Reshaping neuron-glial communication at hippocampal synapses

Neuron-glial interactions in the nervous system are of fundamental importance to many processes including neural migration,axon guidance, myelination and synaptic transmission. At synapses in the CNS, the physiological and structural relationship between neurons and astrocytes is particularly complex. The juxtaposition of astrocytic membranes with presynaptic and postsynaptic elements is important for regulating synaptic transmission and plasticity. Recent investigations demonstrate that the morphology of both neuronal and glial components show rapid, continuous structural remodeling in the hippocampus.These physical modifications are likely to have a significant functional impact upon neurotransmission and indicate that there modeling of astrocytic morphology might be crucial for the dynamic regulation of the synapse and its microenvironment. In this review, we focus on the structural complexities of astrocyte-synapse interactions in the hippocampus and their implications for understanding synaptic physiology, behavior and disease.     

Mobile NMDA receptors at hippocampal synapses

Glutamate receptors are concentrated in the postsynaptic complex of central synapses. This implies a highly organized and stable postsynaptic membrane with tightly anchored receptors. Recent reports of rapid AMPA receptor insertion and removal at synapses have challenged this view. We examined the stability of synaptic NMDA receptors on cultured hippocampal neurons using the open-channel blockers (+)-MK-801 and ketamine to tag synaptic NMDA receptors. NMDA receptor-mediated EPSCs showed an anomalous recovery following "irreversible" MK-801 block. The recovery could not be attributed to MK-801 unbinding or insertion of new receptors, suggesting that membrane receptors had moved laterally into the synapse. At least 65% of synaptic NMDA receptors were mobile. Our results indicate that NMDA receptors can move laterally between synaptic and extrasynaptic pools, providing evidence for a dynamic organization of synaptic NMDA receptors in the postsynaptic complex.     

Neurotransmitter release at rapid synapses

The classical concept of the vesicular hypothesis for acetylcholine (ACh) release, one quantum resulting from exocytosis of one vesicle, is becoming more complicated than initially thought. 1) synaptic vesicles do contain ACh, but the cytoplasmic pool of ACh is the first to be used and renewed on stimulation. 2) The vesicles store not only ACh, but also ATP and Ca(2+) and they are critically involved in determining the local Ca(2+) microdomains which trigger and control release. 3) The number of exocytosis pits does increase in the membrane upon nerve stimulation, but in most cases exocytosis happens after the precise time of release, while it is a change affecting intramembrane particles which reflects more faithfully the release kinetics. 4) The SNARE proteins, which dock vesicles close to Ca(2+) channels, are essential for the excitation-release coupling, but quantal release persists when the SNAREs are inactivated or absent. 5) The quantum size is identical at the neuromuscular and nerve-electroplaque junctions, but the volume of a synaptic vesicle is eight times larger in electric organ; at this synapse there is enough ACh in a single vesicle to generate 15-25 large quanta, or 150-200 subquanta. These contradictions may be only apparent and can be resolved if one takes into account that an integral plasmalemmal protein can support the formation of ACh quanta. Such a protein has been isolated, characterised and called mediatophore. Mediatophore has been localised at the active zones of presynaptic nerve terminals. It is able to release ACh with the expected Ca(2+)-dependency and quantal character, as demonstrated using mediatophore-transfected cells and other reconstituted systems. Mediatophore is believed to work like a pore protein, the regulation of which is in turn likely to depend on the SNARE-vesicle docking apparatus.     

Neurotransmitter release at central synapses

Our understanding of synaptic transmission has grown dramatically during the 15 years since the first issue of Neuron was published, a growth rate expected from the rapid progress in modern biology. As in all of biology, new techniques have led to major advances in the cell and molecular biology of synapses, and the subject has evolved in ways (like the production of genetically engineered mice) that could not even be imagined 15 years ago. My plan for this review is to summarize what we knew about neurotransmitter release when Neuron first appeared and what we recognized we did not know, and then to describe how our views have changed in the intervening decade and a half. Some things we knew about synapses--"knew" in the sense that the field had reached a consensus--are no longer accepted, but for the most part, impressive advances have led to a new consensus on many issues. What I find fascinating is that in certain ways nothing has changed--many of the old arguments persist or recur in a different guise--but in other ways the field would be unrecognizable to a neurobiologist time-transported from 1988 to 2003.     

Neurotransmitter release at fast synapses

We wish to thank Raya Khanin for her assistance in gathering the papers cited here and Chaim Mayerson for reading and editing this review.     

Different priming states of synaptic vesicles underlie distinct release probabilities at hippocampal excitatory synapses

1. Download : Download high-res image (181KB)
2. Download : Download full-size image

     

Synaptic plasticity at hippocampal mossy fibre synapses

The dentate gyrus provides the main input to the hippocampus. Information reaches the CA3 region through mossy fibre synapses made by dentate granule cell axons. Synaptic plasticity at the mossy fibre-pyramidal cell synapse is unusual for several reasons, including low basal release probability, pronounced frequency facilitation and a lack of N-methyl-D-aspartate receptor involvement in long-term potentiation. In the past few years, some of the mechanisms underlying the peculiar features of mossy fibre synapses have been elucidated. Here we describe recent work from several laboratories on the various forms of synaptic plasticity at hippocampal mossy fibre synapses. We conclude that these contacts have just begun to reveal their many secrets.     

Properties of fast endocytosis at hippocampal synapses

Regulation of synaptic transmission is a widespread means for dynamic alterations in nervous system function. In several cases, this regulation targets vesicular recycling in presynaptic terminals and may result in substantial changes in efficiency of synaptic transmission. Traditionally, experimental accessibility of the synaptic vesicle cycle in central neuronal synapses has been largely limited to the exocytotic side, which can be monitored with electrophysiological responses to neurotransmitter release. Recently, physiological measurements on the endocytotic portion of the cycle have been made possible by the introduction of styryl dyes such as FM1-43 as fluorescent markers for recycling synaptic vesicles. Here we demonstrate the existence of fast endocytosis in hippocampal nerve terminals and derive its kinetics from fluorescence measurements using dyes with varying rates of membrane departitioning. The rapid mode of vesicular retrieval was greatly speeded by exposure to staurosporine or elevated extracellular calcium. The effective time-constant for retrieval can be < 2 seconds under appropriate conditions. Thus, hippocampal synapses capitalize on efficient mechanisms for endocytosis and their vesicular retrieval is subject to modulatory control.     

Shrinkage of dendritic spines associated with long-term depression of hippocampal synapses

Activity-induced modification of neuronal connections is essential for the development of the nervous system and may also underlie learning and memory functions of mature brain. Previous studies have shown an increase in dendritic spine density and/or enlargement of spines after the induction of long-term potentiation (LTP). Using two-photon time-lapse imaging of dendritic spines in acute hippocampal slices from neonatal rats, we found that the induction of long-term depression (LTD) by low-frequency stimulation is accompanied by a marked shrinkage of spines, which can be reversed by subsequent high-frequency stimulation that induces LTP. The spine shrinkage requires activation of NMDA receptors and calcineurin, similar to that for LTD. However, spine shrinkage is mediated by cofilin, but not by protein phosphatase 1 (PP1), which is essential for LTD, suggesting that different downstream pathways are involved in spine shrinkage and LTD. This activity-induced spine shrinkage may contribute to activity-dependent elimination of synaptic connections.     

Integrins modulate fast excitatory transmission at hippocampal synapses

The present study provides the first evidence that adhesion receptors belonging to the integrin family modulate excitatory transmission in the adult rat brain. Infusion of an integrin ligand (the peptide GRGDSP) into rat hippocampal slices reversibly increased the slope and amplitude of excitatory postsynaptic potentials. This effect was not accompanied by changes in paired pulse facilitation, a test for perturbations to transmitter release, or affected by suppression of inhibitory responses, suggesting by exclusion that alterations to alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA)-type glutamate receptors cause the enhanced responses. A mixture of function-blocking antibodies to integrin subunits alpha(3), alpha(5), and alpha(v) blocked ligand effects on synaptic responses. The ligand-induced increases were (i) blocked by inhibitors of Src tyrosine kinase, antagonists of N-methyl-d-aspartate receptors, and inhibitors of calcium calmodulin-dependent protein kinase II and (ii) accompanied by phosphorylation of both the Thr(286) site on calmodulin-dependent protein kinase II and the Ser(831) site on the GluR1 subunit of the AMPA receptor. N-Methyl-d-aspartate receptor antagonists blocked the latter two phosphorylation events, but Src kinase inhibitors did not. These results point to the conclusion that synaptic integrins regulate glutamatergic transmission and suggest that they do this by activating two signaling pathways directed at AMPA receptors.     

Long-term potentiation at hippocampal perforant path-dentate astrocyte synapses

Accumulated evidence indicates that astroglial cells actively participate in neuronal synaptic transmission and plasticity. However, it is still not clear whether astrocytes are able to undergo plasticity in response to synaptic inputs. Here we demonstrate that a long-term potentiation (LTP)-like response could be detected at perforant path-dentate astrocyte synapses following high-frequency stimulation (HFS) in hippocampal slices of GFAP-GFP transgenic mice. The potentiation was not dependent on the glutamate transporters nor the group I metabotropic glutamate receptors. However, the induction of LTP requires activation of the NMDA receptor (NMDAR). The presence of functional NMDAR was supported by isolating the NMDAR-gated current and by identifying mRNAs of NMDAR subunits in astrocytes. Our results suggest that astrocytes in the hippocampal dentate gyrus are able to undergo plasticity in response to presynaptic inputs.     

Facilitation of transmitter release at squid synapses

Facilitation is shown to decay as a compound exponential with two time constants (T1, T2) at both giant and non-giant synapses in squid stellate ganglia bathed in solutions having low extracellular calcium concentrations ([Ca++]o). Maximum values of facilitation (F1) were significantly larger, and T1 was significantly smaller in giant than non-giant synapses. Decreases in [Ca++]o or increases in [Mn++]o had variable effects on T1 and F1, whereas decreases in temperature increased T1 but had insignificant effects on F1. The growth of facilitation during short trains of equal interval stimuli was adequately predicted by the linear summation model developed by Mallart and Martin (1967. J. Physiol. (Lond.). 193:676--694) for frog neuromuscular junctions. This result suggests that the underlying mechanisms of facilitation are similar in squid and other synapses which release many transmitter quanta.     

Leptin reverses long-term potentiation at hippocampal CA1 synapses

The hormone leptin crosses the blood brain barrier and regulates numerous neuronal functions, including hippocampal synaptic plasticity. Here we show that application of leptin resulted in the reversal of long-term potentiation (LTP) at hippocampal CA1 synapses. The ability of leptin to depotentiate CA1 synapses was concentration-dependent and it displayed a distinct temporal profile. Leptin-induced depotentiation was not associated with any change in the paired pulse facilitation ratio or the coefficient of variance, indicating a post-synaptic locus of expression. Moreover, the synaptic activation of NMDA receptors was required for leptin-induced depotentiation as the effects of leptin were blocked by the competitive NMDA receptor antagonist, D-aminophosphovaleric acid (D-AP5). The signaling mechanisms underlying leptin-induced depotentiation involved activation of the calcium/calmodulin-dependent protein phosphatase, calcineurin, but were independent of c- jun NH₂ terminal kinase. Furthermore, leptin-induced depotentiation was accompanied by a reduction in α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptor rectification indicating that loss of glutamate receptor 2 (GluR2)-lacking AMPA receptors underlies this process. These data indicate that leptin reverses hippocampal LTP via a process involving calcineurin-dependent internalization of GluR2-lacking AMPA receptors which further highlights the key role for this hormone in regulating hippocampal synaptic plasticity and neuronal development.     

Structural changes at dendritic spine synapses during long-term potentiation

Two key hypotheses about the structural basis of long-term potentiation (LTP) are evaluated in light of new findings from immature rat hippocampal slices. First, it is shown why dendritic spines do not split during LTP. Instead a small number of spine-like dendritic protrusions may emerge to enhance connectivity with potentiated axons. These 'same dendrite multiple synapse boutons' provide less than a 3% increase in connectivity and do not account for all of LTP or memory, as they do not accumulate during maturation. Second, polyribosomes in dendritic spines served to identify which of the existing synapses enlarged to sustain more than a 30% increase in synaptic strength. Thus, both enhanced connectivity and enlarged synapses result during LTP, with synapse enlargement being the greater effect.     

Neurotransmitter release at ribbon synapses in the retina

The synapses of photoreceptors and bipolar cells in the retina are easily identified ultrastructurally by the presence of synaptic ribbons, electron-dense bars perpendicular to the plasma membrane at the active zones, extending about 0.5 microm into the cytoplasm. The neurotransmitter, glutamate, is released continuously (tonically) from these 'ribbon synapses' and the rate of release is modulated in response to graded changes in the membrane potential. This contrasts with action potential-driven bursts of release at conventional synapses. Similar to other synapses, neurotransmitter is released at ribbon synapses by the calcium-dependent exocytosis of synaptic vesicles. Most components of the molecular machinery governing transmitter release are conserved between ribbon and conventional synapses, but a few differences have been identified that may be important determinants of tonic transmitter release. For example, the presynaptic calcium channels of bipolar cells and photoreceptors are different from those elsewhere in the brain. Differences have also been found in the proteins involved in synaptic vesicle recruitment to the active zone and in synaptic vesicle fusion. These differences and others are discussed in terms of their implications for neurotransmitter release from photoreceptors and bipolar cells in the retina.     

Multivesicular release at climbing fiber-Purkinje cell synapses.

Synapses driven by action potentials are thought to release transmitter in an all-or-none fashion; either one synaptic vesicle undergoes exocytosis, or there is no release. We have estimated the glutamate concentration transient at climbing fiber synapses on Purkinje cells by measuring the inhibition of excitatory postsynaptic currents (EPSCs) produced by a low-affinity competitive antagonist of AMPA receptors, gamma-DGG. The results, together with simulations using a kinetic model of the AMPA receptor, suggest that the peak glutamate concentration at this synapse is dependent on release probability but is not affected by pooling of transmitter released from neighboring synapses. We propose that the mechanism responsible for the elevated glutamate concentration at this synapse is the simultaneous release of multiple vesicles per site.     

Actin-dependent regulation of neurotransmitter release at central synapses

Depolymerization of actin by latrunculin A transiently promotes neurotransmitter release. The mean rate of mEPSCs increases by a Ca2+-independent process, without a concomitant change in the mean amplitude. The readily releasable vesicle pool size and the rate of refilling of the readily releasable pool remain unaltered by latrunculin treatment. Evoked neurotransmitter release also increases in a manner consistent with an increase in vesicle release probability. The observed enhancement of neurotransmitter release is specific to actin depolymerization mediated by latrunculin A and is not caused by cytochalasin D. Our findings indicate that actin participates in a regulatory mechanism that restrains fusion of synaptic vesicles at the active zone.     

Visualizing postendocytic traffic of synaptic vesicles at hippocampal synapses.

We have investigated mechanisms in postendocytic processing of synaptic vesicles at hippocampal synapses, using synaptobrevin/vesicle-associated membrane protein (VAMP) tagged with variants of the green fluorescent protein. Following exocytosis, VAMP is retrieved at synaptic and adjoining axonal regions. Retrieved VAMP-containing vesicles return to synaptic vesicle clusters at a rate slower than endocytosis. Vesicles containing a different protein, synaptophysin, recluster at a similar rate, suggesting common vesicular intermediates for the two proteins. Activity prolongs the time taken by endocytosed vesicles to return to synapses. Exogenous calcium buffers slow endocytosis but have no additional effect on the time course of reclustering. In contrast, the protein kinase inhibitor staurosporine does not affect endocytosis but slows reclustering. Finally, since VAMP can move freely on surface membranes, sustained synaptic activity leads to mixing of this vesicular component between adjacent synapses.     

Rapid reuse of readily releasable pool vesicles at hippocampal synapses

Functional presynaptic vesicles have been subdivided into readily releasable (RRP) and reserve (RP) pools. We studied recycling properties of RRP vesicles through differential retention of FM1-43 and FM2-10 and by varying the time window for FM dye uptake. Both approaches indicated that vesicles residing in the RRP underwent rapid endocytosis (tau approximately 1s), whereas newly recruited RP vesicles were recycled slowly (tau approximately 30 s). With repeated challenges (hypertonic or electrical stimuli), the ability to release neurotransmitter recovered 10-fold more rapidly than restoration of FM2-10 destaining. Finding neurotransmission in the absence of destaining implied that rapidly endocytosed RRP vesicles were capable of reuse, a process distinct from repopulation from the RP. Reuse would greatly expand the functional capabilities of a limited number of vesicles in CNS terminals, particularly during intermittent bursts of activity.