Bilateral processing in chemical synapses with electrical 'ephaptic' feedback: a theoretical model

I have developed a detailed biophysical model of the chemical synapse which hosts voltage-dependent presynaptic ion channels and takes into account the capacitance of synaptic membranes. I find that at synapses with a relatively large cleft resistance (e.g., mossy fiber or giant calyx synapse) the rising postsynaptic current could activate, within the synaptic cleft, electrochemical phenomena that induce rapid widening of the presynaptic action potential (AP). This mechanism could boost fast Ca(2+) entry into the terminal thus increasing the probability of subsequent synaptic releases. The predicted difference in the AP waveforms generated inside and outside the synapse can explain the previously unexplained fast capacitance transient recorded in the postsynaptic cell at the giant calyx synapse. I propose therefore the mechanism of positive ephaptic feedback that acts between the postsynaptic and presynaptic cell contributing to the basal synaptic transmission at large central synapses. This mechanism could also explain the supralinear voltage dependence of EPSCs recorded at hyperpolarizing membrane potentials in low extracellular calcium concentration.     

Fast and slow excitation of inhibitory cells in the CA3 region of the hippocampus

Pyramidal cells form excitatory synaptic connections with local inhibitory neurons in the hippocampus. This recurrent synapse plays a crucial stabilizing role in the control of hippocampal activity, since it transforms pyramidal cell population. Using a combination of dual recording from presynaptic and postsynaptic cells and anatomical techniques, we show that these synaptic connections often comprise a single site for liberation of excitatory transmitter. The resulting excitatory postsynaptic potentials (EPSCs) have a fast time course and a similar amplitude to miniature EPSCs recorded in tetrodotoxin and cobalt. In contrast, activation of metabotropic glutamate receptors (mGluRs) by transmitter liberated during repetitive activation of these synapses produces an excitation with a much slower time course. In addition to somatodendritic mGluRs, which excite inhibitory cells, a different species of mGluR is present on inhibitory cell terminals. This mGluR is activated by higher concentrations of the agonist t-1-amino-cyclopentyl–1,3-decarboxylate and acts to reduce γ-aminobutyric acid release. mGluRs, thus, have a dual action to enhance and to depress synaptic inhibition in the hippocampus. © 1995 John Wiley & Sons, Inc.     

Putative synaptic mechanisms of inhibition in Limulus lateral eye

Serotonin (5-HT) perfusion of a thin section of Limulus lateral eye hyperpolarizes retinular and eccentric cell membrane potential, and blocks spike action potentials fired by the eccenteric cell. The indoleamine does not directly affect retinular cell receptor potential or eccenteric cell generator potential in response to light stimuli. LSD perfusion blocks both this inhibitory action of 5-HT and light-evoked, synaptically mediated, lateral inhibition. Iontophoretic application of 5-HT to the synaptic neuropil produces shorter latency and duration and larger amplitude of inhibition than does the perfusion technique. This inhibition is dose dependent; the accompanying inhibitory postsynaptic potential (IPSP) appears to have an equilibrium potential more hyperpolarized than normal resting potential levels of ca. -50 mV. IPSP amplitude is sensitive to extracellular potassium ion concentration: it increases with decreased [K+]0 and decreases with increased [K+]0. LSD blocks the inhibition produced by iontophoretic application of 5-HT. Interaction between light-evoked, natural synaptic transmitter-mediated IPSP's and 5-HT IPSP's suggests a common postsynaptic receptor or transmitter-receptor-permeability change mechanism.     

Phosphatases modulate transmission and serotonin facilitation at synapses: studies with the inhibitor okadaic acid.

We examined the role of phosphatases in synaptic transmission using the permeant phosphatase inhibitor okadaic acid (OA). In the crayfish neuromuscular junction (NMJ), postsynaptic effects including increases in input resistance occurred at doses greater than 5 microM OA. At lower doses (0.5-5 microM) the effects were solely presynaptic and transmitter release increased over three-fold despite small reductions in amplitude and duration of presynaptic action potentials. Potentiating effects of serotonin on transmitter release, which depend on phosphorylation, were increased by OA. Frequency facilitation was reduced but its decay was not affected. In frog NMJs, OA increased spontaneous and evoked release two-fold through presynaptic mechanisms. An inactive analog of OA, OA tetra-acetate, had no effect on transmitter release at frog and crayfish NMJ. Therefore, phosphatases have a strong modulating influence on synaptic transmission.     

Specificity of serotoninergic inhibition in Limulus lateral eye

The receptor specificity for synaptically mediated lateral inhibition in Limulus lateral eye retina was studied by structure-activity correlations of the action of the putative indoleaminergic neurotransmitter, serotonin (5-HT), and its isomers and structural analogs, tryptamine (TRYP), 6-hydroxytryptamine (6HT), 5,6-dihydroxytryptamine (5,6-DHT), 5-hydroxydimethyltryptamine (5-HDMT), and 5-hydroxytryptophan (5-HTP). The 5-HT blockers, lysergic acid diethylamide (LSD), bromo-LSD (BOL), and cinanserin, were also tested. The inhibitory action of the indoleaminergic agonists is highly structure-specific. An hydroxyl group in the 5 position of the indole nucleus, sterically unencumbered by hydroxyls in neighboing positions, is essential. In order of decreasing potency, 5-HT, 5-HDMT, and 5-HTP are active agonists; TRYP, 6-HT, and 5,6-DHT are inactive. Configuration and mobility of the side chains of the active agonists also affect the interaction, and these side-chain characteristics correlate with agonist potency. The receptors for inhibitory action and for transmembranal transport in reuptake are different. Both active agonists and inactive analogs appear to be taken up (Adolph and Ehinger, 1975. Cell Tissue Res. 163:1-14). LSD and BOL have bimodal actions: direct inhibition and agonist blockade. These actions may be mediated via low-specificity presynaptic uptake receptor sites rather than highly specific, postsynaptic, agonist receptor sites.     

Physiological roles of adenosine derivatives which are released during neurotransmission in mammalian brain.

1. Experiments using synaptosome beds suggested that ATP was released from presynaptic sites and degraded to adenosine in the synaptic cleft and that the resulting adenosine was taken up again into nerve endings where it was re-phosphorylated to ATP. 2. Adenosine derivatives in the synaptic cleft inhibited the postsynaptic potentials in olfactory cortex slices in vitro, presumably by the inhibition of Ca2+ influx into nerve endings which resulted in the reduction of transmitter release. 3. The adenosine derivatives also increased the level of cyclic AMP in the slices under the same conditions as above. 4. Although the nature of the "adenosine receptors" for both functions was remarkably similar, the increase of cyclic AMP did not mediate the inhibitory action, but the presynaptic increase of cyclic AMP induced by adenosine derivatives might mediate the facilitation observed in the olfactory cortex. 5. Possible physiological roles of extracellular adenosine derivatives in mammalian brain were classified, at different sites of action around the synapses, with different time courses and modes of action, directly or via the increase of intracellular cyclic AMP.     

The ultrastructure of giant fibre and serial synapses in crayfish

Summary The ultrastructure of synapses between the cord giant fibres (lateral and medial) and the motor giant fibres in crayfish, Astacus pallipes, third abdominal ganglia have been examined. These electrotonic synapses are asymmetrical, they have synaptic vesicles only in the presynaptic fibre, and they have synaptic cleft widths normally of about 100 Å but narrowed to about 50 Å in restricted areas. Localized increases in density of the synaptic cleft and adjacent membranes also occur within a synapse, and synaptic vesicles are most tightly grouped at the membrane in such areas. Tight or gap junctions with 30 Å or narrower widths have not been found, but the junctions probably function in a similar way to gap junctions.Three small nerves are closely associated with the synapses between the giant fibres. One of these small nerves has round synaptic vesicles and is thought to be excitatory on morphological grounds; one has flattened vesicles and is thought to be inhibitory; and one is postsynaptic to the lateral giant and the two small presynaptic nerves. It is proposed that these small nerves modulate activity in the much larger giant fibre synapse.     

Comparison of the effect of gamma-aminobutyric acid and its derivative baclofen on the activity of Purkinje cells of frog cerebellum in vitro

The inhibitory effect of gamma-aminobutyric acid (GABA) and its synthetic derivative baclofen were compared in frog cerebellum in vitro. Baclofen inhibited synaptic transmission from parallel fibres to the Purkinje cells in EC50 concentrations approximately 200-fold lower than for GABA. In addition to its inhibitory effect, GABA induced temporary facilitation of responses in the dendrite zone by a mechanism dependent on the presence of a normal Cl- concentration; the inhibitory phase was only partly sensitive to reduction of the Cl- concentration in the medium and to the administration of picrotoxin. The action of baclofen, which was unaffected by these treatments, requires an intact catecholamine and serotonin pool, since it is ineffective in reserpine-treated animals. Both substances also influence the excitability of parallel fibres. In solutions with a high Mg2+ and a low Ca2+ concentrations GABA inhibits the spontaneous activity of Purkinje cells by acting on the postsynaptic membrane of the soma and the primary dendrites. The effect of baclofen is evidently the outcome of inhibition of transmitter release from presynaptic endings.     

Anatomical circuitry of lateral inhibition in the eye of the horseshoe crab, Limulus polyphemus

Lengthy uninterrupted series of sections of the neural plexus in the compound eye of the horseshoe crab, Limulus polyphemus, have been used to reconstruct all the arborizations and their synaptic interconnections in a neuropil knot. This one microglomerulus contains the axons of 19 retinular cells, which pass by without contacts; 13 efferent fibres with 44 synapses to and from eccentric cell collaterals; and arborizations from 54 eccentric cells with 577 synapses. Eccentric cell axons are devoid of synaptic input. Their collaterals ramify in synaptic knots and subserve both pre- and postsynaptic functions simultaneously. Arborizations near the axon of origin have a highly branched pattern (up to 20 bifurcations), a high synaptic input: output ratio (up to about 9:1), and high synaptic density (a maximum of 12 per micrometre of neurite length). The opposite extreme is represented by sparsely branched eccentric cell collaterals distant from their axons of origin with very little synaptic input and sparse output. Spatially graded lateral inhibition is the apparent outcome of a radially decreasing distribution of inhibitory synapses on the arborizations of eccentric cell collaterals combined with possible decremental signal transmission in the plexus. The synaptic analysis has a bearing on most physiological aspects of lateral inhibition that have been studied in the Limulus eye. Implied in the results is the suggestion that synapse formation is an intrinsic property of the presynaptic element, but that the connectivity is governed by the electrical activity of target neurons.     

High-density presynaptic transporters are required for glutamate removal from the first visual synapse

Reliable synaptic transmission depends not only on the release machinery and the postsynaptic response mechanism but also on removal or degradation of transmitter from the synaptic cleft. Accumulating evidence indicates that postsynaptic and glial excitatory amino acid transporters (EAATs) contribute to glutamate removal. However, the role of presynaptic EAATs is unclear. Here, we show in the mouse retina that glutamate is removed from the synaptic cleft at the rod to rod bipolar cell (RBC) synapse by presynaptic EAATs rather than by postsynaptic or glial EAATs. The RBC currents evoked by electrical stimulation of rods decayed slowly after pharmacological blockade of EAATs. Recordings of the evoked RBC currents from EAAT subtype-deficient mice and the EAAT-coupled anion current reveal that functional EAATs are localized to rod terminals. Model simulations suggest that rod EAATs are densely packed near the release site and that rods are equipped with an almost self-sufficient glutamate recollecting system.     

Riluzole Enhances Glutamate Uptake in Rat Astrocyte Cultures

1. Riluzole is used for the treatment of amyotrophic lateral sclerosis and reported to have neuroprotective effects in animal models of Parkinson's disease, Huntington's disease, and brain ischemia. The neuroprotective action of riluzole has been attributed to its ability to inhibit glutamate release (A. Doble, Neurology 47(4):233S-241S, 1996). 2. The effect of riluzole on L-[2,3-3H] glutamate uptake was investigated in rat cortical astrocyte cultures. 3. Riluzole showed a biphasic concentration-dependent effect on basal glutamate uptake. At low concentrations (1 and 10 microM) riluzole significantly increased glutamate uptake, whereas from 100 microM promoted a slight reduction. 4. Considering the large range of glutamate levels in the synaptic cleft, we studied the 1 microM riluzole effect on uptake of glutamate at different concentrations (1-1000 microM). Riluzole was more effective at low glutamate concentrations (10 microM), enhancing the basal glutamate uptake up to 42%. 5. The action of riluzole on astrocytic glutamate uptake could be an additional mechanism to its neuroprotective role, perhaps suggesting a modulatory action on glutamatergic system involving glutamate clearance from synaptic cleft.     

Synaptic interactions between nonspiking local interneurones in the terminal abdominal ganglion of the crayfish

Nonspiking local interneurones are the important premotor elements in arthropod motor control systems. We have analyzed the synaptic interactions between nonspiking interneurones in the crayfish terminal (6th) abdominal ganglion using simultaneous intracellular recordings. Only 15% of nonspiking interneurones formed bi-directional excitatory connections. In 77% of connections, however, the nonspiking interneurones showed a one-way inhibitory interaction. In these cases, the presynaptic nonspiking interneurones received excitatory synaptic inputs from the sensory afferents innervating hairs on the surface of the uropods and the postsynaptic nonspiking interneurones received inhibitory synaptic inputs that were partly mediated by the inputs to the presynaptic nonspiking interneurones. The membrane hyperpolarization of the postsynaptic nonspiking interneurones mediated by the presynaptic nonspiking interneurones was reduced in amplitude when the hyperpolarizing current was injected into the postsynaptic interneurones, or when the external bathing solution was replaced with one containing low calcium and high magnesium concentrations. The role of these interactions in the circuits controlling the movements of the terminal appendages is discussed.Abbreviations AL antero-lateral - epsp excitatory postsynaptic potential - ipsp inhibitory postsynaptic potential - PL postero-lateral     

A metabotropic glutamate receptor regulates transmitter release from cone presynaptic terminals in carp retinal slices.

The role of group III metabotropic glutamate receptors (mGluRs) in photoreceptor-H1 horizontal cell (HC) synaptic transmission was investigated by analyzing the rate of occurrence and amplitude of spontaneous excitatory postsynaptic currents (sEPSCs) in H1 HCs uncoupled by dopamine in carp retinal slices. Red light steps or the application of 100 microM cobalt reduced the sEPSC rate without affecting their peak amplitude, which is consistent with hyperpolarization or the suppression of Ca(2+) entry into cone synaptic terminals reducing vesicular transmitter release. Conversely, postsynaptic blockade of H1 HC AMPA receptors by 500 nM CNQX reduced the amplitude of sEPSCs without affecting their rate. This analysis of sEPSCs represents a novel methodology for distinguishing between presynaptic and postsynaptic sites of action. The selective agonist for group III mGluRs, l-2-amino-4-phosphonobutyrate (L-APB or L-AP4; 20 microM), reduced the sEPSC rate with a slight reduction in amplitude, which is consistent with a presynaptic action on cone synaptic terminals to reduce transmitter release. During L-APB application, recovery of sEPSC rate occurred with 500 microM (s)-2-methyl-2-amino-4-phosphonobutyrate (MAP4), a selective antagonist of group III mGluR, and with 200 microM 4-aminopyridine (4-AP), a blocker of voltage-dependent potassium channels. Whole-cell recordings from cones in the retinal slice showed no effect of L-APB on voltage-activated Ca(2+) conductance. These results suggest that the activation of group III mGluRs suppresses transmitter release from cone presynaptic terminals via a 4-AP-sensitive pathway. Negative feedback, operating via mGluR autoreceptors, may limit excessive glutamate release from cone synaptic terminals.     

Transmitter concentration profiles in the synaptic cleft: an analytical model of release and diffusion.

A three-dimensional model for release and diffusion of glutamate in the synaptic cleft was developed and solved analytically. The model consists of a source function describing transmitter release from the vesicle and a diffusion function describing the spread of transmitter in the cleft. Concentration profiles of transmitter at the postsynaptic side were calculated for different transmitter concentrations in a vesicle, release scenarios, and diffusion coefficients. From the concentration profiles the receptor occupancy could be determined using alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor kinetics. It turned out that saturation of receptors and sufficiently fast currents could only be obtained if the diffusion coefficient was one order of magnitude lower than generally assumed, and if the postsynaptic receptors formed clusters with a diameter of roughly 100 nm directly opposite the release sites. Under these circumstances the gradient of the transmitter concentration at the postsynaptic membrane outside the receptor clusters was steep, with minimal cross-talk among neighboring receptor clusters. These findings suggest that for each release site a corresponding receptor aggregate exists, subdividing an individual synapse into independent functional subunits without the need for specific lateral diffusion barriers.     

Kainic acid and synaptic transmission in the stellate ganglion of the squid.

Kainate, a conformational analogue of glutamate, blocks synaptic transmission across the giant synapse of the squid. In the presence of blocking doses of kainate, impulses continue to propagate into the nerve terminal, but action potentials are slightly reduced in size and the subsequent hyperpolarization is greatly diminished. Kainate depolarizes the postsynaptic axon. Since the depolarizing action of kainate is confined to the postsynaptic membrane, it appears that kainate can combine with the receptors which are normally activated by the transmitter. This results in a diminished effect of the transmitter released by a presynaptic nerve impulse.     

Correlation of Transmitter Release with Membrane Properties of the Presynaptic Fiber of the Squid Giant Synapse

Depolarization of the presynaptic terminal by current produced a postsynaptic potential (PSP) which increased with increasing presynaptic polarization and then reached a plateau. Iontophoretic injection of tetraethylammonium ions (TEA) into the presynaptic axon near the terminal produced a prolonged presynaptic spike. The resulting PSP is increased in size and its time course closely followed that of the presynaptic spike. The presynaptic fiber no longer exhibited rectification and strong depolarizations revealed that the PSP reached a maximum with about 110 mv depolarization. Further depolarization produced a decrease in PSP amplitude and finally transmission was blocked. However, a PSP then always appeared on withdrawal of the depolarizing current. Under the conditions of these experiments, the PSP could be considered a direct measure of transmitter release. Bathing the TEA-injected synapse with concentrations of tetrodotoxin (TTX) sufficient to block spike activity in both pre- and postsynaptic axons did not greatly modify postsynaptic electrogenesis. However, doubling TTX concentration reversibly blocked PSP. Thus the permeability changes to Na and K accompanying the spike do not appear necessary for transmitter release. Some other processes related to the level of presynaptic polarization must be involved to explain the data. The inhibition of transmitter release by strong depolarizations appears to be related to Ca action. A membrane Ca current may also be necessary for normal transmitter release.     

GABA and neuroligin signaling: linking synaptic activity and adhesion in inhibitory synapse development

GABA-mediated synaptic inhibition is crucial in neural circuit operations. In mammalian brains, the development of inhibitory synapses and innervation patterns is often a prolonged postnatal process, regulated by neural activity. Emerging evidence indicates that gamma-aminobutyric acid (GABA) acts beyond inhibitory transmission and regulates inhibitory synapse development. Indeed, GABA(A) receptors not only function as chloride channels that regulate membrane voltage and conductance but also play structural roles in synapse maturation and stabilization. The link from GABA(A) receptors to postsynaptic and presynaptic adhesion is probably mediated, partly by neuroligin-reurexin interactions, which are potent in promoting GABAergic synapse formation. Therefore, similar to glutamate signaling at excitatory synapse, GABA signaling may coordinate maturation of presynaptic and postsynaptic sites at inhibitory synapses. Defining the many steps from GABA signaling to receptor trafficking/stability and neuroligin function will provide further mechanistic insights into activity-dependent development and possibly plasticity of inhibitory synapses.     

Synapse proteomics: current status and quantitative applications

Chemical synapses are key organelles for neurotransmission. The coordinated actions of protein networks in diverse synaptic subdomains drive the sequential molecular events of transmitter release from the presynaptic bouton, activation of transmitter receptors located in the postsynaptic density and the changes of postsynaptic potential. Plastic change of synaptic efficacy is thought to be caused by the alteration of protein constituents and their interaction in the synapse. As a first step toward the understanding of the organization of synapse, several proteomics studies have been carried out to profile the protein constituents and the post-translational modifications in various rodent excitatory chemical synaptic subdomains, including postsynaptic density, synaptic vesicle and the synaptic phosphoproteome. Quantitative proteomics have been applied to examine the changes of synaptic proteins during brain development, in knockout mice model developed for studies of synapse physiology and in rodent models of brain disorders. These analyses generate testable hypotheses of synapse function and regulation both in health and disease.     

Synapse proteomics: current status and quantitative applications

Chemical synapses are key organelles for neurotransmission. The coordinated actions of protein networks in diverse synaptic subdomains drive the sequential molecular events of transmitter release from the presynaptic bouton, activation of transmitter receptors located in the postsynaptic density and the changes of postsynaptic potential. Plastic change of synaptic efficacy is thought to be caused by the alteration of protein constituents and their interaction in the synapse. As a first step toward the understanding of the organization of synapse, several proteomics studies have been carried out to profile the protein constituents and the post-translational modifications in various rodent excitatory chemical synaptic subdomains, including postsynaptic density, synaptic vesicle and the synaptic phosphoproteome. Quantitative proteomics have been applied to examine the changes of synaptic proteins during brain development, in knockout mice model developed for studies of synapse physiology and in rodent models of brain disorders. These analyses generate testable hypotheses of synapse function and regulation both in health and disease.