (5-HT) release in the striatum of freely moving rats has been studied using microdialysis technique. NMDA-induced increase in 5-HT release was significantly inhibited by selective nitric oxide synthase (nNOS) inhibitor S-methylthiocitrulline (S-Me-TC), ONOO- scavenger L-cysteine (L-cys), and guanylate cyclase (GC) inhibitor 1H[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ). These data suggest that modulation of 5-HT levels is linked to the formation of NO produced by NMDA receptor activation and that endogenously produced NO increases 5-HT concentrations both by stimulating formation of 3'-5'-cyclic monophosphate (cGMP) and conversion of ONOO-.

Ionotropic glutamate receptors regulating labeled acetylcholine release from rat striatal tissue in vitro: possible involvement of receptor modulation in magnesium sensitivity

This study evaluated the role of glutamate ionotropic receptors on the control of [3H]acetylcholine ([3H]ACh) release by the intrinsic striatal cholinergic cells. [3H]-choline previously taken up by chopped striatal tissue and converted to [3H]ACh, was released under stimulation by glutamate, N-methyl-d-aspartate (NMDA), kainate and a-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA). Experiments were conducted in the absence of choline uptake inhibitors or acetylcholinesterase inhibitors. A paradigm of two stimulations was employed, the first in control conditions and the second after 9 min of perfusion with the test agents MK-801, 2-amino-5-phosphonopentanoic acid (AP-5), tetrodotoxin (TTX), 6,7-dinitroquinoxaline-2,3-dione (DNQX), 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo-[f]quinoxaline-7-sulfonamide (NBQX), glycine and magnesium. Our results support that (1) in the absence of Mg2+, NMDA is the most effective agonist to stimulate [3H]ACh release from striatal slices (2) magnesium effectively antagonized kainate and AMPA stimulation suggesting that at least part of the kainate and AMPA effects might be attributed to glutamate release (3) besides NMDA, kainate receptors showed a more direct involvement in [3H]ACh release control based on the smaller dependence on Mg2+ and less inhibition by TTX and (4) stimulation of ionotropic glutamate receptors may induce long lasting biochemical changes in receptor/ion channel function since the effects of TTX and/or Mg2+ ions on [3H]ACh release were modified by previous exposure of the tissue to agonists.

Protein kinase A and calcium/calmodulin-dependent protein kinase II regulate D-[3H]aspartate release in auditory brain stem nuclei

We noted previously that after unilateral cochlear ablation (UCA) in young adult guinea pigs, plastic changes in glutamatergic transmitter release in several brain stem auditory nuclei depended on protein kinase C. In this study, we assessed whether such changes depended on protein kinase A (PKA) and calcium/calmodulin-dependent protein kinase II (CaMKII). The electrically-evoked release of D-[3H]aspartate (D-[3H]Asp) was quantified in vitro as an index of glutamatergic transmitter release in the major subdivisions of the cochlear nucleus (CN) and the main nuclei of the superior olivary complex (SOC). In tissues from intact animals, dibutyryl-cyclic adenosine monophosphate (DBcAMP), a PKA activator, elevated D-[3H]Asp release by 1.9-3.7-fold. The PKA inhibitor, H-89 (2 microM), did not alter the evoked release but blocked the stimulatory effects of DBcAMP. These findings suggested that PKA could positively regulate glutamatergic transmitter release. Seven days after the ablation of one cochlea and its cochlear nerve, the stimulatory effect of DBcAMP remained evident. After 145 postablation days, H-89 blocked the plastic elevations of D-[3H]Asp release in the ipsilateral CN and lateral (LSO) and medial (MSO) superior olive. A CaMKII inhibitor, KN-93, produced similar blocks, suggesting that the postablation plasticities in these nuclei depended on PKA or CaMKII. Both H-89 and KN-93 elevated release in the medial nucleus of the trapezoid body (MNTB) and the contralateral MSO, suggesting that either kinase could be used by endogenous mechanisms in these nuclei to downregulate glutamatergic release.

Cyclic AMP-mediated regulation of striatal glutamate release: interactions of presynaptic ligand- and voltage-gated ion channels and G-protein-coupled receptors

The presynaptic regulation of striatal glutamate transmission was investigated using D-[3H]aspartate and mouse striatal slices. Functional changes in voltage-dependent and glutamate receptor-gated ion channels were elicited by pharmacologically modifying intracellular cyclic AMP formation via G-protein-coupled receptor stimulation. The kainate (KA)-evoked release was potentiated by the stimulatory G-protein (G(s))-coupled beta-adrenoceptor agonist isoproterenol (ISO) in a concentration-dependent manner. This effect was mimicked by the specific calmodulin (CaM) antagonists trifluoperazine and calmidazolium. Tetrodotoxin (TTX), a blocker of Na(+) channels, did not affect the basal release but inhibited to the same degree the releases evoked by kainate alone and by kainate and isoproterenol together. Vinpocetine, a blocker of voltage-dependent Na(+) channels, did not alter the basal or the evoked release. The Na(+) channel activator veratridine enhanced the basal release in a concentration-dependent manner and isoproterenol attenuated this effect. The opposite effects of isoproterenol on the kainate- and veratridine-evoked releases may reflect prevention of the cyclic AMP-protein kinase A (PKA) phosphorylation cascade in striatal glutamatergic signal transduction. In addition, the calmidazolium-induced potentiation of kainate-evoked release was thwarted by LY354740 and L-2-amino-4-phosphonobutanoate, agonists of the inhibitory G-protein (G(i))-coupled metabotropic group II and III glutamate receptors (mGluRs). Vinpocetine, which inhibits the CaM-dependent phosphodiesterase (PDE1), was likewise inhibitory. In turn, selective agonists and antagonists of the G(q)-protein-coupled group I mGluRs and (S)-3,5-dihydroxyphenylglycine (3,5-DHPG) and (RS)-1-aminoindan-1,5-dicarboxylate (AIDA), which modulate the intracellular Ca(2+) levels, did not alter the kainate-evoked release. The beta-adrenoceptor-mediated cyclic AMP accumulation seems to downregulate Na(+) channels but to enhance glutamate release by means of upregulation of kainate receptors. This regulation of presynaptic ligand- and voltage-gated ion channels is affected by the cAMP-protein kinase A-dependent phosphorylation cascade and controlled by G(i)-protein-coupled mGluRs.