Feed-forward facilitation of glutamate release by presynaptic GABA(A) receptors

Disynaptic GABAergic inputs from Schaffer collateral (SC) afferents on to the soma of glutamatergic CA1 pyramidal neurons are involved in feed-forward inhibition in the hippocampal neural circuits. Here we report the functional roles of presynaptic GABA(A) receptors on SC afferents projecting to CA1 pyramidal neurons. Muscimol (0.5 microM), a selective GABA(A) receptor agonist, increased SC-evoked EPSC amplitude and decreased paired-pulse ratio in the slice preparation, in addition, it facilitated spontaneous glutamate release on to mechanically dissociated CA1 pyramidal neurons in an external Ca2+-dependent manner. In field recordings, muscimol at low concentrations (< or = 0.5 microM) increased not only the excitability of SC afferents but glutamate release, however, it at high concentrations (> or = 1 microM) changed bidirectionally. These results suggest that the moderate activation of presynaptic GABA(A) receptors depolarizes SC afferents and enhances SC-mediated glutamatergic transmission. When endogenous GABA was disynaptically released by brief trains of stimulation of SC afferents, the axonal excitability in addition to glutamate release was increased. The effects of endogenous GABA on the excitability of SC afferents were blocked by either SR95531 or AMPA receptor blockers, which would be expected to block disynaptic feed-forward neural circuits. The present results provide a novel form of presynaptic modulation (feed-forward facilitation) of glutamatergic transmission by presynaptic GABA(A) receptors within the intrinsic hippocampal neural circuits.     

Is GABA release modulated by presynaptic excitatory amino acid receptors?

The purely GABAergic nature of spontaneous synaptic activity in cultures from the neonatal rat superior colliculus (SC) is of great advantage in investigations aimed at characterizing presynaptic factors regulating GABAergic synaptic transmission. Using SC-derived cultures it was confirmed that excitatory amino acids (EAA) can induce a marked increase in the frequency of spontaneous synaptic Cl- currents (ICl(GABA)SYN). However, this tetrodotoxin-resistant facilitation of Ca2(+)-dependent GABA release required application of EEA to several neurons (multiple cell superfusion). In contrast, no frequency increase of Icl(GABA)SYN was seen with restricted access of EAA to only one neuron and the presynaptic axonal terminals (single cell superfusion). It is therefore concluded that the strong facilitatory effect of glutamate (Glu) and kainate (KA) on GABAergic synaptic activity, as observed under the condition of multiple cell superfusion, is mediated via somatodendritic excitatory amino acid receptors (EAARs).     

5-HT1B receptor-mediated presynaptic inhibition of GABA release in the suprachiasmatic nucleus

The suprachiasmatic nucleus (SCN) receives a dense serotonergic innervation that modulates photic input to the SCN via serotonin 1B (5-HT1B) presynaptic receptors on retinal glutamatergic terminals. However, the majority of 5-HT1B binding sites in the SCN are located on nonretinal terminals and most axonal terminals in the SCN are GABAergic. We therefore tested the hypothesis that 5-HT1B receptors might also be located on SCN GABAergic terminals by examining the effects of the highly selective 5-HT1B receptor agonist CP-93,129 on SCN miniature inhibitory postsynaptic currents (mIPSCs). Whole cell patch-clamp recordings of mIPSCs were obtained from rat and mouse SCN neurons in hypothalamic slices. Using CsCl-containing microelectrodes with QX314, we isolated mPSCs that were sensitive to the GABAA receptor antagonist, bicuculline. Bath application of CP-93,129 (1 microM) decreased the frequency of mIPSCs by an average of 22% (n = 7) in rat SCN neurons and by an average of 30% (n = 8) in mouse SCN neurons with no clear effect on mIPSC amplitude. In mice lacking functional 5-HT1B receptors, CP-93,129 (1 microM) had no clear effect on the frequency or the amplitude of mIPSCs recorded in any of the cells tested (n = 4). The decrease in the frequency of mIPSCs of SCN neurons produced by the selective 5-HT1B receptor agonist CP-93,129 is consistent with the interpretation that 5-HT1B receptors are located on GABA terminals in the SCN and that 5-HT inhibits GABA release via a 5-HT1B presynaptic receptor-mediated mechanism.     

Kainate acts at presynaptic receptors to increase GABA release from hypothalamic neurons.

Recent reports suggest that kainate acting at presynaptic receptors reduces the release of the inhibitory transmitter GABA from hippocampal neurons. In contrast, in the hypothalamus in the presence of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) receptor antagonists [1-(4-methyl-7,8-methylenedioxy-5H-2,3-benzodiazepine (GYKI 52466) and D,L-2-amino-5-phosphonopentanoic acid (AP5)], kainate increased GABA release. In the presence of tetrodotoxin, the frequency, but not the amplitude, of GABA-mediated miniature inhibitory postsynaptic currents (IPSCs) was enhanced by kainate, consistent with a presynaptic site of action. Postsynaptic activation of kainate receptors on cell bodies/dendrites was also found. In contrast to the hippocampus where kainate increases excitability by reducing GABA release, in the hypothalamus where a much higher number of GABAergic cells exist, kainate-mediated activation of transmitter release from inhibitory neurons may reduce the level of neuronal activity in the postsynaptic cell.     

Corticocortical inhibition of peripheral inputs within primary somatosensory cortex: the role of GABA(A) and GABA(B) receptors

A conditioning-test pulse paradigm was used in combination with microiontophoresis to examine the corticocortical modulation of somatosensory processing. Single-cell recordings were made in the glabrous digit representation of primary somatosensory (S1) cortex in anesthetized raccoons. Test stimulation of the periphery (the on-focus digit) was preceded by conditioning stimulation of the cortical area that represents an adjacent digit at interstimulus intervals ranging from 5 to 200 ms. An early and prolonged inhibitory modulation was produced in most of the 61 neurons examined, and an early facilitation followed by inhibition was produced in about one-third of the cells. Microiontophoretic administration of a potent GABA(B) receptor antagonist, CGP 55845, blocked the inhibition and in many cases revealed a facilitation of the sensory response. Microiontophoretic administration of a GABA(A) receptor antagonist, gabazine, blocked inhibition at short interstimulus intervals and reduced the longer inhibition by half. These results indicate that connections between glabrous digit representations within S1 cortex produce predominantly inhibitory modulation of sensory input and that both GABA(A) and GABA(B) receptors contribute to this modulation. The relevance of these connections to the effects of peripheral nerve injury and subsequent reorganization is discussed.     

Zinc-mediated inhibition of GABA(A) receptors: discrete binding sites underlie subtype specificity

Zinc ions are concentrated in the central nervous system and regulate GABA(A) receptors, which are pivotal mediators of inhibitory synaptic neurotransmission. Zinc ions inhibit GABA(A) receptor function by an allosteric mechanism that is critically dependent on the receptor subunit composition: alphabeta subunit combinations show the highest sensitivity, and alphabetagamma isoforms are the least sensitive. Here we propose a mechanistic and structural basis for this inhibition and its dependence on the receptor subunit composition. We used molecular modeling to identify three discrete sites that mediate Zn2+ inhibition. One is located within the ion channel, and the other two are on the external amino (N)-terminal face of the receptor at the interfaces between alpha and beta subunits. We found that the characteristically low Zn2+ sensitivity of GABA(A) receptors containing the gamma2 subunit results from disruption to two of the three sites after receptor subunit co-assembly.     

Retrograde activation of presynaptic NMDA receptors enhances GABA release at cerebellar interneuron-Purkinje cell synapses

Synaptic inhibition is a vital component in the control of cell excitability within the brain. Here we report a newly identified form of inhibitory synaptic plasticity, termed depolarization-induced potentiation of inhibition, in rodents. This mechanism strongly potentiated synaptic transmission from interneurons to Purkinje cells after the termination of depolarization-induced suppression of inhibition. It was triggered by an elevation of Ca(2+) in Purkinje cells and the subsequent retrograde activation of presynaptic NMDA receptors. These glutamate receptors promoted the spontaneous release of Ca(2+) from presynaptic ryanodine-sensitive Ca(2+) stores. Thus, NMDA receptor-mediated facilitation of transmission at this synapse provides a regulatory mechanism that can dynamically alter the synaptic efficacy at inhibitory synapses.     

Hypothesis:presynaptic receptors controlling renin release.

Presynaptic receptors have recently been identified on nerves supplying vascular smooth muscle. Renin is stored in juxtaglomerular cells in the wall of the afferent glomerular arteriole and the JG cell is probably derived from vascular smooth muscle. Nervous stimuli exert an important influence on release of renin, but the nature of the receptors involved is not agreed. We propose that there are presynaptic receptors associated with the JG apparatus as with vascular muscle and that a mechanism of this sort gives a better account of data on renin release.     

Postsynaptic and presynaptic GABA(B) receptor-mediated inhibition in rat superficial superior colliculus neurons

It has been reported that GABA(B) receptors are abundantly expressed in the superficial superior colliculus (sSC). To elucidate the action of GABA(B) receptors in the sSC, we tested the effect of a specific GABA(B) receptor agonist baclofen on sSC neurons in slices obtained from rats (16-22 days postnatal) using the whole-cell patch clamp technique. Bath application of baclofen (10-30 microM) elicited hyperpolarization associated with an increase in potassium conductance in morphologically different types of neurons including putative excitatory and inhibitory neurons. In addition, baclofen presynaptically suppressed electrically induced excitatory postsynaptic currents and inhibitory postsynaptic currents. These effects of baclofen were antagonized by a specific GABA(B) receptor antagonist CGP55845A (1 microM). The results suggest that the GABA(B) receptors have an inhibitory action at both post- and pre-synaptic sites in the sSC.     

Drug interactions at GABA(A) receptors

Neurotransmitter receptor systems have been the focus of intensive pharmacological research for more than 20 years for basic and applied scientific reasons, but only recently has there been a better understanding of their key features. One of these systems includes the type A receptor for the gamma-aminobutyric acid (GABA), which forms an integral anion channel from a pentameric subunit assembly and mediates most of the fast inhibitory neurotransmission in the adult vertebrate central nervous system. Up to now, depending on the definition, 16-19 mammalian subunits have been cloned and localized on different genes. Their assembly into proteins in a poorly defined stoichiometry forms the basis of functional and pharmacological GABA(A) receptor diversity, i.e. the receptor subtypes. The latter has been well documented in autoradiographic studies using ligands that label some of the receptors' various binding sites, corroborated by recombinant expression studies using the same tools. Significantly less heterogeneity has been found at the physiological level in native receptors, where the subunit combinations have been difficult to dissect.This review focuses on the characteristics, use and usefulness of various ligands and their binding sites to probe GABA(A) receptor properties and to gain insight into the biological function from fish to man and into evolutionary conserved GABA(A) receptor heterogeneity. We also summarize the properties of the novel mouse models created for the study of various brain functions and review the state-of-the-art imaging of brain GABA(A) receptors in various human neuropsychiatric conditions.The data indicate that the present ligands are only partly satisfactory tools and further ligands with subtype-selective properties are needed for imaging purposes and for confirming the behavioral and functional results of the studies presently carried out in gene-targeted mice with other species, including man.     

Activation of GABA(A) receptors in subthalamic neurons in vitro: properties of native receptors and inhibition mechanisms.

The subthalamic nucleus (STN) influences the output of the basal ganglia, thereby interfering with motor behavior. The main inputs to the STN are GABAergic. We characterized the GABA(A) receptors expressed in the STN and investigated the response of subthalamic neurons to the activation of GABA(A) receptors. Cell-attached and whole cell recordings were made from rat brain slices using the patch-clamp technique. The newly identified epsilon subunit confers atypical pharmacological properties on recombinant receptors, which are insensitive to barbiturates and benzodiazepines. We tested the hypothesis that native subthalamic GABA(A) receptors contain epsilon proteins. Applications of increasing concentrations of muscimol, a selective GABA(A) agonist, induced Cl(-) and HCO currents with an EC(50) of 5 microM. Currents induced by muscimol were fully blocked by the GABA(A) receptor antagonists, bicuculline and picrotoxin. They were strongly potentiated by the barbiturate, pentobarbital (+190%), and by the benzodiazepines, diazepam (+197%) and flunitrazepam (+199%). Spontaneous inhibitory postsynaptic currents were also significantly enhanced by flunitrazepam. Furthermore, immunohistological experiments with an epsilon subunit-specific antibody showed that the epsilon protein was not expressed within the STN. Native subthalamic GABA(A) receptors did not, therefore, display pharmacological or structural properties consistent with receptors comprising epsilon. Burst firing is a hallmark of Parkinson's disease. Half of the subthalamic neurons have the intrinsic capacity of switching from regular-firing to burst-firing mode when hyperpolarized by current injection. This raises the possibility that activation of GABA(A) receptors might trigger the switch. Statistical analysis of spiking activity established that 90% of intact neurons in vitro were in single-spike firing mode, whereas 10% were in burst-firing mode. Muscimol reversibly stopped recurrent electrical activity in all intact neurons. In neurons held in whole cell configuration, membrane potential hyperpolarized by -10 mV whilst input resistance decreased by 50%, indicating powerful membrane shunting. Muscimol never induced burst firing, even in neurons that exhibited the capacity of switching from regular- to burst-firing mode. These molecular and functional data indicate that native subthalamic GABA(A) receptors do not contain the epsilon protein and activation of GABA(A) receptors induces membrane shunting, which is essential for firing inhibition but prevents switching to burst-firing. They suggest that the STN, like many other parts of the brain, has the physiological and structural features of the widely expressed GABA(A) receptors consisting of alphabetagamma subunits.     

Muscarinic receptors mediate direct inhibition of GABA release from rat striatal nerve terminals

The effects of acetylcholine (ACh) on the depolarization-evoked release of [3H]gamma-aminobutyric acid ([3H]GABA) have been investigated using synaptosomes prepared from rat corpus striatum and depolarized by superfusion with 9 mM KCl. Acetylcholine inhibited the [3H]GABA overflow in a concentration-dependent manner. The maximal effect was about 50%. The IC50 value (concentration producing half-maximal effect) amounted to 1 microM, in the absence of acetylcholinesterase inhibitors. The effect of ACh on the K(+)-evoked [3H]GABA release was counteracted by the muscarinic receptor antagonist atropine, but not by the nicotinic receptor antagonist mecamylamine or by the selective M1 antagonist pirenzepine. The data show that muscarinic receptors with low affinity for pirenzepine are localized on GABAergic nerve endings in rat corpus striatum where they may directly inhibit the release of GABA.     

Calcium influx through presynaptic 5-HT3 receptors facilitates GABA release in the hippocampus: in vitro slice and synaptosome studies

Serotonin 5-hydroxytryptamine type 3 receptors (5HT3R) are Ca2+-permeant, non-selective cation channels that have been localized to presynaptic terminals and demonstrated to modulate neurotransmitter release. In the present study the effect of 5-HT on GABA release in the hippocampus was characterized using both electrophysiological and biochemical techniques. 5-HT elicited a burst-like, 6- to 10-fold increase in the frequency of GABAA receptor-mediated inhibitory postsynaptic currents (IPSCs) measured with whole-cell voltage-clamp recordings of CA1 neurons in hippocampal slices. When tetrodotoxin was used to block action potential propagation, the 5-HT-induced burst of IPSCs was still observed. Stimulation of hippocampal synaptosomes with 5-HT resulted in a significant increase in the amount of [3H]GABA released by hyperosmotic saline. In both preparations, the 5-HT effect was shown to be mediated by 5HT3Rs, as it was mimicked by the selective 5HT3R agonist m-chlorophenyl biguanide and blocked by the selective 5HT3R antagonist 3-tropanylindole-3-carboxylate hydrochloride. The 5HT3R-mediated increase in GABA release was blocked by 100 microM cadmium or by omitting Ca2+ in external solutions, indicating the Ca2+-dependence of the effect. The high voltage-activated Ca2+ channel blockers omega-conotoxin GVIA and omega-conotoxin MVIIC and 10 microM cadmium had no significant effect on the 5-HT3R-mediated enhancement of GABA release, indicating that Ca2+ influx through the 5-HT3R facilitates GABA release. Taken together, these data provide direct evidence that Ca2+ entry via presynaptic 5HT3Rs facilitates the release of GABA from hippocampal interneurons.     

Peripheral GABA(A) receptors: evidence for peripheral primary afferent depolarization.

We propose that the primary afferent depolarization that follows GABA(A) receptor activation in the spinal cord also occurs in the periphery. As evidence, the present study localizes beta2/beta3 and alpha1 subunits of the GABA(A) receptor on 10-14% of the unmyelinated primary afferents axons in the glabrous skin of the cat paw. Behavioral studies demonstrate that local peripheral injection of the GABA(A) agonist muscimol at a low concentration (2.0 microM) attenuates, and at a high concentration (1 mM) enhances, formalin-induced nociceptive behaviors. Intraplantar injection of muscimol alone at a high dose evokes thermal hyperalgesia. Bicuculline, a GABA(A) antagonist, prevents these muscimol-induced changes in behavior. The muscimol-induced effects are due to local rather than systemic or central activation of GABA(A) receptors, as such effects are not observed in the contralateral paw. We interpret these findings to indicate that activation of GABA(A) receptors by low concentrations of muscimol depolarizes peripheral primary afferent terminals, a phenomenon we call peripheral primary afferent depolarization, in turn reducing the size of the peripheral action potentials and concomitantly reducing the amount of algogenic substances released from the peripheral terminals of these fibers. This sequence of events presumably results in a reduction in nociceptor activation. Higher concentrations of muscimol further depolarize GABA(A) receptor-containing terminals, which then initiates action potentials in nociceptors analogous to the appearance of dorsal root reflexes that arise following activation of GABA(A) receptors on central primary afferent terminals. These latter events reverse the analgesic effects of GABA(A) ligands and lead to potentiation of nociceptive input. Thus, the present study provides anatomical and behavioral evidence supporting a bimodal role for GABA(A) receptors in the modulation of peripheral nociceptive transmission.     

Cocaine inhibition of GABA(A) current: role of dephosphorylation

Acute cocaine toxicity is frequently associated with seizures. The mechanisms underlying the convulsant effect of cocaine are not well understood. Previously, we have shown that cocaine depresses whole-cell current evoked by gamma-aminobutyric acid (GABA) in hippocampal neurons freshly isolated from rats. Cocaine's effect was voltage-independent and concentration-dependent. In the present study, using whole-cell patch-clamp recording on rat neurons freshly isolated from hippocampus, we examined the intracellular mechanisms involved in cocaine's action. Increasing intracellular Ca(2+) concentration ([Ca]i) from 0.01 to 5 microM strongly increased the depressant effect of cocaine. By contrast, 1-[N, O-bis (5-isoquinolinesulfonyl)-N-methyl-L-tyrosyl]-4-phenylpiperazine (KN-62), a specific antagonist of Ca/calmodulin-dependent protein kinase (CaMKII), attenuated or enhanced cocaine's action in different neurons: in three out of nine neurons dialysed with 5 microM KN-62,1 mM cocaine depressed GABA current by only 33%, but in another three out of nine neurons, cocaine depressed GABA current by as much as 83%. Chelerythrine (a specific CaCa(2+)/phospholipid-dependent protein kinase C [PKC] antagonist) had minimal effect on cocaine's action. We suggest that cocaine induces an increase in [Ca]i, which stimulates phosphatase activity and thus leads to dephosphorylation of GABA receptors. This dephosphorylation-mediated disinhibitory action may play a role in cocaine-induced convulsant states.     

Ovarian cycle-linked changes in GABA(A) receptors mediating tonic inhibition alter seizure susceptibility and anxiety

Disturbances of neuronal excitability changes during the ovarian cycle may elevate seizure frequency in women with catamenial epilepsy and enhance anxiety in premenstrual dysphoric disorder (PMDD). The mechanisms underlying these changes are unknown, but they could result from the effects of fluctuations in progesterone-derived neurosteroids on the brain. Neurosteroids and some anxiolytics share an important site of action: tonic inhibition mediated by delta subunit-containing GABA(A) receptors (deltaGABA(A)Rs). Here we demonstrate periodic alterations in specific GABA(A)R subunits during the estrous cycle in mice, causing cyclic changes of tonic inhibition in hippocampal neurons. In late diestrus (high-progesterone phase), enhanced expression of deltaGABA(A)Rs increases tonic inhibition, and a reduced neuronal excitability is reflected by diminished seizure susceptibility and anxiety. Eliminating cycling of deltaGABA(A)Rs by antisense RNA treatment or gene knockout prevents the lowering of excitability during diestrus. Our findings are consistent with possible deficiencies in regulatory mechanisms controlling normal cycling of deltaGABA(A)Rs in individuals with catamenial epilepsy or PMDD.     

Redox modulation of recombinant human GABA(A) receptors

We previously reported that GABA-evoked currents of rat retinal ganglion cells were modulated by redox agents. In this study, we further characterized the effects of redox modulation on GABA receptors using recombinant human subunits in the Xenopus oocyte expression system with two-electrode voltage-clamp recording. GABA receptors composed of subunits alpha(1-3), beta(1-3), gamma(1), gamma(2S,) and rho(1) were expressed. The sulfhydryl reducing agent dithiothreitol reversibly potentiated the responses of various combinations of functional recombinant GABA(A) subunits, whether expressed as triplets (alpha(1)beta(1-3)gamma(1,2S)), pairs (alpha(1-3)beta(1-3); beta(1-3)gamma(1,2S)), or singly (beta(2)). These effects of dithiothreitol were rapidly reversible, and the oxidizing agent 5-5'-dithiobis-2-nitrobenzoic acid exerted the opposite effect. In contrast to these effects on GABA(A) receptors, dithiothreitol had no effect on the responses of homomeric GABA rho(1) (GABA(C)) receptors. The degree of dithiothreitol potentiation of GABA(A) receptor responses depended on subunit composition. Co-expression of gamma(2S) with alpha(1)beta(1-3) subunits resulted in markedly less dithiothreitol potentiation of GABA-evoked currents than that observed for alpha(1-3)beta(1-3) subunits in the absence of gamma(2S). None the less, the magnitude of dithiothreitol potentiation could be restored by using a combination of lower GABA concentrations (5-10 microM) and higher dithiothreitol concentrations (5-20mM). N,N,N', N'-tetrakis(2-pyridyl-methyl)ethylenediamine, a high-affinity Zn(2+) chelator, also potentiated GABA(A) receptor currents. However, the potentiation produced by 10mM dithiothreitol was larger than that produced by saturating concentrations of N,N,N', N'-tetrakis(2-pyridyl-methyl)ethylenediamine (100 microM), implying that at least part of the effect of dithiothreitol was due to redox modulation rather than Zn(2+) chelation. Dithiothreitol also potentiated the spontaneous current of homomeric GABA(A) receptors composed of beta subunits. Mutation of a single cysteine residue in the M3 domain, yielding homomeric beta(3)(C313A) receptors, abrogated dithiothreitol potentiation of the spontaneous current.In summary, this study further characterizes the modulatory effects of redox agents on recombinant GABA(A) receptors. The degree of redox modulation of GABA(A) receptors depended on subunit composition. In contrast to their effect on GABA(A) receptors, redox agents were not found to modulate GABA(C) receptors composed of homomeric rho(1) subunits. Using site-directed mutagenesis, a cysteine residue was located in the beta(3) subunit which may comprise one of the redox-active sites that underlies the modulation of heteromeric GABA(A) receptors by reducing and oxidizing agents.